We hypothesized that inferior disease-free survival (DFS) seen in older patients undergoing αβ/CD19-T-cell depleted (AB-TCD) haploidentical hematopoietic cell transplantation (HCT) for patients with hematologic malignancies was due to excessive exposure to rabbit antithymocyte globulin (rATG; Thymoglobulin®). Between 2015-2023, 163 patients with a median age of 13 years (range, 0.4-27.4) underwent AB-TCD haploidentical HCT for treatment of ALL (n=98), AML/MDS (n=49), or other malignancies (n=16) at nine centers on two prospective trials. Exposures of rATG pre- and post-HCT were predicted with a validated pharmacokinetic (PK) model. ROC curves were used to identify optimal target windows of rATG exposure related to outcomes. We identified four quadrants of rATG exposure - quadrant 1 (n=52): high pre-HCT AUC (≥50 AU*day/mL) and low post-HCT (<12 AU*day/L); quadrant 2 (n=47): both low pre-HCT and post-HCT AUCs, quadrant 3 (n=13): low pre-HCT AUC and high post-HCT, and quadrant 4 (n=51): both high pre- and post-HCT AUCs. Quadrant 1 had a 3-year DFS of 86.5% (95% CI, 76.3-96.7%), compared to quadrant 2 (64.6%; 95% CI, 49.1-80.1%), quadrant 3 (32.9%; 95% CI, 0.1-80.5%) or quadrant 4 (48.2%; 95% CI, 22.1-63.3%) (p<0.001). Adjusted regression analysis demonstrated additional factors associated with increased hazard for worse DFS: MRD-positivity (HR=2.45; 1.36-4.41; p=0.003) and CMV R+/D- serostatus (HR=3.33; 1.8-6.16; p<0.001). Non-optimal rATG exposure exhibited the strongest effect in unadjusted (HR=4.24; 1.79-10.03; p=0.001) and adjusted (MRD status or CMV serostatus) analyses (HR=3.84, 1.63-9.05; p=0.002). High exposure to rATG post-HCT is associated with inferior DFS following AB-TCD haploidentical HCT for pediatric patients with hematologic malignancies. Model-based dosing of rATG to achieve optimal exposure may improve DFS. Clinical trials: NCT02646839 & NCT04337515
