Content uploaded by Charles O Obonyo
Author content
All content in this area was uploaded by Charles O Obonyo on Oct 11, 2017
Content may be subject to copyright.
Transactions of the Royal Society of Tropical Medicine and Hygiene (2007) 101, 117—126
available at www.sciencedirect.com
journal homepage: www.elsevierhealth.com/journals/trst
Amodiaquine combined with
sulfadoxine/pyrimethamine versus artemisinin-based
combinations for the treatment of uncomplicated
falciparum malaria in Africa: a meta-analysis
Charles O. Obonyoa,∗, Elizabeth A. Jumaa, Bernhards R. Ogutu b,
John M. Vululea, Joseph Lauc
aCentre for Vector Biology & Control Research, Kenya Medical Research Institute, P.O. Box 1578-40100, Kisumu, Kenya
bCentre for Clinical Research, Kenya Medical Research Institute, Nairobi, Kenya
cInstitute for Clinical Research and Health Policy Studies, Tufts-New England Medical Center, Boston, MA 02111, USA
Received 25 November 2005; received in revised form 5 July 2006; accepted 5 July 2006
Available online 15 September 2006
KEYWORDS
Malaria;
Amodiaquine;
Sulfadoxine-
pyrimethamine;
Artemisinin;
Combination drug
therapy;
Drug resistance;
Sub-Saharan Africa
Summary Drug resistance in Plasmodium falciparum is a major obstacle to malaria control.
Artemisinin-based combination therapy (ACT) is being advocated to improve treatment efficacy
and to delay development of resistance. Here we summarise the available data on the efficacy of
amodiaquine plus sulfadoxine/pyrimethamine (AQ+SP) versus ACTs in the treatment of uncom-
plicated malaria in sub-Saharan Africa. We searched for randomised trials in which patients
with uncomplicated malaria treated with AQ+SP were compared with those treated with either
amodiaquine plus artesunate (AQ+AS), artesunate plus sulfadoxine/pyrimethamine (AS+SP) or
artemether/lumefantrine (AL). Medline, EMBASE, Cochrane Central Register of Controlled Tri-
als and reference lists up to July 2005 were searched. Two reviewers independently extracted
the data. The primary outcome measure was treatment failure by Day 28. Outcome measures
were combined using a random effects model. Seven randomised trials of 4472 children were
included. Trial quality was generally high. Treatment failure of AQ+SP was significantly reduced
compared with AS+SP (relative risk (RR) = 0.56, 95% CI 0.42—0.75), but increased compared
with AL (RR = 2.80, 95% CI 2.32—3.39). The overall failure rate of AQ+SP was similar compared
with AQ+AS (RR = 1.12, 95% CI 0.81—1.54), but there was significant heterogeneity of results
across the studies. All the treatment regimens were safe and well tolerated. AQ+SP should be
considered in some settings before the full implementation of an ACT.
© 2006 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights
reserved.
∗Corresponding author. Tel.: +254 57 202 2924, +254 733 837 969; fax: +254 57 202 2940.
E-mail address: cobonyo@kisian.mimcom.net (C.O. Obonyo).
0035-9203/$ — see front matter © 2006 Royal Society of Tropical Medicine and Hygiene. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.trstmh.2006.07.001
118 C.O. Obonyo et al.
1. Introduction
Although malaria is a global disease, over 90% of the bur-
den is borne by populations in sub-Saharan Africa where
Plasmodium falciparum particularly affects young children
and pregnant women (Snow et al., 2005). At least 300 mil-
lion clinical cases of malaria occur each year, resulting in
more than a million childhood deaths globally. Early diagno-
sis and prompt effective treatment remain the cornerstone
strategy for reduction of malaria-related morbidity and mor-
tality (WHO, 1993). The parasite has developed resistance
to most of the standard affordable antimalarial drugs. Drug
resistance is a major obstacle to malaria control in sub-
Saharan Africa, accounting for most of the current increases
in the incidence of malaria-specific morbidity and mortality
(Trape et al., 1998). Despite the huge malaria problem in
sub-Saharan Africa, the resources available to national con-
trol programmes are inadequate.
Treatment of malaria is in transition: for over 50
years, malaria was treated successfully using chloroquine
(CQ). With the emergence of CQ resistance, sulfadox-
ine/pyrimethamine (SP) replaced CQ in many settings, but
unfortunately the effectiveness of SP has decreased in some
areas within a few years of its introduction (Sibley et al.,
2001). Malaria can no longer be treated using a single drug;
however, the list of available antimalarials is limited. More-
over, development of new antimalarial drugs has not kept
pace with the rate of development of drug resistance. Care-
ful use must therefore be made of the available antimalarial
drugs. The incidence of antimalarial drug resistance can
be slowed by the use of combination therapy (White and
Olliaro, 1996). In combination therapy, two or more drugs
with different modes of action are administered simultane-
ously to improve treatment efficacy and to extend the useful
therapeutic life of the constituent drugs by reducing the rate
at which resistance develops (White, 1998).
Of the available antimalarial drugs, the artemisinins are
the most potent and the WHO specifically advocates the
use of artemisinin-based combination therapy (ACT) as the
standard policy in the treatment of uncomplicated falci-
parum malaria (WHO, 1998). ACT results in rapid clinical
and parasitological response, may delay the development
of resistance and may reduce malaria transmission by killing
gametocytes. However, ACTs are relatively expensive, are
untested in sub-Saharan Africa, are currently unavailable
on a wide scale and have complex dosing regimens com-
pared with single-dose SP (Bloland, 2003). By 2004, over
16 countries in sub-Saharan Africa had selected an ACT for
first-line therapy, but implementation of the revised poli-
cies was delayed by both the cost and the unavailability
of ACTs. It is unclear whether ACTs should be used for
empirical management of febrile illnesses outside the for-
mal health sector without laboratory confirmation, in the
same way that CQ and SP were utilised for several years.
Combination regimens including artesunate plus sulfadox-
ine/pyrimethamine (AS+SP) or amodiaquine plus artesunate
(AQ+AS) have demonstrated superior treatment efficacy
compared with that of SP or amodiaquine (AQ) monother-
apy (Adjuik et al., 2002, 2004). However, when artemisinin
derivatives were combined with a standard drug that was
already failing, the efficacy of that combination was greatly
compromised (Obonyo et al., 2003).
Most malaria control programmes are currently revising
their treatment policies to replace the current first-line
failing monotherapies with combination therapy, but they
may be challenged to afford the effective but more expen-
sive ACTs. In these circumstances, the control programmes
may need to consider alternatives to ACT, which include
non-artemisinin-based combination therapy (NACT). Poten-
tial candidates for NACT include amodiaquine plus sulfa-
doxine/pyrimethamine (AQ+SP) or clindamycin plus quinine.
The combination of clindamycin plus quinine is effective,
well tolerated and safe, although it is rarely used (Lell
and Kremsner, 2002). There has been renewed interest in
AQ+SP (Gasasira et al., 2003; Schellenberg et al., 2002). The
WHO’s Roll Back Malaria programme has recommended that
the combination AQ+SP can be used to treat uncomplicated
malaria in settings where the efficacy of both component
drugs is high. In some studies, AQ+SP was shown to be as
effective as or better than ACT in treating uncomplicated
malaria (Dorsey et al., 2002; Rwagacondo et al., 2003). Non-
artemisinin-based drug combinations may offer an attractive
option during a transition period while securing resources
for ACT implementation; in the right setting, the combina-
tion may be relatively effective and the drugs are familiar,
available and affordable.
Changing the malaria treatment policy is a complex,
lengthy and expensive process (Shretta et al., 2001). Many
ministries of health in sub-Saharan Africa may wish to opt
for an efficacious ‘interim’ regimen before fully imple-
menting the ACT policy. The ideal ‘transitional’ antimalar-
ial drug combination has not been found (Kremsner and
Krishna, 2004). Hence, there is uncertainty regarding the
optimal choice of antimalarial drug combinations to use in
the interim and, similarly, there are insufficient data to
guide the choice of ACT. The background treatment fail-
ure rates are likely to determine significantly the timing
of a policy change: higher failure rates with monotherapies
(CQ, SP and AQ) have been documented in eastern African
compared with the western Africa regions. In this system-
atic review, we summarise the available evidence on the
efficacy of one potential transitional regimen (AQ+SP) com-
pared with antimalarial drug combinations containing an
artemisinin derivative in the treatment of uncomplicated
falciparum malaria in Africa.
2. Materials and methods
2.1. Criteria for selecting studies
All randomised controlled trials conducted in Africa that
compared AQ+SP with an ACT in participants with uncom-
plicated falciparum malaria were included. The ACTs were
specified as AQ+AS, AS+SP or artemether/lumefantrine (AL).
Articles published both in English and non-English languages
were included. The primary outcome measure was prospec-
tively defined as parasitological treatment failure by Day 28
after starting treatment. Treatment failure was considered
as the sum of early and late treatment failures, as defined
by the WHO (2002). Secondary outcomes were Day 28
measurements of recrudescence, gametocyte carriage,
mean change in haemoglobin and the incidence of adverse
events. Adverse events were defined as any unfavourable or
Treatment regimens for uncomplicated falciparum malaria 119
unintended sign, symptom or disease temporally associated
with the use of a medicinal product, whether or not it
was considered as related to the medicinal product. A
serious adverse event was defined as a sign, symptom
or intercurrent illness that was fatal, life threatening or
required admission to hospital.
Review papers, studies in pregnant women or participants
with signs of severe malaria, studies comparing an ACT with
another ACT or with monotherapy with standard antimalar-
ial drugs, economic analyses, pharmacokinetic studies and
studies performed outside Africa were excluded.
2.2. Search strategy
Using the OVID platform, the following electronic databases
were searched according to the Cochrane collaboration’s
optimum search strategy for randomised controlled trials
(Alderson et al., 2004): Medline (1966 to July 2005); EMBASE
(1980 to July 2005); and the Cochrane Central Register
of Controlled Trials (CENTRAL) on the Cochrane Library
(Issue 3, 2005). The search terms used included the fol-
lowing: amodiaquine, sulfadoxine/pyrimethamine, fansidar,
artesunate, artemether, benflumetol, co-artemether, lume-
fantrine and malaria. These terms were combined with the
search strategy for retrieving randomised trials (Alderson et
al., 2004). The corresponding authors of each included trial
and experts on this subject were contacted for details of
unpublished or ongoing trials. The reference lists of included
studies and review articles were searched for additional
studies.
2.3. Quality assessment
The following aspects of methodological quality from each
included trial report were independently assessed by two
reviewers, without using a score or being masked: gen-
eration of the allocation sequence; adequacy of conceal-
ment of the allocation of treatment; degree of blinding;
and completeness of follow-up. Generation of the alloca-
tion sequence and allocation concealment were classified
as adequate, inadequate or unclear (Juni et al., 2001).
Blinding was classified as open, single or double. The pro-
portion of patients lost to follow-up was computed and con-
sidered adequate if <10%. The included studies were also
assessed regarding whether a sample size was determined
using power calculations and whether an intention-to-treat
analysis was performed.
2.4. Data extraction
Two reviewers independently screened the results of the
literature search and selected eligible studies according to
pre-set criteria. Differences over inclusion of studies were
resolved by discussion. The following information from each
trial report was abstracted: date of trial; location of trial;
background drug failure rates; study design; inclusion cri-
teria; participants; interventions; and outcomes. The lead
authors were contacted to seek additional information if
data from the published study reports were insufficient or
missing.
2.5. Statistical analysis
Data were extracted to allow for an intention-to-treat anal-
ysis (the analysis includes all participants in the groups to
which they were originally randomised). Information from
comparable trials was combined using a random effects
model (DerSimonian and Laird, 1986). Meta-analyses were
performed on the Day 28 parasitological failure rate as well
as the failure rate corrected by parasite genotyping. In
addition, data were analysed on gametocyte carriage and
haemoglobin changes. Estimates of treatment efficacy were
compared using the relative risk (RR) for dichotomous data
and weighted mean differences (WMD) for continuous data.
RR was computed as the proportion of treatment failures (or
gametocyte carriers) that received AQ+SP divided by those
who received an ACT. The pooled point estimates are pre-
sented together with the 95% CIs. Heterogeneity between
trials was assessed using the I2statistic, which is derived
from Cochran’s Qheterogeneity statistic and describes the
percentage of total variation in the effect estimates due to
heterogeneity rather than chance (Higgins and Thompson,
2002). A value >50% was considered substantial hetero-
geneity. When heterogeneity was detected, explanations
were sought using sensitivity analyses. The following factors
were pre-specified as possible reasons for heterogeneity:
patient age (<5 years and ≥5 years); treatment dosage;
transmission intensity; trial quality; study size; and event
rates of treatment failure. Data were analysed using the
Cochrane collaboration software (Review Manager 4.2.8;
http://www.cochrane.org).
3. Results
3.1. Characteristics of included studies
Sixteen studies were identified comparing AQ+SP with other
antimalarial drugs. Nine of these studies were excluded: the
comparison group was CQ+SP (three trials) or monother-
apy with CQ or SP (five trials), and one study was not a
randomised trial. Thus, a total of seven studies met our
inclusion criteria (Table 1). None of the seven corresponding
authors knew of any other ongoing or unpublished studies.
A total of 4472 children were enrolled in these studies. Indi-
vidually, the trials enrolled between 276 and 1541 children.
AQ+SP was compared with AQ+AS in four trials (Studies I, IV,
VI, VII), AS+SP in four trials (Studies I, II, III, V) and AL in one
trial (Study IV). In two trials (Studies I, IV), AQ+SP was com-
pared with two different artemisinin-based combinations.
One study each was conducted in Rwanda (Study V),
Ghana (Study III), Mozambique (Study I) and Tanzania (Study
IV), and three were carried out in Uganda (Studies II, VI, VII).
The studies were all undertaken between 2001 and 2004 and
included children aged between 4 months and 59 months in
five studies (Studies I, II, III, IV, V) and between 6 months
and 10 years in two studies (Studies VI, VII). The proportion
of children lost to follow-up was <10% in all the included
trials. In all the trials except one (Study V), the method-
ological quality was high and the randomisation sequence
was computer-generated. Two trials were open (Studies I,
IV), three trials were single-blinded (Studies II, V, VII) and
two trials were double-blinded (Studies III, VI). The basis
120 C.O. Obonyo et al.
Table 1 Characteristics of included trials
Study Location Transmission
intensity
No. of participants randomisedaNo. lost to
follow-up
Background drug
failure rate
Inclusion criteria
AQ+SP AQ+AS AS+SP AL
Abacassamo et al.
(2004) (Study I)
Mozambique Seasonal 64 61 60 — AQ+SP (3),
AS+AS (3)
AQ (25.9%)c,SP
(21.4%)c
Age 6—59 months,
temperature >37.5 ◦C, Pf
density 2000—100 000 l
Dorsey et al. (2002)
(Study II)
Uganda Mesoendemic 164b— 198b— 0 SP (32%)cAge 6—59 months,
temperature ≥38.0 ◦C, Pf
density >500 l
Mockenhaupt et al.
(2005) (Study III)
Ghana Hyperendemic 148 — 145 — AQ+SP (1),
AS+AS (4)
SP (23%) Age 6—59 months,
temperature >37.5 ◦C, Pf
density 2000—200 000 l
Mutabingwa et al.
(2005) (Study IV)
Tanzania Holoendemic 507 515 — 519 AQ+SP (41),
AQ+AS (38),
AL (36)
AQ (42%)c,AQ
(76%)
Age 4—59 months,
symptomatic, Pf density
>2000 l
Rwagacondo et al.
(2003) (Study V)
Rwanda Seasonal 132 — 144 — 0 AQ (23%) Age 6—59 months,
temperature >37.5 ◦C, Pf
density 1000—100 000 l
Staedke et al. (2004)
(Study VI)
Uganda Mesoendemic 139 139 — — AQ+SP (1),
AQ+AS (0)
SP (14%)cAge 6 months to 10 years,
temperature ≥38.0 ◦C, Pf
density 500—200 000 l
Yeka et al. (2005)
(Study VII)
Uganda Jinja (low),
Arua (high),
Tororo (high),
Apac (very
high)
770 767 — — AQ+SP (22),
AQ+AS (18)
NR Age 6 months to ≥5 years,
axillary temperature
>37.5 ◦C, Pf density
2000—200 000 l
AQ: amodiaquine; SP: sulfadoxine/pyrimethamine; AS: artesunate; AL: artemether/lumefantrine; Pf: Plasmodium falciparum; NR: not reported.
aThe dosages were: AS, 4 mg/kg once daily for 3 days; SP, 1.25 mg/kg of pyrimethamine and 25 mg/kg of sulfadoxine, given as a single dose; AQ, 30 mg/kg over 3 days in three trials
(Studies I, III, V) and 25 mg/kg over 3 days (i.e. 10 mg/kg daily for 2 days and 5mg/kg on the third day) in four trials (Studies II, IV, VI, VII); AL, six-dose regimen over 3 days, given as one
tablet for children weighing 10—15 kg, two tablets for those 15—25 kg, three tablets for those 25—35kg and four tablets for those >35 kg.
bDenotes the number of treatments given rather than patients.
cDenotes Day 14 failure rates.
Treatment regimens for uncomplicated falciparum malaria 121
Table 2 Study features analysed for methodological quality of randomised controlled trials
Study Generation
of allocation
sequence
Allocation
concealment
Blinding Follow-up Sample size
calculation
Intention-to-treat
analysis
Dorsey et al. (2002) (Study II) Adequate Adequate Single Adequate Yes No
Rwagacondo et al. (2003) (Study V) Not described Unclear Single Adequate No No
Staedke et al. (2004) (Study VI) Adequate Adequate Double Adequate Yes No
Abacassamo et al. (2004) (Study I) Adequate Adequate Open Adequate Yes No
Mutabingwa et al. (2005) (Study IV) Adequate Adequate Open Adequate Yes Yes
Mockenhaupt et al. (2005) (Study III) Adequate Adequate Double Adequate Yes No
Yeka et al. (2005) (Study VII) Adequate Adequate Single Adequate Yes No
for the sample size studied was provided in six of the trials
(Studies I, II, III, IV, VI, VII). Table 2 summarises the aspects
of methodological quality assessed.
In all the included studies, the primary treatment out-
come was early or late treatment failure or adequate clinical
and parasitological response (Figure 1). Six trials (Studies
II, III, IV, V, VI, VII) also presented the treatment failure
data as corrected by parasite genotyping to distinguish new
from recrudescent infections (Figure 2). All trials reported
data on haemoglobin levels during follow-up, but only four
trials presented the data as mean change in haemoglobin
(±SD) (Studies III, IV, VI, VII). All the trials except Study
V reported data on the prevalence of gametocyte carriage
at study enrolment and during follow-up (7—14 days after
treatment). Figure 3 shows the gametocyte carriage rates
during follow-up. Two studies (Studies I, II) did not assess
for adverse events (Table 3). The study from Tanzania (Study
IV) was an effectiveness trial, whilst the study from Mozam-
bique (Study I) assessed treatment failure at Day 21. The
study by Dorsey et al. (Study II) had a follow-up period of 1
year. One study (Study VII) from Uganda was conducted in
four geographical areas with different transmission inten-
sities. The authors presented their results according to the
different study areas. We have similarly presented the meta-
analysis results from this study as four different substudies
of the same trial.
3.2. Amodiaquine plus sulfadoxine/pyrimethamine
vs. artesunate plus sulfadoxine/pyrimethamine
Four trials compared AQ+SP with AS+SP in 1055 children
(Studies I, II, III, V). Overall, there was a significant reduc-
tion in the risk of treatment failure by Day 28 in chil-
dren treated with AQ+SP compared with those treated with
Figure 1 Parasitological failure rates in randomised trials comparing amodiaquine plus sulfadoxine/pyrimethamine (AQ+SP) ver-
sus artemisinin-based combination therapies (ACT). AQ+AS: amodiaquine plus artesunate; SP+AS: sulfadoxine/pyrimethamine plus
artesunate; RR: relative risk.
122 C.O. Obonyo et al.
Figure 2 Parasitological failure rates corrected by genotyping in the randomised controlled trials comparing amodiaquine plus
sulfadoxine/pyrimethamine (AQ+SP) versus artemisinin-based combination therapies (ACT). AQ+AS: amodiaquine plus artesunate;
SP+AS: sulfadoxine/pyrimethamine plus artesunate; RR: relative risk.
Figure 3 Proportion of children with gametocyte carriage in the randomised controlled trials comparing amodiaquine plus sulfa-
doxine/pyrimethamine (AQ+SP) versus artemisinin-based combination therapies (ACT). AQ+AS: amodiaquine plus artesunate; AS+SP:
artesunate plus sulfadoxine/pyrimethamine; RR: relative risk.
Treatment regimens for uncomplicated falciparum malaria 123
Table 3 Adverse events reported in the included studies
Study Adverse events Treatment arm
AQ+SP AQ+AS AS+SP AL
Rwagacondo et al. (2003) (Study V) Any 0 — 0 —
Serious 0 — 0 —
Mockenhaupt et al. (2005) (Study III) Any 4 — 3 —
Serious 2 — 2 —
Mutabingwa et al. (2005) (Study IV) Any 2 1 — 1
Serious 1 1 — 1
Staedke et al. (2004) (Study VI) Any 8 2 — —
Serious 6 1 — —
Yeka et al. (2005) (Study VII) Any 11 4 — —
Serious 11 4 — —
AQ: amodiaquine; SP: sulfadoxine/pyrimethamine; AS: artesunate; AL: artemether/lumefantrine.
AS+SP (RR = 0.56, 95% CI 0.42—0.75; I2= 0%). There was a
non-significant reduction in the recrudescence risk follow-
ing AQ+SP treatment compared with AS+SP (RR = 0.58, 95%
CI 0.32—1.04; I2= 46.3%) in the three trials (Studies II, III, V)
that performed genotyping.
Compared with AS+SP, treatment with AQ+SP was asso-
ciated with a significant increase in the risk of gametocyte
carriage (RR = 4.03, 95% CI 1.35—12.03; I2= 48%) in three tri-
als (Studies I, II, III). The mean haemoglobin change at Day
28, reported in only one trial (Study III), was not signifi-
cantly different between the two regimens (WMD = —0.01,
95% CI −0.35 to 0.33). There was no difference in the num-
bers (2/280 vs. 2/289) of children who developed serious
adverse events between those treated with AQ+SP compared
with AS+SP.
3.3. Amodiaquine plus sulfadoxine/pyrimethamine
vs. amodiaquine plus artesunate
Meta-analysis of four trials (Studies I, IV, VI, VII) involving
2962 children that compared AQ+SP with AQ+AS found no dif-
ference in treatment failure (RR = 1.12, 95% CI 0.81—1.54).
However, there was significant heterogeneity across the
studies (2= 55.79; P< 0.00001; I2= 89.2%).
One of the four studies was a large trial (Study VII) per-
formed in 1537 children between 6 months and ≥5 years
of age in four areas with different malaria transmission
intensities. Combining the four areas in this study, there
was a non-significant reduction in failure risk following
AQ+SP compared with AQ+AS (RR = 0.91, 95% CI 0.71—1.17;
I2= 77.5%), but with significant heterogeneity across the four
study sites. Similarly, when stratified by age, treatment
with AQ+SP was associated with a non-significant reduc-
tion in failure risk in children <5 years (RR = 0.90, 95% CI
0.74—1.10; I2= 66.5%) and those aged ≥5 years (RR = 0.70,
95% CI 0.31—1.60; I2= 36.8%) compared with AQ+AS. Strat-
ifying by transmission intensity, AQ+SP was associated with
a significant reduction in failure risk in the high transmis-
sion areas (RR = 0.75, 95% CI 0.65—0.86; I2= 0%) but a non-
significant increase in the two low—moderate transmission
sites (RR =1.15, 95% CI 0.85—1.57; I2=50.6%) compared with
AQ+AS.
In the meta-analysis of three studies (Studies IV, VI, VII)
that adjusted the failure risk by genotyping, AQ+SP was asso-
ciated with a higher risk of recrudescence compared with
AQ+AS (RR = 1.70, 95% CI 1.16—2.49; I2= 54.6%).
All the four trials reported on gametocyte carriage dur-
ing follow-up. Compared with AQ+AS, AQ+SP was associ-
ated with a significantly higher risk of gametocyte carriage
(RR = 1.33, 95% CI 1.13—1.57; I2= 33.4%). The mean change
in haemoglobin by Day 28 reported in two trials (Studies VI,
VII) was similar between the two regimens (WMD = 0.02, 95%
CI −0.31 to 0.35).
The incidence of serious adverse events reported in three
trials (Studies IV, VI, VII) was comparable (18/1416 vs.
6/1421; P= 0.23) between those treated with AQ+SP com-
pared with AQ+AS, respectively.
3.4. Amodiaquine plus sulfadoxine/pyrimethamine
vs. artemether/lumefantrine
Only one study of 1026 children compared AQ+SP with AL
(Study IV). Compared with AL, treatment with AQ+SP signif-
icantly increased the risk of treatment failure (RR = 2.80,
95% CI 2.32—3.39), recrudescence (RR = 11.26, 95% CI
4.54—27.90) and gametocyte carriage (RR = 3.74, 95% CI
2.31—6.03). AQ+SP was associated with a significantly lower
weighted mean change in haemoglobin on Day 14 compared
with AL (WMD = −0.07, 95% CI −0.09 to −0.05). Serious
adverse events were rare (1/507 vs. 1/519).
4. Discussion
In this systematic review and meta-analysis, we sum-
marise the available data from seven randomised trials
on the comparative efficacy and safety between AQ+SP
and artemisinin-based combinations for the treatment
of uncomplicated falciparum malaria in African children.
We found that treatment failure risks were significantly
124 C.O. Obonyo et al.
reduced following AQ+SP compared with AS+SP but were
increased compared with AL, and similar compared with
AQ+AS. However, the comparison with AQ+AS should be
interpreted with caution because there was significant
heterogeneity across the studies. Gametocyte carriage was
consistently and significantly reduced in all the treatment
arms with an artemisinin-based combination. The incidence
of serious adverse events was similar in all the treatment
arms, although the numbers were small.
With the emergence of resistance to commonly used
and affordable antimalarial drugs, control programmes must
decide when to change the first-line treatment for uncom-
plicated malaria and what to change it to. Combination
therapy offers great promise for extending the useful ther-
apeutic life and improving the efficacy of available drugs,
but the choices are limited. Consequences of not chang-
ing from the current monotherapies include increasing costs
of treating patients with treatment failures and increased
morbidity and mortality. In this review, we compared the
efficacy of the WHO-recommended ACTs with that of AQ+SP.
All the included trials were generally of high methodologi-
cal quality. They were done in children and hence should be
extrapolated to adults with caution. The treatment outcome
evaluated between 14 days and 28 days after therapy essen-
tially documents the risk of late treatment failure and is an
indirect measure of recurrent parasitaemia or risk of rescue
treatment. Genotyping assisted in differentiating whether
these recurrent infections were due to recrudescence or
new infections. Recrudescent infections are likely to stim-
ulate the production of gametocytes that accelerate the
transmission of drug resistance.
The assumption that ACTs are always more efficacious
than NACTs is not always true. In our review, AS+SP was
not as efficacious as AQ+SP, consistent with the findings of a
previous meta-analysis that compared 3 days of AS+SP with
SP monotherapy (Adjuik et al., 2004). We speculate that
AQ+SP performed better than AS+SP in the included trials
because of the post-treatment prophylactic effect of AQ and
because of the extensive use of SP in many settings following
the emergence of CQ resistance. Because of the rapid elim-
ination of artesunate (AS) from the bloodstream, a 3-day
course of AS+SP treatment will essentially be equivalent to
SP monotherapy. Currently, only a few settings remain where
SP monotherapy is still efficacious.
In a meta-analysis of four trials, AQ+SP was as effica-
cious as AQ+AS, but with significant heterogeneity across
the studies. AQ+SP was significantly more efficacious than
AQ+AS in high transmission areas. In areas of high malaria
transmission, it may be beneficial to use an antimalarial
drug with post-treatment prophylactic properties, but this
benefit must be weighed against the risk of selecting for
resistance, which is common with drugs with a long ter-
minal elimination half-life (Hastings et al., 2002). The risk
of recrudescence was significantly higher with AQ+SP, but
the results were still heterogeneous. It may be argued that
evidence from large trials is preferable to meta-analysis
of small trials. Our review included a large trial (Study
VII) conducted in sites with different malaria transmis-
sion intensities. The large trial was associated with signif-
icant heterogeneity and, surprisingly, showed a treatment
effect similar to that found by a meta-analysis of three
smaller trials (Studies I, IV, VI). The transmission inten-
sity explained a large proportion of the heterogeneity in
the treatment effect in the large trial, demonstrating the
difficulties commonly encountered in interpreting multicen-
tre studies conducted in sites with different transmission
intensities. Drug-resistant P. falciparum is thought to spread
faster in high transmission areas, but this is still controver-
sial and should be explored in future studies (Hastings and
D’Alessandro, 2000; Hastings and Watkins, 2005; Talisuna et
al., 2002).
Selecting an effective first-line therapy is only the first
step in the decision-making process for the implementa-
tion of a new treatment policy. Other activities include
considerations of financing, diagnostics (clinical vs. labora-
tory confirmation), development of guidelines, training of
health workers, advocacy, drug procurement, drug distribu-
tion, quality assurance, regulation and pharmacovigilance
(Rational Pharmaceutical Management Plus Program, 2005;
Williams et al., 2004). An ideal antimalarial drug combina-
tion should be cheap, administered over 3 days (at most),
safe and have an effect on all stages of the parasite, includ-
ing gametocytes (Kremsner and Krishna, 2004). Our meta-
analysis found evidence for the preference of AQ+SP over
AS+SP, and AL over AQ+SP. Compared with AQ+SP, all the
ACTs evaluated significantly reduced gametocyte carriage,
a beneficial property of artemisinin derivatives that may
be useful for transmission reduction (Price et al., 1996;
Sutherland et al., 2005). Our results indicate that for coun-
tries changing from SP, alternatives include AQ+AS or AL.
None of these combinations is perfect. AQ+SP is currently
not co-formulated, but the once-daily dosing may enhance
adherence, although increasing resistance to AQ monother-
apy may shorten the useful therapeutic life of the combina-
tion. AL is currently the only co-formulated ACT, but the high
cost and twice-daily dosing may cause problems with adher-
ence. Of the alternatives, AQ+AS may be a rational choice
for first-line therapy, reserving the more costly AL for those
who fail therapy with AQ+AS (second-line) and quinine for
severe malaria. The efficacies of AQ and SP are still high in
many areas of the West African region, and in those settings
AQ+SP or AS+SP should be considered as first-line therapy
and AQ+AS as second-line. Compared with ACTs, AQ+SP was
more effective at reducing the incidence of new infections
(data not shown), implying that in areas with high malaria
transmission an ACT policy should be accompanied by an
integrated programme that includes vector control interven-
tions. The implementation of a malaria control programme
using effective vector control (DDT spraying) and treatment
with AL has resulted in a marked reduction in malaria mor-
bidity and mortality in KwaZulu-Natal in South Africa (Barnes
et al., 2005).
Combination therapy will inevitably be more costly com-
pared with either CQ or SP monotherapy. Consequently, cost
is the main reason for selecting to implement an interim
option or delaying the switch to ACT in most malaria-
endemic settings. Currently, ACTs cost up to 10 times the
price of CQ or SP (Arrow et al., 2004). A recent cost-
effectiveness analysis has compared the effect of moving
from the current CQ or SP monotherapy to ACT and moving
to ACT through an affordable although partially effective
(e.g. AQ+SP) interim antimalarial regimen (Laxminarayan,
2004). An interim regimen was only cost-effective in the
short-term (<5 years). The main caveat would be the need
Treatment regimens for uncomplicated falciparum malaria 125
to change policy again within a short time, which is a time-
consuming and expensive process. However, considering
among other factors the logistical implications of changing
a treatment policy as well as the malaria-specific morbidity
and mortality averted, it was more cost-effective to move
straight to an ACT regimen (Laxminarayan, 2004). The
Global Fund for HIV/AIDS, Tuberculosis and Malaria (GFATM)
is one initiative that provides support to malaria-endemic
countries for expanding access to ACT and accelerating the
change in policy to ACTs. Some countries may be hesitant to
accept the GFATM support owing to lack of structures to sus-
tain the purchase and distribution of ACTs once this support
ends. A possible solution to the high cost of ACTs is a global
subsidy at the high levels of drug procurement (Arrow et al.,
2004).
Our review has several limitations. We found only a small
number of studies, with a preponderance of trials from the
Eastern Africa region where SP resistance is increasingly
widespread; there was only one study (Study III) from West
Africa, conducted in an area with relatively high SP sensi-
tivity and therefore our conclusions may not apply to other
settings with optimal therapeutic efficacy of AQ or SP. Sim-
ilarly, there was only one study that compared AQ+SP with
AL (Study IV). Reporting of adverse events, haemoglobin or
background failure rates for SP and AQ was not consistently
done in all the trials that we found. Only two of the stud-
ies (Studies III, VI) had performed a double-blind design. For
the primary outcome, we used an intention-to-treat analy-
sis, which may have underestimated the treatment effect.
Several questions remain unanswered. The safety of ACTs
is not established for infants, pregnant women and patients
infected with HIV. Whether the reduction in malaria trans-
mission observed in Southeast Asia following widespread
deployment of AS plus mefloquine would be experienced
with ACTs in areas of high malaria inoculation is still
unknown. Similarly, there are no data that ACTs will delay
the emergence of drug resistance in sub-Saharan Africa.
Efficacy data alone are not sufficient for changing treat-
ment policy; only limited data are available on the effec-
tiveness of ACTs. The role of ACTs in home management
of malaria (fever) or in intermittent presumptive therapy
is still unclear. The efficacy of AL should be studied in
other settings: there are unpublished data showing that it
may perform differently in settings with higher AQ+SP effi-
cacy. Similarly, there is an urgent need to evaluate other
co-formulated ACTs, e.g. dihydroartemisinin/piperaquine,
which is in the pipeline. There is also evidence of declining
efficacy of AQ+SP in some settings. Overall, for malaria con-
trol in sub-Saharan Africa, effective tools (e.g. insecticide-
treated bed nets, intermittent preventive treatment and
combination therapies) have become available, and there
is a global interest to expand coverage and to maximise
impact.
In conclusion, our results indicate that AQ+SP is infe-
rior to AL, superior to AS+SP, but comparable with AQ+AS.
There is an urgent need to evaluate the utility of combining
ACTs with vector control interventions. At best, the AQ+SP
combination should be considered as an interim regimen,
and it may play an important role in settings where both
components have maintained high efficacy, where malaria
diagnostics are unavailable or when there are shortages of
ACTs.
Conflicts of interest statement
The authors have no conflicts of interest concerning the work
reported in this paper.
Acknowledgements
This work was supported by Pfizer Global Pharmaceuti-
cals. We acknowledge the comments of Drs Anne Gasasira,
Ambrose Talisuna and Francois Nosten on earlier drafts
of this manuscript. We thank the Director, Kenya Medical
Research Institute (KEMRI), for permission to publish the
results of this study.
References
Abacassamo, F., Enosse, S., Aponte, J.J., Gomez-Olive, F.X., Quinto,
L., Mabunda, S., Barreto, A., Magnussen, P., Ronn, A.M., Thomp-
son, R., Alonso, P.L., 2004. Efficacy of chloroquine, amodi-
aquine, sulphadoxine—pyrimethamine and combination therapy
with artesunate in Mozambican children with non-complicated
malaria. Trop. Med. Int. Health 9, 200—208.
Adjuik, M., Agnamey, P., Babiker, A., Borrmann, S., Brasseur, P.,
Cisse, M., Cobelens, F., Diallo, S., Faucher, J.F., Garner, P.,
Gikunda, S., Kremsner, P.G., Krishna, S., Lell, B., Loolpapit,
M., Matsiegui, P.B., Missinou, M.A., Mwanza, J., Ntoumi, F.,
Olliaro, P., Osimbo, P., Rezbach, P., Some, E., Taylor, W.R.,
2002. Amodiaquine—artesunate versus amodiaquine for uncom-
plicated Plasmodium falciparum malaria in African children: a
randomised, multicentre trial. Lancet 359, 1365—1372.
Adjuik, M., Babiker, A., Garner, P., Olliaro, P., Taylor, W., White, N.;
International Artemisinin Study Group, 2004. Artesunate com-
binations for treatment of malaria: meta-analysis. Lancet 363,
9—17.
Alderson, P., Green, S., Higgins, J. (Eds), 2004. Highly sensitive
search strategies for identifying reports of randomised con-
trolled trials in MEDLINE. Cochrane Reviewers’ Handbook 4.2.2.
Appendix 5b. In: The Cochrane Library, Issue 2, The Cochrane
Collaboration. John Wiley and Sons, Ltd, Chichester, UK.
Arrow, K.J., Panosian, C.B., Gelband, H. (Eds), 2004. Saving lives,
buying time: economics of malaria drugs in an age of resistance.
National Academies Press, Washington, DC.
Barnes, K.I., Durrheim, D.N., Little, F., Jackson, A., Mehta,
U., Allen, E., Dlamini, S.S., Tsoka, J., Bredenkamp, B.,
Mthembu, D.J., White, N.J., Sharp, B.L., 2005. Effect of
artemether—lumefantrine policy and improved vector control
on malaria burden in KwaZulu-Natal, South Africa. PLos Med.
2, e330.
Bloland, P.B., 2003. A contrarian view of malaria therapy policy in
Africa. Am. J. Trop. Med. Hyg. 68, 125—126.
DerSimonian, R., Laird, N., 1986. Meta-analysis in clinical trials.
Control. Clin. Trials 7, 177—188.
Dorsey, G., Njama, D., Kamya, M.R., Cattamanchi, A., Kyabayinze,
D., Staedke, S.G., Gasasira, A., Rosenthal, P.J., 2002. Sulfadox-
ine/pyrimethamine alone or with amodiaquine or artesunate for
treatment of uncomplicated malaria: a longitudinal randomised
trial. Lancet 360, 2031—2038.
Gasasira, A.F., Dorsey, G., Nzarubara, B., Staedke, S.G., Nassali,
A., Rosenthal, P.J., Kamya, M.R., 2003. Comparative efficacy of
aminoquinoline—antifolate combinations for the treatment of
uncomplicated falciparum malaria in Kampala, Uganda. Am. J.
Trop. Med. Hyg. 68, 127—132.
Hastings, I.M., D’Alessandro, U., 2000. Modelling a predictable dis-
aster: the rise and spread of drug-resistant malaria. Parasitol.
Today 16, 340—347.
126 C.O. Obonyo et al.
Hastings, I.M., Watkins, W.M., 2005. Intensity of malaria trans-
mission and the evolution of drug resistance. Acta Trop. 94,
218—229.
Hastings, I.M., Watkins, W.M., White, N.J., 2002. The evolution of
drug-resistant malaria: the role of drug elimination half-life. Phi-
los. Trans. R. Soc. Lond. B Biol. Sci. 357, 505—519.
Higgins, J.P., Thompson, S.G., 2002. Quantifying heterogeneity in a
meta-analysis. Stat. Med. 21, 1539—1558.
Juni, B., Altman, D.G., Egger, M., 2001. Systematic reviews of
healthcare: assessing the quality of controlled clinical trials. BMJ
323, 42—46.
Kremsner, P.G., Krishna, S., 2004. Antimalarial combinations.
Lancet 364, 285—294.
Laxminarayan, R., 2004. ACT now or later? Economics of malaria
resistance. Am. J. Trop. Med. Hyg. 71 (Suppl. 2), 187—195.
Lell, B., Kremsner, P.G., 2002. Clindamycin as an antimalarial drug:
review of clinical trials. Antimicrob. Agents Chemother. 46,
2315—2320.
Mockenhaupt, F.P., Ehrhardt, S., Dzisi, S.Y., Bousema, J.T., Was-
silew, N., Schreiber, J., Anemana, S.D., Cramer, J.P., Otch-
wemah, R.N., Sauerwein, R.W., Eggelte, T.A., Bienzle, U.,
2005. A randomised, placebo-controlled, double-blind trial on
sulfadoxine—pyrimethamine alone or combined with artesunate
or amodiaquine in uncomplicated malaria. Trop. Med. Int. Health
10, 512—520.
Mutabingwa, T.K., Anthony, D., Heller, A., Hallet, R., Drakeley,
C., Greenwood, B.M., Whitty, C.J.M., 2005. Amodiaquine
alone, amodiaquine+sulfadoxine—pyrimethamine, amodi-
aquine+artesunate, and artemether—lumefantrine for
outpatient treatment of malaria in Tanzanian children:
a four-arm randomised effectiveness trial. Lancet 365,
1474—1480.
Obonyo, C.O., Ochieng, F., Taylor, W.R., Ochola, S.A., Mugitu,
K., Olliaro, P., ter Kuile, F., Oloo, A.J., 2003. Artesunate
plus sulfadoxine—pyrimethamine for uncomplicated malaria
in Kenyan children: a randomised, double-blind, placebo-
controlled trial. Trans. R. Soc. Trop. Med. Hyg. 97, 585—591.
Price, R., Nosten, F., Luxemburger, C., ter Kuile, F.O., Phaipun,
L., Chongsuphajaisiddhi, T., White, N.J., 1996. Effects of
artemisinin derivatives on malaria transmissibility. Lancet 347,
1654—1658.
Rational Pharmaceutical Management Plus Program, 2005. Changing
malaria treatment policy to artemisinin-based combinations: an
implementation guide. Submitted to the U.S. Agency for Inter-
national Development by the RPM Plus Program. Management
Sciences for Health, Arlington, VA.
Rwagacondo, C.E., Niyitegeka, F., Sarushi, J., Karema, C.,
Mugisha, V., Dujardin, J.-C., Van Overmeir, C., Van Den
Ende, J., D’Alessandro, U., 2003. Efficacy of amodiaquine
alone and combined with sulfadoxine—pyrimethamine and of
sulfadoxine—pyrimethamine combined with artesunate. Am. J.
Trop. Med. Hyg. 68, 743—747.
Schellenberg, D., Kahigwa, E., Drakeley, C., Malende, A., Wigayi,
J., Msokame, C., Aponte, J.J., Tanner, M., Mshinda, H.,
Menendez, C., Alonso, P.L., 2002. The safety and efficacy of
sulfadoxine—pyrimethamine, amodiaquine, and their combina-
tion in the treatment of uncomplicated Plasmodium falciparum
malaria. Am. J. Trop. Med. Hyg. 67, 17—23.
Shretta, R., Omumbo, J., Rapuoda, B., Snow, R.W., 2001. Using
evidence to change antimalarial drug policy in Kenya. Trop. Med.
Int. Health 5, 755—764.
Sibley, C.H., Hyde, J.E., Sims, P.F., Plowe, C.V., Kublin, J.G., Mberu,
E.K., Cowman, A.F., Winstanley, P.A., Watkins, W.M., Nzila, A.M.,
2001. Pyrimethamine—sulfadoxine resistance in Plasmodium fal-
ciparum: what next? Trends Parasitol. 17, 582—588.
Snow, R.W., Guerra, C.A., Noor, A.M., Myint, H.Y., Hay, S.I., 2005.
The global distribution of clinical episodes of Plasmodium falci-
parum malaria. Nature 434, 214—217.
Staedke, S.G., Mpimbaza, A., Kamya, M.R., Nzarubara, B.K., Dorsey,
G., Rosenthal, P.J., 2004. Combination treatments for uncom-
plicated falciparum malaria in Kampala, Uganda: randomised
clinical trial. Lancet 364, 1950—1957.
Sutherland, C.J., Ord, R., Dunyo, S., Jawara, M., Drakeley, C.J.,
Alexander, N., Coleman, R., Pinder, M., Walraven, G., Targett,
G.A., 2005. Reduction of malaria transmission to Anopheles
mosquitoes with a six-dose regimen of co-artemether. PLos Med.
2, e92.
Talisuna, A.O., Langi, P., Bakyaita, N., Egwang, T., Mutabingwa,
T.K., Watkins, W., Van Marck, E., D’Alessandro, U., 2002. Inten-
sity of malaria transmission, antimalarial drug use and resistance
in Uganda: what is the relationship between these three factors?
Trans. R. Soc. Trop. Med. Hyg. 96, 310—317.
Trape, J.F., Pison, G., Preziosi, M.P., Enel, C., Desgrees du Lou, A.,
Delaunay, V., Samb, B., Lagarde, E., Molez, J.F., Simondon, F.,
1998. Impact of chloroquine resistance on mortality. C. R. Acad.
Sci. III 321, 689—697.
White, N.J., 1998. Preventing antimalarial drug resistance through
combinations. Drug Resist. Updat. 1, 3—9.
White, N.J., Olliaro, P.L., 1996. Strategies for the prevention of
antimalarial drug resistance: rationale for combination therapy
for malaria. Parasitol. Today 12, 399—401.
WHO, 1993. Implementation of the Global Malaria Control Strat-
egy. Report of a WHO Study Group on the implementation of
the global plan of action for malaria control, 1993—2000. World
Health Organization, Geneva, Technical Report Series No. 839.
WHO, 1998. The use of artemisinin and its derivatives as antimalar-
ial drugs: a report of a joint CTD/DMP/TDR Informal Consulta-
tion. World Health Organization, Geneva, WHO/MAL/98.1086.
WHO, 2002. Monitoring antimalarial drug resistance. A report of a
WHO consultation, Geneva, Switzerland, 3—5 December 2001.
World Health Organization, Geneva, WHO/CDS/RBM/2002.39.
Williams, H.N., Durrheim, D., Shretta, R., 2004. The process
of changing national malaria treatment policy: lessons from
country-level studies. Health Policy Plan. 19, 356—370.
Yeka, A., Banek, K., Bakyaita, N., Staedke, S.G., Kamya, M.R.,
Talisuna, A., Kironde, F., Nsobya, S.L., Kilian, A., Slater, M.,
Reingold, A., Rosenthal, P.J., Wabwire-Mangen, F., Dorsey, G.,
2005. Artemisinin versus nonartemisinin combination therapy for
uncomplicated malaria: randomised clinical trials from four sites
in Uganda. PLos Med. 2, e190.
