Plasmodium falciparum and Plasmodium vivax Genotypes and Efficacy of Intermittent Preventive Treatment in Papua New Guinea

Abstract

Intermittent preventive treatment of infants (IPTi) reduces early childhood malaria-related morbidity. While genotypic drug resistance markers have proven useful in predicting the efficacy of antimalarial drugs in case management, there are few equivalent data relating to their protective efficacy when used as IPTi. The present data from an IPTi trial in Papua New Guinea demonstrate how these markers can predict protective efficacy of IPTi for both Plasmodium falciparum and Plasmodium vivax.

TEXT

Intermittent preventive treatment of infants (IPTi) has reduced early childhood malaria-related morbidity in many settings (1, 2). Although its precise mechanism of action has been debated, it is now generally agreed that the intermittent short courses of slowly eliminated antimalarial drugs provide protection due to prolonged prophylactic activity (2). Therefore, the pharmacokinetic and pharmacodynamic properties that determine a drug's suitability for IPTi may differ from those important for therapeutic efficacy in case management.

Genotypic drug resistance markers correlate well with in vivo therapeutic response in the case management of acute infections. Mutations of Plasmodium falciparum and Plasmodium vivax dihydrofolate reductase (pfdhfr and pvdhfr) and dihydropteroate synthase (pfdhps and pvdhps) genes are associated with resistance to pyrimethamine and sulfadoxine, respectively (3–6). Parasite resistance to 4-aminoquinolines is associated with mutations in the chloroquine resistance transporter gene (pfcrt) (7) and multidrug resistance gene 1 (pfmdr1) in P. falciparum infections and with P. vivax multidrug resistance gene 1 (pvmdr1) mutations in some studies of vivax malaria (8, 9) but not others (10).

IPTi effectiveness is likely to be highly dependent on drug resistance patterns in local parasite populations. Current IPTi regimens utilize either 4-aminoquinoline or antifolate drugs (2), and so existing genotypic resistance markers should predict protective efficacy in a given setting without the need for large expensive field-based efficacy studies. However, data directly relating drug resistance genotypes to IPTi efficacy is limited to African IPTi trials of sulfadoxine-pyrimethamine (SP) dual therapy, which have demonstrated that the pfdhfr triple mutant genotype predicts reduced efficacy (11, 12). No studies have examined the relationship between IPTi efficacy and genotypic markers in P. vivax infections nor between 4-aminoquinoline resistance markers and IPTi efficacy in falciparum malaria. In addition, there are concerns that the high burden of antimalarial drug use through IPTi will drive the selection and spread of parasite drug resistance, with a consequent reduction in public health benefit (13). Serial monitoring of validated genotypic markers could help identify this phenomenon at an early stage (12).

In Papua New Guinea (PNG), an IPTi study of 1,121 infants conducted from 2006 to 2010 showed that amodiaquine plus SP (AQ-SP) reduced P. falciparum and P. vivax incidence by 35% and 23%, respectively, compared with placebo (14). Artesunate-SP (ART-SP) had a similar efficacy (31%) against P. falciparum but none against P. vivax infections. Protective efficacy of both SP-containing regimens therefore fell in the midrange of comparative African trials of IPTi using SP (median protective efficacy, 30.3%; range, 20 to 59%) (1), suggesting a likelihood of highly cost-effective public health benefit. However, protective efficacy against P. vivax was dependent on the partner drug used alongside SP. We therefore hypothesized that therapy-specific efficacy in P. falciparum and P. vivax infections reflects both pharmacokinetic differences between ART and AQ and intrinsic differences in the sensitivities of P. falciparum and P. vivax parasites to SP. In the present study, we investigated whether the frequency of mutations associated with 4-aminoquinoline and SP resistance could explain differences in IPTi efficacy of AQ-SP and ART-SP and explored the impact of different IPTi regimens on resistance marker prevalence.

The IPTi trial study sample and clinical methods have been described in detail elsewhere (14). All participants were enrolled at age 3 months and had blood samples collected at 3, 6, 9, and 12 months after enrollment (i.e., passive detection in children aged 6 to 15 months) in each study arm (AQ-SP, ART-SP, and control). The study was approved by the PNG Medical Research Advisory Committee (approval no. 05.20) and the PNG Institute of Medical Research Institutional Review Board (approval no. 06.01).

Drug resistance genotyping was performed using a ligase detection reaction-fluorescent microsphere assay (LDR-FMA). Briefly, fragments of pfcrt, pfmdr1, pfdhfr, and pfdhps and fragments of pvmdr1, pvdhfr, and pvdhps were amplified from P. falciparum- and P. vivax-positive samples, respectively. LDR was performed using allele-specific primers as previously described (15–17). Products were run through a Bio-Plex 200 reader, and fluorescence signals were analyzed to determine which alleles were present in each sample. A positivity cutoff was determined for each locus by the use of methodology developed previously to account for an allele-specific background (15, 18). Positive controls (genomic DNA [gDNA] from P. falciparum reference strains or P. vivax mdr1, dhfr, and dhps plasmids) displaying wild -type and mutant alleles were run in each assay. Statistical analyses were performed using R (BiostaTGV) to compare genotypic frequencies in each treatment group, using chi-square and Fisher's exact tests.

The proportions of individuals with infections with wild-type and mutant genotypes of the different parasite genes are shown in Fig. 1. The pfcrt 76T and pfmdr 86Y mutations appeared close to fixation in P. falciparum, and the proportion of P. vivax infections with the pvmdr1 976F mutation was 73.8%, suggesting compromised AQ efficacy against P. falciparum (17, 19, 20) and possibly (given the currently uncertain functional significance of pvmdr1 [8–10]) against P. vivax.

FIG 1.

Proportions of infections with wild-type and mutant genotypes in all P. falciparum and P. vivax infections.

While 98.9% of P. falciparum-infected individuals carried at least one pfdhfr mutation, only mutations at codons 59 (C→R) and 108 (S→N) were identified. The 51I mutant was not detected, so none had the pfdhfr triple mutation (51I-59R-108N) nor the pfdhfr/pfdhps quintuple mutant haplotype associated with high-level SP resistance in Africa (20). Similarly, the pfdhfr I164L mutation which is associated with high-level SP resistance in Africa, even when present as a single mutation (21), was not detected. Mutations in pfdhps (codons 437, 540, and 581) were present in 29.8% of the infected individuals, with 17.5% and 11.7% of monoclonal infections having pfdhfr or pfdhps triple and quadruple mutant haplotypes, respectively (see Table S1 in the supplemental material).

For P. vivax, 57.3% of the individuals with monoclonal infections showed the quadruple mutant genotype (57L-58R-61M-117T) associated with very high rates of in vivo SP treatment failure (22, 23) (see Table S1 in the supplemental material). Only 3.0% of all P. vivax infections showed a mutation at pvdhps codons 383 and/or 553.

Overall parasite genotype patterns suggested some retention of SP efficacy against P. falciparum, given the absence of the quintuple mutant haplotype and pfdhfr I164L mutant. The high prevalence of the pvdhfr quadruple mutant genotype suggested poor or absent SP efficacy against P. vivax. Any AQ-SP efficacy against P. falciparum would therefore be due largely to the SP component, and conversely, any efficacy against P. vivax would result from the use of AQ. For ART-SP, given the negligible posttreatment prophylactic duration of ART, genotypic prevalence predicts retention of some SP efficacy against P. falciparum but virtually none against P. vivax, consistent with the differential protective efficacies against the two species demonstrated in the IPTi trial (14). The present genotypic data are therefore concordant with the observed clinical outcomes (14), explaining both the similar protective efficacies of AQ-SP and ART-SP against P. falciparum and the discordant efficacies of AQ-SP and ART-SP against P. vivax.

There were no differences between P. falciparum mutant genotype frequencies in the control AQ-SP and ART-SP study arms that might have suggested selection of more resistant strains (Table 1). However, the low population diversity shown by the dominance of one or two pfdhfr, pfdhps, pfcrt, and pfmdr1 alleles means that few of the genes evaluated would be under selection pressure. Therefore, we cannot exclude selection of resistance in other contexts, with different balances of parasite genotypes. Indeed, selection of P. falciparum drug-resistant mutants has previously been shown following single- or three-dose SP IPTi in Africa (13, 24). No between-group differences were seen for pvdhfr or pvdhps mutations (Table 1).

TABLE 1.

Proportions of individuals infected with wild-type and mutant P. falciparum or P. vivax genotypes^a

Sample type and gene	Codon(s)	Amino acid(s)b	% of individuals infected with mutant genotype from the indicated study arm			P valued
			Placebo	AQ-SP	ART-SP
P. falciparum
			n = 70	n = 59	n = 49
pfdhfr	51	Nc	100	100	100	NS
	59	C/R	1.4/98.6	1.7/98.3	0/100	NS
	108	S/N	1.4/98.6	1.7/98.3	0/100	NS
	164	Ic	100	100	100	NS
			n = 63	n = 51	n = 47
pfdhps	436–437	SA/SG	69.8/30.2	64.7/37.3	76.6/25.5	NS
	540	K/E	87.3/12.7	86.3/13.7	91.5/10.6	NS
	581	A/G	100/0	100/2.0	100/0	NS/NS
	613	A	100	100	100	NS
			n = 66	n = 55	n = 44
pfcrt	72–76	CVMNK/SVMNT/CVIET	1.5/98.5/0	3.6/96.4/1.8	4.5/100/0	NS
			n = 59	n = 44	n = 39
pfmdr1	86	N/Y	10.2/89.8	9.1/90.9	5.1/94.9	NS
	184	Y/F	93.2/6.8	100/0	100/0	NS
	1034	Sc	100	100	100	NS
	1042	N/D	96.6/3.4	100/0	100/0	NS
	1246	Dc	100	100	100	NS
P. vivax
			n = 177	n = 167	n = 205
pvdhfr	57/58/61	FST/FRT/LRT/LRM	15.3/11.3/41.8/66.1	18.0/7.2/47.9/60.5	16.6/7.8/46.3/56.1	NS
	117	S/N/T	51.4/9.6/74.6	55.7/6.0/68.9	52.2/6.8/66.3	NS
	173	Ic	100	100	100	NS
			n = 170	n = 160	n = 195
pvdhps	382–383	SC/SG/CC	98.8/3.5/0	98.1/1.9/0.6	98.5/3.1/0	NS
	553	A/G	100/0	100/0.6	100/0	NS/NS
	647	A/S/P	8.8/1.2/93.5	5.0/0/96.9	6.2/0.5/97.9	NS
			n = 173	n = 170	n = 211
pvmdr1	976	Y/F	46.2/74.0	54.1/72.9	46.0/74.4	NS

^an values represent the numbers of individuals in each group.

^bBoldface indicates mutant amino acids.

^cpfdhfr 51N and 164L, pfdhps 436–437FG and 613S, pfmdr1 1034C and 1246Y, and pvdhfr 173L genotypes were not observed.

^dNS, not significant.

We have demonstrated concordance of P. falciparum and P. vivax drug resistance genotypic prevalence with the protective efficacy of two IPTi drug combinations. Drug resistance genotyping is a valuable public health tool for monitoring antimalarial drug resistance and the early detection of compromised therapeutic efficacy in case management. Analysis of population prevalence of genotypic markers could similarly help predict the protective efficacy of IPTi drug regimens in a wide variety of settings where malaria is endemic. It is clear from the present and previous studies that the efficacy of IPTi with SP is highly dependent on prevalence of SP resistance at a population level (2, 11, 12). However, the complex genetics of antifolate drug resistance makes it challenging to establish clear, simple guidelines to help policy makers decide on the appropriateness or other aspects of using SP for IPTi in their national programs.

To be useful in practice, such guidelines require a clear cutoff to recommend its use (2). Current WHO guidelines use a threshold of pfdhps codon 540 mutant prevalence of ≤50% to endorse SP use in IPTi in areas of moderate to high P. falciparum transmission (25). These recommendations would therefore have endorsed the use of SP in IPTi in our PNG setting, and this is supported by the results of the clinical component of this study. However, more sophisticated guidelines will be needed for situations in which 4-aminoquinolines are considered for IPTi in settings in which the pfdhfr I164L mutation is prevalent and for settings outside Africa where P. vivax is endemic. In order to enhance guideline development, it is important, therefore, that drug resistance markers be considered a routine assessment in future IPTi efficacy evaluations.