We commenced an epidemiological study related to malaria drug resistance in 2002 in East Sepik, PNG [29, 30]. After a long interval from 2004 to 2015, the study was restarted in 2016 [14]. In 2017, we identified three pfk13 C580Y mutants [13]. This raised concern that ART-resistant P. falciparum parasites may have already emerged and spread in the study area. In this study, the prevalence of pfk13 C580Y was relatively low, but it had increased significantly since the first detection. In general, drug-resistant malaria increases slowly at the beginning of an emergence, but increases rapidly as the frequency of resistant parasites increases. Indeed, C580Y frequencies gradually increased approximately five years from the initial detection, but then rapidly expanded and even overtook the other pfk13 alleles in Cambodia and Western Thailand [5, 9, 10]. Therefore, even though the current prevalence of pfk13 C580Y is low in PNG, rapid expansion in the near future could be anticipated.
On the other hand, regional malaria epidemiological factors in PNG may suppress the rapid increase in ART-resistant parasites. First, residents in our study area developed higher levels of herd immunity to malaria than those in the GMSR because of higher malaria transmission intensity in PNG [34, 35]. This can considerably influence the clearance of ART-resistant parasites from human hosts [36] and may slow the rate of increase in the C580Y allele in the region. Second, because ART is primarily used in ACT, the presence of resistant parasites to ACT partner drugs significant affects the diffusion rate of C580Y. In the GMSR, parasites resistant to partner drug(s), mefloquine, and piperaquine have already emerged and spread [37, 38]. In particular, parasites harboring both pfk13 C580Y and plasmepsin 2/3 copy number variants, which are molecular markers for piperaquine resistance, have rapidly increased in West Cambodia [5, 39]. In contrast, in PNG there is no evidence that parasites are resistant to lumefantrine, the currently used partner drug of ACT. Our previous ex vivo drug susceptibility study from 2016 to 2018 also demonstrated that the average IC50 to lumefantrine was 4.6 nM and no parasite fulfilled the criteria of ex vivo lumefantrine resistance [14]. Furthermore, we found that no patient exhibited parasite positivity on day 3, although the follow-up number was small.
The effects of the C580Y mutation on ART resistance are important determinants of the survival of drug-resistant parasites. However, the introduction of C580Y into P. falciparum clones did not substantially increase the level of in vitro ART resistance compared to other mutations, such as R539T [40]. In addition, drug-resistant mutations generally confer a decrease in parasite fitness, which often leads to a survival disadvantage [41–44]. Several laboratory studies have demonstrated that the growth rates of C580Y harboring transgenic parasites were equal to or less than those of transgenic parasites with other pfk13 mutations, such as R561H, E252Q, and G538V, suggesting that C580Y incurs at least a similar level of fitness impairment to the other pfk13 mutations [40, 45, 46]. These laboratory findings are inconsistent with the field observations in the GMSR where the C580Y mutant has outcompeted other mutants [5, 9, 10]. This implies that some unique background genetic changes in the South-East Asia (SEA) parasites play a beneficial role in the survival of the SEA parasites harboring the C580Y allele. This might include compensation for the harmful effects of the C580Y mutation. Several single nucleotide polymorphisms (SNPs) have been identified in SEA pfk13 isolates [47]. However, among these SNPs, only one (ferredoxin D193Y) was found in our PNG C580Y mutants [13], suggesting that PNG C580Y mutants do not possess the same background genetic changes as SEA C580Y mutants.
As for chloroquine susceptibility, we previously reported that the average IC50s values were still high (80.5–106.6 nM) with a low prevalence of pfcrt K76 wild-type (2.3–11.7%) during 2016–2018 [14]. In 2020, however, the prevalence of pfcrt K76 wild-type rapidly increased to 46.7%. Since a significant association between the pfcrt K76 wild-type allele and ex vivo chloroquine susceptibility was confirmed in our study area [14], the observed rapid increase in the pfcrt K76 allele suggests that chloroquine sensitivity has been recovering. The significant increase in pfmdr1 N86 wild-type from 59.5–91.4% in the three years to 2020 also suggests the potential resurgence of chloroquine sensitivity. Although this phenomenon has been widely observed in African countries [15–26, 28], it is very rare in SEA [48].
We previously found that decreased ex vivo susceptibility to lumefantrine was significantly associated with pfmdr1 N86 [14], consistent with a previous transfection study showing that an allelic change from N86Y to N86 increased IC50 for lumefantrine three to four times [49]. Thereare considerable reports that pfmdr1 N86 wild-type is selected by AL treatment [50, 51]. Furthermore, a recent meta-analysis of 60 AL clinical trials revealed that only 38% of patients treated with AL were symptomatic when the infection recurred [52]. If patients are asymptomatic at the time of recurrence, they are unlikely to seek treatment, resulting in the persistence of parasitemia. This would increase the chance that the parasites could be transmitted to other human hosts, which could subsequently spread the drug-resistant mutations in the parasite population. In addition, a specific pfmdr1 haplotype (multicopy pfmdr1 in addition to N86 and Y184F) has increased and become predominant in Cambodia and Vietnam [53]. Gene editing experiments showed that this haplotype significantly reduces parasite susceptibility to lumefantrine [53]. In our study site, although this haplotype was not detected [13], all these genetic changes (multicopy pfmdr1, N86, and Y184F) were individually observed. Therefore, in addition to the assessment of the relapse rate following AL, it is necessary to monitor the appearance of this haplotype.