The CXCL16/CXCR6 axis is linked to immune effector cell-associated neurotoxicity in chimeric antigen receptor (CAR) T cell therapy.

Abstract

Background: Immune effector cell-associated neurotoxicity syndrome (ICANS) is a common and potentially life-threatening complication of chimeric antigen receptor (CAR) T cell therapy. The underlying mechanisms of ICANS remain incompletely understood and are unlikely to be explained by cytokine excess alone.

Methods: We analyzed paired peripheral blood and cerebrospinal fluid (CSF) samples from CAR T cell-treated patients who developed ICANS (n = 11) within 5-21 days post-infusion. ICANS severity was graded as follows: grade 1 (n = 3), grade 2 (n = 4), grade 3 (n = 1), and grade 4 (n = 3). Control samples were obtained from patients with idiopathic intracranial hypertension, functional neurological disorders, and multiple sclerosis. We employed single-cell RNA sequencing (scRNA-seq) and flow cytometry to profile immune cell populations and performed multi-modal spatial transcriptomics and immunofluorescence on postmortem choroid plexus and brain tissue from a patient with fatal grade 4 ICANS.

Results: We identified a distinct population of proliferating, cytotoxic T cells characterized by CXCR6 expression, enriched in CD4 + CAR T cells and predominantly localized in ICANS CSF. These CXCR6 + T cells were largely absent from control CSF samples. Spatial mapping of postmortem brain tissue revealed widespread infiltration of myeloid cells and a striking spatial association between CXCR6 + T cells and CXCL16-expressing myeloid cells in both the choroid plexus and brain parenchyma. Notably, CSF levels of CXCL16 positively correlated with ICANS severity across the cohort, from grade 1 to grade 4.

Conclusions: Our findings implicate the CXCL16/CXCR6 axis in the recruitment of cytotoxic CAR CD4 + T cells to the central nervous system (CNS) during ICANS. This interaction may be linked to neuroinflammatory processes and severity stratification in ICANS pathogenesis. These results provide a mechanistic rationale for exploring CXCL16/CXCR6 as a potential biomarker and therapeutic target in CAR T cell-associated neurotoxicity.