Purpose: Mycophenolic acid (MPA) area under the curve (AUC) has been reported as an important marker of graft outcome. Pharmacokinetic (PK) tools for the MPA AUC monitoring have been proposed, but mainly in renal transplantation. The aim of our study was to develop multiple pharmacokinetic tools that could improve AUC estimation in heart transplant patients.

Methods: This study was performed using 56 MPA PK profiles collected from 39 adult heart transplant patients given mycophenolate mofetil associated to a calcineurine inhibitor, in the first year post-transplantation. Using 37 PK profiles (i.e. building dataset), 3 different Bayesian estimators (BEs) were developed on the basis of 3 independent, parametric (nonlinear mixed-effect modeling in NONMEM® and iterative two-stage Bayesian) or nonparametric (nonparametric adaptive grid with PMETRICS) population modeling approaches. The predictive performance of the BEs was assessed in the remaining 19 PK profiles (i.e., validation dataset) by comparing: (i) The AUC estimated by BE using a limited number of samples collected at 0.33, 1, 3 and 6 hours post-dose and the reference AUC (calculated using all concentrations) and: (ii) dose recommendations to achieve a 45 mg.h/L target (30-60 mg.h/L therapeutic window) using either AUCs. Results: Despite good predictive performance (mean relative bias = 0.04±0.24, -0.01±0.19 and -0.01±0.19 in NONMEM®, PMETRICS and ITS, respectively), none of the BEs were able to estimate accurately all the AUCs, leading to misevaluations of exposure and inadequate dose adjustments for some patients. However, when combining the information provided by the 3 BEs, the dose proposed was within ± 250mg of those obtained using the reference AUCs for all the patients. Moreover, on the basis of this combined information, dose adjustment following this recommendation would have led to an AUC in the therapeutic target for 17/19 patients, while 2 AUCs would have probably slightly exceeded it. Conclusion: Bayesian estimators allowing the determination of MPA AUC using a limited number of blood samples have been developed using 3 totally independent population modelling approaches, the combination of which now increases the accuracy of our results.

Sinnasse-Raymond, G.: Employee, Salary. Marquet, P.: Grant/Research Support, Research Support.

« Back to 2013 American Transplant Congress