Paper Details

ABSTRACT

Introduction: Relapse of B-ALL after CD19 CAR T cells (CART19) is predominantly driven by antigen escape or inadequate CAR T cell persistence. We hypothesized that co-infusing CD22 CAR T cells (CART22-65s) with humanized CART19 (huCART19) could counter both mechanisms of CART19 failure. We report interim Phase 1/2 trial results using this dual-targeted strategy (NCT05674175).

Methods: Patients <30 years with relapse after prior CART19 (retreatment cohort) or no prior CART19 exposure (CAR-naïve cohort) received co-infused CART22-65s and huCART19 after fludarabine/cyclophosphamide lymphodepletion. The initial dose (DL1) was 2x10⁶ CAR T cells/kg of each product. Higher rates of severe immune effector cell-associated HLH-like syndrome (IEC-HS) prompted a reduction in the CART22-65s dose to 0.5x10⁶ CAR T cells/kg (DL-1). Phase 1 assessed safety, while phase 2 assessed the complete remission rate (CR). Secondary endpoints included relapse-free survival (RFS; relapse defined as ≥0.01% marrow blasts or extramedullary disease [EMD]), overall survival (OS) and CAR T cell expansion and persistence.

Results: From March 2023 to April 2026, 46 patients were infused (DL1, n=27; DL-1, n=19). The retreatment cohort (n=35) was highly refractory: 57% were aged ≥18 years, 57% were in ≥3^rd relapse, 31% had prior hematopoietic cell transplant (HCT), inotuzumab, blinatumomab, and CART19, and 46% had ≥5% blasts or EMD at infusion. The CAR-naïve cohort (n=11) was younger (55% <13 years), mostly in 1^st relapse (82%), and had lower disease burden (18% ≥5% blasts).  

In vivo cellular kinetics by qPCR demonstrated robust expansion of both CARTs: huCART19 peaked at day 10 (median 30,357 copies/mg DNA) while CART22-65 peaked at day 21 (187,174 copies/mg DNA at DL1, 35,415 at DL-1). These kinetics aligned with distinct toxicity events. CRS clustered around huCART19 expansion, whereas a 2^nd CRS event or IEC-HS aligned with CART22-65s expansion. CRS occurred in 30 (65%) patients (grade 3-4, 7.5%) and ICANS in 1 patient (grade 3).

IEC-HS was frequent at DL1, occurring in 13/27 (48%) patients, with grade 3-4 events in 5 (21%). IEC-HS rates decreased after the CART22-65s dose reduction, with 4/19 (21%) developing IEC-HS and no grade ≥3 events. Median onset was day 23 (IQR 14-28) and median duration was 28 days (IQR 19-33). Laboratory correlates of IEC-HS occurrence and severity included elevated ferritin, LDH, CXCL9, IFNγ, and IL-18. Eight patients received IEC-HS directed therapy, including anakinra (n=7), corticosteroids (n=4), emapalumab (n=4), ruxolitinib (n=3), eculizumab (n=3), cytosorb (n=1) and/or camoteskimab (n=1). Cyclophosphamide 1.5g/m² was administered to a patient with grade 4 IEC-HS complicated by multisystem organ failure, resulting in complete resolution of symptoms and transient CART ablation. HuCART19 then re-expanded at 5 weeks without recurrent toxicity while CART22-65s remained undetectable. The patient remains in CR with huCART19 persistence for >1 year.

All patients were evaluable for response. At day 28, measurable residual disease (MRD)-negative CR rates by flow cytometry were 89% (31/35) in the retreatment and 100% (11/11) in the CAR-naïve cohort. With a median follow-up of 20 months, 2 year OS and RFS were 69% (95% CI 53-90) and 54% (95% CI 38-78) in the retreatment cohort compared with 100% (95% CI 100) and 89% (95% CI 71-100) in the CAR-naïve cohort. Six patients underwent consolidative HCT. Among 14 relapses, 9 were CD19+/CD22+ and 5 were CD19-/CD22+, 4 of which had evidence of CD19 downregulation pre-infusion.

Functional persistence was evaluated using competing risks analysis of time to B cell recovery (BCR) or CD19+/CD22+ relapse. At 6 months post-infusion, the cumulative incidence of BCR was 30% (95% CI 16-47) in the retreatment and 30% (95% CI 6-59) in the CAR-naïve cohort. Despite distinct early expansion profiles, qPCR measurements of CAR transgenes from months 2-12 demonstrated largely parallel persistence of both CARs.  

Conclusions: Dual-targeting with co-infused CART22-65s and huCART19 is an effective salvage therapy in children and young adults with B-ALL refractory to prior CART19 and multiple cellular/immunotherapies. Early outcomes in CAR-naïve patients are also encouraging. Reducing the CART22-65s dose improved safety, with no grade ≥3 IEC-HS events at DL-1. Ongoing follow-up focuses on long-term dual-CAR T cell persistence and optimization of IEC-HS prevention and management strategies.