Global burden of transmitted HIV drug resistance and

HIV-exposure categories: a systematic review and

meta-analysis
Quang D. Phama,b, David P. Wilsona, Matthew G. Lawa,
Anthony D. Kellehera and Lei Zhanga
Objectives: Our aim was to review the global disparities of transmitted HIV drug
resistance (TDR) in antiretroviral-naive MSM, people who inject drugs (PWID) and
heterosexual populations in both high-income and low/middle-income countries.
Design/methods: We undertook a systematic review of the peer-reviewed English
literature on TDR (19992013). Random-effects meta-analyses were performed to pool
TDR prevalence and compare the odds of TDR across at-risk groups.
Results: A total of 212 studies were included in this review. Areas with greatest TDR
prevalence were North America (MSM: 13.7%, PWID: 9.1%, heterosexuals: 10.5%);
followed by western Europe (MSM: 11.0%, PWID: 5.7%, heterosexuals: 6.9%) and
South America (MSM: 8.3%, PWID: 13.5%, heterosexuals: 7.5%). Our data indicated
disproportionately high TDR burdens in MSM in Oceania (Australia 15.5%), eastern
Europe/central Asia (10.2%) and east Asia (7.8%). TDR epidemics have stabilized in
high-income countries, with a higher prevalence (range 10.912.6%) in MSM than in
PWID (5.28.3%) and heterosexuals (6.49.0%) over 19992013. In low/middleincome countries, TDR prevalence in all at-risk groups in 20092013 almost doubled
than that in 20042008 (MSM: 7.8 vs. 4.2%, P 0.011; heterosexuals: 4.1 vs. 2.6%,
P < 0.001; PWID: 4.8 vs. 2.4%, P 0.265, respectively). The risk of TDR infection was
significantly greater in MSM than that in heterosexuals and PWID. We observed
increasing trends of resistance to non-nucleoside reverse transcriptase and protease
inhibitors among MSM.
Conclusion: TDR prevalence is stabilizing in high-income countries, but increasing in
low/middle-income countries. This is likely due to the low, but increasing, coverage of
antiretroviral therapy in these settings. Transmission of TDR is most prevalent among
2014 Wolters Kluwer Health | Lippincott Williams & Wilkins
MSM worldwide.

AIDS 2014, 28:27512762

Keywords: heterosexual people, HIV, MSM, people who inject drugs,
transmitted drug resistance

Introduction
We live in an era in which antiretroviral drugs have become
widely available for people living with HIV (PLHIV)
globally. However, the widespread use of antiretroviral
therapy (ART) over time (Fig. 1) has contributed to the

emergence of epidemics of transmitted HIV drug resistance (TDR) [1]. The spread of TDR can substantially
reduce therapeutic choices of ARTand increase the chance
of virological treatment failure among undiagnosed
patients on treatment [25]. Consequently, further
accumulation of drug-resistant mutations may exhaust

The Kirby Institute, University of New South Wales, Sydney, New South Wales, Australia, and bDepartment for Disease Control
and Prevention, The Pasteur Institute, Ho Chi Minh City, Vietnam.
Correspondence to Lei Zhang, PhD, The Kirby Institute, University of New South Wales, Level 6, Wallace Wurth Building, UNSW
Australia, Sydney, NSW 2052, Australia.
Tel: +61 2 9385 0869; fax: +61 2 9385 0920; e-mail: lzhang@kirby.unsw.edu.au
Received: 18 June 2014; revised: 5 September 2014; accepted: 16 September 2014.
DOI:10.1097/QAD.0000000000000494

ISSN 0269-9370 Q 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins

2751

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

AIDS

2014, Vol 28 No 18
Global scale-up of ART for PLHIV
in LMIC with the 3 by 5 initiative
Primary infection with zidovudineresistant HIV strains was first
diagnosed in a homosexual man

WHO and US CDC piloted a strategy for prevention

and assessment of HIV drug resistance in 2004 and
recommended it worldwide in 2008
A Swiss team warned of
an emerging epidemic of
HIV drug resistance in
PLHIV in Switzerland

Shafer et al. published the first

list of HIV-1 protease and
reverse transcriptase mutations
for surveillance in 2007 and
WHO updated it in 2009

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

There is strong evidence

of TDR contributing to
virological failure in
both HIC (2011) and
LMIC (2012/13)

1993

1992

1991

1990

1989

1988

First antiretroviral drug (zidovudine)

was approved for use by US FDA

1987

2752

A global review indicated an

increasing TDR prevalence in
untreated people in LMIC

HAART was introduced

The global strategy for the

surveillance and monitoring of
HIV drug resistance was revised

Mutations resisting to zidovudine were

first reported in patients receiving
prolonged zidovidine monotherapy

US and UK initiated drug resistance

testing as a part of routine care for
newly HIV diagnosed people
Expanding TDR epidemics were
documented in both European and
North American settings

WHO and UNAIDS launched the

treatment 2.0 framework for action in
2011, which was scaled-up in 2013

Fig. 1. Timeline of the development of transmitted HIV drug resistance and responses. ART, antiretroviral therapy; CDC, Centers
for Disease Control and Prevention; HIC, high-income countries; LMIC, low/middle-income countries; PLHIV, people living with
HIV; TDR, transmitted HIV drug resistance; UNAIDS, Joint United Nations Programme on HIV/AIDS; US FDA, United States Food
and Drug Administration.

available drug options and lead to a poorer prognosis for

PLHIV and transmission of TDR strains to the wider
community [6].
There has been a strong and growing interest in
understanding the global trend of TDR epidemics in
recent years [7]. In western Europe, about 35.3% of PLHIV
are infected through homosexual exposure, a substantial
proportion through heterosexual contacts (41.7%) and a
smaller proportion through sharing of injecting equipment
(5.1%) [8]. Developed countries have reported stabilizing
TDR epidemics among the HIV-infected population who
are antiretroviral-naive [9], but whether equilibrium has
been reached in specific at-risk populations remains
unclear. In contrast, low/middle-income countries in East
Africa, parts of Asia and South America have experienced
rapidly increasing TDR prevalence levels over the past
decade [1012]. Reports in these settings have also shown
a disproportionate burden of TDR among MSM and
people who inject drugs (PWID) in comparison with the
overall population of PLHIV (e.g. China: 15.0% in MSM
vs. 3.8% in all PLHIV [13,14]; India: 10.8% in PWID vs.
2.1% in HIV-infected pregnant women [15,16]; and
Kenya: 13.8% in PWID vs. 1.1% in all PLHIV [17,18]).
Despite the numerous published TDR studies, the global
disparities in TDR-related disease burdens among at-risk
populations are yet to be investigated.

The objectives of this review were to provide a

comprehensive assessment of the temporal and geographical trends of TDR globally. It also aimed to identify
the TDR-related disease burdens in various HIV at-risk
populations according to a countrys income categories,
to inform future HIV-care interventions.

Methods
Search strategy and selection criteria
We conducted a systematic review by searching available
electronic English literature on PubMed, Embase, Scopus
and Web of Science databases, according to the following
search strategy: HIV and (drug or antiretroviral or
antiviral or ARV or ART) and (resistance or resistant
or mutation or mutant) not (review or modelling).
We further searched the peer-reviewed abstracts from
major conferences including the International AIDS
Conference, the Conference on HIV Pathogenesis,
Treatment and Prevention and the International HIV
Drug Resistance Workshop during 20012013. Searches
were conducted between June and August 2013.
Quantitative TDR studies with any study designs that
recruited participants at any stage of HIV infection were
eligible for inclusion in this review. These were screened

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

Global transmitted drug resistance Pham et al.

against two inclusion criteria: infection with TDR had to

be confirmed by genotypic assays in an antiretroviralnave adult cohort and the sample size of successful
genotyping must be greater than 30 for the overall study
and 10 for subpopulations classified by HIV transmission
modes; the prevalence rates or odds ratios (ORs) had to be
stratified by modes of HIV transmission, namely male-tomale sex, sharing of injection equipment, heterosexual
intercourse and others. The reference lists of eligible
publications were manually examined for further relevant
articles. We used the Newcastle-Ottawa Scale to assess the
quality of the eligible publications, marked from 1 (poor
quality) to 9 (excellent quality) [19]. We excluded
publications having a quality score of less than 3.

Data extraction
From each selected publication, a single unblinded
investigator (Q.D.P.) extracted relevant study-level
information [see more details in Table S3 (appendix),
http://links.lww.com/QAD/A588]. For each at-risk
group, we collected the numbers of samples sequenced,
the number of people diagnosed with TDR and resistance to nucleoside reverse transcriptase inhibitor
(NRTIs), non-nucleoside reverse transcriptase inhibitors
(NNRTIs) and protease inhibitors to calculate the
relevant prevalence levels. We categorized countries as
high-income and low/middle-income countries according to the 2012 classification of the World Bank [20].
Statistical data analysis
The distributions of TDR prevalence were highly rightskewed so we used the FreemanTukey double arcsine
transformation to normalize these rates before metaanalysis [21,22]. We conducted DerSimonianLaird
random-effects meta-analysis [23] to pool transformed
rates and 95% confidence intervals (CIs). We assessed
cross-study heterogeneity with both Cochrans Q tests
and the I2 statistics (values of I2 <25%, 2575% and
>75% indicating low, moderate and high heterogeneity,
respectively) [24]. In our exploratory analysis, we
identified moderate-to-high levels of heterogeneity
(MSM: Q test P < 0.001, I2 77.8%; PWID:
P < 0.001, I2 66.7%; heterosexuals: P < 0.001,
I2 76.3%) and conducted meta-regression analysis to
identify sources of these heterogeneities. We identified
that large geographical and country income levels
influenced between-study heterogeneity in transformed
prevalence rates in all at-risk populations. Therefore, we
then conducted subgroup analysis by these two factors.
Notably, stages of HIV infection for recruits, study
design, sampling locations and sampling method were not
significant contributors to heterogeneity. Stratified
analyses by drug class were also conducted when data
were available. We also compared TDR frequency across
key transmission modes of HIV. Odds ratios (ORs) were
calculated with a standard adjustment of 0.5 in the
contingency table if it contained a zero value. The effect
rates of odds of TDR and resistance to specific drug

2753

classes were estimated with random-effects models

stratified by countrys income category. Publication bias
was assessed by the methods of Egger et al. [25] and Begg
and Manzudar [26]. Statistical analyses were carried out
by Comprehensive Meta-Analysis 2.0 (Biostat, Englewood, New Jersey, USA) and Stata 12.0 (StataCorp,
College Station, Texas, USA).

Results
Study characteristics
A comprehensive search of 17 866 publications resulted in
212 selected studies (Fig. S1, appendix, http://links.lww.com/QAD/A588). Overall, information on antiretroviral drug resistance of 84 595 antiretroviral-naive PLHIV,
including 37 846 MSM (89.2% from high-income
countries), 6531 PWID (81.3% from high-income
countries) and 26 934 heterosexuals (70.7% from highincome countries), were extracted from samples in 46
low/middle-income countries (94 studies) and 34 highincome countries (118 studies). The median numbers of
sequenced samples in MSM, PWID and heterosexuals
were 110 (range 103220), 46 (range 101358) and 69
persons (range 112874), respectively. About onequarter of the studies (56/212) recruited participants at
an early (acute/recent) stage of HIV infection and men
were predominant (
50% of the samples) in over 80% of
the studies (160/198). About 70% (148/212) of the
studies reported the mean or median age of participants
(range 1844 years). The mean or median CD4 T-cell
count and viral load at recruitment were within the ranges
of 56807 cells/ml in 116 studies and of 2.66.4
log10 copies/ml in 106 studies when data were available,
respectively. In 74/165 studies (45%), 50% or more of the
participants were infected with HIV non-B subtypes.
This was mostly observed among studies in low/middleincome countries. Most samples used traditional population-based genotypic assays (201/212 studies) and
nearly half (93/212) of the selected studies interpreted
detected drug resistance-associated mutations according
to the TDR mutational lists of Shafer et al. [27] or the
WHO [28]. Study-level characteristics stratified by time
period, country income and region are further summarized in Table 1.
Global disease burdens of transmitted HIV drug
resistance
Disease burdens of TDR varied substantially across
geographical regions (Fig. 2). North America exhibited
consistently high TDR prevalence in all populations [MSM: 13.7% (95% CI 12.315.2%); PWID:
9.1% (8.010.2%); heterosexuals: 10.5% (8.612.4%)],
followed by western Europe [MSM: 11.0% (9.912.2%);
PWID: 5.7% (4.56.9%); heterosexuals: 6.9% (5.68.3%)]
and South America [MSM: 8.3% (6.010.9%); PWID:
13.5% (7.620.4%); heterosexuals: 7.5% (6.38.8%)].
Significant publication biases were found in TDR

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

146 [113220], (37)

84 [10359], (24)

195 (1)
141 [261778], (12)
78 [15104], (7)
368 [270466], (2)

40
30
31
1

16
18

74 [22116], (6)
35 [26165], (6)

66 [191358], (18)
16 [1065], (5)
58 (1)
23 (1)

39 [12348], (31)
44 [1883], (10)

117 (23596), (11)

54 (141025), (12)

77 [151516], (22)
85 [15299], (24)
62 [32354], (30)
35 (1)

78 [112874], (46)
25 [1467], (10)

2/2 (100%)

16/16 (100%)
11/15 (73%)

38/39 (97%)
24/28 (86%)
3/31 (10%)
1/1 (100%)

51/52 (98%)
14/14 (100%)

3539 (2)

3144 (14)
2039 (12)

2043 (25)
2439 (20)
1844 (20)
34 (1)

2839 (41)
2439 (13)

2043 (84)
1844 (64)

2/2 (100%)

1/16 (6%)
7/18 (39%)

15/40 (38%)
4/30 (13%)
4/31 (13%)
0/1 (0%)

20/58 (34%)
3/16 (19%)

41/118 (35%)
15/94 (16%)

12/21 (57%)
19/64 (30%)
25/127 (20%)

128 [121487], (53)

48 [10120], (7)

109/111 (98%)
51/87 (59%)

2839 (14)
1939 (40)
1844 (94)

58
16

71 [112874], (83)
65 [141025], (73)

16/17 (94%)
47/58 (81%)
97/123 (79%)

128 [113220], (107) 51 [121358], (53)

83 [10359], (36)
37 [10165], (25)

33 [14172], (9)
53 [12490], (50)
79 [112874], (97)

118
94

52 [1483], (7)
44 [12206], (26)
48 [101358], (45)

40 [11270], (16)
118 [151487], (43)
125 [103220], (84)

Range mean
or median
years of
age
(studies)

21
64
127

No. studies
with
50%
male
patients/total
studies (%)

110 [103220], (143) 46 [101358], (78) 69 [112874], (156) 160/198 (81%) 1844 (148) 56/212 (26%)

Median [minmax]
no. heterosexual
samples genotyped
per study (studies)

212

PWID, people who inject drugs.

All studies
Published period
19992003
20042008
20092013
Country income
High-income
Low/middleincome
Region
Western Europe
Eastern Europe
and central
Asia
North America
South America
Sub-Saharan Africa
North Africa and
Middle East
East Asia
South and southeast Asia
Oceania

No.
study

Median [minmax]
no. PWID samples
genotyped
per study (studies)

No. studies
recruiting
patients at
acute or
recent
infection/
total studies
(%)

0/2 (05)

9/16 (56%)
16/16 (100%)

0/18 (0%)
5/23 (22%)
29/29 (100%)
0/1 (0%)

6/46 (13%)
9/14 (64%)

10/82 (12%)
64/83 (77%)

2/14 (14%)
19/46 (41%)
53/105 (50%)

74/165 (45%)

No. studies
with
50%
patients
infected
HIV-1 subtype
non-B/total
studies (%)

Range mean
or median
level of
viral load,
log10
copies/ml
(studies)

524 (1)

206409 (9)
56614 (10)

220602 (18)
179560 (17)
285656 (12)
350 (1)

230807 (41)
261417 (7)

139807 (72)
56656 (44)

283807 (8)
139678 (31)
56656 (77)

6.4 (1)

4.65.5 (8)
3.85.9 (7)

4.16.1 (18)
3.65.4 (15)
2.65.7 (8)

4.25.6 (41)
4.55.3 (8)

4.16.4 (73)
2.65.7 (33)

4.36.4 (8)
3.85.6 (30)
2.66.1 (68)

56807 (116) 2.66.4 (106)

Range mean
or median
no. CD4
T-cell
count,
cells/ml
(studies)

AIDS

Median [minmax]
no. MSM samples
genotyped
per study (studies)

Table 1. Summary of study-level characteristics of included literature.

2754
2014, Vol 28 No 18

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

Global transmitted drug resistance Pham et al.

2755

(a) TDR prevalence in MSM

Eastern Europe and Central Asia
10.2% (4.817.0%)*
North America
13.7% (12.315.2%)

Western Europe
11.0% (9.912.2%)
East Asia
7.8% (5.410.6%)

South America
8.3% (6.010.9%)

South and Southeast Asia

2.8% (1.05.3%)

Country prevalence
5.0%
5.0110.0%
10.0115.0%
15.0120.0%
>20.0%
Study conducted, data not reported
No data

Oceania
15.5% (12.818.3%)

(b) TDR prevalence in PWID

Eastern Europe and Central Asia

0.5% (0.01.7%)
North America
9.1% (8.010.2%)

Western Europe
5.7% (4.56.9%)
East Asia
2.7% (1.24.7%)

South America
13.5% (7.620.4%)

South and Southeast Asia

2.4% (0.45.4%)

Country prevalence
5.0%
5.0110.0%
10.0115.0%
15.0120.0%
>20.0%
Study conducted, data not reported
No data

(c) TDR prevalence in heterosexuals

Eastern Europe and Central Asia

2.3% (0.25.8%)
North America
10.5% (8.612.4%)

Western Europe
6.9% (5.6 8.3%)
East Asia
4.5% (2.47.0%)

South America
7.5% (6.38.8%)
Country prevalence
5.0%
5.0110.0%
10.0115.0%
15.0120.0%
>20.0%
Study conducted, data not reported
No data

Sub-Saharan Africa
2.2% (1.43.2%)

South and Southeast Asia

2.3% (1.53.1%)

Fig. 2. Global prevalence of transmitted HIV drug resistance in HIV-infected antiretroviral-nave MSM, people who inject
drugs and heterosexual populations by regions (19992013). PWID, people who inject drugs; TDR, transmitted HIV drug
resistance. () Prevalence estimate was yielded from a meta-analysis of seven studies with only small numbers of MSM (10120
men).

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

2756

AIDS

2014, Vol 28 No 18

prevalence among heterosexuals in western Europe (Egger

test, P 0.003) and MSM in North America (Egger test,
P 0.047). In eastern Europe and central Asia, east Asia,
south and south-east Asian regions, where infections are
predominately transmitted through sharing of equipment,
there was a much lower TDR prevalence among PWID
(range 0.52.7%) and heterosexuals (range 2.34.5%).
TDR prevalence among heterosexuals in sub-Saharan
Africa was 2.2% (1.43.2%). Notably, Australia (in
Oceania) had the highest TDR prevalence among
MSM [15.5% (12.818.3%)] globally, whereas TDR
was also emerging in east Asia [7.8% (5.410.6%)]. We also
documented a disproportionately high prevalence of TDR
among MSM in eastern Europe and central Asia [10.2%
(4.817.0%)]. However, this figure must be interpreted
with caution as it was derived from a meta-analysis of seven
studies with only small numbers of MSM (range 10120
individuals). There has been limited information on TDR
in MSM and PWID in African and Middle East settings.
Argentina, Brazil, Canada, Germany, Poland, Spain and
United States were among the most TDR-affected
countries (Table S1, http://links.lww.com/QAD/A588).
Temporal trends of TDR epidemics in high-income and
low/middle-income settings are distinct. In high-income
countries, TDR epidemics have been established and have
potentially stabilized with a significantly higher prevalence
(10.912.6%) among MSM than PWID (5.28.3%) and
heterosexual people (6.49.0%) (Fig. 3b). Whereas widely
available potent ART options and regular drug resistance
testing may have contributed to stabilizing these epidemics
in high-income settings [29], the smaller but rapidly
emerging TDR epidemics in low/middle-income countries are alarming [1]. Data were limited in low/middleincome settings prior to 2004, but TDR prevalence across
all at-risk populations during 20092013 almost doubled
the levels during 20042008 (MSM: 7.8 vs. 4.2%,
P 0.011; heterosexuals: 4.1 vs. 2.6%, P < 0.001; PWID:
4.8 vs. 2.4%, P 0.265, respectively) (Fig. 3a). TDR
epidemics have most severely affected MSM in low/
middle-income countries, particularly in China and Cuba
(Fig. 2 and Table S1, http://links.lww.com/QAD/A588).
We compared the risk of TDR infection across all
transmission routes (Table 2). In a high-income context,
the odds of TDR infection was 28% higher in MSM than
in heterosexual persons [OR 1.28 (1.141.42)]; there was
potential presence of publication bias (Egger test,
P 0.006) and existence of moderate and significant
heterogeneity between studies (I2 38.4%, Q test
P < 0.001). Similarly, the corresponding risk was 42%
higher in MSM in low/middle-income settings compared
to that in heterosexual persons [OR 1.42 (1.111.82)];
heterogeneity between studies was low (I2 0.0%, Q test
P 0.551). Notably, 19 of 24 included studies concerned
South America and their findings may not be fully
representative for the actual risk for TDR among MSM in
low/middle-income settings worldwide. Differences in the

risk of TDR infection were relatively consistent between

MSM and heterosexuals in high-income countries over
19992013. In low/middle-income settings, although
differences were insignificant during 19992008, MSM
were reportedly more likely to harbour TDR compared to
heterosexual persons [OR 1.49 (1.131.99), I2 5.1%, Q
test P 0.394] during 20092013. We also observed a
significantly higher risk of TDR infection in MSM
compared to PWID in high-income countries [OR 1.32
(1.101.58), Egger test, P 0.003, I2 33.1%, Q test
P 0.017], but not in low/middle-income settings. Metaregression also suggested significant modification effects of
some study-level characteristics on the pooled risks of
TDR (Table S3S4, http://links.lww.com/QAD/A588).
In particular, seven studies in high-income countries
conducted before 1996 showed greater odds of TDR
infection in MSM compared with heterosexuals than more
recent studies [OR 2.46 (1.494.04) vs. 1.24 (1.111.39)].
In low/middle-income settings, after excluding one major
study (1340 participants) [30], the difference in the risk of
TDR between MSM and heterosexuals became insignificant [OR 1.20 (0.911.57)]. Interestingly, after
excluding six studies in these settings, which did not
interpret the frequency of resistance in PWID according to
the WHO standardized drug resistance mutation list,
the pooled odds of TDR was significantly higher in
PWID compared to that in heterosexual people [OR 2.11
(1.173.79)].

Transmitted HIV drug resistance to various

antiretroviral drug classes
The pooled prevalence estimates of NRTI-resistant strains
in MSM, based on available data regardless of geographical region, significantly decreased from 6.7% during
19992003 to 4.1% during 20042008 and 5.0% during
20092013 (P 0.045), but the prevalence of NNRTI/
protease inhibitor-resistant strains significantly increased
(NNRTI: 1.9, 2.3 and 4.1%, P < 0.001; pro-tease inhibitor: 0.6, 1.1 and 1.8%, P < 0.001, respectively) (Fig. 3c).
Among PWID and heterosexual populations, TDR
prevalence in all drug classes remain low during 1999
2013 (PWID: 1.32.2, 01.5 and 0.20.8%; heterosexuals: 1.22.7, 1.02.0 and 0.11.9% for NNRTIs, NRTIs
and protease inhibitors, respectively) (Fig. 3d and e).

Discussion
We demonstrate a high but stable TDR prevalence across
MSM, PWID and heterosexuals in high-income
countries. Conversely, TDR levels are relatively low
but rapidly increasing in low/middle-income countries.
Notably, MSM is the population most affected by TDR
regardless of the countries economic status worldwide.
Transmitted HIV drug resistance prevalence is largely
heterogeneous across continental regions, particularly
between high-income and low/middle-income settings.

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

Global transmitted drug resistance Pham et al.

(a)

(b)

20.0

20.0

19992003
20042008
20092013

Resistance prevalence (%)

Resistance prevalence (%)

2757

15.0

P = 0.011
10.0
P = 0.265

7.8

P < 0.001
4.8

5.0

4.2

4.1
2.6

2.4
No data

0.0

No data

625 3,346

MSM

447

No data 1,199 6,420

700

PWID

19992003
20042008
20092013

P = 0.296

10.9

10.0

9.0
8.3

NRTI

P = 0.045

2,002
4,230

3,259
2,446

779 7,690 23,264

0.0

1,257

20092013

730

229 1,232 3,394

238 3,52713,721

PWID

MSM

Heterosexuals

(e)

n
251

n
399

581

2,275

938

5,023

5,514

19992003
20042008

5.6

5.2

5.0

715
P < 0.001

NNRTI

8.0
6.4

(d)
n
730

P = 0.296

12.6

11.7

Heterosexuals

(c)

P = 0.121

15.0

19992003

P = 0.024

19992003
891

431
20042008
624

20042008
1,292

20092013

251

399

2,202

581

2,275

4,230

807

5,023

20092013

P < 0.001

PI

10

10

10

Resistance prevalence (%)

Fig. 3. Prevalence of transmitted HIV drug resistance in HIV-infected antiretroviral-naive MSM, people who inject drugs and
heterosexual people by time periods. (a) and (b) present prevalence levels in low/middle-income and high-income countries,
respectively, in which P values were obtained from Q tests. (ce) present prevalence of resistance to antiretroviral drug classes.
Each column chart series represents the prevalence trends of resistance to a certain antiretroviral drug class (NRTI, NNRTI and PI)
among MSM [represented in (c)], people who inject drugs [PWID, represented in (d)] and heterosexual people [represented in (e)]
using data from three periods (19992003, 20042008 and 20092013). P values were obtained from meta-regression analysis
examining the slopes of resistance prevalence levels over three time periods. Sample size for each meta-analysis is presented at the
bottom of each column. NRTI, nucleoside reverse transcriptase inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor;
PI, protease inhibitor.

Geographical difference is mostly associated with the

circulation of acquired HIV drug resistance, which is
largely dependent on ART coverage, drug regimens used,
clinical experience and patient adherence on treatment.
High ART coverage (>75%) among treatment-eligible
PLHIV has been consistently reported for years in highincome settings [31]. In contrast, just over one-third
(34%) of the estimated 28.6 million people eligible for
ART in low/middle-income countries were on ART by
2012 [32]. The ART programmes implemented in low/
middle-income settings, consisting of potent combination
therapy, are more effective in suppressing in-vivo viral

replication and limiting the selection of drug-resistant

strains than the primitive single/dual therapies introduced
in high-income settings in the early stages of their ART
programmes. It is therefore anticipated that the population
levels of acquired antiretroviral resistance in low/middleincome settings are substantially lower than those in the
high-income settings. The low levels of acquired drug
resistance have limited the transmission of drug-resistant
strains in these settings [33,34]. However, an increase in
TDR prevalence levels across populations in low/middleincome settings has been reported here, likely as a result of a
number of factors including increased ART coverage, late

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

1.35
1.38
1.24
1.28
0.91
0.97
1.49
1.42

22/204
260/3333
968/10 808
1250/14 345
15/172
19/446
147/2688
181/3306

Heterosexuals

(0.342.44)a, n/a
(0.412.28), 0.0%
(1.131.99)b, 5.1%
(1.111.82)c, 0.0%

(0.702.58), 0.0%
(1.121.70)b, 25.7%
(1.081.42)b,e, 49.0%
(1.141.42)d,e,h, 38.4%

OR (95% CI), I2

6/75

75/934
81/1009

4/17
678/5474
1890/15 754
2572/21 245

MSM

0/14

12/172
12/186

5/25
107/1176
279/3161
391/4 362

PWID
(0.285.45)a, n/a
(1.061.70)b, 0.0%
(0.991.72)f,h, 51.2%
(1.101.58)c,f,h, 33.1%

2.71 (0.1450.87)a, n/a

0.80 (0.401.60), 0.0%

0.85 (0.441.67), 0.0%

1.23
1.34
1.31
1.32

OR (95% CI), I2

0/14
1/66
27/470
28/550

16/96
101/1121
272/3029
389/4 246

PWID

15/172
8/229
91/2267
114/2668

9/54
219/2641
808/8945
1036/11 640

Heterosexuals

0.35
0.41
1.66
1.45

1.07
1.08
0.97
1.01

(0.026.16)a, n/a
(0.053.13), 0.0%
(0.982.78), 0.0%
(0.892.39), 0.0%

(0.392.94), 0.0%
(0.811.43), 0.0%
(0.791.19)g, 25.1%
(0.881.17)g, 5.9%

OR (95% CI), I2

No. patients with

TDR/total patients

CI, confidence interval; n/a, not applicable; OR, odds ratio; PWID, people who inject drugs; TDR, transmitted HIV drug resistance.
a
Estimate was obtained from a single study.
b
P value < 0.05.
c
P value < 0.01.
d
P-value < 0.001.
e
This also included four studies (1601 MSM and 1557 heterosexual people) reporting only crude odds ratios and its 95% confidence intervals.
f
This also included two studies (361 MSM and 42 PWID) reporting only crude odds ratios and its 95% confidence intervals.
g
This also included two studies (413 PWID and 1445 heterosexual people) reporting only crude odds ratios and its 95% confidence intervals.
h
Signficant potential presence of publication bias: MSM versus heterosexuals over the entire study period (Eggers test, P 0.006; Begg and Mazumdar test, P 0.138) and MSM versus PWID over
20092013 (Eggers test, P 0.027; Begg and Mazumdar test, P 0.134), as well as the entire study period (Eggers test, P 0.003; Begg and Mazumdar test, P 0.125) in high-income countries.

High-income countries
19992003
24/194
20032008
755/6418
20092013 2094/18 766
Overall
2873/25 378
Low/middle-income countries
19992003
6/75
20032008
10/228
20092013
182/2130
Overall
198/2433

MSM

No. patients with

TDR/total patients

AIDS

No. patients with

TDR/total patients

Table 2. Comparison of pooled odds ratios of transmitted HIV drug resistance infection across HIV-infected antiretroviral-nave MSM, people who inject drugs and heterosexual populations by
country-income category and published period.

2758
2014, Vol 28 No 18

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

Global transmitted drug resistance Pham et al.

treatment initiation, limited availability of antiretroviral

regimens, frequent interruption of drug supply, suboptimal
adherence and absence of viral load monitoring and
therefore late diagnosis of treatment failure [1,35].
Consistent with prior observational studies [36,37], this
review confirms a disproportionately higher risk of TDR
among MSM compared with other populations. Antiretroviral-naive but TDR-infected homosexual partners of
MSM are the primary source of TDR variants [38,39], via
unprotected sexual activities during the early stage of HIV
infection when HIV is more likely to be undiagnosed and
may have TDR variants [40]. Another source of HIV drug
resistance among MSM is from ART-treated patients, who
have developed acquired drug-resistant strains caused by
interruption to therapy due to substance abuse such as
methamphetamines [41]. However, strong biological
support for these links remains undefined.
We found that over the past decade, TDR levels in MSM
remained at moderate-to-high levels in both low/middleincome and high-income areas, emphasizing an urgent
need to promote MSM-targeted TDR surveillance, prevention and treatment guidelines in settings where
infections are primarily transmitted among this group.
Importantly, alarming surges of NNRTI and protease
inhibitor drug resistance have been observed. Enhancement of existing pre-therapy drug resistance testing is of
great importance for MSM living with HIV in highincome countries. A study reported that less than half
(41.7%) of the MSM in New York City had been tested for
antiretroviral drug resistance at diagnosis [42], despite a
wide availability of HIV genotyping for PLHIV in the
United States [43]. This suggests that providing the service
for MSM at the time of diagnosis remains a significant
challenge.
In accordance with a previous meta-analysis [44], the risk
of TDR was not significantly different between PWID
and heterosexual people. Notably, efforts to determine
whether injecting drug use is associated with an elevated
risk of TDR have failed in studies conducted in low/
middle-income countries with high prevalence of
injecting drug use. Of the included studies in low/
middle-income countries, in which detected TDR
patterns were interpreted according to the WHO
standardized mutational list, significantly higher odds of
TDR infection were found among PWID. This finding
adds to current global recommendations which stated that
this list should be used to interpret the frequency of TDR
and a separate TDR survey should be conducted for
PWID in areas of the world where PWID are thought to
be at a higher risk of TDR infection [45].
A major limitation of this review is related to the
significant heterogeneities across the selected studies. In
the course of this study, detailed subgroup analyses of
TDR disease burden according to key HIV transmission

2759

modes in different geographical areas and country income

levels have been conducted. Despite this, cross-study
heterogeneities in TDR frequencies remain significant.
This may further attribute to the variations in study
design and sample characteristics of the eligible articles.
Although little evidence of publication bias was
discovered, we acknowledge the possible under-representation of data that was not published due to negative
results or language criteria used to select the literature
reviewed herein. Sutton et al. [46] additionally indicated
that most methods for assessing potential publication bias
in a systematic review have low statistical power, so we
may have underestimated its effect on this analysis.
The current understanding of drug resistance-associated
mutations is heavily based on research among HIV-1 B
subtype-infected persons in high-income countries [47],
and many studies included in this review were conducted
before the first standardized mutational list for TDR
surveillance became available globally in 2007 [27]. The
inability to adjust the current analysis for studies that
recruited people infected with non-B subtypes of HIV-1
or used different TDR mutational lists may lead to biases
in estimating the frequency of TDR [48,49].
The reported findings have limited generalizability
globally. Our strict inclusion criteria may have excluded
important TDR studies with a large sample size, in which
TDR-prevalence levels were unavailable for different HIV
risk categories. In contrast, some studies with a small
number of sub-populations were included in this review.
Studies recruiting MSM participants may be more prone to
include clusters of recent infections, leading to a higher
chance of detecting TDR variants with standard population-based sequencing than studies recruiting other
subgroups.
The lack of data on risk behaviours and resistance to
particular drug classes in key affected populations limited
our attempts to more definitively define the transmission
patterns of TDR. The overlapping self-reporting risk
behaviours among the recruits may lead to the
miscategorization of the infection route, leading to biases
in examining the risk of TDR infection. Generally, a
substantial proportion of HIV-infected MSM also inject
drugs [50,51], but this important information was
available in only six of 143 (4%) included studies on
MSM. The dearth of individual-level data on risk-taking
behaviours and biological characteristics further limited
the possibility to identify underlying factors for differences in infection risk observed across at-risk populations
in this review. With limited data on TDR by antiretroviral
drug class across HIV transmission modes, we were not
able to investigate differences in infection risks according
to drug classes. The absence of specific information on
mutational patterns and the resistance level induced by
the mutations limited our investigation in the trend and
severity of the TDR mutations over time.

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

2760

AIDS

2014, Vol 28 No 18

Despite improvements in managing PLHIV entering the

ART programmes in most countries in recent years, a
large information gap remains in estimating the number
of PLHIV eligible for ART. This poses major difficulties
in providing an accurate assessment of treatment coverage
over time among various at-risk populations. As a result,
this analysis could not adjust for ART coverage despite its
likely association with TDR prevalence [1].
Our findings are likely to have important implications for
the continued roll-out of ARTworldwide. In high-income
settings, with a high background level of resistance, pretherapy HIV genotyping has been recommended for all
PLHIV to guide the choice of antiretroviral regimens
[52,53]. Ideally it should also be implemented elsewhere,
but based on resource availability and the local TDR
epidemiological context. Nevertheless, given a higher risk
of TDR among MSM, available resources should be
prioritized to provide targeted HIV genotyping test for this
population. An optimal therapy based on genotypic
resistance results, along with existing viral load monitoring
strategy over the course of treatment, could limit the
emergence of acquired HIV drug resistance and help to
curb the subsequent transmission of TDR. On the
contrary, due to constraints in laboratory infrastructure
and financial revenues in most low/middle-income
countries, the inclusion of both routine drug resistance
testing prior to ART initiation and viral load monitoring
during ART for all PLHIV is unlikely in the near future.
Containing the expanding TDR epidemics in these
settings is challenging. Clearly, further representative
studies, financial commitments and innovative and costeffective interventions, such as a newly developed
technique to determine drug resistance in dried blood
spots [54,55], are essential for any effective prevention of
TDR in these settings. Expansion of existing harm
reduction programmes may indicate the need for MSM
receiving ART in both low/middle and high-income
contexts to enhance their safe sexual practices and reduce
the possible transmission of their acquired drug-resistant
strains [56]. Moreover, standardized guidelines are required
for collecting behavioural information among PLHIV
recruited in TDR surveillance surveys and their partners.
Though data confirm the efficacy and effectiveness of
treatment as prevention [5759], other data suggest that
TDR is one element that could limit the efficacy of this
approach in the near future when a broader roll-out of this
strategy is employed [60,61]. This strongly suggests the
need for ongoing surveillance of TDR in all settings to
regularly update knowledge of the levels of TDR and to
ensure timely initiation of therapies which effectively
suppress viral load.

Acknowledgements
The authors wish to thank Louisa Wright for assistance
with English language editing and proofreading, and

Andrew Craig for assistance with figures. The Kirby

Institute is funded by the Australian Government,
Department of Health and Ageing. The views expressed
in this publication do not necessarily represent the position
of the Vietnamese or the Australian Governments. The
Kirby Institute is affiliated with UNSW Australia.
Q.D.P. led the design of the study, collected data,
performed all statistical analysis, interpreted the data and
wrote the first draft of the manuscript. L.Z., D.P.W.,
M.G.L. and A.D.K. assisted in the interpretation of
findings, reviewed the manuscript and contributed critical
revision of the manuscript. L.Z. and D.P.W. overviewed
the study. All authors have approved the final version.
Funding: This study was conducted as part of the PhD
programme of QDP and was not funded by any particular
organization.
Q.D.P. is supported by an AusAID PhD scholarship.
D.P.W.s work was funded in part by the Australian
National Health and Medical Research Council.

Conflicts of interest
The authors declare that there is no conflict of interest.

References
1. Gupta RK, Jordan MR, Sultan BJ, Hill A, Davis DH, Gregson J,
et al. Global trends in antiretroviral resistance in treatmentnaive individuals with HIV after rollout of antiretroviral treatment in resource-limited settings: a global collaborative study
and meta-regression analysis. Lancet 2012; 380:12501258.
2. Wittkop L, Gunthard HF, de Wolf F, Dunn D, Cozzi-Lepri A, de
Luca A, et al. Effect of transmitted drug resistance on virological
and immunological response to initial combination antiretroviral therapy for HIV (EuroCoord-CHAIN joint project): a European
multicohort study. Lancet Infect Dis 2011; 11:363371.
3. Bertagnolio S, Perno CF, Vella S, Pillay D. The impact of HIV
drug resistance on the selection of first- and second-line art in
resource-limited settings. J Infect Dis 2013; 207:S45S48.
4. Phanuphak P, Sirivichayakul S, Jiamsakul A, Sungkanuparph S,
Kumarasamy N, Lee MP, et al. Transmitted drug resistance and
antiretroviral treatment outcomes in nonsubtype B HIV1- infected patients in South East Asia. J Acquir Immune Defic Syndr
2014; 66:7479.
5. Hamers RL, Schuurman R, Sigaloff KCE, Wallis CL, Kityo C,
Siwale M, et al. Effect of pretreatment HIV-1 drug resistance
on immunological, virological, and drug-resistance outcomes
of first-line antiretroviral treatment in sub-Saharan Africa:
a multicentre cohort study. Lancet Infect Dis 2012; 12:307
317.
6. Cambiano V, Bertagnolio S, Jordan MR, Lundgren JD, Phillips A.
Transmission of drug resistant HIV and its potential impact on
mortality and treatment outcomes in resource-limited settings.
J Infect Dis 2013; 207:S57S62.
7. Frentz D, Boucher CA, van de Vijver DA. Temporal changes in
the epidemiology of transmission of drug-resistant HIV-1
across the world. AIDS Rev 2012; 14:1727.
8. European Centres for Disease prevention and Control. HIV/AIDS
surveillance in Europe, 2012. http://www.ecdc.europa.eu/en/
publications/_layouts/forms/Publication_DispForm.aspx?List4f
55ad51-4aed-4d32-b960-af70113dbb90&ID971. [Accessed 9
August 2014]
9. Vercauteren J, Wensing AM, van de Vijver DA, Albert J, Balotta
C, Hamouda O, et al. Transmission of drug-resistant HIV-1 is
stabilizing in Europe. J Infect Dis 2009; 200:15031508.

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

Global transmitted drug resistance Pham et al.

10. Hamers RL, Sigaloff KC, Kityo C, Mugyenyi P, de Wit TF.
Emerging HIV-1 drug resistance after roll-out of antiretroviral
therapy in sub-Saharan Africa. Curr Opin HIV AIDS 2013;
8:1926.
11. Sohn AH, Srikantiah P, Sungkanuparph S, Zhang F. Transmitted
HIV drug resistance in Asia. Curr Opin HIV AIDS 2013; 8:27
33.
12. Pineda-Pena AC, Bello DC, Sussmann O, Vandamme AM,
Vercauteren J, van Laethem K, et al. HIV-1 transmitted drug
resistance in Latin America and the Caribbean: what do we
know? AIDS Rev 2012; 14:256267.
13. Zhang X, Li S, Li X, Li X, Xu J, Li D, et al. Characterization of
HIV-1 subtypes and viral antiretroviral drug resistance in men
who have sex with men in Beijing, China. AIDS 2007; 21 (Suppl
8):S59S65.
14. Liao L, Xing H, Shang H, Li J, Zhong P, Kang L, et al. The
prevalence of transmitted antiretroviral drug resistance in
treatment-naive HIV-infected individuals in China. J Acquir
Immune Defic Syndr 2010; 53 (Suppl 1):S10S14.
15. Iqbal HS, Solomon SS, Madhavan V, Solomon S, Balakrishnan
P. Primary HIV-1 drug resistance and polymorphic patterns
among injecting drug users (IDUs) in Chennai, Southern India. J
Int Assoc Physicians AIDS Care 2009; 8:323327.
16. Thorat SR, Chaturbhuj DN, Hingankar NK, Chandrasekhar V,
Koppada R, Datkar SR, et al. Surveillance of transmitted HIV
type 1 drug resistance among HIV type 1-positive women
attending an antenatal clinic in Kakinada, India. AIDS Res
Hum Retroviruses 2011; 27:12911297.
17. Osman S, Lihana RW, Kibaya RM, Ishizaki A, Bi X, Okoth FA,
et al. Diversity of HIV type 1 and drug resistance mutations
among injecting drug users in Kenya. AIDS Res Hum Retroviruses 2013; 29:187190.
18. Hassan AS, Mwaringa SM, Obonyo CA, Nabwera HM, Sanders
EJ, de Wit TFR, et al. Low prevalence of transmitted HIV type 1
drug resistance among antiretroviral-naive adults in a rural
HIV clinic in Kenya. AIDS Res Hum Retroviruses 2013; 29:129
135.
19. Wells G, Shea B, OConnell D, Peterson J, Welch V, Losos M,
et al. The Newcastle-Ottawa Scale (NOS) for assessing the
quality of nonrandomised studies in meta-analyses. 2000.
http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. [Accessed 20 July 2013]
20. The World Bank. World Bank list of economies. 2012. http://
www.healthsystemsglobal.org/Portals/0/files/World_bank_list_july2012.pdf. [Accessed 15 July 2013]
21. Freeman MF, Tukey JW. Transformations related to the angular
and the square root. Ann Math Statist 1950; 21:607611.
22. Miller JJ. The inverse of the FreemanTukey double arcsine
transformation. Am Stat 1978; 32:1381138.
23. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control
Clin Trials 1986; 7:177188.
24. Huedo-Medina TB, Sanchez-Meca J, Marn-Martnez F, Botella
J. Assessing heterogeneity in meta-analysis: Q statistic or I 2
Index? Psychol Methods 2006; 11:193206.
25. Egger M, Smith GD. Misleading meta-analysis. Br Med J 1995;
310:752754.
26. Begg CB, Mazumdar M. Operating characteristics of a rank
correlation test for publication bias. Biometrics 1994; 50:1088
1101.
27. Shafer RW, Rhee SY, Pillay D, Miller V, Sandstrom P, Schapiro
JM, et al. HIV-1 protease and reverse transcriptase mutations
for drug resistance surveillance. AIDS 2007; 21:215223.
28. Bennett DE, Camacho RJ, Otelea D, Kuritzkes DR, Fleury H,
Kiuchi M, et al. Drug resistance mutations for surveillance of
transmitted HIV-1 drug-resistance: 2009 update. PLoS One
2009; 4:e4724.
29. Jordan MR, Bennett DE, Wainberg MA, Havlir D, Hammer S,
Yang C, et al. Update on World Health Organization HIV drug
resistance prevention and assessment strategy: 20042011.
Clin Infect Dis 2012; 54 (Suppl 4):S245S249.
30. Kiertiburanakul S, Chaiwarith R, Sirivichayakul S, Ditangco R,
Jiamsakul A, Li PCK, et al. Comparisons of primary HIV-1 drug
resistance between recent and chronic HIV-1 infection within
a sub-regional cohort of Asian patients. PLoS One 2013; 8:.
31. Bollerup AR, Donoghoe MC, Lazarus JV, Nielsen S, Matic S.
Access to highly active antiretroviral therapy (HAART) in the
WHO European Region 20032005. Scand J Public Health
2008; 36:183189.

2761

32. Joint United Nations Programme on HIV/AIDS (UNAIDS).

Global report: UNAIDS report on the global AIDS epidemic
2013. http://www.unaids.org/en/media/unaids/contentassets/
documents/epidemiology/2013/gr2013/UNAIDS_Global_
Report_2013_en.pdf. [Accessed 3 March 2014]
33. Blower SM, Aschenbach AN, Gershengorn HB, Kahn JO. Predicting the unpredictable: transmission of drug-resistant HIV.
Nat Med 2001; 7:10161020.
34. Smith RJ, Okano JT, Kahn JS, Bodine EN, Blower S. Evolutionary
dynamics of complex networks of HIV drug-resistant strains:
the case of San Francisco. Science 2010; 327:697701.
35. Stadeli KM, Richman DD. Rates of emergence of HIV drug
resistance in resource-limited settings: a systematic review.
Antivir Ther 2013; 18:115123.
36. Descamps D, Assoumou L, Chaix M-L, Chaillon A, Pakianather
S, de Rougemont A, et al. National sentinel surveillance of
transmitted drug resistance in antiretroviral-naive chronically
HIV-infected patients in France over a decade: 20012011.
J Antimicrob Chemother 2013; 68:26262631.
37. Wang X, He C, Xing H, Liao L, Xu X, He J, et al. Emerging
transmitted HIV type 1 drug resistance mutations among
patients prior to start of first-line antiretroviral therapy in
middle and low prevalence sites in China. AIDS Res Hum
Retroviruses 2012; 28:16371639.
38. Brenner BG, Roger M, Moisi DD, Oliveira M, Hardy I, Turgel R,
et al. Transmission networks of drug resistance acquired in
primary/early stage HIV infection. AIDS 2008; 22:25092515.
39. Bezemer D, van Sighem A, Lukashov VV, van der Hoek L, Back
N, Schuurman R, et al. Transmission networks of HIV-1 among
men having sex with men in the Netherlands. AIDS 2010;
24:271282.
40. Marks G, Crepaz N, Senterfitt JW, Janssen RS. Meta-analysis of
high-risk sexual behavior in persons aware and unaware they
are infected with HIV in the United States: implications for HIV
prevention programs. J Acquir Immune Defic Syndr 2005;
39:446453.
41. Colfax GN, Vittinghoff E, Grant R, Lum P, Spotts G, Hecht FM.
Frequent methamphetamine use is associated with primary
nonnucleoside reverse transcriptase inhibitor resistance. AIDS
2007; 21:239241.
42. Torian LV, Forgione LA. Transmitted antiretroviral drug resistance in New York City, 20062010: the first five years of
routine genotype surveillance. J Acquir Immune Defic Syndr
2013; 63:e119e122.
43. Grant PM, Zolopa AR. The use of resistance testing in the
management of HIV-1-infected patients. Curr Opin HIV AIDS
2009; 4:474480.
44. Werb D, Mills EJ, Montaner JS, Wood E. Risk of resistance to
highly active antiretroviral therapy among HIV-positive injecting drug users: a meta-analysis. Lancet Infect Dis 2010; 10:464
469.
45. Bennett DE, Myatt M, Bertagnolio S, Sutherland D, Gilks CF.
Recommendations for surveillance of transmitted HIV drug
resistance in countries scaling up antiretroviral treatment.
Antivir Ther 2008; 13:2536.
46. Sutton AJ, Song F, Gilbody SM, Abrams KR. Modelling publication bias in meta-analysis: a review. Stat Methods Med Res
2000; 9:421445.
47. Wainberg MA, Zaharatos GJ, Brenner BG. Development of
antiretroviral drug resistance. N Engl J Med 2011; 365:637
646.
48. Martnez-Cajas JL, Pant-Pai N, Klein MB, Wainberg MA. Role of
genetic diversity amongst HIV-1 non-B subtypes in drug resistance: a systematic review of virologic and biochemical
evidence. AIDS Rev 2008; 10:212223.
49. Green H, Tilston P, Fearnhill E, Pillay D, Dunn DT. The impact
of different definitions on the estimated rate of transmitted HIV
drug resistance in the United Kingdom. J Acquir Immune Defic
Syndr 2008; 49:196204.
50. Marcellin F, Lorente N, Demoulin B, Carrieri MP, Suzan-Monti
M, Roux P, et al. Comparison of risk factors in HIV-infected
men who have sex with men, coinfected or not with hepatitis C
virus (ANRS VESPA2 French cross-sectional national survey).
Sex Transm Infect 2014(in press).
51. Ghanem A, Little S, Drumright L, Liu L, Morris S, Garfein R.
High-risk behaviors associated with injection drug use among
recently HIV-infected men who have sex with men in San
Diego, CA. AIDS Behav 2011; 15:15611569.

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

2762

AIDS

2014, Vol 28 No 18

52. Thompson MA, Aberg JA, Hoy JF, Telenti A, Benson C, Cahn P,
et al. Antiretroviral treatment of adult HIV infection: 2012
recommendations of the International Antiviral Society: USA
panel. JAMA 2012; 308:387402.
53. Sax PE, Islam R, Walensky RP, Losina E, Weinstein MC, Goldie
SJ, et al. Should resistance testing be performed for treatmentnaive HIV-infected patients? A cost-effectiveness analysis. Clin
Infect Dis 2005; 41:13161323.
54. Parkin N, de Mendoza C, Schuurman R, Jennings C, Bremer J,
Jordan MR, et al. Evaluation of in-house genotyping assay
performance using dried blood spot specimens in the global
World Health Organization laboratory network. Clin Infect Dis
2012; 54:S273S279.
55. Monleau M, Aghokeng AF, Eymard-Duvernay S, Dagnra A,
Kania D, Ngo-Giang-Huong N, et al. Field evaluation of dried
blood spots for routine HIV-1 viral load and drug resistance
monitoring in patients receiving antiretroviral therapy in
Africa and Asia. J Clin Microbiol 2014; 52:578586.
56. Wilson T, Vlahov D, Crystal S, Absalon J, Klein S, Remein R,
et al. Integrating HIV prevention activities into the HIV medical care setting: a report from the NYC HIV Centers Consortium. J Urban Health 2006; 83:1830.

57. Tanser F, Barnighausen T, Grapsa E, Zaidi J, Newell M-L. High

coverage of ART associated with decline in risk of HIV acquisition in rural KwaZulu-Natal, South Africa. Science 2013;
339:966971.
58. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour
MC, Kumarasamy N, et al. Prevention of HIV-1 infection
with early antiretroviral therapy. N Engl J Med 2011; 365:
493505.
59. Das M, Chu PL, Santos G-M, Scheer S, Vittinghoff E, McFarland
W, et al. Decreases in community viral load are accompanied
by reductions in new HIV infections in San Francisco. PLoS One
2010; 5:e11068.
60. Nichols BE, Sigaloff KCE, Kityo C, Mandaliya K, Hamers RL,
Bertagnolio S, et al. Averted HIV infections due to
expanded antiretroviral treatment eligibility offsets risk of
transmitted drug resistance: a modeling study. AIDS 2014;
28:7383.
61. Sood N, Wagner Z, Jaycocks A, Drabo E, Vardavas R. Test-andtreat in Los Angeles: a mathematical model of the effects of
test-and-treat for the population of men who have sex with
men in Los Angeles County. Clin Infect Dis 2013; 56:1789
1796.

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