Parasite Viability as a Measure of In Vivo Drug Activity in Preclinical and Early Clinical Antimalarial Drug Assessment

Abstract

The rate at which parasitemia declines in a host after treatment with an antimalarial drug is a major metric for assessment of antimalarial drug activity in preclinical models and in early clinical trials. However, this metric does not distinguish between viable and nonviable parasites. Thus, enumeration of parasites may result in underestimation of drug activity for some compounds, potentially confounding its use as a metric for assessing antimalarial activity in vivo. Here, we report a study of the effect of artesunate on Plasmodium falciparum viability in humans and in mice. We first measured the drug effect in mice by estimating the decrease in parasite viability after treatment using two independent approaches to estimate viability. We demonstrate that, as previously reported in humans, parasite viability declines much faster after artesunate treatment than does the decline in parasitemia (termed parasite clearance). We also observed that artesunate kills parasites faster at higher concentrations, which is not discernible from the traditional parasite clearance curve and that each subsequent dose of artesunate maintains its killing effect. Furthermore, based on measures of parasite viability, we could accurately predict the in vivo recrudescence of infection. Finally, using pharmacometrics modeling, we show that the apparent differences in the antimalarial activity of artesunate in mice and humans are partly explained by differences in host removal of dead parasites in the two hosts. However, these differences, along with different pharmacokinetic profiles, do not fully account for the differences in activity. (This study has been registered with the Australian New Zealand Clinical Trials Registry under identifier ACTRN12617001394336.).

Keywords: Plasmodium falciparum; antimicrobial activity; artesunate; clinical trials; drug activity; malaria; preclinical drug studies; viability.

FIG 1

In vitro and in vivo validation of the regrowth assay (RA) using a limiting dilution assay (LDA). (A) The left panel shows the estimated viable parasites (as a proportion of the viable parasites in the undiluted sample) at different dilutions of the infected red blood cells (RBCs). The right panel shows the estimated absolute number of viable parasites. The dashed diagonal line is the expected number of viable parasites. The red and blue circles are the estimated parasite viability from the LDA and the regrowth assay, respectively. The error bar indicates the 95% CI of these estimates. (B) Comparison of viability estimates from the regrowth assay and the limiting dilution assay after in vitro exposure of parasites to artesunate for different exposure times (also known as the parasite reduction ratio [PRR] assay) (21). The left panel shows the parasite viability estimates from the regrowth assay and the LDA normalized to the parasite viabilities estimated after 0-h exposure to the drug. The right panel compares the regrowth assay and the LDA. The dashed line indicates a one-to-one correspondence line. The continuous line represents the fitted correspondence from the parasite viabilities estimated from LDA and regrowth assay. (C) In vivo validation of the methods for estimating parasite viability. The figure compares the viability estimates obtained from the LDA and regrowth assay after parasites are exposed to treatment in vivo and collected and assessed ex vivo. These estimates were obtained using data from four experiments containing a total of 14 mice. The four colors and three symbols represent the four experiments and the three treatments, respectively. The triangles, diamonds, and squares represent the 50-mg/kg single-dose, 50-mg/kg/day 4-dose, and 200-mg/kg single-dose regimens, respectively.

FIG 2

Panel A shows the comparison of the drop in the number of circulating parasites (red) and the number of viable parasites (black) at 24 h posttreatment for all treatment groups. Panels B and C show the comparison of the drop in the number of circulating parasites and the number of viable parasites at 24 h posttreatment grouped by treatment. (D) Daily fold drop in viable parasite numbers following each of the daily doses of 50 mg/kg for 4 days.

FIG 3

Validation of parasite viability estimate from the regrowth assay (RA). Panel A shows the parasite viability estimates from the RA. Mice are designated as follows: mouse 1.1 is the first mouse treated from experiment 1, mouse 1.2 is the second mouse treated from experiment 1, mouse 2.3 is the third mouse treated from experiment 2, etc. Treatment 1, (top two rows, diamonds) 50 mg/kg/day for 4 days; treatment 2, (bottom row, squares), single 200-mg/kg dose. Note that mice treated with a single 50-mg/kg dose of artesunate were not included in this analysis because cure (parasitemia below the limit of detection) was not consistently achieved, and thus estimating recrudescence kinetics was not possible. The circles represent the measured circulating parasites. Each color represents an experiment. (B) Comparison of the minimum parasite viability estimated from the regrowth assay and the number of viable parasites that could lead to the recrudescence we saw in the mice.

FIG 4

Model of drug effects in humans and mice. The total circulating parasite concentration (black) and viable parasite concentration (blue) are shown for each human and mouse (designated as described in the legend to Fig. 3) infected with P. falciparum 3D7 parasites and treated with artesunate (human data from reference 19). The crosses represent the measured data. Viable and total parasite concentrations are reported as parasites per milliliter in humans and as a percentage of initial parasitemia in mice. Measurements of parasite concentration in humans were performed in triplicate. The lines are the fitted model of parasite killing and clearance. The dotted lines are fitted from a model where the maximum killing rates are the same in mice and humans but the EC₅₀s are different, while the continuous line represent a model where the EC₅₀ is the same, but the maximum killing rates are different between mice and humans. The dotted horizontal lines indicate limits of detection of 20 parasites/mL and 32 parasites/mL for viable and circulating parasites, respectively, and 0.01% parasitemia for circulating parasites in mice.