Population pharmacokinetics and clinical response for artemether-lumefantrine in pregnant and nonpregnant women with uncomplicated Plasmodium falciparum malaria in Tanzania

Abstract

Artemether-lumefantrine (AL) is the first-line treatment for uncomplicated malaria in the second and third trimesters of pregnancy. Its efficacy during pregnancy has recently been challenged due to altered pharmacokinetic (PK) properties in this vulnerable group. The aim of this study was to determine the PK profile of AL in pregnant and nonpregnant women and assess their therapeutic outcome. Thirty-three pregnant women and 22 nonpregnant women with malaria were treated with AL (80/480 mg) twice daily for 3 days. All patients provided five venous plasma samples for drug quantification at random times over 7 days. Inter- and intraindividual variability was assessed, and the effects of covariates were quantified using a nonlinear mixed-effects modeling approach (NONMEM). A one-compartment model with first-order absorption and elimination with linear metabolism from drug to metabolite fitted the data best for both arthemether (AM) and lumefantrine (LF) and their metabolites. Pregnancy status and diarrhea showed a significant influence on LF PK. The relative bioavailability of lumefantrine and its metabolism rate into desmethyl-lumefantrine were, respectively, 34% lower and 78% higher in pregnant women than in nonpregnant patients. The overall PCR-uncorrected treatment failure rates were 18% in pregnant women and 5% in nonpregnant women (odds ratio [OR] = 4.04; P value of 0.22). A high median day 7 lumefantrine concentration was significantly associated with adequate clinical and parasitological response (P = 0.03). The observed reduction in the relative bioavailability of lumefantrine in pregnant women may explain the higher treatment failure in this group, mostly due to lower posttreatment prophylaxis. Hence, a modified treatment regimen of malaria in pregnancy should be considered.

FIG 1

Relationship between parasite density at enrollment in this study and residual levels of sulfadoxine in plasma before treatment in 14 pregnant women.

FIG 2

(A) Observed artemether (AM) (left panel) and dihydroartemisinin (DHA) (right panel) concentrations in plasma. Filled and empty circles represent pregnant and nonpregnant women, respectively. The solid line represents the average predicted concentrations, and the dashed lines show the 95th prediction interval. (B) Observed plasma lumefantrine (LF) (top panels) and debutyl-lumefantrine (DLF) (bottom panels) concentrations in pregnant and nonpregnant women. Triangles show residual plasma LF and DLF concentrations found prior to treatment initiation. The solid lines represent the mean population prediction, and the dashed lines show PI_95%.

FIG 3

Predicted median concentration of lumefantrine after administration of six 480-mg doses over 3 (continuous line) and 5 days (dotted line) in pregnant women. The median predicted concentrations on day 7 (168 h) (circles) with their PI_95% are shown for the two dosage regimens.

FIG 4

(A) Day 7 concentration of lumefantrine in plasma in pregnant (n = 32) and nonpregnant (n = 22) women. (B) Day 7 plasma lumefantrine concentration in women with ACPR (n = 48) and women with treatment failure (n = 6). The day 7 lumefantrine concentration could not be assessed in one woman with treatment failure, since a rescue treatment with quinine was given on day 1 because of early treatment failure.