Output... Links... Efficacy of artesunate pl...

Transcription

1 Efficacy of artesunate plus pyrimethamine-sulphadoxine for uncomplicated malaria in Gambian children: a double-blind, randomised, controlled trial [Articles] von Seidlein, Lorenz; Milligan, Paul; Pinder, Margaret; Bojang, Kalifa; Anyalebechi, Chukwudi; Gosling, Roland; Coleman, Rosalind; Ude, Justin Ifeanyichukwu; Sadiq, Abubakar; uraisingh, Manoj; Warhurst, avid; Alloueche, Ali; Targett, Geoffrey; McAdam, Keith; Greenwood, Brian; Walraven, Gijs; Olliaro, Piero; oherty, Tom Medical Research Council Laboratories, The Gambia (L von Seidlein M, P J M Milligan Ph, M Pinder Ph, K Bojang MRCP, C Anyalebechi MB, R Gosling M, R Coleman MRCP, J Ifeanyichukwu Ude MB, A Sadiq MB, Prof K P W J McAdam FRCP, G Walraven M, J F oherty M); London School of Hygiene and Tropical Medicine, London (M uraisingh Ph, Prof C Warhurst FRCPath, A Alloueche Ph, Prof G A T Targett Sc, Prof B M Greenwood FRS) WHO, Geneva (P Olliaro M) Correspondence to: Lorenz von Seidlein, Farafenni Field Station, Medical Research Council Laboratories, The Gambia ( lseidlein@mrc.gm)

2 Outline Summary Introduction Methods Patients Study design Laboratory methods Statistical analysis Results iscussion Contributors Acknowledgments References Graphics Figure 1: Trial prof... Table 1: Baseline ch... Table 2: Proportion... Figure 2: Proportion... Figure 3: Proportion... Summarytoctoc Background: Resistance to cheap effective antimalarial drugs, especially to pyrimethaminesulphadoxine (Fansidar), is likely to have a striking impact on childhood mortality in sub- Sharan Africa. The use of artesunate (sodium Output... Links... javascript:openweblink(' id.com:80/ovidweb.cgi?s=injhkajlbm GO00&HC=fulltext&HR=1','OvidHelp', 750) javascript:openweblink(' id.com:80/ovidweb.cgi?s=injhkajlbm GO00&HC=fulltext&HR=1','OvidHelp', 750) off=1&s=injhkajlbmgo00 off=1&s=injhkajlbmgo00 History... Efficacy of artesunate pl... artesunate) in combination with pyrimethamine-sulphadoxine may delay or prevent resistance. We investigated the efficacy, safety, and tolerability of this combined treatment. Methods: We did a double-blind, randomised, placebo-controlled trial in The Gambia. 600 children with acute uncomplicated Plasmodium falciparum malaria, aged 6 months to 10 years, at five health centres were randomly assigned pyrimethamine-sulphadoxine (25 mg/500 mg) with placebo; pyrimethamine-sulphadoxine plus one dose of artesunate (4mg/kg bodyweight); or pyrimethamine-sulphadoxine plus one dose 4 mg/kg bodyweight artesunate daily for 3 days. Children were visited at home each day after the start of treatment until parasitaemia had cleared. Findings: The combined treatment was well tolerated. No adverse reactions attributable to treatment were recorded. By day 1, only 178 (47%) of 381 children treated with artesunate were still parasitaemic, compared with 157 (81%) of 195 children in the pyrimethamine-sulphadoxine alone group (relative risk 1.7 [95% CI ], p<0.001). Treatment-failure rates at day 14 were 3.1% in the pyrimethamine-sulphadoxine alone group, and 3.7% in the one-dose artesunate group (risk difference -0.6% [-4.2 to 3.0]) and 1.6% in the three-dose group (1.5 [ ], p=0.048). Symptoms resolved faster in children who received artesunate, but there was no

3 additional benefit for three doses of artesunate over one close. Children given artesunate were less likely to be gametocytaemic after treatment. Interpretation: The combined treatment was safe, well tolerated, and effective. The addition of artesunate to malaria treatment regimens in Africa results in lower gametocyte rates and may lower transmission rates. Introductiontoctoc Resistance to cheap efficient antimalarial drugs poses an increasing threat in Africa.1 This increase is especially worrying in sub-saharan Africa, where many health services depend on chloroquine as the first-line treatment of uncomplicated malaria, despite chloroquine resistance becoming more common and being linked to an increase in childhood mortality.2,3 Pyrimethamine-sulphadoxine (25 mg/500 mg, Fansidar) is widely used in sub-saharan Africa as a replacement for chloroquine. The rapid emergence of resistance against pyrimethaminesulphadoxine, already seen in east Africa,4 is likely to have a striking impact on childhood mortality in many regions where no obvious replacement for pyrimethamine-sulphadoxine is available. Combination therapy against highly drug-resistant infectious diseases such as tuberculosis or HIV-1 infection has been accepted worldwide. A combination of mefloquine and artesunate was successful for the treatment of multidrug-resistant Plasmodium falciparum malaria in Thailand.5-7 The combining of an artemisinin derivative with pyrimethamine-sulphadoxine might delay or prevent the emergence of resistance to pyrimethamine-sulphadoxine. Artemisinin derivatives achieve substantial and rapid decreases in parasite load but have a short half-life; combination with a longer-acting drug, such as pyrimethamine-sulphadoxine, which acts on a different target, protects against the emergence of artemisinin-resistant parasites.1 In addition, the use of combination therapy could have an impact on malaria transmission by lowering rates of gametocytaemia after treatment.8 In areas where there is currently little pyrimethaminesulphadoxine resistance, early use of combination regimens will increase the chance of preventing the spread of resistance. In such areas, it is therefore important to establish whether combination therapy is at least as effective as pyrimethamine-sulphadoxine alone. We investigated in a double-blind, randomised, controlled study the safety, tolerability, and efficacy of pyrimethamine-sulphadoxine and artesunate (sodium artesunate) compared with pyrimethamine-sulphadoxine alone in Gambian children with acute uncomplicated malaria. Methodstoctoc Patientstoctoc Children who weighed more than 5 kg and who were younger than 10 years were eligible for enrolment if: they were infected with Plasmodium falciparum at a density of 500/µL or more;

4 had a history of fever; lived within 20 km of a trial centre; and a parent or guardian gave informed consent. Children were excluded if they required parenteral treatment, had been treated within the previous 2 weeks with pyrimethamine-sulphadoxine, had a haematocrit lower than 15%, or had any evidence of chronic disease. Study designtoctoc We did the study in five health centres in The Gambia. One trial centre was in Fajara, a semiurban coastal area, and the other four sites Keneba, Njaba Kunda, Ngeyen Sanjal, and Basse were in the rural region inland. The study protocol was the same at all sites. All eligible children received half a tablet of pyrimethamine-sulphadoxine if their bodyweight was lower than 10 kg, and an additional quarter tablet for every 5 kg increment. Children were randomly assigned pyrimethamine-sulphadoxine plus placebo, pyrimethamine-sulphadoxine plus one dose 4 mg/kg bodyweight artesunate, or pyrimethamine-sulphadoxine plus one dose 4 mg/kg bodyweight artesunate daily for 3 days (figure 1). The artesunate and placebo tablets were identical in appearance. All children received the first dose under direct supervision. Children were monitored for 1 h to check for vomiting, and a replacement dose was given if necessary. Children who vomited after the starting dose and the replacement dose were given parenteral chloroquine. Each child received 10 mg/kg paracetamol under direct supervision and parents were instructed to give paracetamol every 6 h until the symptoms subsided. Children who had haematocrits of more than 30% were given 15 mg ferrous sulphate per kg bodyweight daily for 14 days. Artesunate and placebo tablets were packaged individually and labelled with a randomisation number for each child in blocks of 12. Investigators, children, and parents or guardians remained masked to treatment status throughout the study FF1&S=INJHKAJLBMGO FF1&S=INJHKAJLBMGO00 [Help with image viewing] Figure 1: Trial profile Field workers visited children to supervise the ingestion of the second and third doses of treatment, assess the children's health, and obtain blood films. The field workers recorded axillary temperature with digital thermometers, the use of antipyretics in the previous 24 h, and the occurrence of any adverse effects. aily visits were continued until blood films were free from asexual P falciparum. Field workers made further visits at days 7 and 14 to take blood films, and at day 28 to obtain blood films and haematocrits. Parents were encouraged to return to the health centre at any time if the child was unwell. We gave rescue therapy with quinine to children who developed symptoms without clearance of parasitaemia or who returned with a second episode of parasitaemia with a density higher than 500/µL. We asked the first 60 children enrolled at the Fajara centre to return for clinical assessment 7 days after the start of treatment. 2 ml venous blood samples were taken at enrolment and on day 7 to measure serum alanine aminotransferase and serum creatinine.

5 The primary endpoints were tolerability and safety, and parasitological and clinical cure by day 14. Secondary endpoints were parasite clearance rates, fever clearance, parasitological and clinical cure by day 28, gametocyte rates on days 7, 14, and 28, and haematocrit on day 28. The study was approved by the ethics review boards of The Gambia government/medical Research Council Laboratories, the London School of Hygiene and Tropical Medicine, London, UK, and the WHO Steering Committee for Research Involving Human Subjects. Laboratory methodstoctoc Thick blood films were stained with giemsa. Parasite density was measured as the number of parasites per 200 leucocytes on a thick film, assuming a total leucocyte count of /L. If gametocytes were seen, the gametocyte count was extended to 1000 leucocytes. All slides were read by the same two experienced microscopists, and discrepancies were reviewed by a third senior microscopist. Haematocrit was measured by microhaematocrit centrifugation. Serum creatinine and alanine aminotransferase were measured with a Cobas Mira serum chemistry analyser (Roche, Basel, Switzerland). Blood samples were taken on to filter paper for analysis of merozoite surface proteins MSP1, MSP2, glutamine-rich protein, dihydrofolate reductase, and dihydropteroate synthase from all children on each occasion they were seen. If a child had a second episode of parasitaemia, samples from the first and second episodes were assayed. NA was purified as described previously.9 We used a nested PCR to analyse three polymorphic genetic markers from P falciparum: the three-sequence families of the MSP1 block 2 repeat region, the two-sequence families of the MSP2 repeat region, and the RII region of glutamine-rich protein. We used a second nested PCR to amplify the regions of dihydrofolate reductase and dihydropteroate synthase associated with resistance to pyrimethamine and sulfadoxine, respectively, with previously reported primers.10 A recrudescent infection was defined as one that showed a complete match in allelic size for all the genes MSP1, MSP2 and glutamine-rich protein between the first and second samples. If any clone of a polyclonal primary infection was detected during a second episode, it was taken to be a recrudescence. Statistical analysistoctoc The parasitological failure rate by day 14 and by day 28 after the start of treatment were the outcomes for which sample sizes were calculated. Previous estimates of parasitological failure rates in The Gambia with pyrimethamine-sulphadoxine alone have been 2% 14 days after treatment,11 and 10-13% 28 days after treatment.12,13 A trial with 200 children in each group, which allowed up to 10% loss to follow up, had 80% power (with a two-sided significance level of 5%) to detect a difference in failure rates at day 28 between the pyrimethamine-sulphadoxine alone group and either artesunate group, if the true rates were 10% in the pyrimethaminesulphadoxine alone group and 2% in each artesunate group. More generally, we wanted to establish whether combination therapy is at least as good as pyrimethamine-sulphadoxine alone for these endpoints. We had 80% power to ensure that the upper 95% confidence limit for the difference between failure rates would be less than 3% if the true rates were 10% and less than 5%.14

6 Treatment was taken to have failed at days 14 and 28 if children had received rescue treatment before or were parasitaemic at those times. All children for whom outcomes were known were included in analyses. We used [chi]2 analysis to compare the failure rates in the three groups. A 95% CI for risk difference was calculated for pair-wise comparisons. In a secondary analysis, children were taken not to be treatment failures if parasitaemia on day 28 was caused by a new, rather than recrudescent, infection. We compared the distribution of time to parasite clearance with the Wilcoxon's test, stratified by site. The date of clearance was taken as the day of the first negative blood film. For this analysis, observations on children lost to follow-up, and children who received rescue treatment, were censored on the date they were lost or treated. Observations for children who had no blood films for 2 or more consecutive days were censored at the start of the missing period. We pooled data for the two artesunate groups for some analyses, since the treatments were the same on day 1. Analyses were done with Stata (version 6). Resultstoctoc 2184 children who presented with fever or a history of fever at a trial site were screened during November and ecember, 1998; 600 eligible children were enrolled (figure 1). For logistic reasons, not all eligible children could be enrolled. At enrolment, the groups were similar for districutions of sex, age, weight, and clinical features (table 1). Two children who had negative blood films at the time of enrolment were included in safety and tolerability analyses but were excluded from efficacy analyses. By day 14, 18 children had been withdrawn from the study because of negative blood films at enrolment (two), receiving rescue therapy (11), or vomiting after treatment (five). By day 28, two further children had been withdrawn after receiving rescue therapy TT1&S=INJHKAJLBMGO TT1&S=INJHKAJLBMGO00 [Help with image viewing] Table 1: Baseline characteristics 11 children received rescue treatment during the first 7 days, three in the pyrimethamine-sulphadoxine alone group, six in the one-dose artesunate group, and two in the three-dose artesunate group. Five children vomited after receiving the starting and replacement doses and were given parenteral chloroquine, four in the one-dose and one in the three-dose artesunate groups. A boy aged 2 years with a high parasite density ( /mL) treated with pyrimethamine-sulphadoxine alone died 24 h after enrolment from severe malaria. One child in the three-dose group had convulsions within 7 h of enrolment and was treated with parenteral quinine. Two children (one in the pyrimethamine-sulphadoxine alone group and one in the one-dose artesunate group) required blood transfusions in the first 4 days of treatment, and also received parenteral therapy. One child deteriorated clinically 2 days after treatment with pyrimethamine-sulphadoxine and received parenteral quinine. In one child in the one-dose artesunate group, parasitaemia did not clear, and quinine was administered on day

7 6. Treatment outcome by day 14 was unknown for 26 children. 17 of these children had cleared parasitaemia before day 14, and only nine were lost to follow-up, on day 1 or 2, before parasitaemia had cleared. Treatment outcome by day 14 was known for 572 (96%) children, and by day 28 for 565 (94%) children. Pyrimethamine-sulphadoxine and artesunate were well tolerated. No severe adverse reactions attributable to treatment were recorded. Children presented most frequently with headache, anorexia, and abdominal pain. Less frequently, children complained on enrolment of vomiting, dizziness, and diarrhoea. Malaria symptoms, indistinguishable from mild adverse reactions, but potentially related to trial medications were found in 328 (56%) of 585 children after treatment (table 2). All symptoms, except anorexia and vomiting, were more frequently seen in children not receiving artesunate TT2&S=INJHKAJLBMGO TT2&S=INJHKAJLBMGO00 [Help with image viewing] Table 2: Proportion of children with symptoms at enrolment and on days 1 and 2 At enrolment, five (8%) of 60 children who were tested had raised concentrations of alanine aminotransferase (>45 IU/L), which resolved in all children within 7 days. All children had normal creatinine concentrations before and after treatment. None of the children had clinical signs suggestive of renal or liver disease. Five children (three in the pyrimethamine-sulphadoxine alone group, one in the one-dose artesunate group, and one in the three-dose artesunate group) who had previously cleared parasitaemia had returned with second episodes by day 14. Overall, six (3%) of 194 children treated with pyrimethamine-sulphadoxine alone received rescue treatment before day 14 or were parasitaemic on day 14, compared with seven (4%) of 189 in the one-dose artesunate group (risk difference -0.6% [95% CI -4.2 to 3.0]) and three (2%) of 189 in the three-dose group (risk difference 1.5% [-1.5 to 4.5]). The proportion of treatment failures by day 28 was 7.3% (14 of 192) in the pyrimethaminesulphadoxine alone group, 7.0% (13 of 186) in the one-dose artesunate group, and 2.1% (four of 187) in the three-dose artesunate group (p=0.048). The differences in risk were: 0.4% (-4.8 to 5.6) between the pyrimethamine-sulphadoxine alone and the one-dose groups; 5.2% ( ) between the pyrimethamine-sulphadoxine alone and the three-dose groups; and 4.9% ( ) between the one-dose and three-dose groups. Second episodes of parasitaemia were seen in 23 children between days 14 and 28, 12 (6%) of 187 in the pyrimethamine-sulphadoxine alone group, and eight (4%) of 180 in the one-dose and three (2%) of 185 in the three-dose artesunate groups (test of equality of the three proportions, p=0.070). From 18 of these 23 children, the genotype of P falciparum from paired blood samples was determined by PCR amplification. There was genetic heterogeneity in eight of the pairs, which suggested that new infections were the cause of these second episodes; in the other ten

8 pairs no heterogeneity was detected, which suggested that these were recrudescences. One child in the three-dose artesunate group had a recrudescence, compared with three in the one-dose artesunate group and six in the pyrimethamine-sulphadoxine alone group. The proportion of treatment failures by day 28, excluding new infections, was eight (4%) of 186 in the one-dose group, and three (2%) of 187 in the three-dose group. In the pyrimethaminesulphadoxine alone group, the estimate was between 12 (6%) of 192 (with the assumption that the four untyped infections were recrudescent) and eight (4%) of 192 (with the assumption that the untyped infections were new). Clearance of parasites was significantly faster for children in each of the artesunate groups (p<0.001) than for those in the pyrimethamine-sulphadoxine alone group (figure 2). By day 1, 178 (47%) of 381 children treated with artesunate were still parasitaemic compared with 157 (81%) of 195 children in the pyrimethamine-sulphadoxine alone group (relative risk 1.7 [ ]). By day 2, 36 (19%) of 192 children treated with pyrimethamine-sulphadoxine only were parasitaemic, compared with 16 (8%) of 190 in the one-dose artesunate group, and two (1%) of 190 in the three-dose group. By day 3, alone one child treated with artesunate was still parasitaemic, compared with six in the pyrimethamine-sulphadoxine alone group. The time to clearance between the one-dose and three-dose artesunate groups did not differ significantly (p=0.6) FF2&S=INJHKAJLBMGO FF2&S=INJHKAJLBMGO00 [Help with image viewing] Figure 2: Proportion of children with patent asexual P falciparum parasitaemiavertical lines are 95% CI. A temperature of 37.5 C or higher was judged a fever. 58% of children were febrile at enrolment. On day 1, 17% had fever in the pyrimethaminesulphadoxine alone group compared with 8% in the one-dose and three-dose artesunate groups combined (p=0.001). By day 2, 11% of the pyrimethamine-sulphadoxine alone group had fever, compared with 2% in the combined artesunate groups (p<0.001). The number of doses of paracetamol taken by children in the three groups was similar. Parasites with dihydropteroate synthase markers associated with sulphadoxine resistance (glycine at codon 437) were found in nine (45%) of 20 first episodes and in 15 (88%) of 17 second episodes for which parasites were typed (p=0.006). The pyrimethamine-resistance allele HFR Asn108 was found in 14 (70%) of 20 children at the time of the first episode and in 15 (88%) of 17 at the time of the second episode (p=0.2). A dihyrofolate reductase genotype, classified as resistant if codon 108 was asparagine, and codon 51 was isoleucine or codon 59 was arginine,15 was found in 13 (65%) of 20 children on enrolment and in 15 (88%) of 17 at second episodes. At enrolment, gametocytes were detected in 106 (18%) of 596 children (slides from two children were lost). On day 7, gametocytes were detected in 126 (67%) of 189 children treated with

9 pyrimethamine-sulphadoxine alone, compared with 46 (25%) of 182 children treated with pyrimethamine-sulphadoxine and one dose of artesunate and 38 (20%) of 185 children who received three doses of artesunate (p<0.001 each, figure 3). Of children who had no gametocytaemia at baseline, 93 (60%) of 154 were gametocytaemic on day 7 in the pyrimethamine-sulphadoxine alone group, 26 (17%) of 149 in the one-dose artesunate group, and 15 (10%) of 152 in the three-dose group. Gametocyte densities on days 7 and 14 were significantly higher in the pyrimethamine-sulphadoxine alone group than in the other two groups (p<0.001). An estimate of mean duration of gametocytaemia, equal to the area under the curve in figure 3, was 13 days in the pyrimethamine-sulphadoxine alone group, compared with 4.2 days in the one-dose and 3.4 days in the three-dose artesunate groups (p<0.001) FF3&S=INJHKAJLBMGO FF3&S=INJHKAJLBMGO00 [Help with image viewing] Figure 3: Proportion of children with gametocytesvertical lines are 95% CI. children who received haematinics was similar in the three groups. At enrolment, the median haematocrit was 29%; 294 (50%) of 587 children had haematocrits lower than 30%. By day 28, the median haematocrit was 34% in each group, with 69 (13%) of 550 children having haematocrits of less than 30%. The proportion of iscussiontoctoc No severe adverse reaction was attributable to combined treatment with pyrimethaminesulphadoxine and artesunate in any child. Mild adverse reactions to drugs might be indistinguishable from the signs and symptoms caused by malaria, but these were more frequently reported by children treated with pyrimethamine-sulphadoxine only and were, therefore, unlikely to be attributable to the combined regimen. Malaria parasites acquire resistance to antimalarial drugs through mutation. Rapid parasite clearance may, therefore, lower the chance of drug-resistant strains emerging.16 Artemisinin derivatives are the most potent and rapidly acting antimalarial drugs.1 The combination of a short-acting artemisinin derivative with a longer-acting drug, such as pyrimethaminesulphadoxine, has the advantage that any parasites remaining after the artemisinin derivative has taken effect are eliminated by a drug with a different mode of action, but are no longer exposed to the artemisinin. Theoretically, therefore, emergence of resistance to each drug is less likely. In this study, 65% of children who developed a second episode of parasitaemia were initially infected with P falciparum that had dihydrofolate reductase markers associated with pyrimethamine-sulphadoxine resistance, and 45% had dihydropteroate synthase markers of resistance. A change in dihydropteroate synthase at codon 437 from alanine to glycine confers a low degree of resistance against sulphadoxine, whereas isolates with asparagine at codon 108 in dihydrofolate reductase have a 1000-fold increase in resistance to pyrimethamine.15,17 Parasites with additional mutations at dihydrofolate reductase codons 51 and 59 are fold more

10 resistant to pyrimethamine than parasites with the asparagine 108 mutation alone of 17 of these second episodes had the dihydropteroate synthase marker of resistance and 15 of 17 isolates had dihydrofolate reductase resistance markers. The higher prevalence of the resistant genotypes at the time of the second episodes than at that of the primary infections suggests a selection of pre-existing or acquired clones resistant to pyrimethamine, sulphadoxine, or both. Resistant genotypes are present in The Gambia and a rapid selection of pyrimethaminesulphadoxine-resistant strains can be expected should treatment with pyrimethaminesulphadoxine alone become widespread. To achieve maximum benefit, combination therapy should be instituted before one of the drugs in the combination has lost its effectiveness. It is not clear whether this selection for pyrimethamine-sulphadoxine resistance also applies to gametocytes, but the high frequency of gametocytaemia after treatment with pyrimethaminesulphadoxine alone may explain the explosive spread of pyrimethamine-sulphadoxine-resistant strains in east Africa. After treatment with only pyrimethamine-sulphadoxine, we found gametocytes in more than half of children (67%). The addition of artesunate lowered this rate three-fold, as well as gametocyte densities and duration of gametocyte carriage. The introduction of artemisinins in combination with pyrimethamine-sulphadoxine or other long-acting antimalarial drugs, may, therefore, decrease P falciparum transmission. In the Northern part of Thailand, the introduction of artemisinin derivatives as first-line treatment for malaria was associated with a 40% reduction in the incidence of the disease in the first 2 years, which the authors suggested was due to lowered transmission.8 These findings have far-reaching implications for the choice of antimalarials in areas of Africa where few currently available drugs are affordable for widespread use. The current practice of replacing one antimalarial drug with another once resistance has developed is unsustainable. In a country such as The Gambia, where pyrimethamine-sulphadoxine resistance is uncommon, the addition of artesunate to each malaria treatment may help avert the predicted malaria disaster through rapid clearance of parasitaemia and low rates of treatment failure. Although three doses of artesunate led to lower densities of asexual parasites on day 2 and fewer second episodes of parasitaemia during 28-day follow-up, a single dose is easier to administer and cheaper. Whether a single-dose regimen can yield a benefit similar to a three-dose regimen over extended periods needs further study. Trials using similar protocols to this study to assess antimalarial combination therapies are in progress throughout sub-saharan Africa. Contributorstoctoc Lorenz von Seidlein contributed to the design and coordination of the study, supervised enrolment of children, and analysed genotypes. Paul Milligan contributed to the design of the study and analysed the data. Margaret Pinder contributed to design and coordination of the study. Kalifa Bojang contributed to the design and coordination of the study, and supervised enrolment and follow-up of children in Njaba Kunda. Roland Gosling contributed to coordination of the study and supervised enrolment and follow-up of children in Keneba. Rosalind Coleman contributed to coordination of the study and supervised enrolment and follow-up of children in Ngeyen Sanjal. Justin Ifeanyichukwu supervised enrolment and follow-up of children in Basse. Abudbakar Sadiq supervised enrolment and follow-up of children in Fajara. Manoj uraisingh analysed and interpreted HFR and HSP alleles. avid Warhurst supervised the analysis of

11 HFR and HSP alleles and their interpretation. Ali Alloueche analysed and interpreted genotypes. Brian Greenwood contributed to the concept of the artesunate combination studies and the specific design and coordination of this study, supervised the analysis of genotypes, and interpreted the data. Geoffrey Targett contributed to the design and coordination of the study. Keith McAdam contributed to the design and coordination of the study. Gijs Walraven contributed to design and coordination of the study and supervised enrolment of children. Piero Olliaro contributed to the concept of artesunate combination studies and the specific design and coordination of this study, and coordinated the provision of artesunate. Tom oherty contributed to the design and coordination of the study and supervised enrolment of children. All investigators contributed to the writing of the paper. Acknowledgmentstoctoc We thank N White, W Watkins, M Molyneux (Wellcome Trust, Bangkok, Kilifi and Blantyre), F ter Kuile (CC, Kisumu), P Lange, (Ministry of Health, Uganda), and R Peto (Oxford, UK), the additional members of the Protocol evelopment team and rug Safety Monitoring Board, for their work and contributions; T Corrah for his help as a safety monitor; S Obaro and A Collinson for their help in Basse and keneba; W Taylor (WHO, Geneva), external monitor; P Njai for pharmaceutical support; K Okunoye for data management; and B Jaiteh for biochemical analyses. We thank also S Jaffar and F Nosten for advice and encouragement; the fieldworkers, represented by S Baldeh and L Manneh; and the microscopists P Bojang and A Bah for their help, which was essential for the completion of this trial; and J P ucret and M H Corel (Sanofi, France) for providing artesunate and placebo. This study was supported by UNP/World Bank/WHO Special Programme for Research and Training in Tropical iseases (TR), Geneva, Switzerland. Referencestoctoc 1 White NJ, Nosten F, Looareesuwan S, et al. Averting a malaria disaster. Lancet 1999; 353: [Fulltext Link] [Medline Link] [Context Link] 2 Marsh K. Malaria disaster in Africa. Lancet 1998; 352: [Fulltext Link] [Medline Link] [Context Link] 3 Trape JF, Pison G, Preziosi MP, et al. Impact of chloroquine resistance on malaria mortality. C R Acad Sci III 1998; 321: [Medline Link] [Context Link] 4 Ronn A, Msangeni H, Mhia J, Wernsdorfer W, Bygbjerg I. High level of resistance of Plasmodium falciparum to sulfadoxine-pyrimethamine in children in Tanzania. Trans R Soc Trop Med Hyg 1996; 90: [Medline Link] [Context Link] 5 Price RN, Nosten F, Luxemburger C, et al. Artesunate/mefloquine treatment of multi-drug resistant falciparum malaria. Trans R Soc Trop Med Hyg 1997; 91: [Medline Link] [Context Link]

12 6 Nosten F, Luxemburger C, ter Kuile FO, et al. Treatment of multidrug-resistant Plasmodium falciparum malaria with 3-day artesunate-mefloquine combination. J Infect is 1994; 170: [Medline Link] [Context Link] 7 Bunnag, Kanda T, Karbwang J, Thimasarn K, Pungpak S, Harinasuta T. Artemether or artesunate followed by mefloquine as a possible treatment for multidrug resistant falciparum malaria. Trans R Soc Trop Med Hyg 1996; 90: [Medline Link] [Context Link] 8 Price RN, Nosten F, Luxemburger C, et al. Effects of artemisinin derivatives on malaria transmissibility. Lancet 1996; 347: [Fulltext Link] [Medline Link] [Context Link] 9 Irion A, Felger I, Abdulla S, et al. istinction of recrudescences from new infections by PCR- RFLP analysis in a comparative trial of CGP and chloroquine in Tanzanian children. Trop Med Int Health 1998; 3: [Fulltext Link] [Medline Link] [Context Link] 10 uraisingh MT, Curtis J, Warhurst C. Plasmodium falciparum: detection of polymorphisms in the dihydrofolate reductase and dihydropteroate synthetase genes by PCR and restriction digestion. Exp Parasitol 1998; 89: 1-8. [Medline Link] [Context Link] 11 von Seidlein L, Bojang K, Jones P, et al. A randomized controlled trial of artemether/benflumetol, a new antimalarial and pyrimethamine/sulfadoxine in the treatment of uncomplicated falciparum malaria in African children. Am J Trop Med Hyg 1998; 58: [Context Link] 12 Muller O, van Hensbroek MB, Jaffar S, et al. A randomized trial of chloroquine, amodiaquine and pyrimethamine-sulphadoxine in Gambian children with uncomplicated malaria. Trop Med Int Health 1996; 1: [Fulltext Link] [Medline Link] [Context Link] 13 Bojang KA, Schneider G, Forck S, et al. A trial of Fansidar plus chloroquine or Fansidar alone for the treatment of uncomplicated malaria in Gambian children. Trans R Soc Trop Med Hyg 1998; 92: [Medline Link] [Context Link] 14 Farrington CP, Manning G. Test statistics and sample size formulae for comparative binomial trials with null hypothesis of non-zero risk difference or non-unity relative risk. Stat Med 1990; 9: [Medline Link] [Context Link] 15 Watkins WM, Mberu EK, Winstanley PA, Plowe CV. The efficacy of antifolate antimalarial combinations in Africa: a predictive model based on pharmacodynamic and pharmacokinetic analyses. Parasitol Today 1997; 13: [Context Link] 16 White NJ. Preventing antimalarial drug resistance. rug Resist Updates 1998; 1: 3-9. [Context Link] 17 Sixsmith G, Watkins WM, Chulay J, Spencer HC. In vitro antimalarial activity of tetrahydrofolate dehydrogenase inhibitors. Am J Trop Med Hyg 1984; 33: [Medline Link] [Context Link]

13