To our knowledge, this study represents the first investigation of the disposition kinetics of artesunate and DHA in HIV-infected adults with and without NVP containing ART. Overall, despite a shorter T1/2
for both artesunate and DHA, we found an increase in overall exposure (AUC0–96
) of artesunate in patients receiving NVP compared to those not on ART (105 versus 69
L/hr; respectively; P
= 0.02) and no difference in the overall exposure to DHA. While the clinical relevance of these results remains unclear, it is noteworthy that the half-life of DHA was significantly shorter when given with NVP, and the conversion of artesunate to DHA was lower in the NVP group. It is possible that a negative impact of NVP on the disposition kinetics of artesunate and DHA may be detected in larger studies. This demands an observant approach to malaria therapy in individuals on NVP containing ART until further investigation into the impact of this interaction can be performed.
Given the metabolic pathways of artesunate (hydrolysis and CYP2A6) and DHA (UGT 1A9 and 2B7), the observed impact on artesunate and DHA pharmacokinetics is unexpected. Nevirapine is well known for decreasing exposure to coadministered medications due to induction of the CYP3A4 and 2B6 isoenzymes [13
]. Interestingly, one other ACT-nevirapine interaction study described an in vivo
pharmacokinetic interaction where NVP both increased and decreased exposure to the coadministered ACT [17
]. Kredo and colleagues described the interaction between NVP and artemether-lumefantrine in HIV-infected subjects in South Africa in which lumefantrine Day 7 concentrations and AUC0–inf
were increased in patients on NVP compared to HIV-infected controls [17
]. These directional changes seen with the lumefantrine parameters when combined with NVP are similar to our artesunate results, despite different metabolic pathways of the two antimalarial agents. Contrary to our artemisinin pharmacokinetic results, Kredo and colleagues found that the artemether and DHA AUC0–inf
were lower in the NVP group compared to controls [17
]. Notably, different CYP enzyme pathways metabolize artesunate (CYP2A6) and artemether (CYP3A4), which may account for the difference in artemisinin pharmacokinetic findings observed in our study of artesunate compared to the results of artemether plus NVP. Although the current study was not designed to evaluate the mechanism of this interaction, our observation of a lower conversion of artesunate to DHA in the NVP group (DHA: artesunate AUC0–96
= 5.6 versus 8.5 in NVP and control groups, respectively, P
= 0.008) is noteworthy. Further investigation into the underlying mechanism of this unexpected change is warranted.
The rate of malaria parasite clearance has been associated with the overall exposure to both parent drug and DHA for other artemisinins [18
]; hence reduction in the blood concentrations of either or both components may negatively impact on the antimalarial activity of the artemisinin therapy. Reassuringly, our findings suggest that although the T1/2
was shorter, the overall exposure of both artesunate and DHA was similar compared to our control group and indeed higher for artesunate. Artesunate is generally a well-tolerated medication, particularly in comparison to other nonartemisinin antimalarial medications [19
]. Dizziness, nausea, vomiting, and anorexia have been reported in patients with malaria who were treated with artemisinin monotherapy [19
]. However, these toxicities were typically transient and resolved after 1-2 days, raising some question as to the relationship of the toxicity to the medication versus the underlying infectious process. Given the relative safety of artesunate, the observed increase in drug exposure would not be expected to cause additional toxicity; however vigilance for excess toxicity may be warranted.
Artesunate and DHA are known to have wide interpatient variability in their pharmacokinetic parameters, and artesunate and DHA exposure are both decreased by the co-administration of amodiaquine [20
]. Additionally, the pharmacology of these agents is known to be different between patients with acute malaria and healthy volunteers. DHA total exposure was shown to be approximately 2-fold higher in patients with active malaria than healthy volunteers (4,024 versus 1,763
]. Additionally, the protein binding of DHA may change during acute malaria infection related to plasma pH and circulating α
-1-acid glycoprotein [22
]. Complicating the evaluation of these important drug interactions further, differences in antiretroviral pharmacokinetics and pharmacodynamics exist between healthy volunteers and HIV infected patients [22
]; therefore, it is conceivable that HIV-infection may impact antimalarial drug concentrations as well.
There are some limitations in the present study that must be considered. In addition to noncompartmental analysis of artesunate and DHA, a comodelling approach that combines the parent and metabolite is currently underway to more fully describe the pharmacokinetic implications of chronic NVP therapy on artesunate and DHA. The pharmacokinetics of concurrent amodiaquine will be also described to fully understand the impact of NVP on antimalarial treatment with artesunate-amodiaquine. Although we have accommodated for the potential impact of HIV infection on the pharmacokinetics of artesunate and DHA by evaluating this interaction in an HIV-infected population, the pharmacokinetic impact of this interaction may be different in patients with acute malarial infection.