Prevalence of Antimalarial Resistance Mediators

(See the Major Article by Asua et al, on pages 985–94.)

Surveillance of DNA and RNA of pathogens has played an increasing role in control of infectious diseases in the last 30 years. These approaches have been widely applied to track the incidence of malaria disease and, more recently, to support efforts to eliminate the disease completely. Malaria currently has no efficacious vaccine, so prevention and treatment of disease both depend on the efficacy of antimalarials. This dependence means that evolution or spread of drug resistance in parasite populations is also a major threat to control and elimination of disease. Molecular methods have improved rapidly in recent years, so identification of the genetic changes potentially responsible for drug resistance and systematic surveillance for changes in marker prevalence are relatively easy. Their role is to give an early warning of antimalarial resistance. If the prevalence of parasites carrying a molecular marker is increasing, direct clinical assessment of drug efficacy can be launched.

The markers associated with resistance to most antimalarials in Plasmodium falciparum are relatively restricted and simple: single point mutations that alter activity of a targeted enzyme or transport protein [1]. However, since 2006, artemisinin-based combination therapies (ACTs) have been recommended. Unfortunately, resistance to the artemisinin component is complex. Some mutations in the propeller region of a key parasite protein, Kelch 13 (K13), diminish the speed of parasite clearance after treatment [2]. Increased prevalence and spread of about 20 of these K13 mutations in Southeast Asia have contributed to a rapid decline in ACT efficacy [3, 4]. Hundreds of parasites that carry a mutation that changes a single amino acid in the K13 propeller region have been identified at African sites, most at very low prevalence, [5]. Until recently, there has been little evidence for increased prevalence of any of these K13 alleles in Africa. A more focused approach for discovery of K13 alleles that might signal diminished response to artemisinin is clearly needed.

Asua and colleagues in the current issue of The Journal of Infectious Diseases present a focused application of molecular surveillance that may provide a way forward. The authors followed the changes in the prevalence of molecular markers in P. falciparum populations in 18 districts in Uganda between 2018 and 2019. In addition to chloroquine and sulfadoxine-pyrimethamine (SP), 2 older antimalarials, the group assessed K13 propeller mutants, and their partner drugs in ACTs [3]. With this focused approach, they identified 6 parasite populations with rapid changes in prevalence of markers associated with resistance to SP and 3 districts with parasites carrying K13 mutants associated with slow clearance in Southeast Asia.

Of most concern are changes in K13 markers. Recently, low numbers of K13 mutants matching an Asian genotype have been observed in Rwanda [6, 7], Tanzania [8], and the north of Uganda [9]. In this focused study, 3 districts in northern Uganda show a worrisome trend: >10% prevalence of parasites with either of 2 K13 mutants with “Asian” K13 genotypes. Even more remarkable, 1 allele was at very low prevalence among parasites at all 3 sites in 2018, but its prevalence increased to 20%, 30%, and almost 40% in these 3 districts in 2019. The high prevalence and very rapid changes in these 2 alleles are potential danger signs of diminishing efficacy of the ACTs that are the foundation of current malaria treatment in Africa [10]. Despite sharing the same K13 propeller mutations observed in Southeast Asia, these African mutants do not share other genomic signatures with Asian parasites, so they appear to have evolved and been selected in Africa by artemisinin use.

The second example concerns response to SP. Efficacy of SP depends on a synergistic action of the 2 drugs, and changes in either dihydrofolate reductase (DHFR) or dihydrpteroate synthase (DHPS) influence the response. Parasites with a specific trio of mutant codons in the pfdhfr gene and a pair of mutations in the pfdhps gene respond so poorly to SP that the drug was abandoned for treatment of clinical malaria as early as 2002 [11]. In this study, a cluster of 6 districts in southwest Uganda had parasites with a fourth mutant amino acid in DHFR, and several districts registered prevalence of these parasites between 40% and 80%. Prevalence of a third mutant amino acid in DHPS was between 20% and 40%. At 1 site studied, there was a jump from a prevalence of 20% in 2018 to 80% in 2019. Parasites with these additional mutations were first identified in Africa at low prevalence in southwest Uganda [12]. They were labeled “super resistant” because where they were prevalent, SP was no longer even effective for protection of pregnant women against malaria in Uganda [13, 14].

In both cases, these rapid increases in prevalence of molecular markers within or between nearby districts identify parasite populations that may be under strong directional selection for resistance to artemisinin or SP, and efficiently identify populations in efficacy of either ACTs or SP is under threat.

What distinguishes this study from many previous studies that assessed the speed and magnitude of changes in marker prevalence in Africa? First, for >20 years, this group has worked closely with Ugandan colleagues and studied malaria drug efficacy in central and eastern Uganda [15]. The current study greatly expanded the surveillance of molecular markers to 18 districts. Their experience meant that they could choose new sampling locations based on local knowledge and establish collaborations at these sites. As a result, they identified sites that varied in the intensity of malaria transmission, history of antimalarial use, human population density and movement, and history. Second, the samples were collected at the peak of malaria incidence in 2018 and 2019, avoiding differences that might simply reflect seasonal trends in the parasite populations. Since the group compared prevalence of key markers during a single year, 2018 to 2019, the short interval revealed several locations where major changes in marker prevalence occurred within this single year.

The focus and clarity of this study also stimulate ideas for further analysis. For example, the maps show some very abrupt changes in prevalence of markers in particular districts. They further look at the data and combine them with information about movement of people from outside Uganda or between districts. In some districts, marker information has been tracked for several decades and correlated with drug use and demographic parameters. Are there clues in these longitudinal data sets? These studies and many others could be pursued by anyone interested because all of the molecular data have been made freely available for use and can be accessed through the online Supplementary Material.

There is an additional informative aspect of this study. The Uganda field-based team has used the ligase detection reaction fluorescent microsphere (LDR-FM) assay for a long time. This method relies on detection of specific genotypes by capture of each target sequence by linking to a specific fluorescent microsphere. Then the prevalence of each sequence can be measured by the level of fluorescence of its microsphere [16]. There is a possible alternative approach that has been developed in the last year or so. This method uses molecular inversion probes (MIPs) to capture and amplify specific sequences of DNA, ready for next-generation sequencing (NGS) of the pools of these amplicons [17]. Both methods have the efficiency of highly multiplexed analysis of samples. Collaboration between the Ugandan-based field team using LDR-FM, and Jeffery Bailey’s group, which has expertise in the MIP probes and NGS sequencing, allowed the direct comparison of the success rate and sensitivity of mutant detection of the 2 approaches. A more conventional dideoxy-sequencing approach was also used as 1 method for detection of propeller mutants in the K13 gene since important unknown mutations within this region of the sequence might need to be identified. The methods for the MIP probes and analysis are detailed in the Supplementary Material provided by the authors. Analyses of samples from each year and each site were directly compared with respect to success of the method with sequencing each sample and the measured prevalence of each molecular marker.

Overall, the article by Asua et al presents a focused approach that may facilitate identification of hot spots where resistance to particular antimalarials may be on the rise. The combination of well-defined sampling sites, very precise and short time frame, and high-efficiency analysis of the samples suggests a possible way forward. The data sets used in this study are available in the Supplementary Material provided online, and they are a gold mine!

Notes

Potential conflicts of interest. The author: No reported conflicts of interest. The author has submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest.

References