EEG as a predictive biomarker of neurotoxicity in anti-CD19 CAR T-cell therapy

Objective Immune effector cell-associated neurotoxicity syndrome (ICANS) is a potentially fatal complication of CD19-directed CAR T-cell therapy. The aim of this study was to investigate the role of EEG as a predictive biomarker of ICANS. Methods In this prospective, monocentric, cohort study, consecutive refractory B-cell non-Hodgkin lymphoma patients undergoing CAR T-cell therapy had EEG assessments at fixed time points pre- and post-infusion. The risk of ICANS was evaluated according to EEG findings detected qualitatively, using a grading scale ranging from 0 (normal) to 3 (severely abnormal), and quantitatively, using power spectral and connectivity measures. Results 307 EEGs from 68 patients have been qualitatively evaluated, of whom 238 were eligible for quantitative analysis. Neurotoxicity manifested in 22/68 (32.4%) patients. Pre-infusion EEG abnormalities (grade 1 and 2) were qualitatively detected in 8/68 (11.7%) patients, emerging as a risk factor for ICANS [HR 5.8 (95%CI 2.6–12.9)]. Quantitative analysis of pre-infusion EEGs did not yield significative results. Post-infusion qualitative EEG abnormalities were associated to a higher risk of ICANS development [HR 11.6 (4.4–30.5) for grade 2; HR 9.7 (2.6–36.6) for grade 3]. Concerning the quantitative analysis, in post-infusion EEGs higher theta energy [HR 1.10 (1.03–1.16)] and delta + theta/alfa ratio [HR 1.37 (1.11–1.67)] were associated to higher risk of ICANS, while higher beta energy resulted protective [HR 0.91 (0.85–0.97)]. Conclusions Our study establishes EEG as a predictive tool for identifying patients at risk for ICANS before CAR T-cell infusion, who may benefit from prophylactic treatments, and anticipating ICANS onset following infusion, enabling early intervention.