Population pharmacokinetics and Bayesian estimation of mycophenolic acid concentrations in stable renal transplant patients.

backgroundTherapeutic drug monitoring of mycophenolic acid (MPA) may minimise the risk of acute rejection after transplantation. Area under the curve (AUC) rather than trough concentration-based monitoring is recommended and models for AUC estimation are needed.

objectivesTo develop a population pharmacokinetic model suitable for Bayesian estimation of individual AUC in stable renal transplant patients. PATIENTS AND

methodsThe population pharmacokinetics of MPA were studied using nonlinear mixed effects modelling (NONMEM) in 60 patients (index group) receiving MPA on a twice-daily basis. Ten blood samples were collected at fixed timepoints from ten patients and four blood samples were collected at sparse timepoints from 50 patients. Bayesian estimation of individual AUC was made on the basis of three blood concentration measurements and covariates. The predictive performances of the Bayesian procedure were evaluated in an independent group of patients (test group) comprising ten subjects in whom ten blood samples were collected at fixed timepoints.

resultsA two-compartment model with zero-order absorption best fitted the data. Covariate analysis showed that bodyweight was positively correlated with oral clearance. However, the weak magnitude of the reduction in variability (from 34.8 to 28.2%) indicates that administration on a per kilogram basis would be of limited value in decreasing interindividual variability in MPA exposure. Bayesian estimation of pharmacokinetic parameters using samples drawn at 20 minutes and 1 and 3 hours enabled estimation of individual AUC with satisfactory accuracy (bias 7.7%, range of prediction errors 0.43-15.1%) and precision (root mean squared error 12.4%) as compared with the reference value obtained using the trapezoidal method.

conclusionThis paper reports for the first time population pharmacokinetic data for MPA in stable renal transplant patients, and shows that Bayesian estimation can allow accurate prediction of AUC with only three samples. This method provides a tool for therapeutic drug monitoring of MPA or for concentration-effect studies. Its application to MPA monitoring in the early period post-transplantation needs to be evaluated.