Pharmacokinetics and pharmacodynamics of multiple-dose terbinafine

Data from clinical trials of terbinafine for the treatment of onychomycosis were analyzed with the following two objectives: 1) to identify demographic predictors of the duration and extent of systemic drug exposure; and 2) to explore whether increased systemic exposure or demographic predictors of increased exposure were associated with altered safety or efficacy. Demographic predictors of exposure were identified by a model-free, nonparametric approach applied to the sparse pharmacokinetic data from the onychomycosis studies. Those covariates were then incorporated into a multicompartmental nonlinear mixed effects model. Post hoc parameter estimates from the nonlinear mixed effects model provided individual measures of exposure. Safety scores were derived for adverse events that were frequently attributed to drug exposure and for liver function tests. Terbinafine was found to have an average terminal half-life (t(1/2)) Of approximately 3 weeks. That terminal elimination phase contributed so little to the total exposure, however, that average concentrations accumulated only approximately two-fold at steady state with once daily dosing. Age and concomitant hypertension were predictors of higher plasma concentrations of terbinafine; smokers had lower levels than nonsmokers. Although some statistically significant associations between adverse events and systemic exposure were found, in all cases the actual frequency of the adverse events was low, and there were no trends in severity with respect to exposure. Above-normal levels of gammaglutamyl transferase were associated with exposure, but there was no trend in severity with respect to exposure. No other liver function test abnormalities were associated with exposure, nor were there any significant associations between adverse events or liver-function abnormalities and demographic subgroups that differed with respect to exposure. Among patients taking the active drug there were no significant associations between exposure levels and efficacy, nor were there differences in efficacy between demographic subgroups that differed with respect to exposure.