Sirolimus population pharmacokinetic/pharmacogenetic analysis and Bayesian modelling in kidney transplant recipients

Objectives: The objectives of the present study were: (i) to analyse the population pharmacokinetics of sirolimus in renal transplant recipients co-administered mycophenolate mofetil, but no calcineurin inhibitor over the first 3 months post-transplantation and study the influence of different potential covariates, including genetic polymorphisms of cytochrome P450 (CYP) metabolic enzymes and active transporters, on pharmacokinetic parameters; and (ii) to develop a Bayesian estimator able to reliably estimate the individual pharmacokinetic parameters and exposure indices in this population. Methods: Twenty-two adult renal transplant patients treated with sirolimus participated in this study. Ninety concentration-time profiles (938 sirolimus whole blood samples) were collected at days 7 and 14, and months I and 3 post-transplantation. The population pharmacokinetic study was conducted using the nonlinear mixed effects model software, NONMEM, and validated using both the bootstrap and the cross-validation approaches. Finally, a Bayesian estimator based on a limited sampling strategy was built using the post hoc option. Results: A two-compartment open model with first-order elimination and Erlang's distribution (to describe the absorption phase) best fitted the data. The mean pharmacokinetic parameter estimates were 5.25 h(-1), 218L and 292L for the transfer rate constant, the apparent volume of the central and peripheral compartments, respectively. The CYP3A5*1/*3 polymorphism significantly influenced the apparent oral clearance: mean oral clearance = 14.1 L/h for CYP3A5 non expressers (CYP3A5*3/*3 genotype) versus 28.3 L/h for CYP3A5 expressers (CYP3A5*/1*3 and *1/*1 genotypes). The standard errors of all the parameter estimates were < 15%. Maximum a posteriori Bayesian forecasting allowed accurate prediction of sirolimus area under the concentration-time curve from 0 to 24 hours using a combination of only three sampling times (0, 1 and 3 hours post-dose), with a non-significant bias of -2.1% (range -22.2% to +25.9%), and a good precision (root mean square error =10.3%). This combination is also easy to implement in clinical practice. Conclusion: This study presents an accurate population pharmacokinetic model showing the significant influence of the CYP3A5*1/*3 polymorphism on sirolimus apparent oral clearance, and a Bayesian estimator accurately predicting sirolimus pharmacokinetics in patients co-administered mycophenolate mofetil, but no calcineurin inhibitor.