Advances and roadblocks in the treatment of malaria

The deployment of artesunate for severe malaria and the artemisinin combination therapies (ACTs) for uncomplicated malaria has been a major advance in antimalarial therapeutics. These drugs have reduced treated mortality, accelerated recovery and reduced treatment failure rates and transmission from the treated infection. Artemisinin derivatives remain highly effective against falciparum malaria in most malaria endemic areas, but significant resistance has emerged in the Greater Mekong subregion of Southeast Asia. Resistance to artemisinins was followed by resistance to the ACT partner drugs, and fit multidrug resistant parasite lineages have now spread widely across the region. ACTs remain highly effective against P. vivax and the other malaria species. Recent studies have shown that radical curative regimens of primaquine (to prevent relapse) can be shortened to 7 days, and that the newly introduced single dose tafenoquine is an alternative, although the currently recommended dose is insufficient in Southeast Asia and Oceania. Targeted malaria elimination using focal mass treatments with dihydroartemisinin‐piperaquine have proved safe and effective malaria elimination accelerators, but progress overall towards malaria elimination is slow. Indeed since 2015 overall malaria case numbers globally have risen. As new drugs will not become widely available in the near future, active measures to preserve the current antimalarials should be given the highest priority.

The deployment of artesunate for severe malaria and the artemisinin combination therapies (ACTs) for uncomplicated malaria has been a major advance in antimalarial therapeutics. These drugs have reduced treated mortality, accelerated recovery and reduced treatment failure rates and transmission from the treated infection.
Artemisinin derivatives remain highly effective against falciparum malaria in most malaria endemic areas, but significant resistance has emerged in the Greater Mekong subregion of Southeast Asia. Resistance to artemisinins was followed by resistance to the ACT partner drugs, and fit multidrug resistant parasite lineages have now spread widely across the region. ACTs remain highly effective against P. vivax and the other malaria species. Recent studies have shown that radical curative regimens of primaquine (to prevent relapse) can be shortened to 7 days, and that the newly introduced single dose tafenoquine is an alternative, although the currently recommended dose is insufficient in Southeast Asia and Oceania.
Targeted malaria elimination using focal mass treatments with dihydroartemisininpiperaquine have proved safe and effective malaria elimination accelerators, but progress overall towards malaria elimination is slow. Indeed since 2015 overall malaria case numbers globally have risen. As new drugs will not become widely available in the near future, active measures to preserve the current antimalarials should be given the highest priority.

K E Y W O R D S
antimalarial drugs, artemisinin, malaria, resistance

| INTRODUCTION
The treatment of malaria has improved substantially in the past 15 years, and morbidity and mortality have declined as a result, but significant challenges lie ahead. 1 The major advance in antimalarial therapeutics has been the deployment of drugs derived from artemisinin (qinghaosu). 2 This unusual compound (a sesquiterpene lactone peroxide) is derived from the leaves of the plant Artemesia annua.
The derivatives of artemisinin, dihydroartemisinin (DHA), artesunate and artemether, now form the cornerstone of current antimalarial treatment. They are the most rapidly acting of the available antimalarial drugs and they are very well tolerated, but resistance has now emerged in Southeast Asia, and it has spread, and there are worrying early reports of foci in other regions. The artemisinins are partnered in fixed-dose combinations (artemisinin combination therapies, ACTs) with more slowly eliminated antimalarials for the treatment of uncomplicated malaria. New antimalarial drugs are on the horizon, but they are unlikely to become generally available within the next few years, so treatments now and in the immediate future must rely upon the artemisinin derivatives. This review presents some of the recent advances in antimalarial therapeutics and some of the obstacles to progress in controlling and eliminating malaria.
1.1 | Advance 1: Improvements in the treatment of severe malaria In the two largest randomized controlled trials conducted in patients hospitalized with severe falciparum malaria artesunate was shown to reduce mortality substantially. Compared to parenteral quinine, the previous first-line treatment, the mortality reduction (95% confidence interval) was 34.7% (18Á5-47Á6%) in Southeast Asian adults and children and 22.5% (8Á1-36Á9%) in African children 3,4 (Figure 1). In African children it can be difficult to distinguish severe malaria from bacterial sepsis with incidental parasitaemia. In those children with a high likelihood of having severe malaria, based on parasite biomass estimation, the reduction in mortality was the same (ie, one-third) as observed in the Asian series. 7 Artesunate was not more expensive than parenteral quinine, it was better tolerated (less hypoglycaemia), and it was easier to administer (intravenous injection rather than controlled rate infusion, and no pain or local toxicity following intramuscular injection).
Importantly there were also fewer neurological sequelae in the survivors, so lives were not saved by artesunate at the expense of neurological deficits. Artemether is nearly as active as artesunate in vitro against Plasmodium falciparum but, being an oil-based intramuscular injection, is slowly and erratically absorbed from the intramuscular injection site in vivo (particularly in shocked patients). 8 In contrast the water-soluble artesunate is rapidly and reliably absorbed after intramuscular injection. This delay in reaching parasiticidal blood concentrations probably explains why the severe malaria mortality following artemether in randomized trials was higher than that following artesunate treatment 5,6 ( Figure 1). Community based pre-referral rectal artesunate was also shown to reduce malaria attributable mortality by 25% in children who were unable to take oral antimalarial medications. 9 Since these trials reported over 10 years ago parenteral artesunate has become the generally recommended first-line treatment for severe malaria 1 and usage has increased substantially, although unfortunately quinine is still the only available parenteral antimalarial in some malaria endemic areas.
Following drug administration the artemisinin derivatives are rapidly and reliably converted back to DHA in vivo, which is then eliminated very rapidly (t 1/2 ≈ 1 hour), mainly by glucuronidation. 10 Despite this the drugs are highly efficacious when given once daily. 11 The main pathological process in severe falciparum malaria is the sequestration of erythrocytes containing mature forms of the parasite in the vascular beds of vital organs. 7,12 This reduces microcirculatory blood flow and probably markedly disturbs endothelial function. The key pharmacodynamic advantage of the artemisinin derivatives, which mediates their life-saving advantage over quinine, is the killing of the younger circulating stages of P. falciparum before they sequester. 12 Unfortunately, this property is substantially reduced in artemisinin resistance (see below Obstacle 2: Artemisnin resistance).
Apart from the prompt initiation of renal replacement therapies (preferably haemofiltration) in acute kidney injury, 12,13 no adjuvant therapies in severe malaria have proved to be beneficial, and many (including aspirin, corticosteroids, heparin, mannitol, high dose phenobarbitone, anti-TNF antibody and rapid fluid loading) were found to be harmful. 12 All except artesunate-SP are available in combined formulations, and all but artemether-lumefantrine are taken once daily. These drugs are rapidly effective and generally well tolerated. 1,10,16 Early concerns over potential neurotoxicity and teratogenicity have receded with increasing evidence of safety. 12 Worries over piperaquine cardiotoxicity (QT prolongation, risk of Torsade de Pointes) have also declined, with a large meta-analysis showing no increase in the rate of sudden death. 17 ACTs are now recommended as first-line treatments for all patients with falciparum malaria, including in pregnancy. 1 They are an alternative to chloroquine for infections caused by the other malaria species, allowing a single treatment to be deployed for all malaria infections. Costs have been reduced and generics developed.
Hundreds of millions of treatments are dispensed annually.
The main current concerns are ensuring access to diagnosis and effective treatment, and containing emerging resistance in P. falciparum to the artemisinin derivatives. Artemisinin resistance (see below Obstacle 2: Artemisinin resistance) manifests as slowing of parasite clearance because of reduced ring stage (the younger asexual forms) parasite susceptibility. 18 The discovery of a molecular F I G U R E 1 Mortality by treatment arms in randomized comparative controlled trials in strictly defined severe falciparum malaria (which together enrolled 2874 adults and 7424 children). The size of the circle is approximately proportional to the size of the trial and the error bars are 95% confidence intervals. The adults were enrolled mainly in Southeast Asia and the children mainly in Africa 3-6 marker, mutations in the propeller region of the P. falciparum kelch gene on chromosome 13, has greatly facilitated characterization and epidemiological assessments. 19,20 Reduced parasite killing in artemisinin-resistant malaria infections places greater selective "pressure" on the ACT partner drug. This is because the number of parasites which remain in the circulation after the artemisinin component in an ACT has been eliminated is many orders of magnitude greater, and the probability of selecting resistant mutants is correspondingly higher (Figure 2A). Indeed, ACT partner drug resistance has followed artemisinin resistance in the Greater Mekong subregion (GMS) of Southeast Asia. [21][22][23][24] Fortunately artemisininresistant P. falciparum are still largely confined to this one region, 25 although there are increasing reports that clusters of kelch mutant parasites have been identified elsewhere. 26,27 One potential therapeutic solution, which could be deployed now, is to combine an artemisinin derivative with two slowly eliminated antimalarials (triple artemisinin combination treatments, TACTs). 23 This solves the pharmacokinetic mismatch whereby the rapidly eliminated artemisinin component leaves the slowly eliminated partner drug "unprotected" for days or weeks after the second post-treatment asexual parasite cycle (ie, >3 days after starting the ACT). With TACTs there are now two slowly eliminated partner drugs providing mutual protection against the selection of resistance ( Figure 2B). The two TACTs under current development, artemether-lumefantrine-amodiaquine and dihydroartemisinin-piperaquine-mefloquine, exploit fortuitous reciprocal susceptibilities whereby resistance to one of the slowly eliminated components is associated with increased susceptibility to the other. In large-scale trials TACTs have proved well tolerated, safe and highly effective. 24

| Advance 3: Chemoprevention in malaria endemic areas
For many years pregnant women living in a malaria endemic region were advised to take chloroquine chemoprophylaxis to reduce the adverse effects of falciparum malaria on the developing foetus (mainly low birthweight). Then, as chloroquine resistance worsened, chemoprophylaxis in Africa was replaced by intermittent presumptive treatment with sulphadoxine-pyrimethamine (IPT-SP). 1 This imperfect chemoprophylaxis involves giving full treatment doses at intervals of 1 month or more. Although preventive efficacy is much greater than treatment efficacy against resistant P. falciparum, IPT-SP is threatened by worsening resistance, both to the antifol and the sulphonamide components. 28 There is increasing evidence that dihydroartemisininpiperaquine (DP) provides excellent antimalarial chemoprevention for approximately 1 month, is well tolerated and appears to be safe in pregnancy. 29,30 To provide continuous suppressive prophylaxis DP needs to be given at least monthly and preferably weekly. 31 The IPT-SP concept has also been advocated in infants, where treatment doses of SP are to be given together with the routine expanded programme on immunization (EPI) vaccines at the ages of 2, 3 and 9 months. This is not widely practiced as the benefits are relatively small and SP resistance is widespread. A more effective strategy, which is now implemented widely across the Sahel region of Africa (where intense malaria transmission is largely confined to the 3-4-month rainy season), is seasonal malaria chemoprevention (SMC). This is monthly administration of treatment doses of amodiaquine together with SP to all children aged between 6 and 59 months. 1,32 SMC prevents symptomatic reinfections and substantially reduces the F I G U R E 2 In artemisinin-sensitive malaria infections (AS) the 3-day artemisinin regimen in ACTs (A), results in rapid parasite killing and consequent decline in parasitaemia. The logarithmic scale vertical axis shows the total number of malaria parasites in the body of an adult with approximately 2% parasitaemia. The ACT partner drug has only approximately 1000 parasites to remove in this example (green triangle). In contrast, in an artemisinin-resistant infection (AR) there is substantially less parasite killing initially and so the ACT partner drug now has approximately 100 million parasites to remove (alone) with a substantially greater risk of treatment failure (recrudescence) and thus selective pressure to the emergence of partner drug resistance. In B, with TACTs there are now two slowly eliminated drugs providing a potentially greater antimalarial effect in resistant infections and ensuring mutual protection against the emergence of resistance. The detection limit (black dotted line) is the limit for microscopy to identify a malaria infection malaria burden. Mass drug administrations with azithromycin have been associated with reduced all-cause mortality in African children, particularly in Niger. 33 However, adding azithromycin to SMC with amodiaquine and SP in West Africa was shown to provide no additional benefit. 34 Resistance to both components of SMC is widespread in East Africa, but whether resistance is impacting on the chemoprophylactic activity of SMC is uncertain currently. More information is needed on this critical point to guide policy.

| Advance 4: Mass treatment as a malaria elimination accelerator
Where malaria transmission is low, the prospects for elimination increase. In the GMS, an area of low seasonal malaria transmission which harbours the most drug-resistant P. falciparum in the world, there is a general consensus that the only way to counter multidrug resistance effectively is to eliminate all falciparum malaria. The urgency to counter the threat of artemisinin resistance in this region has prompted evaluation of radical approaches to eliminate malaria which have relevance for other areas of the world with low malaria transmission. Targeted malaria elimination, even in the most remote and inaccessible areas, has been very effective. 35,36 The key to successful elimination is the support of village health workers in every village (usually 300-800 people) to provide diagnosis of malaria with a rapid diagnostic test and then treatment with an effective ACT. 37 In foci of higher transmission (sometimes called "hot spots"), where a significant proportion of the healthy population have asymptomatic parasitaemias, mass treatments with dihydroartemisinin-piperaquine have proved very effective and well tolerated "accelerators" of elimination. 35,36 Mass screen and treat approaches are less effective and they are not recommended. 38 However, whether these aggressive and more effective approaches to malaria elimination will be deployed by malaria control programmes remains to be seen. Multidrug-resistant P. falciparum is also prevalent in parts of South America. In Africa the most resistant P. falciparum parasites are found in Rwanda and Uganda. It is remarkable how drug resistance keeps emerging from these two relatively small areas of Asia and Africa. In general P. falciparum in Africa is more drug-sensitive, with higher levels of resistance in East compared with West Africa. 39 Resistance is generally less in the other malarias, although antifol resistance in P. vivax is widespread, and high levels of chloroquine resistance in P. vivax are found throughout Indonesia and Papua New Guinea. 40 Antifol resistance in both P. falciparum and P. vivax results from stepwise accumulation of mutations in the dhfr gene which encodes the drug target dihydrofolate reductase (S108N, N51I, C59R). Sulphonamide resistance results from accumulation of mutations in the dhps gene encoding the drug target dihydropteroate synthase (A437G, K540E, A581G). In general, the more of these mutations there are, the more resistant is the P. falciparum infection. 41 The highest level of antifol resistance is conferred by the Pfdhfr I164L mutation (found in Southeast Asia and South America). This mutation renders parasites completely resistant to pyrimethamine. Fortunately, except for Rwanda and adjacent Uganda, the 164 mutation has not established in Africa. 42 Resistance to chloroquine and the structurally related antimalarials which interfere with haem detoxification results from mutations in the transporter Pfcrt, and to a lesser extent mutations in Pfmdr There is serious concern that these "fit" multidrug resistant parasites could spread westward, or that artemisinin resistance could emerge de novo elsewhere and derail global aspirations to control and eliminate malaria.

| Obstacle 3: Underuse of primaquine
Primaquine is a very important antimalarial. It is recommended both as a single dose gametocytocide in falciparum malaria and in multipledose "radical cure" regimens to prevent relapse in vivax and ovale malaria. 1 But primaquine is underused. This is because of concerns over haemolytic toxicity in glucose-6-phosphate dehydrogenase (G6PD) deficiency. 46 Gene frequencies for the X-linked G6PD deficiency average 8-10% in tropical areas (although because G6PD deficiency protects against P. vivax infections prevalences are lower in patients presenting with vivax malaria). Unfortunately, screening tests to identify G6PD deficient patients are not widely available.
Relapses are recurrences of P. vivax or P. ovale malaria which follow complete cure of the blood stage infection. They derive from dormant parasite forms (called hypnozoites) which persist in the liver.
Hypnozoites are resistant to all current antimalarial drugs except the 8-aminoquinolines. 1 Without radical cure relapse rates vary between 20% and 80%. They are often multiple, and they are a major cause of morbidity and mortality in higher transmission settings. 47,48 Primaquine has usually been given in 7-or 14-day "radical cure" courses. As these regimens cause predictable haemolysis in G6PD deficient patients, G6PD testing is recommended before starting the radical cure regimen. 1 The recent development of rapid G6PD deficiency screening tests is a significant advance which should enable wider safe use of primaquine for radical cure, and thereby make elimination a more achievable target. 49 Recent very large studies confirm that the treatment durations, even for the higher dose primaquine regimens (total 7 mg/kg), can be condensed into a 1-week course.
With G6PD testing to exclude deficient patients, these are well tolerated. If these treatments are adhered to, radical curative efficacy is very high (>95%). [50][51][52] F I G U R E 3 The parasite clearance half-lives (PC 50 ) associated with Pfkelch mutations in patients with acute falciparum malaria studied in the TRAC1 study. 20 WT, wild type (note parasite clearance half-life estimates can still exceed 5 hours in Pfkelch wild-type infections). Mutations in the "propeller "region are usually associated with slow parasite clearance, the phenotypic hallmark of artemisinin resistance, although there is substantial interindividual variation and some mutations (A578S, pink arrow) are clearly not associated with artemisinin resistance. In the GMS parasite lineages associated with the F446I mutation have spread widely in Myanmar, and a lineage associated with C580Y was common along the Thailand-Myanmar border before targeted elimination activities. In the Eastern GMS lineages associated with R539T and C580Y both spread, but in recent years a C580Y lineage (termed PfPailin) has dominated 44 Figure 5). Unfortunately, there are no plans currently to rectify this.

| Obstacle 4: Medicine quality
Poor medicine quality is often ignored in discussions of disease control but the problem is massive, and it affects particularly the antimalarial drugs. In many countries the private sector is the main source of antimalarials and there is weak regulation of pharmaceuticals. 56 A recent systematic review and meta-analysis estimated that 12.4% of antibiotics and 19.1% of antimalarials in low-income and F I G U R E 4 The spread of artemisininresistant P. falciparum parasite lineages across the GMS. A single long pfKelch C580Y haplotype (from −50 to +31.5 kb either side of the Pfkelch gene), which emerged in Western Cambodia in or before 2008 (PfPailin), has spread across the Eastern GMS. In Myanmar C580Y parasites of a different lineage have spread widely but not dominated and a single pfKelch F446I haplotype, which probably originated in the north of Myanmar, has spread widely across the country. Modified from Imwong et al. 45 with permission middle-income countries were substandard or falsified, with an estimated economic impact ranging from US$10 billion to $200 billion. 57 Clearly this major obstacle to malaria control and elimination deserves more attention. increased steadily since then. 58 There has been no in-depth analysis to explain this reversal, and no clear evidence that providing more funding without reforms will reverse this trend. Most of these drugs are in phase 2 testing. If some of these compounds do proceed successfully through phase 3 studies and regulatory approval, likely in combinations, and these new combination therapies are well tolerated, effective and affordable, then they will be a welcome addition to the antimalarial armamentarium, but this is will not happen in the next few years.

| CONCLUSION
There have been substantial advances in the control of malaria in the past 20 years. Insecticide-treated bed nets and artemisinin combination treatments account for the majority of the millions of lives saved.
Unfortunately, the promise of a highly effective and long-lasting F I G U R E 5 Radical cure of vivax malaria with tafenoquine. Individual patient meta-analysis 55 of freedom from recurrence of P. vivax malaria (relapse prevention) in the two tafenoquine pivotal phase 3 studies in adults. These compared tafenoquine single dose (300 mg) with a low-dose primaquine regimen (15 mg base day for 14 days). 54,55 The dashed vertical line represents the prespecified noninferiority margin of an odds ratio for recurrence of 1.45 (tafenoquine vs primaquine). In Southeast Asia, which has high relapse rates, tafenoquine was significantly inferior (orange highlighting) to the low-dose primaquine regimen (which is considered inferior to a high-dose primaquine regimen of 30 mg base/day). Modified from Llanos-Cuentas et al. 55 with permission malaria vaccine, which dominated malaria research from the 1980s onwards, has not been fulfilled. Furthermore, with success in malaria control the sense of urgency has diminished and progress in many areas has stagnated. It is critically important that the efficacy of the current generation of antimalarial drugs is preserved for as long as possible, as the new compounds in clinical development are still several years from widescale deployment, and even then there is no guarantee that they will be as safe and effective as the drugs we have now.

| Nomenclature of targets and ligands
Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY.