Neurotoxicity following CD19/CD28ζ CAR T-cells in children and young adults with B-cell malignancies

Abstract

Background: Neurotoxicity is an established toxicity of CD19 CAR T-cell therapy; however, there is little information on neurotoxicity in children, adolescents, and young adults (CAYA) receiving CD19/CD28ζ CAR T-cells for B-cell malignancies.

Methods: We analyzed neurotoxicity of CD19/CD28ζ CAR T-cells in CAYA treated on a phase I study (NCT01593696). Assessments included daily inpatient monitoring, caregiver-based neuro-symptom checklist (NSC), exploratory neurocognitive assessments, clinically-indicated imaging, CSF analysis, and systematic cytokine profiling, outcomes of which were associated with cytokine release syndrome (CRS) and treatment response postinfusion. Patients with active CNS leukemia were included.

Results: Amongst 52 patients treated, 13 patients had active CNS leukemia at infusion. Neurotoxicity was seen in 11/52 (21.2%) patients, with an incidence of 29.7% (11/37) in patients with CRS. Neurotoxicity was associated with the presence and severity of CRS. Those with neurotoxicity had higher levels of peak serum IL-6, IFNγ, and IL-15. Additionally, CNS leukemia was effectively eradicated in most patients with CRS. Pilot neurocognitive testing demonstrated stable-to-improved neurocognitive test scores in most patients, albeit limited by small patient numbers. The NSC enabled caregiver input into the patient experience.

Conclusions: This is the first systematic analysis of neurotoxicity utilizing a CD19/CD28ζ CAR construct in CAYA, including in those with active CNS involvement. The experience demonstrates that the neurotoxicity profile was acceptable and reversible, with evidence of anti-leukemia response and CNS trafficking of CAR T-cells. Additionally, neurocognitive testing, while exploratory, provides an opportunity for future studies to employ systematic evaluations into neurotoxicity assessments and validation is needed in future studies.

Keywords: CAR T-cell; cytokine release syndrome; neurotoxicity.

Published by Oxford University Press on behalf of the Society for Neuro-Oncology 2022.

Figures

Fig. 1

Neurotoxicity assessments and timeline. A. Venn Diagram demonstrating the number of patients who received each part of the neurotoxicity assessments. B. Trial flow diagram for neurotoxicity assessments performed. NTX, neurotoxicity; AE, adverse events; NSC, neuro-symptom checklist; PSI, Weschler processing speed index. C. The swimmer plot shows the time course of cytokine release syndrome and neurotoxicity in each patient who developed neurotoxicity (n=11) through D28 post-CD19/CD28ζ CAR infusion. Colors indicate peak severity of neurotoxicity, width of the box represents duration of any grade neurotoxicity, and dotted line represents the median day of neurotoxicity onset in this cohort. D. Number of patients with each grade of CRS and neurotoxicity. E. % of T-cells that are CAR T-cells in the CSF in patient 11 over time. Peak % CD19 CAR T-cells in the CSF corresponded to onset of neurotoxicity at D+11.

Fig. 2

Systemic inflammation and CAR trafficking. A. % of T-cells that are CAR T-cells in the CSF at Day 28 without significant differences between those with and without neurotoxicity. B–D. Peak serum cytokine concentrations of IL-6, IFNγ, and IL-15 in all patients with CRS, comparing those with and without neurotoxicity which were significantly elevated in those who had CRS and neurotoxicity (n = 11) as compared to those who had CRS alone (n = 26). E–G. Peak serum cytokine concentrations of IL-6, GMCSF, and IL-15 restricted to those who had complete responses comparing those with and without neurotoxicity which were significantly elevated in those who had complete response and neurotoxicity. H. CSF IL-6 cytokine concentrations in patients at D28, comparing those who received tocilizumab for CRS to those who did not, demonstrated significantly elevated CSF IL-6 in those who received tocilizumab. Red boxes: patients who had neurotoxicity in each group; blue circles: patients who did not have neurotoxicity. I–L. Peak or nadir values of coagulation including prothrombin time (PT), activated partial thromboplastin time (aPTT), D-Dimer, Fibrinogen in those with and without neurotoxicity. For all graphs, *, P < .05; **, P < .01; ***, P < .001.

Fig. 3

Approach to diagnosing and managing patients for neurotoxicity. Neurotoxicity algorithm and treatment recommendations for CAYA patients undergoing CAR T-Cell Therapy. *Thorough medical history including prior treatment-related neurotoxicity, or neurological comorbidities (eg seizures); @Baseline evaluations should include neurocognitive testing, observer-reported symptom checklists, and ASTCT ICANS grading using ICE/CAPD tools; ASTCT, American Society for Transplantation and Cellular Therapy; ICANS, Immune effector cell-associated neurotoxicity syndrome; ICE, immune cell effector-associated encephalopathy score; CAPD, Cornell Assessment of Pediatric Delirium; CRS, Cytokine Release Syndrome; ICP, intracranial pressure; $Baseline Magnetic Resonance Imaging (MRI) and Neurology Consult should be done in all high-risk (previous history of neurotoxicity) patients; Consider baseline imaging in all patients if available; ^If not already performed; #Until abnormalities resolve.