EDITORIAL REVIEW

The role of coinfections in HIV epidemic trajectory

and positive prevention: a systematic review and
meta-analysis
Ruanne V. Barnabasa,c,M, Emily L. Webbb,M, Helen A. Weissb
and Judith N. Wasserheita,c
Objectives: Recurrent or persistent coinfections may increase HIV viral load and,
consequently, risk of HIV transmission, thus increasing HIV incidence. We evaluated
the association between malaria, herpes simplex virus type 2 (HSV-2) and tuberculosis
(TB) coinfections and their treatment on HIV viral load.
Design: Systematic review and meta-analysis of the association of malaria, HSV-2 and
TB coinfections and their treatment with HIV viral load.
Methods: PubMed and Embase databases were searched to 10 February 2010 for
studies in adults that reported HIV plasma and/or genital viral load by coinfection status
or treatment. Meta-analyses were conducted using random-effects models.
Results: Forty-five eligible articles were identified (six malaria, 20 HSV-2 and 19 TB).
There was strong evidence of increased HIV viral load with acute malaria [0.67 log10
copies/ml, 95% confidence interval (CI) 0.15–1.19] and decreased viral load following
treatment (
0.37 log10 copies/ml, 95% CI
0.70 to
0.04). Similarly, HSV-2 infection
was associated with increased HIV viral load (0.18 log10 copies/ml, 95% CI 0.01–0.34),
which decreased with HSV suppressive therapy (
0.28 log10 copies/ml, 95% CI
0.36
to
0.19). Active TB was associated with increased HIV viral load (0.40 log10 copies/ml,
95% CI 0.13–0.67), but there was no association between TB treatment and viral load
reduction (log10 copies/ml
0.02, 95% CI
0.19 to 0.15).
Conclusion: Coinfections may increase HIV viral load in populations where they are
prevalent, thereby facilitating HIV transmission. These effects may be reversed with
treatment. However, to limit HIV trajectory and optimize positive prevention for HIV-
infected individuals pre-antiretroviral therapy, we must better understand the mech-
anisms responsible for augmented viral load and the magnitude of viral load reduction
required, and retune treatment regimens accordingly.
ß 2011 Wolters Kluwer Health | Lippincott Williams & Wilkins

AIDS 2011, 25:1559–1573

Keywords: herpes simplex virus-2, HIV, HIV coinfections, malaria, tuberculosis,

viral load

a
Departments of Global Health and Medicine, University of Washington, Seattle, Washington, USA., bMedical Research Council
Tropical Epidemiology Group, Faculty of Epidemiology and Population Health, London School of Hygiene & Tropical Medicine,
London, UK, and cVaccine and Infectious Disease Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, USA.
Correspondence to Ruanne V. Barnabas, UW Box 359927, 325 Ninth Ave., Seattle, WA 98104, USA.
E-mail: [email protected]

Ruanne V. Barnabas and Emily L. Webb contributed equally to the writing of this article.
Received: 27 November 2010; revised: 28 April 2011; accepted: 24 May 2011.

DOI:10.1097/QAD.0b013e3283491e3e

ISSN 0269-9370 Q 2011 Wolters Kluwer Health | Lippincott Williams & Wilkins 1559
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1560 AIDS 2011, Vol 25 No 13

Background (coinfection rates for TB and HSV-2 are 33% [7] and
85% [8], respectively, and HIV increases the frequency
The HIV pandemic has carved starkly different trajec- and severity of acute malaria [9,10]). Further, effective
tories around the globe in the last three decades. In 2009, treatment options are available. The impact of helminth
33.3 million people were living with HIV and coinfections on HIV viral load was recently reviewed
approximately 2.6 million people were newly infected elsewhere [11]. Finally, we discuss opportunities to
[1]. Sub-Saharan Africa is the most severely affected enhance HIV prevention and care by treatment of these
region, home to two-thirds of those living with HIV (an coinfections.
estimated HIV prevalence in adults of 5%) and almost
70% of new infections. In contrast, about 12% of HIV-
infected people live in south and south-east Asia, whereas
just less than half this number live in North America (5%) Methods
and Latin America (4%), with smaller proportions in
western and central Europe (2%), representing estimated We followed Cochrane Collaboration guidelines in
HIV prevalences in adults of 0.3–0.6% [1]. conducting our review [12] and Preferred Reporting
Items for Systematic Reviews and Meta-Analyses
The differential spread of HIV between populations (PRISMA) guidelines in reporting results [13].
results from multiple factors which drive transmission
[2]. One key determinant of HIV transmission is HIV Criteria for considering studies for this review
viral load [3,4]. A growing body of evidence suggests that The a priori criteria for considering studies for the review
recurrent or persistent coinfections, such as malaria, are tabulated in the appendix (Table S1, http://links.lww.
tuberculosis (TB), herpes simplex virus type 2 (HSV-2) com/QAD/A149). Both observational studies and
infection and helminths, may increase HIV viral load randomized controlled trials (RCTs) were eligible. To
and thus facilitate HIV transmission at both individual ensure comparability between groups in observational
and population levels by increasing the person-time at studies, we searched for studies that controlled for key
viral load levels associated with risk of ongoing spread. confounders of viral load, including time from infection
In areas where these coinfections are common, they or CD4 cell count. We excluded the following from
may increase viral load during the protracted interval analyses: studies in which all participants were on ART,
between acute HIV infection and AIDS or initiation of were pregnant women, children or HIV-2-infected
antiretroviral therapy (ART), and the duration at this individuals; studies in which the intervention modified
elevated level, driving HIV transmission. Infections HIV viral load with and without coinfection; and studies
may also increase susceptibility to HIV, as seen with in which the control group was not proven negative for
the association between HSV-2 infection, and HIV the coinfection. For studies in which a subgroup of
acquisition [5]. participants was on ART, pregnant, aged less than 16 or
HIV-2 positive, results were extracted excluding these
Although HIV prevention strategies initially focused participants. The only exception was episodic HSV-2
primarily on HIV-negative individuals, the advent of therapy for which three of the four trials had small
ART has radically reframed the opportunities for and numbers on ART (<4% of all participants) and it was not
potential efficiencies of prevention in HIV positives possible to extract data on ART-naive participants only.
[6]. The growing emphasis on ‘positive prevention’
recognizes that combinations of interventions targeted Search strategy for identification of studies
at HIV-infected people could simultaneously improve Electronic searches of PubMed and EMBASE databases
survival and reduce transmission. Indeed, interventions were conducted on 31 January 2009 and updated on 10
such as ART that reduce HIV viral load are likely February 2010. In PubMed, the following Medical
not only to delay HIV progression but also to limit Subject Heading (MeSH) search terms were used:
secondary infections. In individuals diagnosed with HIV ‘HIV infections’ and ‘malaria/HSV-2/TB’ and ‘adult’.
who are not yet eligible for ART, treatment of In Embase, the following search terms were used: (HIV
coinfections may offer an important alternative approach infection and malaria/HSV-2/TB and adult). The
to reducing HIV viral load and thereby slow disease searches were done separately for each coinfection,
progress, delay ART initiation and decrease HIV included all languages, and were limited to human studies.
transmission. Because the search for TB yielded over 6000 abstracts,
many of which reported on clinical management, the
We therefore explore the potential role of three following additional filters were used independently:
persistent or recurrent infections (malaria, HSV-2 clinical trial, viral load or viral shedding, and disease
and TB) on HIV transmission and acquisition by susceptibility. Reference lists in articles were hand
systematically reviewing the literature. These three searched as were infectious disease conference abstract
coinfections were selected because they are common books. Finally, correspondence with authors yielded one
among HIV-infected persons in sub-Saharan Africa PhD thesis [14] and two in press articles [15,16].

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 HIV coinfections and viral load Barnabas et al. 1561

Selection of studies, data extraction and statistic [28], defined as the percentage of total variation in
synthesis the study estimates that is due to between-study
Abstracts were reviewed and full-text articles of heterogeneity. For example an I2 value of 50% would
potentially relevant studies were examined independently mean that half of the total variability among effect sizes is
by two authors (R.V.B. and J.N.W. for the initial search; caused not by sampling error, but by true heterogeneity
E.L.W. and H.A.W. for the updated search) against between studies. Funnel plots [29] were visually examined
prespecified selection criteria (Table S1, appendix, and Egger tests [30] were conducted to assess the
http://links.lww.com/QAD/A149). Data were extracted possibility of publication bias. All statistical analyses were
independently by R.V.B., J.N.W. and E.L.W. for the conducted in STATA version 11 (StataCorp, College
original search, and by E.L.W. and H.A.W. for the Station, Texas, USA).
updated search, using a data extraction form. Discre-
pancies were discussed and consensus reached.

When data from the same individuals were reported in Results

multiple publications, we used the more informative
publication. When multiple timepoints were reported, We identified 351 potentially relevant abstracts for
we extracted results based on all timepoints provided; malaria, 380 for HSV-2 and 812 for TB (Fig. 1). Of
results based on repeated measures analyses were used these, 45 studies met the prespecified selection criteria for
only if the authors reported no evidence of a change in the systematic review (six malaria, 20 HSV-2 and 19 TB).
treatment effect over time, otherwise we report data for Forty of these studies were included in the meta-analysis
the timepoint most compatible with other studies for that (six malaria, 19 HSV-2 and 15 TB). Of the five studies not
disease. For TB, this was the earliest timepoint after included in the meta-analysis, the results of three are
conclusion of treatment or the latest timepoint during reported in the tables: one HSV-2 study that reported
treatment (if no data were collected posttreatment). For only GVL [31] and two studies of latent TB [18,31]. For
four studies [17–20], we calculated mean differences the remaining two studies it was not possible to
directly from the raw data provided. Unpublished data distinguish between the effects of TB infection and
were requested from authors of eight articles [17,21–27], treatment [22,33] and therefore these studies were not
primarily to obtain 95% confidence intervals (CIs) for included in the tables.
the mean difference. Three provided the requested
data [17,24,25]. We estimated CIs based on standard We identified no studies of the impact of HSV-2 on HIV
deviations reported or displayed in the articles for the acquisition that were published since the last systematic
remainder. review of this topic [5]. No studies were identified
examining the effect of malaria or TB on HIVacquisition.
The methodological quality of included studies was
reviewed by R.V.B. and E.L.W. based on the selection Malaria
criteria listed above. Studies were rated at low; moderate; All six eligible malaria studies were observational. Four
or high risk of bias. A key factor in this assessment was were prospective cohort studies [14,25,34,35], one was a
how studies attempted to account for time since HIV nested case–control study [36] and one was a challenge
seroconversion, because viral load is dynamic and changes study with Plasmodium vivax [20]. The sample sizes
over time. Studies that reported HIV viral load stratified ranged from 10 to 89 (Table 1). All studies except one
by CD4 cell count and, in addition, adjusted for other from Guangzhou, China, were conducted in Africa.
viral load cofactors such as sex were generally rated as ‘A.’ Recruitment in each study was facility based, through
In contrast studies that reported and compared the mean hospitals, clinics, HIV screening or AIDS support
CD4 cell counts of each group and did not adjust for organizations. All participants were HIV-infected adults
other cofactors were rated as ‘B.’ Studies that did not with or without malaria, followed through the course of
adjust for CD4 cell count were rated as ‘C.’ malaria infection and subsequent treatment. Malaria was
diagnosed through standard tests and treated appropri-
Statistical methods ately. One study [35] used four definitions of malaria, and
Plasma viral load (PVL) and genital viral load (GVL) were in this study we analyzed data for those with parasitemia
considered as separate outcomes. HIV PVL is reported and fever to be consistent with the other studies.
as log10 copies/ml and HIV GVL as log10 copies/ml
or/swab. The measure of effect was the mean difference All four studies that assessed the association between
(and 95% CI) by coinfection or treatment status malaria coinfection and HIV viral load (Table 1a) found
respectively, adjusted for potential confounders. increased PVL in the presence of acute malaria
[14,20,34,35], with a summary estimate of the PVL
Random-effects meta-analyses were performed separ- increase of 0.67 log10 copies/ml (95% CI 0.15–1.19)
ately for each coinfection and for each coinfection (Fig. 2a). All estimates showed an increase in PVL, but
treatment. Heterogeneity was quantified using the I2 there was substantial heterogeneity in the magnitude of

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1562 AIDS 2011, Vol 25 No 13

Malaria HSV-2 TB

335 unique 16 records 364 unique 16 records 804 unique 8 records

records identified identified through records identified identified through records identified identified through
through databases other sources through databases other sources through databases other sources

201 records 221 records 604 records

351 abstracts screened 380 abstracts screened 812 abstracts screened
excluded excluded excluded

144 full-text 139 full-text 189 full-text

150 full-text articles 159 full-text articles 208 full-text articles
articles articles articles
assessed for eligibility assessed for eligibility assessed for eligibility
excluded1 excluded2 excluded3

6 studies included in 20 studies included in 19 studies included in

systematic review systematic review systematic review

6 studies included in 19 studies included in 15 studies included in

meta-analysis meta-analysis meta-analysis

1
Malaria: pregnancy (55), no VL data (51), review/comment (16), subjects on ART/malaria prophylaxis (4), genetics (4), malaria and mortality/disease progression
(not VL) (4), biological mechanism of interaction (3), mathematical model (3), immunology (2), children (1), no data on CD4 count (1)

2
HSV-2: no VL data (56), review/comment (25), identified in previous review of HIV acquisition (20), pregnancy (10), HSV-2 is outcome (7), biological mechanism of
interaction (5), mathematical model (5), HIV acquisition but with no adjustment for other factor (3), subjects on ART/other intervention (3), HSV-2 and
mortality/disease progression (not VL) (3), baseline data for RCT (1), not HSV-2 (1)

3
TB: no VL data (63), review/comment (43), risk of TB acquisition (17), subjects on ART/other treatment (16), biological mechanism of interaction (14), genetics
(9), no data on CD4 count (9), immunology (8), TB and mortality/disease progression (not VL) (4), pregnancy (3), mathematical model (2), control group had
suspected TB (1).

Fig. 1. Systematic review flowchart.

effect (I2 ¼ 94.7%), with estimates from 0.13 log10 copies/ the other studies, including number of bites, inoculum
ml (95% CI
0.06 to 0.32) to 1.23 log10 copies/ml (95% size and protection offered by malaria prevention
CI 1.01–1.45). interventions.

Six studies assessed the impact of malaria treatment on Herpes simplex virus type 2
PVL (Table 1b). The impact of malaria treatment was We identified 20 studies of the effect of HSV-2 infection
measured as the mean difference in PVL before and after and treatment and HIV viral load. Of these, eight assessed
malaria treatment. Of these, five found that malaria the association between HSV-2 infection and PVL
treatment was associated with decreased PVL [14,20,34– (Table 2a); three were cohort studies [21,37,38], four
36], with estimates from
0.89 log10 copies/ml (95% CI were cross-sectional [39–42] and one was a nested case–

1.03 to
0.75) to
0.10 (95% CI
0.31 to 0.11). The control study [43]. Cross-sectional and case–control
remaining study, which showed an increase in PVL [25], studies compared PVL among participants with or
had a follow-up period of 9 days compared with a without HSV-2, and cohort studies measured PVL
minimum of 28 days for the other studies. The summary through the course of a HSV-2 clinical episode. Three
estimate showed a trend towards a decrease in PVL of articles reported results from separate substudies accord-

0.25 log10 copies/ml (95% CI
0.59 to 0.10) associated ing to stage of HSV-2 and HIV-1 infections [21,42] or sex
with treatment (Fig. 2b), although there was substantial [39] (Table 2a). One of the two cohorts analyzed in Gray
heterogeneity between studies (I2 ¼ 92.3%). Excluding et al. [40] was reported on more fully in another article
the study with short follow-up [25] the summary estimate [43]; only the latter, more informative data were included
was a mean decrease of
0.37 log10 copies/ml (95% CI in the meta-analysis. The effect of incident HSV-2

0.70 to
0.04) with I2 ¼ 92.5%. Further, when the infection (i.e. recent seroconversion) on PVL was
challenge study was excluded [20] there was little reported in two articles, neither of which found any
evidence of heterogeneity in the malaria treatment association [21,37]. Eight studies examined the associ-
studies; I2 ¼ 20.3%. This decrease in heterogeneity ation between prevalent HSV-2 infection and HIV PVL.
may be explained by differences between the challenge The summary estimate revealed a mean increase in PVL
study design compared with naturally acquired malaria in of 0.20 log10 copies/ml (95% CI 0.00–0.41). Pooling

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Table 1. Association between malaria and HIV-1 viral load.
Mean
Study difference
Study Location Period Study design Endpoint/follow-up Population Malaria diagnosis size (95% CI) Viral load assay

(a) Malaria infection and HIV-1 plasma viral load

Chen Guangzhou, No details Challenge Acute malaria HIV-1 positive individuals P. vivax þ 10 10 1.23 bDNA
et al. [20] China (10 febrile infected with malaria febrile episodes (1.01–1.45)
episodes)
Hoffman Malawi 1997 Cross-sectionala n/a Hospital-based HIV-1 P. falciparum þ 89 0.83 NASBA
et al. [34] positive individuals with malaria symptoms (0.44–1.22)
or without malaria
Kublin Malawi 2000–1 Cohort Acute malaria after HIV-1/malaria coinfected P. falciparum þ fever 36 0.42 RT-PCR
et al. [35] median of 112 individuals recruited (0.24–0.60)
malaria-free days through HIV screening
French [14] Uganda Cohort No details Hospital-based HIV-1/malaria P. falciparum þ fever 10 0.13 RT-PCR
coinfected individuals (
0.06 to 0.32)

Mean
Malaria Study difference Viral load
Study Location Period Study design Endpoint/follow-up Population diagnosis Treatment size (95% CI) assay

(b) Malaria treatment and HIV-1 plasma viral load

Chen Guangzhou, No details Challenge 1 month after HIV-1 positive individuals P. vivax þ Chloroquine 10
0.89 (
1.03 bDNA
et al. [20] China treatment infected with malaria 10 febrile episodes to
0.75)
Hoffman Malawi 1997 Cohort 28 days Hospital-based HIV-1/malaria P. falciparum þ malaria SP 27
0.25 (
0.50 MASBA
et al. [34] post-treatment coinfected individuals symptoms to 0.00)
Kublin Malawi 2000–1 Cohort Mean 65 days HIV-1/malaria coinfected P. falciparum þ fever SP 36
0.37 (
0.55 RT-PCR
et al. [35] after treatment individuals recruited to
0.19)
through HIV screening
Tatfeng Benin City, 2004–2005 Cohort 9 days after HIV-1/malaria coinfected P. falciparum þ fever Dihydroartemisinin 144 1.12 (0.19–2.05) RT-PCR
et al. [25] Nigeria treatment individuals, recruited
from clinics
Van Ndola, 2004–5 Nested 28 days HIV-1/malaria coinfected P. falciparum >1000 SP or AL 68
0.10 (
0.31 RT-PCR
Geertruyden Zambia case– post-treatment individuals, recruited parasites/ml þ fever to 0.11)
et al. [36] control from clinics
French [14] Uganda Cohort No details Hospital-based HIV-1/malaria P. falciparum þ fever SP 10
0.20 (
0.51 RT-PCR
coinfected individuals to 0.11)

AL, artemether-lumefantrine; CI, confidence interval; NASBA, nucleic-acid sequence-based-analysis; P. vivax, Plasmodium vivax; RT, real time; SP, sulfadoxine/pyrimethamine.
a
Baseline measurements from cohort study.
HIV coinfections and viral load Barnabas et al.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1563
1564 AIDS 2011, Vol 25 No 13

(a) Malaria infection (b) Malaria treatment

Study Year Difference (95% CI) Study Year Difference (95% CI)

Hoffman 1999 -0.25 (-0.50, 0.00)

Hoffman 1999 0.83 (0.44, 1.22)
Chen 2003 -0.89 (-1.03, -0.75)
Chen 2003 1.23 (1.01, 1.45) French 2004 -0.20 (-0.51, 0.11)

French 2004 0.13 (-0.06, 0.32) Kublin 2005 -0.37 (-0.55, 0.19)

Van Geertruyden 2006 -0.10 (-0.31, 0.11)

Kublin 2005 0.51 (0.29, 0.73)
Tatfeng 2007 1.12 (0.19, 2.05)
Overall (I-squared = 94.7%) 0.67 (0.15, 1.19)
Overall (I-squared = 92.3%) -0.25 (-0.59, 0.10)

0 0.5 1 -1 -0.5 0 0.5 1

(c) HSV-2 infection (d) Suppressive HSV-2 treatment

Study Year Difference (95% CI)

Study Year Difference (95% CI)

Mole 1997 0.75 (0.30, 1.19)

Gray 2004 -0.09 (-0.28, 0.10) Schacker 2002 -0.28 (-0.54, -0.02)
Serwadda 2004 0.55 (0.18, 0.92) Nagot 2007 -0.53 (-0.72, -0.35)
Duffus 2005 0.30 (0.07, 0.53) Zuckerman 2007 -0.33 (-0.42, -0.23)
Chu 2006 0.10 (-0.40, 0.60)
Baeten 2008 -0.26 (-0.33, -0.19)
Chu 2006 0.50 (0.00, 1.00)
Dunne 2008 -0.43 (-0.56, -0.29)
Barbour 2007 0.10 (-0.19, 0.39)
Barbour 2007 0.10 (-0.27, 0.47) Delany 2009 -0.27 (-0.41, -0.13)
Cachay 2007 -0.17 (-0.73, 0.44) Celum 2010 -0.25 (-0.29, -0.22)
Cachay 2007 -0.12 (-0.45, 0.20)
Tanton 2010 -0.02 (-0.09, 0.13)
Cachay 2008 -0.08 (-0.30, 0.46)
Overall (I-squared = 83.1%) -0.28 (-0.36, -0.19)
Overall (I-squared = 60.3%) 0.18 (0.01, 0.34)

-1 -0.5 0 0.5 1 -1 -0.5 0

(e) Active TB infection (f) TB treatment

Study Year Difference (95% CI) Reference Year Difference (95% CI)

Goletti 1996 -0.80 (-1.37, -0.22)

Goletti 1996 1.01 (-0.07, 2.08) Wallis 1996 0.17 (-0.27, 0.61)
Whalen 2000 -0.10 (-0.31, 0.11) Lawn 1999 -0.20 (-0.90, 0.50)
Whalen 2000 0.49 (0.02, 0.96)
Toossi (CD4>500) 2001 1.30 (0.44, 2.16)
Dean 2002 -0.94 (-1.78, -0.10)
Toossi (CD4<=500) 2001 -0.04 (-0.36, 0.28) Morris 2003 -0.07 (-0.40, 0.28)
Toossi (cohort) 2001 0.43 (0.05, 0.81) Kalou 2005 0.64 (0.32, 0.96)
Kizza 2005 -0.20 (-0.30, -0.10)
Hung 2003 0.24 (0.08, 0.40)
Schon 2006 -0.09 (-0.38, 0.20)
Kizza 2005 0.70 (0.18, 1.22) Toossi 2007 -0.08 (-0.26, 0.09)
Lopez-Gatell 2008 0.76 (0.39, 1.10) Nikolaeva 2008 0.09 (-0.03, 0.22)
Swaminathan 2008 -0.15 (-0.34, 0.04)
Overall (I-squared = 78.8%) 0.40 (0.13, 0.67)
Overall (I-squared = 77.7%) -0.03 (-0.19, 0.13)

0 0.5 1 1.5 2 -1.5 -1 -0.5 0 0.5 1

Fig. 2. Forest plots of associations between coinfections and their treatment and plasma HIV viral load (log10 copies/ml).
CI, confidence interval; HSV-2, herpes simplex virus type 2; TB, tuberculosis.

studies of HSV-2 prevalence and incidence, the summary regimens. Seven studies were RCTs [16,44–48,50] and
estimate showed a mean PVL increase of 0.18 log10 one was an observational cohort study [49]. Sample sizes
copies/ml (95% CI 0.01–0.34), with substantial hetero- ranged between 12 and 3302 participants and trials were
geneity (I2 ¼ 60.3%; Fig. 2c). conducted in the United States, Thailand, Burkina Faso,
South Africa, Uganda, Tanzania, Botswana, Zambia,
Eight studies assessed the impact of HSV-2 suppressive Kenya, Rwanda and Peru. Seven of these studies found
therapy on PVL (Table 2b). Of these, half evaluated decreases in PVL associated with suppressive therapy, with
regimens using acyclovir 400 mg twice daily, whereas the effect estimates from
0.53 log10 copies/ml (95% CI
remainder tested higher dose acyclovir or valacyclovir
0.72 to
0.35) to
0.25 log10 copies/ml (95% CI

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 Table 2. Association between herpes simplex virus type 2 and HIV-1 viral load.
Study Mean difference Viral load
Study Location Period Study design Endpoint/follow-up Population size (95% CI) assay

(a) HSV-2 infection and HIV-1plasma viral load

Barbour et al. [21] Sao Paulo, Brazil No details Cohort Quarterly visits, median HSV-2/HIV-1 coinfected adults 186 0.10 (
0.27 to 0.47) No details
follow-up 381 days HIV-1 infected adults 47 0.10 (
0.19 to 0.39) No details
with acute HSV-2 infection
Cachay San Diego, USA 1996-2005 Cross-sectional1 NA HSV-2 seropositive men 85
0.17 (
0.73 to 0.44) RT-PCR
et al. [42] with incident HIV-1 infection
HSV-2 seropositive men 209
0.12 (
0.45 to 0.20) RT-PCR
with early HIV-1 infection
Cachay San Diego, USA 1996-2005 Cohort Median follow-up HIV-1 infected men 9 0.08 (
0.30 to 0.46) RT-PCR
et al. [37] 779 days with acute HSV-2 infection
2
Chu Bangkok, Thailand 2000–1 Cross-sectional NA HIV-1/HSV-2 coinfected men 69 0.10 (
0.39 to 0.59) NASBA
et al. [39] HIV-1/HSV-2 coinfected 71 0.50 (0.03–0.97) NASBA
women
Duffus 2 sites, Uganda 1997–9 Cross-sectional NA HIV-1/HSV-2 coinfected 339 0.30 (0.07–0.53) No details
et al. [41] individuals
Gray Rakai, Uganda 1994–8 Cross-sectional NA HIV-1/HSV-2 coinfected 345
0.09 (0.28 to 0.10) RT-PCR
et al. [40] individuals
Mole Palo Alto, USA 1991–4 Cohort Single endpoint, HIV-infected males, with 8 0.75 (0.30–1.19) bDNA
et al. [38] 30–45 days acute or reactive HSV-2
infection
Serwadda Rakai, Uganda 1994–8 Nested NA Acute HIV-1 infected individuals, 219 0.55 (0.16–0.94) RT-PCR
et al. [43] case–control from a CRT of STI control
Mean
Study Study difference
Study Location Period design Population Therapy Endpoint/follow-up size (95% CI) VL assay

(b) HSV-2 treatment and HIV-1 plasma viral load

Suppressive therapy trials
Baeten Lima, Peru 2005 RCT HIV-1/HSV-2 Suppressive (valacyclovir 500 mg Weekly visits, 18 weeks 20
0.26 (
0.33 RT-PCR
et al. [44] seropositive women twice daily for 8 weeks) (cross-over) to
0.19)
Celum 14 sites in 2004–7 RCT Heterosexual HIV Suppressive (acyclovir 400 mg Quarterly visits, 24 months 3302
0.25 (
0.29 RT-PCR
et al. [45] 7 countries discordant couples twice daily for 24 months) to
0.22)
in Africa
Delany Johannesburg, 2005–6 RCT HIV-1/HSV-2 Suppressive (acyclovir 400 mg Monthly visits, 3 months 288
0.27 (
0.41 RT-PCR
et al. [46] South Africa seropositive women twice daily for 3 months) to
0.13)
Dunne Chiang Rai, No detailsRCT HIV-1/HSV-2 Suppressive (acyclovir 800 mg Monthly visits, 128
0.43 (
0.56 RT-PCR
et al. [47] Thailand coinfected women twice daily for 1 month) 3 months (cross-over) to
0.29)
Nagot Burkino Faso 2004–5 RCT HIV-1/HSV-2 Suppressive (valacyclovir twice Thrice weekly 136
0.53 (
0.72 RT-PCR
et al. [48] coinfected women, daily for 3 months) visits, 3 months to
0.35)
Schacker Seattle, USA 1994–6 Cohort Facility-recruited Suppressive (acyclovir 800 mg Weekly visits, 8 weeks 12
0.28 (
0.54 bDNA
et al. [49] HIV/HSV-2 thrice daily for 8 weeks) to
0.02)
coinfected individuals
Tanton Tanzania 2004, 2006 RCT HIV-1/HSV-2 Suppressive (acyclovir 400 mg 6, 12 and 24 month 419 0.02 (
0.09 RT-PCR
et al. [16] coinfected women twice daily until censoring) visits, 24 months to 0.13)
attending mobile clinic
Zuckerman Lima, Peru 2003–4 RCT HIV-1/HSV-2 Suppressive Weekly visits, 20
0.33 (
0.42 RT-PCR
et al. [50] coinfected MSM (valacyclovir 500 mg 18 weeks (cross-over) to
0.23)
twice daily for 8 weeks)
HIV coinfections and viral load Barnabas et al.

(continued overleaf )

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1565
 1566

Table 2 (continued )
AIDS

Mean
Study Study difference
Study Location Period design Population Therapy Endpoint/follow-up size (95% CI) VL assay

Episodic therapy trials

Mayaud 3 sites in 2003–7 RCT HIV-1 infected women Episodic (acyclovir 400 mg Single endpoint, 28 days 93 0.09 (
0.10 RT-PCR
et al. [51] Ghana/CAR with HSV-2 ulcers thrice daily for 5 days) to 0.30)
Mole Palo Alto, USA 1991–4 Cohort HIV infected males, Episodic (acyclovir Single endpoint, 30–45 days 8
0.48 (
0.77 bDNA
et al. [38] with HSV-2 ulcers 200 mg 5 times to
0.20)
a day for 10 days)
Paz-Bailey Gauteng, 2005–6 RCT HIV-infected males Episodic (acyclovir 400 mg Single endpoint, 28 days 295
0.12 (
0.23 RT-PCR
2011, Vol 25 No 13

et al. [52] South Africa with HSV ulcers thrice daily for 5 days) to
0.01)
Phiri et al. [15] Lilongwe, 2004–6 RCT HIV-1 infected Episodic (acyclovir 800 mg Single endpoint, 28 days 244 0.09 (
0.08 to 0.26) RT-PCR
Malawi individuals with twice daily for 5 days)
HSV-2 ulcers

Mean
Study difference
Study Location Period Population Therapy Endpoint/follow-up Specimen size (95% CI) VL assay

(c) HSV-2 treatment and HIV-1 genital viral load

Suppressive therapy trials
Baeten et al. [44] Lima, Peru 2005 HIV-1/HSV-2 Suppressive (valacyclovir 500 mg Weekly visits, 18 weeks Self-collected 20
0.67 (
1.08 RT-PCR
seropositive women twice daily for 8 weeks) (cross-over) genital swab to
0.26)
Endocervical swab 20
0.35 (
0.46 RT-PCR
specimen to
0.25)
Delany Johannesburg, 2005–6 HIV-1/HSV-2 Suppressive (acyclovir 400 mg Monthly visits, 3 months Cervicovaginal 288
0.13 (
0.28 RT-PCR
et al. [46] South Africa seropositive twice daily for 3 months) lavage to 0.03)
women
Dunne Chiang Rai, No details HIV-1/HSV-2 Suppressive (acyclovir 800 mg Monthly visits, 3 months Cervicovaginal 128
0.32 (
0.48 RT-PCR
et al. [47] Thailand coinfected women twice daily for 1 month) (cross-over) lavage to
0.19)
Nagot et al. [48] Burkino Faso 2004–5 HIV-1/HSV-2 Suppressive (valacyclovir Thrice weekly Cervicovaginal 136
0.29 (
0.44 RT-PCR
coinfected women twice daily for 3 months) visits, 3 months lavage to
0.15)
Tanton Tanzania 2004, 2006 HIV-1/HSV-2 Suppressive (acyclovir 400 mg 6, 12 and 24 month Cervicovaginal 425 0.03 (
0.11 RT-PCR
et al. [16] seropositive women twice daily until censoring) visits, 24 months lavage to 0.16)
Zuckerman Lima, Peru 2003–4 HIV-1/HSV-2 Suppressive (valacyclovir 500 mg Thrice weekly visits, Anoscopy with 20
0.16 (
0.25 RT-PCR
et al. [50] coinfected men twice daily for 8 weeks) 18 weeks (cross-over) Snostrips to
0.07)
Zuckerman Lima, Peru No details HIV-1/HSV-2 Suppressive (valacyclovir 500 mg Weekly visits, 18 weeks Semen 19
0.29 (
0.48 RT-PCR
et al. [31] coinfected men twice daily for 8 weeks) (cross-over) to
0.11)
Episodic therapy trials
Mayaud 3 sites in 2003–5 HIV-1 infected women Episodic (acyclovir 400 mg Day 7, 28 days Cervicovaginal 89
0.06 (
0.40 RT-PCR
et al. [51] Ghana/CAR with HSV-2 ulcers thrice daily for 5 days) lavage to 0.30)
Paz-Bailey Gauteng, 2005–6 HIV-1 infected men Episodic (acyclovir 400 mg Day 7, 28 days Ulcer lavage 193
0.82 (
1.45 RT-PCR
et al. [52] South Africa with HSV-2 ulcers thrice daily for 5 days) to
0.18)
Phiri Lilongwe, 2004–6 HIV-1 infected men Episodic (acyclovir 800 mg Day 14, 28 days Semen 62
0.14 (
0.72 RT-PCR
et al. [15] Malawi with HSV-2 ulcers twice daily for 5 days) to 0.44)
HIV-1 infected women Cervical swab 41
0.08 (
0.66 RT-PCR
with genital ulcers to 0.50)

CI, confidence interval; CRT, community randomized trial; HSV-2, herpes simplex virus type 2; NA, not applicable; NASBA, , nucleic-acid sequence-based-analysis; RCT, randomized controlled trial;
RT, real time; STI, sexually transmitted infection; VL, viral load.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 HIV coinfections and viral load Barnabas et al. 1567

0.29 to
0.22). The remaining suppressive treatment The infection studies were all observational, with three
study [16] found no evidence of an effect possibly due to cohort [19,26,53], three-nested case–control [18,32,54],
the comparatively low adherence (123/232 participants and three cross-sectional studies [26,55,56]. Two of these
[16] compared with 2363/2924 participants [45] with reported on the association of latent TB with viral load,
>90% adherence) and relatively low dose of acyclovir one finding a significant decrease [32] and the other
(400 mg twice daily by mouth). Overall, the summary finding no effect [18]. Of the six articles on active TB,
estimate of the decrease in PVL was
0.28 (95% CI three were pulmonary [26,54,55], two were both

0.36 to
0.19), with I2 ¼ 83.1% (Fig. 2d). pulmonary and extrapulmonary TB [53,56] and the
sixth article did not specify site of infection [19]. Overall,
Four studies (three RCTs and one cohort study) assessed active TB was associated with a mean increase in viral load
the relationship between episodic therapy (5–10 days of of 0.40 log10 copies/ml (95% CI 0.13–0.67) (Fig. 2e),
acyclovir treatment at a range of doses) and PVL up to although there was again substantial heterogeneity
45 days after treatment initiation [15,38,51,52]. Sample between studies (I2 ¼ 78.8%).
sizes ranged from 8 to 422 participants, and studies
included participants in the United States, Malawi, South Of the 12 studies that investigated associations between
Africa, Ghana and the Central African Republic. Three treatment of active TB and PVL, ten were cohort studies
of the four trials [15,51,52] had small numbers on ART and two were RCTs of non-TB treatment (immune
(<4% of all participants), and it was not possible to extract modulators pentoxifylline and Immunoxel) and standard
data on ART naive participants only. Two of the studies TB treatment versus standard TB treatment alone
using higher dose acyclovir regimens reported a [17,57]. In both trials, the intervention arm showed a
significant decrease in PVL associated with episodic significantly greater reduction in PVL than controls.
therapy, whereas the other two studies found no evidence We, therefore, used data only from the standard TB
of an association (summary estimate of decrease treatment arms of these RCTs. Duration of treatment

0.08 log10 copies/ml, 95% CI
0.28 to 0.11). and study follow-up varied considerably (2–8 months;
and 2–12 months after treatment initiation, respec-
Eleven studies measured GVL [15,16,31,39,44,46– tively). The summary estimate of the mean difference in
48,50–52]. There was wide variation in treatment PVL was a decrease of
0.03 log10 copies/ml (95% CI
regimens and in both the site and method of collection
0.19 to 0.13) (Fig. 2f), which did not change
of samples for this outcome. Collection methods included significantly when studies with less than 6 months of
cervicovaginal lavage, semen samples and endocervical follow-up were excluded.
and rectal swabs. Ten of the 11 studies were RCTs of the
effect of HSV-2 treatment (seven suppressive; three Two additional studies [22,33] reported change in PVL
episodic). The remaining was a cross-sectional study [39] from before active TB onset compared with after TB
that found no difference in GVL by HSV-2 infection treatment, so that it was not possible to distinguish
status. Six of the RCTs found that treatment significantly between the effects of infection and treatment. However,
decreased GVL (five suppressive; one episodic) both showed a significant increase in viral load associated
[31,44,47,48,50,52]. All of these evaluated valacyclovir with infection/treatment (results not shown).
or high dose acyclovir regimens. The remaining studies
found no significant effects. A formal meta-analysis was Risks of bias within and across studies
not undertaken for this outcome, due to numerous The methodological quality of studies varied. For
sources of heterogeneity, including the range of treatment malaria, the majority of studies were cohort studies
regimens, the variety of techniques used for sample which reported viral load before, during and after acute
collection, the variable lower limits of detection in the malaria infection and were therefore rated A or B,
GVL assays used and day-to-day variations in detectable depending on whether there was an external control
levels of viral load present in genital samples. Individual group. One cross-sectional study was rated C for
study results for the RCTs are summarized in Table 2c. methodological quality because associations between
treatment and PVL did not adjust for difference in CD4
Tuberculosis cell count between the comparison groups. Studies of
Eight articles (one containing results from two studies, of HSV-2 treatment were of high methodological quality,
which one study was divided into participants with CD4 the majority being RCTs. However, studies examining
cell count 500 or less and more than 500 [26]) reported HSV-2 infection were of variable methodological
on the effect of TB infection on PVL (Table 3a), and quality. Two did not adjust for potential confounders
12 examined the association of TB treatment and PVL [39,40] and were, therefore, rated C. For TB, three
(Table 3b). Sample sizes varied from seven participants to cross-sectional studies were rated C due to extremely
276 (median 20). All studies were facility-based (hospital, small sample size or failure to adjust for time since
clinic or TB treatment center) and located in urban or infection. For each coinfection, removal of studies rated
periurban settings. The majority were conducted in C from the meta-analysis did not significantly change
Africa. results.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 1568

Table 3. Association between tuberculosis and HIV-1 viral load.

AIDS

Mean
Study difference
Study Location Period Study design Population Follow-up TB size (95% CI) VL assay TB diagnosis

(a) TB infection and HIV-1 plasma viral load

Goletti et al. [19] 3 sites in No details Cohort HIV-1/TB coinfected 6–10 months Active 2 1.01 (
0.07 to 2.08) bDNA Culture positive
USA and Italy patients for MTB
2
Hung et al. [56] Taiwan 1994–2002 Cross-sectional HIV infected individuals Active 276 0.24 (0.08 to 0.40) RT-PCR Culture positive
for MTB
Kizza et al. [55] Kampala, Uganda 2000–1 Cross-sectional2 HIV-1/TB-coinfected Active 40 0.70 (0.18 to 1.22) RT-PCR Culture positive
individuals for MTB
2011, Vol 25 No 13

Lopez-Gatell USA 1984-2005 Cohort HIV-1/TB-coinfected Median Active 15 0.76 (0.39 to 1.10) RT-PCR Culture, cytology,
et al. [53] men 5.4 years clinical or
radiology
confirmed
Manoff et al. [18] Baltimore, USA 1990–4 Nested HIV-1 infected Latent 6
0.14 (
0.94 to 0.66) bDNA TST >5 mm
case–control individuals
Mawa et al. [32] Entebbe, Uganda 2000 Nested HIV-1 infected Latent 29
0.5 (
0.99,
0.01) RT-PCR TST > 5 mm
case–control individuals
Toossi et al. [26] Kampala, Uganda 1993–5 Cross-sectional HIV-1/TB coinfected Active 51
0.04 (
0.36 to 0.28) RT-PCR Culture positive
individuals with for MTB
CD4<500
Cross-sectional HIV-1/TB coinfected Active 23 1.30 (0.44 to 2.16) RT-PCR Culture positive
individuals with for MTB
CD4 cell count > 500
Cohort HIV-1 infected 6 months Active 10 0.43 (0.05 to 0.81) RT-PCR Culture positive
individuals for MTB
who developed
pulmonary TB
Whalen et al. [54] Kampala, Uganda 1993–4 Nested HIV infected individuals Active 40
0.10 (
0.31 to 0.11) RT-PCR Culture positive
case–control with and without TB for MTB

Mean
Study difference VL TB
Study Location Period Population Therapy Follow-upb TB size (95% CI) assay diagnosis

(b) TB treatment and HIV-1 plasma viral loada

Dean et al. [23] South-East UK 1996–9 HIV-1/TB coinfected Multiple 12 months Active 20
0.94 (
1.78 to
0.10) Multiple Culture positive
patients for MTB
Goletti et al. [19] 3 sites in US and Italy No details HIV-1/TB coinfected 3–10 months Active 5
0.80 (
1.37 to
0.22) bDNA Culture positive
patients for MTB
Kalou et al. [59] Abidjan, Ivory Coast 1995–8 HIV-1 infected 2HRZ/4HR 12 months Active 44 0.64 (0.32 to 0.96) RT-PCR 2 positive sputum
patients with smears
newly diagnosed TB
Kizza et al. [55] Kampala, Uganda 2000–1 HIV-1/TB-coinfected 2HRZE/4HR 6 months Active 27
0.20 (
0.30 to
0.10) RT-PCR Culture positive
individuals for MTB
Lawn [58] Ghana no details HIV-1/TB coinfected 2SHRZ/6HE 3 months Active 20
0.20 (
0.90 to 0.50) RT-PCR At least 2 out of 3
individuals positive sputum
smears
Morris et al. [60] South Africa 1997–8 HIV-1/TB-coinfected 2HRZE/4HR 6 months Active 57
0.07 (
0.40 to 0.26) RT-PCR Sputum positive for
individuals AFB and radiology

(continued overleaf )

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 HIV coinfections and viral load Barnabas et al. 1569

for AFB, or radiology/

smear and radiology
Sputum smear positive
Funnel plots and the Egger test for each outcome did not
indicate any evidence of publication bias (results not

positive for AFB

positive sputum

histopathology
shown).
Culture positive

Culture positive

Culture positive

Culture positive
Sputum smear
for MTB, or

for MTB

for MTB

for MTB
diagnosis
TB

Discussion
RT-PCR

RT-PCR

RT-PCR

RT-PCR

RT-PCR

RT-PCR
Malaria, HSV-2 and TB contribute significantly to the
assay

global burden of infectious disease, with high prevalence

VL

areas geographically overlapping with HIV endemic

regions. Here, dual infection occurs commonly, making
0.09 (
0.03 to 0.22)

0.15 (
1.03 to 0.72)

0.09 (
0.38 to 0.20)

0.08 (
0.26 to 0.09)

0.17 (
0.27 to 0.61)

0.49 (0.02 to 0.96)

the role of coinfections and their treatment in HIV

AFB, acid fast bacillus; MTB, Mycobacterium tuberculosis; RT, real time; TB, tuberculosis; TST, tuberculin skin test; US, United States; VL, viral load.
difference
(95% CI)

transmission salient questions. We identified 45 studies of

Mean

the effect of these coinfections on HIV viral load (six for

malaria, 20 for HSV-2 and 19 for TB). We found
significant increases in HIV PVL associated with acute
malaria (0.67 log10 copies/ml; 95% CI 0.15–1.19), HSV-
2 (0.18 log10 copies/ml; 95% CI 0.01–0.34) and active
Study
size

20

14

20

15

20

20

TB (0.40 log10 copies/mL; 95% CI 0.13–0.67). Overall

treatment for malaria appeared to have limited impact on
Active

Active

Active

Active

Active

Active

viral load (
0.25 log10 copies/ml, 95% CI
0.59 to 0.10);

TB

however, when the study with less than 10 days of follow-

up was excluded, viral load decreased significantly with
Follow-upb

malaria treatment (
0.37 log10 copies/ml; 95% CI
0.70

12 months

12 months
2 months

2EHRZ3/4RH3 6 months

2 months

6 months

to
0.04). These results are consistent with an earlier

study suggesting that the full impact of malaria treatment
on HIV viral load does not occur within 4 weeks after
treatment [34], Kublin et al. [35] chose 8 weeks after
2HZER/6HR

HIV infected individuals 2HRZE/6HR

2HZES/6HE

malaria treatment as the follow-up time necessary to see a

2HRZSE
Therapy

return of HIV viral load to baseline, suggesting that a

9-day follow-up period would be too short. Suppressive
treatment for HSV-2 produced significant decreases in
HIV viral load (
0.28 log10 copies/ml, 95% CI
0.36 to
with and without TB
Up to 1995 HIV-1/TB-coninfected
HIV-1/TB-coinfected

HIV-1/TB-coinfected

No details HIV-1/TB-coinfected

0.19); however, there was no significant impact of

HIV/TB coinfected

either episodic HSV-2 treatment (
0.08 log10 copies/ml,

95% CI
0.28 to 0.11) or TB therapy (
0.03 log10
individuals

individuals

individuals

individuals

individuals
Population

copies/ml, 95% CI
0.19 to 0.13) on HIV viral load.

Collectively, these studies suggest that coinfections
increase HIV viral load, and that some of these effects
may be reversed by treatment using current treatment
1999-2000

regimen.
No details
no details

1993–4
Period

We found that TB treatment did not significantly

reduce viral load, an unexpected finding because active
TB was associated with a 0.40 log10 copies/ml increase
Kampala, Uganda

Kampala, Uganda

Whalen et al. [54] Kampala, Uganda

in viral load. There are at least two possible reasons

Gondor, Ethiopia
Chennai, India

for this finding. First, the long course of TB treatment

(6–9 months) and variable follow-up times make
Location

comparisons challenging because both the outcome of

Ukraine

interest (viral load) and its main modifier (CD4 cell count)
All are cohort studies.
Time since diagnosis.

may change over the course of treatment. However,

Table 3 (continued )

excluding studies with less than 6 months of follow-up did

Wallis et al. [57]
Toossi et al. [27]
Schon et al. [61]

not change our results. Second, standard TB treatment

Swaminathan
et al. [17]

et al. [24]

regimens may not address the underlying mechanisms that

Nikolaeva

trigger increased PVL. Assessing the impact of treatment

Study

on chronic coinfection will require careful study designs to

account for modifiers of effect that change over time.
b
a

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1570 AIDS 2011, Vol 25 No 13

Our results add to those in a recent review of the impact rate above 1 in the population. ART may be able to
of coinfection treatment [62]. Our findings for impact of supersede other factors in forcing viral load back below
malaria and HSV-2 treatment support those of the review, these levels. However, for the vast number of HIV-
but our TB results do not. The Modjarrad review infected individuals who are in the pre-ART window
included two studies of malaria treatment, two of TB or who do not have access to ART despite being
treatment and six of HSV-2 treatment. In contrast, we eligible, interventions to prevent and treat coinfections
identified six studies of malaria treatment, 12 of TB may be an important strategy to tip the balance in favor of
treatment and eight of HSV-2 suppressive treatment, as the host not only by limiting progression and delaying
well as studies of coinfection prior to treatment. There are ART initiation, but also by maintaining community viral
several differences between the reviews, including the load at levels that limit transmission and help curb
inclusions/exclusion criteria, methods of analysis, and epidemic trajectory. It is noteworthy that the potential
data extracted. For example, the Modjarrad review impact of coinfections on total community viral load
required studies to include a control group to adjust for over time is a function not only of the magnitude of the
natural history changes in viral load over time, but most of increase in HIV viral load observed in coinfected
the additional studies included in our review were cohort individuals, but also of the prevalence of the infection
studies conducted over relatively short periods during among HIV-positives and the average duration or
which viral load would not be expected to change proportion of time during an interval such as one year
substantially in the absence of coinfection or interven- that coinfection is likely to exist. When the results of
tion. We, therefore, included pre-post studies without our meta-analysis are examined from this perspective,
control groups. Further, the Modjarrad review used the the potential impact of HSV-2 coinfection on total
standardized mean difference (SMD) as the measure of community viral load over a year is likely to markedly
effect, rather than the actual viral load, which tends to outstrip that of malaria and TB, even though the increase
exaggerate differences between studies because the in HIV viral load in coinfected individuals with malaria
dimensionless scale of the SMD is not directly is more than three times that observed with HSV-2,
interpretable. Finally, for the two TB studies that were and the increase with TB is more than twice that with
included in both this and the Modjarrad review [19,55], HSV-2.
the data extracted in the Modjarrad review used a
subgroup of patients, while we included data from Although treatment of coinfections has been demon-
all patients. strated to delay HIV disease progression and ART
initiation, it has yet to be shown to reduce HIV
Impact on HIV transmission transmission in intervention trials. Although an RCT of
Although our review found only modest changes in PVL HSV-2 suppressive therapy found a 16% reduction in HIV
associated with coinfections and their treatment progression as measured by CD4 cell count less than 200,
(0.2 log10 to 0.7 log10 for coinfections and
0.3 log10 ART initiation or death [hazard ratio 0.83 (0.71–0.98)],
to
0.4 log10 for treatment), modeling studies have there was no effect on HIV transmission [45,67]. This
reported that small changes in mean viral load at may be due to persistent immune activation, including
population level could translate to significant reductions HIV target cells in the genital tract as much as 2 months
in HIV incidence – a decrease in PVL of 0.3 log10 was after ulcer healing on standard HSV suppressive therapy
estimated to decrease HIV transmission by 20% and [68]. Hence, although HSV-2 suppressive therapy
HIV progression by 25% [63]. Lingappa et al. [64] decreases PVL sufficiently to help reduce the impact of
produced similar estimates in their analysis of PVL HIV disease on infected individuals, further work is
among 108 genetically linked HIV transmission events, needed to fully understand the mechanisms by which
suggesting that a 0.3 log10 reduction in PVL would HSV boosts HIV viral load, better define the viral load
reduce HIV transmission risk by 25%. reductions needed to curb transmission, and design HSV
treatment regimens that reliably achieve these targets.
Fraser et al. [65] proposed that HIV evolves to reach One recent model indicates that at least a 0.75 log10
a balance between sufficiently high viral load for reduction in HIV viral load is needed to decrease
successful transmission with each exposure and suffi- transmission by 50% [64]. Trials are needed to evaluate
ciently low viral load for a protracted asymptomatic the impact of carefully redesigned treatment regimens for
period to maximize transmission opportunities. We HSV and other coinfections on HIV transmission, as well
posit that coinfections help HIV achieve that ‘sweet as on the surrogate of HIV viral load. Evidence of an
spot’ viral load through immune activation [58,66], impact of TB and malaria on HIV progression is limited,
increasing the cumulative time spent at the optimum with mixed findings for TB [69,70] and no effect for
viral load for both asymptomatic disease and HIV malaria [71,72]. In evaluating non-ART regimens to
transmission. From a population perspective, this may reduce HIV disease progression we suggest using
accelerate HIV epidemic trajectory by increasing the outcomes as defined by Lingappa et al. [67]: ART
aggregate person-time at viral loads at which the risk for initiation, CD4 cell count less than 200 and HIV-related
transmission is sufficient to maintain the reproductive death.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 HIV coinfections and viral load Barnabas et al. 1571

Study limitations HIV coinfections may offer important opportunities to

Methodological quality of individual studies was variable. reduce viral load, and thereby both curb HIV trans-
Several studies (especially of malaria and TB) were cross- mission and extend the period before ART is required.
sectional and thus particularly susceptible to bias. Our results suggest that malaria prevention and treatment
However, nearly all controlled for potential confounders should be evaluated as part of multicomponent HIV
such as CD4 cell count and sex, and results were similar prevention packages in appropriate epidemiologic con-
when restricted to these studies. Although we found no texts. Although HSV-2 infection and active TB increase
evidence of publication bias, we may not have had viral load, we do not yet have HSV-2 treatment regimens
sufficient power to detect this. Studies included used with demonstrated efficacy in reducing HIV trans-
different diagnostic methods and definitions for acute/ mission, or TB treatment regimens capable of reducing
active infection, which could affect our findings HIV PVL. However, due to their prevalence and
(although when examined explicitly for malaria [35], persistence, HSV-2 coinfections may be particularly
no material difference was found in the results). Further, important drivers of total community viral load over time.
assays for measuring viral load and CD4 cell count also For HSV-2 infection, we have assumed that regimens
varied between studies and standardization of tests would with well-documented efficacy in reducing clinical
likely narrow the confidence intervals. symptoms and HSV-2 genital shedding would perform
equally well in reducing HIV viral load and transmission.
We used a random effects model and found that It is now clear that we must better understand the
heterogeneity was significant, which we would expect mechanisms responsible for augmented plasma and GVL,
given variability in study design, populations, sample sizes and retune HSV-2 treatment regimens accordingly.
and follow-up times. Follow-up times varied from 9 to Similar approaches may well be key to development of
65 days after malaria treatment, 28 days to 24 months for strategies to reduce the impact of malaria or TB on HIV
HSV-2 therapy and 2 to 12 months for TB treatment. In transmission. We must also better define the level of viral
addition, differences in treatment regimens contributed load reduction that we must achieve with these treatment
to the observed heterogeneity. Thus, our results are regimens to limit HIV transmission.
interpreted with caution but provide strong stimulus for
further research on the impact of coinfections on HIV Understanding the role of coinfection in the trajectory of
viral load and transmission. HIV epidemics and in positive prevention is critical to the
development of evidence-based HIV prevention and care
Our study focused on PVL, a reproducible quantitative policies and programs, and to the rational allocation of
measure of HIV infectiousness. However, because HIV is resources to implement them with the requisite quality and
sexually transmitted in regions bearing the greatest scale. As we develop multicomponent combination
burden of the epidemic, it would be most relevant to packages of HIV prevention interventions tailored to
determine the impact of coinfections on GVL. In a people across the full range of HIV status and risk,
study among 2521 African HIV serodiscordant couples, prevention and treatment of coinfections may be
Baeten et al. [73] demonstrated that higher GVL was particularly relevant for HIV-infected individuals who
associated with greater risk of HIV transmission – they are not on ART, and may offer an important entry point for
found that a 1 log10 increase in GVL was associated with broader HIV prevention and care services in communities
roughly a two-fold increased risk of HIV transmission with high burdens of these intertwined diseases.
(2.2 fold, 95% CI 1.60–3.04 for endocervical swabs
and 1.8, 95% CI 1.30–2.47 for semen) [73]. To date,
data on GVL are limited in studies of the effect of
coinfections and their treatment on HIV viral load. Acknowledgements
Indeed, GVL was not measured in any of the malaria or
TB studies that we reviewed. However, it is plausible We are grateful to Ann Marie Clark, Director, and the
that compared with systemic infections, sexually trans- outstanding staff, especially Alison Nobis, at the Arnold
mitted infections increase GVL more than PVL due to Library, Fred Hutchinson Cancer Research Center, and
local inflammatory factors in the genital tract. A key area Aleta Elliott in the Department of Global Health for help
for future research is elucidating the HIV viral load with electronic searches and interlibrary loans.
dynamics between plasma and the genital compartment
for both systemic and sexually transmitted infections that Conflicts of interest
modulate viral load. Improved approaches to accurately R.V.B. is funded by the National Center for Research
and reproducibly measure GVL will help answer these Resources/National Institutes of Health (NIH) (5 KL2
questions. RR025015) and the Qatar National Research Fund
(NPRP 08–068–3–024). R.V.B., H.A.W. and J.N.W.
Conclusion acknowledge support from the NIH (1R01 AI083034).
In combination with other interventions, prevention
and treatment of some frequently recurrent or persistent The authors have no conflicts of interest to declare.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
1572 AIDS 2011, Vol 25 No 13

References 22. Day JH, Grant AD, Fielding KL, Morris L, Moloi V, Charalam-
bous S, et al. Does tuberculosis increase HIV load? J Infect Dis
2004; 190:1677–1684.
1. UNAIDS. AIDS epidemic update. Geneva: WHO; 2009:100. 23. Dean GL, Edwards SG, Ives NJ, Matthews G, Fox EF, Navaratne
2. Buve A, Carael M, Hayes RJ, Auvert B, Ferry B, Robinson NJ, L, et al. Treatment of tuberculosis in HIV-infected persons in
et al. The multicentre study on factors determining the differ- the era of highly active antiretroviral therapy. AIDS 2002;
ential spread of HIV in four African cities: summary and 16:75–83.
conclusions. AIDS 2001; 15 (Suppl 4):S127–S131. 24. Swaminathan S, Deivanayagam CN, Rajasekaran S, Venkatesan
3. Donnell D, Baeten JM, Kiarie J, Thomas KK, Stevens W, Cohen P, Padmapriyadarsini C, Menon PA, et al. Long term follow up
CR, et al. Heterosexual HIV-1 transmission after initiation of of HIV-infected patients with tuberculosis treated with
antiretroviral therapy: a prospective cohort analysis. Lancet 6-month intermittent short course chemotherapy. Natl Med J
2010; 375:2092–2098. India 2008; 21:3–8.
4. Quinn TC, Wawer MJ, Sewankambo N, Serwadda D, Li C, 25. Tatfeng YM, Ihongbe JC, Okodua M, Oviasogie E, Isibor J,
Wabwire-Mangen F, et al. Viral load and heterosexual trans- Tchougang S, et al. CD4 count, viral load and parasite density
mission of human immunodeficiency virus type 1. Rakai Pro- of HIV positive individuals undergoing malaria treatment with
ject Study Group. N Engl J Med 2000; 342:921–929. dihydroartemisinin in Benin City, Edo state, Nigeria. J Vector
5. Glynn JR, Biraro S, Weiss HA. Herpes simplex virus type 2: a key Borne Dis 2007; 44:111–115.
role in HIV incidence. AIDS 2009; 23:1595–1598. 26. Toossi Z, Mayanja-Kizza H, Hirsch CS, Edmonds KL, Spahlinger
6. Bunnell R, Mermin J, De Cock KM. HIV prevention for a T, Hom DL, et al. Impact of tuberculosis (TB) on HIV-1 activity
threatened continent: implementing positive prevention in in dually infected patients. Clin Exp Immunol 2001; 123:233–
Africa. JAMA 2006; 296:855–858. 238.
7. WHO. Tuberculosis Facts. 2009. http://www.who.int/tb/publi 27. Toossi Z, Mayanja-Kizza H, Lawn SD, Hirsch CS, Lupo LD,
cations/2009/tbfactsheet_2009update_one_page.pdf. Butera ST. Dynamic variation in the cellular origin of HIV
8. Weiss H. Epidemiology of herpes simplex virus type 2 infection type 1 during treatment of tuberculosis in dually infected
in the developing world. Herpes 2004; 11 (Suppl 1):24A–35A. subjects. AIDS Res Hum Retroviruses 2007; 23:93–100.
9. Grimwade K, French N, Mbatha DD, Zungu DD, Dedicoat M, 28. Higgins JP, Thompson SG. Quantifying heterogeneity in a
Gilks CF. HIV infection as a cofactor for severe falciparum meta-analysis. Stat Med 2002; 21:1539–1558.
malaria in adults living in a region of unstable malaria trans- 29. Light RJ, Pillemer DB. Summing up. The science of reviewing
mission in South Africa. AIDS 2004; 18:547–554. research. Cambridge, MA: Harvard University Press; 1984.
10. French N, Nakiyingi J, Lugada E, Watera C, Whitworth JA, Gilks 30. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-
CF. Increasing rates of malarial fever with deteriorating analysis detected by a simple, graphical test. BMJ 1997;
immune status in HIV-1-infected Ugandan adults. AIDS 315:629–634.
2001; 15:899–906. 31. Zuckerman RA, Lucchetti A, Whittington WL, Sanchez J,
11. Walson JL, John-Stewart G. Treatment of helminth co-infection Coombs RW, Magaret A, et al. HSV suppression reduces semi-
in individuals with HIV-1: a systematic review of the literature. nal HIV-1 levels in HIV-1/HSV-2 co-infected men who have sex
PLoS Negl Trop Dis 2007; 1:e102. with men. AIDS 2009; 23:479–483.
12. Higgins JPT, Green S. Cochrane Handbook for Systematic 32. Mawa PA, Pickering JM, Miiro G, Namujju PB, Watera C,
Reviews of Interventions 4.2.6. Chichester, UK: John Wiley & Anyaegani G, et al. The effect of tuberculin skin testing on
Sons, Ltd.; 2006. viral load and antimycobacterial immune responses in HIV-1-
13. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, infected Ugandan adults. Int J Tuberc Lung Dis 2004; 8:586–
Ioannidis JP, et al. The PRISMA statement for reporting sys- 592.
tematic reviews and meta-analyses of studies that evaluate 33. Wolday D, Hailu B, Girma M, Hailu E, Sanders E, Fontanet AL.
healthcare interventions: explanation and elaboration. BMJ Low CD4R T-cell count and high HIV viral load precede the
2009; 339:b2700. development of tuberculosis disease in a cohort of HIV-positive
14. French N. Pneumoccal disease and its prevention by vaccina- Ethiopians. Ethiop Med J 2003; 41 (Suppl 1):67–73.
tion in HIV-infected Ugandan adults. Liverpool: University of 34. Hoffman IF, Jere CS, Taylor TE, Munthali P, Dyer JR, Wirima JJ,
Liverpool; 2004. et al. The effect of Plasmodium falciparum malaria on HIV-1
15. Phiri S, Hoffman IF, Weiss HA, Martinson F, Nyirenda N, RNA blood plasma concentration. AIDS 1999; 13:487–494.
Kamwendo D, et al. Impact of aciclovir on ulcer healing, 35. Kublin JG, Patnaik P, Jere CS, Miller WC, Hoffman IF, Chimbiya
lesional, genital and plasma HIV-1 RNA among patients with N, et al. Effect of Plasmodium falciparum malaria on concen-
genital ulcer disease in Malawi. Sex Transm Infect 2010; tration of HIV-1-RNA in the blood of adults in rural Malawi: a
86:345–352. prospective cohort study. Lancet 2005; 365:233–240.
16. Tanton C, Weiss HA, Rusizoka M, Legoff J, Changalucha J, 36. Van Geertruyden JP, Mulenga M, Kasongo W, Polman K,
Baisley K, et al. Long-term impact of acyclovir suppressive Colebunders R, Kestens L, D’Alessandro U. CD4 T-cell count
therapy on genital and plasma HIV RNA in Tanzanian women: and HIV-1 infection in adults with uncomplicated malaria.
a randomized controlled trial. J Infect Dis 2010; 201:1285– J Acquir Immune Defic Syndr 2006; 43:363–367.
1297. 37. Cachay ER, Frost SD, Poon AF, Looney D, Rostami SM, Pacold
17. Nikolaeva LG, Maystat TV, Pylypchuk VS, Volyanskii YL, ME, et al. Herpes simplex virus type 2 acquisition during recent
Masyuk LA, Kutsyna GA. Effect of oral immunomodulator HIV infection does not influence plasma HIV levels. J Acquir
Dzherelo in TB/HIV co-infected patients receiving antituber- Immune Defic Syndr 2008; 47:592–596.
culosis therapy under DOTS. Int Immunopharmacol 2008; 38. Mole L, Ripich S, Margolis D, Holodniy M. The impact of active
8:845–851. herpes simplex virus infection on human immunodeficiency
18. Manoff SB, Farzadegan H, Munoz A, Astemborski JA, Vlahov D, virus load. J Infect Dis 1997; 176:766–770.
Rizzo RT, et al. The effect of latent Mycobacterium tuberculosis 39. Chu L, Jiamton S, Pepin J, Cowan F, Mahakkanukrauh B, Suttent
infection on human immunodeficiency virus (HIV) disease R, et al. Association between HSV-2 and HIV-1 viral load in
progression and HIV RNA load among injecting drug users. semen, cervico-vaginal secretions and genital ulcers of Thai
J Infect Dis 1996; 174:299–308. men and women. Int J STD AIDS 2006; 17:681–686.
19. Goletti D, Weissman D, Jackson RW, Graham NM, Vlahov D, 40. Gray RH, Li X, Wawer MJ, Serwadda D, Sewankambo NK,
Klein RS, et al. Effect of Mycobacterium tuberculosis on HIV Wabwire-Mangen F, et al. Determinants of HIV-1 load in
replication. Role of immune activation. J Immunol 1996; subjects with early and later HIV infections, in a general-
157:1271–1278. population cohort of Rakai, Uganda. J Infect Dis 2004;
20. Chen X, Xiao B, Shi W, Xu H, Gao K, Rao J, Zhang Z. Impact of 189:1209–1215.
acute vivax malaria on the immune system and viral load of HIV- 41. Duffus WA, Mermin J, Bunnell R, Byers RH, Odongo G, Ekwaru
positive subjects. Chin Med J (Engl) 2003; 116:1810–1820. P, Downing R. Chronic herpes simplex virus type-2 infection
21. Barbour JD, Sauer MM, Sharp ER, Garrison KE, Long BR, and HIV viral load. Int J STD AIDS 2005; 16:733–735.
Tomiyama H, et al. HIV-1/HSV-2 co-infected adults in early 42. Cachay ER, Frost SD, Richman DD, Smith DM, Little SJ. Herpes
HIV-1 infection have elevated CD4R T cell counts. PLoS One simplex virus type 2 infection does not influence viral dynamics
2007; 2:e1080. during early HIV-1 infection. J Infect Dis 2007; 195:1270–1277.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
 HIV coinfections and viral load Barnabas et al. 1573

43. Serwadda D, Gray RH, Sewankambo NK, Wabwire-Mangen F, 57. Wallis RS, Nsubuga P, Whalen C, Mugerwa RD, Okwera A,
Chen MZ, Quinn TC, et al. Human immunodeficiency virus Oette D, et al. Pentoxifylline therapy in human immuno-
acquisition associated with genital ulcer disease and herpes deficiency virus-seropositive persons with tuberculosis: a rando-
simplex virus type 2 infection: a nested case-control study in mized, controlled trial. J Infect Dis 1996; 174:727–733.
Rakai, Uganda. J Infect Dis 2003; 188:1492–1497. 58. Lawn SD. AIDS in Africa: the impact of coinfections on the
44. Baeten JM, Strick LB, Lucchetti A, Whittington WL, Sanchez J, pathogenesis of HIV-1 infection. J Infect 2004; 48:1–12.
Coombs RW, et al. Herpes simplex virus (HSV)-suppressive 59. Kalou M, Sassan-Morokro M, Abouya L, Bile C, Maurice C,
therapy decreases plasma and genital HIV-1 levels in HSV-2/ Maran M, et al. Changes in HIV RNA viral load, CD4R T-cell
HIV-1 coinfected women: a randomized, placebo-controlled, counts, and levels of immune activation markers associated
cross-over trial. J Infect Dis 2008; 198:1804–1808. with anti-tuberculosis therapy and cotrimoxazole prophylaxis
45. Celum C, Wald A, Lingappa JR, Magaret AS, Wang RS, Mugo N, among HIV-infected tuberculosis patients in Abidjan, Cote
et al. Acyclovir and transmission of HIV-1 from persons d’Ivoire. J Med Virol 2005; 75:202–208.
infected with HIV-1 and HSV-2. N Engl J Med 2010; 362:427– 60. Morris L, Martin DJ, Bredell H, et al. Human immunodeficiency
439. virus-1 RNA levels and CD4 lymphocyte counts, during treat-
46. Delany S, Mlaba N, Clayton T, Akpomiemie G, Capovilla A, ment for active tuberculosis, in South African patients. J Infect
Legoff J, et al. Impact of aciclovir on genital and plasma HIV-1 Dis 2003; 187:1967–1971.
RNA in HSV-2/HIV-1 co-infected women: a randomized 61. Schon T, Wolday D, Elias D, Melese E, Moges F, Tessema T,
placebo-controlled trial in South Africa. AIDS 2009; 23:461– et al. Kinetics of sedimentation rate, viral load and TNF-alpha in
469. relation to HIV co-infection in tuberculosis. Trans R Soc Trop
47. Dunne EF, Whitehead S, Sternberg M, Thepamnuay S, Med Hyg 2006; 100:483–488.
Leelawiwat W, McNicholl JM, et al. Suppressive acyclovir 62. Modjarrad K, Vermund SH. Effect of treating co-infections on
therapy reduces HIV cervicovaginal shedding in HIV- and HIV-1 viral load: a systematic review. Lancet Infect Dis 2010;
HSV-2-infected women, Chiang Rai, Thailand. J Acquir 10:455–463.
Immune Defic Syndr 2008; 49:77–83. 63. Modjarrad K, Chamot E, Vermund SH. Impact of small reduc-
48. Nagot N, Ouedraogo A, Foulongne V, Konate I, Weiss HA, tions in plasma HIV RNA levels on the risk of heterosexual
Vergne L, et al. Reduction of HIV-1 RNA levels with therapy to transmission and disease progression. AIDS 2008; 22:2179–
suppress herpes simplex virus. N Engl J Med 2007; 356:790– 2185.
799. 64. Lingappa JR, Hughes JP, Wang RS, Baeten JM, Celum C, Gray
49. Schacker T, Zeh J, Hu H, Shaughnessy M, Corey L. Changes in GE, et al. Estimating the impact of plasma HIV-1 RNA reduc-
plasma human immunodeficiency virus type 1 RNA associated tions on heterosexual HIV-1 transmission risk. PLoS One 2010;
with herpes simplex virus reactivation and suppression. J Infect 5:e12598.
Dis 2002; 186:1718–1725.
65. Fraser C, Hollingsworth TD, Chapman R, de Wolf F, Hanage
50. Zuckerman RA, Lucchetti A, Whittington WL, Sanchez J,
WP. Variation in HIV-1 set-point viral load: epidemiological
Coombs RW, Zuniga R, et al. Herpes simplex virus (HSV)
analysis and an evolutionary hypothesis. Proc Natl Acad Sci
suppression with valacyclovir reduces rectal and blood plasma
U S A 2007; 104:17441–17446.
HIV-1 levels in HIV-1/HSV-2-seropositive men: a randomized,
double-blind, placebo-controlled crossover trial. J Infect Dis 66. Renia L, Potter SM. Co-infection of malaria with HIV: an
2007; 196:1500–1508. immunological perspective. Parasite Immunol 2006; 28:589–
51. Mayaud P, Legoff J, Weiss HA, Gresenguet G, Nzambi K, 595.
Bouhlal H, et al. Impact of acyclovir on genital and plasma 67. Lingappa JR, Baeten JM, Wald A, Hughes JP, Thomas KK,
HIV-1 RNA, genital herpes simplex virus type 2 DNA, and ulcer Mujugira A, et al. Daily acyclovir for HIV-1 disease progression
healing among HIV-1-infected African women with herpes in people dually infected with HIV-1 and herpes simplex virus
ulcers: a randomized placebo-controlled trial. J Infect Dis type 2: a randomised placebo-controlled trial. Lancet 2010;
2009; 200:216–226. 375:824–833.
52. Paz-Bailey G, Sternberg M, Puren AJ, Markowitz LE, Ballard R, 68. Zhu J, Hladik F, Woodward A, Klock A, Peng T, Johnston C, et al.
Delany S, et al. Improvement in healing and reduction in HIV Persistence of HIV-1 receptor-positive cells after HSV-2 reac-
shedding with episodic acyclovir therapy as part of syndromic tivation is a potential mechanism for increased HIV-1 acquisi-
management among men: a randomized, controlled trial. Infect tion. Nat Med 2009; 15:886–892.
Dis 2009; 200:1039–1049. 69. Del Amo J, Perez-Hoyos S, Hernandez Aguado I, Diez M,
53. Lopez-Gatell H, Cole SR, Margolick JB, Witt MD, Martinson J, Castilla J, Porter K. Impact of tuberculosis on HIV disease
Phair JP, Jacobson LP. Effect of tuberculosis on the survival of progression in persons with well documented time of HIV
HIV-infected men in a country with low tuberculosis inci- seroconversion. J Acquir Immune Defic Syndr 2003; 33:184–
dence. AIDS 2008; 22:1869–1873. 190.
54. Whalen CC, Nsubuga P, Okwera A, Johnson JL, Hom DL, 70. Straetemans M, Bierrenbach AL, Nagelkerke N, Glaziou P,
Michael NL, et al. Impact of pulmonary tuberculosis on survival van der Werf MJ. The effect of tuberculosis on mortality in
of HIV-infected adults: a prospective epidemiologic study in HIV positive people: a meta-analysis. PLoS One 2010;
Uganda. AIDS 2000; 14:1219–1228. 5:e15241.
55. Kizza HM, Rodriguez B, Quinones-Mateu M, Mirza M, Aung H, 71. Hewitt K, Steketee R, Mwapasa V, Whitworth J, French N.
Yen-Lieberman B, et al. Persistent replication of human Interactions between HIV and malaria in nonpregnant adults:
immunodeficiency virus type 1 despite treatment of pulmonary evidence and implications. AIDS 2006; 20:1993–2004.
tuberculosis in dually infected subjects. Clin Diagn Lab 72. Slutsker L, Marston BJ. HIV and malaria: interactions and
Immunol 2005; 12:1298–1304. implications. Curr Opin Infect Dis 2007; 20:3–10.
56. Hung CC, Chen MY, Hsiao CF, Hsieh SM, Sheng WH, Chang 73. Baeten JM, Kahle E, Lingappa JR, Coombs RW, Delany-
SC. Improved outcomes of HIV-1-infected adults with tuber- Moretlwe S, Nakku-Joloba E, et al. Genital HIV-1 RNA predicts
culosis in the era of highly active antiretroviral therapy. AIDS risk of heterosexual HIV-1 transmission. Sci Transl Med 2011;
2003; 17:2615–2622. 3:77ra29.

Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.