Chimeric Antigen Receptor (CAR) therapy targeting CD19 achieves complete remission (CR) rates of 70%-90% in relapsed/refractory B-ALL. Relapse due to loss of the CD19 targeted epitope presents a therapeutic challenge as evidenced by the largest global pediatric CD19-CAR experience which showed 15 of 16 relapses to be explained by CD19 downregulation (Maude et al, NEJM 2018). Alternatively targeting CD22 using CD22-CAR therapy has demonstrated a CR rate of approximately 70% in both CD19⁺ and CD19^- B-ALL, however relapse due to CD22 downregulation limits the curative potential of singularly-targeting CD22 (Fry et al, Nat Med. 2018). We hypothesized that simultaneous targeting of CD19 and CD22 via a bispecific CAR-T cell would be a safe and tolerable treatment strategy in relapsed/refractory B-cell ALL and address immune evasion. Here, we report the first clinical experience in pediatric patients using bispecific CD19-CD22 CAR T cells.

We describe a single institution phase I dose escalation study in pediatric patients with relapsed or refractory B cell ALL. We utilized lentiviral transduction of a bivalent CAR construct incorporating the fmc63 CD19 and m971 CD22 single chain variable fragments (scFvs) used in clinically tested CAR constructs and a 41BB costimulatory endodomain (Fry et al, Nat Med. 2018). Our primary objectives are feasibility of production of this bivalent CAR and safety at 3 dose escalation levels (1x10⁶, 3x10⁶ and 1x10⁷ CAR T cells/kg). Clinical response assessment is evaluated as a secondary aim. All patients described received lymphodepletion with fludarabine (25mg/m² x 3 days) and cyclophosphamide (900mg/m² x 1) followed by fresh or cryopreserved CAR T cell infusion after a 7-9 day production time. Patients were prospectively monitored at predefined intervals for disease response and correlative assessments.

Four pediatric patients with precursor-B ALL, age 2-17, have been enrolled and treated with CD19/CD22 bispecific CAR T cells at dose level 1 (1x10⁶) [Table 1]. Three patients entered CAR therapy with low disease burden detected by minimal residual disease (MRD) alone and 1 patient initiated therapy with 12% bone marrow blasts. All patients were CNS1 at time of treatment. The toxicity profile mirrored that of the singular CD19 and CD22 CAR experience with 3 patients experiencing reversible CRS (2 Grade I, 1 Grade II), onset day 3-8, and 2 patients experiencing grade I neurotoxicity, onset day 3-9. In our cohort, we experienced lower grade toxicities than previously reported, likely due to a mean lower disease burden. Only 1 patient with CRS met criteria for tocilizumab and this patient was the singular study patient treated with higher burden disease. Neurotoxicity was managed with supportive care and fully reversible. Peripheral blood flow cytometry analysis detects circulating CAR by day 6 in all patients and demonstrates peak CAR expansion between day 6-10. Peak CAR T expansion reached levels of 10-25% of total T cells with inter-patient variability in CD4 and CD8 predominance, favoring CD8 expansion in 3 of 4 patients. Clinical symptoms and inflammatory markers expectedly correlate with peak CAR expansion. Four of 4 patients achieved complete remission (CR) at day 28 post-CD19/CD22 bispecific CAR therapy. Three of 4 patients demonstrated MRD^- remissions by flow cytometry and of these, next generation sequencing (NGS) was negative where available (N=2). Multi-parametric CyTOF studies permitting CAR T cell phenotyping in conjunction with single cell TCR tracking, proteomics, epigenomics and cytokine profiling are ongoing and will be used to further …