Predictors of Early Safety Outcomes with Axicabtagene Ciloleucel (axi-cel) in Patients with Relapsed or Refractory (R/R) Large B-Cell Lymphoma (LBCL)

Background: Cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) are known safety concerns following chimeric antigen receptor (CAR) T-cell therapy. Axi-cel, a CAR T-cell therapy approved in the US for adults with R/R LBCL after ≥ 1 line of prior therapy (2L+), has been administered in the outpatient setting, where a common cause for post-infusion hospital admissions were CAR T-cell-related toxicities (Furqan et al., 2024). Due to reimbursement policies in the US, patients at higher risk for safety events requiring inpatient care within 3 days of infusion may not be ideal candidates for outpatient care. We sought to identify what pre-infusion characteristics may put patients more at risk for developing early CRS and ICANS after axi-cel.

Methods: Patients receiving axi-cel for R/R LBCL 2L+ between 01/2021−10/2023 registered in the Center for International Blood and Marrow Transplant Research (CIBMTR) observational database were potentially eligible for analysis. Patients with prior non-transplant cellular therapy or missing data for analysis were excluded. Primary endpoints were onset of early CRS and early ICANS, defined as an any grade event within 3 calendar days of axi-cel infusion (date of infusion is defined as Day 0 and onset of CRS/ICANS before the end of Day 3 is an early event). Early Grade ≥ 2 and early Grade ≥ 3 CRS and ICANS (onset of CRS/ICANS before the end of Day 3 and maximum Grade ≥ 2 and ≥ 3 per ASTCT Criteria) were also evaluated. Patient demographics, ECOG PS, comorbidities, disease characteristics at diagnosis and prior to infusion, line of therapy (LoT), lab assessments prior to infusion, and infusion-related attributes were evaluated by univariate screening and multivariable regression for development of early CRS or early ICANS.

Results: A total of 888 patients from 105 centers were included in the analysis. Median age was 63 years (45% ≥ 65 years) and 66% were male. At LBCL diagnosis, 22% of patients had a histological transformation. Prior to infusion, 5% of patients had ECOG PS ≥ 2 and 70% had at least 1 clinically significant comorbidity. Half (50%) of patients received axi-cel with 1 prior LoT (2L) and 10% of patients underwent a prior autologous hematopoietic cell transplant. Median time from leukapheresis to infusion was 30 days (interquartile range [IQR] 27−34). Bridging therapy of any type was administered in 59%, and systemic bridging therapy in 48%, of patients and 14% of patients were intended to be treated in the outpatient setting. Thirty percent of patients had chemo-sensitive disease prior to infusion. Median follow-up post-infusion was 12.3 months (range 3−38).

Median time to onset of CRS was 4 days (IQR 2−6) and 84% of patients experienced CRS at any time. Early any grade (Grade ≥ 3) CRS was observed in 47% (3%) of patients. In the multivariate analysis of early any grade CRS, only prior LoT (≥ 2 vs 1) was associated with lower risk (OR 0.71; 95% CI 0.55−0.94), but this association was not seen in the analyses of early Grade ≥ 2 nor early Grade ≥ 3 CRS. Notably, 47% and 52% of patients with ≥ 2 vs 1 prior LoT, respectively, only experienced early Grade 1 CRS (defined as fever only). In the analysis of early Grade ≥ 3 CRS, only ECOG PS 0−1 was associated with lower risk vs patients with ECOG PS ≥ 2 (OR 0.17; 95% CI 0.06−0.48).

Median time to onset of ICANS was 7 days (IQR 5−9) and 47% of patients experienced ICANS at any time. Early any grade (Grade ≥ 3) ICANS occurred in 9% (5%) of patients. After multivariate adjustment, age < 65 vs ≥ 65 (OR 0.38 [95% CI 0.23−0.62]) and 1 vs ≥ 2 prior LoT (OR 0.58 [95% CI 0.36−0.95]) were associated with lower risk of early any grade ICANS. Age < 65 vs ≥ 65 was also associated with lower risk of early Grade ≥ 3 ICANS (OR 0.31 [95% CI 0.15−0.61]), as was ECOG PS 0−1 vs ECOG PS ≥ 2 (OR 0.30 [95% CI 0.11−0.83]). Comorbidities were evaluated and no significant associations with early CRS nor early ICANS were found.

Conclusions: This analysis of readily available clinical variables in a real-world contemporary cohort of patients treated with axi-cel demonstrates that younger patients and patients with better ECOG PS have lower risk of early CRS and early ICANS. Axi-cel in 2L setting, vs 3L+, is associated with lower risk of early any grade ICANS and comparable early Grade ≥ 2 CRS, but a higher risk of early fever. These results further enable a more informed decision of patient selection for physicians, which is critical for the success of outpatient administration.