Role of K13 Mutations in Artemisinin-Based Combination Therapy

To the Editor—We read with interest the article by Phyo et al [1] on the role of K13 mutations (markers of partial artemisinin resistance) in artemisinin-based combination therapy (ACT) failures in the Greater Mekong Subregion (GMS). Though we agree that Plasmodium falciparum parasites carrying both K13 mutations and genomic signatures of partner drug resistance have a multiplicative effect on ACT failure rates, we have a different interpretation of the apparent role of K13 mutation alone in ACT failures.

Artesunate-mefloquine (ASMQ) was implemented in Cambodia in 2001 in the presence of both artemisinin and mefloquine resistance. Contrary to what happened at the Thailand–Myanmar border, ASMQ rapidly encountered failure rates >10% at the Thailand–Cambodia border, and soon thereafter throughout Cambodia [2, 3]. ASMQ was replaced in 2008 by dihydroartemisinin-piperaquine (DP), first in Pailin province and later nationwide. Retrospective analysis of DP efficacy from 2009 to 2015 (n = 725) showed that treatment failure rates at day 42 in the presence of wild-type parasites lacking both artemisinin and piperaquine resistance markers, or in the presence of parasites carrying K13 mutations only, were 0% and 3.4%, respectively (Witkowski et al, unpublished data). On the contrary, high DP failure rates (46.4%) were observed in the presence of mutant parasites harboring both artemisinin and piperaquine resistance markers. Simultaneous detection of decreased mefloquine resistance (ie, decreased in vitro 50% inhibitory concentration, decreased pfmdr1 copy number) coincided with massive increases in both piperaquine resistance and DP failures. Unsurprisingly, ASMQ efficacy in 5 Cambodian provinces where DP was failing in 2014–2016 was 100% at day 42 (n = 305), even though 94.2% (278/295) of parasites carried the K13 C580Y mutation (Amaratunga et al, unpublished data; Rithea et al, unpublished data). High ACT efficacy despite high K13 mutation prevalence has also been reported in neighboring countries [4–7].

As components of ACT, the artemisinin derivative's role is to decrease the biomass, whereas the efficacious partner drug's role is to eliminate any remaining parasites. Therefore, we were surprised that Phyo et al reported that the presence of K13 mutations alone could lead to higher ASMQ failure rates than the presence of multiple pfmdr1 copy numbers alone. One potential explanation for this finding is that some admission isolates were misclassified as having a single copy of pfmdr1; these isolates may instead comprise a mixture of parasites having a single copy or 2 copies of pfmdr1 (such isolates would show a raw data pfmdr1 copy number between 1.2 and 1.5). This possibility is suggested by the observation that recurrent isolates were significantly more likely to have multiple pfmdr1 copies (53/65 [81.5%]) than admission isolates (377/726 [51.9%]; P < .001, Fisher exact test). Another explanation is that parasites having a single pfmdr1 copy may be overexpressing pfmdr1 mRNA, as described in South America [8]. These 2 hypotheses require investigation.

We agree that elimination of all parasites is the best way to rid the GMS of artemisinin-resistant parasites. To this end, the World Health Organization has supported the development of a malaria elimination strategy for this region, and is currently integrating it into the national malaria control programs of all GMS countries [9].