Artemisinin Combination Therapy for Malaria: Beyond Good Efficacy

In the first half of the 20th century, substantial reductions in the burden of malaria were achieved through aggressive vector control. These early successes were enhanced further in the 1950s by the widespread use of chloroquine. An enthusiastic malaria elimination program was started; however, within a decade the overwhelming challenges of achieving this ambitious goal and the resilience of the Plasmodium parasite had become all too apparent. The rest of the century witnessed a losing battle with the parasite as it developed resistance to chloroquine and then to other antimalarials. Although technological advances delivered new therapeutic agents, the efficacy of such agents was often short-lived because resistance rapidly emerged and spread through an increasingly connected world. The widespread use of partially effective treatment policies had devastating consequences for the malaria endemic world, resulting in an increasing burden of death and disease [1].

Prospects at the start of the 21st century are brighter. Huge financial resources have been made available to develop the infrastructure and resources required to tackle the parasite. One of the cornerstones of control programs today is the early diagnosis of malaria and its treatment with highly effective drugs. New combination therapies containing artemisinin derivatives are central to this approach, providing practical treatment regimens with high cure rates and transmission-blocking potential. The rationale is that the short-acting but highly potent artemisinin derivative delivers a rapid reduction in parasite biomass, with the remaining parasites being removed by the intrinsically less active but more slowly eliminated partner drug [2]. Endorsed by the World Health Organization, artemisinin-based combination therapies are now being used in >50 countries where malaria is endemic.

The most common combinations are artemether-lumefantrine, artesunate-amodiaquine, artesunate-sulfadoxine-pyrimethamine, artesunate-mefloquine, and dihydroartemisinin-piperaquine. But how does one choose among such regimens? Numerous factors come into play, including safety, tolerability, adherence, availability, coformulation, cost, and effectiveness [3]. Of the newer combinations, artemether-lumefantrine and dihydroartemisinin-piperaquine have been shown to be well tolerated, with excellent efficacy in almost all studies conducted. Both treatments are coformulated, facilitating adherence and protecting against misuse. The main difference between these antimalarial combinations is their duration of action. Lumefantrine has a terminal elimination half-life of ~4 days [4], whereas piperaquine has a terminal elimination half-life of 21–28 days [5]. Suppressive concentrations of the drugs last ~16 days and 28 days, respectively. The nature of this pharmacokinetic mismatching has both advantages and disadvantages.

Clinical trials in uncomplicated malaria usually report the overall risk of recurrent parasitemia and the polymerase chain reaction–adjusted risk of recrudescence. The latter is often used as the primary end point and the main indicator of a regimen’s local efficacy. However, the overall recurrence rate should not be ignored. After eradication of the asexual stages of the parasite from the peripheral blood, patients who remain in an endemic area are at risk of additional infection or, in areas where Plasmodium vivax is endemic, of relapses arising from liver stage hypnozoites. This risk can be considerable, with up to 50% of patients having recurrent infection within 28 days [6, 7]. Slowly eliminated antimalarial drugs (eg, piperaquine or mefloquine) exert a greater posttreatment prophylactic effect against additional infection and relapse than more rapidly eliminated drugs, such as lumefantrine. This explains why the overall treatment failure of artemether-lumefantrine and dihydroartemisinin-piperaquine, two highly effective antimalarial combinations, can differ so greatly in comparative drug trials [7–9]. If one continues to follow up patients beyond day 42, the survival curves for these 2 medications will eventually converge as the plasma concentrations decrease below levels able to suppress parasite growth. What, therefore, is the clinical benefit of this transient additional protection? A simple answer, often overlooked by academics, comes from the patients themselves when asked whether they would like their next bout of malaria in 3 weeks or 6 weeks. Delaying the next recurrence gives patients a longer period free of symptomatic malaria, allowing a greater time for hematologic recovery, which in turn reduces the cumulative risk of anemia [9]. Furthermore, in areas endemic for P. vivax, long-acting antimalarials provide the only practical means of delaying the first relapse. However, the magnitude of these benefits will be short-lived and dependent on the background risks of additional infection and relapse.

The public health implications of posttreatment prophylaxis can only be answered by studies with long follow-up, such as the study by Arinaitwe et al in this issue of Clinical Infectious Diseases [10]. Their study, conducted in Eastern Uganda, enrolled and randomized infants presenting to a local clinic to repeated treatment with artemether-lumefantrine or dihydroartemisinin-piperaquine. Children were followed up intensely for 28 days and then monthly for up to 1 year. In this area of high malaria transmission, the 232 children randomized experienced 671 episodes of uncomplicated malaria. As expected, the unadjusted risk of late parasitological failure at day 28 was significantly higher in the artemetherlumefantrine arm (20%), compared with the dihydroartemisinin-piperaquine arm (7.4%; hazard ratio, 3.6). These treatment failures were made up almost entirely of additional infections. By day 63, treatment failure had reached 60% and 63%, respectively, and although the difference was still significant, the hazard ratio had decreased to 1.4. During the year-long study period, no significant difference was found in the mean number of malaria episodes between infants randomized to artemether-lumefantrine and those randomized to dihydroartemisinin-piperaquine (4.82 vs 4.61; P = .63), a surprising result given the high level of malaria transmission in the study area and the significantly longer terminal elimination half-life of the dihydroartemisinin-piperaquine combination. All infants were given bed nets at the start of the study (the use of which was very high). Could this have reduced the force of infection for the participants? The overall incidence of malaria was <1 episode per 2 months, lower than one might expect given the duration of parasiticidal activity of both drugs used and the high level of transmission. Irrespective of this, no apparent differences were found in secondary end points, such as anemia rates or gametocyte carriage, although the study may have been underpowered to detect these. Therefore, the hypothesis that longer postexposure prophylaxis should be associated with a reduction in the risk of additional infection in highly endemic areas is not supported. The authors propose that the short-lived benefit of dihydroartemisinin-piperaquine over artemether-lumefantrine was overwhelmed by a deluge of recurrent infections.

This important study presents strong evidence that there is little to choose from between artemether-lumefantrine and dihydroartemisinin-piperaquine in such areas of high transmission. Whether this remains so in regions of more moderate transmission has yet to be tested, although studies in such areas suggest that dihydroartemisinin-piperaquine may offer sustained benefit at least until day 42 [9, 11]. However, any such benefits of greater postexposure prophylaxis must be balanced against the potential disadvantages of using a long half-life antimalarial. Drugs and combinations with substantial pharmacokinetic mismatching increase the risk of selecting drug-resistant isolates [12]. Although current theory proposes that combination therapies reduce the risk of de novo resistance, should a mutation conferring resistance to a long-acting partner drug occur, the prolonged subtherapeutic tail may enhance selective transmission and lead to the drug’s ultimate demise [13].

Arinaitwe and colleagues should be applauded for a meticulously designed and analyzed study that addresses an important question in a logistically challenging environment. Their study paves the way for longitudinal cohorts and population-level impact studies that address the practicalities of widespread deployment and recurrent exposure to different artemisinin-based combination therapies. As we move toward the elimination of malaria, it is crucial that policymakers and research groups endorse treatment regimens with high efficacy but do not ignore the additional implications of posttreatment prophylaxis.