Sophie Huby, Philippe Gelisse, Jean-Jacques Tudesq, Pierre Labauge, Claire Duflos, Guillaume Cartron, Marc-Antoine Gallerand, Laura Platon, Stephanie Badiou, Sylvain Lamure, Nicolas Menjot de Champfleur, Xavier Ayrignac, Guillaume Taieb

Chimeric antigen receptor (CAR) T-cell therapies have dramatically improved the prognosis of patients with relapsed or refractory hematologic malignancies; however, cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome (ICANS) occur in ∼100 and 50% of patients, respectively. This study aimed to determine whether EEG patterns may be considered as diagnostic tools for ICANS. Patients who received CAR T-cell therapy at Montpellier University Hospital between September 2020 and July 2021 were prospectively enrolled. Neurologic signs/symptoms and laboratory parameters were monitored daily for 14 days after CAR T-cell infusion. EEG and brain MRI were performed between day 6 and 8 after CAR T-cell infusion. EEG was performed again on the day of ICANS occurrence, if outside this time window. All collected data were compared between patients with and without ICANS. Thirty-eight consecutive patients were enrolled (14 women; median age: 65 years, interquartile range: [55-74]). ICANS was observed in 17 of 38 patients (44%) after a median time of 6 days after CAR T-cell infusion (4-8). The median ICANS grade was 2 (1-3). Higher C-reactive protein peak (146 mg/L [86-256], p = 0.004) at day 4 (3-6), lower natremia (131 mmol/L [129-132], p = 0.005) at day 5 (3-6), and frontal intermittent rhythmic delta activity (FIRDA, p < 0.001) on EEG between days 6 and 8 after infusion were correlated with ICANS occurrence. FIRDA was only observed in patients with ICANS (N = 15/17, sensitivity of 88%) and disappeared after ICANS resolution, usually after steroid therapy. Except for hyponatremia, no other toxic/metabolic marker was associated with FIRDA (p = 0.002). The plasma concentration of copeptin, a surrogate marker of antidiuretic hormone secretion, assessed at day 7 after infusion, was significantly higher in patients with (N = 8) than without (N = 6) ICANS (p = 0.043). FIRDA is a reliable diagnostic tool for ICANS, with a sensitivity of 88% and a negative predictive value of 100%. Moreover, as this EEG pattern disappeared concomitantly with ICANS resolution, FIRDA could be used to monitor neurotoxicity. Finally, our study suggests a pathogenic pathway that starts with increased C-reactive protein, followed by hyponatremia and eventually ICANS and FIRDA. More studies are required to confirm our results. This study provides Class III evidence that FIRDA on spot EEG accurately distinguishes patients with ICANS compared with those without after CAR T-cell therapy for hematologic malignancy.