Model-predicted fludarabine exposure and outcomes after tisagenlecleucel in children and young adults with B-Cell lymphoblastic leukemia.
AI interpretation is pending for this paper.
Open original publication →What the AI sees
Not AI summarized yet.
Research significance
Pending deeper interpretation.
Source abstract
BACKGROUND: Tisagenlecleucel (tisa-cel) is a chimeric antigen receptor T-cell (CAR-T) therapy that provides meaningful responses and survival for children and young adults with relapsed/refractory B-cell acute lymphoblastic leukemia (r/r B-ALL). Although more than 80% of those treated achieve complete remission, lack of CAR-T cell persistence remains a concern, and 40-50% experience relapse within a year. Lymphodepleting chemotherapy (LD) with fludarabine (FLU) and cyclophosphamide (Cy) is essential for CAR-T expansion and persistence. In children, FLU pharmacokinetic (PK) variability occurs using conventional body-surface-area-based dosing. Population pharmacokinetic (popPK) modeling enables individualized prediction of FLU exposure using patient-specific characteristics and administered FLU dosing information. We hypothesized that popPK-model-predicted FLU cumulative area under the curve (cAUC; mg·hr/L) would be associated with clinical outcomes after tisa-cel infusion in pediatric and AYA recipients. We sought to characterize exposure-response relationships and identify FLU exposure ranges associated with improved clinical outcomes. METHODS: We conducted a retrospective, single-center cohort study of pediatric patients receiving standard LD (FLU + Cy) before tisa-cel between November 2017 and October 2024 at the University of Minnesota. Patients were administered FLU at a dose of 30 mg/m2 intravenous daily for 4 days per institutional standard of care. Individual FLU cAUC values were retrospectively calculated using a validated popPK model implemented through Bayesian estimation. Outcomes included 1-year relapse-free survival (RFS), overall survival (OS), and relapse-free, loss-of-B-cell-aplasia (LBCA)-free survival. Cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) were assessed through Day 100. Pre-LD disease burden was categorized as high (>5%), low (<5%), or negative by flow cytometry. Hazard Ratio (HR) function-derived exposure threshold analysis was used to identify FLU cAUC cutoffs for 1-year RFS. Survival and toxicity analyses employed Kaplan-Meier estimates, cumulative incidence with competing risks, and Fine-Gray regression. Group comparisons used Wilcoxon rank-sum and chi-square tests. RESULTS: Thirty-four patients were included (median age 14.4 years, range 1.2-24.4; median weight 50.0 kg, range 9.8-97.3). HR function-derived exposure threshold analysis identified 3 FLU exposure groups: LOW (<18.5 mg·hr/L, n = 8), OPTIMAL (18.5-21.7, n = 11), and HIGH (≥21.7, n = 15). Age, weight, and disease burden distributions were similar across exposure groups. In univariate analysis (UVA), 1-year RFS differed significantly by FLU exposure group, with rates of 25% (95% CI, 4-56%) in LOW, 81% (95% CI, 42-95%) in OPTIMAL, and 60% (95% CI, 32-80%) in HIGH exposure groups (p = 0.04). In multivariable analysis adjusted for disease burden, patients in the OPTIMAL and HIGH exposure groups demonstrated a trended towards improved 1-year RFS compared to LOW exposure group (HR 0.16 (95% CI, 0.03-0.83) and 0.42 (0.13-1.34); p = 0.07)). Patients with low or negative disease burden had significantly better 1-year OS (90%, 95% CI, 65-98%) compared with those with high burden (55%, 95% CI, 26-76%; p < 0.01). CONCLUSION: Lower model-predicted FLU exposure was associated with inferior relapse-related outcomes following tisa-cel infusion in pediatric and AYA patients. These findings support further evaluation of popPK-guided model-based dosing strategies to individualize FLU dosing and avoid underexposure during LD as a practical approach to improving long-term CAR-T effectiveness.