Population-based approach to analyze sparse sampling data in biopharmaceutic and pharmacokinetic studies using nonmem and Monolix

Abstract

Although it has been developed since 1972, the application of a population-based modeling approach in Indonesia, particularly to analyze biopharmaceutic and pharmacokinetic data remains very limited. This study was aimed to evaluate the performance of Monolix and NONMEM, two of the popular software packages in a population-based modeling approach, to analyze the limited data (sparse data) of the time profiles of the simulated plasma drug concentration of a theoretical compound. Monolix and NONMEM were used to model the limited data (40 data points) resulting from random selection of 180 data points of plasma drug concentrations (Cp) in 20 subjects at 0.25; 0.5; 0.75; 1; 1.5; 3; 6; 12 and 18 hours after per-oral administration of a 100mg of a theoretical compound. Values of the absorption rate constant (Ka), the elimination rate constant (Kel) and the distribution volume (Vd) of sparse data estimated using Monolix and NONMEM, were compared to the respective values of rich data obtained by a conventional two-stage approach using PKSolver. The calculation system of a nonlinear mixed effect model in Monolix and NONMEM, successfully describes the sparse data, based on the visual evaluation of the goodness of fit. Comparison of parameter estimates of population values in Monolix and NONMEM are in the range of 94 to 108% of the real values of the rich data analyzed by PKSolver. A population-based modeling can adequately describe limited or sparse data, demonstrating its capability as an important tool in clinical studies, involving patients.