1. IntroductionThe Greater Mekong Subregion (GMS) refers to six countries and regions in the Mekong River Basin, including Cambodia, Vietnam, Laos, Myanmar, Thailand, and China’s Yunnan and Guangxi provinces. China achieved the goal of malaria elimination in 2021, while the rest of the countries are trying to eliminate malaria by 2030 [1,2,3]. Myanmar has the highest malaria burden among the GMS countries [4]. Political instability and military conflict in Myanmar have driven hundreds of thousands of citizens into resettlement camps along its border, where malaria transmission is high [5]. While the camps’ poor sanitation and makeshift structures are conducive to malaria transmission, massive population movements to these camps have also contributed to malaria introduction [6]. Furthermore, intensive malaria transmission at the borders also poses a major threat to malaria introduction to neighboring countries [7]. Therefore, malaria surveillance, a core intervention promoted by the World Health Organization’s Global Technical Strategy for Malaria 2016–2030, must be strengthened to prevent the cross-border spread of malaria.The GMS is a breeding ground for multidrug-resistant Plasmodium falciparum, which has developed resistance to almost all commonly used antimalarials [8]. Due to the difference in the adoption of drug policies, the epidemiological backgrounds of P. falciparum and P. vivax malaria, the public health infrastructure, and accessibility to treatment in the GMS countries, drug resistance exhibits considerable geographical variations. Even P. falciparum populations from closely located transmission “pockets” separated by malaria-free zones can display distinct resistance phenotypes [9]. This geographical difference highlights the necessity of resistance surveillance and management in multiple sentinel sites of malaria endemicity.The Laiza town of Kachin State is in northeastern Myanmar and borders China. Throughout Myanmar’s history, this region has been relatively independent and has had little communication with the central government. Therefore, it has adopted a malaria prevention and treatment policy similar to that in China. Per local government policy, dihydroartemisinin–piperaquine (DHA-PPQ) has been used as the major artemisinin-based combination therapy (ACT) to treat falciparum malaria [10]. In addition, other antimalarial drugs, such as naphthoquine (NQ) and pyronaridine (PND), were also available for malaria treatment in the private sector. Muse is a border city of Shan State, Myanmar, 125 km south of Laiza. In recent years, economic trade between China and Myanmar has increased population movement to this region. As a result, malaria in this area is more migration-related. Malaria treatment here follows Myanmar’s antimalarial treatment policy [11], with artemether–lumefantrine designated as the first-line therapy for uncomplicated P. falciparum malaria [12]. Since we have installed malaria surveillance in these border areas, we were interested in determining whether P. falciparum parasites collected from the two border areas represent different parasite populations. After adapting clinical parasites to long-term in vitro culture, we profiled their in vitro susceptibilities to a panel of antimalarials and genotyped their drug resistance genes. This study revealed significant differences in drug resistance between these parasite populations, even though they were collected from two adjacent regions, emphasizing the importance of monitoring the sources of parasites at the ports of introduction. 4. Discussion

Multidrug-resistant P. falciparum is a significant challenge to the global efforts of malaria eradication. Although P. falciparum incidence in the GMS has continually declined, monitoring drug resistance has remained critical for updating regional antimalarial drug policies. This study represents our efforts to monitor drug resistance in sentinel sites along the China–Myanmar border using in vitro drug assays and molecular surveillance. By assessing the in vitro susceptibilities of parasites from two adjacent but separated areas to 11 common antimalarial drugs, we identified their significantly divergent drug susceptibility profiles, with parasites from Laiza having significantly higher IC50s to NQ, PY, AS, DHA, AM, PND, and LMF than those from Muse. Consistent with the in vitro phenotypic difference, parasites from the two regions were also distinctive in resistance-conferring mutations, possibly reflecting different origins of parasite populations and different drug histories.

CQ and antifolate drugs have long been withdrawn from treating falciparum malaria, but parasites from the two regions remained highly resistant to these drugs. Consistent with this, resistance-conferring mutations in pfcrt, pfdhfr, and pfdhps were highly prevalent, with many remaining fixed in the parasite populations. Persistent CQ resistance has been speculated to be due to continued CQ selection pressure from the widespread use of CQ to treat sympatric P. vivax infections [29] and fixation of mutations in pfcrt that mediate CQ resistance in the parasite populations [32]. Since parasites from the hypoendemic regions are predominantly monoclonal, and intrahost competition is low, these mutations are likely preserved even though they inflict high fitness costs. In comparison, the pfmdr1 N86Y mutation, which was also associated with CQ resistance [33], was infrequent in the GMS parasite populations. On the other hand, the persistent resistance to antifolates may be related to their use to treat bacterial infections [34,35,36]. For example, trimethoprim–sulfamethoxazole, used to treat acute respiratory infections, presents cross-resistance with pyrimethamine and sulfadoxine [34,35].Differential use of antifolate drugs in Laiza and Muse may account for the >4-fold differences in PY susceptibility and different frequencies of key mutations in pfdhfr (N51I) and pfdhps (S436A and A581G) between the two populations.Partial artemisinin resistance, displayed as delayed parasite clearance following treatment with an ACT, has been detected in all the GMS regions, and the K13-propeller mutations have been used widely to track the emergence and spread of ART-resistant P. falciparum [31,37]. Within the GMS, pfk13 mutations are diverse and region-specific [38,39]. F446I was most prevalent in northern Myanmar and the China–Myanmar border [38,40], and this study further confirmed this. Although the N-terminal NN insertion was associated with altered susceptibility to ART drugs [41] and its prevalence has increased dramatically over the years along the China–Myanmar border [42], we did not identify the association of this insertion with changes in in vitro IC50 values to DHA and AS. Siddiqui et al. showed that the N458Y mutation, which occurs at the China–Myanmar border, confers ART resistance with a significant increase in RSA values in vitro [43]. In our study, the median RSA was higher in parasites with the N458Y mutation (5.0%) than in the K13 WT group (2.1%), but the difference was not significant, probably because the number of samples with the N458Y mutation was limited and therefore it was difficult to draw firm conclusions about the association of certain K13 genotypes with reduced ART susceptibility. Consistent with previous reports [43], F446I is associated with significantly higher RSA values than the WT parasites. Globally, K189T was identified at a relatively high proportion in the Amazon basin [44]. This mutation showed a similar prevalence in Africa, but was rarely described in Southeast Asia [45,46,47,48]. The study by Wu et al. in 2020 showed that K189T was first discovered in Myanmar [49]. Reports showed that K189T mutation is associated with delayed parasite clearance, but there is no clear correlation with ART resistance [48,50,51]. However, in our study, the RSA values of the samples with K189T mutation were significantly higher than those of WT type, which provides a direction for future studies.In Cambodia, AS-MFQ was the first ACT introduced, but its efficacy steadily declined in the early 2000s [52], resulting in the switch to DHA-PPQ in 2008. However, with partial ART resistance emerging in the region, this ACT was soon found to be ineffective [53,54]. The concurrent recovery of the sensitivity of the parasites to AS-MFQ led to the consideration of recycling this ACT. ACT failures in Cambodia appear largely due to resistance to partner drugs. MFQ and PPQ seem to impose opposing selection on drug targets, especially pfmdr1 copy number: MFQ is associated with pfmdr1 amplification, whereas PPQ selects single-copy pfmdr1. Such opposite selection on the same target has been the basis for including these drugs in a triple ACT design [55]. The main ACT used in the China–Myanmar border area was DHA-PPQ, which remained highly efficacious [10,56]. Our recent study using an in vitro drug assay also showed that parasites from the China–Myanmar border area were largely susceptible to PPQ [57]. Consistently, parasites from this region did not contain plasmepsin2/3 amplification [58,59] or the new pfcrt mutations (H97Y, F145I, M343L, and G353V) conferring PPQ resistance [60,61,62], while the parasites also contained predominantly single-copy pfmdr1. It is noteworthy that the I356T mutation fixed in the study populations was significantly associated with decreased QN sensitivity and increased MFQ sensitivity in P. falciparum parasites from Africa [20]. In addition, I356T and N326S might be the background mutations on which pfk13 mutations emerged [63].The ATP-binding cassette (ABC) transporters, including pfmdr1 and pfmrp1, are involved in parasite resistance to multiple antimalarial drugs [64]. The Y184F mutation was only detected in Laiza, and the F1226Y mutation in Muse. The Y184F mutation was associated with increased resistance to AS, LMF, and MFQ, while the F1226Y mutation was correlated with increased susceptibility to PND and NQ, suggesting that the divergent pfmdr1 mutations may be partially responsible for the differences in drug susceptibility between the two neighboring sites. MFQ has not been deployed in the China–Myanmar border area. Consistently, parasites from the two areas showed similar sensitivity to MFQ and pfmdr1 amplification associated with MFQ resistance was not detected [65]. While this and an earlier study identified that several mutations in pfmrp1 were associated with altered sensitivities to a number of drugs [23], they were similarly prevalent in the two populations. Thus, the significance of the pfmrp1 mutations in the context of pfcrt and pfmdr1 haplotypes warrants further investigation.Antimalarial therapy is one of the pillars of malaria control and elimination. Updated knowledge about antimalarial resistance in malaria parasites is critical for delivering effective frontline treatment. Myanmar has the heaviest malaria burden in the GMS and is also a gateway to the Indian subcontinent. Thus, effective management of border malaria is essential for preventing cross-border spillover of resistant parasites, especially to neighboring countries that have just become malaria-free. This and an earlier study have demonstrated drastic differences in drug resistance between neighboring parasite populations during the elimination phase in the GMS [9]. A limitation of this study is the relatively small sample size. As malaria incidence has plummeted in recent years, collecting laboratory culture samples has become difficult. Differences in anthropology, administration, public health infrastructure, access to treatment, and intensity of malaria transmission across borders may contribute to the observed differences in susceptibility to antimalarial drugs. This may require timely adjustment of the antimalarial drug policy between different strata of malaria endemicity.