Abstract
Corticosteroids are commonly used to manage cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) following chimeric antigen receptor (CAR) T-cell therapy, and yet, their dose-specific impact on outcomes remains uncertain. We retrospectively evaluated 276 adults with large B-cell lymphoma (LBCL) treated with CD19-directed CAR-T therapy (axi-cel, tisa-cel, or liso-cel) between 2016 and 2023 at a single institution. Cumulative corticosteroid dose was defined as the total dose administered within 21 days of infusion. Corticosteroids were administered to 105 patients (38%), initiated at a median of 5 days post-infusion (interquartile range [IQR] 3-7) for CRS (38%), ICANS (15%), or both (47%). Use was more frequent with axi-cel (P < 0.001), although the cumulative dose and duration were similar across products. To assess the impact of corticosteroid exposure, we carried out a 21-day landmark analysis. Corticosteroid exposure, modeled as a time-dependent covariate, was not significantly associated with infection risk, non-relapse mortality, or inferior overall survival (OS) or progression-free survival (PFS). However, in a landmark analysis, patients receiving above-median cumulative corticosteroid doses had a significantly higher risk of infection compared to those receiving below-median corticosteroid doses or no corticosteroids (P = 0.042). In a landmark multivariable analysis, corticosteroid cumulative dose was associated with increased late hematologic toxicity (adjusted hazard ratio [HR] 1.02, 95% CI 1.02-1.03). Finally, in a sensitivity analysis excluding patients with Grade ≥4 CRS/ICANS, corticosteroid cumulative dose remained unassociated with OS or relapse, but was linked to shorter PFS (adjusted HR 1.04, 95% CI 1.01-1.06). These findings support the safe yet judicious use of corticosteroids to manage CAR-T toxicities in LBCL.