Tacrolimus population pharmacokinetic model-informed precision dosing in adult liver transplant patients

We aimed to explore tacrolimus population pharmacokinetic (PPK) characteristics in adult liver transplantation patients and develop individualized dosing software for precision dosing. Data were retrospectively extracted from adult liver transplantation patients receiving tacrolimus at Chinese PLA General Hospital and Beijing Friendship Hospital. The PPK model was established using Phoenix, with the final model developed through forward inclusion-backward elimination. Bootstrap and visual predictive check (VPC) were used to validate the final model. External validation was conducted, and the mean error (ME), mean absolute error (MAE), and root mean square error (RMSE) were calculated. The software, developed in C# language, predicted drug concentrations three times for each of the 10 patients, calculating the predictive error (PE) and the absolute predictive error (APE). The application of this software for dosing regimen recommendations was also elucidated. Fifty-seven patients were included, with 633 blood drug concentrations collected. Data from 41 patients (Chinese PLA General Hospital) were used for modeling, and a one-compartment model with first-order absorption was built. Postoperative days and γ-glutamyl transferase affected clearance. The final model parameters were within Bootstrap's prediction range, and VPC prediction results aligned with the observations. Data from 16 patients (Beijing Friendship Hospital) were used for external validation, ME, MAE, and RMSE were - 0.32, 2.07, and 2.76 ng/mL, respectively, indicating robust predictive capability. PE and APE decreased with an increase in the number of blood drug concentrations. The developed software accurately predicts drug concentrations and the accuracy of these predictions increases with the number of drug concentrations used. A robust PPK model was established for liver transplant adults. The individualized dosing software not only predicts drug concentrations with increasing precision but also facilitates the practical application of model-informed precision dosing, offering customized dosing regimens that are poised to optimize therapeutic outcomes in liver transplant patients.