External evaluation of a population pharmacokinetic model of tacrolimus after hematopoietic stem cell transplantation in children with thalassemia

Abstract

Abstract Objective Tacrolimus (TAC), a narrow therapeutic window drug, several population pharmacokinetic (PopPK) models of TAC have been established to explore its optimized treatment regimen after hematopoietic stem cell transplantation (HSCT). However, there is no evidence of the predictive performance of these models when extrapolated to different clinical centers. We aimed to perform a systematic external evaluation of the published TAC PopPK models and to identify underlying influence elements. Methods Published population pharmacokinetic models of oral TAC after HSCT were searched for external assessment using TAC blood trough concentrations of pediatrics with β-thalassemia major (β-TM) after HSCT at the First Affiliated Hospital of Guangxi Medical University. The effect of prior information on model predictability was determined by Prediction error (PE%)、Visual predictive check (VPC)、Normalized prediction distribution error (NPDE) test and Bayesian prediction. Results Two published population pharmacokinetic models were evaluated by including 296 oral TAC blood concentrations from 46 children. Evaluation results on the basis of prediction errors, simulations, and Bayesian founded that population of published model may affect the transferability of the model. TAC bioassay methods in different centers may also result in an undesirable predictive performance of model extrapolation. Conclusion Individualized dosing with frequent therapeutic drug monitoring during the use of TAC was confirmed to be of clinical value. The two published models of oral TAC after HSCT have performed unsatisfactorily in all aspects of predictive diagnostics and are therefore not suitable for direct extrapolation to use in children with β-TM in our center.