Background

The rising threat of Plasmodium falciparum resistance to monotherapies prompted the World Health Organization (WHO) 2006 guidelines for the treatment of malaria to recommend that combinations of antimalarials be used to treat malaria caused by P. falciparum [1]. Since artemisinin derivatives are the only class of anti-malarial agents to which resistance outside the Thai-Cambodia border region has not been reported in vivo [1], the guidelines specifically recommend the use of artemisinin-based combination therapy (ACT) [1]. WHO criterion for adequate efficacy of ACT in malaria is the achievement of an average cure rate of ≥ 95% in clinical trials [1].

Artemisinin derivatives have the most potent and rapid onset of anti-parasitic activity of any anti-malarial drug available today and are active against all Plasmodium species that infect humans. Importantly, they allow more parasite clearance than any other anti-malarial drug (parasite numbers can be reduced by a factor of 105 per asexual cycle, compared with 102-103 with other anti-malarial drugs) [2]. When combined with efficacious anti-malarials with slower elimination rates, such as lumefantrine, shorter courses of treatment (three days) become effective [2].

The combination of artemether (an artemisinin derivative) and lumefantrine in a 1:6 ratio was the first fixed-dose ACT to meet the WHO's prequalification criteria for efficacy, safety and quality [3]. Artemether/lumefantrine (AL), as Coartem®, now comprises nearly 75% of the 100 million or so ACT treatments used each year [4]. Coartem® was approved by Swissmedic in 1999 and was recently approved by the USA FDA [5]. Coartem® Dispersible was approved by Swissmedic in December 2008. More than 40 malaria-endemic countries in sub-Saharan Africa have rapidly scaled up malaria prevention and treatment and now recommend the use of ACT as first-line treatment for uncomplicated falciparum malaria [6]. Both artemether and lumefantrine are blood schizonticides with complementary pharmacokinetics and dissimilar modes of action, and hence provide synergistic anti-malarial activity [7]. Artemether is rapidly eliminated from plasma with a half-life of two to three hours, whereas lumefantrine is eliminated more slowly with a half-life of three to six days and provides a high long-term cure rate after a short treatment course [7]. The combination thus provides rapid clearance of parasitemia and most malaria-related symptoms, coupled with prevention of recrudescence.

AL tablets have been included on the WHO model list of Essential Medicines since March 2002 and on the first WHO Model List of Essential Medicines for Children since October 2007 [8, 9].

The six-dose AL regimen is currently the approved treatment regimen for acute, uncomplicated P. falciparum malaria in adults and paediatric patients with a body weight ≥ 5 kg, irrespective of the immune status of the patients to P. falciparum and of the local multidrug resistance situation, in the majority of the 83 countries in Africa, Asia, Europe and Latin America where the drug is registered.

The objective of this paper is to review the literature investigating the drug's clinical efficacy. In this respect, it is important to note that early clinical studies used a four-dose AL regimen (two doses per day for two days).

However, because this regimen did not provide optimal efficacy in some areas, such as Thailand, where multi-drug resistant P. falciparum malaria is prevalent, it was changed to its current six-dose format (three-day course: dosing at 0, 8, 24, 36, 48 and 60 hours) in 1997. This article will focus on data involving the six-dose regimen only. In addition to the key clinical trials that support the drug's indication for the treatment of uncomplicated P. falciparum malaria, over 40 independent trials have produced data supporting the excellent efficacy of this ACT.

Study design considerations

The active comparators used in the controlled studies summarized below were the standard therapies for each country as recommended by the WHO at the time. Placebo-controlled studies are not practical for ethical reasons because untreated P. falciparum malaria may progress rapidly, with a potentially fatal outcome. All comparative studies were randomized, and double-blind designs were used where practical.

Efficacy evaluations

The key criterion used to assess the efficacy of an anti-malarial agent is the elimination of malaria parasites, which in turn leads to resolution of symptoms such as fever. Unless otherwise stated, the primary efficacy endpoint in the studies outlined below was the 28-day parasitological cure rate. This describes the proportion of patients with clearance of asexual parasitaemia within seven days of initiating study treatment without recrudescence at day 28, based on blood smears. In most studies, the 28-day cure rate was also corrected by polymerase chain reaction (PCR) to differentiate between recurrence of the initial infection and a new infection. This is particularly important in highly endemic regions, as is the case in many parts of sub-Saharan Africa, where reinfection is common.

Several secondary efficacy endpoints were also evaluated. These included: 7-day, 14-day, and/or 42-day parasitological cure rates; fever clearance time (FCT) in patients who had fever at baseline; parasite clearance time (PCT) and gametocyte clearance time (GCT).

The evaluable population included all patients with confirmed P. falciparum malaria who received at least one dose of study drug and had parasite counts performed at the pre-specified time points, including day 28, or who discontinued due to unsatisfactory therapeutic effect.

The primary efficacy endpoint was also evaluated in the modified intent-to-treat (mITT) population, which included all patients with confirmed P. falciparum malaria who received at least one dose of study drug. Since patients who did not have a parasite count performed at day 7 or day 28 were classified as a treatment failure in the mITT analysis, the evaluable population analysis gives the most clinically relevant result as it better reflects the true activity of the drug (i.e. it avoids classifying as treatment failure patients who did not have a parasite count performed but who may have been cured).

Efficacy of AL studied in children and adults

Six key clinical studies, summarized in Table 1[10–15], have been conducted to evaluate the efficacy of the six-dose AL regimen in a range of patient populations and geographic regions with varying levels of drug resistant P. falciparum and malaria endemicity. In all studies, patients were either non-immune or semi-immune to P. falciparum. Some of the studies included other anti-malarial drugs or combinations as active comparators and some also allowed the inclusion of patients with mixed infections that included P. falciparum at baseline.

Table 1 Clinical studies evaluating the efficacy and safety of the six-dose regimen of artemether/lumefantrine
Full size table

Study AO25

This randomized, double-blind study enrolled 359 adults and children above two years of age with uncomplicated P. falciparum in two centres in Thailand. It compared the AL four-dose regimen to two different six-dose regimens, one administered over 60 hours and one administered over 96 hours [10].

AL achieved a 28-day PCR-corrected cure rate of 96.9% in the evaluable population (Figure 1) [10]. All patients cleared peripheral parasitaemia rapidly - the median parasite clearance time was 44 hours [10]. Median fever clearance time was 35 hours, and no patient developed gametocytaemia following treatment [10].

Figure 1
figure 1

28-day PCR-corrected cure rate for artemether/lumefantrine in Studies A025,A026 and A028 [10–12].

Full size image

Since the efficacy was similar between the 96 and 60-hour six-dose regimens, the 60-hour regimen was chosen for further development to facilitate better compliance with a shorter treatment duration.

Studies A026 and A028

These two randomized, open-label studies were both conducted in Thailand and included mefloquine plus artesunate (MAS) as an active control arm [11, 12].

Study 026 enrolled patients ≥ 2 years of age [11], and Study 028 enrolled patients >12 years of age [12]. The open-label design for both studies was chosen because of the excessive number of placebo tablets (24) patients would have had to take if a double-blind/double-dummy design had been used.

In Study 026, 150 patients received AL and 50 patients received mefloquine and artesunate. Patient characteristics were comparable between treatment groups, and treatment comparisons were not affected by a difference in baseline parasite density. The 28-day PCR-corrected cure rates were >95% in the evaluable population for both treatment groups (Figure 1). The median fever and parasite clearance times were identical in the two treatment groups [11].

In Study 028, 164 patients received AL and 55 patients received MAS. Patient characteristics were similar to those seen in Study 026 and were comparable across treatment groups, except for parasite density which was slightly lower than in Study 026 [11, 12].

Once again, the 28-day cure rates were >95% in the evaluable population and were comparable in both treatment groups (Figure 1). The lower boundary of the confidence interval was greater than 85%. Median fever and parasite clearance times were in the range of 20 to 30 hours in both groups [11, 12].

Study A2403

This was an open-label non-comparative study performed in collaboration with the WHO. It was designed to examine the safety and efficacy of AL in 310 infants and young children (5-25 kg body weight) in sub-Saharan Africa (Kenya, Nigeria, and Tanzania) [13]. In the absence of an ethical comparator agent for paediatric patients with a body weight of 5 to <10 kg at the time of the study, it was a single-arm study. The median age was 24 months, ranging from two months to 10 years. AL tablets were crushed and mixed with sterile water to facilitate administration to those children who were unable to swallow whole tablets [13]. In all cases, treatment was given after feeding.

The overall 28-day cure rate, corrected for reinfection, was 93.9% for the intent-to-treat population [13]. The median time to parasite clearance ranged from 24 to 36 hours, with 98.4% of patients achieving parasite clearance within 48 hours of their first dose [13]. The median time to fever clearance was less than eight hours [13], and none of the children had gametocytes after day 14 [13].

Study B2303

To facilitate administration of AL to children, a sweet-tasting dispersible formulation has been developed. The primary objective of Study B2303, which enrolled 899 African children (5-<35 kg body weight), was to demonstrate non-inferiority of the dispersible tablet to the crushed tablet based on the 28-day PCR-corrected cure rate in the evaluable population. The median age of children enrolled was 36 months, ranging from two months to 12 years, and the median body weight was 13 kg [14].

In the evaluable population, the 28-day PCR-corrected cure rate for the group receiving crushed tablets was 98.5%, compared with 97.8% for the group receiving the dispersible formulation of AL (Coartem® Dispersible) [14]. The PCR-adjusted cure rates for the two groups were also comparable at day 14 and day 42 [14]. Notably, the cure rates were comparable across body weight groups (Figure 2) [14]. In addition, there were no significant differences between those receiving the dispersible formulation and those receiving crushed tablets with regard to median time to parasite clearance (Figure 3A) and median time to fever clearance (Figure 3B) [14].

Figure 2
figure 2

Efficacy of dispersible AL formulation across body weight groups [14].

Full size image
Figure 3
figure 3

Median time to (A) parasite clearance and (B) fever clearance for dispersible artemether/lumefantrine formulation and crushed tablets [14].

Full size image

Study A2401

This study was designed to assess the efficacy of AL in non-immune populations (e.g. travellers). To date, it is the largest study investigating a drug for the treatment of malaria in travellers. Because of practical considerations related to the availability of patients, it was an open-label, single-arm study. A total of 165 patients who had been diagnosed with non-endemic malaria were enrolled from Europe and non-malarious areas of Colombia. Non-immune patients were defined as those who had not spent the first five years of their life, nor the five years prior to study entry, in a malaria-endemic area, and had not had acute P. falciparum malaria diagnosed during the past 5 years.

A high 28-day cure rate of 96% (both uncorrected and PCR-corrected) was observed in the evaluable population (Figure 4) [15]. The median time to parasite clearance was 41.5 hours and the median time to fever clearance was 36.8 hours [15].

Figure 4
figure 4

28-day PCR-corrected cure rate for AL in Study A2401 [15].

Full size image

A total of 599 adult patients (>16 years) and 877 children (<16 years) were treated with the six-dose AL regimen in the above studies. The 28-day PCR-corrected cure rate in evaluable patients consistently met WHO criteria for efficacy (>95%) in both adults and children. Rapid clearance of parasitaemia and fever was also evident, although noticeably quicker in children.

Comparative results from independent studies

A large number of independent studies have compared the efficacy of AL with other anti-malarial treatments. These studies conclusively show that AL is highly efficacious and at least as effective as other ACTs and combinations of antimalarials over a wide range of geographical areas and in a variety of populations. A summary of these trials is presented in Tables 2 and 3.

Table 2 Published studies with the six-dose AL regimen in Africa
Full size table
Table 3 Published studies with the six-dose artemether/lumefantrine regimen in Asia
Full size table

Studies in Africa

In studies reported from Africa, summarized in Table 2[16–43], the six-dose AL regimen appeared to be at least as effective as most other forms of ACT included in comparative studies and was generally more effective than other, non-ACT, comparators. In the few studies where comparators were associated with lower risks of treatment failure than AL, the studies were typically conducted in areas with very intense malaria transmission, and the comparators included anti-malarials with a long half-life, which could prevent re-infection late in the study. Indeed, the lower crude failure rates with the comparators in these few studies are the result of lower re-infection rates - recrudescence rates were generally similar for AL and comparators.

Several of the publications report the potent effects of ACT treatment on gametocyte carriage. One study investigated gametocyte carriage in detail and it was found that in AL-treated patients, not only was the rate of carriage of gametocytes significantly reduced as compared with that in patients receiving CQSP, but the carriage time of gametocytes and the infectivity of the gametocytes for mosquitoes were also significantly decreased [37]. These findings potentially have considerable implications for public health, given the possibility of reducing transmission of malaria by extensive use of ACT.

Supervised vs unsupervised AL in acute, uncomplicated falciparum malaria

Piola et al [44] conducted a randomized trial to compare the efficacy, safety, and pharmacokinetics of AL when given in a supervised (all doses observed with fatty food intake; n = 313) or unsupervised (first dose supervised followed by outpatient treatment with nutritional advice; n = 644) setting to patients of all ages (weight >10 kg) with acute, uncomplicated falciparum malaria in Mbarara, Uganda. The primary endpoint was 28-day, PCR-adjusted, parasitological cure rate.

28-day cure rates were very similar (100% in both groups in an evaluable patient analysis). The only difference between supervised and unsupervised treatment was that blood lumefantrine levels in the unsupervised patients were statistically significantly lower (p < 0.001) at Days 3 and 7.

Studies in Asia

The studies reported from South-East Asia, summarized in Table 3[45–53], show that in general AL is associated with similar efficacy to other forms of ACT in the treatment of P. falciparum malaria, with high parasitological cure rates and rapid clearance of parasitaemia and resolution of fever. In one study in which P. vivax infections were included, the comparator ACT (DP) was significantly more effective in terms of reducing the risk of recurrent P. vivax infection, probably due to the long half-life of piperaquine offering protection against re-infection for longer than lumefantrine: in the same study, PCR-corrected P. falciparum parasitological cure rates were almost identical for AL and DP [50].

Studies from South Asia show high parasitological cure rates and rapid clearance of fever and parasitaemia with the six-dose AL regimen, comparable with those for MAS in the one study where this combination was used as a comparator. AL was more effective than non-ACT comparators in these studies.

Meta-analysis

A meta-analysis of 32 published randomized studies performed predominantly in Africa but also in South America and Asia has evaluated the efficacy of a number of different forms of ACT using a Bayesian random effects approach. The analysis showed that AL was one of the most effective ACT with a 28-day PCR-corrected parasitological cure rate of 97.4% (Table 4) [54].

Table 4 Results of a meta-analysis comparing the efficacy of several anti-malarial combinations [54]
Full size table

Safety and tolerability of artemether/lumefantrine

The safety and tolerability of AL has been confirmed in both adults and children in several studies [10–15]. These data support the use of a six-dose regimen of AL as a safe and well-tolerated treatment in a wide range of patient populations including children.

Safety and tolerability are covered in detail in a companion article in this supplement [55].

Conclusion

Artemether/lumefantrine has consistently met the WHO criteria for efficacy (>95%) in both adults and children in randomized, controlled trials in malaria-endemic regions [10–15]. Efficacy appeared to be maintained across different populations and regions (Tables 2 and 3); it was unaffected by bodyweight in infants and children and was comparable for dispersible and crushed tablets of AL [14].

Effectiveness studies comparing supervised and unsupervised treatment [44] also demonstrated comparable efficacy data. The 28-day uncorrected cure rate in the per protocol population was also >95% in non-immune adult travellers with P. falciparum malaria [15].

A large number of independent studies show the six-dose AL regimen is at least as effective as other artemisinin-based combination therapies and combinations of anti-malarials. A meta-analysis of 32 comparative trials showed AL to be one of the most effective ACTs currently available [54].

The combination of artemether and lumefantrine also has good gametocidal properties, which may help to limit the spread of resistance [10, 12]. To date, no resistance to AL in vivo has been reported in Africa [1].