Simulation of the impact of rifampicin on darunavir/ritonavir PK and dose adjustment strategies in HIV-infected patients: a population PK approach

INTRODUCTION Treatment of HIV/TB co-infection is challenging due to high drug-drug interaction potential between antiretrovirals and rifamycins, such as rifampicin (RIF). The PK interaction between darunavir/ritonavir (DRV/RTV) and RIF has not been studied. Utilizing other protease inhibitor data, population PK modelling and simulation was applied to assess the impact of RIF on DRV/RTV PK and generate alternative dosing strategies to aid future clinical trial design. MATERIALS AND METHODS A previously developed model describing DRV/RTV PK including data from three studies in HIV patients was used [n=51, 7 female, DRV/RTV 800/100 mg (n=32) or 900/100 mg once daily (qd; n=19) (1). The PK interaction between DRV/RTV and RIF was assumed to mimic that observed in HIV-infected, TB negative patients receiving lopinavir (LPV)/RTV (n=21) (2). Simulations of DRV/RTV 800/100 mg qd (n=1000) were performed (-RIF). The model was adapted to increase the typical value of apparent oral clearance (CL/F) by 71% and 36% and decrease relative bioavailability (F) by 20% and 45% for DRV and RTV, respectively (2); 1000 simulations were generated (+RIF). Dose adjustments of DRV/RTV 1200/200 mg qd, 800/100 mg and 1200/150 mg twice daily (bid) were simulated to overcome the interaction. DRV trough (Ctrough) for each dosing scenario was compared to the reference (-RIF) by GMR (90% CI). RESULTS DRV and RTV were described by a 1 and 2-compartment model, respectively. A maximum effect model, with RTV inhibiting DRV CL/F, best described the relationship between the drugs. Compared to the reference (-RIF), simulated DRV Ctrough was 70%, 46% and 20% lower for 800/100 mg qd, 1200/200 mg qd and 800/100 mg bid all +RIF, respectively. Ctrough was 38% higher with 1200/150 mg bid +RIF (Table 1). CONCLUSIONS Modelling and simulation was used to investigate the theoretical impact of RIF on DRV/RTV PK. Based on simulations, 800/100 mg and 1200/150 mg both bid could largely overcome the impact of the interaction. However, the risk of increased RTV-related side effects and higher pill burden should be considered. In vitro work is ongoing to develop a physiologically based model characterizing the interaction and informing simulations.