Background
Foreword
Since the first Cochrane Review on artemether-lumefantrine was published (Omari 2002), the six-dose regimen treatment has become the standard, since the four-dose regimen is acknowledged to be associated with treatment failures (Nosten 2003). Trials are generally using the six-dose regimen, and the evidence for this regimen is maintained in a separate review (Omari 2005). This review evaluates the four-dose regimen. We do not intend to update this review as the question has been answered.
Malaria
Malaria is a major health problem with at least 300 to 500 million people diagnosed with the illness every year (WHO 2000a). The main cause is Plasmodium falciparum, one of the four species of malaria parasites found in humans. Uncomplicated malaria occurs in the majority of those affected, and is the form of the illness which presents with such symptoms as fever, headache, muscle pain (myalgia), vomiting, mild diarrhoea, anaemia, and enlarged spleen (splenomegaly). Children also commonly present with rapid breathing (tachypnoea), cough, and convulsions.
Antimalarial drug resistance
The resistance of malaria parasites to antimalarials emerged in South-East Asia and South America (White 1999a), and then spread to Africa and western Oceania. Sulfadoxine-pyrimethamine has replaced chloroquine as the first-line treatment in some African countries (such as Malawi and Kenya), but resistance to this is also emerging (WHO 2000a). Resistance to sulfadoxine-pyrimethamine is relatively common in South-East Asia (WHO 2001b), where resistance and declining sensitivity to mefloquine have also been reported (WHO 2000a). Mefloquine is contraindicated in areas of intensive malaria transmission, such as sub-Saharan Africa, because its long half life may expose parasites to subcurative doses that could result in the development of resistant strains (WHO 2000a).
Artemisinin drugs, including artemether and artesunate, are now used as first-line treatments in some countries in South-East Asia, but they are recommended only as combination treatment (WHO 2000a). Such combination therapy affords rapid clinical response and higher cure rates when compared with other antimalarial combinations (White 1999a). Combination therapy may slow the development of resistance (White 1999b).
Artemether-lumefantrine combination
The fixed-dose combination of artemether-lumefantrine, called co-artemether, contains 20 mg of artemether and 120 mg of lumefantrine (previously called benflumetol). It was initially developed by scientists at the Academy of Military Medical Sciences in China before the pharmaceutical company Novartis (Switzerland) became a partner and was licensed to market it as Coartem® or Riamet®. This oral preparation is designed for use against chloroquine-resistant falciparum malaria. Artemether has a rapid onset of action and the short half life of two to three hours (Lefèvre 1999) means that it is rapidly eliminated from the plasma. Lumefantrine is cleared more slowly and has a longer elimination half life of about 4.5 days (Ezzet 1998). The rationale behind this combination is that artemether initially provides rapid symptomatic relief by reducing the number of parasites present before lumefantrine eliminates any residual parasites. This is thought to minimize development of resistance because the malaria parasites are never exposed to artemether alone (due to its rapid elimination). Although they may be exposed to lumefantrine alone, the probability of resistance developing simultaneously to both drugs used in combination is thought to be low (Bloland 2000). Artemether-lumefantrine also reduces gametocyte carriage and thus should have an impact on malaria transmission (Van Vugt 1998b).
There has been some concern about the possible risk of neurotoxicity with artemisinin derivatives arose from animal studies using high doses of lipid-soluble preparations given intramuscularly (WHO 1999). No serious adverse or persistent neurotoxic adverse events have been documented (Novartis 2005). There has been concern that the lumefantrine component could have adverse cardiac effects due to its similar structure to halofantrine (Bindschedler 2000). Artemether-lumefantrine causes minimal QTc prolongation, but it was not associated with adverse clinical cardiac events (Novartis 2005). These potential adverse effects have to be considered when assessing the drug combination.
Artemether-lumefantrine has been added to the WHO Model List of Essential Medicines and is being promoted in Africa as first-line treatment for malaria by the World Health Organization. The World Health Organization has commended Novartis for providing the drug at discounted prices for developing countries in malaria endemic areas (WHO 2001a).
This review aims to summarize the existing evidence of different doses of artemether-lumefantrine and how it compares with other antimalarial drugs for treating uncomplicated falciparum malaria, including mefloquine, sulfadoxine-pyrimethamine, and chloroquine.
For our primary outcome measure, we use total failure by day 28 as the primary outcome measure or day 42 for sulfadoxine-pyrimethamine and day 63 for mefloquine because of their long half lives. In areas where malaria transmission is intense, recurrence of parasites by day 28 could also be due to re-infection, so we also examine the polymerase chain reaction (PCR), which is thought to distinguish between a new infection and a recurrence of malaria (recrudescence) due to drug resistance.
Objectives
To evaluate the four-dose regimen of artemether-lumefantrine for treating uncomplicated falciparum malaria.
Methods
Criteria for considering studies for this review
Types of studies
Randomized controlled trials.
Types of participants
Adults and children with acute uncomplicated malaria, as defined in WHO 2000b, with asexual P. falciparum parasitaemia confirmed from blood slides.
Types of interventions
Intervention
Four doses of artemether-lumefantrine.
Control
- Standard treatment regimens (single drug or combination).
- Six doses of artemether-lumefantrine.
Types of outcome measures
Primary
Total failure by day 28, day 42 (for sulfadoxine-pyrimethamine), or day 63 (for mefloquine); defined as a recurrent malaria infection with or without clinical malaria.
Secondary
- Total failure adjusted by PCR to exclude new infections (recrudescent infections) by day 28.
- Parasite clearance time (PCT), defined as the time between commencing treatment and the first negative blood test when negativity persists for more than 48 hours; PCT 50, defined as the time taken for parasites to be reduced to 50% of first test value; and PCT 90, defined as the time taken for parasites to be reduced to 10% of first test value.
- Fever clearance time, defined as the time between starting treatment and the temperature returning back to normal, and remaining normal for more than 48 hours.
- Gametocyte carriage on days 14 and 28.
- Gametocyte clearance time, defined as the time taken for gametocytes to disappear (if present in the blood initially) after commencing treatment.
Adverse events
- Adverse events requiring discontinuation of treatment or are fatal, life-threatening, or require hospitalization.
- Other adverse events (all other adverse events).
Search methods for identification of studies
We attempted to identify all relevant trials regardless of language or publication status (published, unpublished, in press, and in progress).
Databases
We searched the following databases using the search terms and strategy described in Table 1: Cochrane Infectious Diseases Group Specialized Register (October 2005); Cochrane Central Register of Controlled Trials (CENTRAL), published in The Cochrane Library (2005, Issue 3); MEDLINE (1966 to October 2005); EMBASE (1974 to October 2005); and LILACS (1982 to October 2005).
Conference proceedings
We searched the following conference proceedings for relevant abstracts: The Third Multilateral Initiative on Malaria Pan-African Conference, 18 to 22 November 2002, Arusha, Tanzania; and the Second European Congress on Tropical Medicine (Liverpool, UK, 14 to 18 September 1998).
Researchers, organizations, and pharmaceutical companies
We contacted researchers working in the field, the World Health Organization, and the pharmaceutical company Novartis for unpublished and ongoing trials.
Data collection and analysis
Selection of studies
Aika Omari (AO) screened the results of the search strategy to identify potentially relevant trials. AO and Carrol Gamble (CG) independently assessed the eligibility of these trials for inclusion in the review using the stated inclusion criteria. Any differences in opinion between the authors were discussed with the third author (PG).
Data extraction and management
AO and CG independently extracted data of trial characteristics including methods, participants, interventions, and outcomes, and recorded the data on standard forms. Where data from the published papers were insufficient or missing, we contacted the trial authors for additional information.
Where possible, we extracted data to allow an intention-to-treat analysis (the analysis should include all the participants in the groups to which they were originally randomly assigned). This approach is considered to be more pragmatic as it attempts to estimate the effectiveness of the treatment in routine practice rather than in the context of a clinical trial. To allow the intention-to-treat principle to be applied, all participants should be followed for the duration of the trial irrespective of whether or not the treatment course was completed or other protocol deviations. Any reason for dropping out of the trial or being excluded from the trial should be documented (WHO 1996).
If the number of participants randomized and the numbers analysed were inconsistent, we calculated the percentage loss to follow up. For dichotomous outcomes, we recorded the number of participants experiencing the event in each group of the trial. For continuous outcomes, we extracted arithmetic means and standard deviations and combined means using mean difference for each group where possible. If the data were reported using geometric means, we extracted standard deviations on the log scale, and extracted and reported the medians and ranges.
Assessment of risk of bias in included studies
We assessed the generation of allocation sequence and concealment of allocation as adequate, inadequate, or unclear according to Jüni 2001. We described whether the trial was open, single blind, or double blind. We assessed the inclusion of all randomized participants in the analysis of the reported primary outcome to be adequate if 90% or more were included in the analysis, inadequate if less than 90%, or unclear.
Data synthesis
We analysed data using Review Manager 5. We compared outcome measures for dichotomous data using risk ratio (RR), which is the risk of achieving an outcome in the artemether-lumefantrine group relative to that in the control group, and 95% confidence intervals (CI). We used total failure (clinical or parasitological failure by day 28) as our main outcome, and we also conducted analysis excluding re-infection where PCR data were available. As the value of the risk ratio is constrained to lie between 0 and 1/CGER (control group event rate), large values of the risk ratio are impossible when events are common, so failure is preferred to treatment success.
For total failure with trials that had conducted PCR analysis, we classified the infections into: 1. recrudescent infection (matching genotypes on day 0 and day of recurrence); 2. new infection (different genotypes on day 0 and day of recurrence); and 3. missing values. We conducted an analysis of parasitaemia corrected for PCR on days 14 and 28. We intended to conduct a sensitivity analysis around PCR examining the effect of missing data, but there were too few trials for us to do this.
We intended to explore age, level of malaria transmission, and level of drug resistance as potential sources of heterogeneity, and to conduct sensitivity analyses according to blinding, allocation concealment, whether the trials used an intention-to-treat analysis, and total failure adjusted by PCR to exclude new infections, but data were insufficient.
Results
Description of studies
See: Characteristics of included studies; Characteristics of excluded studies.
We identified 30 potentially relevant studies. Seven trials (2057 participants) met the inclusion criteria (see 'Characteristics of included studies'), with one trial, Hatz 1998, reported across two publications. We excluded 19 studies for the reasons given in the 'Characteristics of excluded studies'. We have requested data since 2003 on four studies from Novartis (cited in Novartis 1999) but have not yet received a response.
Trial location
Trials were conducted in The Gambia (Von Seidlein 1998), Tanzania (Hatz 1998), India (Kshirsagar 2000), travellers from the Tropics returning to France and the Netherlands (Van Agtmael 1999a), and Thailand (Van Vugt 1998a; Looareesuwan 1999; Van Vugt 1999).
Trial funding
Four trials reported that they were sponsored by Novartis (Van Vugt 1998a; Von Seidlein 1998; Looareesuwan 1999; Van Vugt 1999). The other three trials did not mention funding.
Participants
Two trials included 547 children aged one to five years (Hatz 1998; Von Seidlein 1998), two trials included 976 adults and children (Van Vugt 1998a; Van Vugt 1999), and three trials included 534 participants over 13 years of age (Looareesuwan 1999; Van Agtmael 1999a; Kshirsagar 2000).
Interventions
All trials gave the four doses of artemether-lumefantrine over 48 hours. Two trials compared it with chloroquine (Hatz 1998; Kshirsagar 2000). Other comparisons were with sulfadoxine-pyrimethamine (Von Seidlein 1998), mefloquine (Looareesuwan 1999), halofantrine (Van Agtmael 1999a), mefloquine plus artesunate (Van Vugt 1998a), and two six-dose regimens of artemether-lumefantrine (Van Vugt 1999).
Antimalarial drug resistance
Multiple-drug resistance was reported for three trials conducted in Thailand; one trial specified resistance to chloroquine, mefloquine, amodiaquine, and sulfadoxine-pyrimethamine (Looareesuwan 1999). Chloroquine resistance was reported in the Tanzanian trial, which was located in Ifakara where 20% of P. falciparum strains were resistant in 1988 (Hatz 1998). Kshirsagar 2000, which took place in Mumbai, India, was stopped early because the overall cure rate with chloroquine decreased to 50% due to a higher degree of chloroquine resistance present than expected; it was not clear from the trial report whether this was the result of a formal preplanned stopping rule. The Gambian trial, Von Seidlein 1998, was conducted in an area of chloroquine resistance and less than 5% of the P. falciparum strains were resistant to sulfadoxine-pyrimethamine.
Outcome measures
(See Table 2). Total failure by day 28 was the most frequently reported outcome (five of the seven studies). The results of PCR analyses used to exclude new infections were reported in five trials. Five trials reported on fever clearance times. Time to parasite clearance was reported in six trials: two trials reported it as a time-to-event analysis (Looareesuwan 1999; Van Agtmael 1999a); the other trials calculated the time to clearance on those participants in which the event occurred. Six trials reported on gametocyte carriage, and all seven trials reported adverse events.
Risk of bias in included studies
See Table 3 for the quality assessment and the 'Characteristics of included studies' for details.
Generation of allocation sequence
All the trials were reported as randomized. Four trials reported adequate generation of the allocation sequence. The remaining three trials mentioned randomization, but they did not report the methods used to generate the allocation sequence.
Concealment of allocation
Allocation concealment was adequate in the six trials that used a priori numbered or coded drug containers, central randomization, or sequentially numbered and sealed, opaque envelopes. One trial did not describe the method of allocation concealment (Van Agtmael 1999a).
Blinding
Four trials reported as double blind used the double dummy technique, while the other three trials were described as open. Three of the seven trials used adequate methods to generate the allocation sequence and conceal allocation, and used double blinding (Von Seidlein 1998; Looareesuwan 1999; Kshirsagar 2000).
Inclusion of randomized participants in the analysis
None of seven trials had complete data for all participants randomized into the trial for the duration of follow up. This was partly because researchers stopped follow up after a participant withdrew. An intention-to-treat analysis was therefore not possible for the trial investigators or for this review because the necessary data were not collected. All trials gave results of analyses based on evaluable participants, that is, participants still on treatment at each time point. Five of the trials, however, also claimed to have reported cure rates as an 'intention-to-treat' analysis (Hatz 1998; Von Seidlein 1998; Van Agtmael 1999a; Van Vugt 1999; Kshirsagar 2000). These are not the results of an intention-to-treat analysis and differed from their analysis of evaluable participants by assuming that all participants withdrawn from treatment or lost to follow up still had parasitaemia at all remaining time points. At the end of follow up, the number of participants evaluable for the primary outcome was more than 90% in one trial (Looareesuwan 1999), 90% to 85% in two trials (Von Seidlein 1998; Van Vugt 1999), and less than 85% in four trials (Hatz 1998; Van Vugt 1998a; Van Agtmael 1999a; Kshirsagar 2000).
Effects of interventions
1. Versus sulfadoxine-pyrimethamine (287 participants, 1 trial)
Total failure and parasite clearance time
Von Seidlein 1998, which was conducted in The Gambia, did not report any results for total failure by day 28 or 42. By day 14, there was no statistically significant difference in the number of participants with parasites between the two groups, but numbers were small (247 participants, Analysis 1.1). PCR results were available for six of the eight treatment failures in the artemether-lumefantrine group (all new infections), but no PCR results were available for the three treatment failures in the sulfadoxine-pyrimethamine group ( Table 4). Adjusting for new infections (with missing samples or failed tests classified as treatment failures) on day 14 did not change the statistical significance, but it did change the direction of the effect ( Analysis 1.2). However, parasite clearance time was faster with artemether-lumefantrine compared with sulfadoxine-pyrimethamine (PCT 50 mean 11.6 h, 144 participants; mean 21.1 h, 143 participants), although the trial authors reported that this difference was statistically significant (P < 0.0001, test not stated).
Gametocyte carriage
By day four there were fewer participants carrying gametocytes in the artemether-lumefantrine group than in the sulfadoxine-pyrimethamine group (115 participants, Analysis 1.3). A similar pattern was observed at day 15 (253 participants, Analysis 1.4).
Adverse events
Severe adverse events reported by the trialists are documented in Table 5.
2. Versus mefloquine (252 participants, 1 trial)
Total failure, and parasite, fever, and gametocyte clearance times
Looareesuwan 1999 reported that total failure was higher with artemether-lumefantrine by day 28 (RR 1.74, 95% CI 1.08 to 2.80; 233 participants, Analysis 2.1). However, parasite and fever clearance times were faster with artemether-lumefantrine ( Table 6 and Table 7). Gametocyte clearance was also faster with artemether-lumefantrine, but this was not statistically evident in the first 72 hours (252 participants, Analysis 2.2 and Table 8).
Adverse events
The majority of adverse events reported were mild or moderate ( Table 9), but there were reports of some severe adverse events ( Table 5). In particular, cardiac monitoring was reported with no difference in the QTc interval between the treatment groups.
3. Versus halofantrine (103 participants, 1 trial)
The participants in Van Agtmael 1999a were travellers returning to France and the Netherlands from the tropics.
Total failure
Although the number of participants failing treatment by day 28 was higher with artemether-lumefantrine, it was not statistically significant (86 participants, Analysis 3.1). The wide confidence interval reflects considerable imprecision, therefore clinically important results cannot be excluded in either direction.
Parasite clearance time
All participants on halofantrine cleared their parasites. The trial authors reported that parasite clearance time was statistically significant faster with artemether-lumefantrine ( Table 6).
Fever clearance time
Fever clearance was faster with artemether-lumefantrine, but this was not statistically significant ( Table 7).
Gametocyte carriage
There was no statistically significant difference between the two treatments within the first 72 hours (14 participants, Analysis 3.2).
Adverse events
Most of the reported adverse events were mild or moderate ( Table 9), but four separate severe adverse events were also documented ( Table 5). Some of the occurrences resulted in stopping treatment. The results of cardiac monitoring showed a significant increase in the overall QTc interval in participants treated with halofantrine (26/52) compared with artemether-lumefantrine (8/51).
4. Versus chloroquine (439 participants, 2 trials)
Kshirsagar 2000, a double blind trial, was conducted in an area of India with an unexpectedly high chloroquine resistance, which resulted in the trial being terminated before completion. Hatz 1998, which was an open trial in Tanzania, was also conducted in an area of high levels of chloroquine resistance.
Total failure
Both trials showed that artemether-lumefantrine was associated with fewer failures on day 28, but the size of this effect varied greatly between the two trials ( Analysis 4.1): Hatz 1998 reported a risk ratio of 0.38 (95% CI 0.30 to 0.49) with 237 participants, and Kshirsagar 2000 reported a risk ratio of 0.06 (95% CI 0.02 to 0.17) with 141 participants. The significant heterogeneity could be attributed to age, transmission, or degree of chloroquine resistance. As there is no benefit of increased precision for this meta-analysis, it is more useful to consider the individual trial results. On days seven and 14, Hatz 1998 reported fewer parasitological failures with artemether-lumefantrine.
The trial authors did PCR analyses on subsets of participants who failed treatment on days 14 and 28 ( Table 4). Less than half of the treatment failures in Hatz 1998 were tested, so any statistical inferences must be made with extreme caution. In Kshirsagar 2000, all treatment failures in the artemether-lumefantrine group were classified as recrudescent infections although no PCR was done.
Parasite clearance time and fever clearance time
Kshirsagar 2000 reported that parasite and fever clearance times were statistically significantly faster with artemether-lumefantrine ( Table 6 and Table 7).
Gametocytes
The pooled results from the two trials suggested a reduction in gametocyte carriage with artemether-lumefantrine within the first 72 hours (RR 0.79, 95% CI 0.63 to 0.99; 427 participants, Analysis 4.2). Gametocyte clearance was faster with artemether-lumefantrine compared with chloroquine ( Table 8).
Adverse events
Most reported adverse events were mild or moderate ( Table 9), but a few were severe ( Table 5). No deaths reported by either trial were associated with trial medication. Hatz 1998 reported occurrences of adverse events that resulted in stopping treatment. Kshirsagar 2000 reported cardiac monitoring with no difference in the QTc interval between treatment groups. One participant on halofantrine was withdrawn from treatment due to a prolonged QTc interval.
5. Versus mefloquine plus artesunate (617 participants, 1 trial)
Total failure
Van Vugt 1998a, conducted in Thailand, reported that more participants experienced total failure with artemether-lumefantrine at day 28 than with mefloquine plus artesunate (RR 6.77, 95% CI 3.12 to 14.67; 537 participants, Analysis 5.1). The results remained similar when new infections were excluded (RR 6.45, 95% CI 2.78 to 14.95; 537 participants, Analysis 5.2). The total failure on day 63 was also higher with artemether-lumefantrine (RR 2.67, 95% CI 1.55 to 4.57; 478 participants, Analysis 5.3).
PCR was conducted on 80% (72/89) of participants with parasites on day 63. There was a trend to fewer new infections in the artemether-lumefantrine group, but there were more also recrudescent infections in the artemether-lumefantrine group.
Parasite and fever clearance times
The trial authors reported the median parasite clearance times in both groups as two days (range: one to three) and the median fever clearance time as one day (range: one to four) with no difference between the treatment groups.
Gametocyte carriage
There was no statistically significant difference in gametocyte carriage within the first 72 hours between the two groups (RR 1.41, 95% CI 0.86 to 2.32; 616 participants, Analysis 5.4).
Adverse events
Van Vugt 1998a specifically sought serious neurotoxicity, but it reported no such adverse event.
6. Versus artemether-lumefantrine (six doses) (359 participants, 1 trial)
Van Vugt 1999, which was conducted in Thailand, compared the four-dose regimen of artemether-lumefantrine with two six-dose regimens given over 60 and 96 hours, respectively.
Total failure and parasite clearance time
The four-dose regimen had higher parasitological failure rates at day 28 compared with six-dose regimen (RR 7.71, 95% CI 2.99 to 19.88; 60 hour and 96 hour regimens combined, 306 participants, Analysis 6.1).
A PCR analysis was conducted during the 28-day follow up, but the results were reported according to the treatment site and not the treatment group. Twenty-four of the 25 treatment failures were evaluated by PCR genotyping. There were no re-infections in the 13 recurrent infections seen in the hospital, and four of the 11 recurrent malaria episodes at the campsite were new infections.
Parasite and fever clearance times
The median parasite clearance time was 44 hours for all three groups ( Table 6). There was no statistically significant difference in the median fever clearance times between the four-dose regimen and the two six-dose regimens ( Table 7).
Gametocytes
For comparisons with the six-dose regimens, 66 (18.4%) of all the participants had gametocytes detected during the first 72 hours. Gametocyte carriage was reported according to treatment centre. There was no statistically significant difference in the gametocyte clearance time in comparisons between the four-dose and six-dose regimens (P = 0.5; trial authors' calculation).
Adverse events
None were reported.
Discussion
Four of the seven trials described the methods used to generate the allocation sequence, and six trials described the concealment of allocation. Five trials reported an 'intention-to-treat' analysis for total failure, the primary outcome measure, but it was actually a limited form of sensitivity analysis because they made the assumption that all participants lost to follow up were treatment failures. As results were not based on an intention-to-treat analysis, they are subject to attrition bias and the clinical effectiveness may be biased.
Where PCR results were available, the total failure by day 14 and day 28 were corrected for new infections, with missing samples or failed tests classified as treatment failures. This did not materially affect the results for chloroquine and mefloquine plus artesunate, but it had a more substantial impact on sulfadoxine-pyrimethamine on day 14 by changing the direction of the effect to favour artemether-lumefantrine.
The four-dose regimen had significantly more treatment failures than the six-dose regimen. There was no significant difference in the parasite, fever, and gametocyte clearance times between the regimens, and the six-dose regimens were not associated with increased adverse events. The trial with the six-dose regimen as the comparator was conducted in an area with multiple-drug resistance, which is one of the recommendations for using this regimen. This review has shown that six-dose regimens are superior to the four-dose regimen in terms of total failure rates.
For parasitological outcomes, artemether-lumefantrine did not perform better than the other antimalarials apart from chloroquine. The trial comparing sulfadoxine-pyrimethamine with artemether-lumefantrine did not report on total failure on day 42. The fewer recrudescent and new infections seen in the sulfadoxine-pyrimethamine group may be explained by the long half life of sulfadoxine-pyrimethamine. Artemether-lumefantrine showed no marked advantage over mefloquine for parasitaemia in one trial of high quality (Looareesuwan 1999). Halofantrine had fewer treatment failures compared with artemether-lumefantrine, which was not statistically significant. However, this trial was small and the participants had varying degrees of immunity (they were from different countries), which could have confounded their response to medication (ter Kuile 1995).
Artemether-lumefantrine was only superior to chloroquine for total failure on day 28. However, there was a high withdrawal rate (more than 50% of the participants in the chloroquine group) in both trials, and the unexpected higher resistance to chloroquine decreased the cure rate to 50% in Kshirsagar 2000. This comparison would have been inappropriate had the actual chloroquine resistance been documented prior to commencing the trial since a clinical failure rate greater than 25% of a first-line antimalarial indicates the treatment should be changed (WHO 1994).
Parasite, fever, and gametocyte clearance times were shorter for artemether-lumefantrine when compared with the other antimalarial drugs, which suggests that clinical symptoms may resolve faster. With the exception of Looareesuwan 1999 and Van Agtmael 1999a, the trials reported clearance times as medians and percentiles, and not as time-to-event analyses. Time-to-event analyses are more informative as data on participants who did not reach the event would have been included in the analysis.
It is important to acknowledge that the trials were not designed to evaluate safety. Although some trials reported adverse cardiac events, the evidence was insufficient to address concerns about the possible risk of cardiotoxicity. We, therefore, cannot justifiably comment on adverse events reported apart from reporting the details.
Authors' conclusions
Implications for practice Artemether-lumefantrine showed no advantage in relation to total failure when compared with mefloquine, mefloquine plus artesunate, halofantrine, and sulfadoxine-pyrimethamine (result not adjusted for new infections). There is evidence to suggest the four-dose artemether-lumefantrine regimen is superior to chloroquine in areas where chloroquine resistance is common. In these circumstances, artemether-lumefantrine is an alternative to chloroquine as first-line treatment where policy makers wish to avoid sulfadoxine-pyrimethamine or other standard regimens. The six-dose artemether-lumefantrine regimen is superior to the four-dose regimen. 2008: This review will not be updated because the four-dose regimen is no longer recommended. Only the six-dose regimen, reviewed in Omari 2005, is recommended (WHO 2007). |
Implications for research Since the artemether-lumefantrine is no longer used as a four-dose regimen, there is no need for further research. 2008: This review will not be updated because the four-dose regimen is no longer recommended. Only the six-dose regimen, reviewed in Omari 2005, is recommended (WHO 2007). |
Acknowledgements
This document is an output from a project funded by the UK Department for International Development (DFID) for the benefit of developing countries. The views expressed are not necessarily those of DFID.
Data and analyses
- Top of page
- Background
- Objectives
- Methods
- Results
- Discussion
- Authors' conclusions
- Acknowledgements
- Data and analyses
- What's new
- History
- Contributions of authors
- Declarations of interest
- Sources of support
- Index terms
| |||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||
| |||||||||||||||||||||||||
| |||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||
What's new
Last assessed as up-to-date: 19 January 2006.
| |||||||||||||||
History
Protocol first published: Issue 3, 2001
Review first published: Issue 2, 2006
| |||||||||
Contributions of authors
Aika Omari and Carrol Gamble extracted and analysed data, and drafted the review. Paul Garner helped write the review.
Declarations of interest
Paul Garner and Carrol Gamble (né Preston) were unpaid technical advisers to a World Health Organization meeting (19 and 20 February 2001) considering efficacy and effectiveness studies of artemether-lumefantrine. The World Health Organization paid for their travel and accommodation, and a representative of Novartis chaired the international meeting.
Aika Omari: none known.
Sources of support
Internal sources
- Liverpool School of Tropical Medicine, UK.
External sources
- Department for International Development, UK.
Index terms
Medical Subject Headings (MeSH)
Antimalarials [adverse effects; *therapeutic use]; Artemisinins [adverse effects; *therapeutic use]; Drug Combinations; Fluorenes [adverse effects; *therapeutic use]; Malaria, Falciparum [*drug therapy]; Randomized Controlled Trials as Topic; Sesquiterpenes [adverse effects; *therapeutic use]; Treatment Failure
MeSH check words
Humans
* Indicates the major publication for the study