Tropical Medicine and International Health doi:10.1111/tmi.

12919

volume 22 no 9 pp 1186–1195 september 2017

Effect of cotrimoxazole prophylaxis on malaria occurrence

among HIV-infected adults in West Africa: the MALHIV Study
e1,2,3, Fr

Serge P. Eholi
eric N. Ello1,2, Patrick A. Coffie1,2,3, Ars
ed
ema4, Daouda K. Minta5 and
ene H

4
Adrien Sawadogo
1 D
epartement de Dermatologie-Infectiologie, Unit
e de Formation et de Recherche des Sciences M
edicales, Universit
e F
elix Hou-
phouet-Boigny, Abidjan, C^ ote d’Ivoire
2 Service des Maladies Infectieuses et Tropicales, Centre Hospitalier Universitaire de Treichville, Abidjan, C^
ote d’Ivoire
3 PAC-CI Program, Centre Hospitalier Universitaire de Treichville, Abidjan, C^ ote d’Ivoire
4 Service des Maladies Infectieuses et Tropicales, H^
opital de Jour Centre Hospitalier Universitaire Sourou Sanon, Bobo-Dioulasso,
Burkina Faso
5 Service des Maladies Infectieuses et Tropicales, Centre Hospitalier Universitaire du Point G, Bamako, Mali

Abstract introduction Cotrimoxazole (CTX) should be given to all HIV-infected adults with mild or severe
HIV-disease or those with CD4 counts below 350/mm3 according to 2006 WHO guidelines. We
assessed the impact of CTX prophylaxis on the risk of malaria episodes in HIV-1-infected adults
from four West African countries with different patterns of malaria transmission.
method Multicentric cohort study, conducted between September 2007 and March 2010 in four
West African cities. Antiretroviral therapy (ART) na€ıve HIV-infected adults started CTX at enrolment
(CTX group) if they had CD4 < 350 cells/mm3 or were at WHO clinical stage ≥2. For patients who
did not start CTX at enrolment (non-CTX group) and started CTX afterwards, follow-up was
censored at CTX initiation. We used Cox’s proportional hazard model to compare the risk of malaria
between CTX groups.
results A total of 514 participants (median CD4 count 238 cells/mm3) were followed for a median of
15 months. At enrolment, 347 started CTX, and 261 started ART. During the follow-up, 28 started
CTX. The incidence of malaria was 8.7/100 PY (95%CI 6.3–11.5) overall, 5.2/100 PY (95%CI 3.1–8.3)
in the CTX group and 15.5/100 PY (95%CI 10.3–22.1) in the non-CTX group. In multivariate analysis,
CTX led to a 69% reduction in the risk of malaria (aHR 0.31, 95%CI 0.10–0.90).
conclusion Patients in the CTX group had an adjusted risk of malaria three times lower than
those in the non-CTX group. The prolonged large-scale use of CTX did not blunt the efficacy of
CTX to prevent malaria in this region.

keywords HIV, malaria, cotrimoxazole, West Africa, adults

Malaria and HIV can influence each other’s pathogeni-

Introduction
In 2015, 60% of the 214 million cases of malaria world-
wide occurred in sub-Saharan Africa, where 26 million of
the 36.7 million HIV-infected individuals live [1, 2]. This
strong overlap between malaria and HIV displays some
heterogeneicity [3, 4]. For instance, transmission of
malaria is unstable in East or Southern Africa compared
to Western or central regions, while the former has a
much higher HIV-prevalence [4–6]. Although the attribu-
table mortality of malaria is much higher in young chil-
dren, recent reports suggested increasing frequency and
severity of Plasmodium falciparum infection among older
children and adults living in African urban settings who
also happen to bear the highest HIV-prevalence [7–9].
sis. HIV infection increases the risk of malaria acquisition
and its severity, with a more pronounced effect as
immunosuppression becomes more severe [10–13].
Malaria episodes are associated with transient lymphope-
nia and a decrease in CD4+ cell count among individuals
with increased levels of HIV RNA replication [14–16].
These negative effects must be dealt with by reinforcing
interventions against both diseases [17].
In HIV-infected persons living in tropical areas, malaria
is one of the targets of cotrimoxazole (CTX) prophylaxis,
together with invasive bacterial diseases, pneumocystosis,
toxoplasmosis and isosporosis. In the era prior to
antiretroviral therapy (ART), a randomised controlled trial
in C^ote d’Ivoire demonstrated that CTX prophylaxis

1186 © 2017 John Wiley & Sons Ltd

Tropical Medicine and International Health
S. P. Eholi
e et al. Cotrimoxazole malaria prophylaxis for HIV patients
decreased by 90% the risk of malaria episodes in HIV-
infected adults [18]. These findings were further confirmed
in studies performed in East Africa, where a reduction in
clinical malaria with CTX in HIV-infected adults and chil-
dren ranged from 40% to 95%, when accounting for the
effect of Insecticide-treated bednets (ITNs) [19–29].
Based on this evidence, WHO recommended in 2006
that CTX should be given to all HIV-infected adults with
CD4 cell counts below 350/mm3 and/or at WHO clinical
stage 2–4 [30], and then revised the guidelines in 2014 to
recommend initiating CTX regardless of CD4 cell count
and WHO stage in settings where malaria and/or severe
bacterial infections were highly prevalent [31]. However,
some unresolved issues remain regarding malaria and
CTX in HIV-infected individuals. Firstly, most of the
national guidelines still recommend starting CTX when
CD4 cells count is below 350 cells/mm3, even in settings
with high malaria burden [30]. Secondly, national recom-
mendations are hardly applied in the field, where many
HIV-infected patients eligible for CTX prophylaxis still
do not receive it [32]. Thirdly, most of the evidence for
CTX prophylaxis efficacy against malaria in West Africa
was obtained in the pre-ART era. In this region, CTX
prophylaxis was recommended as early as 2000, after the
results of the randomised trials in C^ ote d’Ivoire [18] were
published. Some authors then hypothesised that large-
scale use of CTX might accelerate parasitic resistance
and limit its efficacy in the long term [22, 33].
The aim of this study was to assess the impact of CTX
prophylaxis on the risk and severity of malaria episodes
in HIV-1-infected adults from four West African coun-
tries with different patterns of malaria transmission.
Methods
Study design and settings
MALHIV was a multicentre cohort study conducted
between September 2007 and March 2010 in four West
African cities: Abidjan (C^ote d’Ivoire), Bamako (Mali),
Bobo-Dioulasso (Burkina Faso) and Dakar (Senegal). The
study protocol was approved by national ethics commit-
tees of the countries. Study centres were Infectious and
Tropical Disease Departments at university teaching hos-
pitals within each city.
Study participants
Eligibility criteria were as follows: HIV-1 infection, at least
18 years old, no ongoing CTX or other sulfamide treat-
ment, ART-na€ıve and no concomitant self-medication.
Exclusion criteria were patients with HIV-2 monoinfection
© 2017 John Wiley & Sons Ltd
volume 22 no 9 pp 1186–1195 september 2017
or dually reactive HIV-1/2 infection, any acute infectious
episode, first trimester of pregnancy, history of allergy to
sulfamides, haemoglobin level <7.5 g/ml, and neutrophil
count <750 cells/ml. All patients gave written informed
consent before inclusion.
Enrolment and follow-up
Patients enrolled in the study were requested to complete
a baseline visit at enrolment, and then every 12 weeks.
Between scheduled visits, patients had free access to the
study centre in case of fever or other symptoms. Socio-
demographic, ITN use and clinical data were recorded at
baseline. At each visit, clinical data, including symptoms
of malaria, were assessed. Adherence to ART and CTX
was measured by pill count. Adverse events were graded
in accordance with the WHO grading table [34]. Severe
morbidity other than malaria was defined as the occur-
rence of any WHO stage 3 or 4 defining HIV-disease or
any clinical event leading to at least one day of hospitali-
sation and/or death.
Treatment of HIV and malaria
Patients initiated CTX (800 mg sulfamethoxazole and
160 mg trimethoprim, once daily) at enrolment (CTX
group) if they were symptomatic for HIV (WHO clinical
stage 2 to 4) or had a CD4 count <350 cells/mm3,
according to national and WHO 2006 guidelines [35].
Patients initiated ART at enrolment based on WHO
2006 criteria (CD4 count <200 cells/mm3, WHO clinical
stage 4, or WHO clinical stage 2–3 and CD4 count <350
cells/mm3), or if they had a plasma HIV RNA >105
copies/ml. The latter criterion was included in the study
protocol, although it was not included in the national
guidelines in any of the countries.
Two clinical forms of malaria were distinguished:
uncomplicated malaria, defined by fever without clinical
or biological severity criteria, and severe malaria, defined
by fever with at least one major criterion of severity
according to WHO definition [36].
Based on national guidelines, patients with an uncom-
plicated malarial episode received a fixed-dose combina-
tion of artesunate plus amodiaquine (AS200 mg/AQ50 mg,
one pill BID for three consecutive days) as first-line treat-
ment, or a fixed-dose combination of artemether plus
lumefantrine (A120 mg L20 mg, four pills BID for three
consecutive days) as an alternative. In the event of severe
malaria, intravenous infusion of quinine (8 mg/8 h per
day for 5–7 days) was prescribed as first-line treatment,
or intramuscular artemether (2.4 mg/12 h for 5–7 days)
as an alternative. The efficacy of antimalarial treatment
1187
Tropical Medicine and International Health
S. P. Eholi
e et al. Cotrimoxazole malaria prophylaxis for HIV patients
was evaluated according to WHO 28 day-protocol [37].
In case of treatment failure, merozo€ıte proteins 1 and 2
(msp1 and msp2) genotyping were performed to distin-
guish recurrence from new infection [38].
For women who became pregnant during follow-up
and were already on CTX prophylaxis, CTX was main-
tained and no other treatment was given. If the woman
had no eligibility criteria for CTX, she received a prophy-
laxis with proguanil+atovaquone in fixed-dose combina-
tion [30].
Laboratory procedures
Biological tests, including complete blood count, liver
and renal function tests, CD4 counts, plasma HIV viral
load, thick and thin blood smears (Giemsa Stain), were
performed at enrolment and every six months. CD4 cells
were counted by FACS Count Flow cytometer (Fascan
Becton-Dickinson, San Carlos, CA, USA). Plasma HIV-1
RNA levels were quantified using the Amplicor 1.5
Roche assay (Rotkreuz, Switzerland), detection threshold
50 copies/ml or the Abbott Real Time HIV-1 m2000rt
quantitative assay (Abbott Laboratories, Chicago, IL,
USA, detection threshold 40 copies/ml).
Patients who presented with documented fever (axillary
temperature ≥37.5 °C) or history of fever in the previous
24 h had a finger prick for thick and thin blood smears
for Plasmodium species detection. Thick smears were
screened for the presence of malaria parasites according
to WHO procedures. Parasite density was estimated by
counting the number of asexual parasites per 200 white
blood cells (WBC) and was expressed as the number of
parasites per microlitre (ll) of blood. Parasite species
were confirmed by Plasmodium small subunit ribosomal
RNA (ssrRNA) gene amplification using nested-PCR as
previously described [38].
If the smear was positive for any Plasmodium sp par-
asite, the patient was considered as having malaria
regardless of parasite density. For every confirmed case
of malaria, the 2003-revised WHO 28-day evaluation
protocol for antimalarial drug efficacy was applied. It
included clinical examination, thick blood smears on
days 0, 1, 2, 3, 14, 21 and 28 after treatment initiation,
other biological tests (complete blood cell counts, liver
and renal function tests) on days 0, 14 and 28 and CD4
count and HIV viral load measurements on days 0 and
28.
Statistical analysis
Patients were analysed in two groups, defined by the
prescription (CTX group) or absence of prescription
1188
volume 22 no 9 pp 1186–1195 september 2017
(non-CTXgroup) of CTX at enrolment. For patients who
were not prescribed CTX at enrolment (non-CTX group)
and who started CTX later, follow-up was censored at
the date of the first CTX prescription. For patients who
were not prescribed ART at inclusion and who started
ART later, follow-up was censored at the date of the first
ART prescription. Analyses were performed on intent-to-
treat basis, meaning that all patients from CTX group
were considered receiving CTX during their entire
follow-up period.
For baseline characteristics, comparisons between
groups were performed using Student’s t-test,
Mann–Whitney U test or variance analysis for continuous
variables, and chi-square test or Fisher’s exact test for
categorical variables.
The rate of malaria was estimated with its 95% confi-
dence interval (CI). The rate was calculated as the num-
ber of patients having at least one incident episode of
malaria per 100 person-years (PY) of ‘at-risk follow-up’.
The ‘at-risk period’ began at baseline and ended at the
earliest among the following events: first malaria episode,
death, lost to follow-up, CTX initiation (for non-CTX
group only), ART initiation (for patients who did not
start ART at baseline only) or end of study (at
month-18).
Multivariable Cox’s proportional hazard regression
method was used to compare the risk of malaria
between CTX groups, adjusting for baseline CD4
count, ART, HIV viral load, WHO clinical stage and
any other variables found to be associated with the risk
of malaria with a P < 0.25 in univariable analysis.
Adjusted hazard ratios (aHR) were reported with 95%
CI. P-values were two-sided. Analyses were performed
using SAS software version 9.1 (SAS Institute, Cary,
NC, USA).
Results
Baseline characteristics
A total of 514 patients were enrolled: 184 (35.8%) in
Bobo-Dioulasso, 157 (30.5%) in Abidjan, 99 (19.3%) in
Bamako and 74 (14.4%) in Dakar. Overall, 347 (67.5%)
were in CTX group and 167 (32.5%) in non-CTX group.
The proportion of patients in CTX group was similar
across sites (range 62.6–72.3%). ART was initiated at
inclusion in 261 (50.8%) patients overall, 245 (70.6%)
in CTX group and 16 (9.6%) in non-CTX group
(P < 0.01). All first-line ART regimens were composed of
two nucleos(-t)ide reverse transcriptase inhibitors (NRTI),
combined with a non-nucleos(-t)ide reverse transcriptase
inhibitor (NNRTI, n = 243, 93.1%), a third NRTI
© 2017 John Wiley & Sons Ltd
Tropical Medicine and International Health volume 22 no 9 pp 1186–1195 september 2017

S. P. Eholi
e et al. Cotrimoxazole malaria prophylaxis for HIV patients

(n = 14, 5.4%), or a protease inhibitor (n = 4, 1.5%). Follow-up

Compared to patients in the CTX group, those in the
Participants were followed for a total of 583 person-years
non-CTX group were younger (P ˂ 0.01), were more fre-
(median 15.3 months, Interquartile range (IQR): 10.1–
quently women (P = 0.02), had higher CD4 cell count
18.0). Of the 167 patients in CTX group, 28 started
(P ˂ 0.01), and lower HIV viral load (P ˂ 0.01)
CTX during follow-up after a mean of 8.8 months. At
(Table 1).

Table 1 Baseline and follow-up characteristics

Total (N = 514) Non-CTX group (N = 167) CTX group (N = 347) P

Baseline
Age (years)
Median (IQR) 35 (29–42) 32 (27–39) 37 (31–43) <0.01
Gender, N (%)
Male 153 (29.8) 38 (22.8) 115 (33.1) 0.02
Female 361 (70.2) 129 (77.2) 232 (66.9)
Body mass index (kg/m²)
Median (IQR) 21.0 (19.0–23.7) 22.2 (19.7–25.3) 20.7 (18.8–23.3) <0.01
Schooling, n (%)
No or primary school 332 (64.6) 105 (62.9) 227 (65.4) 0.62
Secondary or more 182 (35.4) 62 (37.1) 120 (34.6)
CDC clinical stage, n (%)
A or B 468 (91.0) 157 (94.0) 311 (89.6) 0.14
C 46 (9.0) 10 (6.0) 36 (10.4)
ART initiated at baseline, n (%)
Yes 261 (50.8) 16 (9.6) 245 (70.6) <0.01
No 253 (49.2) 151 (90.4) 102 (29.5)
Use of bednet, n (%)
Yes 381 (74.1) 121 (72.5) 260 (74.9) 0.59
No 133 (25.9) 46 (27.5) 87 (25.1)
CD4+ T lymphocyte count (cells/mm3)
Median (IQR) 238 (130–398) 494 (406–664) 176 (92–253) <0.01
<100, n (%) 101 (19.7) 6 (3.6) 95 (27.4) <0.01
100–199, n (%) 105 (20.4) 2 (1.2) 103 (29.7)
200–349, n (%) 156 (30.3) 7 (4.2) 149 (42.9)
≥350, n (%) 152 (29.6) 152 (91.0) 0 (0.0)
Plasma HIV-1 RNA (log10copies/ml)
Median log10copies/ml (IQR) 5.0 (3.8–5.6) 4.6 (3.3–5.2) 5.3 (4.2–5.8) <0.01
<105, n (%) 258 (50.2) 114 (68.2) 144 (41.5) <0.01
≥105, n (%) 256 (49.8) 53 (31.8) 203 (58.5)
Haemoglobin (g/dl)
Median (IQR) 11.1 (10.0–12.3) 11.7 (10.7–12.6) 10.9 (9.7–12.1) <0.01
Plasma ALAT (IU/l)
Median (IQR) 17 (12–26) 15 (10–23) 19 (13–29) <0.01
Follow-up
Follow-up, median (IQR) 15.3 (10.1–18.0) 15.6 (10.6–18.0) 15.2 (9.8–18.0) 0.42
Status at study termination, n (%)
Death 13 (2.5) 3 (1.8) 10 (2.9) 0.07
Alive, not censored 290 (56.4) 83 (49.7) 207 (59.6)
Censored 211 (41.1) 81 (48.5) 130 (37.5)
CTX and/or ART initiation 109 (51.7) 44 (54.3) 65 (50.0)
Transferred out 11 (5.2) 4 (4.9) 7 (5.4)
Voluntary study discontinuation 9 (4.3) 4 (4.9) 5 (3.8)
Lost to follow-up 82 (38.9) 29 (35.8) 53 (40.8)

CDC, centers for disease control; IQR, interquartile range; ALAT, alanine aminotransferase; IU, international unit; ART, antiretroviral
treatment.

© 2017 John Wiley & Sons Ltd 1189

Tropical Medicine and International Health volume 22 no 9 pp 1186–1195 september 2017

S. P. Eholi
e et al. Cotrimoxazole malaria prophylaxis for HIV patients

the end of the study, 290 (56.4%) patients were alive Table 2 Incidence of malaria according to country, cotrimoxa-
and not censored during the follow-up (CTX group zole group and antiretroviral treatment
559.6%, non-CTX group 49.7%), 13 (2.5%) were
Incidence
known to be dead (CTX group 2.9%, non-CTXgroup Person- rate per
1.8%), and 211 (41.1%) have had an event other than years 100 95%
malaria that led to censor their follow-up in the final of person- confidence
analysis (CTX group 37.5%, non-CTXgroup 48.5%). N n follow-up years interval
Reasons for follow-up being censored were as follows:
Total 514 42 485 8.7 6.3–11.5
CTX and/or ART initiation (n = 109), transfer to another Study sites
centre (n = 11), voluntary withdrawal from the study Abidjan (C^ote 157 14 127 11.0 6.2–17.8
(n = 9) or loss to follow-up (n = 82) (Table 1). d’Ivoire)
Bamako (Mali) 99 2 111 1.8 0.2–6.4
Bobo-Dioulasso 184 25 174 14.4 9.5–20.5
Incidence of malaria (Burkina Faso)
Dakar (Senegal) 74 1 73 1.4 0.1–7.4
Of the 154 episodes of fever, 50 (32.4%) were confirmed
Cotrimoxazole
as malaria episodes among 42 patients (CTX group, CTX group 347 17 324 5.2 3.1–8.3
n = 17; non-CTX group, n = 25). One episode that Non-CTX group 167 25 161 15.5 10.3–22.1
occurred in a pregnant woman in the non-CTX group ART initiated at baseline
was classified as severe. All other episodes were classified No 253 25 188 13.3 8.8–19.0
as uncomplicated. Plasmodium falciparum was the only Yes 261 17 297 5.7 3.4–9.0
species identified. The mean parasitaemia was 55724 per N, number of patients; n, number of patients with at least one
ll (standard deviation 141701) in the CTX group and malaria episode; ART, antiretroviral treatment.
15671 per ll (standard deviation 27163) in the non-CTX
group (P = 0.83). The median time between enrolment ‘on-treatment analysis’, CTX significantly reduced the
and first malaria episode was 12.1 months (IQR 4.0– risk of malaria by 57% (aHR 0.43, 95%CI 0.21–0.85)
17.8) overall, 12.0 months (IQR 3.0–17.9) in the CTX after adjusting for baseline CD4+ cell count, WHO clini-
group and 12.1 months (IQR 6.5–17.7) in the non-CTX cal stage, plasma HIV RNA and ART (data not shown).
group (P = 0.22). The incidence of malaria was 8.7/100
PY (95%CI 6.3–11.5) overall, with substantial variation
Evolution of CD4 cell count and plasma HIV RNA
across the four cities (Table 2).
during malaria episodes
The incidence of malaria was lower in the CTX group
(5.2/100 PY) than in the non-CTX group (15.5/100 PY) A decrease in CD4 cells/mm3 between the last available
(Incidence Rate ratio [IRR] 0.36, 95%CI 0.15–0.80, pre-malaria CD4 count (all within 3 months) and the M-
P < 0.01). The cumulative probability of remaining free Day-0 CD4 count was observed for both groups, with a
of malaria at month-18 was 88.1% overall (95%CI 83.3– median CD4 count decreasing from 389 to 254 CD4/
91.6); 93.3% in the CTX group (95%CI 88.3–96.2) vs. mm3 in the CTX group and from 499 to 315 CD4 cells/
77.2% (95%CI 65.1–85.6) in the non-CTX group (log- mm3 in the non-CTX group. Between M-Day-0 and M-
rank: P < 0.0001) (Figure 1a); and 93.0% in patients who Day-28 of malaria, the median CD4 cell count increased
started ART at enrolment (95%CI 88.0–96.0) vs. 79.2% from 254 to 402/mm3 in the CTX group and from 315
in those who did not start ART at enrolment (95%CI to 492/mm3 in the non-CTX group) (Figure 2a). A simi-
67.7–87.0) (log-rank: P = 0.02) (Figure 1b). lar trend between pre-malaria, M-Day-0 and M-Day-28
In multivariate analysis (Table 3), CTX significantly was observed for the total lymphocyte counts (data not
reduced the risk of malaria by 69% (adjusted Hazard shown). The median plasma HIV RNA tended to
Ratio [aHR] 0.31, 95%CI 0.105–0.90) after adjusting for decrease between pre-malaria and M-Day-0, and between
baseline CD4+ cell count, WHO clinical stage, plasma M-Day-0 and M-Day-28 (Figure 2b).
HIV RNA, ART and study sites. There was no indepen-
dent association between ART and the risk of malaria
Antimalarial treatment efficacy and tolerance
(aHR 0.85, 95%CI 0.44–4.15). When restricting the
analysis to patients residing in areas with high transmis- In the only case of severe malaria, intravenous quinine
sion of malaria (Abidjan and Bobo-Dioulasso), CTX was was administered. All other patients were given AS/AQ.
still associated with a 76% reduction of malaria (aHR In these AS/AQ-treated patients, the median time to fever
0.24, 95%CI 0.71–0.80). Also when we performed an clearance was 48 h (IQR 24–72), and the PCR-adjusted

1190 © 2017 John Wiley & Sons Ltd

Tropical Medicine and International Health volume 22 no 9 pp 1186–1195 september 2017

S. P. Eholi
e et al. Cotrimoxazole malaria prophylaxis for HIV patients

(a) Cotrimoxazole groups

(b) Antiretroviral treatment

Probability
Probability
1.0
CTX+ 1.0
ARV+
0.75 CTX– ARV–
0.75

0.50
0.50

0.25 0.25 Log-rank: P = 0.02

Log-rank: P < 0.0001

0 0
0 6 12 18 0 5 15 18
Time since enrollment (months) Time since enrollment (months)

Month-6 Month-12 Month-18 Month-6 Month-12 Month-18

CTX⊕ 99.0 (97.1 – 99.7) 97.4 (94.6 – 98.8) 93.3 (88.3 – 96.2) ARV+ 98.0 (95.4 – 99.2] 97.1 (94.0 – 98.6) 93.0 (88.0 – 96.0)
321 189 109 224 188 107
CTX – 93.0 (96.0 – 87.6) 88.3 (81.5 – 92.7) 77.2 (65.1 – 85.6) ARV– 94.6 (90.2 – 97.1) 90.7 (84.8 – 94.4) 79.2 (67.7 – 87.0)
137 88 45 154 98 47

Figure 1 Kaplan–Meier probability of remaining free of malaria, according to cotrimoxazole (a) and antiretroviral treatment (b) pre-
scription.

adequate clinical and parasitological response by Day 28
was 98.1% (95%CI 89.6–99.9). One case of late treat-
ment failure was observed at Day 21, but was reclassified
as reinfection after PCR adjustment. AS/AQ was well tol-
erated, with grade 3–4 neutropenia occurring in four
(9.5%) patients: three were in the CTX group and
received concomitant ART with zidovudine, and one was
in the non-CTX group and did not received zidovudine.
Discussion
Our data clearly suggest that CTX prophylaxis still has a
protective effect against malaria in West African HIV-
infected individuals, particularly for those living in stable
transmission area. As expected, patients in the CTX group
at baseline had an adjusted risk of malaria three times-
lower than those in the non-CTX group.
In the four West African countries, CTX was recom-
mended when CD4 fell below 350 cells/mm3. This was
based on the results of two CTX placebo-controlled trials
performed in C^ ote d’Ivoire at the end of the 1990s [30, 39,
40]. One of these trials clearly demonstrated that malaria
was one of the diseases prevented by CTX [39]. During the
2000s, many studies conducted in eastern and southern
Africa, confirmed that CTX had a strong preventive effect
on malaria in HIV-infected adults and children [19, 20,
22–24, 27, 29]. In two four-year cohorts of HIV-infected
adults in Uganda and Malawi, CTX prophylaxis reduced
malaria incidence from 51 to 9 episodes per 100 PY, with
an additional reduction to 3.5 and 2.1 per 100 PY when
ART and bednets were further introduced [41].
Having low-CD4 cell count number and living in an
area of unstable transmission have been previously
reported to be associated with malaria severity [11, 12,
42]. This differs from our findings, where most of the
episodes occurred in areas of stable transmission (Burkina
Faso and Cote d’Ivoire), and were observed in patients
with CD4 counts above 350 cells/mm3. Moreover we did
not observe severe malaria cases even in the cities with
unstable transmission.
Drops in CD4 cell counts during malaria episodes
have been previously reported in HIV-infected patients
[14, 15]. Our study confirms this phenomenon. The
most important conclusion to be drawn from it in the
field is that any change in CD4 counts in a context of
concomitant acute malaria should clearly not be inter-
preted as a sign of non-adherence or ART failure. In
our study, the concentration of plasma HIV RNA
decreased between baseline and malaria episode while
it increased in a study conducted in rural Malawi [15].
This difference could be explained by several factors.
Firstly, the median time between enrolment and first
malaria episode was higher in our study (12 months vs.
4 months). Thus, the baseline HIV RNA in our study
does not reflect the pre-malaria measurement. Secondly,
in our study, 40% of those who had at least one

© 2017 John Wiley & Sons Ltd 1191

Tropical Medicine and International Health volume 22 no 9 pp 1186–1195 september 2017

S. P. Eholi
e et al. Cotrimoxazole malaria prophylaxis for HIV patients

Table 3 Factors associated with the risk of malaria

Univariate analysis Multivariate analysis

No. of patients
with data Malaria HR 95% CI P aHR 95% CI P

Age 0.94
<29 115 9 (7.8) REF - -
29–34 120 11 (9.2) 1.32 0.54–3.22 0.54
35–42 157 12 (7.6) 0.95 0.42–2.74 0.88
43 122 10 (8.2) 1.12 0.45–2.81 0.80
Gender
Female 361 27 (7.5) REF - -
Male 153 15 (9.8) 1.53 0.81–2.94 0.19
BMI 0.69
<18 75 3 (4.0) REF - -
355 32 (9.0) 1.61 0.49–5.30 0.44
25 84 7 (8.3) 1.52 0.39–5.93 0.54
Education level
No or primary school 332 22 (6.6) REF - -
Secondary or more 182 20 (11.0) 1.60 0.88–2.96 0.13
Study sites 0.001 0.001
Abidjan (C^ote d’Ivoire) 157 14 (8.9) REF - -
Dakar (Senegal) 74 1 (1.35) 0.19 0.02–1.47 0.11 0.23 0.03–1.81 0.16
Bamako (Mali) 99 2 (2.0) 0.19 0.03–0.85 0.03 0.22 0.05–1.03 0.06
Bobo-Dioulasso (Burkina Faso) 184 25 (13.6) 1.93 0.94–3.93 0.07 2.89 1.32–6.33 0.01
WHO clinical stage
A or B 468 38 (8.1) REF - - REF - -
C 46 4 (8.7) 1.47 0.52–4.13 0.47 1.91 0.58–8.41 0.29
CD4 count (cells/mm3) 0.08
<200 206 10 (4.9) REF - - REF - -
200 308 32 (10.4) 1.99 0.98–4.06 0.06 1.70 0.67–4.30 0.26
Antiretroviral treatment
No 253 25 (9.9) REF - - REF - -
Yes 261 17 (6.5) 0.49 0.26–0.93 0.03 0.85 0.44–4.15 0.35
Use of cotrimoxazole
Non-CTX group 167 25 (15.0) REF - - REF - -
CTX group 347 17 (4.9) 0.40 0.21–0.75 0.004 0.31 0.10–0.90 0.03
Use of bednet
No 133 9 (6.8) REF - -
Yes 381 33 (8.7) 1.07 0.51–2.24 0.86
Plasma HIV-1 RNA (log10copies/ml)
<105 258 21 (8.1) REF - - REF - -
105 256 21 (8.2) 0.91 0.50–1.67 0.76 1.09 0.56–2.13 0.79

n, number of patients; N (%), number (percentage) of patients with at least one malaria episode; HR, hazard ratio; aHR, adjusted hazard ratio.

malaria episode were on ART while no patient was on
ART in the other study. Indeed, ART might diminish
the effect of malaria on HIV RNA concentrations.
After the malaria episode, the concentration of plasma
HIV-RNA decreased in our study as reported in the
previous studies [14, 15].
As expected, artesunate-amodiaquine combination ther-
apy was efficacious [43, 44]. Unlike in previous studies, we
did not observe any delay in clinical response and para-
sitaemia clearance. AS/AQ combination was well toler-
ated, even when patients received ART. The only grade 3
or 4 adverse events were neutropenia. They were rare and
transient. Severe neutropenia with AS/AQ was previously
reported as more frequent in HIV-infected children receiv-
ing ART as compared to HIV-negative ones [45]. We did
not observe severe liver toxicity, even among patients
exposed to NNRTI, contrary to previous report of
increased hepatotoxicity with concomitant use of amodi-
aquine with NNRTI [46].
Our study has some limitations. First, the sample size
was small, especially in the CTX group. However, the
association between CTX use and a lower risk of malaria

1192 © 2017 John Wiley & Sons Ltd

Tropical Medicine and International Health volume 22 no 9 pp 1186–1195 september 2017

S. P. Eholi
e et al. Cotrimoxazole malaria prophylaxis for HIV patients

(a) CD4+ cell count (b) plasma HIV RNA

8
800

6
600

Viral load log

CD4 count

4
400
200

2
0
0

Pre-malaria Malaria day-0 Malaria day-28 Pre-malaria Malaria day-0 Malaria day-28

CTX+ CTX– CTX+ CTX–

CD4 Pre-malaria Malaria day-0 Malaria day-28 Plasma HIV Pre-malaria Malaria day-0 Malaria day-28
count/mm3 RNA
CTX⊕ log10 copies/ml
Median 389 254 402 CTX⊕
IQR 209 – 510 153 – 389 193 – 620 Median 3.8 2.7 2.7
CTX – IQR 2.3 – 25.6 2.6 – 3.6 2.0 – 2.8
Median 499 315 492 CTX –
IQR 383 – 601 225 – 376 463 – 610 Median 4.7 3.7 3.1
IQR 2.6 – 5.4 2.6 – 4.1 2.2 – 3.9

Figure 2 Distribution of CD4+ cell count (a) and plasma HIV RNA (b) prior to, at the onset, and at the end of the episode of malarial
in the 42 patients who had an episode of malaria.

was strong. Second, only three cases of malaria were
observed in the area of unstable transmission. This did
not allow us to assess CTX efficacy by transmission area.
Third, we compared two groups with very different base-
line characteristics. However, patients who received CTX
had more advanced HIV-disease, suggesting that the study
might have been biased towards an underestimation of the
protective effect of CTX against malaria, rather than an
overestimation. Fourth, the follow-up duration only
allows us to draw medium term conclusions. Fifth, we
could not document the prevalence of resistance – confer-
ring Plasmodium falciparum mutations. These mutations,
in case of presence, could have the effect of underestimat-
ing the efficacy of CTX. Despite these limitations, this
study has two main strenghts. First, it took place in four
West African countries with different transmission pat-
terns of malaria. Second, few patients who did not start
CTX at baseline started it afterwards. This allowed us to
compare the incidence of malaria between patients who
started CTX prophylaxis at enrolment and those who
were followed-up with no CTX prophylaxis.
Conclusion
In conclusion, this study confirms the impact of CTX
prophylaxis on malaria in HIV-infected adults with
varying degrees of immunosuppression in West Africa
in the ART era. It provides evidence that CTX prophy-
laxis reduces the incidence of malaria in this region.
Ours results emphasise the latest WHO guidelines to
prescribe CTX regardless of CD4 cell count in settings
with high prevalence of malaria and bacterial infections
[31].
Acknowledgements
We are grateful to all study participants for their cooper-
ation. We thank all members of the study team, including
Zelica Diallo, Cl
ement Adj
e, Hamsatou Ciss
e, Ibrahima
Ba, Mohamed Haidara for the for tireless work of data
entry and verification; Adou-Bryn Dao, Djimd
e Abdou-
laye, Amagana Dolo for the parasitological assessments;
Colette Niamekeh for administrative support; Maryam
Kassambara-Sow, Amadou Diarra, John Damonti for
Secure the Future. The MALHIV STUDY received finan-
cial support from the Programme Secure the Future of
the Bristol Myers Squibb Foundation. The sponsor of the
study had no role in the study design, data collection,
data analysis, data interpretation or writing of the report.
The corresponding author had full access to all study and
had final responsibility for the decision to submit for
publication.

© 2017 John Wiley & Sons Ltd 1193

Tropical Medicine and International Health
S. P. Eholi
e et al. Cotrimoxazole malaria prophylaxis for HIV patients
References
1. WHO: World malaria report 2015. Available from: http://
www.who.int/malaria/publications/world-malaria-report-
2015/report/en/. [9 December 2016].
2. WHO: WHO. HIV/AIDS, Fact sheet. Updated November
2016. Available from: http://www.who.int/mediacentre/fact
sheets/fs360/en/. [9 December 2016].
3. Abu-Raddad LJ, Patnaik P, Kublin JG. Dual infection with
HIV and malaria fuels the spread of both diseases in sub-
Saharan Africa. Science 2006: 314: 1603–1606.
4. Korenromp EL, Williams BG, de Vlas SJ et al. Malaria attri-
butable to the HIV-1 epidemic, sub-Saharan Africa. Emerg
Infect Dis 2005: 11: 1410–1419.
5. Cuadros DF, Branscum AJ, Crowley PH. HIV-malaria co-
infection: effects of malaria on the prevalence of HIV in East
sub-Saharan Africa. Int J Epidemiol 2011: 40: 931–939.
6. Hay SI, Smith DL, Snow RW. Measuring malaria endemic-
ity from intense to interrupted transmission. Lancet Infect
Dis 2008: 8: 369–378.
7. Bouyou-Akotet MK, Offouga CL, Mawili-Mboumba DP,
Essola L, Madoungou B, Kombila M. Falciparum malaria as
an emerging cause of fever in adults living in Gabon, Cen-
tral Africa. Biomed Res Int 2014: 2014: 351281.
8. Cotter C, Sturrock HJ, Hsiang MS et al. The changing epi-
demiology of malaria elimination: new strategies for new
challenges. Lancet 2013: 382: 900–911.
9. Mayor A, Aponte JJ, Fogg C et al. The epidemiology of
malaria in adults in a rural area of southern Mozambique.
Malar J 2007: 6: 3.
10. Alemu A, Tsegaye W, Golassa L, Abebe G. Urban malaria
and associated risk factors in Jimma town, south-west Ethio-
pia. Malar J 2011: 10: 173.
11. Cohen C, Karstaedt A, Frean J et al. Increased prevalence of
severe malaria in HIV-infected adults in South Africa. Clin
Infect Dis 2005: 41: 1631–1637.
12. Grimwade K, French N, Mbatha DD, Zungu DD, Dedicoat
M, Gilks CF. HIV infection as a cofactor for severe falci-
parum malaria in adults living in a region of unstable
malaria transmission in South Africa. AIDS 2004: 18: 547–
554.
13. Patnaik P, Jere CS, Miller WC et al. Effects of HIV-1
serostatus, HIV-1 RNA concentration, and CD4 cell count
on the incidence of malaria infection in a cohort of adults in
rural Malawi. J Infect Dis 2005: 192: 984–991.
14. Hoffman IF, Jere CS, Taylor TE et al. The effect of Plas-
modium falciparum malaria on HIV-1 RNA blood plasma
concentration. AIDS 1999: 13: 487–494.
15. Kublin JG, Patnaik P, Jere CS et al. Effect of Plasmodium
falciparum malaria on concentration of HIV-1-RNA in the
blood of adults in rural Malawi: a prospective cohort study.
Lancet 2005: 365: 233–240.
16. Van Geertruyden JP, Mulenga M, Kasongo W et al. CD4 T-
cell count and HIV-1 infection in adults with uncomplicated
malaria. J Acquir Immune Defic Syndr 2006: 43: 363–367.
17. Kamya MR, Byakika-Kibwika P, Gasasira AF et al. The
effect of HIV on malaria in the context of the current
1194
volume 22 no 9 pp 1186–1195 september 2017
standard of care for HIV-infected populations in Africa.
Future Virol 2012: 7: 699–708.
18. Anglaret X, Chene G, Attia A et al. Early chemoprophylaxis
with trimethoprim-sulphamethoxazole for HIV-1-infected
adults in Abidjan, Cote d’Ivoire: a randomised trial. Cot-
rimo-CI Study Group. Lancet 1999: 353: 1463–1468.
19. Bwakura-Dangarembizi M, Kendall L, Bakeera-Kitaka S et al.
A randomized trial of prolonged co-trimoxazole in HIV-
infected children in Africa. N Engl J Med 2014: 370: 41–53.
20. Campbell JD, Moore D, Degerman R et al. HIV-infected
ugandan adults taking antiretroviral therapy with CD4
counts >200 cells/muL who discontinue cotrimoxazole pro-
phylaxis have increased risk of malaria and diarrhea. Clin
Infect Dis 2012: 54: 1204–1211.
21. Denoeud-Ndam L, Zannou DM, Fourcade C et al. Cotri-
moxazole prophylaxis versus mefloquine intermittent preven-
tive treatment to prevent malaria in HIV-infected pregnant
women: two randomized controlled trials. J Acquir Immune
Defic Syndr 2014: 65: 198–206.
22. Gasasira AF, Kamya MR, Ochong EO et al. Effect of
trimethoprim-sulphamethoxazole on the risk of malaria in
HIV-infected Ugandan children living in an area of wide-
spread antifolate resistance. Malar J 2010: 9: 177.
23. Kamya MR, Gasasira AF, Achan J et al. Effects of trimetho-
prim-sulfamethoxazole and insecticide-treated bednets on
malaria among HIV-infected Ugandan children. AIDS 2007:
21: 2059–2066.
24. Kapito-Tembo A, Meshnick SR, van Hensbroek MB, Phiri
K, Fitzgerald M, Mwapasa V. Marked reduction in preva-
lence of malaria parasitemia and anemia in HIV-infected
pregnant women taking cotrimoxazole with or without sul-
fadoxine-pyrimethamine intermittent preventive therapy dur-
ing pregnancy in Malawi. J Infect Dis 2011: 203: 464–472.
25. Klement E, Pitche P, Kendjo E et al. Effectiveness of co-tri-
moxazole to prevent Plasmodium falciparum malaria in
HIV-positive pregnant women in sub-Saharan Africa: an
open-label, randomized controlled trial. Clin Infect Dis
2014: 58: 651–659.
26. Mermin J, Ekwaru JP, Liechty CA et al. Effect of co-trimox-
azole prophylaxis, antiretroviral therapy, and insecticide-
treated bednets on the frequency of malaria in HIV-1-
infected adults in Uganda: a prospective cohort study. Lan-
cet 2006: 367: 1256–1261.
27. Mermin J, Lule J, Ekwaru JP et al. Effect of co-trimoxazole
prophylaxis on morbidity, mortality, CD4-cell count, and
viral load in HIV infection in rural Uganda. Lancet 2004:
364: 1428–1434.
28. Suthar AB, Granich R, Mermin J, Van Rie A. Effect of cotri-
moxazole on mortality in HIV-infected adults on antiretrovi-
ral therapy: a systematic review and meta-analysis. Bull
World Health Organ 2012: 90: 128C–138C.
29. Walker AS, Ford D, Gilks CF et al. Daily co-trimoxazole
prophylaxis in severely immunosuppressed HIV-infected
adults in Africa started on combination antiretroviral ther-
apy: an observational analysis of the DART cohort. Lancet
2010: 375: 1278–1286.
© 2017 John Wiley & Sons Ltd
Tropical Medicine and International Health
S. P. Eholi
e et al. Cotrimoxazole malaria prophylaxis for HIV patients
30. WHO: Guidelines on co-trimoxazole prophylaxis for HIV-
related infections among children, adolescents and adults.
Recommendations for a public health approach. August
2006. Available from: http://www.who.int/hiv/pub/plhiv/ctx/
en/. [9 December 2016].
31. WHO: Guidelines on Post-Exposure Prophylaxis for HIV
and the Use of Co-Trimoxazole Prophylaxis for HIV-Related
Infections Among Adults, Adolescents and Children: Recom-
mendations for a Public Health Approach: December 2014
supplement to the 2013 consolidated guidelines on the use
of antiretroviral drugs for treating and preventing HIV infec-
tion. Available from: http://www.ncbi.nlm.nih.gov/pubmed/
26042326. [9 December 2016].
32. Date AA, Vitoria M, Granich R, Banda M, Fox MY, Gilks
C. Implementation of co-trimoxazole prophylaxis and isoni-
azid preventive therapy for people living with HIV. Bull
World Health Organ 2010: 88: 253–259.
33. Iyer JK, Milhous WK, Cortese JF, Kublin JG, Plowe CV.
Plasmodium falciparum cross-resistance between trimetho-
prim and pyrimethamine. Lancet 2001: 358: 1066–1067.
34. WHO: Toxicity grading Scale for determining the severity of
adverse events. Available from: http://www.icssc.org/Docu
ments/Resources/AEManual2003AppendicesFebruary_06_
2003%20final.pdf. [9 December 2016].
35. WHO: Antiretroviral therapy for HIV infection in adults
and adolescents. Recommendations for a public health
approach (2006 revision). Available from: http://www.who.
int/hiv/pub/arv/adult/en/. [9 December 2016].
36. WHO: Guidelines for the tretament of malaria. Third Edi-
tion 2015. Available from: http://apps.who.int/iris/bitstream/
10665/162441/1/9789241549127_eng.pdf?ua=1. [9 Decem-
ber 2016].
37. WHO: Methods for surveillance of antimalarial drug effi-
cacy. Geneva 2009. Available from: http://www.who.int/ma
laria/publications/atoz/9789241597531/en/. [9 December
2016].
Corresponding Author Serge P. Eholi
e, Service des Maladies Infectieuses et Tropicales, Centre Hospitalier Universitaire de Tre-
ichville, Abidjan, C^
ote d’Ivoire, BP V3, Abidjan, C^
ote d’Ivoire. Tel.: 00 225 08 64 10 80; E-mails: speholie@afnet.net and serge-
holie@yahoo.fr
© 2017 John Wiley & Sons Ltd
volume 22 no 9 pp 1186–1195 september 2017
38. Singh B, Bobogare A, Cox-Singh J, Snounou G, Abdullah
MS, Rahman HA. A genus- and species-specific nested poly-
merase chain reaction malaria detection assay for epidemio-
logic studies. Am J Trop Med Hyg 1999: 60: 687–692.
39. Anglaret X, Toure S, Ouassa T, Dabis F, N’Dri-Yoman T.
Thresholds of CD4 cells for initiating trimethoprim-sulfa-
methoxazole prophylaxis in West Africa. AIDS 2000: 14:
2628–2629.
40. Wiktor SZ, Sassan-Morokro M, Grant AD et al. Efficacy of
trimethoprim-sulphamethoxazole prophylaxis to decrease
morbidity and mortality in HIV-1-infected patients with
tuberculosis in Abidjan, Cote d’Ivoire: a randomised con-
trolled trial. Lancet 1999: 353: 1469–1475.
41. Kasirye R, Baisley K, Munderi P, Grosskurth H. Effect of
cotrimoxazole prophylaxis on malaria occurrence in HIV-
infected patients on antiretroviral therapy in sub-Saharan
Africa. Trop Med Int Health 2015: 20: 569–580.
42. Noor AM, Kinyoki DK, Mundia CW et al. The changing
risk of Plasmodium falciparum malaria infection in Africa:
2000-10: a spatial and temporal analysis of transmission
intensity. Lancet 2014: 383: 1739–1747.
43. Birku Y, Mekonnen E, Bjorkman A, Wolday D. Delayed
clearance of Plasmodium falciparum in patients with human
immunodeficiency virus co-infection treated with artemisi-
nin. Ethiop Med J 2002: 40(Suppl 1): 17–26.
44. Kamya MR, Gasasira AF, Yeka A et al. Effect of HIV-1
infection on antimalarial treatment outcomes in Uganda: a
population-based study. J Infect Dis 2006: 193: 9–15.
45. Gasasira AF, Kamya MR, Achan J et al. High risk of neu-
tropenia in HIV-infected children following treatment with
artesunate plus amodiaquine for uncomplicated malaria in
Uganda. Clin Infect Dis 2008: 46: 985–991.
46. German P, Greenhouse B, Coates C et al. Hepatotoxicity
due to a drug interaction between amodiaquine plus arte-
sunate and efavirenz. Clin Infect Dis 2007: 44: 889–891.
1195