Population pharmacokinetic models for armodafinil and its major metabolites, R-modafinil acid and modafinil sulfone, were developed, and selected covariates were investigated. Data from 583 healthy subjects and patients with bipolar I disorder in 11 phase 1-3 studies (8027 concentrations) of armodafinil, given as single or multiple once-daily doses (50- to 400-mg tablet or capsule), were pooled. A previously developed 1-compartment model with first-order absorption without covariate effects was initially applied to pooled phase 1 and 2 data. After covariate analysis, the phase 3 data were pooled with the phase 1 and 2 data set and the model was refined again using a second backward elimination step. Population modeling was performed with NONMEM version 7 with the first-order conditional estimation method. Estimated armodafinil apparent oral clearance (CL/F), volume of distribution (Vc/F), and absorption t½ were 2.01 L/h, 45 L, and 0.226 hours, respectively. Armodafinil CL/F and Vc/F increased with weight; predicted steady-state area under the curve was 16.4% higher and 29.1% lower in a patient weighing 50 or 150 kg, respectively, relative to a 70-kg patient. Female participants had 10.2% lower armodafinil Vc/F compared with male participants. Age, race (white vs nonwhite), health status (healthy vs bipolar I disorder), liver function, and renal function were not statistically significant predictors of armodafinil pharmacokinetics. CL/F and Vc/F for R-modafinil acid and modafinil sulfone were 16.7 L/h and 8.95 L and 6.82 L/h and 12.4 L, respectively. Weight did not affect exposure of either metabolite. These population pharmacokinetic models were from the largest population of adults reported to date and provide a robust characterization of the pharmacokinetics of armodafinil, R-modafinil acid, and modafinil sulfone in adults.

By Jill Fiedler-Kelly, Elizabeth A Ludwig