CD19 CAR T cell therapy is an effective strategy for first CNS relapse in pediatric b ALL

Abstract Introduction: The 4-year disease-free survival (DFS) of children with first isolated central nervous system (iCNS) relapse treated on the most recent Children's Oncology Group AALL1331 randomized phase 3 trial was extremely poor at 24%, and the addition of blinatumomab failed to improve outcomes. Cranial radiation (cXRT) is the standard of care (SOC) for CNS relapse but can have profound neurologic consequences. More effective and less toxic strategies are urgently needed. We hypothesized that CD19-targeted CAR T cell therapy (CART19) could lead to durable remissions in patients with CNS-relapsed B-ALL and obviate the need for cXRT or hematopoietic cell transplant (HCT). We conducted a Phase 2 trial of CTL019, the construct that was FDA approved as tisagenlecleucel, for this population (NCT04276870). Methods: Patients 0-29 years (y) of age with CNS relapse of B-ALL without receipt of cXRT for this relapse were lymphodepleted with fludarabine/cyclophosphamide prior to infusion with 5x106 CART19 cells/kg. Results: From March 2020 to April 2025, 43 patients were enrolled and infused with CART19; 39 (91%) had first CNS relapse (3 second, 1 third). Thirty-six patients (84%) had iCNS relapse. Median age at infusion was 8y (range 1-26y); 44% were female. Six (14%) patients were Hispanic; of those not Hispanic, 72% were White, 11% other, 8% Black, 6% unknown, 3% Asian. Only the 4 patients in ≥2nd relapse had received prior cXRT and/or total body irradiation (TBI). Ten (23%) received prior blinatumomab. At infusion, 6 patients (14%) had marrow disease, 8 patients (19%) had CNS2 disease (cerebrospinal fluid (CSF) white blood cell count (WBC) 5 with blasts), and 2 (5%) had CNS disease by imaging. No patients had CNS3 disease by CSF evaluation (CSF WBC ≥5 with blasts). Of the 42 evaluable patients 28 days after CART19 infusion, 41 (98%) were in CNS remission, and all with marrow disease at infusion were MRD negative. One patient was CNS2 at day 28 and cleared at month 2 without intervention for a best overall response rate of 100%. With a median follow-up of 12 months (m) and mean of 22m (range 1-55m), 1y and 2y EFS were both 90% (95% CI 81-100%) when censoring for alternative therapy. Shorter CART19 persistence, as evidenced by B-cell recovery (BCR), led to 12 patients pursuing other therapy (HCT, n=2; other CART, n=10). EFS without censoring for alternative therapy was 90% (95% CI 81-100%) at 1y and 85% (95% CI 73-99%) at 2y. Among the 14 patients with 4 years of potential follow-up time at data cutoff, 4y EFS was 65% (95% CI 42-100%) and 71% (95% CI 51-99%) with and without censoring for alternative therapy, respectively, and OS was 93% (95% CI 80-100%). Four patients experienced morphologic relapse (1 medullary, 1 combined, 1 CNS2, 1 lineage switch); 2 had MRD-level relapse. Four relapses were CD19 positive, and 2 were CD19 negative. Three patients were successfully bridged to HCT, 1 went to other CART, and 2 were being bridged to HCT at the time of data cut off. All 5 patients who went to HCT after CART19 (early BCR, n=2; relapse, n=3) received TBI; all other patients avoided cXRT. Overall survival (OS) at 1y was 100% and 90% (95% CI 78-100%) at 2y. CART19 proliferation by qPCR in peripheral blood peaked at a median of 9 days (range 6-14) post-infusion with a median of 31,620 copies/μg genomic DNA. CART19 was detectable by qPCR in CSF at day 28 in 98% of evaluable patients. Manufacturing optimization reduced the cumulative incidence of early BCR at 6m from 67% (95% CI 35-85%) for the first 16 patients treated to 39% (95% CI 19-59%) for the subsequent 27 patients treated after the changes, representing enhanced CART19 persistence. Cytokine release syndrome (CRS) was observed in 33 patients (77%) after initial infusion, only 2 of which had higher grade events (1 grade 2; 1 grade 4). Four patients (93%) had grade 3 neurotoxicity events: focal cerebral edema, seizure, and encephalopathy (n=2). All fully recovered; no grade 4 or 5 toxicity occurred. Conclusions:In this Phase 2 trial, CART19 achieved an 85% 2-year EFS in CNS-relapsed B-ALL with 88% (38/43) of patients avoiding toxic cranial radiation, the SOC for CNS relapse for decades. While longer follow-up is needed, the 4y EFS of 71% and OS of 93% in the 14 patients with 4+ years of follow-up is promising and is superior to the outcomes on AALL1331. This study demonstrates that CART19 is an effective and less toxic strategy for ≥1st CNS relapse of pediatric B-ALL.