Background: CNS relapse of B-ALL is difficult to treat after cranial radiation or multiple relapses. Durable remissions of relapsed/refractory (r/r) B-ALL have been seen with CD19 CAR T cells; however, most trials excluded patients with active CNS disease. As we observed CAR trafficking into the CSF, we hypothesized that CD19 CAR T cells could control CNS B-ALL. Methods: We identified children and young adults with r/r CNS B-ALL treated on 4 clinical trials of CD19 CAR T cells, CTL019 or CTL119. NCT01626495 and NCT02435849 excluded active CNS disease, while the former in an amendment as well as NCT02374333 and NCT02906371 permitted active CNS disease controlled on therapy. All trials permitted CNS disease that cleared and excluded bulky intracranial disease that did not improve. We analyzed outcomes (CR, RFS) and safety. Results: We identified 65 patients 1-29y (median 10y) with r/r CNS B-ALL (CNS+) of 182 treated with CTL019/CTL119. There were no differences in age, sex, history of SCT or neurologic comorbidities in the CNS+ and CNS- cohorts. CNS+ patients were more likely to be in ≥2^nd relapse (74% vs 46%, p < 0.01), to have received cranial radiation (58% vs 11%, p < 0.01), to have detectable CNS disease (p = 0.02) and less bone marrow disease pre-infusion (p < 0.01). At 1 mo post infusion, 62 (95%) CNS+ and 110 (94%) CNS- patients were in CR; 1 in each cohort died of sequelae of CRS and was inevaluable for response. All patients with CNS disease detected pre-infusion cleared by mo 3, including 9 in the CNS+ cohort [5 CNS2 ( < 5 CSF WBC with blasts), 4 CNS3 ( > 5 CSF WBC with blasts or exam/imaging evidence)] and 8 in the CNS- cohort (isolated CNS2 status pre-infusion). There was no difference in RFS (p = 0.28) in the CNS+ and CNS- cohorts [24-mo RFS: 61% (95% CI 46-73%) and 60% (95% CI 48-70%)]. There were 4 CNS relapses in the CNS+ cohort, and 1 in the CNS- cohort. Encephalopathy rate and grade was similar in the CNS+ and CNS- cohorts (52% vs 40% any grade; 12% vs 11% grade 3/4; p = 0.41). There were no deaths due to neurotoxicity (NT) and no statistically significant differences in incidence or severity of any NT or CRS in the CNS+ and CNS- cohorts. Conclusions: The CD19 CAR T cell therapies CTL019/CTL119 are effective at clearing CNS disease and inducing durable remissions in children and young adults with r/r CNS B-ALL. CNS relapse rates are low ( < 3%). Most CD19 CAR T cell trials excluded patients with active CNS disease, primarily due to the risk of NT. We show that patients with r/r CNS B-ALL that is adequately controlled prior to infusion can be safely treated with CD19 CAR T cells, with no increased risk of NT. Clinical trial information: NCT01626495, NCT02435849, NCT02374333, NCT02906371

None