|Home | About | Journals | Submit | Contact Us | Français|
Dihydroartemisinin-piperaquine (DP) offers prolonged protection against malaria, but its impact on Plasmodium falciparum drug sensitivity is uncertain. In a trial of intermittent preventive treatment in schoolchildren in Tororo, Uganda, in 2011 to 2012, monthly DP for 1 year decreased the incidence of malaria by 96% compared to placebo; DP once per school term offered protection primarily during the first month after therapy. To assess the impact of DP on selection of drug resistance, we compared the prevalence of key polymorphisms in isolates that emerged at different intervals after treatment with DP. Blood obtained monthly and at each episode of fever was assessed for P. falciparum parasitemia by microscopy. Samples from 160 symptomatic and 650 asymptomatic episodes of parasitemia were assessed at 4 loci (N86Y, Y184F, and D1246Y in pfmdr1 and K76T in pfcrt) that modulate sensitivity to aminoquinoline antimalarials, utilizing a ligase detection reaction-fluorescent microsphere assay. For pfmdr1 N86Y and pfcrt K76T, but not the other studied polymorphisms, the prevalences of mutant genotypes were significantly greater in children who had received DP within the past 30 days than in those not treated within 60 days (86Y, 18.0% versus 8.3% [P = 0.03]; 76T, 96.0% versus 86.1% [P = 0.05]), suggesting selective pressure of DP. Full sequencing of pfcrt in a subset of samples did not identify additional polymorphisms selected by DP. In summary, parasites that emerged soon after treatment with DP were more likely than parasites not under drug pressure to harbor pfmdr1 and pfcrt polymorphisms associated with decreased sensitivity to aminoquinoline antimalarials. (This study has been registered at ClinicalTrials.gov under no. NCT01231880.)
Malaria, in particular infection with Plasmodium falciparum, remains a huge public health problem, with the highest disease burden in sub-Saharan Africa (1, 2). Important advances have been made in malaria control recently, with a significant decrease in malaria burden and progress toward elimination noted in some areas (3). Among key tools in the control of malaria is intermittent preventive treatment (IPT), the provision of full treatment courses at regular intervals to high-risk populations (4). IPT is standard practice during pregnancy (IPTp), is recommended for children living in seasonal malaria transmission settings as seasonal malaria chemoprevention (5), and is being investigated in other populations (6,–9). However, currently IPT is advocated only with sulfadoxine-pyrimethamine (SP) or a combination of SP and amodiaquine (SP-AQ) (5, 10), regimens that are severely compromised by drug resistance in much of Africa (11,–13). For malaria treatment, older regimens have been replaced by artemisinin-based combination therapies (ACTs), and a similar change may be warranted for IPT.
Dihydroartemisinin-piperaquine (DP), which provides rapid killing of most parasites by dihydroartemisinin, prolonged action against any remaining parasites by piperaquine, and protection for weeks after therapy due to the long half-life of piperaquine, has recently been investigated for IPT. Compared to IPTp with SP, IPTp with DP was associated with lower risks of P. falciparum infection and symptomatic malaria during pregnancy in Kenya (14) and Uganda (15). In Ugandan schoolchildren, monthly IPT with DP was associated with a reduced incidence of malaria and reduced prevalence of parasitemia and anemia compared to DP given approximately once every 3 months or placebo (6, 16). Similar results were observed in Ugandan infants when monthly IPT with DP was compared with daily trimethoprim-sulfamethoxazole or monthly SP (7). Thus, DP is a promising alternative to SP or SP-AQ for IPT, but its benefits may be undone by the emergence of P. falciparum resistance to either component of the combination.
Mediators of decreased drug sensitivity and selective pressures for resistance are quite well understood for some antimalarial drugs. Resistance to the aminoquinolines chloroquine and amodiaquine is mediated largely by polymorphisms in putative drug transporters encoded by pfcrt and pfmdr1 (13, 17), and these polymorphisms are selected in new infections that emerge soon after therapy with artesunate-AQ (AS-AQ) (18, 19). Piperaquine is a bisaminoquinoline related to chloroquine and amodiaquine. Resistance to piperaquine was widely reported during the preartemisinin era in China (20), and recently clinically relevant resistance, with frequent recrudescences after therapy with DP, has been noted in Cambodia (21,–23). However, mechanisms of resistance to piperaquine are uncertain. Use of DP for treatment (24) or chemoprevention (25) did not select for the polymorphisms associated with chloroquine resistance in Burkina Faso, but in Uganda recent treatment with DP selected for pfmdr1 mutations associated with decreased sensitivity to aminoquinolines (26). Interestingly, some other antimalarials, notably lumefantrine, which is a component of the Ugandan first-line antimalarial regimen artemether-lumefantrine (AL), exert the opposite selective pressure. Thus, new infections emerging within 2 months of treatment with AL showed selection of wild-type sequences at the pfcrt K76T and pfmdr1 N86Y and D1246Y alleles (26,–29); mutant sequences are selected at these same alleles by aminoquinolines. Of recent concern has been resistance to artemisinins, manifest as delayed parasite clearance after therapy, in Southeast Asia (22, 30,–32), but recent studies utilizing clinical, parasitological, and molecular markers (33, 34) suggest that the artemisinin-resistant phenotype is not yet prevalent in Uganda (26, 35, 36) or other parts of Africa (37, 38).
Taken together, the available evidence suggests that DP may select for the same P. falciparum polymorphisms as other aminoquinolines, leading to decreased treatment or preventive efficacy of DP, but data on the effects of IPT with DP are very limited. We therefore assessed the prevalences of key polymorphisms in isolates that emerged at different intervals after treatment with DP using samples from a recent trial evaluating IPT with DP in Ugandan schoolchildren.
Study samples were from a randomized, double-blinded, placebo-controlled trial conducted in Tororo, Uganda, from 2011 to 2012 (6, 39). In brief, 740 schoolchildren aged 6 to 14 years from one primary school in Mulanda subcounty, Tororo District, were enrolled and randomized 1:1:1 to one of three study arms: DP monthly, DP once per school term (four treatments over 12 months), or placebo. DP was administered according to weight-based guidelines, and treatment was directly observed. Finger-prick blood samples were obtained at enrollment, every month, and with every episode of fever to assess for malaria infection by thick blood smear and for storage on filter paper. Episodes of uncomplicated malaria were treated with AL. Children were followed for 12 months. The trial was approved by the Uganda National Council for Science and Technology and the Makerere University School of Medicine Research and Ethics Committee and registered at ClinicalTrials.gov (NCT01231880). Molecular studies were also approved by the University of California, San Francisco, Committee on Human Research.
We considered all samples that were positive for P. falciparum parasitemia based on evaluation of Giemsa-stained thick blood smears, as previously described (6). A total of 160 symptomatic and 1,522 asymptomatic episodes of P. falciparum parasitemia were documented. The number of samples analyzed was determined by estimating the power for two-sample comparison of proportions using effect sizes observed for each mutant polymorphism in a recent study in Tororo (0.34 for pfmdr1 N86Y, 0.11 for pfmdr1 D1246Y, 0.04 for pfmdr1 184F, and 0.09 for pfcrt K76), fixing α at 0.05 (26). The sample size giving the maximum power was considered in the analysis. From these estimates, we analyzed all 160 samples from symptomatic episodes, all 50 samples from children with recurrent parasitemia within 13 to 30 days of prior therapy with DP, and 600 samples randomly selected from children with either recurrent parasitemia >30 days after prior therapy with DP or from the control arm of the study. All samples were analyzed for 4 common P. falciparum polymorphisms known to be associated with drug sensitivity: pfcrt K76T, and pfmdr1 N86Y, Y184F, and D1246Y. A subset of 25 samples from children with prior DP therapy within 13 to 30 days and 25 randomly selected paired samples from children in the control arm (each pair matched for collection within 15 days of each other) were subjected to sequencing of the complete pfcrt gene.
DNA was extracted from filter paper blood spots into 100 μl of water using Chelex-100 as previously described (40). Gene fragments spanning all loci of interest were amplified in nested reactions (26), and failed reactions were repeated. To detect polymorphisms, multiplex ligase detection reaction-fluorescent microsphere assays were performed as previously described (26, 41).
For a subset of samples, pfcrt was sequenced from DNA samples as previously described (42) with minor modifications. Briefly, pfcrt was amplified in 3 nested PCRs, covering exons 1 and 2, 3 to 8, and 9 to 13, using the published primer sequences. For both rounds of PCR, each 25-μl reaction mixture contained 2 mM MgSO4, 200 μM each deoxynucleoside triphosphate (dNTP), 1 μM each primer, 1× PCR buffer, and 2 U Platinum Taq high-fidelity DNA polymerase (Invitrogen). Conditions for all reactions were 94°C for 2 min, 30 cycles of 94°C for 20 s, 47°C for 10 s, and 60°C for 3 min, and a final extension at 60°C for 5 min. Amplicons were cloned with the TOPO-TA cloning kit for sequencing and transfected into One Shot TOP10 chemically competent Escherichia coli (Invitrogen) according to the manufacturer's instructions. Colonies were grown overnight under kanamycin selection, picked, and incubated in LB broth with kanamycin. Plasmid DNA was purified using the PureLink quick plasmid miniprep kit (Invitrogen), digested with EcoRI to confirm the insert size, and then sequenced (Eurofins) using M13 forward and reverse primers. DNA sequence data were assembled and edited, and mutations were detected by alignment and comparison to the expected sequence using CodonCode Aligner v. 5.1.5. Multiple clones were sequenced to distinguish true polymorphisms from PCR errors, including at least 3 clones for all but 3 fragments, for which 2 clones were sequenced.
Data analysis was done using Stata version 14 (StataCorp). Outcomes of interest were the prevalence of pure mutant alleles for each locus of interest. The exposure variable of interest was duration since prior DP dose, evaluated as a categorical variable split into 13 to 30, 31 to 60, and >60 days (including the no-treatment control group) since the last treatment. Associations between outcomes and duration since last treatment and differences between prevalences of pfcrt alleles were measured using Fisher's exact test and expressed as relative risk. In all analyses, a 2-tailed P value of <0.05 was considered statistically significant.
A total of 740 schoolchildren aged 6 to 14 years were randomized to one of the three study arms in the parent study and followed for 1 year from 2011 to 2012. As previously reported, compared to either DP once per school term (approximately every 3 months) or placebo, monthly DP offered strong protective efficacy against malaria (6). For this substudy, samples collected from children with blood smears positive for P. falciparum were analyzed (Table 1). As expected due to the protective efficacy of monthly DP, fewer samples were available from this study arm than from children who received placebo or DP once per school term. A total of 810 samples from 160 symptomatic and 650 asymptomatic episodes of parasitemia were assessed (Table 1). Samples were analyzed for common polymorphisms in pfmdr1 and pfcrt. Genotyping results were available for pfcrt K76T in 806 (99.5%) samples and for pfmdr1 N86Y, N184Y, and D1246Y in 800 (98.8%), 810 (100%), and 784 (96.8%) samples, respectively, and these results were included in the analysis.
The prevalence of the 4 studied polymorphisms was similar to that in contemporaneous samples from Tororo that were reported previously (43). For two polymorphisms, pfcrt K76T and pfmdr1 N86Y, the prevalence of mutant genotypes was significantly higher in samples from children who had received DP within 30 days than in those from children who had not received DP within 60 days (Table 2). For the other studied polymorphisms, the prevalence of genotypes did not differ between children who had or had not received recent therapy with DP. Matching for duration since a prior episode, there was no difference in the prevalence of pfcrt and pfmdr1 mutant alleles between samples from children with symptomatic or asymptomatic parasitemia (data not shown).
As DP may select for additional polymorphisms in pfcrt, we sequenced the gene in a subset of 25 parasitemic samples under strong selective pressure as indicated by emergence within 30 days of prior therapy with DP and in 25 paired samples collected near the same date from children who did not receive DP. We successfully sequenced the full gene in 17 pairs. We identified 9 polymorphisms, 6 of which are commonly reported in African isolates (see Table S1 in the supplemental material). All isolates had the pfcrt 72-76 CVIET or a mix of the CVIET and CVMNT haplotype, except for one isolate that had the pfcrt 72S mutation, resulting in the SVIET haplotype (in all 6 clones from a patient not receiving DP). Two additional polymorphisms, L50P and F112I, were each identified in at least 2 clones from a single isolate, the 50P mutation in a control isolate and the 112I mutation in an isolate from a child recently treated with DP (see Table S2 in the supplemental material). We found 9 pfcrt haplotypes; the majority (76% in the DP arm and 65% in the control arm) were mutant at the six loci that are commonly mutant in Africa (74I, 75E, 76T, 220S, 271E, and 371I) (17). Overall, we saw no evidence that DP selected for novel pfcrt polymorphisms in Ugandan children.
Monthly IPT with DP was highly efficacious in reducing the risks of symptomatic malaria, parasitemia, and anemia in Ugandan schoolchildren (6). However, the chemoprophylactic benefits of a long-acting antimalarial such as piperaquine may be accompanied by selection of drug-resistant parasites (13). We tested whether DP selected for parasites with genotypes associated with altered sensitivity to aminoquinolines. Compared to parasites not under drug pressure, those that emerged within 30 days of IPT with DP were more likely to harbor two mutations, pfmdr1 86Y and pfcrt 76T; these mutations are associated with resistance to chloroquine and amodiaquine (36, 43,–45). Thus, the marked preventive efficacy of IPT with DP may be accompanied by selection of decreased sensitivity to aminoquinolines.
Resistance to chloroquine and amodiaquine is mediated primarily by polymorphisms in putative drug transporters encoded by pfcrt and pfmdr1 (13, 17). The pfcrt 76T and pfmdr1 86Y and 1246Y mutations are selected in new infections that emerge soon after therapy with regimens including chloroquine or amodiaquine (46). Piperaquine is a related bisaminoquinoline, but mechanisms of resistance are uncertain, and studies of the selective pressure exerted by DP have yielded conflicting results. Specifically, use of DP for treatment (24) or chemoprevention (25) did not select for the polymorphisms associated with aminoquinoline resistance in Burkina Faso, but in Uganda, recent treatment with DP selected for the pfmdr1 86Y and 1246Y mutations (26). Our new results shed additional light on this area. In the setting of IPT in schoolchildren, recent receipt of DP was associated with selection of the pfmdr1 86Y and pfcrt 76T mutations but not the pfmdr1 1246Y mutation. The differing results may have been due to the changing baseline of polymorphism prevalence in Uganda, with decreasing prevalence of pfmdr1 1246Y and pfcrt 76T over time. Differences in results between West and East Africa may also be explained by differences in parasite backgrounds; of note, the pfmdr1 1246Y mutation, which until recently was widespread in Uganda, has consistently been uncommon in Burkina Faso (24, 25, 28).
Importantly, although we lack a head-to-head comparison, it appears that DP does not select as readily as other ACTs for key transporter mutations. In multiple studies, the selective pressure of AS/AQ was marked (47), including a recent trial that showed the prevalence of the pure pfmdr1 86Y mutation to rise from 59% at baseline to 99% in recurrent infections within 1 month of treatment (48). AL also exerts strong selective pressure, but in the opposite direction, with selection of wild-type pfcrt K76 and pfmdr1 N86 and N1246 sequences in parasites that emerge soon after therapy (19, 29). Our recent findings indicate that DP selects for resistance in a manner similar to that of the other aminoquinolines, but associations between recent therapy and transporter polymorphisms were less marked, suggesting that the selective pressure of DP is lower than that of other regimens. This difference might be due to different mechanisms of transport for piperaquine, a much larger molecule than chloroquine or amodiaquine.
We were concerned that IPT with DP might select for additional resistance-mediating P. falciparum polymorphisms. Polymorphisms in addition to those commonly described in African isolates have been identified in other regions, in some cases with biochemical and clinical consequences (49, 50). Sequencing of pfcrt in a subset of samples either under or not under the selective pressure of DP identified a few previously unidentified pfcrt mutations, but it did not suggest that additional polymorphisms were selected by DP.
Our results have important implications for the use of DP for IPT. Although it offers great promise for decreasing the malaria burden, DP use may be accompanied by selection of parasites with decreased sensitivity to DP and also to the related ACT AS-AQ. Consideration of the opposite resistance pressures of different antimalarials has led some to recommend multiple or rotating first-line antimalarial regimens (51). For example, AS-AQ and AL have opposite selective pressures on pfcrt and pfmdr1 such that each regimen should blunt selection of resistance to the other. Our results are consistent with a prior study in Uganda indicating that DP has selective pressure similar to that of AS-AQ. Thus, considering resistance selection, using DP in IPT might be best advised when the standard treatment regimen is AL, such that the treatment and IPT regimens offer mutual protection against selection of resistance. Further, our results suggest that, with changing treatment and control practices, continued surveillance for clinical, biochemical, and molecular markers of antimalarial drug resistance in Africa is an important priority.
We thank the participants in the clinical trials from whom the samples were collected, their parents and guardians, and our clinical study team, including Pauline Mary Amuge Kayiga, Sarah Nabwire, Victor Asua, Immaculate Mary Nandera, Twaha Yiga, Godfrey Odoi, and Daniel James Owino.
This work was supported by the Malaria Capacity Development Consortium (WT084289MA), the Bill & Melinda Gates Foundation (51941), and grants from the National Institutes of Health (AI075045 and AI089674). Bonnie Wandera is a THRiVE fellow supported by the Wellcome Trust (087540), and Simon J Brooker was supported by a Wellcome Trust Senior Fellowship in Basic Biomedical Sciences (098045).
All authors report no conflicts of interest.
This work, including the efforts of Joaniter I. Nankabirwa, was funded by Malaria Capacity Development Consortium through a grant from Bill & Melinda Gates Foundation (51941) and a grant from Wellcome Trust (WT084289MA). This work, including the efforts of Bonnie Wandera, was funded by THRiVE through a grant from Wellcome Trust (087540). Part of the lab work for this study was funded by HHS | National Institutes of Health (NIH) (AI075045 and AI089674). This work, including the efforts of Simon J. Brooker, was funded by Wellcome Trust (098045).
Supplemental material for this article may be found at http://dx.doi.org/10.1128/AAC.00920-16.