Identification of independent prognostic factors in relapsed pediatric B-ALL patients receiving CD19 CAR-T without consolidative HSCT.

Abstract

249 Background: After CD19 chimeric antigen receptor (CAR) T-cell therapy (CAR-T), some B cell acute lymphoblastic leukemia (B-ALL) patients do not proceed to consolidative hematopoietic stem cell transplantation (HSCT) for various reasons. However, the specific characteristics of patients who can achieve long-term survival without subsequent HSCT remain to be elucidated. This study aimed to identify independent predictors of Relapse-Free Survival (RFS) in patients who achieved complete remission after CD19 CAR-T but did not undergo subsequent HSCT. Methods: We conducted a single-center retrospective study in a cohort of relapsed pediatric B-ALL patients (n=52) receiving CD19 CAR-T without consolidative HSCT. Candidate variables were first screened using the Least Absolute Shrinkage and Selection Operator (Lasso) regression. Selected variables were then analyzed in a multivariable penalized Firth Cox proportional hazards model to address the issues of small sample size and complete separation. Results: The median follow-up was 39 months. The 3-year RFS was 40.9% (95% CI: 28.5-58.6%). The Lasso regression selected four variables including B-cell aplasia (BCA) duration after infusion, minimal residual disease (MRD) level before infusion, time to initial relapse and relapse site for the final model. The subsequent stratified Firth Cox model was highly significant (Likelihood Ratio Test p = 8.1e-07). BCA duration (≥ 6 months vs. < 6 months) was a strong independent protective factor (Hazard Ratio, HR = 0.032; 95% CI: 0.00025 - 0.230; p < 0.001). Relapse site (bone marrow relapse vs. isolated extramedullary relapse) was a significant independent risk factor (HR = 3.47; 95% CI: 1.06 – 11.57; p = 0.039). The effects of MRD and the stratified time to relapse were not statistically significant. Conclusions: Patients with BCA duration ≥ 6 months after CD19 CAR-T infusion and isolated extramedullary relapse have a significantly more favorable prognosis, defining a potential low-risk subgroup. This finding requires validation in studies with comparative treatment arms to assess implications for therapeutic strategy.