See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/6508378
Mefloquine treatment for uncomplicated falciparum malaria in young children
6-24 months of age in northern Ghana
Article  in  The American journal of tropical medicine and hygiene · February 2007
DOI: 10.4269/ajtmh.2007.76.224 · Source: PubMed
CITATIONS READS
8 24
8 authors, including:
David J Fryauff Seth Owusu-Agyei
Naval Health Research Center University of Health and Allied Sciences
81 PUBLICATIONS   2,969 CITATIONS    342 PUBLICATIONS   6,744 CITATIONS   
SEE PROFILE SEE PROFILE
J. Kevin Baird Kwadwo A Koram
University of Oxford University of Ghana
311 PUBLICATIONS   10,641 CITATIONS    169 PUBLICATIONS   4,974 CITATIONS   
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Mental Health View project
Malaria View project
All content following this page was uploaded by Seth Owusu-Agyei on 26 September 2015.
The user has requested enhancement of the downloaded file.
Am. J. Trop. Med. Hyg., 76(2), 2007, pp. 224–231
Copyright © 2007 by The American Society of Tropical Medicine and Hygiene
MEFLOQUINE TREATMENT FOR UNCOMPLICATED FALCIPARUM MALARIA IN
YOUNG CHILDREN 6–24 MONTHS OF AGE IN NORTHERN GHANA
DAVID J. FRYAUFF,* SETH OWUSU-AGYEI, GREGORY UTZ, J. KEVIN BAIRD, KWADWO A. KORAM,
FRED BINKA, FRANCIS NKRUMAH, AND STEPHEN L. HOFFMAN
US Naval Medical Research Unit No. 3, Cairo, Egypt; Navrongo Health Research Center, Navrongo, Upper East Region, Ghana;
Naval Medical Research Center, Silver Spring, Maryland; Noguchi Memorial Institute for Medical Research, University of Ghana,
Legon, Ghana; Department of Public Health and Epidemiology, School of Public Health, University of Ghana, Legon, Ghana
Abstract. Mefloquine (MQ) single dose 20 mg/kg treatment of falciparum malaria was evaluated in 186 children of
6–24 months of age in northern Ghana. There were 15 RII/RIII-type parasitologic failures, all with Day 2 MQ blood
levels significantly lower than children whose parasitemias cleared before Day 7 and remained clear through 28 days.
Predictors of RII/RIII parasitologic response were vomiting after MQ dosing, Day 2 MQ levels < 500 ng/mL, and
undetectable Day 2 levels of the carboxymefloquine metabolite. There were 50 cases of delayed RI parasitologic failure,
but 71% of these cases had undetectable Day 28 blood levels of MQ and drug levels in the remaining 29% ranged below
the 620 ng/mL level that suppresses MQ sensitive strains of P. falciparum. Drug levels among infants that tolerated MQ
well were not associated with age, weight, hemoglobin, parasitemia, and pre-existing symptoms of vomiting or diarrhea.
An observed recurrent parasitemia of 34,400 trophozoites/
L against a MQ blood concentration of 550 ng/mL was taken
as indication of tolerance to suppressive levels of the drug at this location.
INTRODUCTION sis of their MQ blood levels would be predictive of the MQ
prophylaxis failures that occurred among clinically-immune
Under the conditions of intense malaria transmission com- adults in this same community 2 years later. In other words,
monly seen throughout rural sub-Saharan Africa, Plasmo- we sought details of recurrent parasitemias that occurred in
dium falciparum rapidly developed resistance to easily afford-
1,2 MQ-treated children, confirmation of dosing, and a determi-able drug treatments based on chloroquine or Fansidar.
nation of MQ levels in blood at the time of any early or late
Consequently, mefloquine (MQ) became a sensible and
popular first-line drug for the prevention of malaria in trav- treatment failures.
elers, and where cost was not a primary limiting factor, this Because of cost and availability, MQ has not been widely
drug was also considered a safe and effective single-dose used for treatment in Africa, and taken collectively, trials of
treatment of uncomplicated P. falciparum infection. How- MQ treatment in African children since the late 1980s have
ever, despite limited and carefully supervised use of MQ in not produced clear indications of either resistant infections or
combination with sulphadoxine and pyrimethamine (MSP), successful treatments. A number of studies have reported that
treatment failure rates in Thailand rose 20-fold in just 4 years young children tolerate single oral doses of the drug poorly,
and forced an abandonment of the MSP treatment strategy.3 leading some authorities to advise splitting the dose and to10,11
In vitro tests from locations throughout Africa gave early guard against vomiting. Several studies have indicated
indication that even without prior exposure or drug pressure, that diarrhea, either a preexisting condition or an adverse side
strains of malaria circulated that were inherently resistant to effect of the MQ therapy, is also associated with inadequate12,13
MQ.4,5 The drug was 94% more effective than chloroquine MQ drug levels. In a study of infant children in French
for preventing malaria among Peace Corps volunteers in Guyana pre-treated with metoclopramide before gastric tube
West Africa during 1992, but cases of MQ prophylaxis failure, administration of a single 25 mg/kg dose of MQ, early vom-
confirmed by high drug levels in blood (744–1,275 ng/mL), iting occurred in 11.5%, and diarrhea was found to be the14
have now been reported in African states ranging geographi- most common adverse event of therapy. Host genetic fac-
cally from Tanzania to Sierra Leone.6,7 While the majority of tors, disease condition, and age have also been variably con-
these have been isolated cases, spanning a dozen years, an sidered responsible for significant pharmacokinetic differ-
unexpectedly high MQ prophylaxis failure rate of 14% was ences of MQ in whole blood and sera. In one comparative
reported during 1998 at a single location in northern Ghana.8 study, healthy adult Swiss men achieved peak plasma MQ
This high prophylaxis failure rate was even more unusual levels 3-fold lower than parasitemic adult male Thais given15
because it occurred among “malaria-immune” adult Ghana- the same dose of MQ. Thai children 5–10 years of age
ians who had been randomized to receive standard weekly achieved a mean Day 2 blood level of 2,031 ng/mL that was
dosing of MQ in a rigidly supervised double-blind, placebo- virtually the same as the mean serum level of 2,165 ng/mL3,16
controlled prophylaxis trial. This trial had been preceded in measured in Thai adults, but pharmacokinetic parameters
1996 by studies of seasonal malaria incidence in young chil- of Cmax and area under the curve (AUC) for MQ in Thai
dren and against the likelihood of chloroquine and Fansidar infants with malaria were reported to be much higher than in
failures, MQ was the drug of choice selected for single dose adult Thai patients.
17 African children with malaria seem to
treatment of their uncomplicated malaria infections.9 We hy- attain lower levels of MQ in blood, making treatments and in
pothesized that a retrospective evaluation of the MQ treat- vivo determinations of resistance problematic. Cohorts of
ment outcomes in these non-immune children, and an analy- Malawian children younger than 5 years of age given single
15- or 25-mg/kg doses of MQ reportedly attained comparable
mean Day 2 blood levels of 1,043 and 1,420 ng/mL, respec-
* Address correspondence to David J. Fryauff, US Naval Medical tively, but mean levels in these young children by Day 7 fell
Research Unit No. 3, PSC 452 Box 52, FPO AE 09835-0007. E-mail: to 670 and 718 ng/mL, respectively, with individual measures
fryauffd@nmrc.navy.mil ranging from 26 to 1,716 ng/mL.10 Senegalese children > 1
224
MEFLOQUINE TREATMENT IN GHANA 225
year old treated with a single oral dose of 12.5 mg/kg achieved mL of suspension/kg body weight (20 mg base/kg) as a one-
average Day 2 MQ and CMQ blood levels of 955 and 407 time oral dose administered by calibrated plastic syringe.
ng/mL, respectively.18 A more recent evaluation of the MQ Each child was monitored thereafter for 1 hour, and dosing
15-mg/kg single dose in young Malawian children reported a was repeated if vomiting occurred. The 20-mg/kg body weight
mean Day 2 serum level of only 633 ng/mL.19 Given such dose selected was a compromise between intention to reduce
variables, uncertainties, and contradictions, we considered the likelihood of vomiting the dose in the young subjects and
that our provision of malaria therapy to Ghanaian infants intention to affect a full rapid cure. Children unable to toler-
afforded an opportunity to broaden the knowledge base of ate MQ were treated with Fansidar or quinine.
MQ treatment in the age and ethnic group at greatest risk of Follow-up. Field workers conducted visits three times
infection, morbidity, and death by malaria. weekly to monitor each child. Malaria blood smears were
made on days 0, 2, 7, and 28 or at any occasion of illness
MATERIALS AND METHODS consistent with malaria. Children who were symptomatic and
parasitemic were brought to the hospital for evaluation by a
Study site. The Kassena Nankana District (KND), situated physician. Children with unremitting or recurrent parasit-
in the Upper East Region of Ghana at its northern border emias during the MQ in vivo test were provided Fansidar or
with Burkina Faso, is an administrative territory of 1,675 km2 quinine treatment. Heparinized capillary tubes were used to
with a population of ∼140,000 residents. The flat, open wood- collect 100 
L of whole blood on days 2 and 28 or any day of
land ecology is termed Guinea Savannah, with annual rainfall parasitemia that prompted alternative therapy. Capillary
of ∼900 mm occurring from May to October. The ethnic Kas- blood was blotted onto a Whatman no. 2 filter paper that was
sem and Nankan tribes are predominantly subsistence farm- air-dried and thereafter kept refrigerated in separate plastic
ers cultivating seasonal millet; however, a dam constructed in ziplock bags. In the absence of evidence suggesting poor com-
1980 provides irrigation for year-round rice and vegetable pliance, emesis, or diarrhea at the time of dosing, parasitemias
cultivation for about one fifth of the district population. Ma- that persisted or recurred within the 7 test days after MQ
laria is holoendemic, transmitted by Anopheles gambiae sensu treatment were classified according to World Health Organi-
strictu, during the wet season and by An. funestus at a lower, zation (WHO) criteria as RII- or RIII-type resistance. Clear-
but continuous, rate across both seasons.20 The incidence of ance followed by recurrence of parasitemia between days 7
malaria infection among young children was determined to be and 28 of the test represented either RI-type resistance or
9.1 infections per person-year during the wet season and 4.7 re-infection.
infections per person-year during the dry season.9 The MQ Analysis of MQ levels in blood. Coded filter paper blood
treatment component of the dry season incidence study con- blots were sent to the Department of Clinical Chemistry,
stitutes the subject of this report. Falun Central Hospital, Falun, Sweden, where high-
Study subjects. Informed consenting parents enrolled their performance liquid chromatography (HPLC) was used to de-
children during November 1996 into a prospective study mea- termine concentrations of MQ and its main metabolite, car-
suring dry season malaria attack rates. Institutional boards of boxymefloquine (CMQ).21 Sensitivity of this assay was 95
the US Navy and the Ghanaian Ministry of Health conducted ng/mL (0.25 
mol/L) for MQ and 75 ng/mL (0.20 
mol/L) for
ethical review and granted approval for the conduct of this CMQ. It was reported that 5–10% degradation of MQ could
study. Approximately 28% of children screened were ex- be expected from our handling of the samples before extrac-
cluded, mainly because of severe anemia (hemoglobin [Hb] < tion and HPLC. Based on confirmed parasitemia, patient’s
6.0 g/dL), and were referred for treatment. The enrolled dry illness and dosing history, concentrations of MQ in blood at
season cohort consisted of 259 breast-feeding children 6–24 the time of parasitemia, and the threshold of 620 ng/mL (1.67
months of age, all of whom received curative therapy consist- 
mol/L) blood concentration of MQ considered to be an ef-
ing of quinine sulfate (10 mg/kg orally, three times a day, days fective barrier against the appearance of sensitive strains of P.
1–4), Fansidar (5–10 kg, 1/2 tablet; 11–20 kg, 1 tablet; Day 5), falciparum in the bloodstream,7,22 a determination of sensi-
and primaquine (0.25 mg/kg orally, four times a day, days tivity, resistance, or poor drug absorption was made.
5–18), except in the case of 17 glucose-6-phosphate dehydro- Statistical analysis. Descriptive statistics were performed
genase–deficient children. All medications were given as sus- for baseline characteristics of the enrolled children and their
pensions in a fruit-flavored sugar syrup. Children were visited parasitemias. Student t test was used to compare levels of MQ
regularly, with blood films made every 2 weeks or at any and CMQ in cured cases and those considered early or late
occasion of illness consistent with malaria (fever, chills, nau- treatment failures. 2 and Fisher exact tests were used to
sea, vomiting, and malaise). The presence of asexual stages of compare proportions. Simple linear regression analysis was
P. falciparum in a Giemsa-stained blood film prompted im- used to explore the relationship between drug levels and vom-
mediate clinical evaluation and treatment. Uncomplicated iting or diarrhea. For graphing and statistical purposes, MQ
clinical malaria and parasitemias < 1% were treated with MQ. and CMQ values below the limits of detection were arbitrarily
Children averaged 6.9 (95% CI, 6.5, 7.3) parasite-free weeks assigned a value midway between that threshold value and
from the end of radical cure to re-infection and the initiation zero.
of MQ treatment. Loss of premunition by the radical curative
therapy was thought responsible for significantly greater par-
asitemias in children at the time of reinfection.9 RESULTS
Chemotherapy. Mefloquine hydrochloride (Lariam; Hoff-
man-LaRoche, Basel, Switzerland) uncoated tablets contain- A total of 193 children with uncomplicated, slide-confirmed
ing 250 mg base were used to prepare a sweet syrup suspen- falciparum malaria were evaluated and enrolled into the MQ
sion containing 40 mg MQ base/mL. Children were given 0.5 in vivo test. No symptoms were reported or observed for 29
226 FRYAUFF AND OTHERS
infants with parasitemias ranging from 40 to 26,000/
L.
Among 164 infants with illness, fever was their dominant
symptom, reported by 81% of mothers. Additional symptoms
or conditions reported were chills/rigor (46%), diarrhea
(39%), vomiting (38%), apathy/listless (26%), respiratory ill-
ness (6.7%), and altered mental state (2%). Two cases were
excluded because of concomitant anti-malarial use, and five
individuals were lost to follow-up. Enrollment characteristics
of 186 children (Table 1) that were followed to an endpoint
show that anemia (Hb < 8.0 g/dL), measured fever (> 37.5°C),
and high-density parasitemia (> 20,000 asexual forms/
L)
were conditions present in about one third of the children.
Therapy was supervised and directly observed in 98.7% of the
cases. Vomiting within the post-dosing observation period
and re-dosing occurred in 18 (9.3%) children.
Parasitemias cleared by Day 2 in 51% (93/181) and by Day
7 in 92% (170/185) of treated cases. Among 15 cases classified
as RII (9) and RIII (6) failures (Table 2) with persistent (13)
or recurrent, (2) parasitemias observed on or before Day 7,
records revealed histories of vomiting (3), diarrhea (2), or
both (4), emesis of the dose (2), and unmonitored dosing (1).
Geometric mean (GM) parasitemia for this group was 3,083/

L at the start of therapy and 349/
L on Day 7. Figure 1A
shows that blood concentrations of MQ in these 15 cases
ranged on Day 2 from below detectable level to 350 ng/mL,
with a mean significantly below that of a random selection of
“sensitive” cases (Table 3) that cleared and remained clear
through the 28 day test period (RII/RIII: 172 ng/mL versus S:
561 ng/mL; P < 0.0001). Day 2 MQ and/or CMQ levels were
below the limits of detection in eight cases. There was no
apparent cause for early treatment failure among three chil-
dren with observed compliance and no record of vomiting or
diarrhea. Day 2 levels of MQ (210, 255, and 305 ng/mL) in
these exceptions, while being among the highest in this group,
were far below the lower 95% confidence limit of the drug in
whole blood (467 ng/mL) from children who were cured by
the treatment.
Late recurrent parasitemias ranging from 40 to 64,000/
L
(GM, 10,560/
L) developed in 50 children (50/186 
 27%).
Twelve of these cases were symptomatic and appeared be-
tween days 15 and 26. An additional 38 recurrent, but largely
asymptomatic, RI-type parasitemias were detected in the
scheduled Day 28 screen. MQ blood level analysis selected
TABLE 1
Characteristics of the study population at enrollment
Characteristic Values
No. enrolled 186
Males:females 90:96
Mean age (mo) 15.51 (95% CI: 14.7–16.3)
Age range 6–25
Mean body weight (kg) 8.29 (95% CI: 8.1–8.5)
Mean hemoglobin (g/dL) 8.6 (95% CI: 8.3–8.8)
No. anemic (Hb < 8.0) 69 (37%)
Geometric mean (GM)
parasite density/
L 3,363 (95% CI: 2,430–4,653)
No. febrile (%) 69 (37%)
GM parasite density
of febrile cases 8,670 (95% CI: 5,433–13,836)
GM parasite density
of non-febrile cases 2,138 (1,268–3,606)
No. parasitemias > 20,000/
L 52 (28%)
TABLE 2
Patient descriptions, symptoms at enrollment, follow-up parasitemia, parasitologic classification, and whole blood levels of MQ and CMQ metabolite in MQ-treated cases of falciparum malaria
with RII/RIII parasitologic outcomes
Patient characteristics at enrollment Symptoms before treatment Parasite densities/
L Drug and metabolite levels (ng/mL)
ID# Sex/age Hb Weight Temp. Fever Vomiting Diarrhea Score* D0 D2–6 D7 Class D2-MQ D2-CMQ D28-MQ D28-CMQ Comments
ETF-1 M/19 9.1 6 36.1 N N Y 1 120 Negative 160 RII 47.5 125 47.5 37.5 Suspect incomplete dosing
ETF-2 M/7 14.9 6 36.1 N Y Y 3 1,200 1,640 320 RIII 47.5 37.5 47.5 37.5 Vomiting reported
ETF-3 F/22 10.5 12 37.1 Y N Y 3 30,960 22,120 120 RII 47.5 37.5 47.5 37.5 Vomiting reported
ETF-4 M/19 9.1 7 36.6 N N N 0 480 200 840 RII 95 37.5 47.5 37.5 Dose vomited
ETF-5 F/19 8.9 9 38.5 Y Y N 3 36,000 20,440 560 RII 130 37.5 47.5 37.5 Vomiting reported
ETF-6 F/11 9.3 8 38.4 Y Y Y 5 17,360 800 160 RII 140 37.5 47.5 37.5 Dose confirmed vomited
ETF-7 M/13 11.2 10 36.6 N N N 0 80 Negative 2,000 RIII 145 115 165 37.5 Dose vomited
ETF-8 F/10 6.6 9 35.1 Y Y N 3 18,000 1,600 160 RII 155 37.5 375 37.5 Vomiting reported
ETF-9 F/22 7.1 10 36.3 N N N 0 120 600 5,880 RIII 165 230 47.5 37.5 Unsupervised dose of MQ
ETF-10 F/21 6 8 36.3 Y Y N 3 13,200 2,200 240 RII 190 80 47.5 37.5 Vomiting reported
ETF-11 F/16 8.4 8 38.0 Y N N 2 12,400 2,160 120 RII 210 125 47.5 37.5 No apparent cause for failure
ETF-12 F/22 6.8 10 36.8 N N N 0 120 520 1,000 RIII 255 300 47.5 85 No apparent cause for failure
ETF-13 M/20 9.3 7 36.8 N N Y 1 16,680 400 400 RII 295 37.5 47.5 37.5 Mother reported diarrhea
ETF-14 M/19 6.6 8 38.0 Y N N 3 36,480 34,000 40 RIII 305 185 47.5 37.5 No apparent cause for failure
ETF-15 M/18 9.4 7 36.2 Y Y Y 5 7,840 76,000 D4 S/P RIII 350 140 47.5 37.5 Vomiting reported
Mean 17.2 8.9 8.3 36.8 53% 40% 40% 2.1 3,083 2,632 349 171.8 104.2 77.2 40.7
95% CI 2.4 1.1 0.8 0.5 0.9 48.7 41.6 44.4 6.21
* Illness Score based on number of indivudal symptoms reported by the child’s mother. Weight is given in kilograms, hemoglobin as grams per decaliter (g/dL), and age in months.
MEFLOQUINE TREATMENT IN GHANA 227
FIGURE 1. Scatterplot of whole blood concentrations determined
for MQ and its carboxylic acid metabolite (CMQ) on days 2 (A) and
28 (B) after oral treatment with 20 mg/kg MQ base in young Gha-
naian children whose responses to treatment were classified as sen-
sitive (cure), delayed RI-type parasitologic failures, or RII/RIII early
parasitologic failure. The solid line at 620 ng/mL designates the MQ
blood concentration considered to be the threshold level required for
suppression of sensitive P. falciparum infections. This figure appears
in color at www.ajtmh.org.
the 12 symptomatic cases and 12 others with parasitemias
> 6,680/
L (Table 4). Day 2 levels of MQ > 500 ng/mL char-
acterized 33% of children with RI-type late recurrences com-
pared with 44% of children who remained clear through 28
days. Mean Day 2 drug and metabolite levels in these late
parasitemias were also lower, but not significantly different,
from that of sensitive cases (MQ:, 467RI versus 561S ng/mL;
P 
 0.28; CMQ: 220RI versus 229S ng/mL; P 
 0.90). Day 2
levels of MQ typically exceeded those of CMQ, averaging 2.6,
3.1, and 3.4 times higher in the RII/RIII, RI, and S groups,
respectively.
Day 2 levels of MQ < 500 ng/mL and Day 2 levels of CMQ
below the level of detection were both strongly associated
with vomiting after dosing (P < 0.0007) and RII/RIII (P <
0.001) parasitologic responses. There was a stronger negative
association on Day 2 between levels of CMQ and pre-
treatment vomiting (r 
 −0.41) than between MQ and vom-
iting (r 
 −0.21). We found no association between mother’s
report at baseline of diarrhea in her child and Day 2 levels of
either MQ or CMQ. No association was seen between levels
of MQ or CMQ and parasitemia, age, weight, or hemoglobin.
Across all outcome groups, Day 28 MQ levels (Figure 1B;
Tables 2–4) were well below the putative boundary level of
TABLE 3
Patient descriptions, symptoms at enrollment, follow-up parasitemia, parasitologic classification, and whole blood levels of MQ and CMQ metabolite in MQ-treated cases of falciparum malaria
that were cured
Patient characteristics at enrollment Symptoms before treatment Parasite densities/
L Drug and metabolite levels (ng/mL)
ID# Sex/age Hb Weight Temp. Fever Vomiting Diarrhea Score* D0 D2–6 D7 D8–25 D26–28 Class D2-MQ D2-CMQ D28-MQ D28-CMQ
S-1 F/7 10.6 8 36.1 Y N N 1 320 Negative Negative Negative Negative S 295 135 47.5 37.5
S-2 F/18 9.1 9 35.5 N N N 1 10,040 80 Negative ND Negative S 295 380 125 150
S-3 M/23 6.8 8 36.7 Y N Y 3 23,200 200 Negative ND Negative S 375 120 47.5 125
S-4 M/6 8.3 6 37.2 N N Y 2 24,640 400 Negative ND Negative S 395 125 95 115
S-5 M/24 10.4 11 36.1 Y Y Y 5 88,000 400 Negative ND Negative S 405 105 47.5 80
S-6 F/15 10.4 8 37.9 Y Y N 3 8,840 80 Negative ND Negative S 425 255 115 105
S-7 F/13 9.9 8 35.2 Y Y N 3 1,200 Negative Negative Negative Negative S 475 230 47.5 37.5
S-8 M/7 6.9 8 37.8 Y Y N 4 28,000 Negative Negative Negative Negative S 525 37.5 120 37.5
S-9 F/10 9.9 9 37.1 Y N N 2 21,840 40 Negative ND Negative S 580 375 175 180
S-10 M/29 10.1 11 37.4 Y N N 1 3,880 Negative Negative ND Negative S 595 310 47.5 130
S-11 M/7 11.5 7 35.9 Y Y Y 3 880 Negative Negative ND Negative S 600 190 615 125
S-12 M/18 9.3 10 39.9 Y N Y 4 3,600 120 Negative ND Negative S 710 145 425 115
S-13 F/16 8.8 7 36.3 Y N Y 3 4,960 Negative Negative ND Negative S 755 165 47.5 37.5
S-14 F/21 9.9 10 35.6 N N N 0 1,200 120 Negative ND Negative S 830 345 200 95
S-15 M/18 11.4 12 38.9 Y N N 3 1,800 Negative Negative Negative Negative S 835 520 47.5 150
S-16 F/12 9.3 7 36.8 Y Y Y 3 840 Negative Negative Negative Negative S 875 235 175 37.5
Mean 15.25 9.5 8.7 36.9 81% 38% 44% 2.6 4915 136 560.6 229.5 148.6 103.6
95% CI 3.3829 0.7 0.8 0.6 0.6 95.0 62.7 77.5 26.8
* Illness Score based on number of individual symptoms reported by child’s mother. Weight is given in kilograms, hemoglobin as grams per decaliter (g/dL), and age in months.
228 FRYAUFF AND OTHERS
TABLE 4
Patient descriptions, symptoms at enrollment, follow-up parasitemia, parasitologic classification, and whole blood levels of MQ and CMQ metabolite in MQ-treated cases of falciparum malaria
with delayed RI-type parasitologic outcomes
Patient characteristics at enrollment Symptoms before treatment Parasite densities/
L Drug and metabolite levels (ng/mL)
ID# Sex/age Hb Weight Temp. Fever Vomiting Diarrhea Score* D0 D2–6 D7 D8–25 D26–28 Class D2-MQ D2-CMQ D28-MQ D28-CMQ Comments
LTF-1 F/9 9.7 6 37.1 Y N N 1 400 Negative Negative 200 13,600 RI 47.5 37.5 47.5 37.5 Vomiting reported
LTF-2 F/14 8.8 9 38.1 Y Y N 3 32,200 7,560 Negative 9,200 Negative RI 170 37.5 47.5 37.5
LTF-3 F/8 8.4 7 36.2 Y Y Y 4 200 Negative Negative 560 Negative RI 200 37.5 47.5 37.5 Vomiting reported
LTF-4 F/24 11.2 9 36.3 N N N 0 80 Negative Negative ND 34,400 RI 205 295 550 75
LTF-5 M/16 8.1 10 35.9 Y N N 1 32,000 12,760 Negative D15-28160 RI 270 130 47.5 105
LTF-6 M/24 8.8 9 36.1 Y Y Y 3 23,600 80 Negative 5,200 120 RI 320 85 47.5 37.5
LTF-7 M/18 7.3 7 36.6 Y Y Y 4 3,080 Negative Negative D16-7800 RI 325 105 47.5 37.5
LTF-8 F/16 8.2 9 38.8 Y Y Y 4 128,000 6,400 Negative 960 80 RI 330 37.5 47.5 37.5
LTF-9 F/18 7.9 8 37.5 Y Y N 3 10,800 Negative Negative ND 6,880 RI 355 115 145 37.5
LTF-10 M/16 8.9 9 37.2 Y N Y 2 23,200 Negative Negative ND 8,400 RI 385 140 47.5 37.5
LTF-11 F/12 8 6 37.9 Y N N 3 160 120 Negative 156,000 520 RI 405 120 47.5 75 Vomiting reported
LTF-12 M/22 9.7 7 34.9 N N N 0 22,000 160 Negative ND 26,000 RI 420 260 47.5 140 Vomiting reported
LTF-13 M/24 9.5 9 36.5 Y Y N 2 14,000 640 Negative 5,920 3,440 RI 430 475 47.5 37.5 Vomiting reported
LTF-14 F/10 7.7 8 36.8 Y N N 2 76,000 240 Negative ND 32,400 RI 455 115 170 37.5
LTF-15 F/20 8.7 10 38.1 Y Y N 4 76,800 400 Negative ND 6,680 RI 470 125 47.5 140
LTF-16 F/13 5.6 7 37.4 Y N N 2 9,600 Negative Negative ND 20,280 RI 490 390 135 37.5
LTF-17 F/22 10.3 9 38.5 Y N N 3 21,920 800 Negative D23-24800 RI 535 210 47.5 37.5
LTF-18 F/15 7.6 8 35.7 Y N N 2 24,400 Negative Negative ND 16,400 RI 550 560 47.5 80
LTF-19 M/7 9.5 7 36.2 Y Y N 3 31,440 Negative Negative 8,040 Negative RI 580 85 47.5 37.5
LTF-20 F/21 7.6 9 38.7 Y N N 2 3,400 Negative Negative ND 64,000 RI 750 490 47.5 155
LTD-21 F/24 7.8 11 36.0 Y N Y 2 4,320 Negative Negative ND 6,720 RI 765 570 47.5 37.5
LTF-22 M/23 9 7 36.2 Y N N 1 480 160 Negative ND 36,800 RI 855 290 140 295
LTF-23 M/6 8.7 7 38.9 Y N N 3 34,680 22,800 Negative ND 15,280 RI 890 230 47.5 95 Vomiting reported
LTF-24 F/23 10.5 9 36.1 Y N Y 3 16,480 Negative Negative ND 9,000 RI 1,005 350 47.5 37.5
Mean 16.9 8.6 8.2 37.0 92% 38% 29% 2.4 7,836 863 3,739 7,058 467.0 220.4 85.1 71.8
95% CI 2.4 0.5 0.5 0.4 0.5 96.7 68.4 42.5 24.4
* Illness Score based on number of individual symptoms reported by the child’s mother. Weight is given in kilograms, hemoglobin as grams per decaliter (g/dL), and age in months.
MEFLOQUINE TREATMENT IN GHANA 229
620 ng/mL needed to suppress susceptible strains of P. falci- their populations.6 However, among the in vivo trials that
parum. Drug and/or metabolite levels were below the limits of have followed, reasonable validation based on demonstration
detection in 7 of the 16 “sensitive” cases and in 22 of the 24 of parasitemia and high levels of MQ in blood has only been
late recurrent cases. Parent drug was more frequently de- documented from Northern Cameroon.5 Two separate trials
tected in sensitive cases than in RI cases (S: 56% versus RI: of the 25-mg/kg dose in Nigerian children ranging from 6
22%; P 
 0.02), but mean MQ levels were not significantly months to 10 years of age reported 28-day cure rates of 93%
different in the two groups. The CMQ metabolite was more and 95%, respectively, both of which were supported by in
often detected than MQ (40% versus 31%) and at concentra- vitro measures of reduced MQ susceptibility, but neither were
tions exceeding MQ in 17 of 29 cases (mean ratio of CMQ to validated by MQ blood levels at the time of recurrence.24,25 In
MQ: 1.4) where either parent drug or metabolite was mea- Northern Cameroon, among children 1–10 years of age, 13%
surable. of cases showed persistent or rising parasitemias during the 7
One case of drug tolerance was documented in which a days after MQ 25-mg/kg single dose treatment.5 Based on
parasitemia of 34,400/
L on Day 28 was observed against an Day 3 MQ levels > 500 nmol/L in these six children, RII-RIII
MQ concentration of 550 ng/mL (1.46 
mol/L). With a con- resistance to MQ was considered proven. In the absence of
ceivable 10% loss of MQ from this sample as a result of time prior MQ use in that area and supported by in vitro evidence,
and storage conditions, the whole blood concentration at the investigators hypothesized MQ resistance in Northern Cam-
time of parasitemia may have been as high as 605 ng/mL, with eroon to be a by-product of quinine resistance.5
plasma levels even higher. Surprisingly, among Ghanaian children with good treat-
ment outcomes, the 20-mg/kg MQ dose did not seem to be
well absorbed, because Day 2 concentrations of drug > 500
DISCUSSION ng/mL, a level associated with treatment success,5,10,12,18,19
were measured in only 33% of the children. Mefloquine is
Apart from their critical need for malaria care during this rapidly absorbed, attains maximum blood levels within 24
most vulnerable period in their lives, young children of 6–24 hours of oral dosing, and concentrations of ∼1,000 ng/mL are
months of age are ideal for showing the effect of a drug, typically attained in older children and adults by the 25-mg/kg
alone, against malaria parasites and yield a valid in vivo test treatment dose.26–28 Early treatment failures of MQ in
outcome for drug resistance that is free from the confounding Malawian children were strongly associated with MQ blood
effects of maternally and naturally acquired immunity. Lower levels < 500 ng/mL on Day 2. These low Day 2 drug levels
MQ failure rates in adults compared with young children have were in turn associated with vomiting, which occurred in 40%
been attributed to immunity, owing to comparable Day 2 drug of children given a 15-mg/kg single dose and in 29% of those
levels in the two groups.3 Our findings in this select popula- who received a 25-mg/kg single dose. Surprisingly, among
tion of African infants yield no firm evidence of innate RII- children who did not vomit within 30 minutes of treatment,
RIII resistance in the parasites and indicate that low drug Day 2 blood levels of MQ were comparable in both groups,
levels resulting from emesis, vomiting, poor absorption, and/ but 18% still had Day 2 MQ levels < 500 ng/mL. Because
or incomplete dosing account for all of the 15 RII/RIII para- unremitting or recurrent parasitemias developed within the
sitologic responses that occurred. Based on the low levels of 28-day test period in comparable proportions of low (62%)
MQ measured 48 hours after dosing in children who achieved and high (55%) dose groups, the authors refrained from any
better treatment outcomes, we believe that re-infection, and interpretation of MQ resistance and called attention to the
not late recrudescence, accounted for the majority of late problems of erratic drug absorption and vomiting in very
recurrent parasitemias seen. In the absence of molecular evi- young children.19 In this regard, it seemed unusual that we
dence showing genotype differences between baseline and saw so few RII/RIII-type MQ failures in Ghanaian children,
recurrent parasitemias, additional support for this supposition and we suspect that parasitemias at that time and location
derives from the malaria incidence measured among these were relatively susceptible to the drug.
same children immediately preceding their MQ treatment. To date, the MQ prophylaxis failures that were reported
Assuming full compliance, normal absorption, and a normal among adult Ghanaians have not been verified by drug levels
decline by Day 10 after receiving treatment to MQ blood and may not have occurred in all cases by MQ-resistant para-
levels below the minimum inhibitory threshold of 620 ng/mL, sites. Based on the low blood levels of MQ seen in their
there would have been 3,078 child-days (∼8.4 child-years) of children, wide inter-individual variation in MQ pharmacoki-
re-infection risk leading up to the Day 28 endpoint of our MQ netics, low blood levels associated with lighter parasitemia,
in vivo test. From the calculated 4.7 infections per child-year older age, and ethnic/genetic factors, it seems possible that
determined during that dry season,9 an expected 40 infections the high failure rate of MQ prophylaxis in 6 of 46 Ghanaian
would have occurred in our children. The difference between adults also resulted from poor drug absorption and inad-
the 50 late recurrent parasitemias observed and the 40 that equate blood levels.
were expected is presumed to relate to less than ideal absorp- Mefloquine performed poorly in Ghanaian infants, an age
tion of the drug and reduced sensitivity to MQ in a fraction of group already prone to vomiting and diarrhea, and wide-
the circulating P. falciparum strains. spread use of this drug, alone, for treatment of malaria may
Well before any use of MQ in Africa, P. falciparum strains exacerbate any existing low level resistance in the circulating
from Ghana and Ivory Coast showed tolerance to the highest malaria. The mean Day 2 blood level of MQ we measured in
levels of MQ in the standard WHO in vitro micro-test.23 In Ghanaian infants who tolerated MQ dosing well (mean: 501
vitro test results from Senegal, Mali, Cameroon, and Nigeria ng/mL; 95% CI: 431–571; range: 170–1,005 ng/mL) was lower
also gave indication of reduced sensitivity to MQ in the ab- than that reported for Senegalese (955 ± 74 ng/mL) or Mala-
sence of direct pressure resulting from the use of that drug in wian (633 ± 343 ng/mL) children, and was grossly below
230 FRYAUFF AND OTHERS
the Cmax reported for Thai children (2,031 ± 831 ng/ Received July 26, 2005. Accepted for publication July 6, 2006.
mL).3,16,18,19,28 In retrospect of the globally low drug levels Acknowledgments: The authors thank the parents and children who
achieved, we are also led to consider the possibility that our participated in this study and health workers and the support person-
formulating MQ as a suspension in a high-fructose, pine- nel of the Navrongo Health Research Center. The authors thank
apple-flavored syrup may somehow have altered the drug’s Charles Attiogbe of the Noguchi Memorial Institute of Medical Re-search for work as the study microscopist and Cletus Tindana, Salifu
chemistry and/or subsequent bioavailability. Abdul Rahman, and Paulina Tindana for field supervision. Special
The carboxylic acid metabolite of MQ is inactive against thanks are also extended to the study physicians, Drs. Kweku Enos
malaria but is deemed to have value in monitoring compli- and Mensah-Afful, and to Dr. Alex Nazzar for essential support and
ance during prophylaxis.29 Our purpose in presenting CMQ advice. This research was approved by scientific and ethical reviewboards of the Ghanaian Ministry of Health and the US Navy and was
data for MQ-treated Ghanaian infants is that of verifying conducted in accordance with regulations governing the protection of
drug consumption, providing an indication of the metabolism human subjects in medical research.
of parent MQ in this relatively under-studied group and Financial support: This study was supported by independent research
showing its potential for predicting early treatment failures. Grant WU 34 3C30.001.3601 and the US Department of Defense
Great individual variability is typically seen in blood levels of Global Emerging Infections Surveillance and Response System
MQ and CMQ under both therapeutic and prophylactic regi- (GEIS). The views of the authors expressed herein do not purport to
reflect those of the Ghanaian Ministry of Health, the US Navy, or the
mens, but pregnancy, parasitemia, and age are known deter- US Department of Defense.
minants of MQ absorption, distribution, and elimination.26–
Relevant to our findings in Ghanaian children are those Authors’ addresses: David J. Fryauff and Greg Utz, US Naval Medi-28,30
cal Research Unit No. 3, PSC 452 Box 52, FPO AE 09835-0007,
from Thailand which found that children 6–24 months of age Telephone: 20-2-342-0576, Fax: 20-2-342-7121, E-mai l :
achieve peak levels, tissue distribution, and elimination of FryauffD@nmrc.navy.mil and gcutz@nmcsd.med.navy.mil. Seth
MQ more rapidly than older children and adults.17 From this, Owusu-Agyei, Navrongo Health Research Center, Navrongo, Upper
assuming a 12-hour peak in MQ blood levels, such as that East Region, Ghana, Telephone: 233-742-22380, Fax: 233-742-22310,
E-mail: seth.owusu-agyei@ghana-khrc.org. J. Kevin Baird,
seen in young Thai children, our Day 2 measurement, 36 ALERTAsia Foundation, Jakarta, Indonesia, E-mail: jkevinbaird@
hours later, might be expected to show the heightened accu- yahoo.com. Kwadwo A. Koram and Francis Nkrumah, Noguchi Me-
mulation of CMQ over MQ in Ghanaian children.31 This was morial Institute of Medical Research, University of Ghana, Legon,
not apparent, however, and even at Day 28, we did not see Ghana, Telephone: 233-21-501178, Fax: 233-21-502182, E-mail:
KKoram@noguchi .mimcom.net and FNkrumah@noguchi
ratios of CMQ:MQ that approached those of 2–6 that have .mimcom.net. Fred Binka, School of Public Health, University of
been reported for adults on weekly prophylaxis.29,32 The Ghana, Legon, Ghana, Telephone: 233-21-500799, E-mail:
stronger negative correlation we observed on Day 2 between FBinka@indepth-nework.org. Stephen L. Hoffman, Sanaria Inc.,
CMQ levels and vomiting was noteworthy, and although our 12511 Parklawn Drive, Suite L, Rockvile, MD 20852, Telephone:
301-770-3222, Fax: 301-770-5554, E-mail: slhoffman@sanaria.com.
study was not intended or powered to examine the relation-
ship between drug/metabolite levels and clinical condition of Reprint requests: Research Publications Branch, US Naval Medi-
these children, no correlation was seen between diarrhea at cal Research Unit No. 3, PSC 452 Box 5000, FPO AE 09835-0007.E-mail: KaramE@namru3.med.navy.mil.
baseline and blood levels of either MQ or CMQ measured on
Day 2. Multiple studies have reported an association between
diarrhea and low MQ drug levels in blood.11–14,16, Diarrhea, REFERENCES
anemia, and vomiting as conditions before MQ dosing were
not predictive of low Day 2 levels of MQ or treatment failure 1. World Health Organization, 2001. Drug Resistance in Malaria.Geneva, Switzerland: World Health Organization.
in our Ghanaian children, but vomiting after the MQ dosing 2. Wongsrichanalai C, Prajakwong S, Meshnick SR, Shanks GD,
observation period was clearly related to low drug levels and Thimasarn K, 2004. Mefloquine—its 20 years in the Thai Ma-
early treatment failure. laria Control Program. SE Asian J Trop Med Publ Hlth 35:
In summary, a single oral dose treatment of a 20-mg/kg MQ 300–308.
3. Nosten F, ter Kuile F, Chongsphajaisiddhi T, Luxemburger C,
suspension to breast-feeding Ghanaian infants was well tol- Webster HK, Edstein M, Phaipun L, Thew KL, White NJ,
erated, but resulted in whole blood levels of the drug on Day 1991. Mefloquine-resistant falciparum malaria on the Thai-
2 that were surprisingly low, even among those children that Burmese border. Lancet 337: 1140–1143.
had tolerated the suspension without vomiting. Early RII/ 4. Oduola AMJ, Milhous WK, Salako LA, Walker O, Desjardins
RIII-type parasitologic failures that occurred in 8% of the RE, 1987. Reduced in-vitro susceptibility to mefloquine in
West African isolates of Plasmodium falciparum. Lancet ii:
infants were associated with vomiting after dosing, Day 2 1304–1305.
blood levels of MQ < 500 ng/mL, and undetectable Day 2 5. Brasseur P, Kouamouo J, Moyou-Somo R, Druilhe P, 1992.
levels of CMQ. Vomiting and diarrhea before dosing were Multi-drug resistant falciparum malaria in Cameroon in 1987–
not associated with treatment failure. Despite no prior use of 1988. II. Mefloquine resistance confirmed in vivo and in vitro
MQ for treatment at this location, one instance was found of and its correlation with quinine resistance. Am J Trop Med
Hyg 46: 8–14.
reduced sensitivity by the parasite to this drug. Such early 6. Mockenhaupt FP, 1995. Mefloquine resistance in Plasmodium
evidence does provide some rationale and credence for mul- falciparum. Parasitol Today 11: 248–253.
tiple cases of MQ prophylaxis failure that occurred at the 7. Lobel HO, Varma JK, Mianai M, Green M, Todd GD, Grady K,
same location 3 years later. Ghana thus joins a number of Barber AM, 1998. Monitoring for mefloquine-resistant Plas-
other African nations that pose a risk to travelers of MQ modium falciparum in Africa: implications for travelers’
prophylaxis failure. We consider the collective results of this health. Am J Trop Med Hyg 59: 129–132.
study and others to be a warning against further use and/or 8. Hale BR, Owusu-agyei S, Fryauff DJ, Koram KA, Adjuik M,Oduro AR, Prescott WR, Baird JK, Nkrumah FR, Franke ED,
promotion of treatment by MQ alone for falciparum malaria Binka FN, Richie TL, Horton J, Hoffman SL, 2003. A ran-
in this high transmission area of northern Ghana. domized, double-blind, placebo-controlled dose-ranging trial
MEFLOQUINE TREATMENT IN GHANA 231
of tafenoquine for weekly prophylaxis against Plasmodium fal- 20. Appawu M, Owusu-Agyei S, Dadzie S, Asoala V, Anto F, Koram
ciparum. Clin Infect Dis 36: 541–549. K, Rogers W, Nkrumah FK, Hoffman SL, Fryauff DJ, 2004.
9. Baird JK, Owusu-Agyei S, Utz GC, Koram KA, Barcus MJ, Malaria transmission dynamics at a site in northeastern Ghana
Jones TR, Fryauff DJ, Binka FN, Hoffman SL, Nkrumah FN, proposed for testing malaria vaccine. Trop Med Int Hlth 9:
2002. Seasonal malaria attack rates in infants and young chil- 164–170.
dren in northern Ghana. Am J Trop Med Hyg 66: 280–286. 21. Bergqvist Y, Kabbani JA, Krysén B, Palme IB, Rombo L, 1993.
10. Slutsker LM, Khoromana CO, Payne D, Allen CR, Wirima JJ, High-performance liquid chromatographic method for the si-
Heymann DL, Patchen L, Steketee RW, 1990. Mefloquine multaneous determination of mefloquine and its carboxylic
therapy for Plasmodium falciparum malaria in children under metabolite in 100-ul capillary blood samples dried on paper. J
5 years of age in Malawi: In vivo/in vitro efficacy and correla- Chromatog 615: 297–302.
tion of drug concentration with parasitological outcome. Bull 22. Lobel HO, Miani M, Eng T, Bernard KW, Hightower AW,
WHO 68: 53–59. Campbell CC, 1993. Long-term malaria prophylaxis with
11. ter Kuile FO, Luxemburger C, Nosten F, Thwai KL, Chong- weekly mefloquine. Lancet 341: 848–851.
suphajaisiddhi T, White NJ, 1995. Predictors of mefloquine 23. Hogerzeil HV, Hogewoning AA, Van Doorn JW, Wernsdorfer
treatment failure: a prospective study of 1590 patients with WH, Van der Kaay HJ, 1985. In vitro assessment of sensitivity
uncomplicated falciparum malaria. Trans R Soc Trop Med Hyg of Plasmodium falciparum to chloroquine and mefloquine in
89: 660–664. Ghana. Trans R Soc Trop Med Hyg 79: 808–811.
12. Boudreau EF, Fleckenstein L, Pang LW, Childs GE, Schroeder 24. Sowunmi A, Oduola AM, 1995. Open comparison of mefloquine,
AC, Ratnaratorn B, Phintuyothin P, 1990. Mefloquine kinetics mefloquine/sulfadoxine/ pyrimethamine and chloroquine in
in cured and recrudescent patients with acute falciparum ma- acute uncomplicated falciparum malaria in children. Trans R
laria and in healthy volunteers. Clin Pharmacol Ther 48: 399– Soc Trop Med Hyg 89: 303–305.
409. 25. Okoyeh JN, Lege-Oguntoye L, Ugbode RO, Ogunrinde GO,
13. Karbwang J, White NJ, 1990. Clinical pharmacokinetics of me- 1997. Responses of multidrug-resistant Plasmodium falci-
floquine. Clin Pharmacokinet 19: 264–279. parum parasites to mefloquine in Nigerian childfren. Trop Med
14. Dubos F, Delattre P, Demar M, Carme B, Gendrel D, 2004. Int Hlth 2: 319–324.
Safety of mefloquine in infants with acute falciparum malaria. 26. White NJ, 1988. Drug treatment and prevention of malaria. Eur
Ped Inf Dis J 23: 679–681. J Clin Pharmacol 34: 1–14.
15. Looareesuwan S, White NJ, Warrell DA, Forgo I, Dubach UG, 27. Karbwang J, Na-Bangchang K, 1994. Clinical application of me-
Ranalder UB, Schwartz DE, 1987. Studies on mefloquine bio- floquine pharmacokinetics in the treatment of P. falciparum
availability and kinetics using a stable isotope technique: a malaria. Fundam Clin Pharmacol 8: 491–502.
comparison of Thai patients with falciparum malaria and 28. Nosten F, ter Kuile F, Chongsuphajaisiddhi T, Na Bangchang K,
healthy Caucasian volunteers. Br J Clin Pharmacol 24: 37–42. Karbwang J, White NJ, 1991. Mefloquine pharmacokinetics
16. Fontanet AL, Johnston BD, Walker AM, Bergqvist Y, Hellgren and resistance in children with acute falciparum malaria. Br J
U, Rooney W, 1994. Falciparum malaria in eastern Thailand: Clin Pharmacol 31: 556–559.
A randomized trial of the efficacy of a single dose of meflo- 29. Todd GD, Hopperus Buma APCC, Green MD, Jaspers CAJJ,
quine. Bull WHO 72: 73–78. Lobel HO, 1997. Comparison of whole blood and serum levels
17. Singhasivanon V, Chongsuphajaisiddhi T, Sabcharoen A, At- of mefloquine and its carboxylic acid metabolite. Am J Trop
tanath P, Webster HK, Wernsdorfer WH, Sheth UK, Djaja Med Hyg 57: 399–402.
Lika I, 1992. Pharmacokinetics of mefloquine in children aged 30. Na-Bangchang K, Davis TME, Looareesuwan S, White NJ, Bun-
6 to 24 months. Eur J Drug Metab Pharmacokinet 17: 275–279. nag D, Karbwang J, 1994. Mefloquine pharmacokinetics in
18. Hatin I, Trape J-F, Legros F, Bauchet J, Le Bras J, 1992. Sus- pregnant women with acute falciparum malaria. Trans R Soc
ceptibility of Plasmodium falciparum strains to mefloquine in Trop Med Hyg 88: 321–323.
an urban area in Senegal. Bull WHO 70: 363–367. 31. Franssen G, Rouveix B, Lebras J, Bauchet J, Verdier F, Michon
19. MacArthur JR, Stennies GM, Macheso A, Kolczak MS, Green C, Bricaire F, 1989. Divided-dose kinetics of mefloquine in
MD, Ali D, Barat LM, Kazembe PN, Ruebush TK II, 1998. man. Br J Clin Pharmac 28: 179–184.
Efficacy of mefloquine and sulfadoxine-pyrimethamine for the 32. Wallace MR, Sharp TW, Smoak B, Iriye C, Rozmajzl P, Thorn-
treatment of uncomplicated Plasmodium falciparum infection ton SA, Batchelor R, Magill AJ, Lobel HO, Longer CF,
in Machinga District, Malawi, 1998. Am J Trop Med Hyg 64: Burans JP, 1996. Malaria among United States troops in So-
679–684. malia. Am J Med 100: 49–56.
View publication stats