BACKGROUND

Immunosuppressive therapy remains the primary method for preventing rejection in renal transplant recipients. While multiple population pharmacokinetic (popPK) models of mycophenolate sodium (MPS) have been developed for this population, their predictive performance across different clinical settings remains unverified. This study systematically evaluated published MPS popPK models through external validation to assess their extrapolation potential.

METHODS

Published MPS popPK models for renal transplant recipients were identified through systematic searches of PubMed, Embase and Web of Science. These models were externally evaluated using a cohort of renal transplant patients receiving MPS therapy at the Affiliated Hospital of Qingdao University. Model prediction performance was evaluated using three metrics: the goodness-of-fit method based on model prediction, prediction error test method and visual predictive checks method based on model simulation.

RESULTS

A total of 186 drug concentration data of 31 patients in our hospital were collected, and 4 literature were retrieved, among which 1 were one-compartment models and 3 were two-compartment models. In the goodness-of-fit diagnosis and prediction error test based on model prediction, the population prediction data of all models were not good, while the individual prediction data showed that the fitting result of Model 1 was relatively better. The visual prediction test results based on model simulation show that the fitting result of Model 1 was relatively good, while the distribution deviation between the observed data and the simulation data of the remaining models was large, and the fitting effect was not good.

CONCLUSION

The published models exhibit significant variability and unsatisfactory predictive performance, indicating that therapeutic drug monitoring (TDM) remains an essential requirement for the clinical application of MPS. To advance individualized medication for MPS based on popPK, future research must prioritize the investigation of potential covariates. This will enable identification of key factors influencing MPS model predictability and facilitate the development of a popPK model suitable for patients in our hospital.