Abstract
Anti-thymocyte globulin (ATG) is a polyclonal antibody used to prevent graft-versus-host disease in allogeneic hematopoietic stem cell transplantation. While recent studies have worked towards understanding the pharmacokinetic-pharmacodynamic relationship via population pharmacokinetic modeling, dosing remains weight-based and largely empirical in clinical practice. In this study, conventional and novel ATG dosing regimens were evaluated using Monte Carlo simulation and an in silico population of 10,000 representative adult patients. Individualized ATG dosing was examined using empirical Bayesian estimation and an optimisation function. Post-transplant exposure was calculated, and the probability of target attainment was determined for each regimen. No conventional ATG dosing regimen was able to dose > 25% of patients within the defined therapeutic target range of 60-95 AU·day/mL, with regimens utilizing total ATG doses of 4-5 mg/kg having the highest target attainment. Dose optimisation approaches were only able to improve on target attainment over baseline by up to 7%. In all cases, empirical Bayesian estimation of individual pharmacokinetic parameters was poor, with shrinkage > 30% noted on three of the five parameters. These findings suggest that improving target attainment of ATG by means of model-informed precision dosing is unlikely to be of clinical benefit in hematopoietic stem cell transplant recipients. Further research into factors affecting the pharmacokinetic parameters of ATG is required to allow for dose individualisation and treatment optimisation.