Antithymocyte globulin exposure in CD34(+) T-cell–depleted allogeneic hematopoietic cell transplantation

Traditional weight-based dosing results in variable rabbit antithymocyte globulin (rATG) clearance that can delay CD4(+) T-cell immune reconstitution (CD4(+) IR) leading to higher mortality. In a retrospective pharmacokinetic/pharmacodynamic (PK/PD) analysis of patients undergoing their first CD34(+) T-cell–depleted (TCD) allogeneic hematopoietic cell transplantation (HCT) after myeloablative conditioning with rATG, we estimated post-HCT rATG exposure as area under the curve (arbitrary unit per day/milliliter [AU × day/mL]) using a validated population PK model. We related rATG exposure to nonrelapse mortality (NRM), CD4(+) IR (CD4(+) ≥50 cells per µL at 2 consecutive measures within 100 days after HCT), overall survival, relapse, and acute graft-versus-host disease (aGVHD) to define an optimal rATG exposure. We used Cox proportional hazard models and multistate competing risk models for analysis. In all, 554 patients were included (age range, 0.1-73 years). Median post-HCT rATG exposure was 47 AU × day/mL (range, 0-101 AU × day/mL). Low post-HCT area under the curve (<30 AU × day/mL) was associated with lower risk of NRM (P < .01) and higher probability of achieving CD4(+) IR (P < .001). Patients who attained CD4(+) IR had a sevenfold lower 5-year NRM (P < .0001). The probability of achieving CD4(+) IR was 2.5-fold higher in the <30 AU × day/mL group compared with 30-55 AU × day/mL and threefold higher in the <30 AU × day/mL group compared with the ≥55 AU × day/mL group. In multivariable analyses, post-HCT rATG exposure ≥55 AU × day/mL was associated with an increased risk of NRM (hazard ratio, 3.42; 95% confidence interval, 1.26-9.30). In the malignancy subgroup (n = 515), a tenfold increased NRM was observed in the ≥55 AU × day/mL group, and a sevenfold increased NRM was observed in the 30-55 AU × day/mL group compared with the <30 AU × day/mL group. Post-HCT rATG exposure ≥55 AU × day/mL was associated with higher risk of a GVHD (hazard ratio, 2.28; 95% confidence interval, 1.01-5.16). High post-HCT rATG exposure is associated with higher NRM secondary to poor CD4(+) IR after TCD HCT. Using personalized PK-directed rATG dosing to achieve optimal exposure may improve survival after HCT.