Detection of Artemisinin Resistance Marker Kelch-13 469Y in Plasmodium falciparum, South Kivu, Democratic Republic of the Congo, 2022

ABSTRACT.

Partial artemisinin resistance has emerged in East Africa, posing a threat to malaria control across the continent. The Democratic Republic of the Congo carries one of the heaviest malaria burdens globally, and the South Kivu province directly borders current artemisinin resistance hot spots, but indications of such resistance have not been observed so far. We assessed molecular markers of antimalarial drug resistance in 256 Plasmodium falciparum isolates collected in 2022 in South Kivu, Democratic Republic of the Congo. One isolate carried the P. falciparum Kelch-13 469Y variant, a marker associated with partial artemisinin resistance and decreased lumefantrine susceptibility in Uganda. In addition, the multidrug resistance-1 mutation pattern suggested increased lumefantrine tolerance.

Sub-Saharan Africa (SSA) bears the vast majority of the annual 247 million malaria cases and >0.6 million malaria-related deaths. The Democratic Republic of the Congo (DRC) accounts for more than 12% of this burden.¹ Malaria control relies on prompt and effective treatment, facilitated by artemisinin combination therapies (ACTs), consisting of a rapid-acting artemisinin derivative and a long-lasting partner drug, e.g., artemether-lumefantrine or artesunate-amodiaquine. However, partial artemisinin resistance (ART-R) had already emerged two decades ago in Southeast Asia and was recently observed in East Africa.² In Asia, resistance initially manifested as delayed parasite clearance after treatment. As a consequence of secondary or parallel development of resistance to the non-artemisinin partner drug, ACT failure rates of 50% are now seen in some hot spots.³ Although still far from such figures, the artesunate-lumefantrine efficacy in ACT efficacy studies is increasingly reported to be below the WHO threshold of 90% in SSA.² Potential reasons include the recent emergence and limited spread of ART-R, maintenance of partner drug activity, and population immunity.

Partial artemisinin resistance is mediated by mutations in the Plasmodium falciparum Kelch-13 (PfK13) propeller domain, categorized as validated mutations (associated with delayed parasite clearance in patients and in vitro/ex vivo confirmation) and candidate mutations (associated with delayed parasite clearance only).² Validated mutations associated with ART-R have been demonstrated across East Africa, the mutations of concern being 441 L, 469Y/F, 561H, 622I, and 675V.^2,4–9 In parts of Rwanda and Uganda, the recent prevalence of validated resistance mutations exceeds 20%.^6,8 Susceptibility to the main partner drugs lumefantrine and amodiaquine is impacted by mutations in the P. falciparum multidrug resistance-1 gene (PfMDR1).¹⁰ PfMDR1 N86 (wild type) is associated with decreased sensitivity to lumefantrine but with increased sensitivity to amodiaquine, and the N86-184F-D1246 (NFD) haplotype is associated with decreased artemether-lumefantrine sensitivity.^10,11 In line with that, N86 and D1246 mutations dominate where artemether-lumefantrine has long been the mainstay antimalarial.¹²

The DRC is a neighbor of Rwanda and Uganda and is considered a bridge for antimalarial resistance spread between East and West Africa.¹³ South Kivu, eastern DRC, had >1.5 million malaria cases in 2020, as reported by the National Malaria Program. It contains rural and urban settings at varied elevations. The provincial capital, Bukavu, directly borders Rwanda and experiences intense cross-border movements. National malaria treatment guidelines recommend artemether-lumefantrine or artesunate-amodiaquine as first-line antimalarials, but adherence might deviate and other drugs are frequently used.¹⁴ Considering the local spread and intensification of ART-R in the region,² we assessed PfK13 and PfMDR1 mutations in P. falciparum collected in 2022 from South Kivu.

In July–October 2022, we collected 292 anonymous, P. falciparum-positive malaria rapid diagnostic tests (RDTs; SD Bioline® Malaria Ag Pf [Standard Diagnostics, Inc.] and First Response® Malaria Ag Pf/PAN [Premier Medical Corporation, Ltd.]) from 18 health centers located in three health zones (Miti-Murhesa, Nyantende, and Uvira) as well as 49 anonymous, Giemsa-stained, positive thick blood smears from the Provincial Referral Hospital in Bukavu city. For DNA extraction (QIAamp DNA blood minikit [Qiagen]), the RDT cellulose strip was immersed in 250 µL of phosphate-buffered saline (PBS), 250 µL of AL buffer (Qiagen), and 20 µL of Proteinase K; for smears, blood was scratched from the glass slide with a sterile scalpel and dissolved in 200 µL of PBS, 200 µL of AL buffer, and 20 µL of Proteinase K. After 1 hour of incubation at 56°C, the manufacturer’s standard instructions were followed. PfK13 was amplified by a nested PCR assay¹⁵ but using forward-outer primer 5′-GTGGAAGACATCATGTAACCAGAG-3′ and Hotstart FIREPol DNA polymerase (Solis Biodyne) for RDT DNA and Phusion high-fidelity polymerase (New England Biolabs) for smear DNA. For PfMDR1, nested PCR assays on a random sample subset (N = 131) amplified two regions, one spanning N86Y and Y184F and one spanning D1246Y.¹⁶ PCR amplicons were sequenced using Sanger sequencing (Eurofins Genomics), and sequences were aligned to references PF3D7_1343700 (K13) and PF3D7_0523000 (PfMDR1; https://plasmodb.org) in CodonCode Aligner v. 9.0. Loci with both wild-type and mutant alleles were counted as mutants and considered as such for haplotype counting.

We detected four nonsynonymous PfK13 mutations, each only once, in 256 samples with good-quality reads for K13. The PfK13 mutations comprised 469Y (mixed signal, from Bukavu city), 569S (homozygous signal, Uvira), 581D (mixed signal, Miti Murhesa), and 522R (homozygous signal, Bukavu city). Among these, the 469Y mutation is known in Uganda to be linked to delayed parasite clearance in patients and significantly decreased susceptibility to artemisinin and lumefantrine ex vivo.^7,8 For PfMDR1, reads spanning amino acid positions 85 to 231 were obtained from 97 samples, and reads spanning positions 1042 to 1289 were obtained from 87 samples. PfMDR1 alleles N86, 184F, and D1246 were present in 95.9% (93/97, 86Y signal in all samples was homozygous), 45.3% (44/97, 184F signal was mixed in 9/44 samples), and 95.4% (83/87, 1246Y signal in all samples was homozygous), respectively, and the haplotype NFD was present in 42.4% (31/73). Other haplotypes were NYD (35/73), NYY (2/72), and NFY (1/73). We observed seven additional nonsynonymous PfMDR1 mutations, each once unless otherwise indicated: 86I, 103F, 129I, 199S (2/97), 222I (9/97), 1109 L, and 1151R.

We demonstrated the presence of the artemisinin resistance-associated marker PfK13-469Y in a P. falciparum isolate from Bukavu, South Kivu, DRC, in 2022. Moreover, this marker was recently associated with decreased lumefantrine susceptibility ex vivo in East Africa.⁷ Although seen only once in more than 250 samples, the detection of PfK13-469Y complements the recent report of 561H and 441 L mutations in other parts of the country,¹⁷ now adding DRC to the countries in SSA where ART-R-associated mutations have been detected. Our finding could reflect importation by an infected patient, e.g., from Uganda, or a local occurrence. The other three identified PfK13 mutations are of unknown relevance. The mutation 522C/A (here, 522R) has been observed sporadically in several countries, including DRC, and 569T/G/S (here, A569S) has been observed in Kenya and DRC.^17,18 Overall, PfK13 variation (1.6% nonsynonymous mutations) in South Kivu did not reflect the figures from neighboring Rwanda. In Huye district, southern Rwanda, we found 12% nonsynonymous mutant isolates in 2019, with the validated marker 561H predominating.⁴ This could indicate a limited crossover of P. falciparum parasites from Rwanda to South Kivu, a pending arrival of PfK13 mutant parasites from Rwanda, and/or that PfK13 mutant parasites are selected out in South Kivu because of fitness costs at presumably differing artemisinin pressures.

Regarding the PfMDR1 pattern in South Kivu, wild-type alleles N86 and D1246 are close to fixation, whereas for codon 184, wild-type alleles make up about half. Compared with data from DRC from 2008 to 2013, N86 and D1246 are 10- and 5-fold increased, respectively, whereas 184F occurs at a similar proportion in South Kivu.¹² Together with 43% of the PfMDR1 NFD haplotype, our data suggest the presence of an artemisinin-resistant parasite line in a setting of compromised susceptibility to the non-artemisinin partner drug lumefantrine.

Our data should be interpreted with caution because they might not accurately represent the actual parasite population in South Kivu. Low-density infections may have been missed because positive RDTs and smears were used as a DNA source, and genotyping success was not complete. Further infections might have been overlooked because of P. falciparum HRP2/3 deletions,¹ rendering RDTs falsely negative. Increased PfMDR1 copy number, affecting susceptibility to various antimalarials,¹ was not assessed but was absent from a nationwide study from 2019.¹⁹ We analyzed a limited selection of molecular markers in the present study. Additional molecular determinants potentially affecting partner drug susceptibility should be considered in future work as well, e.g., P. falciparum chloroquine resistance transporter 76T or novel PfMDR1 loci of interest. Study strengths include the rapid generation of urgently needed data for the African Great Lakes region and for DRC in particular. These data form the basis for subsequent monitoring and are valuable for designing containment measures. Eastern DRC directly borders Rwanda and Uganda, apparently the current epicenter of ART-R in SSA. The molecular epidemiology of chloroquine and sulfadoxine-pyrimethamine resistance demonstrated a strong East-West divide in P. falciparum populations, where DRC serves as a bridge for gene flow, including antimalarial resistance genes.¹³ To halt the dissemination of ART-R, eastern DRC appears to be crucially placed. This warrants upscaled molecular surveillance, understanding the pattern of emergence and spread, functional analysis of novel mutations, confirmatory treatment efficacy trials, and piloting of countermeasures. Although still experimental and under debate, potential countermeasures are prolongation of the ACT regimen to 5 days, alternation of different ACTs, and implementation of triple ACT to delay resistance development, as well as single low-dose primaquine treatment to interrupt the transmission of resistant parasites.