Abstract
Capillary samples offer practical benefits compared with venous samples for the measurement of drug concentrations, but the relationship between the two measures varies between different drugs. We measured the concentrations of lumefantrine, mefloquine, piperaquine in 270 pairs of venous plasma and concurrent capillary plasma samples collected from 270 pregnant women with uncomplicated falciparum or vivax malaria. The median and range of venous plasma concentrations included in this study were 447.5 ng/mL (8.81-3,370) for lumefantrine (day 7, n = 76, median total dose received 96.0 mg/kg), 17.9 ng/mL (1.72-181) for desbutyl-lumefantrine, 1,885 ng/mL (762-4,830) for mefloquine (days 3-21, n = 90, median total dose 24.9 mg/kg), 641 ng/mL (79.9-1,950) for carboxy-mefloquine, and 51.8 ng/mL (3.57-851) for piperaquine (days 3-21, n = 89, median total dose 52.2 mg/kg). Although venous and capillary plasma concentrations showed a high correlation (Pearson's correlation coefficient: 0.90-0.99) for all antimalarials and their primary metabolites, they were not directly interchangeable. Using the concurrent capillary plasma concentrations and other variables, the proportions of venous plasma samples predicted within a ±10% precision range was 34% (26/76) for lumefantrine, 36% (32/89) for desbutyl-lumefantrine, 74% (67/90) for mefloquine, 82% (74/90) for carboxy-mefloquine, and 24% (21/89) for piperaquine. Venous plasma concentrations of mefloquine, but not lumefantrine and piperaquine, could be predicted by capillary plasma samples with an acceptable level of agreement. Capillary plasma samples can be utilized for pharmacokinetic and clinical studies, but caution surrounding cut-off values is required at the individual level.CLINICAL TRIALSThis study is registered with ClinicalTrials.gov as NCT01054248.