Significant geographical differences in prevalence of mutations associated with Plasmodium falciparum and Plasmodium vivax drug resistance in two regions from Papua New Guinea

BACKGROUND: Drug resistance remains a major obstacle to malaria treatment and control. It can arise and spread rapidly, and vary substantially even at sub-national level. National malaria programmes require cost-effective and timely ways of characterizing drug-resistance at multiple sites within their countries. METHODS: An improved multiplexed post-PCR ligase detection reaction-fluorescent microsphere assay (LDR-FMA) was used to simultaneously determine the presence of mutations in chloroquine resistance transporter (crt), multidrug resistance 1 (mdr1), dihydrofolate reductase (dhfr) and dihydropteroate synthase (dhps) genes in Plasmodium falciparum (n = 727) and Plasmodium vivax (n = 574) isolates collected in 2006 from cross-sectional community population surveys in two geographically distinct regions (Madang and East Sepik) of Papua New Guinea (PNG) where strong regional differences in in vivo aminoquinoline and antifolate therapeutic efficacy had previously been observed. Data were compared to those of a follow-up survey conducted in 2010. RESULTS: Despite some very low parasite densities, the assay successfully amplified all P. falciparum and P. vivax loci in 77 and 69 % of samples, respectively. In 2006, prevalences of pfdhfr (59R-108 N) double mutation/wild type pfdhps haplotype, pfcrt SVMNT haplotype (72S-76T double mutation), and 86Y pfmdr1 mutation all exceeded 90 %. For P. vivax, 65 % carried at least two pvdhfr mutations, 97 % the 647P pvdhps mutation and 54 % the 976F pvmdr1 mutation. Prevalence of mutant haplotypes was higher in Madang than East Sepik for pfcrt SVMNT (97.4 vs 83.3 %, p = 0.001), pfdhfr (59R-108 N) (100 vs 90.6 %, p = 0.001), pvdhfr haplotypes (75.8 vs 47.6 %, p = 0.001) and pvmdr1 976F (71.2 vs 26.2 %, p < 0.001). Data from a subsequent Madang survey in 2010 showed that the prevalence of pfdhps mutations increased significantly from 30 % (p < 0.001) as did the prevalence of pvdhfr mutant haplotypes (from 75.8 to 97.4 %, p = 0.012). CONCLUSIONS: This LDR-FMA multiplex platform shows feasibility for low-cost, high-throughput, rapid characterization of a broad range of drug-resistance markers in low parasitaemia infections. Significant geographical differences in mutation prevalence correlate with previous genotyping surveys and in vivo trials and may reflect variable drug pressure and differences in health-care access in these two PNG populations.