Population pharmacokinetics of mycophenolic acid in pediatric renal transplant patients using parametric and nonparametric approaches

Abstract

Mycophenolic acid (MPA) is an immunosuppressive drug widely used in the prevention of acute rejection in pediatric renal transplant recipients and is characterized by a wide inter-individual variability in its pharmacokinetics. The aim of this study was to compare population pharmacokinetic modeling of MPA in pediatric renal transplant recipients given mycophenolate mofetil, the ester prodrug of MPA, using parametric and nonparametric population methods. The data from 34 pediatric renal transplants (73 full pharmacokinetic profiles obtained on day 21, months 3, 6 and 9 post-transplant) were analyzed using both the nonlinear mixed-effect modeling (NONMEM) and nonparametric adaptive grid (NPAG) approaches, based on a two-compartment model with first order lagged time absorption and first order elimination. The predictive performance of the two models was evaluated in a separate group of 32 patients. Higher mean population parameter values and ranges of individual pharmacokinetic parameters were obtained with NPAG, especially for the elimination constant ke: mean 1.16 h(-1) (0.26-4.33 h(-1)) and 0.78 h(-1) (0.66-1.15 h(-1)) with NPAG and NONMEM, respectively. With NPAG, the skewness and kurtosis values for ke (2.03 and 7.80, respectively) were far from the theoretical values expected for normal distributions. Such a non-normal distribution could explain the high value of shrinkage (35%) obtained for this parameter with the parametric NONMEM method. Bayesian forecasting of mycophenolic acid exposure using the NPAG population pharmacokinetic parameters as priors yielded a better predictive performance, with a significantly smaller bias than with the NONMEM model (-1.68% vs -9.53%, p<0.0001). In conclusion, in the present study, NPAG was found to be the most adequate population pharmacokinetic method to describe the pharmacokinetics of MPA in pediatric renal transplant recipients.

Pharmacokinetics of Vancomycin in Elderly Patients Aged over 80 Years

“…This algorithm calculates the optimal discrete parameter joint density by successive iterations, with each iteration increasing the likelihood. This algorithm has already shown its usefulness in several pharmacokinetic studies (13)(14)(15). Individual pharmacokinetic parameters were determined by Bayesian estimation for each patient.…”

Therapeutic monitoring of immunosuppressive drugs in pediatric patients: special considerations

“…This is based on population pharmacokinetic data and takes into account the pharmacokinetic characteristics of a typical population, data collected from individual patients, and also considers the variability of pharmacokinetic parameters in the population studied. The prediction is therefore more precise and offers higher flexibility in blood sample collection [12].…”

“…(11) Is the duration of drug therapy sufficient for the patient to benefit from therapeutic drug-monitoring? (12) Will the results of the monitoring assay make a significant difference in the clinical decision-making process (i.e. provide more information than sound clinical judgment alone)?…”

Some Comments and Suggestions Concerning Population Pharmacokinetic Modeling, Especially of Digoxin, and Its Relation to Clinical Therapy

“…In contrast to parametric compartmental models and the above noncompartmental models, rigorous nonparametric (NP) population modeling methods estimate the entire model parameter distributions 2,7-9 . There is no equation required any more to describe any assumed shape of the model distributions.…”

“…The parameter distributions provided by NP modeling methods 2,7-9 obtain a discrete distribution, based on the theorems of Caratheodory, Lindsay, and Mallet 2,7,13 . The NP model parameter distributions thus consist of multiple discrete points (support points), up to 1 per patient studied in the population.…”