You are on page 1of 7


Atovaquone-proguanil versus chloroquine-proguanil for malaria prophylaxis in non-immune travellers: a randomised, double-blind study

Birthe Høgh, Paul D Clarke, Daniel Camus, Hans Dieter Nothdurft, David Overbosch, Matthias Günther, Izak Joubert, Kevin C Kain, Dea Shaw, Neil S Roskell, Jeffrey D Chulay, and the Malarone International Study Team

Background Chloroquine plus proguanil is widely used for malaria chemoprophylaxis despite low effectiveness in areas where multidrug-resistant malaria occurs. Studies have shown that atovaquone and proguanil hydrochloride is safe and effective for prevention of falciparum malaria in lifelong residents of malaria-endemic countries, but little is known about non-immune travellers. Methods In a double-blind equivalence trial, 1083 participants travelling to a malaria-endemic area were randomly assigned to two treatment groups: atovaquoneproguanil plus placebos for chloroquine and proguanil, or chloroquine, proguanil, and placebo for atovaquoneproguanil. Follow-up was by telephone 7 and 60 days after travel and at a clinic at 28 days. Serum samples were tested for antibodies to a malaria circumsporozoite protein. Blood and serum samples of participants with a potential malaria diagnosis were tested in a reference laboratory. Findings 7 days after travel, at least one adverse event was reported by 311 (61%) of 511 participants who received atovaquone-proguanil and 329 (64%) of 511 who received chloroquine-proguanil. People receiving atovaquone-proguanil had a lower frequency of treatmentrelated gastrointestinal adverse events (59 [12%] vs 100 [20%], p=0·001), and of treatment-related adverse events of moderate or severe intensity (37 [7%] vs 56 [11%], p=0·05). There were fewer treatment-related adverse events that caused prophylaxis to be discontinued in the atovaquone-proguanil group than in the chloroquineproguanil group (one [0·2%] vs ten [2%], p=0·015). Interpretation Overall the two preparations were similarly tolerated. However, significantly fewer adverse gastroInternational Travel Vaccination Centre, Copenhagen, Denmark (B Høgh MD); Medical Advisory Services for Travellers Abroad (MASTA), London, UK (P D Clarke FRCP); Institut Pasteur, Lille, France (D Camus MD); Department for Infectious Diseases and Tropical Medicine, Munich, Germany (H D Nothdurft MD); Harbour Hospital and Institute of Tropical Medicine, Rotterdam, Netherlands (D Overbosch MD); Institute for Tropical Medicine, Berlin, Germany (M Günther MD); Travelsafe Clinic, Cape Town, Republic of South Africa (I Joubert MD); Toronto General Hospital, Toronto, Canada (K C Kain MD); and Glaxo Wellcome, Research Triangle Park, NC, USA and Greenford, UK (D Shaw RN, N S Roskell MSc, J D Chulay MD) Correspondence to: Malarone Publication Coordinator, Room 50-3505B, Glaxo Wellcome Inc, 5 Moore Drive, Research Triangle Park, NC 27709, USA (e-mail:

intestinal events were observed in the atovaquoneproguanil group in than in the chloroquine-proguanil group. Lancet 2000; 356: 1888–94

Malaria is one of the greatest threats to health for international travellers. The number of reported cases of malaria in Europe increased from 6840 in 1985, to 8438 in 1995, with a case-fatality rate as high as 3·6%.1 Imported malaria is also a problem in North America, with more than 1000 cases reported annually in the USA and Canada.2,3 Imported malaria mainly occurs when travellers fail to use appropriate chemoprophylaxis or do not use it correctly.3 Widespread parasite resistance renders chloroquine poorly effective in most malaria-endemic countries.4 Mefloquine is highly effective in preventing malaria but has been associated with neuropsychiatric side-effects that may be severe and can restrict its use.5 The combination of chloroquine plus proguanil has fewer serious side-effects than mefloquine6 and is more effective than chloroquine alone, but still less effective than mefloquine.7 An additional drawback with drugs widely used for malaria prophylaxis is that they must be continued for 4 weeks after leaving a malaria-endemic area.8 There is a clear need for better drugs to prevent malaria. In controlled clinical trials with more than 600 participants, a fixed-dose combination of atovaquone and proguanil hydrochloride was 98% effective in prophylaxis of malaria caused by Plasmodium falciparum and had a safety profile similar to that of placebo.9–12 However, these studies were done on lifelong residents of malaria-endemic countries, who may have had some immunity to malaria. Additionally, the frequency and nature of adverse events might differ between travellers and residents of endemic areas. We have undertaken a randomised, double-blind trial to compare the safety and efficacy of atovaquone-proguanil versus chloroquine-proguanil in non-immune travellers. The hypothesis was that the frequency of adverse events in participants receiving atovaquone-proguanil was not higher than in those receiving chloroquine-proguanil. The frequency of treatment-limiting adverse events and efficacy of prophylaxis were secondary endpoints.

Study participants Participants were enrolled into study MAL30011 at 21 travel clinics in Denmark, UK, France, Germany, Netherlands, South Africa, and Canada. They were of both sexes, at least age 14 years, weighed more than


THE LANCET • Vol 356 • December 2, 2000

Copyright © 2000 All Rights Reserved

participants were given a diary card to record details of study drug administration. atovaquone-proguanil or chloroquine-proguanil. headache. information was obtained about the exact dates and locations of travel in malaria-endemic areas. or otherwise seriously jeopardised the participant. or travel to a malaria-endemic area within the previous 60 days. For each active drug. Compliance with study drug use was assessed at the 28 day follow-up visit by interview. or chloroquine. itching. insomnia. At all sites consecutively enrolled individuals who satisfied all entry criteria received the next treatment number. Schleicher and Schull. containing 250 mg atovaquone and 100 mg proguanil hydrochloride. to be opened only if knowledge of study drug assignment was required for management of a medical emergency. severe blood or neurological disorders. Those in the atovaquoneproguanil group received active atovaquone-proguanil and placebos for chloroquine and proguanil. and those in the chloroquine-proguanil group received chloroquine. information was obtained about demographic characteristics. Mississauga. Follow-up surveys People were interviewed by telephone 7 and 60 days after leaving the malaria-endemic area and at a clinic visit at 28 days after departure. or cardiac dysfunction. and a malaria diagnosis kit. and planned to travel for up to 28 days to an area where P falciparum was endemic. psychiatric disorders. resulted in admission. At the 7-day contact. PGN=proguanil. duration. proguanil. Measurement of malarial indices At the baseline and 28 day follow-up visits. or alcoholism. They were also asked an open-ended question about any other medical conditions or adverse events they had had since the last visit. nausea. severe=sought medical advice). and instructions on how to take samples and arrange for prepaid courier service to return samples to a central laboratory. Investigators assessed whether there was a reasonable possibility that the adverse event was caused by the study drug.8 At enrolment. An adverse event was recorded as treatment-emergent if it started while the participant was taking the study drug. Written informed consent was obtained from all individuals or their parents. active proguanil and a placebo for atovaquone-proguanil (figure 1). mouth ulcers. malaria history. depression. Design and procedures Participants were randomly assigned to one of two treatment groups. and counting unused pills. filter strips (Isocode Stix. capsules or film-coated tablets were identical in appearance to the matching placebo. They were instructed to use this kit if they were diagnosed with malaria. a physical examination was done.12 Chloroquine phosphate (Avloclor. NH. 2000 1889 Copyright © 2000 All Rights Reserved 28 days 7 days 7 days . 250 mg tablets equivalent to 155 mg chloroquine base) and proguanil hydrochloride (Paludrine. Atovaquone-proguanil hydrochloride (Malarone tablets. strange or vivid dreams. The participant was instructed to bring the slides and filter strips to the investigator at the follow-up visit. visual difficulties. Reasons for exclusion were: known hypersensitivity to atovaquone. treatment-limiting if it resulted in permanent discontinuation of study drug. THE LANCET • Vol 356 • December 2. and intensity (mild=fairly trivial. were in good general health. dizziness. diarrhoea.ARTICLES 50 kg. a history of seizures. Individuals were told that if they developed malaria they should ask the health care provider making the diagnosis to obtain the samples and send the whole- Before Travel period During After 1–2 days Group ATQ-PGN active one tablet daily Atovaquoneproguanil Chloroquine placebo two capsules per week Proguanil placebo two capsules daily ATQ-PGN placebo one tablet daily Chloroquineproguanil Chloroquine active two capsules per week Proguanil active two capsules daily Figure 1: Drug dosing regimens ATQ=atovaquone. and fits or seizures). Canada) or matching placebo was given according to standard recommendations of the manufacturer (figure 1). Each adverse event was assessed for date of onset. hepatic. generalised psoriasis. and blood was obtained for routine haematological tests (haemoglobin. moderate=interfered with daily activities. life-threatening. vomiting. Chloroquine and proguanil or matching placebos were given according to standard recommendations of WHO (figure 1). malaria within the previous 12 months. participants were asked if they had been diagnosed with malaria and queried about symptoms (fever. current medical conditions and medications. At enrolment and each follow-up assessment. The study was approved by the ethics committee at each study site. anxiety. 100 mg tablets) were manufactured by AstraZeneca and encapsulated by GlaxoWellcome to achieve blinding. renal. The kit contained a card on which a health care provider could record details about the diagnosis. abdominal pain. disabling. Treatment codes were provided to investigators in opaque sealed envelopes. At enrolment. and serious if it were fatal. Keene. serum was collected from participants at all sites. travel destination. USA) on which blood could be spotted for parasite DNA analysis. pregnancy or lactation. slides and a slide holder (to take blood films). A computer-generated code was used to randomly assign a treatment number to the three bottles of study drug for every individual. reviewing the diary card. blood and plasma samples to the central laboratory. GlaxoWellcome. tubes for whole blood (for parasite DNA analysis) and plasma (to measure plasma drug concentrations). and a pregnancy test was carried out for women of childbearing potential.

but blood smears and parasite DNA analyses were negative or missing and antibodies to blood-stage parasites were negative. protocol violation (1). and negative if a diagnosis of malaria was recorded by a health care provider. and adverse event (1). The two groups were well balanced with respect to baseline demographics. Proportions of people with adverse events were compared with Yates’ corrected ␹2 test. The serum was used to measure antibodies to P falciparum circumsporozoite protein by ELISA13 for all participants. concern about potential side-effects (1). analysed in the intent-to-treat population of all participants who received at least one dose of study drug.21 An equivalence trial with 400 suitable participants per group has power of 82% to detect non-inferiority of atovaquone-proguanil compared with chloroquine-proguanil. Statistical analyses The primary study endpoint was the overall frequency of adverse events assessed at 7 days after leaving the malaria-endemic area. The circumsporozoite antibody test was judged positive when the result for a post-travel serum sample was more than 2 SD above the mean of results for the negative control sera and also more than 2 SD above the baseline serum sample result. The average duration of travel was about 2·5 weeks and 642 (63%) people travelled to Africa. drug lost (1). history of malaria. The power is also more than 80% if the overall proportion of adverse events is 60–70%. permanently left country (1). and to measure antibodies to blood-stages of the four human species of plasmodium by immunofluorescence14 for those diagnosed with potential malaria. 16 participants (1·6%) reported a previous episode of malaria a median of 12 (range 2–60) years before enrolment. who had confirmed malaria or who developed anticircumsporozoite antibodies and for whom 60-day efficacy data were available. protocol violation (1). and protocol violation (1). 1083 participants were randomly assigned to receive atovaquone-proguanil (n=540) or chloroquine-proguanil (543). and treatment with another antimalarial drug (1). possible if a diagnosis of malaria was recorded by a health care provider. The overall frequency of adverse events at 28 days after leaving the malaria-endemic area was also determined. the 95% CI around the proportion was calculated from the binomial distribution. were lost to follow-up (nine). 24 participants were more than 65 years old. the number of participants regarded as at high risk was the number 1083 eligible participants R 540 randomised to receive atovaquone-proguanil 543 randomised to receive chloroquine-proguanil 11 6 4 8 did not travel lost to follow-up consent withdrawn other/unknown* 14 3 3 12 did not travel lost to follow-up consent withdrawn other/unknown 511 received study drug 462 completed dosing 49 stopped dosing early 511 received study drug 468 completed dosing 43 stopped dosing early 2 2 1 1 4 did not travel lost to follow-up adverse event consent withdrawn other 2 did not travel 1 lost to follow-up 1 adverse event 501 completed the trial Figure 2: Trial profile 507 completed the trial *Unknown (3). work-related (2). 1008 (99%) completed the trial (figure 2). DNA was extracted from whole blood samples and filter strips. travel destination. but blood smears and parasite DNA analyses were negative or missing and antibodies to blood-stage parasites were missing. 1890 THE LANCET • Vol 356 • December 2. a potential diagnosis of malaria was thought definite if parasite DNA was detected by PCR or parasites were seen on a blood smear slide. The sample size was based on Jones and colleagues’ formula. or other reasons (20). ‡Could not swallow capsules (1). Antibodies to blood-stage parasites were recorded as positive when they were detected at a titre of 1/64 or greater in a post-diagnosis serum sample and the baseline sample was negative. and duration of travel (table 1). if the overall proportion of adverse events is 40% and a 10% difference in adverse event proportions is regarded as clinically important. +10%). withdrew consent (seven). wanted more protection from malaria (1). The target enrolment was increased to 1000 (500 per group) to allow for unsuitable participants. 61 people did not receive their first dose of study drug because they did not travel to a malaria-endemic area (25). Secondary study endpoints were the frequency of treatment-limiting adverse events and the efficacy of prophylaxis. parasite ribosomal RNA genes were amplified by PCR to identify the Plasmodium spp causing malaria. 2000 Copyright © 2000 All Rights Reserved .ARTICLES white blood cell count.15 As defined in the trial protocol.14 Blood smears obtained at the time of malaria diagnosis were examined at the malaria reference laboratory of the Toronto General Hospital. For both minimum and maximum estimates of efficacy. care should be taken when interpreting these comparisons. alanine aminotransferase) from all people at one site. platelet count) and chemistry (creatinine. Percentage efficacy was calculated from: 100ϫ(1Ϫ[number of participants with confirmed malaria/number of participants at risk]). the number of those thought at risk was the number for whom 60-day efficacy data were available. †Unknown (9). Because the p-values are unadjusted and multiple comparisons will inflate the type I error. P falciparum isolates obtained during the course of this study underwent DNA analysis for putative molecular markers of chloroquine and proguanil resistance as described16–20 and by sequencing. Of the 1022 who received at least one dose of study drug. or when there was a 16-fold or greater increase in antibody titre in a postdiagnosis serum sample compared with baseline. 1999. To estimate the maximum efficacy of each drug for prevention of malaria. To estimate the minimum efficacy of each drug. Results From April to October. Non-inferiority was assessed by comparison of the two-sided 95% CI for the adverse event proportion difference against the non-inferiority range (Ϫ100%.

§Other events were allergic reaction in one person receiving atovaquone-proguanil and headache in one participant on chloroquine-proguanil. p=0·024).ARTICLES Characteristic Mean (SD) age (years) Range Male/Female Race* White Black Asian Other Mean (SD) height (cm) Mean (SD) weight (kg) History of malaria Median (range) time since last episode (years) Travel destination† East Africa West Africa Southern Africa Central Africa South America Other Mean (SD) travel duration (days) Atovaquone-proguanil (n=511) 36·0 (13·3) 13–72 251/260 (49/51%) 496 (97%) 6 (1%) 3 (1%) 6 (1%) 173·1 (9·6) 71·4 (13·4) 9 (1·8%) 13 (2–60) Chloroquine-proguanil (n=511) 35·0 (13·3) 13–74 276/235 (54/46%) 486 (95%) 12 (2%) 7 (1%) 6 (1%) 173·7 (9·7) 72·3 (14·5) 7 (1·4%) 11 (2–30) Event Number of participants with adverse events Atovaquone-proguanil Chloroquine-proguanil (n=511) (n=511) p-value 0·024 0·001 0·113 0·001 0·033 0·350 0·002 0·754 0·865 0·479 0·498 0·999 0·370 0·999 0·872 0·999 Any adverse event* Any gastrointestinal event Diarrhoea Nausea Abdominal pain Mouth ulcers Vomiting Any neuropsychiatric event Dizziness Strange or vivid dreams Insomnia Visual difficulties Anxiety Depression Headache Itching 110 (22%) 59 (12%) 24 (5%) 9 (2%) 15 (3%) 18 (4%) 0 49 (10%) 17 (3%) 19 (4%) 8 (2%) 10 (2%) 1 (<1%) 3 (<1%) 21 (4%) 6 (1%) 142 (28%) 100 (20%) 37 (7%) 34 (7%) 30 (6%) 25 (5%) 11 (2%) 53 (10%) 19 (4%) 14 (3%) 12 (2%) 10 (2%) 4 (<1%) 2 (<1%) 19 (4%) 5 (1%) 132 (26%) 90 (18%) 77 (15%) 15 (3%) 50 (10%) 151 (30%) 16·9 (7·5) 152 (30%) 102 (20%) 68 (14%) 12 (2%) 39 (8%) 140 (27%) 17·6 (6·9) *Among participants who reported a drug-related adverse event. dysphagia. mouth ulcers. travel and post-travel periods was: 484 (95%). The mean (SD) duration of treatment was 26 (8) days for atovaquone-proguanil. One participant receiving chloroquine-proguanil had an epileptic seizure. respectively. p=0·003). The difference was significant for gastrointestinal events but not for neuropsychiatric events (table 2). and 45 (10) days for proguanil. one or more adverse events were reported by 311 of 511 (61%) participants in the atovaquone-proguanil group and 329 of 511 (64%) in the chloroquine-proguanil group. or consent withdrawn. †p=0·015. +2·4%. p=0·001) and proguanil (difference=30 [6%]. The difference in number of participants who reported adverse events while receiving active treatment was 33 (Ϫ6·5%) (95% CI Ϫ12·4%. At 7 days after return from a malaria-endemic area. 478 (94%). each of which occurred in one person receiving chloroquineproguanil. p=0·05). 489 (96%). Treatment-emergent adverse events attributed to study drug occurred in more participants who received chloroquine-proguanil than those who received atovaquone-proguanil (142 [28%] vs 110 [22%]. Most adverse events were mild in intensity. participants who opened the chloroquine and proguanil capsules themselves (four) and loss of study drug before completion of postexposure prophylaxis (one). Most events were thought by the investigator to be unrelated to study drug. adverse events were reported by 296 of 511 (58%) of those receiving atovaquone-proguanil and 329 of 511 (64%) receiving chloroquine-proguanil. The difference in frequency of adverse events in the intent-to-treat population was Ϫ3·5% (95% CI Ϫ9·5%. and 473 (93%). and 443 (87%) for proguanil. for atovaquone-proguanil. ‡Other events were burning in pharynx. 2000 1891 Copyright © 2000 All Rights Reserved . and oesophageal reflux. participants randomly assigned to atovaquone-proguanil started chloroquine placebo about 5 days before starting atovaquone-proguanil. Table 2: Treatment-emergent adverse events attributed to study drug *Percentages may not add up to 100 because of rounding. Ϫ0·5%). non-compliance (five vs six). not significant). Event Number of participants with an event Atovaquone-proguanil (n=511) Any treatment-limiting event Any gastrointestinal event Nausea Abdominal pain Diarrhoea Vomiting Other‡ Any neuropsychiatric event Dizziness Anxiety Other§ 1* 0 0 0 0 0 0 0 0 0 1 Chloroquine-proguanil (n=511) 10† 10 6 5 4 2 4 2 1 1 1 *Two additional participants developed a treatment-limiting adverse event while receiving chloroquine placebo before starting active atovaquone-proguanil (rash or abdominal pain and diarrhoea) and one additional person developed a treatment-limiting adverse event (dizziness) while receiving chloroquine and proguanil placebos after completing prophylaxis with active atovaquone-proguanil. Moderate or severe events attributable to study drug occurred in 37 (7%) participants (54 events) receiving atovaquone-proguanil and 56 (11%) (97 events) on chloroquine-proguanil (difference= 19 [4%]. the mean (SD) number of adverse events per person was 1·8 (1·1) while receiving atovaquone-proguanil and 2·2 (1·8) while receiving chloroquine-proguanil. 48 (11) days for chloroquine. 480 (94%). †Some participants visited more than one part of Africa Table 1: Baseline characteristics Participants and study personnel remained unaware of treatment assignment apart from ten individuals (seven in the atovaquone-proguanil group and three in the chloroquine-proguanil group). or other reasons (24 vs 15). Excluding events that occurred while people were receiving only placebo. In the post-travel period the difference from atovaquone-proguanil was significant for chloroquine (difference=65 [13%]. drugs lost or stolen. 466 (91%). and 408 (80%) for chloroquine. These events were severe in five (1%) people (six events) receiving atovaquone-proguanil and 11 (2%) (20 events) on chloroquine-proguanil. Other reasons included: did not travel. protocol violation (nine vs six). Reasons for breaking the treatment code were: an adverse event for which knowledge of study treatment was deemed essential (n=five). constipation. Table 3: Treatment-limiting adverse events attributed to study drug THE LANCET • Vol 356 • December 2. 49 people in the atovaquone-proguanil group and 43 on chloroquine-proguanil discontinued study drug prematurely because of adverse events (11 vs 16). and 461 (90%). Because of the different pretravel dosing regimens. The proportion of participants who took at least 80% of prescribed doses in the pretravel.

isolates were analysed for mutations in genes associated with resistance to chloroquine and proguanilcycloguanil. A potential diagnosis of malaria was made in four participants. Testing of these specimens in reference laboratories allowed us to confirm a diagnosis of falciparum malaria in three people receiving chloroquine-proguanil and in none on atovaquone-proguanil. the proportion of participants in our study who developed a circumsporozoite-antibody response is a minimum estimate of the proportion actually bitten by a malaria-infected mosquito. the proportion of individuals who had at least one adverse event was similar in the two treatment groups. a low frequency of treatment-limiting adverse events is important for the effectiveness of prophylaxis.23. In-vitro drug susceptibility testing of parasite isolates was not feasible but.26 Because we obtained only one postexposure serum sample. all three isolates had N51I and C59R mutations associated with high resistance to pyrimethamine.20 1008 participants completed the 60-day follow-up and had efficacy information recorded.16–18 All three isolates contained the S108N mutation in dhfr. Based on the circumsporozoite antibody results. In one person. compared with ten on chloroquine-proguanil. DNA from the three P falciparum isolates carried the K76T mutation in the pfcrt gene associated with chloroquine resistance. detection of anti-circumsporozoite antibodies by ELISA was reported to have high sensitivity and specificity for identifying individuals who were bitten by a P falciparum-infected mosquito and developed clinical malaria. chemoprophylaxis was discontinued prematurely in only one person receiving atovaquone-proguanil.19 One of three isolates also contained the N86Y allele in pfmdr1. In three. symptoms of malaria began 28 days after completing prophylaxis with atovaquone-proguanil and a subsequent blood smear was reported as positive for P ovale. which has been associated with resistance to pyrimethamine and a 4-fold to 22-fold increase in the 50% inhibitory concentration to cycloguanil. Wolff-ParkinsonWhite syndrome in one. Table 4: Estimates of minimum and maximum efficacy for malaria prophylaxis Among those who discontinued study drug because of an adverse event.6. The minimum efficacy for prevention of P falciparum malaria was estimated to be 100% (95% CI 59–100%) in the atovaquone-proguanil group and 70% (95% CI 35–93%) in the chloroquineproguanil group (table 4). Five treatment-limiting adverse events attributed to study drug arose in four participants. 1892 THE LANCET • Vol 356 • December 2. A serious adverse event occurred in six participants who received atovaquoneproguanil (infectious illnesses in four. In our study. we obtained blood specimens at the time of suspected malaria and serum samples before travel and several weeks after treatment. but evaluation of several indices indicated that atovaquone-proguanil was significantly better tolerated with respect to gastrointestinal events.24 Failure to complete the full course of antimalarial prophylaxis will increase the risk of developing malaria. despite evidence of decreasing effectiveness as the prevalence of multidrug-resistant P falciparum increases. No clinically important laboratory abnormalities were identified. 2000 Copyright © 2000 All Rights Reserved . For the 180 people who had laboratory safety samples collected. We assessed the efficacy of chemoprophylaxis as a secondary endpoint. or 11 days after leaving Mali. and diagnostic increase in blood-stage antibodies (three).ARTICLES Atovaquone-propguanil Number of participants 60-day efficacy data available CS* antibodies present Confirmed P falciparum malaria Minimum efficacy (95% CI) Maximum efficacy (95% CI) 501 7 0 100% (59–100%) 100% (99–100%) Chloroquine-proguanil 507 8 3† 70% (35–93%) 99% (98–99%) *CS=circumsporozoite. and all three had mutations in the dhfr gene linked to antifolate drug resistance. †One participant with confirmed malaria also developed CS* antibodies. P falciparum malaria was diagnosed while they were receiving chloroquoneproguanil. or Uganda. Discussion Chloroquine and proguanil are generally regarded as the safest drugs available for malaria prophylaxis. and depression in one). This assumption is lent support by the observation that circumsporozoite antibodies were detected in only one of three people with confirmed P falciparum malaria. the event was attributed to treatment in 14 of 27. Nigeria.20 Additionally. The findings that treatment-emergent gastrointestinal adverse events attributed to study drug occurred more often in the chloroquine-proguanil group and were the most common reason for premature discontinuation are in accord with those of previous studies showing that gastrointestinal symptoms are the most common adverse events in people receiving chloroquine-proguanil for malaria prophylaxis. Prospective collection of paired serum samples to measure antibodies to the circumsporozoite protein of P falciparum allowed us to estimate the minimum efficacy of chemoprophylaxis. These data provide a potential molecular basis for failure of prophylaxis with chloroquine-proguanil. there were no significant differences in either treatment group between baseline and follow-up values for haematological and clinical chemistry tests. None of these serious events was thought to be caused by study drug. parasite DNA analysis (three).22 The favourable safety profile of these drugs accounts for their continued wide use. In a study in which serum was obtained every 10 days for 2 months from soldiers exposed to malaria. All three isolates carried at least one mutation previously linked to chloroquine resistance. Because many other illnesses can be mistaken for malaria. For this reason. 6. No blood smears or parasite DNA samples were returned to the reference laboratory. 3. 987 had paired samples available for serological testing. circumsporozoite antibodies developed in 15 (1·5%). of these.25 People taking effective antimalarial prophylaxis can develop anticircumsporozoite antibodies after being bitten by infected mosquitoes without developing malaria. and pituitary tumour in one) and in six people who were on chloroquine-proguanil (malaria in three. including one of three people with confirmed P falciparum malaria. the minimum efficacy of chemoprophylaxis was higher for atovaquone-proguanil than for chloroquineproguanil. In this study. The diagnoses were confirmed by review of slides (two). as a surrogate. and are very rarely associated with severe adverse reactions in the recommended doses. randomly assigned to receive atovaquone-proguanil and 26 such events happened in ten randomly assigned to chloroquine-proguanil (table 3). but serological testing for blood-stage antibodies showed a diagnostic titre increase against P ovale. other infectious illnesses in two.

J Beytout (Clermont Ferrand. MMWR CDC Surveill Summ 1999. 1997: 67–78. Lobel HO. and is consistent with the course of events we observed. France). 60: 521–25. Snounou G. Sukwa TY. References 1 Muentener P. 278: 1767–71. Cowman AF. P de Jonge (Netherlands). Germany). Williams HA. and I Joubert did the study and contributed to data review and interpretation of results. Caruana SR. Clin Infect Dis 1998. Gammon S. Nature 2000. M Lebaddi (France). T Hendrickx (Netherlands). Hall EC. K C Kain did the parasite DNA analyses and contributed to interpretation of results and writing the paper.31 Thus atovaquone-proguanil. Mol Cell 2000. Klotz FW. C Brennan (USA). Roberts J. S van Delft (Netherlands). U Bienzle (Berlin. Nomura T. Chisaka N. acts only as a suppressive prophylactic agent against P vivax and P ovale. Bradley DJ. UK). G B Miller (USA). Compliance with and tolerance of mefloquine and chloroquine + proguanil malaria chemoprophylaxis in French short-term travellers to sub-Saharan Africa.7 If the average duration of travel is 2·5 weeks. Evidence for selection for the tyrosine-86 allele of the pfmdr 1 gene of Plasmodium falciparum by chloroquine and amodiaquine. Indirect fluorescent-antibody tests for parasitic diseases. Barrett PJ. Duraisingh MT. Peguet C. H D Nothdurft. a delayed onset of malaria caused by a relapsing type of malaria parasite can therefore occur. P D Clarke. France). Schildknecht J. for the blood-stage antibody assays. 11: 1–8. B Høgh and J D Chulay had primary responsibility for writing the paper. Nature 1990. Neuropsychiatric adverse effects of mefloquine. Both atovaquone and proguanil have causal prophylactic activity directed against developing hepatic forms of P falciparum. Hoffman SL. Lancet 1987. N S Roskell and J D Chulay contributed to study design and data analysis. D Camus. Wilson M. Ebbutt AF. The risk of developing malaria for travellers to East Africa who take no chemoprophylaxis is about 1·2% per month and chloroquine-proguanil has prophylactic effectiveness of 72%. 351: 709–13. Ndjavé M. This study was funded by GlaxoWellcome. van Vugt M. R Marina (Canada). Imported malaria (1985–95): trends and perspectives. the difference in efficacy rates between treatment groups was not analysed statistically. 1: 1277–81. Trials to assess equivalence: the importance of rigorous methods. Saliba KJ. Durrheim DN. Kirk K. Amino acids in the dihydrofolate reductase-thymidylate synthase gene of Plasmodium falciparum involved in cycloguanil resistance differ from those involved in pyrimethamine resistance. France). Adverse effects of antimalarials: an update. South Africa). H Richter (Germany). and T Scott (USA). Germany).31 Contributors B Høgh. Lancet 1993. Mutations in the Plasmodium falciparum digestive vacuole transmembrane protein PfCRT and evidence for their role in chloroquine resistance. Waner S. High sensitivity of detection of human malaria parasites by the use of nested polymerase chain reaction. et al. Foote SJ. et al. R Balschmidt (Denmark). Galatis D. 28 days after completing the standard course of prophylaxis with atovaquone-proguanil. Luzzi GA. USA. Lancet 1998. et al. An evaluation of a thick-smear antigen in the IFA test for malaria antibodies. Viriyakosol S. K Fleischer (Würzburg.30 but neither drug is active against hypnozoites. A randomized. 313: 36–39. Foote SJ. Denmark). double-blind. E Petersen (Copenhagen. et al. One participant developed malaria caused by P ovale. will require additional treatment with primaquine.27 Although our study was clearly underpowered to show higher efficacy than the chloroquine-proguanil combination. JAMA 1997. France). DC. 114: 205–11. Fidock DA. After dosing is discontinued and drug elimination occurs. for the circumsporozoite antibody assays. Schlagenhauf P. Germany. Efficacy and safety of atovaquone/proguanil for suppressive prophylaxis against Plasmodium falciparum malaria. Kremsner PG. like chloroquine and mefloquine. 313: 525–28. Kremsner PG. Scott T. C Hedgley (UK). France). a study in travellers designed to show that a new antimalaria drug with 95% efficacy is better than chloroquine-proguanil would require more than 16 000 participants.ARTICLES Because of the small number of endpoints. Jones B. 48: 1–23. Carme B. With these expected rates. GlaxoWellcome members of the Study Team include: Amod (South Africa). Nosten F. et al. O Bouchaud (Paris. Tennyson S. Am J Trop Med Hyg 1969. V. Reed MB. J Travel Med 1999. Bull World Health Organ 1999. Germany). M Günther. Several allelles of the multidrug-resistance gene are closely linked to chloroquine resistance in Plasmodium falciparum. and Marianna Wilson of the Centers for Disease Control and Prevention. T J L M Goud (Rotterdam. Am J Trop Med Hyg 1999. J Olsen (Denmark). Sulzer AJ. 403: 906–09. Gordon DM. D Shaw supervised the study. the expected rate of malaria in travellers taking prophylaxis would be 0·2% per month with chloroquine-proguanil and 0·035% per month with a drug that has 95% effectiveness. France). WHO. S Toovey (Sunninghill. D Overbosch. Sherwood JA. 1995. Harrington MA. Atovaquone and proguanil hydrochloride for prophylaxis of malaria. Nevez G. Kyle DE. Pgh1 modulates sensitivity and resistance to multiple antimalarials in Plasmodium falciparum. Emmins PD. Keystone JS. 27: 494–99. Update on prevention of malaria for travellers. Drug Safety 1993. USA. Antimalarial 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Acknowledgments We thank Kent Kester and Carolyn Holland of the Walter Reed Army Institute of Research. 345: 255–58. BMJ 1996. Muller O. Kain KC. Scott TR. Mol Biochem Parasitol 1993. 27: 142–49. Peto TE. Lewis JA. Steffen R. Germany). and B Zieger (Dresden. 2000 1893 Copyright © 2000 All Rights Reserved . Jarvis P. Proc Nat Acad Sci USA 1990. Trop Med Int Health 1997. Netherlands). Kozarsky PE. France). Shanks GD. What do we know and what should we do? CNS Drugs 1999. 87: 3014–17. 77: 560–66.29. Talley AK. 6: 861–71. Lell B. B Marchoux (Toulouse. 341: 1299–303. Imported malaria: prospective analysis of problems in diagnosis and management. Clarke PD. Sukwa TY. 61: 315–20. Other members of the Malarone International Study Team include: R H Behrens (London. Drakeley CJ. 18: 199–205. V Masson (Lille. 2: 953–56. Canada). P Inglebert (Lille. South Africa). Tübingen. Geneva: WHO. Clin Infect Dis 1998. and those who develop malaria caused by either of these parasites. H Schilthius (Amsterdam. Mefloquine compared with other malaria chemoprophylactic regimens in tourists visiting East Africa. Zhu XP. M Peters (Hamburg. Comparison of adverse events associated with use of mefloquine and combination of chloroquine and proguanil as antimalarial prophylaxis: postal and telephone survey of travellers. International travel and health. Netherlands). Ballou WR. Malaria surveillance— United States. J Delmont (Marseilles. Fuchs E. atovaquoneproguanil is more effective than these drugs for treatment of falciparum malaria28 and therefore could be more effective than chloroquine-proguanil for prophylaxis in areas where multidrug resistant P falciparum infections occur. France). Washington. 22 23 24 THE LANCET • Vol 356 • December 2. Martin RK. Steffen R. Luckner D.29. S Waner (Johannesburg. et al. Atlanta. B Gachot (Paris. et al. 6 (suppl 1): S21–S27. Mulenga M. E Dutoit (Lille. et al. et al. Safety and efficacy of a recombinant DNA Plasmodium falciparum sporozoite vaccine. M O’Hare (UK). Braack LE. J Knobloch. R Tan (Vancouver. BMJ 1996. Shanks GD. GA. Cowman AF. Travellers with intense exposure to P vivax or P ovale. Kachur SP. and assuming ␣ is 5% and power is 80%. placebo-controlled field trial to determine the efficacy and safety of Malarone (atovaquone/proguanil) for the prophylaxis of malaria in Zambia. K C Kain was supported in part by a career award from the Ontario Ministry of Health and a grant from PSI. Randomised placebo-controlled study of atovaquone plus proguanil for malaria prophylaxis in children. Parasitology 1997. A Goetschel (France). 8: 295–311. Roskell NS.

2nd edn. closer examination revealed Kayser-Fleischer rings. Verhave JP. Ranasinha CD. a diagnosis of Wilson's disease presenting with Fanconi’s syndrome leading to rickets was made. 40: 105–51. Clinical picture Short stature and rickets A S Kashyap. Canfield CJ. muscle pain. and fibular metaphyses with transverse growth arrest lines. S Afr Med J 1999. and genu-valgum. Wilairatana P. 3: 66–69. Prophylactic activity of atovaquone against Plasmodium falciparum in humans. Wirtz RA. Australia. Atovaquone and proguanil hydrochloride followed by primaquine for treatment of Plasmodium vivax malaria in Thailand. 93: 637–40. 2000 Copyright © 2000 All Rights Reserved . clinical and radiological features of rickets. Chulay JD. 1981: 38–42. Department of Medicine. Development of immunity in natural Plasmodium falciparum malaria: antibodies to the falciparum sporozoite vaccine 1 antigen (R32tet32). 60: 533–41. Trop Med Int Health 1998.ARTICLES 25 26 27 28 prophylaxis—use and adverse events in visitors to the Kruger National Park. Permpanich B. moderate pelvic girdle muscle weakness. An experimental investigation undertaken by the L. Medical Research Unit (A. Docters van Leeuwen WM. tibial. et al. India ( V Kumar MD. and laboratory investigations consistent with renal tubular acidosis. On examination he was found to have short stature. 60: 831–36. Trans R Soc Trop Med 1946. Laboratory investigations were consistent with Fanconi syndrome with proximal renal tubular acidosis. abnormal trabecular patterns. Kumar N. Trans R Soc Trop Med 1999.4888) in malaria. Pang LW.H. New York: John Wiley. Vivek Kumar A 14-year-old male was referred for growth arrest. Am J Trop Med Hyg 1999. Glanarongran R. Boudreau EF. 31 Looareesuwan S. Am J Trop Med Hyg 1999. 25: 1002–08. Bwire R. 29 Fairley NH. Fleiss JL. Sookto P. and a painful waddling gait. He was being managed as a case of renal rickets. His pelvic radiograph showed homogeneous ground glass appearance and deformity.Q. A S Kashyap MD) 1894 THE LANCET • Vol 356 • December 2. Bruins J. genu-valgum. J Clin Microbiol 1987. Researches on paludrine (M. Hutchinson DB.I. Pune 411 040. Knee radiographs showed widened femoral. However. Cairns. 89: 170–75. Armed Forces Medical College. Malaria anticircumsporozoite antibodies in Dutch soldiers returning from sub-Saharan Africa.F). The clinical diagnosis of Wilson's disease was confirmed by serum caeruloplasmin levels of 150 mg/L (270–370 mg/L). With the combination of Kayser-Fleischer rings. Looareesuwan S. Barditch-Crovo P. 30 Shapiro TA. Physical examination was otherwise normal. Slootman EJ. Malarone (atovaquone and proguanil hydrochloride): a review of its clinical development for treatment of malaria. Webster HK. and a malunited femoral shaft fracture. Statistical methods for rates and proportions.