Intracerebroventricular B7-H3-targeting chimeric antigen receptor T cells for non-pontine diffuse midline glioma and recurrent/refractory pediatric CNS tumors: A phase 1 trial

Abstract

BACKGROUND High-grade central nervous system (CNS) tumors carry a poor prognosis with limited curative options if first-line therapy fails. B7-H3 is expressed in many of these tumors, and chimeric antigen receptor (CAR) T cell therapy is an emerging immunotherapeutic strategy. METHODS BrainChild-03 (NCT04185038) is a single-center, dose-escalation phase 1 study of repeated intracerebroventricular (ICV) B7-H3 CAR T cells in children and young adults with recurrent/refractory CNS tumors (Arms A, B) and diffuse intrinsic pontine glioma (DIPG, Arm C). Here, we report results from Arm B, in which patients with refractory/relapsed CNS tumors or pre- or post-progression non-pontine diffuse midline glioma (DMG) received repeated ICV infusions. Primary objectives were feasibility and safety/tolerability; secondary objectives included CAR T cell detection, disease response, and survival. RESULTS Of 36 enrolled patients (atypical teratoid rhabdoid tumor n = 5, DMG n = 8, embryonal tumor with multilayer rosettes n = 2, ependymoma n = 4, high-grade glioma n = 6, medulloblastoma n = 8, pineoblastoma n = 3), manufacturing was successful for 35 patients, 26 of whom received therapy. Median age was 10 years (range 1-26). Dose escalation from 1 × 107 to 10 × 107 CAR T cells/dose identified this dose as the maximally tolerated dose regimen, with no dose-limiting toxicities observed. Across 181 total doses (median 7/patient), common adverse events included headache (n = 26), fever (n = 15), and nausea (n = 14). Median survival from first infusion was 11.5 months, ranging from 3.2 months (pineoblastoma, HGG) to 21.4 months (ependymoma); two patients achieved a partial response. CONCLUSIONS Repeated ICV B7-H3 CAR T cell dosing is feasible and tolerable across a spectrum of pediatric CNS tumors, supporting continued investigation in future trials. Children with aggressive brain and spinal cord tumors have few treatment options when standard therapy no longer works. This study tested a new immunotherapy called B7-H3 CAR T cell therapy, which uses a patient's own immune cells to recognize and attack cancer cells. Twenty-six children and young adults received repeated CAR T cell infusions directly into the fluid surrounding the brain. The treatment was successfully manufactured for nearly all patients and was generally well tolerated, with headaches, fever, and nausea being the most common side effects. Two patients experienced meaningful tumor shrinkage, and some patients lived longer than expected. These early results show that this approach is safe and feasible and support further studies to determine how well it can improve outcomes for children with these difficult-to-treat tumors.