Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

Impact of treatment and re-treatment with artemether-lumefantrine and artesunate-amodiaquine on selection of Plasmodium falciparum multidrug resistance gene-1 polymorphisms in the Democratic Republic of Congo and Uganda

  • Vito Baraka ,

    Roles Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing

    vitobaraka@gmail.com

    Affiliations National Institute for Medical Research, Tanga Centre, Tanga, United Republic of Tanzania, Global Health Institute, University of Antwerp, Antwerp, Belgium

  • Hypolite Muhindo Mavoko,

    Roles Conceptualization, Data curation, Methodology, Writing – review & editing

    Affiliations Global Health Institute, University of Antwerp, Antwerp, Belgium, Département de Médecine Tropicale, Faculté de Médecine, Université de Kinshasa, Kinshasa, Democratic Republic of Congo

  • Carolyn Nabasumba,

    Roles Conceptualization, Data curation, Investigation

    Affiliations Global Health Institute, University of Antwerp, Antwerp, Belgium, Epicentre Mbarara Research Base, Mbarara, Uganda

  • Filbert Francis,

    Roles Data curation, Formal analysis, Writing – review & editing

    Affiliation National Institute for Medical Research, Tanga Centre, Tanga, United Republic of Tanzania

  • Pascal Lutumba,

    Roles Conceptualization, Investigation, Methodology

    Affiliation Département de Médecine Tropicale, Faculté de Médecine, Université de Kinshasa, Kinshasa, Democratic Republic of Congo

  • Michael Alifrangis,

    Roles Conceptualization, Validation, Writing – review & editing

    Affiliations Centre for Medical Parasitology at the Department of Immunology and Microbiology, University of Copenhagen, Copenhagen, Denmark, Department of Infectious Diseases, National University Hospital (Rigshospitalet), Copenhagen, Denmark

  • Jean-Pierre Van geertruyden

    Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Validation, Writing – review & editing

    Affiliation Global Health Institute, University of Antwerp, Antwerp, Belgium

Abstract

Background

The emergence of resistance against artemisinin combination treatment is a major concern for malaria control. ACTs are recommended as the rescue treatment, however, there is limited evidence as to whether treatment and re-treatment with ACTs select for drug-resistant P. falciparum parasites. Thus, the purpose of the present study is to investigate the impact of (re-)treatment using artesunate-amodiaquine (ASAQ) and artemether-lumefantrine (AL) on the selection of P. falciparum multidrug resistance-1 (Pfmdr1) alleles in clinical settings.

Methods

P. falciparum positive samples were collected from children aged 12–59 months in a clinical trial in DR Congo and Uganda. Pfmdr1 single nucleotide polymorphisms (SNPs) analysis at codons N86Y, Y184F, and D1246Y were performed at baseline and post-treatment with either AL or ASAQ as a rescue treatment using nested PCR followed by restriction fragment length polymorphism (RFLP) assays.

Results

The pre-treatment prevalence of Pfmdr1 N86 and D1246Y varied significantly between the sites, (p>0.001) and (p = 0.013), respectively. There was borderline significant directional selection for Pfmdr1 184F in recurrent malaria infections after treatment with AL in Uganda site (p = 0.05). Pfmdr1 NFD haplotype did not significantly change in post-treatment infections after re-treatment with either AL or ASAQ. Comparison between pre-treatment and post-treatment recurrences did not indicate directional selection of Pfmdr1 N86, D1246 alleles in the pre-RCT, RCT and post-RCT phases in both AL and ASAQ treatment arms. Pfmdr1 86Y was significantly associated with reduced risk of AL treatment failure (RR = 0.34, 95% CI:0.11–1.05, p = 0.04) while no evidence for D1246 allele (RR = 1.02; 95% CI: 0.42–2.47, p = 1.0). Survival estimates showed that the Pfmdr1 alleles had comparable mean-time to PCR-corrected recrudescence and new infections in both AL and ASAQ treatment arms.

Conclusion

We found limited impact of (re-)treatment with AL or ASAQ on selection for Pfmdr1 variants and haplotypes associated with resistance to partner drugs. These findings further supplement the evidence use of same or alternative ACTs as a rescue therapy for recurrent P.falciparum infections. Continued monitoring of genetic signatures of resistance is warranted to timely inform malaria (re-)treatment policies and guidelines.

Introduction

Global malaria control efforts have expanded significantly. However, malaria remains a major health challenge with approximately 212 million new cases and 429,000 malaria-related mortality in 2015[1]. The success is mainly attributed to scaling up of coverage with insecticide-treated nets (ITNs) and deploying artemisinin-based combination therapies (ACTs) for the treatment of uncomplicated Plasmodium falciparum malaria [2]. The efficacy of ACTs depends on the synergistic action of short-acting artemisinin derivatives and a longer acting partner drug that eliminates the residual parasite load[3]. However, recent reports have documented diminishing efficacy of ACTs in Great Mekong Sub-region (GMS), raising serious concerns regarding the future of malaria control and elimination [46]. Despite these reports, ACTs remain largely efficacious with adequate cure rates in sub-Saharan Africa (SSA) [7,8]. In order to maintain ACTs as the cornerstone for malaria treatment, it is crucial that the artemisinin derivatives and their longer acting partner drugs retain high efficacy. Currently, the most used partner drugs in SSA are amodiaquine and lumefantrine which are combined with artesunate (artesunate-amodiaquine, ASAQ) or artemether (artemether-lumefantrine, AL).

Treatment failure is a complex problem interplay of several factors ranging sub-optimal drugs exposure, poor adherence to treatment, host-genetics factors and/or the emergence of drug resistance. Recrudescence rates are low in most countries in Africa and ACTs retain high efficacy as the first-line therapy for malaria treatment, the few reported recurrent infections are commonly new infections [7,8]. ACTs are increasingly an integral part of malaria control and are the recommended “rescue treatment” or “second-line treatment” for the treatment of recurrent P.falciparum infections in addition to Quinine+antibiotic [9]. However, the emergence of resistance/tolerance against the long-acting partner drugs, such as amodiaquine and lumefantrine, may consequently affect the efficacy of the ACTs and increase the odds for emerging resistance to the short-acting artemisinin-derivatives. Monitoring is, therefore, crucial to understanding the development of a drug-mediated selection of Pfmdr1 alleles in the patients that are treated and re-treated with ACTs as a rescue treatment. In the Democratic Republic of Congo, uncomplicated malaria cases are treated with ASAQ as the first line while in Uganda, AL is the recommended regimen.

Amodiaquine (AQ) is a slow-acting 4-aminoquinoline antimalarial drug with a terminal plasma half-life between 9–18 days, acting through its active metabolite desmethyl amodiaquine [10]. Lumefantrine is an arylamino-alcohol drug with a half-life of 3–5 days and has shared structural similarity to mefloquine and quinine[10]. The molecular targets and mechanisms of action of lumefantrine are yet to be fully unravelled [11]. However, deteriorating clinical outcomes of various antimalarial drugs, including some of the partner drugs, in particular to 4-aminoquinoline such as chloroquine (CQ), amodiaquine (AQ) and piperaquine (PQ) and aryl-amino alcohol based drugs (lumefantrine, quinine and mefloquine) is alarming as may increase rates of treatment failures.

P. falciparum multidrug resistance transporter 1 (Pfmdr1) gene occurs on Chromosome 5 and encodes for P-glycoprotein homologue 1 (Pgh1) protein located in the digestive vacuole of the parasite. Pfmdr1 is of interest because of the described role of the polymorphisms at codon N86Y, Y184F, S1034C, N1042D, and D1246Y in drug resistance. The polymorphisms were associated with altered parasite susceptibility to several classes of antimalarials [1115]. Drug pressure due to artemisinin derivatives partner drugs was shown to exert directional selection of Pfmdr1 86Y, Y184 and 1246Y variants in amodiaquine(AQ) and while AL and Mefloquine (MQ) inversely selected for N86, 184F D1246 variants [12,1618]. Similar observations were documented in in vitro and ex vivo susceptibility assays [1921]. Pfmdr1 SNPs exert altered susceptibility profile to other antimalarials including mefloquine(MQ), chloroquine(CQ) and quinine(QN) [13,14]. Interestingly, it was recently shown that Pfmdr1 86Y increases parasite susceptibility to dihydroartemisinin, which is the active artemisinin metabolite of artemisinin derivatives[15].

In addition, Pfmdr1 haplotypes are an important determinant in modulating the level of resistance. Early treatment failures(ETF) that occur between day 2–3 post-treatment with AL and DHA-PQ, were more significantly shown to harbour Pfmdr1 NFD haplotype in the previous study [22]. Furthermore, Pfmdr1 haplotypes are differentially selected, for instance, Pfmdr1 YYY haplotype (86Y, Y184 and 1246Y) was shown to be selected after exposure to artesunate-amodiaquine(ASAQ) while AL inversely exerted selective pressure of Pfmdr1-NFD haplotype(N86, 184F and D1246) [16,23]. The observation was further exemplified by Malmberg et al [24] that well demonstrated the ability of the parasites harbouring Pfmdr1 NFD haplotype to withstand a 15-folds concentration of lumefantrine compared to Pfmdr1 YYY in re-infecting parasites [24]. Also, evidence suggests an association between the Pfmdr1-NFD and increased risk of gametocyte carriage following treatment with AL [23]. In addition, variation in Pfmdr1 copy number (CNVs) has been shown to modulate drug sensitivity pattern of the partner drugs, mefloquine, lumefantrine and piperaquine, confined in the SEA in the Great Mekong sub-region [12,25,26].

We investigated the impact of subsequent treatment courses with ACTs, either the same or alternative, on the selection of Pfmdr1 polymorphisms in clinical settings in DR Congo and Uganda.

Material and methods

Study design and participants

The protocol and the outcome of the clinical trial for this molecular analysis were published separately [27,28]. Briefly, the trial was a bi-centre, open-label, randomised, three-arm phase 3 trial (2:2:1 ratio) conducted in Lisungi health centre in DR Congo, and Kazo health centre in Uganda in 2012–2014. In the pre-randomized controlled (Pre-RCT) phase of the trial, the patients were treated with the first-line treatment as per respective country i.e. artesunate-amodiaquine (ASAQ) in DR Congo and artemether-lumefantrine (AL) in Uganda and followed up for 42 days. Patients who experienced malaria between day 14–42 were enrolled in randomised control trial (RCT) phase and assigned either ASAQ, AL or Quinine+clindamycin(QnC). The patients that presented with clinical or parasitological failure post day 14 were recruited in post-randomization control (post-RCT) trial phase and were re-treated according to the country malaria treatment guideline [28].

Sample collection

P. falciparum microscopically positive samples for this study were collected in Uganda and DR Congo as part of the QuinACT clinical trial described above [28]. We analysed samples in pre-RCT phase, RCT phase and post-RCT phases from ASAQ and AL arms of the QuinACT trial. The samples from QnC arm were excluded because Pfmdr1 is not associated with Quinine resistance. Fingerprick samples of blood were collected on filter paper (Whatman 3MM, Maidstone, UK), air-dried, labelled and stored in a desiccator containing silica gel until further processing.

DNA extraction

DNA isolation from dried blood spots was done using QIAmp DNA Mini Kit as per manufacturer’s instructions (Qiagen®, Hilden, Germany). Isolated DNA was stored at -20°C until further use in the downstream applications.

PCR and RFLP assays

The extracted samples were amplified by outer and nested PCR protocols for amplification of Pfmdr1 genes targeting single nucleotide polymorphisms (SNPs) at codons N86Y, Y184F and D1246Y followed by restriction fragment length polymorphism (RFLP) digestion using restriction enzymes as previously described [29].

Data analysis

Data management and analyses were performed using Stata, version 13(StataCorp, College Station, TX, USA). Prevalence of Pfmdr1 N86Y, Y184F and D1246Y alleles and haplotypes were calculated and compared between the treatment arms and sites. Mixed alleles are presented in prevalence data, however, excluded when constructing the Pfmdr1 haplotypes. Chi-square or Fisher exact tests were used to assess the associations between categorical variables. McNemar’s test or exact McNemar's test were used to determine the directional selection for Pfmdr1 alleles in patients with P.falciparum recurrent infections in matched analysis between pre-treatment and post-treatment in the different trial phases. Risk Ratios (RR) and adjusted odds ratios (aORs) with 95% confidence intervals (CIs) were used to assess the association between Pfmdr1 SNPs or haplotypes and treatment outcomes. Multivariate logistic regression was used to determine factors associated with PCR-corrected treatment failures and new infections adjusted for such as age, site, anaemia, fever, and parasite density. The cumulative risk of treatment failure or new-infections by Pfmdr1 alleles and haplotypes were analysed using Kaplan-Meier survival analysis and the comparison was made by using log-rank test. P-values were considered significant at ≤0.05.

Ethical considerations

The study was approved by the Ethical Committee of the University of Antwerp, Belgium, the Uganda National Council for Science and Technology, Uganda and the Ethics Committee of the School of Public Health, University of Kinshasa, DR Congo. Informed consent was obtained from all participants through their parents or legal guardians [27]

Results

Demographic and clinical characteristics of patients

A total of 755 samples were analysed for Pfmdr1 SNPs polymorphisms in pre-RCT, RCT and post-RCT phases. Of those, proportions of the successfully genotyped samples were; 732/755 (97%), 729/755 (96.6%) and 736/755 (97.5%) in codon 86, 184 and 1246 respectively. Of those, a total of 242 patients with uncomplicated P.falciparum infections were randomised to either ASAQ (n = 114) or AL (n = 128) (Table 1). The proportion of males and females enrolled was similar between both study sites and also by treatment randomization (Table 1). The mean weight was 12.7kg and 12.0kg in DR Congo and Uganda, respectively and was also similar between treatment arms. In terms of the clinical characteristics, the geometric mean parasitemia was comparable by site and treatment allocation. The proportion of the treatment allocation for patients receiving AL vs ASAQ was 56.5% vs 43.5% and 50% vs 50% in Congo and Uganda, respectively. The bednet ownership was approximately 50% for each site and was marginally higher in AL (58.7) vs ASAQ (41.3) treatment arm, (p = 0.07), (Table 1).

thumbnail
Table 1. Patients demographic and clinical characteristics at baseline of the randomisation phase of the QuinACT trial in DR Congo and Uganda sites.

https://doi.org/10.1371/journal.pone.0191922.t001

Baseline prevalence of Pfmdr1 alleles and combined Pfmdr1 haplotypes by the site during the RCT phase

Overall, in DR Congo the Pfmdr1 SNPs results revealed that about half of the parasites harboured Pfmdr1 N86, whereas in Uganda it was high almost at fixation point(52.8% vs 91%), (χ2 = 38.26, p<0.001). The mutant Pfmdr1 86Y was found at high proportion in DR Congo in comparison to Uganda (37.9% vs 7.5%). Pfmdr1 Y184 (wildtype) variant was observed at high proportions at both sites (56.5% vs 67.2%), (χ2 = 5.04, p = 0.08). The prevalence of Pfmdr1 D1246 was significantly higher in DR Congo in comparison to the Ugandan site (88.9% vs 73.9%), (χ2 = 8.65, p = 0.013). Logically, the Ugandan site harboured higher prevalence of the Pfmdr1 1246Y (9.2% vs 20.9%). The prevalence of the Pfmdr1-NFD and non-NFD haplotypes did not seem to vary between the sites (χ2 = 0.06, p = 0.81), (Table 2).

thumbnail
Table 2. Baseline prevalence of Pfmdr1 polymorphisms and haplotypes among isolates from DR Congo and Uganda site in the randomisation phase.

https://doi.org/10.1371/journal.pone.0191922.t002

Selection of Pfmdr1 alleles and haplotypes associated with ASAQ and AL resistance in different trial phases

The analyses of the matched samples at baseline (D0) in comparison to the post-treatment recurrence infections (PCR-corrected recrudescent and new-infections) at the pre-RCT, RCT and post-RCT phases were performed using McNemar’s test for isolates at Pfmdr1 at codon 86, 184, and 1246 (Table 3). Following treatment with ASAQ in DR Congo in the pre-RCT, there was no evidence to suggest directional selection for Pfmdr1 86Y (p = 0.32), Y184 (p = 0.13) and 1246Y (p = 0.27) alleles in the recurrent infections. Likewise, in the RCT phase where the patients were re-treated with the same treatment (ASAQ) or the alternative treatment (AL), there was no apparent selection of certain Pfmdr1 alleles, albeit the sample size was low. Similar results were found for the post-RCT phase.

thumbnail
Table 3. Drug-mediated directional selection of the Pfmdr1 alleles at codon 86, 184, and 1246 in isolates collected from children with recurrent malaria infections following (re-)treatment with AL or ASAQ.

https://doi.org/10.1371/journal.pone.0191922.t003

Similarly, in Ugandan site, no significant evidence for directional selection of Pfmdr1 N86 (p = 0.07) and D1246 (p = 0.6) in patients with recurrent infections treated with AL treatment in the pre-RCT phase. However, there was there was marginal evidence to suggest a directional selection of Pfmdr1 184F (p = 0.0555), (Table 3). For the RCT phase, patients who received same AL as a rescue treatment and those received alternative treatment, ASAQ, no significant selection for the Pfmdr1 allele was detected, (Table 3). Also, similar findings were observed in post-RCT where patients with recurrent infections and treated using AL. However, the sample size was low to provide statistical power for the recurrent infections in the RCT and Post-RCT phase.

Further, we assessed the selection of Pfmdr1 NFD haplotype (N86- 184F-D1246 vs. non-NFD haplotypes comparing baseline with post-treatment isolates (Fig 1). In the DR Congo-site, the proportion of the Pfmdr1-NFD haplotypes did not significantly change in isolates from patients that received either same ASAQ or AL as an alternative drug in the course of the study (Fig 1). However, for the Ugandan site (with exception of the RCT phase), there was a marginal reduction in the proportion of isolates harbouring Pfmdr1-NFD haplotype in new infections in AL treatment arm (p = 0.052), (Fig 1). Although not statistically significant, the general trend indicates that the proportion of the Pfmdr1 NFD in treatment failure (TF) slightly increased from the baseline in pre-RCT and RCT in AL retreatment arm (Fig 1A). On the contrary, retreatment with ASAQ seems to have the opposite effect on the proportion of NFD haplotype (Fig 1).

thumbnail
Fig 1. Comparison of the distribution of Pfmdr1 NFD and non-NFD haplotypes (95% Confidence intervals) by site in pre- and post-treatment outcomes in AL and ASAQ treatment arms.

Haplotypes present amino acid residues at codon N86Y, Y184F, D1246Y (N = Asparagine, Y = Tyrosine, F = Phenylalanine, D = Aspartic acid, K = Lysine, T = Threonine). D0 = Day 0, TF = PCR-corrected treatment failure, ACPR = Adequate clinical and parasitological response.

https://doi.org/10.1371/journal.pone.0191922.g001

Association between Pfmdr1 alleles and the treatment failure after treatment with AL and ASAQ in randomization phase

We assessed the relative risk (RR) of treatment failure (PCR adjusted) in the RCT phase and of various Pfmdr1 alleles and as well, the haplotypes following treatment with AL and ASAQ (Table 4). The parasites harbouring the Pfmdr1 86Y had significantly lower risk of treatment failure post AL treatment compared to 86N (relative risk [RR] = 0.34; 95% CI = 0.11–1.05, p = 0.04), (Table 4). The Pfmdr1 Y184 and D1246 were not associated with risk of treatment failure after AL treatment, (RR = 0.50; 95% CI: 0.17–1.49, p = 0.26) and (RR = 1.02; 95% CI: 0.42–2.47, p = 1.0), respectively (Table 4). Likewise, the NYY haplotype was not significantly associated with risk of treatment failure relative to the Pfmdr1 NFD haplotype after exposure to AL (RR = 1.16; 95% CI: 0.70–1.92; p = 0.59). None of the Pfmdr1 alleles was associated with ASAQ treatment failure, (Table 4). Similarly, Pfmdr1 NYY and NYD did not seem to affect the risk of ASAQ treatment failure, (RR = 0.54; 95% CI: 0.08–3.64; p = 0.65) and (RR = 1.18; 95% CI: 0.56–2.52; p = 0.78), respectively, (Table 4).

thumbnail
Table 4. Association of Pfmdr1 alleles and haplotypes and risk of RCR-adjusted treatment failure following treatment with Artemether-Lumefantrine and Artesunate-Amodiaquine.

https://doi.org/10.1371/journal.pone.0191922.t004

Predictors for P. falciparum recrudescence and new-infections

Anaemia was a strong predictor of recrudescence in patients treated with AL (aOR = 16.71; 95% CI: 1.86–149.26, p = 0.012), (S1 Table). There was no evidence to suggest the role of Pfmdr1 SNPs in modulating the risk of recrudescence in this AL treatment arm (S1 Table). In addition, in the AL treatment arm, Pfmdr1 184F was associated with an increased risk of new-infections (aOR = 2.92; 95% CI: 0.99–8.53; p = 0.0486). None of the Pfmdr1 variants and were associated with risk of PCR-corrected recrudescent or new infections in the ASAQ arm. Analysis of the Pfmdr1-NFD and non-NFD haplotypes did not reveal significant association trend with either the re-infection or recrudescence in either AL or ASAQ treatment arms, (S1 Table).

Survival estimates to recrudescence and new infections stratified by Pfmdr1 alleles

The 28-days survival estimates of the Pfmdr1 alleles in relation to the occurrence of recrudescence was conducted for AL (n = 110) and ASAQ (n = 104) treatment arms in the RCT phase.

We noted no difference in AL treatment arm. Time to treatment failure tended to occur late (> 20days) after AL treatment and was comparable between Pfmdr1 N86 vs 86Y, (27.7 mean days (md); 95% CI: 27.4–28.0 vs 27.4 md; 95% CI: 26.3–28.5; p = 0.97). Similarly, the mean time to treatment failure in patient harbouring either Pfmdr1 184F and 184Y had comparable survival estimates [27.9md; 95% CI: 27.7–28.1 vs 27.3md; 95% CI: 26.4–28.1; p = 0.75). Similar trend was observed with regard to the Pfmdr1 D1246 and 1246Y, (27.6md; 95% CI: 27.1–27.9 vs 28md; 95% CI: 28–28; p = 0.42), (Fig 2).

thumbnail
Fig 2. Survival estimates of the cumulative risk of treatment failure by Pfmdr1 alleles in patients treated with AL and ASAQ in the randomisation phase.

Fig 2(A1-A3) represent the risk of PCR-adjusted recrudescence in participants who received AL harbouring by Pfmdr1 variant. Similarly, Fig 2(B1-B3) represents the risk of PCR-adjusted recrudescence in participants who received ASAQ by Pfmdr1 variants.

https://doi.org/10.1371/journal.pone.0191922.g002

Treatment failure tended to occur late (>20 days) after ASAQ treatment arm with close similarity between Pfmdr1 N86 vs 86Y alleles, (27.5md; 95% CI: 26.9–27.9 vs 27.3md, 95% CI: 26.4–28.2; p = 0.97). Similarly, patients harbouring either Pfmdr1 184F or Y184 also had comparable mean-time to recrudescence [(27.6md; 95% CI: 27.1–28.0 vs 27.2md; 95% CI: 26.2–28.1; p = 0.75). Furthermore, no difference was observed when comparing Pfmdr1 D1246 and 1246Y (27.4md; 95% CI: 26.9–27.9 vs 27.5md; 95% CI: 26.6–28.4; p = 0.42), (Fig 2). A similar trend was observed when considering new-infections (PCR-adjusted) and crude treatment failure (PCR-unadjusted) in both AL and ASAQ treatment arms, (S1 Fig).

Discussion

Our data suggest a limited drug-mediated selection of Pfmdr1 alleles after exposure to AL and ASAQ suggesting a restricted impact on the selection of Pfmdr1 alleles previously associated with reduced sensitivity to 4-aminoquinolines and aryl-aminoquinolines partner drugs. The results supplement the evidence on the ACTs as a possible rescue treatment for curing recurrent P.falciparum malaria with restricted impact on the selection of drug-resistant malaria parasites [27].

We observed significant variation in the baseline Pfmdr1 among the two sites. In Uganda, the baseline prevalence of the Pfmdr1 N86 was significantly higher while in DR Congo Pfmdr1 86Y was more predominant. However, on both sites, there was no evidence of drug-mediated selection in either phase except the observed borderline directional selection for Pfmdr1 184F and borderline significance for Pfmdr1 N86 in recurrent infections after AL treatment (Table 3). The observed selection could indicate that AL for the Pfmdr1 184F allele in patients with malaria recurrences and could subsequently increase the circulation of parasite harbouring the Pfmdr1 NFD haplotype known to tolerate high concentrations of lumefantrine [24]. The results mirror the previous finding that demonstrated a strong selection of Pfmdr1 184F that persisted for a prolonged duration (2months) after treatment with AL, but not ASAQ in Uganda [30]. Reports in DR Congo suggests a high prevalence of Pfmdr1 86Y, 184F, D1246 [31] while in Uganda evidence suggest a high prevalence of Pfmdr1 N86 in most settings[17,30]. However, we found no marked difference in the frequency of Pfmdr1 NFD and non-NFD haplotypes in Ugandan and DR Congo in patients. Although ASAQ is the recommended first-line treatment in DR Congo, both AL and ASAQ are used in the treatment of uncomplicated malaria exerting a balanced selection pressure on Pfmdr1 alleles compared to Uganda where AL is more widely used. The trend suggests that usage of different or alternating ACTs may delay the development of resistance by exerting inverse selection pressure on the parasites [26]. We found significant evidence increased in relative risk of treatment failure (PCR-adjusted) associated with Pfmdr1-N86 in patients treated with AL, however, none of the Pfmdr1 alleles was associated with ASAQ treatment failure, (Table 2). The findings are in concordance with a meta-analysis data that suggested 5-folds increased the risk of treatment failure in the patients treated with AL[26]. However, we could not associate the increased risk of ASAQ of failure in patients harbouring Pfmdr1 86Y after treatment which compares to other reports elsewhere [16,26]. None of the Pfmdr1 haplotypes was associated with AL or ASAQ treatment, however, there were few Pfmdr1 haplotypes detected in post-treatment to allow meaningful statistical power for comparison.

In multivariate analysis, none of the Pfmdr1 alleles had no significant association with treatment failure or risk of new-infections after treatment with ASAQ. In AL treatment arm, patients harbouring Pfmdr1-184F had significantly higher risk of new-infection, in line with a study that reported an association between lumefantrine and Pfmdr1 F184 in ex-vivo assay [32]. Interestingly, patients with anemia were increasingly at high risk of recrudescence in AL treatment arm, this may be supported the fact that the risk of acquiring anaemia is high in patients with high parasite density (<200,000/μL)) which was shown to be an independent risk factor for clinical treatment failure by up to 60% for the different antimalarial treatment[33]. This is further reinforced by the evidence that falciparum infected patients that responded poorly to the treatment were more likely to recrudesce and consequently increased their risk of anaemia [34]. However, it is reassuring that, with the high efficacy of AL, the risk of treatment failure and selection of drug-resistant parasites after repeated ACT administration is low in most endemic settings in SSA. Additionally, Pfmdr1 NFD haplotype showed increasing trend in the RCT phase in patients with TF administered AL as a rescue treatment, these findings are in line with evidence from previous studies that demonstrated increased level after AL exposure [13,16]. The comparison of the pre- and post-treatment indicate an increasing overall trend indicate that the proportion of the Pfmdr1 NFD in treatment failure(TF) slightly increased from the baseline in pre-RCT and RCT in AL retreatment arm (Fig 1), however, this was not statically significant. On the contrary, retreatment with ASAQ seems to have the opposite trend on the proportion of Pfmdr1 NFD haplotype (Fig 1). In most of the countries where AL is the first-line treatment, the proportion of the Pfmdr1 Y86 seems to decline over time decreasing the overall proportion of NFD haplotypes, further depicting the opposing drug-mediated selection of resistant alleles [29,35,36].

With regard to the impact of treatment and re-treatment on selection for drug resistance markers, we did not observe a directional selection of Pfmdr1 86Y, Y184 and 1246Y in patients that received ASAQ as their first-line and as an alternative therapy in RCT phase in both sites. Similarly, the same trend for with regard to the selection of the Pfmdr1 N86 and D1246 in patients that received AL as the first-line regimen and then randomised to receive the same AL or ASAQ as a rescue therapy in RCT phase. Generally, there is evidence that (re-) treatment with same or alternative ACT offers limited impact in terms of selection for Pfmdr1 resistant alleles and haplotypes as documented in most of the trial phases, pre-RCT, RCT, and post-RCT. This could be further be supported by other studies that documented related fitness cost as it was shown by Fröberg et al [37], that demonstrated that although ASAQ selects for Pfmdr1 mutations associated with ASAQ reduced susceptibility, there is a significant fitness-cost incurred to the parasite rendering them unfit to compete with the wild type variants. Our findings, thus, suggest the limited impact of the ASAQ and AL on selection for Pfmdr1 allele when used as a rescue therapy in (re-)treatment of recurrent infections in real-life settings.

Currently, there are limited systematic studies that evaluated changes of Pfmdr1 alleles and haplotypes in the context of (re-)treatment approach. In a prolonged follow-up in Uganda, there was a clear significant selection of Pfmdr1 N86, 184F and D1246 alleles prior treatment with AL but not after ASAQ treatment [30]. Also, recent findings in Burkina Faso presented conflicting results on select of Pfmdr1 N86 [38,39]. Other studies suggest increased selection of the Pfmdr1 alleles prevalence in the parasite population following treatment with AL and ASAQ [14,18,26,40,41]. The plausible explanation for the observed differences could be the differences in country treatment policy. Also, the shorter half-life of lumefantrine is unlikely to have conferred strong directional selection for Pfmdr1 copy number in sub-Saharan African parasite populations that are known to harbour less copy number compared to SEA where MQ is widely used [12,26]. I addition, the synergistic action of artemether and lumefantrine exerted on the parasite could also play a role, artemether has retained the high efficacy due to the rapid reduction in parasite biomass rendering a relatively low risk to parasite infections being exposed to the selective pressure of the partner's drug.

Furthermore, we presume that other factors such as drug bioavailability, transmission intensity, intrinsic human genetic factors and immunity could also contribute to the observed outcomes, however, these were not explored in the scope of the present study.

As a limitation to the present study, the association between risks of treatment failure in relation to Pfmdr1 genotypes was restricted by the low sample size in the RCT and post-RCT. Furthermore, we did not analyse changes in Pfmdr1 copy number variation (CNV) that has been previously associated with lumefantrine tolerance. In GMS where parasites are resistant to artemisinin derivatives and partner drugs such as Mefloquine (MQ), increased Pfmdr1 CNVs is reportedly high in contrast to most endemic settings in sub-Saharan Africa [12,26]. Nonetheless, despite the observed occurrence of wild haplotype Pfmdr1 N86, Y184, and D1246 alleles in pre- and post-treatment isolates, AL still maintain high 28-day in vivo efficacy (>90% PCR-adjusted efficacy rate) in parasite populations in Uganda and elsewhere in sub-Saharan Africa [8]. In DR Congo, ASAQ retains high sensitivity due to the demonstrated low occurrence of the Pfmdr1 YYY and also the rare occurrence of SVMNT haplotype reported elsewhere[42]. Overall, there is limited reports evidence of declining ACT efficacy, except the demonstrated marginal elongation of parasites clearance time (PCT) and cases of treatment failure in clinical the coast of Kenya and report of imported malaria cases isolated from UK based travellers returning from Africa [43,44]. The high efficacy is depicted in the recent larger multicentre studies that demonstrated distinct clusters of parasite population in Africa and South-East Asian isolates with the limited dispersal of parasites strains carrying Pfkelch-13 key mutations in Africa settings and these include baseline samples from our QuinACT trial[4547]. However, despite these observations, new-infections remain exposed to sub-therapeutic levels of partner’s drugs due to their relative longer elimination half-lives and this may increase drug-mediated selective pressure on the Pfmdr1 gene. Additionally, over-prescription and over-the-counter prescription of ACTs both first-line and the second line should be restricted to reduce unnecessary drug pressure.

Furthermore, our data suggest no difference in time to PCR-adjusted treatment failure in both treatment arms in standard 28-days follow-up during the randomisation phase (Fig 2, and S1 Fig). Also, similar results were observed when considering crude treatment failure (PCR unadjusted treatment failure) and PCR-corrected new-infections. There was a low risk of early PCR-adjusted treatment failure supporting the previous reports that the risk of Pfmdr1 associated new-infections occurs later in >20days [26]. In most settings in Africa, the majority of malaria recurrences are classified as new-infections after PCR-correction. However, we did not observe significant differences in survival estimates between Pfmdr1 wild variants treatment failure or new-infections in both AL and ASAQ arms as reported elsewhere[26].

In light of our findings, we have demonstrated that AL and ASAQ can be used as a rescue treatment to treat recurrent falciparum malaria in real-life settings without exerting additional selection pressure on the Pfmdr1 gene. The observed selection of Pfmdr1 184F in new infections after treatment with AL warrants further investigations as the drug is increasingly used for (re-)treatment of recurrent malaria infections in several endemic countries. It may seem, therefore, plausible to consider the use of alternative ACT with the opposing mechanism of selection against on the Pfmdr1 gene as a stopgap strategy to counterbalance drug-mediated should ACT resistance worsen due to selection pressure. For instance, in countries where treatment policy recommends AL as the first-line therapy, ASAQ or DHA-PQ could be considered reciprocally to reduce the risk of directional drug-mediated selection of resistant alleles, a similar strategy has been previously supported [26,37,48]. Additionally, there is accrued evidence that simultaneous use of multiple first-line ACT therapies (MFT) is an ideal strategy to delay evolution and spread of resistance against artemisinin derivatives and partners drugs, on-going evaluation will inform on the appropriate models to facilitate the implementation and management of MFT strategy for the treatment of uncomplicated malaria[4951]. Thus, to counterbalance the effect of drug-mediated selection, deployment of alternative ACTs with antagonistic effect or implementation of strategies like MFT may offer a way to mitigate resistance against partner drugs. Indeed, a recent clinical trial in Uganda demonstrated the superiority of ASAQ compared to AL in reducing the risk of recurrences, and this was partly attributed to the parasite genetic background [52]. In the meantime, however, using the same ACT as rescue therapy does not seem to compromise treatment outcomes, presumably due to high efficacy of artemisinin derivatives and the fact that majority of malaria recurrences due to new infections. Ensuring patients adherence to (re-)treatment with ACT regimens and providing accurate malaria diagnosis before antimalarial treatment will undoubtedly contribute the most to enhancing the continued efficacy of the drugs.

Of note, de facto ACT resistance is yet to spread or emerge in most endemic settings in Africa, however, future antimalarial drug discovery and clinical trials need to continue evaluation of new treatment combination with opposing mechanisms of selection to reduce the potential risk for selection in malaria endemic transmission settings where patients are exposed to frequent treatment and re-treatment with ACTs. Nonetheless, the present study provides timely and encouraging molecular results on the impact of drug-mediated selection to support the use the same or alternative ACTs for (re-)treatment in endemic countries. In conjunction with these strategies, regular monitoring is warranted to provide evidence to inform policy on the long-term effectiveness of ACT in control and elimination efforts.

Conclusion

In summary, our study demonstrates the limited impact of (re-)treatment with AL or ASAQ on selection for Pfmdr1 alleles and haplotypes. These findings further supplement the evidence use of same or alternative ACTs as a rescue therapy for (re-)treatment of recurrent P.falciparum infections. Continued monitoring of genetic signatures of resistance is warranted to timely inform malaria (re-)treatment policies and guidelines.

Supporting information

S1 Table. Multivariate logistic regression model of the association between Pfmdr1 SNPs and PCR-corrected recrudescence and new infections of patients treated with AL and ASAQ on day 28 follow-up in RCT phase.

https://doi.org/10.1371/journal.pone.0191922.s001

(DOCX)

S1 Fig. Cumulative risk of crude treatment failure (PCR-unadjusted) by Pfmdr1 in patients treated with AL and ASAQ in the randomisation phase.

Figure A1-A3 represent the risk of crude treatment failure in patents who received AL by Pfmdr1 variants. Figure B1-B3 represents the risk of crude treatment failure in participants who received ASAQ by Pfmdr1 variants.

https://doi.org/10.1371/journal.pone.0191922.s002

(TIF)

Acknowledgments

The study participants in the QuinACT trial at Lusungi health centre in DR Congo and Kihuru Health Centre in Uganda are thanked for agreeing to take part in the study. Rashid Madebe (NIMR, Tanzania) and Ulla Abildtrup (Centre for Medical Parasitology, University of Copenhagen, Denmark) are thanked for their technical support during the laboratory analyses. Gwen Lemey and Jan Vervoort (Global Health Institute, Universiteit Antwerpern, Belgium) and Claire Alberte Charasse (Centre for Medical Parasitology, University of Copenhagen, Denmark) are thanked for their logistical support.

References

  1. 1. OMS, (Organisation Mondiale de la Santé). World malaria report 2016. 2016; 186.
  2. 2. World Health Organization, WHO. Guidelines for the treatment of malaria. Guidelines For The Treatment of Malaria. 2015.
  3. 3. Nosten F, White NJ. Artemisinin-based combination treatment of falciparum malaria. 2007;
  4. 4. Amaratunga C, Lim P, Suon S, Sreng S, Mao S, Sopha C, et al. Dihydroartemisinin-piperaquine resistance in Plasmodium falciparum malaria in Cambodia: A multisite prospective cohort study. Lancet Infect Dis. Lancet Publishing Group; 2016;16: 357–365. pmid:26774243
  5. 5. Ariey F, Witkowski B, Amaratunga C, Beghain J, Langlois A-C, Khim N, et al. A molecular marker of artemisinin-resistant Plasmodium falciparum malaria. Nature. Nature Publishing Group, a division of Macmillan Publishers Limited. All Rights Reserved.; 2014;505: 50–55. pmid:24352242
  6. 6. Dhorda M, Cheah PY, Pukrittayakamee S, Ashley EA, Anderson TJC, Nair S, et al. Spread of artemisinin-resistant Plasmodium falciparum in Myanmar: A cross-sectional survey of the K13 molecular marker. Lancet Infect Dis. 2015;15: 415–421. pmid:25704894
  7. 7. Group TPS. Four Artemisinin-Based Treatments in African Pregnant Women with Malaria. http://dx.doi.org/101056/NEJMoa1508606. 2016; pmid:26962727
  8. 8. Sirima SB, Ogutu B, Lusingu JPA, Mtoro A, Mrango Z, Ouedraogo A, et al. Comparison of artesunate–mefloquine and artemether–lumefantrine fixed-dose combinations for treatment of uncomplicated Plasmodium falciparum malaria in children younger than 5 years in sub-Saharan Africa: a randomised, multicentre, phase 4 trial. Lancet Infect Dis. 2016;16: 1123–1133. pmid:27430374
  9. 9. World Health Organization. WHO: Guideline for the treatment of malaria. 2nd ed. Geneva. World Health Organization 2010. Nature. 2010; 284. ISBN 978 92 4 1564403
  10. 10. Eastman RT, Fidock D a. Artemisinin-based combination therapies: a vital tool in efforts to eliminate malaria. Nat Rev Microbiol. Nature Publishing Group; 2009;7: 864–874. Available: http://dx.doi.org/10.1038/nrmicro2239
  11. 11. Petersen I, Eastman R, Lanzer M. Drug-resistant malaria: Molecular mechanisms and implications for public health. FEBS Letters. 2011. pp. 1551–1562. pmid:21530510
  12. 12. Price RN, Uhlemann AC, Brockman A, McGready R, Ashley E, Phaipun L, et al. Mefloquine resistance in Plasmodium falciparum and increased pfmdr1 gene copy number. Lancet. 2004;364: 438–447. pmid:15288742
  13. 13. Holmgren G, Hamrin J, Svärd J, Mårtensson A, Gil JP, Björkman A. Selection of pfmdr1 mutations after amodiaquine monotherapy and amodiaquine plus artemisinin combination therapy in East Africa. Infect Genet Evol. 2007;7: 562–569. pmid:17467344
  14. 14. Sisowath C, Strömberg J, Mårtensson A, Msellem M, Obondo C, Björkman A, et al. In vivo selection of Plasmodium falciparum pfmdr1 86N coding alleles by artemether-lumefantrine (Coartem). The Journal of infectious diseases. 2005. pmid:15717281
  15. 15. Veiga MI, Dhingra SK, Henrich PP, Straimer J, Gna N, Uhlemann A-C, et al. Globally prevalent PfMDR1 mutations modulate Plasmodium falciparum susceptibility to artemisinin-based combination therapies. Nat Commun. 2016; pmid:27189525
  16. 16. Humphreys GS, Merinopoulos I, Ahmed J, Whitty CJM, Mutabingwa TK, Sutherland CJ, et al. Amodiaquine and artemether-lumefantrine select distinct alleles of the Plasmodium falciparum mdr1 gene in Tanzanian children treated for uncomplicated malaria. Antimicrob Agents Chemother. 2007;51: 991–997. pmid:17194834
  17. 17. Dokomajilar C, Nsobya SL, Greenhouse B, Rosenthal PJ, Dorsey G. Selection of Plasmodium falciparum pfmdr1 alleles following therapy with artemether-lumefantrine in an area of Uganda where malaria is highly endemic. Antimicrob Agents Chemother. 2006;50: 1893–1895. pmid:16641472
  18. 18. Baraka V, Tinto H, Valea I, Fitzhenry R, Delgado-Ratto C, Mbonye MKMK, et al. In vivo selection of plasmodium falciparum pfcrt and pfmdr1 variants by artemether-lumefantrine and dihydroartemisinin-piperaquine in Burkina Faso. Antimicrob Agents Chemother. American Society for Microbiology; 2015;59: 734–737. pmid:25403659
  19. 19. Wurtz N, Fall B, Pascual A, Fall M, Baret E, Camara C, et al. Role of Pfmdr1 in In Vitro Plasmodium falciparum Susceptibility to Chloroquine, Quinine, Monodesethylamodiaquine, Mefloquine, Lumefantrine, and Dihydroartemisinin. Antimicrob Agents Chemother. 2014; pmid:25199781
  20. 20. Dama S, Niangaly H, Ouattara A, Sagara I, Sissoko S, Traore OB, et al. Reduced ex vivo susceptibility of Plasmodium falciparum after oral artemether–lumefantrine treatment in Mali. Malar J. BioMed Central; 2017;16: 59. pmid:28148267
  21. 21. Mwai L, Kiara SM, Abdirahman A, Pole L, Rippert A, Diriye A, et al. In vitro activities of piperaquine, lumefantrine, and dihydroartemisinin in Kenyan plasmodium falciparum isolates and polymorphisms in pfcrt and pfmdr1. Antimicrob Agents Chemother. 2009; pmid:19770282
  22. 22. Henriques G, Hallett RL, Beshir KB, Gadalla NB, Johnson RE, Burrow R, et al. Directional selection at the pfmdr1, pfcrt, pfubp1, and pfap2mu loci of Plasmodium falciparum in Kenyan children treated with ACT. J Infect Dis. 2014;210. pmid:24994911
  23. 23. Happi CT, Gbotosho GO, Folarin O a., Akinboye DO, Yusuf BO, Ebong OO, et al. Polymorphisms in Plasmodium falciparum dhfr and dhps genes and age related in vivo sulfadoxine-pyrimethamine resistance in malaria-infected patients from Nigeria. Acta Trop. 2005;95: 183–193. pmid:16023986
  24. 24. Malmberg M, Ferreira PE, Tarning J, Ursing J, Ngasala B, Björkman A, et al. Plasmodium falciparum drug resistance phenotype as assessed by patient antimalarial drug levels and its association with pfmdr1 polymorphisms. J Infect Dis. 2013;207: 842–847. pmid:23225895
  25. 25. Sidhu ABS, Uhlemann A-C, Valderramos SG, Valderramos J-C, Krishna S, Fidock DA. Decreasing pfmdr1 copy number in plasmodium falciparum malaria heightens susceptibility to mefloquine, lumefantrine, halofantrine, quinine, and artemisinin. J Infect Dis. 2006;194: 528–535. pmid:16845638
  26. 26. Venkatesan M, Gadalla NB, Stepniewska K, Dahal P, Nsanzabana C, Moriera C, et al. Polymorphisms in Plasmodium falciparum chloroquine resistance transporter and multidrug resistance 1 genes: Parasite risk factors that affect treatment outcomes for P. falciparum malaria after artemether-lumefantrine and artesunate-amodiaquine. Am J Trop Med Hyg. American Society of Tropical Medicine and Hygiene; 2014;91: 833–843. pmid:25048375
  27. 27. Mavoko HM, Nabasumba C, da Luz RI, Tinto H, D’Alessandro U, Kambugu A, et al. Efficacy and safety of re-treatment with the same artemisinin-based combination treatment (ACT) compared with an alternative ACT and quinine plus clindamycin after failure of first-line recommended ACT (QUINACT): a bicentre, open-label, phase 3, randomise. Lancet Glob Heal. Elsevier; 2017;5: e60–e68.
  28. 28. Muhindo Mavoko H, Nabasumba C, Tinto H, D’Alessandro U, Grobusch MP, Lutumba P, et al. Impact of retreatment with an artemisinin-based combination on malaria incidence and its potential selection of resistant strains: study protocol for a randomized controlled clinical trial. Trials. 2013;14: 307. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3849445&tool=pmcentrez&rendertype=abstract pmid:24059911
  29. 29. Thomsen TT, Ishengoma DS, Mmbando BP, Lusingu JP, Vestergaard LS, Theander TG, et al. Prevalence of single nucleotide polymorphisms in the Plasmodium falciparum multidrug resistance gene (Pfmdr-1) in Korogwe District in Tanzania before and after introduction of artemisinin-based combination therapy. Am J Trop Med Hyg. 2011;85: 979–983. pmid:22144430
  30. 30. Baliraine FN, Rosenthal PJ. Prolonged selection of pfmdr1 polymorphisms after treatment of falciparum malaria with artemether-lumefantrine in Uganda. J Infect Dis. 2011;204: 1120–1124. pmid:21881128
  31. 31. Mobula L, Lilley B, Tshefu AK, Rosenthal PJ. Resistance-mediating polymorphisms in Plasmodium falciparum infections in Kinshasa, Democratic Republic of the Congo. Am J Trop Med Hyg. 2009;80: 555–558. pmid:19346374
  32. 32. Mbaye A, Dieye B, Ndiaye YD, Bei AK, Muna A, Deme AB, et al. Selection of N86F184D1246 haplotype of Pfmrd1 gene by artemether–lumefantrine drug pressure on Plasmodium falciparum populations in Senegal. Malar J. 2016; pmid:27562216
  33. 33. Dorsey G, Gasasira AF, Machekano R, Kamya MR, Staedke SG, Hubbard A. The impact of age, temperature, and parasite density on treatment outcomes from antimalarial clinical trials in Kampala, Uganda. Am J Trop Med Hyg. 2004;
  34. 34. Price R, Simpson J, Nosten F, Luxemburger C, Hkirjaroen L, ter Kuile F, et al. Factors contributing to anemia after uncomplicated falciparum malaria. Am J Trop Med Hyg. 2001;
  35. 35. Eyase FL, Akala HM, Ingasia L, Cheruiyot A, Omondi A, Okudo C, et al. The Role of Pfmdr1 and Pfcrt in Changing Chloroquine, Amodiaquine, Mefloquine and Lumefantrine Susceptibility in Western-Kenya P. falciparum Samples during 2008–2011. PLoS One. 2013;8.
  36. 36. Conrad MD, LeClair N, Arinaitwe E, Wanzira H, Kakuru A, Bigira V, et al. Comparative impacts over 5 years of artemisinin-based combination therapies on Plasmodium falciparum polymorphisms that modulate drug sensitivity in Ugandan children. J Infect Dis. Oxford University Press; 2014;210: 344–353. pmid:24610872
  37. 37. Fröberg G, Ferreira PE, Martensson A, Ali A, Björkman A, Gilb JP. Assessing the cost-benefit effect of a plasmodium falciparum drug resistance mutation on parasite growth in vitro. Antimicrob Agents Chemother. 2013;57: 887–892. pmid:23208719
  38. 38. Sondo P, Derra K, Diallo Nakanabo S, Tarnagda Z, Kazienga A, Zampa O, et al. Artesunate-Amodiaquine and Artemether-Lumefantrine Therapies and Selection of Pfcrt and Pfmdr1 Alleles in Nanoro, Burkina Faso. PLoS One. 2016;11: e0151565. pmid:27031231
  39. 39. Tinto H, Sirima C, Kazienga A, Sombié O, Yara S, Lompo P, et al. Selection of Plasmodium falciparum Pfmdr-1 N86Y alleles by Amodiaquine-Artesunate and Artemether- Lumefantrine in Nanoro, Burkina Faso. 2016;8: 10–14.
  40. 40. Sisowath C, Ferreira PE, Bustamante LY, Dahlström S, Mårtensson A, Björkman A, et al. The role of pfmdr1 in Plasmodium falciparum tolerance to artemether-lumefantrine in Africa. Trop Med Int Heal. 2007;12: 736–742. pmid:17550470
  41. 41. Nsobya SL, Dokomajilar C, Joloba M, Dorsey G, Rosenthal PJ. Resistance-mediating Plasmodium falciparum pfcrt and pfmdr1 alleles after treatment with artesunate-amodiaquine in Uganda. Antimicrob Agents Chemother. 2007;51: 3023–3025. pmid:17562798
  42. 42. Antonia AL, Taylor SM, Janko M, Emch M, Tshefu AK, Meshnick SR. Short report: A cross-sectional survey of Plasmodium falciparum pfcrt mutant haplotypes in the Democratic Republic of Congo. Am J Trop Med Hyg. 2014; pmid:24732459
  43. 43. Borrmann S, Sasi P, Mwai L, Bashraheil M, Abdallah A, Muriithi S, et al. Declining responsiveness of plasmodium falciparum infections to Artemisinin-Based combination treatments on the Kenyan coast. PLoS One. 2011;6.
  44. 44. Sutherland CJ, Lansdell P, Sanders M, Muwanguzi J, Van Schalkwyk DA, Kaur H, et al. pfk13-independent treatment failure in four imported cases of Plasmodium falciparum malaria treated with artemether-lumefantrine in the United Kingdom. Antimicrob Agents Chemother. 2017; pmid:28137810
  45. 45. Ariey FF, Witkowski B, Amaratunga C, Beghain J, Langlois A-CC, Khim N, et al. A molecular marker of artemisinin-resistant Plasmodium falciparum malaria. Nature. Nature Publishing Group, a division of Macmillan Publishers Limited. All Rights Reserved.; 2014;505: 50–55. http://dx.doi.org/10.1038/nature12876
  46. 46. Collet L, Cui L, Thakur G-D, Dieye A, Djallé D, Dorkenoo MA, et al. A Worldwide Map of Plasmodium falciparum K13-Propeller Polymorphisms. N Engl J Med. 2016; pmid:27332904
  47. 47. Amato R, Miotto O, Woodrow CJ, Almagro-Garcia J, Sinha I, Campino S, et al. Genomic epidemiology of artemisinin resistant malaria. Elife. 2016;5: 1–29. pmid:26943619
  48. 48. Spring MD, Lin JT, Manning JE, Vanachayangkul P, Somethy S, Bun R, et al. Dihydroartemisinin-piperaquine failure associated with a triple mutant including kelch13 C580Y in Cambodia: An observational cohort study. Lancet Infect Dis. Lancet Publishing Group; 2015;15: 683–691.
  49. 49. Boni MF, Smith DL, Laxminarayan R. Benefits of using multiple first-line therapies against malaria. Proc Natl Acad Sci. 2008; pmid:18780786
  50. 50. Nguyen TD, Olliaro P, Dondorp AM, Baird JK, Lam HM, Farrar J, et al. Optimum population-level use of artemisinin combination therapies: A modelling study. Lancet Glob Heal. 2015; pmid:26545449
  51. 51. Boni MF, White NJ, Baird JK. The Community As the Patient in Malaria-Endemic Areas: Preempting Drug Resistance with Multiple First-Line Therapies. PLoS Med. 2016; pmid:27022739
  52. 52. Yeka A, Kigozi R, Conrad MD, Lugemwa M, Okui P, Katureebe C, et al. Artesunate/amodiaquine versus artemether/lumefantrine for the treatment of uncomplicated malaria in Uganda: A randomized trial. J Infect Dis. 2016; pmid:26597254