Population pharmacokinetic/pharmacodynamic model for the sedative effects of flibanserin in healthy volunteers.

Flibanserin is being developed for treating hypoactive sexual desire disorder in women; the main side effect is sedation. The analysis objective was to relate flibanserin plasma concentrations with acute sedative effects using a population pharmacokinetic/pharmacodynamic (PK/PD) model. The population model was developed with NONMEM based on data from 24 healthy volunteers. "Drowsiness" was serially assessed by a Visual Analogue Scale (VAS) on a baseline day and after morning oral administration of 100 mg flibanserin together with PK sampling. PK was best described by a three-compartment disposition model and transit compartments accounting for the lag time in absorption. VAS "drowsiness" baseline profiles were modeled using linear splines with three breakpoints located at clock times at first and last observation, and at the median of the observation time across subjects. The drug effect followed a sigmoidal E(MAX) model using predicted effect site concentrations (C(e)). The VAS vs. C(e) relationship was very steep and effect site and plasma concentration-time profiles were very similar thus suggesting little delay between the occurrence of maximum flibanserin plasma concentrations and drowsiness. At effect site concentrations lower than ≈ 200 ng/mL that are reached approximately 4 h after administration, flibanserin shows hardly any effect on the VAS "drowsiness" scale.