Postmarketing safety profile of chimeric antigen receptor (CAR) T cell therapies in diffuse large B-cell lymphoma (DLBCL): Analysis of real-world (RW) AE reporting from the FDA Adverse Event Reporting System (FAERS).

7028 Background: CAR T cell therapies have emerged as effective treatment options with deep and durable responses in patients (pt) with DLBCL. Although efficacy and safety data from clinical trials are usually used for drug approval, potentially relevant AEs may not be captured due to limited study follow-up and population. Surveillance databases like FAERS can further characterize safety of therapeutic biologics by capturing RW AEs. We aimed to characterize the safety profile of CAR T cell therapies in the DLBCL population using FAERS. Methods: FAERS was used to identify AEs in pt with DLBCL treated with 2 commercially available CAR T cell therapies, lisocabtagene maraleucel (liso-cel) or axicabtagene ciloleucel (axi-cel). AEs of interest were cytokine release syndrome (CRS), neurological events (NE), hemophagocytic lymphohistiocytosis (HLH), cytopenia, and infections. The primary analysis examined all case reports from Q4 2017 to Q3 2024, the latest available quarterly release. Two sensitivity analyses adjusting for differences in follow-up after FDA approvals were performed: (1) AEs reported any time after liso-cel FDA approval (02/05/2021), which is later, and (2) AEs reported within 2 years of FDA approval for each CAR T cell therapy. Disproportionality analysis compared relative frequency of AEs. Reporting odds ratios (ROR) and 95% CIs were used to identify significant differences in AEs between treatments (ie, 95% CI did not cross 1). An ROR > 1 indicated higher event frequency for axi-cel vs liso-cel. Results: From Q4 2017 to Q3 2024, 3251 AE reports in pt with DLBCL were associated with liso-cel (n = 232) or axi-cel (n = 3019). In disproportionality analysis, axi-cel had significantly higher ROR for CRS (1.48; 95% CI, 1.13–1.93), NE (1.61; 1.23–2.11), and cytopenia (2.45; 1.41–4.24) than liso-cel. Considering limitations of underreporting and incomplete information inherent in FAERS, no statistically significant difference can be inferred for infections (1.36; 95% CI, 0.75–2.48), seizures (1.36; 0.42–4.40), and HLH (1.18; 0.36–3.83). Observed trends were consistent in both sensitivity analyses, where reporting frequencies remained significantly higher with axi-cel vs liso-cel for CRS (1.60; 95% CI, 1.17–2.17 and 2.02; 1.37–2.98), NE (1.59; 1.18–2.15 and 2.04; 1.39–3.00), and cytopenia (2.00; 1.11–3.60 and 2.60; 1.24–5.46), after adjusting for differences in follow-up durations. Conclusions: This retrospective analysis of FAERS, using spontaneous safety reporting data after approval and broader population beyond clinical trials, demonstrated a favorable RW safety profile for liso-cel vs axi-cel for CRS, NE, and cytopenia. These findings provide valuable insights into the safety profile of CAR T cell therapies in DLBCL to inform clinical decision-making and pt management.