During September–December 2019, we recruited study patients with uncomplicated malaria seeking treatment at the Sovu Health Centre and Kabutare District Hospital, Huye district, Rwanda. Huye district (population ≈390,000) is located on the central plateau of Rwanda (average altitude 1,700 m, yearly rainfall 1,200 mm, mean temperature 19°C). Malaria transmission peaks in October–November and March–May. In 2010, a total of 11.7% of children had microscopically confirmed Plasmodium infection (8).

We obtained written informed consent from all participants or from the caregivers for children; we also obtained written assent from participants 7–18 years of age. The study was approved by the Rwanda National Ethics Committee. Eligibility criteria for participants included age >1 year; a positive result on a rapid diagnostic test, SD Bioline Malaria Ag Pf/Pan (Abbott Global Point of Care, https://www.globalpointofcare.abbott); and a fever (axillary temperature ≥37.5°C) at the time they sought treatment or within 48 hours beforehand (self-reported). We collected whole blood in S-Monovette EDTA (ethylenediaminetetraacetic acid; Sarstedt, https://www.sarstedt.com) tubes and confirmed malaria diagnosis by microscopy of Giemsa-stained thick blood smears; patients were also seen by a physician. We provided a 3-day regimen of artemether/lumefantrine for treatment, the first dose given under observation. All patients were asked to return after 3 days to evaluate residual parasitemia on Giemsa-stained thick blood smears.

Definite parasite density was counted per 200 leukocytes on Giemsa-stained thick blood smears by 2 independent microscopists, assuming a mean leukocyte count of 8,000/µL. We extracted DNA using a QIAamp DNA Blood Mini kit (QIAGEN, https://www.qiagen.com). Plasmodium species were typed by real-time PCRs with commercially available primers and probes for P. falciparum, P. vivax, P. ovale, P. malariae, and P. knowlesi (TIB MolBiol, https://www.tib-molbiol.com) on a Roche LightCycler 480 device (https://lifescience.roche.com). K13 was amplified (codons ≥441≤688) by using nested PCR (3) and sequenced by a commercial provider (Eurofins Genomics, https://www.eurofinsgenomics.com). Sequences were aligned to reference K13 3D7–1343700 (PlasmoDB, https://plasmodb.org) by using Geneious Prime version 2020.1 (https://www.geneious.com). We used R version 3.6.3 (https://cran.r-project.org) for statistical analysis and a binomial logistic regression model to estimate the time-trend of nonsynonymous mutations (p<0.05).

Of 90 patients included in the study, 74 tested positive by microscopy and PCR and 2 by PCR only. Among these patients, 51.3% (39/76) were female and 4 were pregnant; the median age was 18 years (range 2–69 years). P. falciparum was found in 88.2% (67/76), P. vivax in 7.9% (6/76), P. ovale in 7.9% (6/76) and P. malariae in 1.3% (1/76). The geometric mean parasite density, based on microscopy results, was 8,926 parasites/µL (95% CI 5,911–13,478 parasites/µL); mean temperature was 37.4°C (SD ±1.3°C). After 3 days of treatment, 61 malaria patients had a negative blood smear, 1 patient (1.6%, 1/62) had asymptomatic parasitemia (31,520 parasites/µL), and 14 patients did not return for the day 3 checkup.

None of the patients infected with K13 variant parasites tested positive after 3 days of treatment. One pregnant patient sought treatment again. Initially, she had K13 wild-type parasites and was given artemether/lumefantrine; her day 3 microscopy result was negative. Three weeks later, we detected K13 R561H parasites, possibly due to reinfection, and administered quinine.

Samples from 98% (66/67) of P. falciparum isolates were successfully sequenced for the K13 propeller domain. We found 5 different nonsynonymous polymorphisms in 8 isolates (Table); 3 harbored R561H, a validated resistance mutation, and the 2 candidate polymorphisms C469F and A675V (4). This finding suggests that the number of isolates with nonsynonymous K13 mutations had increased significantly over the previous decade (OR 1.4, 95% CI 1.1–1.8; p = 0.003).

Of P. falciparum isolates from symptomatic patients in southern Rwanda, 12% exhibited nonsynonymous K13 mutations, a significant increase (OR 1.4, 95% CI 1.1–1.8; p = 0.003) over the previous decade compared with their absence in 2010 and 4.5% prevalence in 2015 (7). Of note, the validated marker R561H alone occurred in 4.5% of the isolates collected in 2019. Recent studies report 1%–3.5% of nonsynonymous K13 polymorphisms in parasite isolates from East Africa (10), whereas during 2013–2015 in Rwanda, this figure was 6.9% (9).

The R561H artemisinin resistance mutation is regularly observed across Asia (10). A recent study that reported R561H in 7.4% of isolates collected during 2013–2015 in central Rwanda and 0.7% of isolates in south-central Rwanda (9) suggested that this mutation emerged indigenously and independently from Asia 561H strains. We do not have data in our study to support this. None of the K13 variant parasites showed delayed clearance in our study, which may be due to the partner drug lumefantrine still being effective, similar to observations in Southeast Asia (11). In addition, the absence of delayed parasite clearance despite K13 mutations may reflect partial immunity contributing to parasite elimination (12).

We found other nonsynonymous polymorphisms only once among the isolates tested. C469F and A675V are considered artemisinin resistance candidate mutations (4) and have previously been seen in East Africa (7,13,14). G533A and V555A have also been previously reported in Africa but have not yet been evaluated for resistance (7,13). A578S is a common K13 polymorphism across Africa but is not linked to artemisinin resistance (1).

Our study has clear limitations. Data from only 2 healthcare facilities, with limited catchment areas, were included. Adherence to treatment was assessed by patient self-report, and drug susceptibility testing was not performed. Future research should include ring-stage susceptibility assays to contribute to understanding the role of K13 mutations in Africa. Separate testing for each drug in a combination for efficacy and continued surveillance for antimicrobial resistance are needed.

Our results show that K13 mutations are present in Rwanda and that their prevalence in P. falciparum malaria patients in the Huye District increased from 0% in 2010 to >12% in 2019. The validated artemisinin resistance mutation R561H occurs in 4.5% of P. falciparum isolates being transmitted in this area. The emergence of artemisinin resistance–related mutations in Rwanda is alarming because it might indicate developing resistance against commonly used antimalarials in this region. Countermeasures need to be considered early, potentially including 3-drug antimalarial combinations (2).