Introduction: Relapsed or refractory diffuse large b-cell lymphoma (R/R DLBCL) is associated with poor prognosis. Novel treatment modalities such as chimeric antigen receptor (CAR) T-cell therapies and bispecific antibodies (BsAbs) have shown efficacy in R/R DLBCL. Despite their efficacy, these agents have distinct toxicity profiles, requiring close monitoring and prophylactic strategies.

Aims: We aimed to compare the safety profiles of CAR-T cell therapies and BsAbs, and to identify the mortality-associated adverse events (AEs) in patients with DLBCL, using real-world data from FDA Adverse Event Reporting System (FAERS).

Methods: FAERS data from 2018 to 2024 were used to assess the safety profiles of axicabtagene ciloleucel (axi-cel), tisagenlecleucel (tisa-cel), glofitamab, and epcoritamab in patients with DLBCL. Selected AEs were grouped into broader clinical categories such as infection and neurotoxicity based on signal organ class level of MedDRA version 27.1. Disproportionality was evaluated using reporting odds ratios (ROR), and multiple logistic regression was performed to identify mortality-associated AEs. Cumulative incidence of death (CID) was estimated, and inter-agent differences were assessed via Gray's test and Fine–Gray models. All statistical analyses were performed in R version 4.5.1.

Results: A total of 1,022 cases were analyzed, with 3,328 AEs reported. The most frequently reported agent was axi-cel (n = 494, AEs = 1711), followed by glofitamab (n = 246, AEs = 647), epcoritamab (n=185, AEs = 521) and tisa-cel (n=97, AEs = 449). There were 143 (19%) deaths in CAR-T cell therapy and 140 (32%) deaths in BsAbs. Infection rates were higher in patients receiving BsAbs—34.1% with glofitamab and 36.2% with epcoritamab—compared to 22.5% with axi-cel and 17.5% with tisa-cel, resulting in ROR of 1.96 (95% CI: 1.48 – 2.58) for BsAbs. Among reported infections, pneumonia was the most common across all agents, followed by cytomegalovirus and herpesvirus infections. Cytokine release syndrome (CRS) was more frequently reported with CAR-T cell therapies—axi-cel (43.1%) and tisa-cel (58.8%)—compared to BsAbs, with ROR of 2.48 (95% CI: 1.90–3.26). Axi-cel was also associated with higher rates of immune effector cell-associated neurotoxicity syndrome (ICANS) and non-ICANS neurotoxicity, reported in 24.3% and 26.7% of the cases, respectively, with significant disproportionality signals (ROR=4.66 for ICANS and 2.90 for non-ICANS neurotoxicity). Neutropenia and thrombocytopenia were commonly recorded with glofitamab, at rates of 12.2% and 9.3%, respectively, with ROR of 2.12, (95% CI: 1.23–3.65) for thrombocytopenia. At day 30, CID was 28.2% (95%, CI: 23 - 34) for axi-cel and 41.4% (95%, CI: 23 - 59) for tisa-cel, which were higher than the rates observed with glofitamab and epcoritamab: 16.2% (95%, CI: 11-22) and 15.7% (95%, CI: 10-21), respectively (p<0.001). Compared to axi-cel, the hazard ratio (HR) for death was significantly higher with tisa-cel (HR: 1.96; 95% CI, 1.23–3.24; p = 0.008), and significantly lower with glofitamab (HR: 0.60; 95% CI, 0.44–0.81; p < 0.001). In fatal cases most common reported AEs among the agents were infection, disease progression, CRS, non-ICANS neurotoxicity and ICANS. Regression analysis demonstrated that progression significantly increased the odds of death with axi-cel (OR = 3.10, 95% CI: 1.85–5.22, p < 0.001), tisa-cel (OR = 13.3, 95% CI: 4.15–49.6, p < 0.001), and epcoritamab (OR = 2.56, 95% CI: 1.18–5.71, p = 0.019), but not with glofitamab (p = 0.536). Infection appeared to increase the risk of death with axi-cel, but this finding did not reach statistical significance (p = 0.076). Other adverse events were not found to be significantly associated with mortality.

Conclusion: Our real-world data analysis from FAERS revealed distinct toxicity profile of CAR-T cell therapies and BsAbs. CRS, ICAN and non-ICAN neurotoxicity were predominantly associated with CAR-T cell therapies, whereas infection and cytopenias were more frequently reported in the cases of BsAbs. Disease progression emerged as the strongest predictor of death among the other AEs. A comprehensive understanding of these toxicity patterns is crucial for optimizing patient selection, preventing complications, and improving overall management strategies in clinical practice.

This content is only available as a PDF.
2025
Sign in via your Institution