Interaction of darunavir/ritonavir and darunavir/cobicistat with rifampicin in vitro

Background: Clinical management of HIV patients co-infected with tuberculosis (TB) is hampered by drug-drug interactions (DDIs) that limit therapeutic options. Rifampicin (RIF), an important component of anti-TB treatment regimens, is a strong inducer of key metabolic enzymes, and thus can negatively affect antiretroviral bioavailability, clearance and efficacy. The aim of this study was to quantify DDIs between RIF and cobicistat (COBI)-boosted darunavir (DRV), and to compare this with DDIs between RIF and ritonavir (RTV)- boosted DRV (DRV/r) using an in vitro approach. Methods: Cryopreserved primary human hepatocytes plated on collagen-coated cell culture plates were overlaid with Geltrex™ matrix and were treated with RIF (10 μM) alone, or together with RTV (0.1-10 μM) or COBI (0.13-12.76 μM) in Williams' Medium E incubation medium, or were left untreated. Test compounds were replenished each day for a total of 72 hours, after which cells were treated with test compounds together with DRV (5 μM) for one hour. Resultant DRV concentrations were quantified using HPLC-UV. Apparent intrinsic clearance (CLint.app.) of DRV was calculated, and expressed as the mean ± SD (μl/min/106 hepatocytes) of a total of three biological replicates, using cells obtained from three separate donors. Results: Under control conditions where cells treated with DRV alone, DRV CLint.app. was 13.2 ± 1.5 μl/min, while following incubation with 10 μM RIF, DRV CLint.app. increased to 20.5 ± 4.7 μl/min (+55% compared to control). Inclusion of 1 μM RTV, or 1.28 μM COBI, was sufficient to overcome the effect of 10 μM RIF, reducing DRV CLint.app. by -15% and -3% compared to control, respectively. Using regression analysis, log10 RTV and COBI concentrations were found to be associated with percentage inhibition of DRV CLint.app. (β = 20.7, p = 0.001 and β = 11.3, p = 0.001, respectively; Figure 1). Conclusions: DDIs between RIF and both DRV/r and DRV/COBI were quantified using an in vitro human hepatocyte model. RIF-induced elevations in DRV CLint.app. were overcome by co-incubation with RTV or COBI. RTV- and COBI-mediated attenuation of RIF-enhanced DRV CLint.app. occurred in a concentration-dependent manner, but RTV reversed RIF induction more strongly than COBI. These results provide an insight into the relative effect of RTV and COBI as pharmacoenhancers in the presence of RIF, and can be used to inform pharmacokinetic models for optimising regimens in patients receiving concurrent antiretroviral and anti-TB therapy. (Figure Presented).