Nonparametric estimation of population characteristics of the kinetics of lithium from observational and experimental data: individualization of chronic dosing regimen using a new Bayesian approach

Abstract

A population analysis of the kinetics of lithium was performed from experimental and observational data in 113 subjects in order to propose a new approach for lithium dosage individualization. The kinetics of lithium is described by a two-compartment model. Age, body weight, height, and serum creatinine are included as covariates. Population analysis was performed by the nonparametric maximum likelihood method, which provides an estimate of the distribution of the five kinetic parameters and covariates. Mean lithium clearance was 1.50 L/h with a coefficient of variation (CV) of 38%, and was found to increase with body weight. Results were consistent with those of earlier studies and confirm a large interindividual variability. Data from a separate group of 35 patients were used to validate results: the estimated a priori and on covariate conditional distributions of the measured 24-h serum lithium concentration following a single dose were consistent with the corresponding measurements. A Bayesian approach for individualizing dosing schemes is proposed. This approach is based on minimization of a risk function expressing the deviation of the trough concentration at steady-state from the therapeutic range.