• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Infect Dis. Author manuscript; available in PMC May 1, 2011.
Published in final edited form as:
PMCID: PMC2853733

High heritability of malaria parasite clearance rate indicates a genetic basis for artemisinin resistance in Western Cambodia


In Western Cambodia malaria parasites clear slowly from the blood following treatment with artemisinin derivatives, but it is unclear whether this results from parasite, host, or other factors specific to this population. We measured heritability of clearance rate (CR), by examining patients infected with identical or non-identical parasite genotypes, using methods analogous to human twin studies. A substantial proportion (56-58%) of the variation in CR is explained by parasite genetics. This has two important implications: (1) selection with artemisinin derivatives will tend to drive resistance spread, (2) because heritability is high, genes underlying CR may be identified by genome-wide association.

Keywords: Artemisinin, clearance time, heritability, twin studies, resistance, microsatellite

Artemisinin combination therapies (ACTs) are the mainstay of global efforts to control Plasmodium falciparum malaria (1). Typically a three-day treatment course with an ACT reduces parasite densities by a factor of 108, and 95% of patients are malaria blood slide negative 48 hours after treatment (1). Artemisinin derivatives have been used successfully for more than 25 yrs in SE Asia. However, recent reports from Western Cambodia of delayed parasite clearance (2-4) following treatment with an artemisinin derivative have now been confirmed (5) and there is concern that these herald the emergence of parasites resistant to artemisinin derivatives.

While slow clearance rate (CR) in Western Cambodia is now undeniable, it is far from clear whether slow CR results from parasite, host, or other factors specific to this population. Parasites with slow CR after ACT do not show increased resistance to artemisinin compounds with conventional in-vitro testing relative to parasites from Western Thailand showing rapid CR (5). This contrasts with resistance to all other antimalarial drugs where in-vivo resistance is associated with reduced in-vitro susceptibility. There are several viable alternative explanations for slow clearance rates. Host immunity and splenic function are important contributors to parasite clearance following artemisinin treatment (6). Reduction in herd immunity, perhaps resulting from reduced transmission could decrease parasite clearance in Cambodia. Alternatively, host factors, such as hemoglobinopathies (7), or nutritional status, could also play a role.

In humans, identical twin studies provide an effective approach to determining if phenotypic traits have a genetic basis (8;9). If a trait is heritable we would expect identical twins to be more similar than non-identical or unrelated individuals. Such studies are most powerful when identical twins are reared in different households because this reduces common environmental influences. Twin studies demonstrate that traits as varied as height, addictive behaviour, and musical ability have a significant genetic basis. The same analytical framework can be used to analyse heritability in clonal organisms (10). In Southeast Asia malaria parasites that are identical-by-descent across the genome (clonally identical (CI)) are found frequently in different patients due to self-fertilization (11). CI malaria parasites are equivalent to identical twins reared apart, and can be used to measure the heritability (H2) of parasite traits. We measured CR in patients infected with CI or unrelated parasites to determine whether parasite genotype influences CR.

Subjects, materials and Methods

The blood samples and clearance data analyzed came from efficacy trials of artemisinin-based therapy in Pailin, Western Cambodia, in 2007-2008 (5). The trials are registered under ClinicalTrials.gov number NCT00493363 and Current Controlled Trials number ISRCTN15351875. Ethical approval was obtained from the Ministry of Health in Cambodia, the Oxford Tropical Medicine Ethical Committee, the WHO Research Ethics Review Committee, and the Technical Review Group of the WHO Western Pacific Regional Office. Parasite density was measured on admission and every 6 hours until clearance from the peripheral blood after treatment with artemisinin mono- or combination-therapy. The patients enrolled in the drug efficacy trials were given either artesunate [Guilin Pharmaceuticals] for 7 days in a dose of 2 mg/kg or 6 mg/kg (as a single or split dose), or artesunate for 3 days in a dose of 4 mg/kg or 8 mg/kg (as a single or split dose) followed by mefloquine [Medochemie] at 72 h (15 mg/kg) and 96 h (10 mg/kg) after admission (5). We plotted the natural log of parasite density against time for each patient and measured the slope to evaluate first-order clearance rate (CR).

We genotyped 18 microsatellite loci (12). Oligos, genomic location and polymorphism are detailed in the Supplementary online material. Parasites that were identical at all loci are referred as being CI. We compared variance of CR within and among CI parasites recovered from two or more patients and estimated H2 from the mean squares terms in the ANOVA, following methods using for heritability estimation using identical twins or clonal plants (10). In brief, we determined the within and among clone mean squares (MSe and MSb). The total genetic variance in CR (σG2) was estimated as (MSbMSe/n, where n is the weighted mean number of patients infected with each CI genotype. n is calculated as follows: n=[T(ni2T)](N1), where T is the total number of patients, N is the number of different CI genotypes, and ni is the number of patients infected with the ith CI genotype. The environmental variance (σE2) is estimated from the within clone variation as MSe and so H2=σG2(σG2+σE2). This analysis is directly equivalent to methods used in identical twin studies in humans or studies of clonal plants. To account for other covariates that could influence CR we included patient age as a continuous variable and gender and treatment regime as categorical variables in a regression analysis, and used the residuals in the ANOVA.


We analyzed parasite clearance data from 62 patients: 26 of the patients were recruited into the drug-efficacy trials between June-Nov 2007 and 36 between Mar-Oct 2008. Parasite clearance times ranged from 36-114 hrs (median=78hr). Because clearance time is dependent on admission parasitemia which varied 26-fold, we used CR as the primary outcome variable. Plots of the natural log of parasite density against time for each patient fitted well with a linear model (mean r2 = 0.94, range: 0.79-0.99) (Fig 1) and CR ranged from 0.072 to 0.251 (mean = 0.125) (Fig 2b). The majority (84%) of infections showed slow CR (< 0.15), while 16% were cleared rapidly (CR > 0.15). Patient age was significantly related to CR (F=5.373, n=62, p=0.024): older individuals cleared parasites more rapidly than younger individuals. We did not see significant effects of host gender or treatment regime.

Figure 1
Clearance profiles of Clonally Identical parasites. We plotted natural log of parasite density against time post treatment. Each of the nine panels shows clearance profiles for single CI genotype sampled from multiple patients. CI genotypes 1 and 2 (panel ...
Figure 2
(a) Clustering of parasites based on genetic similarity. We constructed a UPGMA tree from a pairwise matrix of allele-sharing (1 – proportion of shared alleles) between parasites using PHYLIP. We observed nine independent 18-locus genotypes that ...

We genotyped parasites collected on admission from each patient using 18 variable microsatellite loci. These loci had 2 - 11 alleles/locus and mean expected heterozygosity of 0.68 (range: 0.28-0.89) (Table S1). 11 infections contained multiple clones and were excluded leaving 51 single-clone infections. Nine CI genotypes, identical at all 18 loci, were found in more than one patient. Four CI genotypes were found twice, three genotypes were found three times, one genotype was found in 5 patients, and another in 6 patients (Fig 2a).

We conducted one-way analyses of variance (ANOVA) with CI genotype as the x-variable and lnCR or the residuals remaining after removing effect of age as the y-variables. Parasite genotype had a strong significant effect on lnCR (F=5.475, d.f.=7, p=0.0017), which remained significant after controlling for patient age (F=5.048, d.f.=7, p=0.0026). Inclusion of CI genotype 9, for which CR in one patient showed a poor fit to the linear model (r2=0.79), did not substantially alter this result (F=4.781, d.f.=8, p=0.0024 and F=4.385, d.f.=8, p=0.0039 after correction for age). The significant effects resulted from faster clearance of two CI genotypes (Fig 2c). We estimated heritability (H2) from the mean squared terms of the ANOVA to determine the contribution of parasite genes to trait variation (10). H2 was 0.58 ± 0.18 (1 s.d.) for lnCR and 0.56 ± 0.18 after correction for age.


These results demonstrate that a substantial proportion (56-58%) of the variation in CR results from parasite genetic factors and provide the first direct evidence that slow CR is a heritable parasite-encoded trait. This has two important consequences. (1) If reduced CR provides a selective advantage to parasites in the face of artemisinin based treatment regimens as suggested by the unusually high treatment failure rates in this area, then the alleles that underlie slow CR are expected to spread within populations, unless associated fitness costs outweigh selective benefits. There is insufficient historical information to evaluate whether CR has increased in Western Cambodia since artemisinin-based therapies were introduced >25 yrs ago or since 2000 when ACTs became the first line drug in Cambodia. However, the dramatic differences in parasite CR observed between western Cambodia and western Thailand, strongly suggest that this is the case. (2) Because CR shows high H2, it should be feasible to locate the parasite genes underlying CR by genome wide association (13). This would provide insights into the mechanism of resistance and mode of drug action, as well as providing markers for surveillance of resistance spread. As IC50 measured by growth inhibition assays are not associated with clearance time (5), measurement of CR is critical for these studies.

Two factors could generate an upward bias to our H2 estimates. If parasite clones are isolated from related patients, then host, rather than parasite factors, may determine clearance rates and result in overestimation of heritability due to parasite genes. To evaluate if this could influence our findings, we determined the origin of patients infected with clones 1 and 2 showing rapid CR. None of the five patients infected with these clones lived in the same household or village suggesting that patient relatedness is unlikely to affect our estimates H2. Transgenerational epigenetic effects - heritable changes in gene expression that are not caused by changes in DNA sequence – provide another possible source of H2 overestimation. In this case, as genetically identical clones share recent common ancestor, they might also be expected to share similar epigenetic modifications. However, recent work suggests that epigenetic changes are unlikely to contribute significantly to heritability unless such changes persist over a very long time scale (14). We believe this is a remote possibility.

The high H2 for CR in Western Cambodia contrasts starkly with comparable results from Western Thailand where more rapid CR is observed (Anderson et al, manuscript). In Western Thailand we found no significant H2 of CR in 61 patients (including 7 clonally identical genotypes) and no significant H2 of parasite reduction ratio at 24 or 48 hours in parasites from 185 patients (including 27 clonally identical genotypes). The observation of strong CR H2 and slow clearance in Western Cambodia and lack of CR H2 and rapid clearance in W. Thailand, is consistent with a focal origin and segregation of alleles that affect CR in Western Cambodia. That parasites with slow CR are in the majority in Western Cambodia suggests that genes conferring slow CR are close to fixation in this population. Effective containment of artemisinin-resistant Plasmodium falciparum should be implemented to prevent spread of this heritable parasite trait and is an urgent global priority (15).

Supplementary Material


We thank the patients for their participation; all staff members of the Pailin Referral Hospital for their dedicated care for the patients and assistance in the study; the Village Malaria Workers in Pailin for their collaboration; all Mahidol-Oxford Tropical Medicine Research Unit staff for help in the execution of the study.

Financial support: Molecular work at SFBR was funded by NIH RO1 AI48071 (TJCA) and conducted in facilities constructed with support from Research Facilities Improvement Program Grant Number C06 RR013556 from the National Center for Research Resources, NIH. The clinical studies received financial support from the Wellcome Trust of Great Britain (Major Overseas Programme – Thailand Unit Core Grant), the Li Ka Shing Foundation (grant nr B9RMXT0-2), and the World Health Organization through grants provided by the Bill and Melinda Gates Foundation (grant nr. 48821) and USAID (umbrella grant nr. AAG-G-00-99-00005).

The funding organizations played no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.


Potential conflicts of interest: All authors: no conflicts.

Reference List

1. White NJ. Qinghaosu (artemisinin): the price of success. Science. 2008 Apr 18;320(5874):330–4. [PubMed]
2. Denis MB, Tsuyuoka R, Lim P, Lindegardh N, Yi P, Top SN, Socheat D, Fandeur T, Annerberg A, Christophel EM, et al. Efficacy of artemether-lumefantrine for the treatment of uncomplicated falciparum malaria in northwest Cambodia. Trop.Med.Int.Health. 2006 Dec;11(12):1800–7. [PubMed]
3. Denis MB, Tsuyuoka R, Poravuth Y, Narann TS, Seila S, Lim C, Incardona S, Lim P, Sem R, Socheat D, et al. Surveillance of the efficacy of artesunate and mefloquine combination for the treatment of uncomplicated falciparum malaria in Cambodia. Trop.Med.Int.Health. 2006 Sep;11(9):1360–6. [PubMed]
4. World Health Oganization Global malaria Control and Elimination: report of a meeting on containment of artemisinin tolerance. World Health Organization; 2009.
5. Dondorp AM, Nosten F, Yi P, Das D, Phyo AP, Tarning J, Lwin KM, Ariey F, Hanpithakpong W, Lee SJ, et al. Artemisinin resistance in Plasmodium falciparum malaria. N.Engl.J.Med. 2009 Jul 30;361(5):455–67. [PMC free article] [PubMed]
6. Chotivanich K, Udomsangpetch R, McGready R, Proux S, Newton P, Pukrittayakamee S, Looareesuwan S, White NJ. Central role of the spleen in malaria parasite clearance. J.Infect.Dis. 2002 May 15;185(10):1538–41. [PubMed]
7. Weatherall DJ. Genetic variation and susceptibility to infection: the red cell and malaria. Br.J.Haematol. 2008 May;141(3):276–86. [PubMed]
8. Galton F. The history of twins as a criterion of relative powers of nature and nurture. Journal of the Royal Anthropological Institute of Great Britain and Ireland. 1875;5:391–406.
9. Boomsma D, Busjahn A, Peltonen L. Classical twin studies and beyond. Nat.Rev.Genet. 2002 Nov;3(11):872–82. [PubMed]
10. Lynch M, Walsh B. Genetics and Analysis of Quantitative Traits. 1 ed Sinauer Associates; Sunderland, MA: 1998. Twins and Clones; pp. 581–92.
11. Anderson TJ, Haubold B, Williams JT, Estrada-Franco JG, Richardson L, Mollinedo R, Bockarie M, Mokili J, Mharakurwa S, French N, et al. Microsatellite markers reveal a spectrum of population structures in the malaria parasite Plasmodium falciparum. Mol.Biol.Evol. 2000 Oct;17(10):1467–82. [PubMed]
12. Anderson TJ, Su XZ, Bockarie M, Lagog M, Day KP. Twelve microsatellite markers for characterization of Plasmodium falciparum from finger-prick blood samples. Parasitology. 1999 Aug;119(Pt 2):113–25. [PubMed]
13. Su X, Hayton K, Wellems TE. Genetic linkage and association analyses for trait mapping in Plasmodium falciparum. Nat.Rev.Genet. 2007 Jul;8(7):497–506. [PubMed]
14. Slatkin M. Epigenetic inheritance and the missing heritability problem. Genetics. 2009 Jul;182(3):845–50. PMCID:PMC2710163. [PMC free article] [PubMed]
15. Maude RJ, Pontavornpinyo W, Saralamba S, Aguas R, Yeung S, Dondorp AM, Day NP, White NJ, White LJ. The last man standing is the most resistant: eliminating artemisinin-resistant malaria in Cambodia. Malar.J. 2009;8:31. PMCID:PMC2660356. [PMC free article] [PubMed]
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...