Comparison of infectious complications in patients with relapsed/refractory large B-cell lymphoma receiving CAR T cells vs bispecific antibodies

Abstract

Abstract INTRODUCTION: T-cell redirecting strategies (TRS), including chimeric antigen receptor (CAR) T cells and bispecific antibodies (BsAbs), have significantly improved the prognosis of patients with relapsed/refractory (R/R) large B-cell lymphoma (LBCL). However, they are associated with a potential risk of severe adverse events (AEs), being infections the leading cause of non-relapse mortality. In this study, we aimed to compare the incidence of infections in patients receiving CAR T cells versus (vs) BsAbs for R/R LBCL. METHODS: This retrospective, multicenter study included adult patients with R/R LBCL treated with CD19-targeted CAR T cells or single-agent BsAbs from November 2017 until July 2025. The primary endpoint was cumulative incidence of grade ≥ 3 infections according to CTCAE (v6) criteria. Secondary endpoints included cumulative incidence of any grade infections, progression-free survival (PFS) and overall survival (OS). The same endpoints were also determined excluding SARS-CoV-2 infections (COVID19-free cohort). Events were classified as microbiologically (MDI) or clinically documented infections (CDI). Infections were recorded from time of CAR-T infusion or BsAbs initiation until progressive disease (PD), last follow-up or death. RESULTS: The study included 228 patients (141 CAR-T, 87 BsAbs) with a median follow-up of 37 months (IQR 32-45). Overall, 145 (64%) were male and median age was 63 years (IQR 52-71). Compared to CAR-T, BsAb recipients were more heavily pretreated (2 vs. 3 previous lines, p<0.001) and had an increased rate of patients with an IPI score of ≥ 3 (45% vs 68%, p=0.002). In the CAR-T cohort, 29 (21%) patients had been previously exposed to BsAbs, with a median of 95 days (IQR 77-134) from last BsAb dose. In the BsAb group, 23 (26%) patients had received prior CAR-T cell therapy, with a median of 145 days (IQR 112-286) from infusion. Concerning efficacy in the CAR-T vs BsAb cohorts, the complete response rate was 51% vs 29%. The 36-month PFS and OS were 34% vs 20% (p<0.001), and 47% vs 32% (p=0.004), respectively. Focusing on toxicity, cytokine release syndrome and neurotoxicity occurred in 101 (72%) vs 45 (52%) patients and 42 (30%) vs 5 (6%) patients, respectively. For AE management, corticosteroids were more frequently used in the CAR-T cohort (33% vs 8.1%, p< 0.001). In terms of infection, 136 (60%) patients had at least one event (62% [88/141] CAR-T and 55% [48/87] BsAbs), with a similar mean number of infections between groups (1.9 [SD 2.6] CAR-T vs. 1.4 [SD 2.3]; p=0.2). Time to the first infection was 1.97 [IQR: 0.85 - 4.57] and 1.45 [IQR: 0.66 - 4.01] months in the CAR-T and BsAb cohorts, respectively. Regarding MDI, most were bacterial (35% CART vs. 31% BsAbs) or viral (23% CART vs. 24% BsAbs), with similar rates of SARS-CoV-2 infection (22% vs 16%); fungal infections were uncommon (3% CART vs 2% BsAbs). Approximately one third of patients in each group had a CDI (30% CART vs 25% ). The cumulative incidence of any-grade infections for CAR-T vs. BsAbs was 26% vs. 25% at 1 month, 47% vs. 35% at 3 months and 59% vs. 49% at 12 months, respectively.In terms of grade ≥3 infections, there were no significant differences in the cumulative incidence at 1, 3 and 12 months (13% vs. 11%, 25% vs. 16% and 37% vs. 26% [HR 0.69, 95%CI 0.44-1.10,p=0.12]). Similar cumulative incidence of any-grade infection and grade ≥3 infections were observed in the COVID19-free cohort. No differences were reported in terms of infection-related mortality rates between CAR-T and BsAbs (12% vs 10%, p=0.07). Among CAR T-cell recipients, a previous exposure to BsAbs was associated with a numerically increased cumulative incidence of any-grade infection (HR 1.34 [95%CI 0.85-2.10]; p=0.2) and a higher number of infectious events (2.8 [SD 3.2] vs. 1.7 [SD 2.4]; p=0.038); however, prior CAR-T therapy did not lead to an increased number of infections in the BsAb cohort (1.5 [2.7] vs. 1.4 [2.1]; p=0.6). In the multivariate analysis, absence of corticosteroid use for AE management was associated with a reduced risk of any-grade infection in the overall cohort (HR 0.50 [95%CI 0.28-0.90]; p=0.021). CONCLUSIONS: There is a high and similar incidence of infections with both CAR T cells and BsAbs. Prior BsAb exposure led to a higher number of infections in CAR-T recipients, as did corticosteroid treatment for AE management in the overall cohort, highlighting the need for intensified prevention strategies in these patients subgroups.