Early intervention for cytokine release syndrome after chimeric antigen receptor T-cell therapy is associated with an increased risk of cytopenias and infections in patients with large B-cell lymphoma

Abstract Introduction: Cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) are common adverse events (AEs) after chimeric antigen receptor (CAR) T-cell therapy in patients with relapsed/refractory (R/R) large B-cell lymphoma (LBCL). Following single-arm studies exploring an earlier AE intervention (Topp MS, BJH 2021 & Oluwole OO, BJH 2021), this approach has been widely adopted in clinical practice. However, large real-world cohorts with granular data focused on the broader effect of this strategy on CRS, ICANS, cytopenias and infections are lacking. Here, we carried out a comprehensive evaluation of the impact that an early CRS mitigation policy has on the full spectrum of CAR T-cell outcomes. Methods: We conducted a multicenter, international study including patients with R/R LBCL treated at 8 centers with commercially-available CAR T-cell products until December 2024. Patients were categorized in 2 groups according to CRS intervention, including early (grade 1 [EM]) vs conventional management (grade >1 [CM]). In the second part of the study, we analyzed 3 groups, dividing the EM cohort according to the post-infusion intervention timepoint (first 24 hours from first fever spike [EM-1] vs later [EM-2]). Detailed information regarding CRS, ICANS, immune effector cell-associated hematotoxicity (N-ICAHT, neutropenia; T-ICAHT, thrombocytopenia) and infections was collected. Results: Among the 589 patients with R/R LBCL treated with axicabtagene ciloleucel (68%) or tisagenlecleucel (32%), 524 (89%) developed any grade CRS and constituted our study dataset. An early vs conventional CRS management was carried out in 159 (30%) vs 365 (70%) patients, respectively. EM included tocilizumab (69%), tocilizumab and steroids (29%) or single-agent steroids (2%). This cohort had a lower rate of patients with IPI 3-5, compared to the CM group (31% vs 47%, p=0.03). Other baseline variables were broadly similar across cohorts, with a median age of 62 years (IQR 51-69) and a male predominance (60%); most patients had a DLBCL diagnosis (80%) and 1-2 previous treatment lines (70% [2% with 1 prior line]). In both EM and CM, the main product was axi-cel (71% vs 73%, p=0.80), CAR-HEMATOTOX was usually high (57% vs 53%, p=0.50) and LDH was increased (54% vs 64%, p=0.10). Focusing on toxicity, EM patients (vs CM) presented lower rates of CRS grade 2+ (21% vs 53%, p<0.01), with similar rates of grade 3+ (5% vs 8%). The incidence of any-grade ICANS was comparable across cohorts (48% vs 48%), as was grade 3+ (18% vs 19%). In terms of steroids, we observed a trend for an increased use in the EM group (64% vs 54%, p=0.06) and a higher cumulative dexamethasone dose (mean 316 vs 180 mg, p<0.01). Concerning cytopenias, N-ICAHT grade 2+ in the first 30 days post-infusion was more common in the EM cohort (69% vs 51%, p<0.01), confirmed in the multivariate analysis. T-ICAHT was similar across groups. Finally, there was a numerically higher rate of EM patients who developed an infection (56% vs 47%, p=0.09) and a higher median number of infections (p=0.04). The rate of ICU admission was similar (16% vs 15%). In terms of efficacy, there were no differences in complete response rate (58% vs 54%) between the EM vs CM cohorts. With a median follow-up of 40 months (95%CI 36-47), progression-free survival (median 5.8 vs 6.6 months, p=0.13) and overall survival (median 23 vs 29 months, p=0.50) were also comparable. In the second part of the study, the EM cohort was split in 2 groups, according to the intervention timepoint: 93 (58%) vs 66 (42%) patients received tocilizumab and/or steroids in the first 24 hours (EM-1) vs later (EM-2). There were no differences in ICANS between the 3 cohorts. In terms of early N-ICAHT grade 2+, there was a progressive decrease across EM-1, EM-2 and CM (73%, 64%, 51%, p<0.01); the EM-1 group also presented a higher rate of late N-ICAHT (68% vs 41% vs 49%, p<0.01). Finally, more patients in the EM-1 group experienced >1 infectious event (25% vs 14% vs 15%, p=0.06). There were no significant differences in efficacy. Conclusions: In this large, real-world CAR-T cohort, an earlier (grade 1) CRS intervention did not result in a reduction of grade 3+ CRS or ICANS, and was associated with higher steroid exposure, cytopenias and infections. These data should be considered in clinical practice, tailoring the management of this patient population.