Development of a population pharmacokinetic database for tianeptine

Two thousand three hundred and thirty five plasma concentrations of tianeptine from 112 patients enrolled in nine studies of tianeptine pharmacokinetics performed prior to the marketing of the drug were pooled for analysis using mixed-effect modeling. Studies represented a combination of single dose and multiple dosing at steady-state. Tianeptine plasma concentration time data were fit to a two compartment model with first order absorption using the NONMEM computer program. The results of this analysis suggested that alcoholism is associated with significant increase in clearance (124% increase) and volume of the central compartment (161% increase). The volume of the peripheral compartment is significantly lower in women (31% decrease) and in depressed patients (59% decrease). The population mean (interindividual variability) clearance was equal to 0.17 l·h^-1·kg^-1 (28.6%), the volume of central compartment was 0.13 l·kg^-1 (60.4%), intercompartmental clearance was 0.07 l·h^-1·kg^-1 (30.1%), volume of the tissue compartment was 1.17 l·kg^-1 (28.3%), and the absorption rate constant was 0.63 h^-1 (21.8%). The residual variability was approximately 30% at concentrations expected during clinical use of the drug. Because of the increased clearance, alcoholic patients would be expected to have significantly reduced concentrations during steady-state dosing. These population parameters provide a basis for developing initial dosing recommendations and for performing bayesian evaluations of drug concentrations obtained in post-marketing studies.