The Lancet Infectious Diseases: Rapid spread of multidrug-resistant malaria in southeast Asia demands urgent action

02 Feb 2018

Genetic surveillance should be incorporated into malaria control programmes to improve treatment and reduce risk of drug-resistant major outbreaks.

The current spread of multidrug-resistant malaria in southeast Asia is likely to be the result of two mutations combining in 2008, according to a retrospective genetic study published in The Lancet Infectious Diseases journal.

The study shows how the multidrug-resistant parasite gained increased biological fitness, spreading rapidly through the region unnoticed for 5 years until the outbreak became apparent in 2013. The authors warn that malaria programmes should closely monitor genetic mutations to mitigate the possibility of the parasite becoming untreatable.

Malaria is caused by Plasmodium parasites, which are transmitted by mosquitoes. Typically, treatment for malaria involves a combination of drugs including artemisinin – a potent and fast-acting antimalarial drug – and a longer-acting partner drug to ensure that all parasites are killed and to prevent the emergence of resistance.

In 2008, Plasmodium falciparum started to become resistant to artemisinin in western Cambodia. Since then, resistance has been observed in other parts of Cambodia, Thailand, Vietnam, Myanmar, and Laos.

From 2013, the frequency of complete treatment failure in patients receiving a drug combination of dihydroartemisinin and piperaquine increased rapidly in Cambodia, northeast Thailand, and Vietnam.

“Malaria policy makers now face a dilemma. On one hand, malaria remains treatable and its prevalence has been reduced to low enough levels to aim to eliminate the disease in Cambodia and neighbouring countries. However, the situation is fragile, and it is unclear how the parasite population will evolve in response to new interventions,” says Roberto Amato, Wellcome Sanger Institute, UK.

While it would be catastrophic if resistance developed in the same way for the last remaining anti-malarial drugs, it is now possible to conduct genetic surveillance of malaria cases, allowing researchers to respond as soon as possible to changes in the parasite population. It is important that we embrace these technologies so that major outbreaks of resistance do not go unnoticed in the future, and to reduce the risk of a global health emergency.” [1]

In the study, the authors analysed the genomes of 1492 P falciparum samples from 11 locations across southeast Asia between 2007-2013, including 464 samples collected in western Cambodia, to determine how resistance developed.

Resistance to artemisinin is caused by mutations in a gene called kelch13, while amplifications of the genes plasmepsin 2 and plasmepsin 3 are linked to resistance to piperaquine.

Overall, 46% (689/1492) P falciparum samples carried kelch13 mutations, and 14% (199/1465) carried amplifications of the plasmepsin 2-3 genes [2], but these mutations were more common in samples from western Cambodia (where 83% [384/464] samples had kelch13 mutations, and 41% [185/456] samples had plasmepsin 2-3 amplifications).

Analysing how these mutations developed, the researchers found 38 genetic origins of artemisinin resistance, and two origins of piperaquine resistance. The researchers next investigated how many P falciparum samples were resistant to dihydroartemisinin–piperaquine and determined how many were caused by the combination of kelch13 and plasmepsin 2-3 mutations.

154 samples from across southeast Asia were resistant to the drug combination, and most of these cases (91% [140/154] of parasites) were caused by one genetic origin of kelch13 mutations (called the KEL1 lineage), combined with one origin of piperaquine resistance (called the PLA1 lineage). The two mutations combined in 2008 in western Cambodia, in the same year that dihydroartemisinin–piperaquine officially became the first-line antimalarial drug.

At this time, 31% (4/13) resistant samples from western Cambodia were of the KEL1/PLA1 co-lineage. This specific combination of mutations then spread rapidly for 5 years before the first clinical reports of a major outbreak of multidrug resistance appeared – by 2013, 92% (22/24) of resistant parasite samples from western Cambodia carried these two mutations.

Further, in 2012-13, the mutations were found in 11 samples from northern Cambodia and one sample from Laos, suggesting that resistance had spread to northern Cambodia.

The rapid spread of the mutation pair suggests that artemisinin-resistant parasites are acquiring increased biological fitness, and it is unclear how much this increases the risk of resistance to other drugs and trans-continental spread. The authors also note that the mutation pair seems to have displaced other artemisinin-resistant parasite lineages, including those that cause resistance to the anti-malarial drug mefloquine.

Writing in a linked Comment, Dr Didier Ménard, Pasteur Institute, France, says:

“The results of this study are reminiscent of the evolution of chloroquine resistance, wherein multiple P falciparum chloroquine resistance transporter (Pfcrt) alleles emerged in southeast Asia before one allele (the CVIET allele) eventually spread to Africa, leading to millions of deaths. Obviously, this scenario should be avoided for artemisinin combination therapy. For chloroquine, the molecular signatures of resistance were only detected in early 2000, long after resistant parasites had spread outside their original focus.

"The spread of strains resistant to artemisinin combination therapy in western Cambodia is underway; however, it is reassuring to learn from this study that genomic tools are available to monitor the onset of this spread and, by contrast with chloroquine resistance, to track resistant parasites in real time. We must take advantage of this situation. One way is to improve understanding of the causes of emergence and selection of resistance to artemisinin combination therapy by progressing analyses of parasite population genetics.”


[1] Quote direct from author and cannot be found in the text of the Article.

[2] There were 1465 samples for plasmepsin 2-3, as results for 27 samples were inconclusive.