Population pharmacokinetics of artesunate and dihydroartemisinin following intra-rectal dosing of artesunate in malaria patients

Abstract

Background: Intra-rectal artesunate has been developed as a potentially life-saving treatment of severe malaria in rural village settings where administration of parenteral antimalarial drugs is not possible. We studied the population pharmacokinetics of intra-rectal artesunate and the relationship with parasitological responses in patients with moderately severe falciparum malaria.

Methods and findings: Adults and children in Africa and Southeast Asia with moderately severe malaria were recruited in two Phase II studies (12 adults from Southeast Asia and 11 children from Africa) with intensive sampling protocols, and three Phase III studies (44 children from Southeast Asia, and 86 children and 26 adults from Africa) with sparse sampling. All patients received 10 mg/kg artesunate as a single intra-rectal dose of suppositories. Venous blood samples were taken during a period of 24 h following dosing. Plasma artesunate and dihydroartemisinin (DHA, the main biologically active metabolite) concentrations were measured by high-performance liquid chromatography with electrochemical detection. The pharmacokinetic properties of DHA were determined using nonlinear mixed-effects modelling. Artesunate is rapidly hydrolysed in vivo to DHA, and this contributes the majority of antimalarial activity. For DHA, a one-compartment model assuming complete conversion from artesunate and first-order appearance and elimination kinetics gave the best fit to the data. The mean population estimate of apparent clearance (CL/F) was 2.64 (l/kg/h) with 66% inter-individual variability. The apparent volume of distribution (V/F) was 2.75 (l/kg) with 96% inter-individual variability. The estimated DHA population mean elimination half-life was 43 min. Gender was associated with increased mean CL/F by 1.14 (95% CI: 0.36-1.92) (l/kg/h) for a male compared with a female, and weight was positively associated with V/F. Larger V/Fs were observed for the patients requiring early rescue treatment compared with the remainder, independent of any confounders. No associations between the parasitological responses and the posterior individual estimates of V/F, CL/F, and AUC0-6h were observed.

Conclusions: The pharmacokinetic properties of DHA were affected only by gender and body weight. Patients with the lowest area under the DHA concentration curve did not have slower parasite clearance, suggesting that rectal artesunate is well absorbed in most patients with moderately severe malaria. However, a number of modelling assumptions were required due to the large intra- and inter-individual variability of the DHA concentrations.

Figure 1. Observed Artesunate Concentrations (Semilogarithmic Scale; ng/ml) versus Time post Dose (h)

Study site: filled circle, Bangkok; open square, Ghana; filled triangle, Mae-Sot; open circle, Malawi; asterisk, South Africa.

Figure 2. Observed and Population Predicted Concentrations (Solid Black Line) of DHA (Semilogarithmic Scale; ng/ml) versus Time post Dose (h)

Study site: filled circle, Bangkok; open square, Ghana; filled triangle, Mae-Sot; open circle, Malawi; asterisk, South Africa. The population predicted concentrations are calculated for a dose of 10 mg/kg.

Figure 3. Residual (log_e Units) versus Population Predicted Concentrations of DHA (log_e Units)

Residual = observed − population predicted concentration of DHA.

Figure 4. Relationship of Estimated Probability of Requiring Rescue Treatment and V/F of DHA (l/kg)

Derived from a multiple logistic regression model with adjustment for the confounders: baseline parasitaemia, baseline PCV, age, and gender.

Figure 5. Observed PCT₅₀ (A) and PCT₉₀ (B) versus Posterior Estimate of DHA AUC_0–6h (ng/ml × h)

For 162 patients (two patients excluded due to baseline parasitaemia <10,000/μl).