Simultaneous population pharmacokinetic modelling of darunavir and ritonavir Once daily in HIV-infected patients: evaluation of lower ritonavir dose

Abstract Purpose of study: Once-daily ritonavir-boosted darunavir (DRV/RTV) is a preferred antiretroviral regimen for treatment-naïve patients. Population pharmacokinetic modelling of the interaction between DRV and RTV allows evaluation of alternative dosing strategies, particularly lower RTV doses (e.g. 800/50mg once daily) and assessment of factors that may influence DRV/RTV PK. Methods: Data were pooled from 3 DRV/RTV PK studies. Fifty-one HIV-infected patients (7 female) stable on DRV/RTV (800/100mg or 900/100mg once daily; n=32 and 19, respectively) were included. Median age, weight and baseline CD4 cell count were 39yr (21-63), 74kg (57-105) and 500cells/mm3 (227-1129), respectively; 49 had undetectable viral load. Nonlinear mixed effects modelling (Monolix v.4.1.2) was applied simultaneously to DRV and RTV to determine PK parameters, interindividual variability and residual error. Covariates evaluated included: age, weight, sex and study. The model was validated by simulation and visual predictive check. DRV/RTV 800/50mg once daily was simulated. Summary of results: RTV and DRV were described by a 1 and 2-compartment model, respectively with first-order absorption and lag-time. A maximum effect model, in which RTV inhibited DRV clearance (CL/F), best described the relationship between the two drugs. A RTV concentration of 0.33mg/L was associated with 50% maximum inhibition of DRV CL/F with the maximum inhibitory effect fixed at 1. The population CL/F of DRV in the absence of RTV was 13.7L/h. Inclusion of weight on RTV CL/F and volume and age on DRV CL/F and study on DRV CL/F, volume and absorption improved the fit. Based on visual predictive check 93% and 91% of observed RTV and DRV concentrations were within the 95% prediction interval, indicative of an adequate model. Of 1000 simulated DRV troughs, 10% and 0% were below the MEC for treatment-experienced