Unraveling the Pathophysiology of Neurotoxicity Induced by CAR T-cells

ABSTRACT We wish to understand how immunotherapy-induced neurotoxicity occurs. Neurotoxicity is the most pernicious side effect of several immunotherapies for B-cell leukemias and lymphomas, including CAR T-cell therapy. In the latter approach, a sample of the patient’s own T-cells are removed, genetically engineered to recognize B- cell tumors, expanded to large numbers, and then reinfused into the patient. The genetically engineered tumor- recognition component is a chimeric antigen receptor (CAR). CARs reprogram T-cells to recognize and kill tumor cells, regardless of the T-cell’s innate specificity. CAR T-cells specific for the B-cell-associated antigen CD19 can induce durable complete remissions in patients with otherwise terminal B-cell malignancies. Like any therapy, though, it has side-effects. CD19-specific CAR T-cells frequently cause a spectrum of neurological adverse effects (NAE) ranging from disorientation to death. They cannot be prevented or treated adequately because their pathophysiology is poorly understood. To that end, we developed a novel, immune- competent humanized mouse model that replicates the anti-tumor efficacy and toxicities (including NAE) caused clinically by CD19-specific CAR T-cells. In our model, mouse B-cells express a human CD19 transgene (hCD19Tg). Transfer of mouse T-cells – called CART19 cells - that express a hCD19-specific CAR into hCD19Tg mice cause NAEs that are very similar to those experienced clinically. Because our findings mirror clinical reports, we suggest the causes of CART19-induced murine NAE will extrapolate to patients treated with CD19-specific CAR T-cells. Our central hypothesis is that blood brain barrier (BBB) disruption following CART19 infusion permits leukocytes, fluids, and systemic cytokines to enter the central nervous system (CNS). Here these systemic cytokines, and to a greater extent cytokines produced in the CNS by CART19 cells, activate resident microglial cells and extravasated myeloid cells. The differentiation of both into proinflammatory cells ultimately causes NAE. We propose two aims to test these hypotheses. The first aim will reveal what causes BBB dysfunction while the second aim will determine what drives NAE. We will learn how CART19 cells cross the BBB and if their persistent activation in the CNS contributes to or drives NAE. Using genetic, immunological, and pharmacological methods, we also will assess the contributions of resident and extravasated peripheral myeloid cells and specific cytokines to NAE. Finally, we will assess how NAE affects gene and protein expression by brain parenchymal cells using single cell approaches. Our proposed project will significantly impact two areas: 1) basic research in CNS pathobiology as it relates to NAE and 2) translational research as it relates to improving CD19-specific CAR T-cell therapy for human B-cell malignancies.