Predictors of Severe Neurotoxicity in Patients Receiving Anti-CD19 CAR-T Cell Therapy for Relapsed/Refractory B-Cell Lymphomas: A Multi-Institutional Real-World Analysis

Introduction/Background/Significance: CD19-directed CAR-T cell therapy has transformed the treatment landscape for relapsed/refractory B-cell lymphomas. Nevertheless, the risk of immune effector cell-associated neurotoxicity syndrome (ICANS), which can be life-threatening, is a significant challenge. Timely identification of factors that can predict adverse neurotoxicity is important for tailoring monitoring and treatment.

Materials and Methods/Case Presentation/Objective: I performed a retrospective cohort study including 108 patients treated with axicabtagene ciloleucel, or tisagenlecleucel, at 4 academic centers from 2019–2024. Patients were stratified by ICANS grade (0–1 vs. 2–4). Baseline characteristics, inflammatory markers, baseline cytokine levels, and early clinical signs were compared across cohorts. Logistic regression modeling was utilized to determine predictors of grade ≥ 2 ICANS. Secondary endpoints included ICU admission, steroid use, and response rates

Results/Description/Main Outcome Measures: Grade ≥2 ICANS occurred in 42.6% of cases. Independent predictors of severe ICANS included high baseline ferritin ( >1,500 ng/mL; OR 3.4, CI 1.5-7.6, p=0.003), early onset fever (<24 hours) (OR 2.7, p=0.01), and pre-treatment CRP >100 mg/L (OR 2.9, p=0.006). Age, ECOG, and burden of disease were not statistically significant. Patients with ICANS stayed longer in the hospital (median 17 days versus 10 days; p< 0.01), had higher odds of being treated with steroids (89% versus 24%), and comparable 6-month complete response rates (67% versus 71%; p=0.54).

Conclusions: Severe ICANS continues to be a major barrier to CAR-T success, with baseline ferritin, CRP, and early fevers being actionable predictors; integrating these metrics into pre-infusion risk models will allow early ICU triage or preventive interventions. Moving forward, the next steps are to validate these findings in prospective trials with uniform management and to incorporate biomarker guided algorithms of toxicity into CAR-T protocols.