The effect of nevirapine-based antiretroviral therapy on the population pharmacokinetics of artemether and dihydroartemesinin in HIV-infected adults
J-S van der Walt [1,2] , T Kredo [1,3], L Wiesner , G Maartens , K Cohen , P Smith , MO Karlsson , KI Barnes 
 Division of Clinical Pharmacology, Department of Medicine, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa; 2] Pharmacometrics Research Group, Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden;  South African Cochrane Centre, South African Medical Research Council, Cape Town, South Africa
Objectives: Artemether-lumefantrine (AL) and nevirapine (NVP) are currently the most widely recommended antimalarials and antiretrovirals in Africa. Artemether (AM) is rapidly demethylated to active dihydroartemesinin (DHA) via CYP3A4/3A5, and may exhibit autoinduction. Co-administration of NVP, an inducer of CYP3A4, may alter the pharmacokinetics of AM and DHA. We investigated the interaction between NVP-based antiretroviral therapy (ART) and AM/DHA in HIV-1 infected adults.
Methods: AM and DHA plasma concentration-time profiles from 2 groups of HIV-infected adults (ART-na´ve [N=18] and NVP-based ART [N=18]) with CD4+ lymphocyte counts>200 cells/ÁL who received AL 80mg/480mg twice daily for 3 days were analyzed using NONMEM7 (FOCE INTER). A published semiphysiological autoinduction model  was adapted by (1) adding a transit compartment model to describe the highly variable absorption of AM and (2) incorporating the conversion of AM to DHA.
Results: NVP-based ART significantly reduced AM and DHA exposure. The autoinduction enzyme kinetics was described by a precursor and an enzyme pool. NVP-based therapy increased both the precursor production rate constant and the slope describing the linear effect of AM liver concentrations on the rate of enzyme precursor production by 66% and 87% respectively. AM absorption was rapid and variable with a typical mean transit time of 0.9 h (IIV 44%, IOV 25%) and 2.6 transit compartments (IIV 133%). In the NVP group the mean transit time was decreased by 38%.This could, in part, be explained by a larger proportion of AM observations below the limit of quantification in the NVP vs the na´ve group (36% vs. 13%). Goodness-of-fit plots confirmed adequate model fit for both AM and DHA. Residual variability in AM observations was high (51%CV) but comparable with previous reports. 
Conclusions: Autoinduction was faster in subjects treated with NVP-based ART compared with ART-na´ve subjects. The decreased bioavailability of AM and DHA may be of concern when using AL for treating malaria in patients co-infected with HIV treated with NVP-based ART.
 Guidelines for the treatment of malaria in South Africa 2008. Department of Health Republic of South Africa. http://www.malaria.org.za/Malaria_Risk/Treatment/treatment_malaria08.pdf (Accessed on 8 March 2011)
 LefŔvre G, Thomsen MS. Clinical pharmacokinetics of artemether and lumefantrine (Riamet«). Clin Drug Invest 1999; 18: 467ľ480
 Faucette SR, Zhang TC, Moore R et al. Relative activation of human pregnane X receptor versus constitutive androstane receptor defines distinct classes of CYP2B6 and CYP3A4 inducers. J Pharmacol Exp Ther 2007 Jan; 320:72-80. Epub 2006 Oct 13
 Kredo mT, Van derWalt J-S, Mauff K, Wiesner L, Maartens G, Cohen K, Smith P, Barnes KI. Nevirapine Increases Lumefantrine Exposure in HIV-infected subjects. 17th Conference on Retroviruses and Opportunistic Infections, San Francisco, 16-19thFebruary 2010 -ABSTRACT N-140
 Gordi T, Xie R, Huong NV, Huong DX, Karlsson MO, Ashton M. A semiphysiological pharmacokinetic model for artemisinin in healthy subjects incorporating autoinduction of metabolism and saturable first-pass hepatic extraction. Br J Clin Pharmacol. 2005 Feb;59(2):189-98