Background: Adults with high-risk (HR), relapsed/refractory (r/r) B-ALL who receive CAR-T often require consolidative allogeneic HCT to achieve long-term remission. Emerging data suggests that adult B-ALL patients (pts) with prior HCT have superior PFS relative to HCT-naïve pts (Roloff JCO 2025). Our group is interested in combining CAR-T cells with allogeneic HCT to determine whether an all-in-one strategy is feasible and efficacious in HR B-ALL. We conducted two successive Phase 1 trials in adults with HR B-ALL using the same CD19/22 CAR construct. The first tested autologous CAR19/22 in adults with r/r or MRD+ B-ALL (Spiegel Nat Med 2021; NCT03233854). The second tested donor-derived, allogeneic CAR19/22 as a component of Orca-T in pts with HR genomics, MRD+, or r/r B-ALL (NCT05507827). We present results comparing toxicity, efficacy, and CAR kinetics of these two approaches.

Methods: All pts received CAR-T cells engineered with a bivalent lentiviral vector to express CAR19/22 (murine anti-CD19 FMC63/fully human anti-CD22 m971 scFv, 41BB co-stim). Autologous CAR recipients (auto cohort) received Flu/Cy followed by CAR19/22 (1 or 3x106cells/kg) +/- IL15 receptor agonist. Allogeneic recipients (allo cohort) received myeloablative conditioning followed by Orca-T (HSPCs plus regulatory T cells on D0; conventional T cells on D+2 as per Meyer Blood 2025; NCT01660607) and donor-derived, allogeneic CD19/22 (1, 2, or 3x106cells/kg) on D+2 from the same donor. Single-agent tacrolimus began on D+3 and continued until ~D+180. Pts with >5% marrow blasts or prior CAR-T were excluded from the current analysis.

Results: Twenty-eight pts with <5% marrow blasts at enrollment were included: 13 auto and 15 allo. 100% of the auto and 60% of the allo cohort received the maximum tested CAR-T dose of 3x106cells/kg. Median age was 36, 50% were female, and 65% were Hispanic; demographics were similar across cohorts. Adverse-risk ALL genomics included Ph-like (auto 23%, allo 20%), complex karyotype (auto 8%, allo 13%), KMT2ar (auto 8%, allo 7%), and TP53m (auto 0, allo 7%). Among the auto cohort, 70% had prior HCT. At enrollment, 92% of auto and 40% of allo pts had detectable MRD.

Toxicities were mild in both cohorts; no grade 3-4 CAR-related toxicities were reported. Among the auto cohort, 69% experienced CRS (46% grade 1, 23% grade 2); among the allo cohort, 100% experienced CRS (80% grade 1, 20% grade 2). ICANS was rare, with only one grade 1 case in the allo cohort. All pts engrafted neutrophils (median days: auto 7, allo 13); platelet recovery occurred in 92% of auto and 100% of allo pts. Bacterial infections were more common in allo (auto 15%, allo 47%) while viral infections were similar (auto 15%, allo 13%); all infections were treated to resolution. Two pts (1 in auto, 1 in allo) developed mod/severe chronic GVHD following CAR19/22. GVHD began in the auto recipient following HCT and was worsened by CAR19/22. Three pts in the auto cohort received post-CAR consolidative HCT in CR. Thus, 12 out of 13 pts (92.3%) in the auto cohort received HCT during B-ALL treatment.

At a median follow-up of 2.5 yrs, 7 pts died in the auto cohort; 6 from refractory B-ALL. There have been no relapses or deaths in the allo cohort. Estimated 18-month PFS was 38.5% and 100% (p<0.001), while OS was 77% and 100% (p=0.176) in the auto and allo cohorts, respectively. Within the first 28 days, peak CAR expansion (auto 1,365 vs allo 1,096, p=0.552) and AUC (auto 13,288 vs allo 13,057, p=0.988) were comparable. Median time to peak expansion was also similar (auto 13 days; allo 13 days). CARs persisted longer in the allo compared to auto recipients. At last follow-up, CARs were detectable by flow (LOD 10-4) in all but 2 pts in the allo cohort, with a median CAR persistence of 270 days. Conversely, median CAR persistence was 28 days in the auto cohort. Product characterization will be available by the time of the ASH Annual Meeting.

Conclusion: In adults with HR B-ALL, Orca-T-allogeneic CAR 19/22 resulted in superior relapse prevention, survival, and CAR persistence relative to autologous CAR 19/22. We hypothesize this is due to tolerance for CAR molecules, resulting in persistent CAR expression and improved antitumor activity. Safety data were comparable between the two cohorts. Most adults with HR B-ALL receive dyssynchronous CAR-T and HCT during their treatment course. Thus, all-in-one HCT-allo-CAR represents a rational approach that warrants additional study.

This content is only available as a PDF.
2025
Sign in via your Institution