This study focuses on a novel CD22-targeted chimeric antigen receptor T-cell therapy (CAR-T CD22-65s) for pediatric and adult patients with B-cell acute lymphoblastic leukemia (B-ALL) who relapsed after CD19-directed immunotherapy. The aim is to address the clinical challenge of poor prognosis and limited treatment options in this patient population, providing new insights and evidence for the treatment of relapsed B-ALL.
Patient Characteristics
A total of 23 patients were screened, of whom 22 met eligibility criteria and had successful CAR-T product manufacture. The pediatric cohort included 19 patients, and the adult cohort included 3 patients (enough to complete enrollment required for adult safety run-in). Two pediatric patients were excluded due to rapid disease progression before infusion, and one adult patient was excluded based on investigator judgment. Ultimately, 17 pediatric and 2 adult patients received infusion and were evaluable for toxicity and response.
Among pediatric patients, 6 received all 3 planned doses, 9 received 2 doses, and 1 received 1 dose. Both adult patients received all 3 doses. Two pediatric patients underwent retreatment: one received only 30% of the initial product (from the supplementary dose of the first manufacture), and the other received a full 3-dose course with newly manufactured product following a second leukapheresis.
At infusion, the pediatric cohort had a median age of 16.1 years (range 3.1–28.2 years). The adult patients were aged 36 and 46 years. All patients had relapsed after prior CD19-directed immunotherapy: 16 pediatric patients (94%) relapsed with CD19-negative disease after CD19 CAR-T therapy, and 1 pediatric patient (6%) relapsed with CD19-negative disease after blinatumomab. Both adult patients relapsed with CD19-positive disease after CD19 CAR-T therapy.
Additionally, 8 pediatric patients (47%) and both adult patients had received inotuzumab ozogamicin (InO) previously, and 9 pediatric patients (53%) had undergone at least one hematopoietic cell transplantation (HCT). At infusion, 9 pediatric patients (53%) and both adult patients had bone marrow blasts >25% (M3 status).
Toxicity
Following initial CAR-T CD22-65s infusion, 15/17 pediatric patients (88%) and both adults experienced cytokine release syndrome (CRS), all grade 1 (n=8) or grade 2 (n=9), even though more than half of patients had M3 marrow disease burden. Median CRS onset was 2 days post-infusion (range 0–14 days), and 5 patients (26%) had multiple CRS episodes. Median total CRS duration was 8 days (range 3–22 days). One pediatric patient received tocilizumab for persistent fever (21 days), resolving within hours; no patients required corticosteroids or other anti-cytokine therapy.
Neurologic toxicity occurred in 24% (4/17) of pediatric patients, all grade 1 (n=3) or grade 2 (n=1), resolving spontaneously without intervention. No adult patients experienced neurotoxicity. One pediatric patient developed repeated inflammatory reactions (fever, chills, hypoperfusion, hypoxemia) associated with platelet transfusions on days 21–22, leading to platelet transfusion refractoriness and ICU admission. Symptoms improved after methylprednisolone and anakinra.
Other grade 3–4 toxicities included febrile neutropenia (n=14), disseminated intravascular coagulation (n=1), hypoxemia (n=1), epistaxis (n=1), and cytopenias persisting >8 weeks post-infusion (neutropenia n=9, anemia n=2, thrombocytopenia n=12). No IEC-HS (immune effector cell–associated hemophagocytic syndrome) and no grade 5 events were observed following initial infusion.
Both patients who received CAR-T CD22-65s retreatment developed CRS and neurotoxicity. One patient (9% marrow blasts at retreatment) experienced grade 1 CRS (two separate episodes) and grade 1 visual blurring, resolving by day 21. On day 35, this patient developed hyperferritinemia, pancytopenia, and arthralgia (without fever), meeting criteria for grade 2 IEC-HS, which resolved with supportive care. Another patient (40% marrow blasts at retreatment) developed grade 3 CRS on day 16, followed by grade 1 encephalopathy that progressed to grade 3 encephalopathy and areflexic flaccid paralysis. MRI revealed diffuse T2 hyperintensity in central and midline spinal cord. The patient was treated with tocilizumab (for CRS), dexamethasone, leucovorin, vitamin B₁₂, and levodopa/carbidopa (for neurotoxicity). CRS resolved, but the patient died of disease progression on day 31 before neurologic symptoms improved. Autopsy of the spinal cord showed diffuse white matter injury without inflammation or leukemic infiltration; the association with CAR-T CD22-65s remains uncertain.
Efficacy
At day 28 post-infusion, the overall response rate (ORR) was 74% (14/19), with 79% (11/14) achieving MRD negativity by multiparameter flow cytometry (MFC). One additional patient spontaneously cleared MRD by month 2, bringing the best overall MFC MRD negativity rate among CR patients to 86%.
Prior exposure to CD22-targeted therapy (InO) did not seem to affect response: ORR was 78% (7/9) in InO-exposed patients vs. 70% (7/10) in InO-naïve patients (Fisher’s exact p=1.0). Notably, 4/6 (67%) patients previously refractory to InO achieved CR after CAR-T CD22-65s.
Among 12 patients who achieved best response of MRD-negative CR, 5 (including 3 with prior HCT) underwent consolidative HCT at a median of 62 days post-infusion (range 50–125 days). Two patients remained in CR without further therapy, one maintained CR after additional InO and donor lymphocyte infusion (DLI) given for NGS-detected MRD, one maintained CR after starting InO upon MFC MRD detection, and one relapsed with CD22-positive disease post-HCT and died of progression.
Of the 7 MRD-negative CR patients who did not undergo HCT (including 5 with prior HCT), one remains in long-term remission beyond 3.5 years post-infusion. The remaining six relapsed at a median of 115 days (range 65–188 days), including 2 CD22-positive, 2 CD22-variable, 1 CD22-negative, and 1 with unknown CD22 status.
Among 2 MRD-positive CR patients, one (CD22-positive) progressed to morphological relapse, and the other (CD22-negative) progressed after rituximab treatment.
All 5 non-responders (NR) (2 CD22-positive, 1 CD22-negative, 1 lineage switch, 1 unknown) died of disease progression at a median of 97 days (range 33–283 days) post-infusion.
With a median potential follow-up of 38 months (range 25–50 months), median event-free survival (EFS) was 4.7 months (95% CI 2.1–not reached), relapse-free survival (RFS) was 5.3 months (95% CI 1.9–not reached), and overall survival (OS) was 12.2 months (95% CI 9.3–not reached). Twelve-month EFS, RFS, and OS rates were 28.3% (95% CI 13.5%–59.3%), 38.4% (95% CI 19.3%–76.5%), and 52.6% (95% CI 34.3%–80.6%), respectively.
Of the 2 retreatment patients, one achieved MRD-negative CR and underwent consolidative HCT but relapsed 140 days post-HCT. The other progressed at day 28.
Key Conclusions
CAR-T CD22-65s demonstrated encouraging efficacy and safety in heavily pretreated pediatric and adult B-ALL patients relapsing after CD19-directed immunotherapy.
- Efficacy: ORR was 74%, with 67% of patients refractory to InO still achieving CR. Some patients achieved durable remission after consolidative HCT, and one patient remains in remission beyond 42 months without HCT.
- Safety: Serious toxicities (≥grade 3 CRS, ICANS, or IEC-HS) were rare after initial infusion, even in patients with high disease burden (>25% blasts). Higher-grade toxicities were observed only during retreatment.
- Limitations: Without consolidative therapy, remission duration was limited (median RFS 5.3 months). Some relapses were associated with downregulation of CD22, suggesting antigen escape as a resistance mechanism.
Clinical Significance and Future Directions
Clinically, CAR-T CD22-65s provides an effective treatment option for B-ALL patients relapsing after CD19-directed therapy, particularly those with CD19-negative relapse, addressing a critical unmet need. Its favorable safety profile, even in high disease burden patients, supports clinical feasibility and paves the way for broader adoption.
The study also identified SIT1 as highly correlated with CAR-T CD22-65s expansion, suggesting potential as a serum biomarker for monitoring CAR-T cell dynamics and guiding therapy.
Future research should focus on extending remission durability, including optimizing consolidative therapy strategies (HCT timing and patient selection) and exploring combination therapies with other immunotherapies or targeted agents. Development of dual-target CAR-T cells (e.g., CD19/CD22) may also help overcome antigen escape, particularly in CD19-negative relapse. Additionally, mechanisms underlying increased toxicity during retreatment require further investigation to improve safety management. Larger multicenter trials with longer follow-up are needed to confirm efficacy and safety, and to define the role of CAR-T CD22-65s across different age groups and disease states, ultimately improving outcomes for relapsed B-ALL patients.
