Antimalarial Resistance Unlikely To Explain U.K. Artemether-Lumefantrine Failures

LETTER

Sutherland et al. focus on antimalarial resistance when explaining the failure of artemether-lumefantrine to cure falciparum malaria in four travelers returning from Africa (1). Recurrences in settings where malaria is not endemic certainly indicate therapeutic failure, and increasing failure rates in an area where malaria is endemic can be a sign of antimalarial resistance (2).

But was this “cluster” of failures a likely indicator of rising levels of resistance? We consider it very unlikely. Artemisinin combination therapies do not cure 100% of patients (3); in large trials and extensive surveys in regions where malaria is endemic and in smaller-scale work in nonimmune populations, artemether-lumefantrine efficacy is approximately 96% (3, 4, 5). Pharmacological factors clearly play a role in recrudescences (6). Furthermore, suppose 400 U.K. travelers are treated with artemether-lumefantrine annually, with 1% experiencing recrudescence (both conservative estimates). Although this averages only four recrudescences per year, we still expect a “cluster” of four recrudescences in 4 months approximately every 2 years, because randomness is clumpy, and all 4-month windows are counted (7).

Consistent with this, there was no robust clinical or laboratory evidence in the case descriptions indicating resistance to artemether-lumefantrine components. Almost no phenotypic data were available relating to parasite responses (parasite clearance rate [8], in vitro susceptibility to lumefantrine and artemisinins, and day 7 lumefantrine levels). Of the molecular markers described, only K13 (no mutations), pfmdr1, and pfcrt have been validated by transfection and/or independent association studies. All parasites carried pfmdr1 haplotypes containing ancestral alleles at the key N86 and D1246 residues (9, 10) and cannot be considered “resistant” unless we also define ancestral Plasmodium falciparum as lumefantrine resistant—which is clearly inappropriate. pfmdr1 gene amplification, a mutation genuinely associated with lumefantrine resistance, was not assessed.

For these reasons, the possibility that these cases represented resistant parasites appears remote, and practically, retreatment with artemether-lumefantrine would have been justified in our view. In African children, retreatment of recurrences (including recrudescences) with the same artemisinin-based combination treatment (ACT) provides similar efficacy to “rescue” treatment (11).

Although we disagree with the authors in terms of the likelihood that resistance explains these cases, we do agree that ongoing studies of resistance are needed. While lumefantrine (and mefloquine) fortunately work particularly well against the chloroquine-resistant parasite populations dominant across Asia and Latin America (except for mainland Southeast Asia), the situation is different in Africa where ancestral chloroquine-sensitive parasites have resurfaced following chloroquine withdrawal. While not resistant per se, these parasites are naturally less sensitive to lumefantrine, a factor potentially contributing to reports of lower cure rates with artemether-lumefantrine in areas where malaria is endemic (12) and in returning travelers (13). Such parasites respond well to artesunate-amodiaquine, but the best way to use these two ACTs providing opposing selective forces over an extended period remains ill defined, particularly in areas with falling transmission. Potential options include multiple first-line therapies, which modeling predicts will delay the emergence of resistance to both artemisinins and their partner drugs (14), or ACTs containing two partner drugs (“triples”) (15, 16). Despite the evidence that ACTs remain highly efficacious across Africa, experience tells us we need to stay ahead of the game.