A 38-year-old Nigerian-born woman, who lived in the Toronto area (and has a good ability to recount her experiences), returned to Lagos, Nigeria, for a visit in January 2009. She did not seek pretravel advice. On arrival in Lagos, the woman purchased artesunate locally and began taking two 50-mg tablets weekly for the 4 weeks of her visit. Immediately on her return to Toronto, the patient experienced myalgia, nausea with vomiting, and chills, ≈7 days after she had taken her last dose of oral artesunate. She sought treatment at the emergency department of a community hospital. Physical examination showed that her temperature was 39.1°C and that she was dehydrated. Laboratory tests showed the following: leukocyte count 3,700 cells/μL, thromocyte count 72 × 10³ cells/μL, hemoglobin level 12.7 g/dL. Her chest radiograph showed that her lungs were clear. An examination of peripheral blood by thick and thin films showed a 0.7% parasitemia with P. falciparum. Her condition was treated with 1,250 mg of oral mefloquine as a single dose. She was treated as an outpatient, and she reported that symptoms promptly resolved over the next 48 hours without side effects.

A blood specimen was placed into culture in the Public Health Laboratory (Toronto), and the P. falciparum isolate was tested for drug susceptibility (6). The 50% inhibitory concentration (IC₅₀) was the following for certain antimicrobial agents (tested in triplicate): chloroquine 170.5 ± 7.8 nmol/L, mefloquine 16.6 ± 0.7 nmol/L, artemisinin 20.1 ± 0.6 nmol/L, artesunate 6.2 ± 1.4 nmol/L, dihydroartemisinin 1.8 ± 0.9 nmol/L, and artemether 21.4 ± 5.3 nmol/L. For this P. falciparum isolate, IC₅₀ was significantly higher for artemisinin, artesunate, and artemether than for other representative P. falciparum isolates imported from Africa (Figure). Because of the short half-life of artesunate, the weekly doses of the oral drug may have led to development of a resistant strain when the patient was in Nigeria. Artesunate-containing drugs therefore should not be used for prophylaxis or single drug therapy. The purchased artesunate may also have been counterfeit and may have contained lower levels of active drug. Although these data suggest that this isolate has reduced susceptibility to artemisinin derivatives, the correlation between in vitro susceptibility and treatment outcomes does not appear to be consistent (4).

Previous studies have reported that resistance to artemisinin is mediated by an increase in gene copy number, mutations within the efflux pump of the P. falciparum multidrug resistance 1 (pfmdr1) gene, or mutations in the calcium transporter pfATPase6 (7,8). When we examined each gene, using a combination of real-time PCR and DNA sequencing, we found that pfmdr1 copy number was elevated in this isolate relative to that of the susceptible control strain 3D7. We also observed nonsynonymous mutations in both pfmdr1 (Y184F) and pfATPase6 (A623E, S769N), whereas other implicated residues remained in the wild-type form (9) (Table). Similar molecular analysis of other representative imported African clinical isolates demonstrated variable mutations for pfmdr1 and pfATPase6 and copy number in relation to IC₅₀ values for key drugs (Table). A trend, albeit weak, was observed in which increased pfmdr1 copy number was correlated with an elevated IC₅₀ to mefloquine (r = 0.52) and artemisinin (r = 0.42). The presence of an asparagine (N) at position 86 of Pfmdr1, when coupled to an elevated pfmdr1 copy number, appeared to correlate well with reduced susceptibility to artemisinin (Table). Chavchich et al. recently demonstrated that increased pfmdr1 copy number occurred in a laboratory strain placed under drug selection pressure with artemisinin derivatives (11). However, Imwong et al. have indicated that genetic polymorphisms and copy number in pfmdr1 do not predict treatment outcome with ACT (10).

Findings in the published literature vary in terms of use of artemisinin derivatives for in vitro drug susceptibility testing. Jambou et al. reported treatment failures with ACT in Cambodia, French Guiana, and Senegal (8). These authors used artemether for testing and showed IC₅₀ values of ≈30 nmol/L in their “resistant” isolates from Senegal. Noedl et al. described treatment failures with ACT in Cambodia, for which IC₅₀ values to dihydroartemisinin were ≈10 nmol/L (12). Dondorp et al. showed IC₅₀ values of 4–6 nmol/L to dihydroartemisinin and 6–8 nmol/L to artesunate in a region of Cambodia and Thailand where ACT treatment failures have occurred (4). Systematic molecular surveillance and standardized drug-testing methods with clinical isolates are required to establish the molecular correlates of reduced susceptibility to antimalarial drugs. In this regard, efforts are ongoing under the auspices of the Worldwide Antimalarial Research Network (13).

The patient’s infection responded to mefloquine when she was back in Canada, possibly because of the high oral dose of mefloquine. Current guidelines from the US Centers for Disease Control and Prevention recommend quinine sulfate plus doxycycline, tetracycline, or clindamycin; or atovaquone-proguanil (Malarone; GlaxoSmithKline, Mississauga, Ontario, Canada) as first- and second-line treatment for uncomplicated P. falciparum malaria. Reduced susceptibility to artesunate is more likely to occur when it is associated with inappropriate use of artemisinin derivatives than because of circulating artemisinin-resistant P. falciparum in sub-Saharan Africa.

In an effort to achieve consensus that artesunate oral monotherapies should not be marketed, the World Health Organization convened the international pharmaceutical sector in April 2006. At that time, 15 companies agreed to cease manufacturing artesunate monotherapies. However, oral artesunate montherapies may still be purchased over the counter in malaria-endemic countries, as this report shows. Thus, strains of P. falciparum malaria are currently at risk of developing reduced susceptibility to artesunate derivatives.