Aims: Artemisinin-resistant falciparum malaria is now prevalent in Southeast Asia. Understanding the pharmacokinetic and pharmacodynamic properties of artesunate and its active metabolite, dihydroartemisinin, in patients with sensitive and resistant malaria is essential to achieve effective and safe treatment. The aim of this work was to characterize the pharmacokinetic properties of artesunate and dihydroartemisinin in patients with sensitive and resistant malaria infections in Myanmar, Cambodia and Thailand.
Methods: A total of 4,217 plasma concentrations of artesunate and dihydroartemisinin from 153 patients from the clinical efficacy study (1) were modeled simultaneously using nonlinear mixed-effects modelling. The model building was conducted using the first-order conditional estimation method with interactions. The first observed serial concentration below the limit of quantification (LLOQ) for each patient was substituted by half of the LLOQ (LLOQ/2), the rest were omitted. Full in vivo conversion of artesunate into dihydroartemisinin was assumed and different structural disposition models were evaluated. Several absorption models were investigated i.e. zero-order, first-order, first-order with lag time, and transit absorption models. A stepwise covariate search was used to evaluate the relationship between pharmacokinetic parameters and patient characteristics. Model diagnostics and evaluation were performed using Xpose version 4.0, Pirana, and Pearl-speaks-NONMEM (PsN; version 4.4.8).
Results: A one-compartment disposition model for both artesunate and dihydroartemisinin provided the best fit to the data. An additional disposition compartment, for both artesunate and dihydroartemisinin, resulted in a decreased objective function value (∆OFV = -205). However, the visual predictive check showed substantial model misspecification in the terminal phase for both compounds, as compared to a simpler one-compartment structural model. Furthermore, the median elimination half-life of dihydroartemisinin was unreasonable long when described using a two-compartment disposition model (16 hours compared to literature values of 30-60 minutes). The one-compartment disposition model was therefore carried forward for both compounds. The absorption of artesunate was best described by a transit absorption model with two transit compartments. Introducing inter-occasion variability on absorption parameters (i.e relative bioavailability and mean transit time) improved the model fit substantially. Body weight was implemented as an allometric function on all clearance (power fixed to 0.75) and volume (power fixed to 1) parameters and resulted in improvement of model fit (∆OFV = -6.65). Alanine aminotransferase and split dosing regimen (i.e twice daily vs once daily dosing) affected artesunate clearance and dihydroartemisinin clearance significantly during the forward covariate addition (p-value < 0.05) but failed to be retained in the more stringent backward elimination step (p-value < 0.001).
Conclusion: The pharmacokinetic properties of artesunate and dihydroartemisinin were adequately described by a flexible transit-compartment absorption of artesunate followed by one-compartment disposition models with full in vivo conversion of artesunate into dihydroartemisinin. The pharmacokinetic properties of artesunate and dihydroartemisinin were not different in patients with sensitive and resistant malaria infection.
1. Das D, Tripura R, Phyo AP, Lwin KM, Tarning J, Lee SJ, et al. Effect of high-dose or split-dose artesunate on parasite clearance in artemisinin-resistant falciparum malaria. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America. 2013;56(5):e48-58.