MAJOR ARTICLE

Treatment of Uncomplicated
Multidrug-Resistant Falciparum Malaria
with Artesunate-Atovaquone-Proguanil
Miche`le van Vugt,1,2,3 Elisabetta Leonardi,1 Lucy Phaipun,1 Thra Slight,1 Kyaw Lay Thway,1 Rose McGready,1,2,4
Alan Brockman,1,2,4 Leopoldo Villegas,1 Sornchai Looareesuwan,2 Nicholas J. White,2,4 and Francois Nosten1,2,4
1
Shoklo Malaria Research Unit, Mae Sot, and 2Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand; 3Academic Medical Centre,
Department of Internal Medicine, Slotervaart Hospital, Amsterdam, The Netherlands; and 4Centre for Tropical Medicine, Nuffield Department
of Medicine, John Radcliffe Hospital, Oxford, United Kingdom

Downloaded from http://cid.oxfordjournals.org/ by guest on June 16, 2014

In an open-label trial carried out on the northwest border of Thailand, 1596 patients with uncomplicated
multidrug-resistant falciparum malaria were randomly assigned to receive atovaquone-proguanil, atovaquone-
proguanil-artesunate, or artesunate-mefloquine and were followed up for 42 days. All 3 regimens were highly
effective and well tolerated. Fever duration and parasite clearance times were significantly shorter among
patients who received artesunate (P ! .001 ). Polymerase chain reaction genotyping confirmed that recrudes-
cence occurred in 13 patients who received artesunate-mefloquine (2.4%), 5 who received atovaquone-
proguanil-artesunate (0.9%), and 15 who received atovaquone-proguanil (2.8%). Adding artesunate to ato-
vaquone-proguanil reduced the risk of failure 3-fold (95% confidence interval [CI], 1.18.2) and subsequent
gametocyte carriage 21-fold (95% CI, 1430). Gastrointestinal complaints in the first 48 h after initiation of
treatment were more common among artesunate recipients, but after day 2, dizziness, sleep disturbance,
nausea, vomiting, and anorexia were more common among mefloquine recipients (P .014 ). Artesunate-
atovaquone-proguanil is a highly effective and well-tolerated treatment for multidrug-resistant falciparum
malaria.

Malaria mortality appears to be increasing in most of
the tropical world. This is attributed to increasing re-
sistance to affordable and available antimalarial drugs
[1]. On the northwestern border of Thailand, Plasmo-
Received 28 June 2002; accepted 12 September 2002; electronically published
3 December 2002.
Institutional approval was obtained from the Ethical Committee of the Faculty
of Tropical Medicine, Mahidol University, Bangkok, and the Karen Refugee
Committee, Mae Sot, Thailand. The purpose and the methods of the study were
explained to all participants in their own language, and informed written consent
was obtained.
Financial support: Wellcome-Trust Mahidol University Oxford Tropical Medicine
Research Programme (funded by the Wellcome Trust of Great Britain). Atovaquone-
proguanil was donated by Glaxo-SmithKline.
Reprints or correspondence: Dr. Francois Nosten, Shoklo Malaria Research Unit,
PO Box 46, Mae Sot, 63110, Thailand (shoklo@cscoms.com).
Clinical Infectious Diseases 2002; 35:1498504
 2002 by the Infectious Diseases Society of America. All rights reserved.
1058-4838/2002/3512-0007$15.00
dium falciparum has developed resistance to all avail-
able antimalarial drugs, with the exception of the
artemisinin derivatives [2]. In camps for displaced per-
sons, the systematic use of the combination of meflo-
quine and artesunate has resulted in a 9-fold reduction
in the incidence of falciparum malaria and has halted
the progression of mefloquine resistance [3]. Artemi-
sinin-based antimalarial combination therapies are now
accepted as the optimum approach to malaria che-
motherapy, and they are increasingly becoming part of
first-line recommendations, especially in countries in
which multidrug-resistant malaria occurs [4]. Never-
theless, new drugs are needed, particularly because
antimalarials such as mefloquine are available widely
as monotherapies and, therefore, selection for resistance
continues.
Atovaquone-proguanil (Malarone; Glaxo-SmithKline)
is a new antimalarial compound, a fixed combination

1498 CID 2002:35 (15 December) van Vugt et al.

of atovaquone and proguanil, with a novel mode of action
unrelated to that of other antimalarial drugs [5]. It is very well
tolerated and effective against multidrug-resistant P. falciparum
isolates [6, 7]. When atovaquone was used alone initially, high-
level resistance was selected rapidly [8]. The addition of pro-
guanil reduced this considerably, but because proguanil is a
relatively weak antimalarial drug, the degree of protection
against resistance may not be sufficient to prevent the devel-
opment of resistance to the combination if it is used widely.
After carrying out a preliminary trial demonstrating the absence
of pharmacological interaction between the 3 drugs [9], we
conducted a very large comparative study to establish the ef-
ficacy and tolerability of atovaquone-proguanil with and with-
out artesunate for the treatment of multidrug-resistant malaria.
METHODS
This study was conducted in the Maela and Mawker Tai malaria
clinics of the Shoklo Malaria Research Unit (Mae Sot, Thailand)
between June 1998 and July 2000. Patients were recruited from
2 populations: displaced people of the Karen ethnic minority
and migrant workers living on the western border of Thailand.
This is an area of low and unstable transmission of Plasmodium
vivax and multidrug-resistant P. falciparum [10]. The trial was
an open-label, 3-arm, randomized study comparing atova-
quone-proguanil-artesunate, atovaquone-proguanil, and the
standard regimen, mefloquine-artesunate, for the treatment of
uncomplicated falciparum malaria.
Patients. We recruited 1596 patients with uncomplicated
P. falciparum infections. The artesunate-mefloquine and ato-
vaquone-proguanil-artesunate groups included 533 patients
each, and the atovaquone-proguanil group included 530; the
age range for all patients was 270 years. Baseline characteristics
were similar in the 3 groups, indicating that randomization was
adequate (table 1). Nearly 92% of the patients in the study
(1458 of 1586) were seen at the day 7 scheduled visit, 83.6%
(1313 of 1571) were seen at day 28, and 88.7% (1352 of 1524)
were seen at day 42 (the total number decreased because some
patients experienced treatment failure or withdrew from the
study). Thirteen patients did not complete the 3-day regimen
(1 from the artesunate-mefloquine group, 7 from the atova-
quone-proguanil-artesunate group, and 5 from the atova-
quone-proguanil group). Their exclusion from the analysis did
not influence any of the results. Three patients died (2 from
the artesunate-mefloquine group and 1 from the atovaquone-
proguanil group) for reasons unrelated to the trial. One of these
patients (from the atovaquone-proguanil group), who had a
history of a heavy amphetamine abuse, died from a stroke 10
days after the end of follow-up. Another (from the artesunate-
mefloquine group) died in a land-mine blast, and the third
(from the artesunate-mefloquine group) died of AIDS.
Artesunate-Atovaquone-Proguanil for Malaria CID 2002:35 (15 December) 1499
Inclusion criteria. Patients 110 kg in weight who had
slide-confirmed acute falciparum malaria were included in the
study, provided that they or their guardians gave fully informed
written consent in their own language, they were not pregnant,
they had not received mefloquine in the previous 63 days, they
were not obtunded or vomiting, and there were no other clinical
or laboratory signs of severe illness [11].
Procedures. On enrollment, a full medical history was re-
corded, and a clinical examination was done. Blood films
stained with Giemsa were read by experienced microscopists.
Patients were seen daily for clinical examination, drug admin-
istration, recording of subjective and objective side effects (in-
cluding a standard neurological examination), and blood smear,
until they were apyrexial and aparasitemic. Thereafter, they
were seen weekly for 6 weeks. At each visit, a questionnaire
was administered to gather information about adverse effects,
and a clinical examination was done. Fingerprick blood samples
Downloaded from http://cid.oxfordjournals.org/ by guest on June 16, 2014
for malaria smears and hematocrit determination were obtained
at each visit. Blood samples for biochemical testing were taken
on day 4 from 30 patients in each group; these patients were
randomly selected and matched for age, weight, and level of
parasitemia.
Drug treatments. Randomization was in blocks of 12. Af-
ter enrollment, a sealed envelope was opened that contained
the treatment allocation. This was either artesunate, 4 mg/kg
once daily for 3 days (day 0 was the first day of treatment),
plus mefloquine, 15 mg/kg on day 1 and 10 mg/kg on day 2;
artesunate, 4 mg/kg, plus atovaquone, 15 mg/kg/day, plus pro-
guanil, 8 mg/kg/day, once daily for 3 days; or atovaquone, 15
mg/kg/day, plus proguanil, 8 mg/kg/day, once daily for 3 days.
These atovaquone-proguanil doses correspond to the follow-
ing numbers of tablets: for patients weighing 1120 kg, 1 tablet;
for patients weighing 2130 kg, 2 tablets; for patients weighing
3140 kg, 3 tablets; and for patients weighing 140 kg, 4 tablets.
Patients were also randomly assigned to receive treatment with
either low-fat (ultra-high temperature low-fat milk) or high-
fat (ultra-high temperature sweetened chocolate milk) supple-
ment as part of a separate population pharmacokinetic study,
which will be reported elsewhere.
Each patient whose axillary temperature was 138C was given
antipyretics and cooled by tepid sponging before drug admin-
istration. After they had received the drugs, patients were ob-
served for 1 h, and if they vomited, the drugs were given again
(one-half of the mefloquine dose was given if vomiting occurred
between 30 and 60 min after the first administration).
Outcome measures. The primary therapeutic outcome
measure in this study was the incidence of microscopically and
genotypically confirmed recrudescent infections in the 3 treat-
ment groups by day 42. Patient data were analyzed on an intent-
to-treat basis. At this study site, parasite population genotyping
for 3 polymorphic gene loci (MSP1, MSP2, and GLURP) by
 Table 1. Admission variables for patients with uncomplicated multidrug-resis-
tant falciparum malaria in a comparison trial of atovaquone-proguanil (AP), ato-
vaquone-proguanil-artesunate (AAP), and artesunate-mefloquine (MAS3).

Patients Patients Patients

receiving MAS3 receiving AAP receiving AP
Variable (n p 533) (n p 533) (n p 530)
Clinic, no. of patients
Maela 486 487 486
Mawker Tai 47 46 44
Male sex, % of patients 69 67 67
Age, years
Mean (SD) 21.6 (13) 22.3 (13.7) 22.6 (13.8)
Range 268 370 265
Age group, no. (%) of patients
!5 years 31 (5.8) 25 (4.7) 22 (4.2)
514 years 138 (25.9) 156 (29.3) 155 (29.2)
15 years 364 (68.3) 352 (66) 353 (66.6)

Downloaded from http://cid.oxfordjournals.org/ by guest on June 16, 2014

Weight, kg
Mean (SD) 41.2 (14.3) 40.5 (14.4) 41.7 (14.4)
Range 1074 1168 1073
Hepatomegaly, % of patients 21.1 19.8 18.9
Splenomegaly, % of patients 33.2 25.4 29.4
Temperature, C
Mean (SD) 37.5 (1) 37.7 (1) 37.6 (1)
Range 3540.5 35.540.5 35.540.8
Fever,a % of patients 43.3 48.7 46.4
Parasitemia, parasites/mL
Geometric mean 4055 4211 3734
Range 7239,883 7204,174 12194,984
Hematocrit, %
Mean (SD) 36.7 (6.9) 37 (6.4) 36.8 (6.8)
Range 1661 1859 1754
a
Temperature 137.5C.

polymerase chain reaction (PCR) allows confident discrimi-
nation between recrudescent and new infections [12]. Second-
ary measures were the immediate treatment responses: parasite
clearance, fever clearance, incidence of adverse effects, and de-
gree of anemia.
A target sample size of 534 in each group was calculated
to allow detection of the difference between the cure rates in
the treatment groups (96% and 90%) with 95% confidence
and 90% power (allowing for a dropout rate of up to 30%
at day 42).
Adverse effects. Adverse effects were symptoms or signs
that were not present at admission and that developed after
the start of treatment. All potential adverse effects, including
those probably related to malaria, were compared among treat-
ment groups. The rates of early vomiting (!1 h) after each dose
and for each drug were recorded and compared among the
groups in the analysis.
Management of recrudescent infections. Patients with un-
complicated recrudescent infections were re-treated and were
followed up for a second period as in the first part of the study.
The second course of treatment included artesunate, 2 mg/kg/
day for 7 days; nonpregnant patients 18 years old also received
doxycycline, 4 mg/kg/day for 7 days. Blood samples were ob-
tained, when possible, from patients with suspected recrudes-
cent infections after atovaquone-proguanil administration for
parasite in vitro sensitivity testing.
Statistical analysis. Data were analyzed using SPSS for
Windows, version 10. Categorical data were compared using
the x2 test with Yates correction or Fishers exact test, as ap-
propriate. Continuous variables conforming to a normal dis-
tribution were compared using Students t test. Data not nor-
mally distributed were log-transformed or compared using the
Mann-Whitney U test. The rates of adverse effects at 3 different
periods (days 12, days 37, and days 1442) were compared

1500 CID 2002:35 (15 December) van Vugt et al.

among treatment groups. For each of the 3 periods, the events
were counted only once (e.g., if a patient vomited on day 1
and day 2, this was counted as 1 adverse event). When the rate
of an adverse event was similar in the 2 groups of patients
receiving atovaquone-proguanil, the groups were pooled, and
the rate was compared with that of the artesunate-mefloquine
group. The risk of treatment failure was evaluated by survival
analysis and compared using the Mantel-Haenszel log-rank test.
Patients in whom parasitemia reappeared by day 42 and for
whom PCR genotyping was either inconclusive or missing were
counted as having treatment failure.

RESULTS
Clinical Findings
Fever cleared more rapidly in the 2 groups that received ar-
tesunate. By day 2, 6 (1.1%) of 533 and 15 (2.8%) of 533
Figure 1. Proportion of patients experiencing treatment failure after
treatment with artesunate-mefloquine (MAS3; artesunate, 4 mg/kg once
daily for 3 days, plus mefloquine, 15 mg/kg on day 1 and 10 mg/kg on
day 2), atovaquone-proguanil-artesunate (AAP; artesunate, 4 mg/kg, plus
atovaquone, 15 mg/kg/day, plus proguanil, 8 mg/kg/day, once daily for

Downloaded from http://cid.oxfordjournals.org/ by guest on June 16, 2014

patients in the artesunate-mefloquine and atovaquone-
proguanil-artesunate groups, respectively, were still febrile,
compared with 50 (9.4%) of 530 patients in the atovaquone-
proguanil group (P ! .001). None of the patients developed
severe malaria. The mean (SD) decrease in hematocrit values
at day 7 from baseline was greater in the groups receiving
artesunate than in the atovaquone-proguanil group: 2.6% (SD,
5.10%; 95% CI, 2.07%3.04%) for the artesunate-mefloquine
group and 3.0% (SD, 4.71%; 95% CI, 2.56%3.45%) for the
atovaquone-proguanil-artesunate group, compared with 1.5%
(SD, 5.54%; 95% CI, 0.98%2.02%) for the atovaquone-pro-
guanil group (P p .003 and P ! .0001, respectively). By day 28,
the mean hematocrit values were equal to or greater than the
baseline values in all 3 groups. There was no difference in the
clearance of hepatosplenomegaly in the 3 treatment groups.
Parasitological Findings
Treatment failure: incidence of recrudescent infections. PCR
confirmed that recrudescence rates were very low in the 3
groups: 5 of 533 patients (0.9%; 95% CI, 0.3%2.1%) in the
atovaquone-proguanil-artesunate group; 13 of 533 (2.4%; 95%
CI, 1.2%4.1%) in the artesunate-mefloquine group; and 15
of 530 (2.8%; 95% CI, 1.6%4.6%) in the atovaquone-
proguanil group (figure 1). The failure rates for artesunate-
mefloquine and atovaquone-proguanil were similar. The rela-
tive risks (RRs) of failure in the atovaquone-proguanil and
artesunate-mefloquine groups, compared with the atovaquone-
proguanil-artesunate group, were 3.0 (95% CI, 1.108.22;
P p .02) and 2.6 (95% CI, 0.937.24; P p .057), respectively.
As in all previous studies at this site, PCR-confirmed treatment
failures were more likely in children than in adults: 2 of 78
patients aged !5 years and 15 of 433 patients aged 514 years
experienced treatment failure, versus 16 of 1069 patients aged
15 years (P p .040; x2 for trend). The median time to patent
3 days), and atovaquone-proguanil (AP; atovaquone, 15 mg/kg/day, plus
proguanil, 8 mg/kg/day, once daily for 3 days).
recrudescence was shorter for artesunate-mefloquine (25 days;
range, 1442 days) than for atovaquone-proguanil-artesunate
(36 days; range, 2842 days; P p .03) and atovaquone-pro-
guanil (29 days; range, 1442 days; P p .14).
The efficacy rates were not significantly influenced by the fat
content of the supplement administered with the drugs. All 33
patients with recrudescent infections were re-treated, but only
12 could be followed up again. Two of these re-treated patients
experienced a second PCR-confirmed treatment failure.
Other parasitological findings. Parasite clearance was
slower in the atovaquone-proguanil group. By day 3, 36 (6.8%)
of 530 patients in the atovaquone-proguanil group still had a
positive blood film, compared with 2 (0.4%) of 533 atova-
quone-proguanil-artesunate recipients and 2 (0.4%) of 533 ar-
tesunate-mefloquine recipients (P ! .001 for each).
During the 42-day follow-up period, 22 new P. falciparum
infections occurred among artesunate-mefloquine recipients and
13 occurred in each of the other 2 groups. Of 1596 patients, 216
(13.5%) had P. vivax parasitemia detected during this period.
There were significantly fewer cases of vivax malaria in the ar-
tesunate-mefloquine group (24 of 533) than in the atovaquone-
proguanil-artesunate (100 of 533) and atovaquone-proguanil (92
of 530) groups (P ! .001). The mean time to appearance of P.
vivax parasitemia was longer in the artesunate-mefloquine group
(38.6 days; SD, 5.2 days) than in the atovaquone-proguanil-
artesunate and atovaquone-proguanil groups pooled together
(30.9 days; SD, 6.8 days; P ! .001).
Overall, 9.1% of patients (145 of 1596) presented with ga-
metocytes at admission. This proportion was similar in the 3
groups. Gametocyte carriage during follow-up was considerably
higher in the atovaquone-proguanil group (62 person-game-

Artesunate-Atovaquone-Proguanil for Malaria CID 2002:35 (15 December) 1501

tocyte-weeks (PGW)/1000 person-weeks vs. 2.9 PGW/1000 per-
son-weeks in the atovaquone-proguanil-artesunate group and
3 PGW/1000 person-weeks in the artesunate-mefloquine
group). Thus, gametocyte carriage was 21 times greater (95%
CI, 1430 times) among patients who did not receive artesunate
than in those who did.
In vitro sensitivity. The median in vitro sensitivity (IC50)
to atovaquone of 10 isolates obtained from patients who had
recrudescence of infection after treatment with atovaquone-
proguanil was 1.05 ng/mL (range, 0.3114.29 ng/mL). This was
similar to the median value for 39 isolates obtained at admission
in the study (from patients who had never received atovaquone-
proguanil), 0.81 ng/mL (range, 0.046.54 ng/mL). None of the
isolates tested had the marked increases in atovaquone IC50
described in earlier reports of atovaquone treatment failures
[8]. However, isolates from 6 patients treated with atovaquone-
proguanil who had recrudescence of infection did not grow in
culture.
Adverse Events
All 3 regimens were generally well tolerated. No serious adverse
effects were reported. In particular, there were no neuropsy-
chiatric reactions and no severe skin rashes, blood dyscrasias,
or liver and renal function (biochemical) abnormalities.
On day 0, vomiting of artesunate tablets within 1 h of ad-
ministration (early vomiting) was more frequent among ato-
vaquone-proguanil-artesunate recipients than among artesun-
ate-mefloquine recipients: 9 (1.7%) of 533 patients versus 1
(0.2%) of 533 patients (RR, 9; 95% CI, 1.1470.8; P p .01).
The corresponding rates of vomiting for atovaquone-proguanil
tablets were 15 (2.8%) of 530 patients in the atovaquone-
proguanil group and 6 (1.1%) of 533 patients in the atova-
quone-proguanil-artesunate group (RR, 2.51; 95% CI, 0.98
6.42; P p .04). On subsequent days (days 1 and 2), the rates
of early vomiting of the drugs were low and did not differ
among groups (data not shown).
On days 1 and 2, the rate of nausea in the atovaquone-
proguanil group was 29 (8.7%) of 334 patients who did not
have this symptom at admission to the hospital, compared with
Table 2. Incidence of adverse events at days 12 among patients with uncompli-
cated multidrug-resistant falciparum malaria in a comparison trial of atovaquone-
proguanil (AP), AP-artesunate (AAP), and artesunate-mefloquine (MAS3).
Patients receiving
Adverse event MAS3, n/N (%)
Dizziness 97/223 (43.5)
Palpitations 58/273 (21.2)
Tremor 13/491 (2.6)
Anorexia 65/165 (39.4)
NOTE. n/N, No. of patients with symptom/no. of patients who did not have that symptom at
admission; RR, relative risk.
1502 CID 2002:35 (15 December) van Vugt et al.
63 (18%) of 350 patients in the artesunate-mefloquine group
(RR, 2.30; 95% CI, 1.443.68; P ! .0001) and 55 (16.7%) of
330 patients in the atovaquone-proguanil-artesunate group
(RR, 2.09; 95% CI, 1.293.38; P p .01). The rate of late vom-
iting (11 h after administration of treatment) in the atova-
quone-proguanil-artesunate group was higher than that in the
atovaquone-proguanil group (51 [12.6%] of 406 versus 34
[7.9%] of 428 patients, respectively; RR, 1.65; 95% CI,
1.052.62; P p .019) but was similar to that among artesunate-
mefloquine recipients (41 [10.0%] of 411; P p .26).
Of 308 patients in the atovaquone-proguanil group who did
not have chills or rigors at admission, 19 (6.1%) complained
of these symptoms on days 1 and 2, compared with 30 (10.4%)
of 286 patients in the artesunate-mefloquine group (RR, 1.30;
95% CI, 1.021.65; P p .039) and 31 (10.2%) of 303 patients
in the atovaquone-proguanil-artesunate group (RR, 1.27; 95%
CI, 1.011.61; P p .046). For other adverse effects, the pooled
Downloaded from http://cid.oxfordjournals.org/ by guest on June 16, 2014
data from atovaquone-proguanil-artesunate and atovaquone-
proguanil recipients were compared with data from artesunate-
mefloquine recipients. Significant results are summarized in
table 2.
During days 37, mefloquine-artesunate recipients had
higher rates of sleeping disorders, dizziness, nausea, vomiting,
and skin rash than did the pooled atovaquone-proguanil-
artesunate and atovaquone-proguanil groups (table 3). During
days 1442, more patients in the atovaquone-proguanil-
artesunate group complained of abdominal pain (29 [7.9%] of
365) than did atovaquone-proguanil recipients (18 [4.7%] of
387; RR, 1.29; 95% CI, 1.021.64; P p .043). There was no
other difference in the incidence of adverse effects among the
3 groups during this late period of follow-up.
Thus, during the first days of treatment, the patients in the
artesunate-mefloquine and atovaquone-proguanil-artesunate
groups had more gastrointestinal side effects than did patients
in the atovaquone-proguanil group. However, during the fol-
lowing days (37), artesunate-mefloquine was associated with
more adverse effects, in particular nausea, vomiting, and sleep-
ing disorders, than were the other 2 treatment regimens (tables
Patients receiving
AP or AAP, n/N (%) P RR (95% CI)
139/453 (30.7) !.0001 1.91 (1.372.60)
75/730 (10.3) .008 1.61 (1.12.32)
10/994 (1) .017 2.60 (1.166.14)
103/354 (29.1) .013 1.58 (1.072.33)
 Table 3. Incidence of adverse events at days 37 among patients with uncomplicated
multidrug-resistant falciparum malaria in a comparison trial of atovaquone-proguanil
(AP), atovaquone-proguanil-artesunate (AAP), and artesunate-mefloquine (MAS3).
Patients receiving
Adverse event MAS3, n/N (%)
Vomiting 9/372 (2.4)
Nausea 14/319 (4.4)
Sleeping disorders 27/329 (8.2)
Dizziness 44/199 (22.1)
Palpitations 16/335 (4.8)
Skin rash 0/479
NOTE. n/N, No. of patients with symptom/no. of patients who did not have that symptom at admission;
RR, relative risk.
2 and 3). No significant abnormalities were noted on neuro-
logical testing.
DISCUSSION
The treatment of malaria is changing. Previously, as one drug
failed, it was replaced eventually by another compound, and
this, after a shorter period of useful life, in turn would fail.
The sequential loss of affordable effective antimalarial drugs to
resistance represents a major threat to the people of tropical
countries. On the western border of Thailand, this process has
been reversed. The systematic use of a combination of arte-
sunate and mefloquine has led to a reversal of mefloquine
resistance and sustained antimalarial efficacy [3]. The present
very large trial confirms that, 8 years after artesunate-meflo-
quine came into wide-scale use in the treatment of falciparum
malaria in this area, its efficacy remains almost 100%. Early
diagnosis and treatment with a drug combination of very high
efficacy that reduces gametocyte carriage has led to a marked
decline in the incidence of falciparum malaria and a reversal
of the previous trend toward increasing mefloquine resistance
[2, 13]. Furthermore, because treatment responses are accel-
erated by artemisinin derivatives, deterioration of the patients
condition to severe malaria is extremely unusual. However, the
continued use of mefloquine monotherapy, even in Thailand,
provides continued selection pressure to the evolution of me-
floquine resistance, which could compromise the effectiveness
of the artesunate-mefloquine combination.
Atovaquone interferes with plasmodial respiration. Because
this mode of action differs from those of existing classes of
antimalarials, there is no significant cross-resistance. Unfor-
tunately, parasites with single-point mutations in the cyto-
chrome b gene show high levels of atovaquone resistance [14],
and these mutants occur naturally at high frequencies [15].
These mutations also confer complete resistance to the com-
bination of atovaquone and proguanil. They would certainly
be selected very rapidly if atovaquone were used alone. Pro-
Artesunate-Atovaquone-Proguanil for Malaria CID 2002:35 (15 December) 1503
Patients receiving
AP or AAP, n/N (%) P RR (95% CI)
5/772 (0.6) .014 3.8 (1.2611.42)
8/612 (1.3) .004 3.46 (1.438.35)
23/692 (3.3) .001 2.6 (1.464.61)
42/421 (10.0) !.0001 2.56 (1.614.06)
17/681 (2.5) .044 1.95 (0.973.92)
10/958 (1) .018 1.49 (1.441.55)
guanil is an antimalarial biguanide; it is biotransformed to the
active metabolite cycloguanil, but it is the less active parent
compound proguanil, rather than the antifol metabolite, that
Downloaded from http://cid.oxfordjournals.org/ by guest on June 16, 2014
synergizes with atovaquone in the combination. This synergy
explains its excellent activity in this study site, despite the high
level of cycloguanil resistance in P. falciparum in the area and
the high prevalence of the CYP2 C19 polymorphism in Asian
populations, which results in markedly reduced biotransfor-
mation of proguanil. However, the extent of the protection
conferred by proguanil against the emergence of atovaquone
resistance is not clear. This study provides some reassurance,
because no atovaquone resistance was detected in 1500 patients
treated with atovaquone-proguanil alone. However, resistance
has been reported after atovaquone-proguanil use, and because
the level of resistance conferred by the single-point mutation
is so high, atovaquone-proguanil must still be considered highly
vulnerable to resistance. The main rationale behind the addition
of artesunate to atovaquone-proguanil was to provide protec-
tion against resistance (and it would be expected to reduce
the probability of resistant mutants emerging by 100 mil-
lionfold). Our results support the idea that highly efficacious
and new antimalarials (such as atovaquone-proguanil) should
be combined with an artemisinin derivative de novo before
resistance to the new drug emerges. In this study, the triple
combination of atovaquone, proguanil, and artesunate pro-
duced a cure rate of almost 100% and was more effective than
artesunate-mefloquine or atovaquone-proguanil alone, even
though atovaquone-proguanil had never been used in the area.
This very high cure rate would help to delay the emergence of
resistance if the drug were used widely.
These antimalarial treatments were generally well tolerated,
and there were no serious adverse effects. This is the largest
single-center randomized antimalarial treatment evaluation,
and it confirms the safety of artesunate, mefloquine, and ato-
vaquone-proguanil. In comparison with the artesunate-meflo-
quine regimen, the 2 regimens that included atovaquone and
proguanil were better tolerated. Despite the larger number of
tablets administered, addition of artesunate to atovaquone and
proguanil may reduce the risk of immediate vomiting of ato-
vaquone-proguanil tablets on the first day of treatment. The
high incidence of chills and rigors reported among the arte-
sunate recipients may be the consequence of faster RBC de-
struction. Taken together with the findings of earlier studies,
these data suggest that atovaquone-proguanil should be used
in combination with artesunate for the treatment of malaria
in areas of endemicity. Unfortunately, this combination is
largely unaffordable in the tropical world.
References
1. Trape JF, Pison G, Spiegel A, Enel C, Rogier C. Combating malaria in
Africa. Trends Parasitol 2002; 18:22430.
2. Brockman A, Price RN, van Vugt M, et al. Plasmodium falciparum
antimalarial drug susceptibility on the north-western border of Thai-
land during five years of extensive use of artesunate-mefloquine. Trans
R Soc Trop Med Hyg 2000; 94:53744.
3. Nosten F, van Vugt M, Price R, et al. Effects of artesunate-mefloquine
combination on incidence of Plasmodium falciparum malaria and me-
floquine resistance in western Thailand: a prospective study. Lancet
2000; 356:297302.
4. World Health Organization (WHO). Antimalarial drug combination
therapy: report of a WHO technical consultation, 45 April 2001.
Geneva: WHO, 2001.
5. Srivastava IK, Rottenberg H, Vaidya AB. Atovaquone, a broad spectrum
antiparasitic drug, collapses mitochondrial membrane potential in a
malarial parasite. J Biol Chem 1997; 272:39616.
1504 CID 2002:35 (15 December) van Vugt et al.
6. Looareesuwan S, Wilairatana P, Glanarongran R, et al. Atovaquone
and proguanil hydrochloride followed by primaquine for treatment
of Plasmodium vivax malaria in Thailand. Trans R Soc Trop Med
Hyg 1999; 93:63740.
7. Bustos DG, Canfield CJ, Canete-Miguel E, Hutchinson DB. Atova-
quone-proguanil compared with chloroquine and chloroquine-sulfa-
doxine-pyrimethamine for treatment of acute Plasmodium falciparum
malaria in the Philippines. J Infect Dis 1999; 179:158790.
8. Looareesuwan S, Viravan C, Webster HK, Kyle DE, Hutchinson DB,
Canfield CJ. Clinical studies of atovaquone, alone or in combination
with other antimalarial drugs, for treatment of acute uncomplicated
malaria in Thailand. Am J Trop Med Hyg 1996; 54:626.
9. van Vugt M, Edstein MD, Proux S, et al. Absence of an interaction
between artesunate and atovaquone-proguanil. Eur J Clin Pharmacol
1999; 55:46974.
10. Luxemburger C, Thwai KL, White NJ, et al. The epidemiology of
malaria in a Karen population on the western border of Thailand.
Trans R Soc Trop Med Hyg 1996; 90:10511.
11. World Health Organization. Control of Tropical Diseases. Severe and
complicated malaria. Trans R Soc Trop Med Hyg 1990; 84(Suppl 2):
185.
12. Brockman A, Paul RE, Anderson TJ, et al. Application of genetic mark-
ers to the identification of recrudescent Plasmodium falciparum infec-
Downloaded from http://cid.oxfordjournals.org/ by guest on June 16, 2014
tions on the northwestern border of Thailand. Am J Trop Med Hyg
1999; 60:1421.
13. Price RN, Nosten F, Luxemburger C, et al. Effects of artemisinin de-
rivatives on malaria transmissibility. Lancet 1996; 347:16548.
14. Korsinczky M, Chen N, Kotecka B, Saul A, Rieckmann K, Cheng Q.
Mutations in Plasmodium falciparum cytochrome b that are associated
with atovaquone resistance are located at a putative drug-binding site.
Antimicrob Agents Chemother 2000; 44:21008.
15. White NJ. Antimalarial drug resistance and combination chemother-
apy. Philos Trans R Soc Lond B Biol Sci 1999; 354:73949.