Dependence of HIV drug resistance on the early warning indicator drug stock out, especially in middle-income countries
By: Mathilda Rudén
Supervisor: Patrik Dinnétz
Södertörn University | School of Natural Sciences, Technology and Environmental Studies.
Master’s thesis, 15 credits
Environmental Science | Spring 2017 Infectious Disease Control
2
Abstract
Background
HIV drug resistance is presumed to be inevitable due to the error-prone nature of the virus.
However, poor adherence to the antiretroviral drugs is proven to be an impending factor for HIV drug resistance development. Of these two explanations, which is the most common reason for HIV drug resistance?
Methods
A total of 40 published studies about HIV drug resistance, were retrospectively collected in Pubmed (May 2017), from 36 different countries for this paper. From each study was
participants, percentage of HIV drug resistance and HIV-1 subtype extracted for analysis. All studies were than classified by either high-income, middle-income or low-income, based on a country income status, defined by the World Bank. HIV drug resistance was tested against:
continents, HIV-1 subtypes, number of study participants, income levels, GDP per capita and EWI’s. All statistical analysis was performed in R: The R project for statistical computing.
Results
This paper show, that HIV drug resistance primarily is caused by poor adherence which is closely associated with drug stock out. Highest HIV drug resistance levels was found in middle-income countries. However, number of participants enrolled per study was important for the outcome and this indicates that HIV drug resistance would be higher in low-income countries if larger studied had been carried out in these settings. This means that there is a large unrecorded prevalence of HIV drug resistance in low-income countries.
Keywords: Error-prone, Poor adherence, Early warning indicators.
3
Table of Contents
Abstract ... 2
Background ... 2
Methods ... 2
Results ... 2
Acronyms and Abbreviations ... 3
Introduction ... 4
Material and methods ... 7
Eligibility criteria ... 7
Method ... 8
Statistical analyses... 8
Results ... 9
Discussion ... 13
Acknowledgements ... 15
References ... 16
Appendix ... 19
References: primary data ... 19
Table: primary data ... 23
Acronyms and Abbreviations
HIV Human Immunodeficiency Virus WHO World Health Organization
NRTI Nucleoside Reverse-Transcriptase Inhibitors NNRTI Nonnucleoside Reverse-Transcriptase Inhibitors INSTI Integrase Inhibitor
PI Protease Inhibitor
EWI Early Warning Indicators LTFU Loss to Follow-Up
ANOVA Analysis of Variance
HSD Honestly Significant Difference GDP Gross Domestic Product
4
Introduction
The idea behind my paper started when I read the course HIV - with focus on the individual from a global perspective at the Karolinska Institute during spring 2017. Briefly, we
discussed drug resistance to antiretroviral drugs and I immediately became interested. I started to update myself on the subject and found conflicting ideas about the cause for resistance. HIV drug resistance is presumed to be unavoidable due to the error-prone nature of the virus, even if medication is taken with appropriate adherence16. However, poor
adherence to antiretroviral drugs is also suggested to be an impending factor for development of HIV drug resistance4.
Antiretroviral drugs are acting as a strong selection pressure selecting for resistance. If HIV drug resistance is caused by the error-prone nature of the virus, then the prevalence of
resistance should be highest in high-income countries where the antiretroviral drugs are most available (Figure 1)? On the other hand, if HIV drug resistance instead is caused by poor adherence, then the prevalence of resistance should be highest in low-income countries were the risk of poor adherence is highest (Figure 1)? The aim of this study is to test which of these two hypotheses for development of HIV resistance.
Figure 1. Visualize the hypothesis. HIV drug resistance is either caused by the error-prone nature of the virus and highest levels of HIV drug resistance should then be found in high-income countries, or caused by poor adherence to antiretroviral drugs and highest levels of HIV drug resistance should then be found in low-income countries.
HIV drug resistance, threatens the effectiveness of the antiretroviral therapy for treatment and preventing further transmission of HIV. At the end of 2015 there were approximately 36.7
5
million people living with HIV worldwide. Fortunately, since the introduction of
antiretroviral therapy HIV related mortality and morbidity have decreased significantly1. In several high-income countries, HIV has become a chronic disease2, with moderate impact on life quality27. However, in many medium, and low-income countries, HIV is still highly endemic and fatal27. Worldwide, only 46 percent of all people living with HIV are receiving antiretroviral therapy, which is a contributing factor to the uneven development3.
Nevertheless, great progress has been achieved through the antiretroviral drugs, but this success is now threatened by the increasing prevalence of HIV drug resistance.
HIV drug resistance can be transmitted directly through infection, but does most commonly arise de novo during viral replication in presence of antiretroviral drugs. The HIV-1 reverse transcriptase is error-prone by nature, and replicate 10.000 to 100.000 times per day in absent of antiretroviral drugs. In the presence of antiretroviral therapy viral particles should be suppressed to undetectable genotyped levels. However, due to the error-prone nature of virus replication, drug use can select for mutations conferring resistance to the antiretroviral drugs4,5. In a continuing replication during drug treatment, resistant viruses can reproduce while the HIV-1 wild type will fail to replicate. This leads to an increase of the fraction of virions over time4,6.
The World Health Organization (WHO) recommend, two Nucleoside Reverse-Transcriptase Inhibitors (NRTI’s) and a Nonnucleoside Reverse-Transcriptase Inhibitor (NNRTI) or an Integrase Inhibitor (INSTI) as the first-line antiretroviral therapy for adults. If one or several antiretroviral drugs would fail to induce viral suppression, second-line antiretroviral therapy should be implemented instead. Viral load suppression failure may occur due to non-
adherence, and can be caused by drug-related toxicity, social and cultural barriers,
forgetfulness, a lack of understanding of treatment benefits, or inconvenient drug regimens with high pill burden5,19. Recommendation for second-line antiretroviral therapy for adults should consist of two NRTI’s and a ritonavir-boosted Protease Inhibitor (PI)7. The
recommendation to use three or more drugs in therapy is to minimize the likelihood that the virus will develop resistance to all drugs simultaneously8. However, cross-resistance to all drugs within a class occurs frequently. This is especially problematic in NNRTI’s, due to their lower genetic barrier for rapid evolution of drug resistance5. The genetic barrier to antiretroviral drugs can be defined as the number of genetic mutations required to develop resistance. Drugs classified as low genetic barrier only need to fail once, for the virus to develop mutations with high-level resistance. Meanwhile, drugs classified as high genetic
6
barrier requires several drug failures before the virus can develop drug resistance. Some PI drugs are high genetic barriers meanwhile several NRTI’s and NNRTI’s drugs are low genetic barriers to develop drug resistance4,29.
Prevalence of HIV drug resistance was in 2010, 12.9 percent in North America, 10.9 percent in Europe, 6.3 percent in South America, 4.7 percent in Africa and 4.2 percent in Asia13,14. The low levels of HIV drug resistance could be associated with less antiretroviral therapy coverage in Africa and Asia. However, since 2010 an antiretroviral drug scale-up in resource limited countries has been carried out. The African region increased their antiretroviral therapy coverage from 21 percent to 47 percent in 2015. Similar progress was seen in South East Asia, their antiretroviral therapy coverage increased from 20 percent to 39 percent in 201515. HIV drug resistance prevalence increased from 4.8 in 2007 to 6.8 in 2010, in low-and middle income countries20. Meanwhile a downward trend in HIV drug resistance is seen in high-income countries14,21.
One of the targets, in the WHO’s global HIV drug resistance surveillance and monitoring strategy, is the annual monitoring of Early Warning Indicators (EWI) to HIV drug resistance.
The EWI’s consist of six different targets shown to be associated with HIV drug resistance.
These are, appropriate prescribing practices, loss to follow-up at twelve months (LTFU), retention at twelve months, on-time pill pick-up, on-time appointment keeping and drug stock out16. The aim is to examine national and regional EWI prevalence estimates to compare performance over time within and across regions. Color-coded scores cards (performance strata), to visualize clinic performance, was used to facilitate identification of gaps in service delivery. In the 2016 report, more than 12.000 clinics from cohorts of patients receiving antiretroviral therapy, from 59 countries reported data between 2004 and 2016. Globally, results showed high levels of appropriate antiretroviral drug prescribing with 99 percent of people prescribed regimens according to international HIV treatment guidelines and LTFU at 12 months during the same period averaged 20 percent, exceeding the WHO-recommended target of 15 percent. Retention on antiretroviral therapy at 12 months averaged 73.5 percent, falling short of the WHO-recommended target of more than 85 percent. Adherence, as
estimated by on-time pill pick-up and on-time appointment keeping, fell below global targets.
Only 1150 clinics monitored drug stock outs and 35.7 percent had at least one drug stock out routinely during respective reporting year, thus failing the WHO-recommended target of no antiretroviral drug stock outs. Regionally, data showed variation and once again was none of
7
the set targets fulfilled16. This indicates that all geographical regions should have high risk of HIV drug resistance.
The effectiveness of antiretroviral therapy and development of HIV drug resistance is strongly associated with poor adherence4,9,10. WHO, among others, advocates that HIV drug resistance is inevitable because of the error-prone characteristic of the virus, even among individuals taking antiretroviral therapy with appropriate adherence6,11,12,18. As mentioned earlier, HIV drug resistance levels, has shown to be high in North America and Europe15. The question is if this is because these regions have higher income levels, which generate wider availability to antiretroviral drugs27, and therefore have a larger population exposed to antiretroviral therapy that can lead to HIV drug resistance? Or if drug resistance is due to poor adherence to antiretroviral therapy, shown by EWI’s? In the latter case, HIV drug resistance should be higher in low-income countries where the EWI’s are likely to be least achieved? In this paper, estimates of HIV drug resistance from 40 studies in different geographical settings, were tested against the variables: continents, income levels, GDP (Gross Domestic Product) per capita, HIV-1 subtypes, participants and EWI’s. The aim is to evaluate which of the variables that are causing HIV drug resistance and see if the result is associated with the characteristics of the virus or the poor-adherence to antiretroviral drugs.
Material and methods Eligibility criteria
A literature search was carried out in Pubmed April 2017. The search terms used were: “HIV drug resistance” or “HIVDR” with no time limit and all fields included. All published studies, written in English, between 2005 and 2017, that had performed a genotyped analysis,
alternatively a low-frequency genotyping analysis, provided data of HIV drug resistance were considered. Studies would preferably have different origin to increase the geographical distribution. If several studies were published from the same country, the first study that met the eligible criteria was included. Studies were excluded if HIV drug resistance could not be calculated among the study-participants (since the research question concerns HIV drug resistance, specifically), or, if a study had a defined population like children/adolescents, pregnant women, men who have sex with men or people who inject drugs (since specific target groups were not important for the research question) or, if more than one study was found from the same country (the idea was to collect as many studies from different countries as possible) or, if the study was published before 2005.
8
Method
The literature search found a total of 1.308 published studies. Of these were 1069 studies published after 2005 and 40 of these fulfilled the eligible criteria and were collected for analysis. The collected studies were then divided into continental origin to visualize the geographical distribution. Ten studies were from Africa, nine from Asia, seven from Europe, eight from Latin America and six from North America. Each study represents a unique country except for North America with three studies from the United Sates and three studies from Canada. This means that, Canada and United States were defined as North American.
Mexico, Cuba, Puerto Rico, Guatemala and Nicaragua were defined as Latin American countries. The aim was to collect as many studies with different origin as possible to increase the geographical distribution.
From each study, participants, percentage of HIV drug resistance and HIV-1 subtype was extracted for analysis. All studies were classified as either high-income, middle-income, or low-income, based on a country income status, defined by the World Bank17. The GDP per capita for each country, collected from the World Bank17, was also extracted for analysis.
WHO’s defintion of EWI’s to HIV drug resistance consists of six risk targets, namely:
prescribing practice, LTFU, Retention on ART, On-time pill pick-up, On-time appointment keeping and drug stock out. These EWI’s were used as a varaible for poor adherence in the statistical analysis. Specifically, data from the 2016 report was used in this paper16. In the report, data was presented through the following regions: Americans, Europe, West Africa, Central Africa, East Africa, South Africa, South East Asia and Western Pacific. Data of HIV drug resistance from the 40 included studies, were later reclassfied from continents to match these regions befor statisctical analysis were perfromed.
Statistical analyses
R: The R project for statistical computing28 was used for all statistical analyses. HIV drug resistance data from the 40 studies was log transformed before statistical analysis. HIV drug resistance was analyzed as a function of income and log number of participants per study with a linear model. Log HIV resistance was also analyzed as a function of income levels in an ANOVA. Results met the assumption of homogeneity of variance and a Tukey’s Honestly Significant Difference (HSD) post hoc test was used to test differences among income levels.
HIV drug resistance was also compared among continents with an ANOVA. Log HIV drug resistance was also analyzed between HIV-1 subtypes with an independent samples t-test. I
9
also analyzed Log HIV drug resistance as a function of GDP per capita with a linear regression. Finally, the correlations among HIV drug resistance and all EWI’s were tested with Pearson correlation tests.
Results
After systematically reviewing 1069 published studies based on the eligible criteria, 40 studies from 36 different countries were included in the analysis. Table 1 shows the characteristics of the 40 studies. Ten studies were collected from Africa, nine from Asia, seven from Europe, eight from Latin America and six from North America. The European studies included the highest number of participants per study, while the Latin American studies had the lowest number of participants, closely followed by the African studies with the second lowest number of participants (Table 1). HIV drug resistance was highest among the Asian- and Latin American participants and lowest among the North American- and European participants (Table 1). Several HIV-1 subtypes were found, but subtype B and subtype C were by far the most common ones (Table 1). Subtype B was seen in all continents except for Africa, where subtype C was most dominant. Income levels were classified as low- income, middle-income and high-income. Five were defined as low-income countries and located in Africa, twenty-one were defined as middle-income countries and were primary located in Asia and Latin America. Fourteen were defined as high-income countries and were primary located in North America and Europe (Table 1).
Table 1. Characteristics of the primary data.
Africa Asia Europe
North America
Latin America
No. of studies 10 9 7 6 8
No. of participants 1.832 9.436 21.289 5.900 1.640
HIV drug resistance (log) 1.20 1.42 1.10 1.13 1.22
Subtype: no. of studies
A 2
B 3 5 6 7
C 3 2
CRF01_AE 2
CRF02_AG 3
F1 1
G 1
Income level: no. of studies
Low-income 5
Middle-income 5 7 2 7
High-income 2 5 6 1
10
There was a significant difference in log HIV drug resistance among countries with different income levels (Table 2). However, the analysis could not show any significant par wise difference in the Tukey post hoc test following the ANOVA.
Table 2. ANOVA of log HIV drug resistance and income levels.
Sum Sq Df F value Pr(>F) Income levels 2 0.80 0.40 3.56 <0.038
Residuals 37 4.16 0.11
The graphical result showed that middle-income countries had the highest amount of HIV drug resistance and high- and low-income countries had the lowest amount (Figure 3). Low- income countries also had the largest CI. A HDS post hoc test showed no statistically significance difference between the different income levels. Closest to a statistically significance result was between high-income and middle-income (p> 0.0756).
Figure 3. Mean log HIV drug resistance ± 95%CI for countries with different economic status. None of the pairwise comparisons were significant with Tukey post-hoc test.
11
For each country, HIV drug resistance and GDP per capita was tested. The results showed no statistically significance (b=0.000005, t=-1.67, p>0.1, Linear Regression).
HIV-1 Subtype B and subtype C were the most common strains and with 60 percent, and 14 percent of all observations, respectively. There was no statistically significant difference in drug resistance between the subtypes (t = -0.90, df= 5.5, p-value > 0.4, t-test)
There was no significant difference for HIV drug resistance between continents (F4,35 = 1.01, p>0.05, ANOVA).
Log transformed levels of HIV drug resistance were also analyzed as function of the log transformed number of participants, income levels and the interaction between log number of participants and income levels using a general linear model. The adjusted R-squared were 0.44, which indicates that almost half of the variation in HIV drug resistance is explained by income levels and number of participants in the study.
Table 3. ANOVA of log HIV drug resistance, log participants and income levels.
Sum Sq Df F Value Pr(>F)
log (Participants) 5.28 1 13.9 <0.001
Income levels 2.37 2 3.13 0.056
log (Participants): income levels 3.94 2 5.21 0.010
Residuals 12.8 34
Middle-income countries, with moderate number of participants had the highest percentage of HIV drug resistance (Figure 4). The regression line is negative for middle-income countries, which indicate, the studies with many participants show a lower percentage of HIV drug resistance. The opposite is shown in low-income countries. Studies with few participants in low-income countries show lower levels of HIV drug resistance and studies with many participants have a higher level of HIV drug resistance. High-income countries also have a negative regression line where studies with many participants have lower levels of HIV drug resistance.
12
Figure 4. Scatterplot illustrating mean log HIV drug resistance as a function of log participants for low-, middle- and high- income countries.
The six different EWI’s and HIV drug resistance was tested with a Pearson correlation tests.
Drug stock out was the only EWI that was significantly positively correlated with mean HIV drug resistance (p < 0.006, Pearson Correlation). The target set for drug stock out was zero, but no region achieved this target in the 2016 global report (Table 4). Western Pacific and South East Asia had the highest frequency of drug stock out, and also the highest mean values for HIV drug resistance. The mean HIV drug resistance for South East Asia is based on four studies, all defined as middle-income countries (Table 4). The mean HIV drug resistance for Western Pacific is based on five studies, where three studies are defined as middle-income and two as high-income countries (Table 4).
13
Table 4. Data from the WHO annual report of Early Warning Indicators, 201616. Values are percentage of clinics meeting targets and mean HIV drug resistance from the 40 studies. Each target has different criteria for target achievement.
Region
Prescription practice
>100 %
LTFU
<15 %
Retention to ART
>85 %
On-time Pill pick- up >90 %
On-time appointment keeping >80 %
Drug stock out 0 %
HIV drug resistance
mean
Americas 40% 67.3% 52.2% 28% 40.9% 58.6% 1.19
Europe 87.9% 87.9% 48.5% 63.6% 93.9% 72.7% 1.10
Western
Africa 61% 20.8% 18% 7.7% 24% 60% 1.25
Central
Africa 96.9% 35.1% 19% 3.8% 30.1% 56.3% 1.25
Eastern
Africa 89.4% 60.6% 26.5% 2.3% 19% 60.1% 1.21
Southern
Africa 88.7% 51.3% 33.3% 17.5% 68.2% 13.2% 0.95
South
East-Asia 85.2% 50.8% 51.6% 65.3% 75% 88.4% 1.44
Western
Pacific 95.5% 89% 82.3% 67.2% 67.3% 94.9% 1.40
Discussion
This study aimed to examine if HIV drug resistance primary is a cause of the error-prone characteristics of the virus or poor adherence to the antiretroviral therapy. The hypothesis was that HIV drug resistance either should be highest among high-income countries due to wider availability and longer use of antiretroviral therapy, or highest in low-income countries where the risk of poor adherence, due to EWI’s is highest? Highest HIV drug resistance was found in middle-income countries, which means that both hypotheses are dismissed. However, this does not rule out adherence as the impending factor to HIV drug resistance development.
There was a significant correlation between drug stock out and mean HIV drug resistance were two regions, South East Asia and the Western Pacific had the highest rates of “Drug stock out”. Of the 40 studies included in statistically analysis, ten studies were from countries located in South East Asia and Western Pacific. A total of 78 percent of these were defined as middle-income countries. The correlation between them indicates that HIV drug resistance may be highest among middle-income countries due to high levels of drug stock out. This can also be interpreted as if poor adherence is a primary reason for developing HIV drug
resistance. This result is consistent with another study which suggest that acquired HIV drug resistance is related to poor adherence and can be minimized if protocols for drug stock were to be improved23. As mentioned earlier, the error-prone nature of the virus can acquire drug resistance even with appropriate adherence. This happens in all geographical settings and is
14
not affected by income levels. The correlation between HIV drug resistance and drug stock out is probably significant because more people becomes affected by drug stock out and therefor cause higher mean HIV drug resistance levels compared to the effect by drugs acting as selective pressure in ongoing treatment scenarios.
Another interesting result showed that the number of participants per study was of different importance for the development of HIV drug resistance in countries with different income levels. Once again, HIV drug resistance was highest in studies from middle-income countries, followed by high-income countries and then low-income countries. Studies from middle- income countries and high-income countries, showed a negative correlation between HIV drug resistance and number of participants. Meanwhile, Low-income countries showed the opposite, a positive correlation between HIV drug resistance and number of participants. The result is hard to interpret, which makes the result very curious. Why would levels of HIV drug resistance be affected by the number of participants in a study, depending on income levels? Is there a representative bias to the low-income studies in this paper? Three out of the five studies30,31,32 classified as low-income countries in this paper, searched for prevalence of transmitted HIV drug resistance and therefore only recruited participants who were
treatment-naïve to antiretroviral drugs. It is known, that strict criteria for inclusion and exclusion of subjects in clinical trials makes it difficult to locate participants33. Could this be an extra difficult problem to HIV, due to the high risk of segregation and discrimination to people living with HIV? And could this even be worse in low-income countries due to factors as, generally lower levels of education, economy, public health care and transportation
possibilities? However, this is only speculations. Regarding these ideas, it would be interesting to follow up with further investigation.
Another interpretation can be, the sample size for the five studies defined as low-income countries are too small. A total of 1.832 participants were included in these studies, which is only five percent of the total number of participants included in my study. If the sample size had been larger in these studies, more HIV drug resistance would perhaps have been found. If this was true, the result indicates that there might be an unrecorded prevalence of HIV drug resistance among people living with HIV in low-income countries. Once again, the result is hard to interpret and therefore is only speculations discussed. However, I hope these
questions will get investigated further.
15
The paper has some concerns that should be highlighted. First, reclassifying of the continents is not ideal but was necessary for the paper to get a representative distribution of studies.
Second, only one study per country were included even if several could be found during the literature search. Because of that, there might be a concern about the reliability of the papers.
However, the aim was to examine HIV drug resistance levels in different geographical settings and therefor an evenly distribution of studies included was necessary. Third, income levels were classified as low-income, middle-income and high-income. However, the World Bank used low-income, lower middle-income, middle-income, upper middle-income and high-income as classifications. To make it easier to manage statistical analysis, all countries defined as either lower middle-income, middle-income and upper middle-income was classified as middle-income in this paper. Fourth, HIV drug resistance levels were collected retrospectively and tested against income levels, GDP per capita, participants, HIV-1 subtypes, continents, and EWI’s as possible variables. Examined variables could not be randomly assign to research participants, which makes this study a Quasi-experiment25. Due to the Quasi-experimental design, causality to findings may be less definitive compared to an experimental test. However, an experimental design in this context would neither be ethical nor feasible to conduct26. Nevertheless, it is crucial to perform this type of research, and findings can still be correctly analyzed and generalized to other scientific data to allow for triangulation or other forms of mixed analyses.
Concluding remarks
The overall conclusion from this study is that HIV drug resistance primarily is caused by poor adherence which is closely associated with drug stock out. Middle-income countries in South East Asia and Western Pacific, showed to have the highest levels of HIV drug
resistance and drug stock out. Another finding, showed an interesting but ambiguous result.
Further investigations are necessary to draw any definitive conclusions.
Acknowledgements
I would like to thank all teachers and students at the master program Infectious Disease Control for an inspiring and fun year. A special thanks to my supervisor Patrik Dinnétz for the support and guidance during the course.
16
References
1. World Health Organization. Health topics. Antiretroviral therapy [Internet]. Geneva:
World Health Organization; 2017. [cited 2017 May 09]. Available from http://www.who.int/topics/antiretroviral_therapy/en/
2. Luo R, Piovoso J.M, Martinez-Picardo J, Zurakowski R. Optimal Antiviral Switching to Minimize Resistance Risk in HIV Therapy. 2011 Nov; 6(11): e27047
3. World Health Organization. 10 Facts on HIV/AIDS [Internet]. Geneva: World Health Organization; 2016. [cited 2017 May 09]. Available from
http://www.who.int/features/factfiles/hiv/en/
4. Gardner ME, Burman JW, Steiner FJ, Anderson LP, Bangsberg RD. Antiretroviral medication adherence and the development of class-specific antiretroviral resistance.
2009 Jun; 1(9): 1035-1046.
5. Petrella M, Brenner B, Loemba H, Wainberg A.M. HIV Drug Resistance and Implications for the Introduction of Antiretroviral Therapy in Resource-poor Countries. 2001 Dec; 4(6): 339-346.
6. Freeman S, Herron C.J. Evolutionary Analysis. Fourth Edition. Upper Saddle River:
Pearson Education International; 2007
7. World Health Organization. Consolidation Guidelines for the use of Antiretroviral Drugs for Treating and Preventing HIV infection: Recommendation for a Public Health Approach. Second Edition. Geneva: World Health Organization; 2016 8. Snedecor JS, Subharshan L, Nedrow K, Bhanegaonkar A, Simpson NK, Haider S, et
al. Burden of Nonnucleoside Reverse Transcriptase Inhibitor Resistance in HIV-1 infected patients: A systematic review and meta-analysis. 2014 Nov; 30(8): 753-768.
9. Riberio M.R, Bonhoeffer S. Production of resistant HIV mutant during antiretroviral therapy. 2000 Jul ;97(14): 7681-7686.
10. Sethi K.A, Celentano D.D, Gange J.S, Moore D.R, Gallant E.J. Association between Adherence to Antiretroviral Therapy and Human Immunodeficiency Virus Drug Resistance. 2003 Oct; 37(8): 1112-1118.
11. World Health Organization. Global Report on Early Warning Indicators of HIV drug Resistance: Technical Report. Geneva: World Health Organization; 2016
12. Wainberg A.M. The Need for Development new HIV-1 Reverse Transcriptase and Integrase Inhibitors in the Aftermath of Antiretroviral Drug Resistance. 2012 Nov;
2012: 1-28.
17
13. Frentz D, Boucher C, Van de vijer D. Temporal Changes in the Epidemiology of Transmission of Drug-Resistant HIV-1 across the World. 2012 Jan-Mar; 14(1):17-27.
14. World Health Organization. HIV Drug Resistance Report 2012: HIV/AIDS Programme. Geneva: World Health Organization; 2012
15. World Health Organization. HIV/AIDS. ART coverage table, 2000-2015 [Internet].
Geneva: World Health Organization; 2017. [cited 2017 May 10]. Available from http://www.who.int/hiv/data/art_table_2000_2015.png?ua=1
16. World Health Organization. Global Report on Early Warning Indicators of HIV drug Resistance: Technical Report. Geneva: World Health Organization; 2016
17. The World Bank. Income levels [Internet]. Washington: The World Bank; 2017 [cited 2017 May 06]. Available from http://data.worldbank.org/country
18. Bissio E, Barbás GM, Bouzas BM, Cudolá A, Salomón H, Espinóla L, et al.
Pretreatment HIV-1 Drug Resistance in Argentina: results from a surveillance study performed according to WHO-proposed new methodology in 2014-2015. 2017 Feb;
72(2): 504-510.
19. Masikini P, Mpondo B.C.T. HIV Drug Resistance Mutations Following Poor Adherence in HIV-infected patients: a case report. 2015 Feb; 3(6): 353-356.
20. Häggbom A. Antiretroviral Drug Resistance and Treatment Outcomes of Human Immunodefiency Virus Type-1, Implications for Low and Middle Income Countries.
Stockholm: Karolinska Institute; 2016 [cited 2017 May 15]. Available from:
https://openarchive.ki.se/xmlui/handle/10616/44964
21. Pham DQ, Wilson PD, Law GM, Kelleher DA, Zhang L. Global Burden of HIV Drug Resistance and HIV-exposure categories: a systematic review and meta-analysis. 2014 Nov; 28(18): 2751-2761.
22. Gregson J, Tang M, Ndembi N, Hamers LR, Rehee S, Marconi VC, et al. Global Epidemiology of Drug Resistance After Failure of WHO Recommended First-line Regimes for Adults HIV-1 Infection: a multicenter retrospective cohort study. 2016 Jan; 16(5): 565-575.
23. Stadeli KM, Richman DD. Rates of Emergence of HIV Drug Resistance in Resource- Limited Settings: a systematic review. 2013; 18(1): 115-123.
24. Hosseinipour CM, Gupta KR, Van Zyl G, Eron JJ, Nachega BJ. Emergence of HIV Drug Resistance during First-and Second-Line Antiretroviral Therapy in Resource- Limited Settings. 2013 Jun; 207(Suppl 2): S49-S56.
18
25. Trochim W. [Internet]. Web Center for Social Research Methods; 2006 [cited 2017 May 16]. Available from http://www.socialresearchmethods.net/kb/quasiexp.php 26. Centre for Innovation in Research and Teaching. Benefits & Limitations of Quasi-
Experimental Research [Internet]. Arizona: Grand Canyon University; 2017 [cited 2017 May 16]. Available from:
https://cirt.gcu.edu/research/developmentresources/research_ready/quasiexperimental/
benefits_limits
27. Hogg R, Lima V, Sterne C.A.J, Grabar S, Battegay M, Bonarek M, et al. Life Expectance of Individuals on Antiretroviral Therapy in High-income Countries: a collaborative analysis of 14 cohort studies. 2008 Jul; 372(9635): 293-299.
28. R Core Team (2016). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R- project.org/.
29. Luber D Andrew. Genetic Barrier to Resistance an Impact on Clinical response. 2005 Jul; 7(3): 69.
30. Mamadou s, Hanki Y, Ali Maazou A.R, Aoula B, Diallo S. Genetic Deversity and Drug Resistance Mutations in HIV-1 from Untreated Patients in Niamey, Niger. 2011 Nov; 2011: 797463.
31. Rusine J, Asiimwe-Kateera B, Van de Wijgert J, Boer R.K, Mukantwali E, Karita E, et al. Low Primary and Secondary HIV Drug-Resistance after 12 Monthes of
Antiretroviral Therapy in Human Immunodeficiency Virus Type 1 (HIV-1)-Infected Individuals from Kigali, Rwanda. 2013 Aug; 8(8): e64345
32. Kasang C, Kalluvya S, Majinge C, Stich A, Bodem J, Kongola G, et al. HIV Drug Resistance (HIVDR) in Antiretroviral Therapy-Naïve Patients in Tanzania Not Eligible for WHO Threshold HIVDR Survey is Dramatically High. 2011 Aug; 6(8):
e23091.
33. Martínez-Mesa J, González-Chica D.A, Duquia R.P, Bonamigo R.R, Bastos J.L.
Sampling: how to select participants in my research study? 2016 May; 91(3): 326-30.
19
Appendix
References: primary data
A
ndargachew M, Maier M, Libert G.U. Resistance 10 Years after Roll-Out of Antiretroviral Therapy in Ethiopia: A Prospective Cohort Study. 2015 Oct; 10(10): e014318.A
vila-Ríos S, García-Morales C, Matías-Florentina M, Tapia-Trejo D, Hernández-Álvarez F.B, Moreira-López E.S, et al. HIV Drug Resistance in Antiretroviral Treatment-naïve Inidviduals in the Largest Public Hospital in Nicaragua, 2011-2015. 2016 Oct; 11(10):e0164156.
A
vila-Ríos S, García-Morales C, Tapia-Trejo D, Meza I.R, Nuñez M.S, Parham L, et al. HIV Drug Resistance Surveillance in Honduras after a Decade of Widespread Antiretroviral Therapy. 2015 Nov; 10(11): e0142604.B
aesi K, Ravanshad M, Ghanbarisfari M,Saberfar E, SeyedAlinaghi S.A, Volk E.J.Antiretroviral Drug Resistance Among Antiretroviral-Naïve and Treatment Experienced Patients Infected with HIC in Iran. 2014 Jul; 86(7): 1093-1098.
B
arrow J.G, Hylton-Kong T, Rodriguez N, Yamamura Y, Figueroa J.P. HIV-1 drugresistance in treatment-naive, chronically infected patients in Jamaica. 2013; 18(7): 941-944.
B
issio E, Barbás G.M, Bouzas B.M, Cudolá A, Salomón H, Espínola L, et al. Pretreatment HIV-1 drug resistance in Argentina: results from a surveillance study performed according to WHO-proposed new methodology in 2014–15. 2017 Feb; 72(2) 504-510.B
rooks I.J, Niznick H, Ofner M, Merks H, Angel B.J. Local Phylogenetic Identifies Distinct Trends in Transmitted HIV Drug Resistance: implications for public health interventions.2013 Oct; 13: 509.
B
urchell N.A, Bayoumi M.A, Rourke B.S, Major C, Gardner S, Sandstrom P, et al. Increase in Transmitted HIV Drug Resistance Among Persons Undergoing Genotypic Resistance Testing in Ontario, Canada, 2002-09. 2012 Nov; 67(11): 2755-65.C
hen X, Zou X, He J, Zheng J, Chiarella J, Kozal J.M. HIV Drug Resistance Mutations (DRMs) Detected by Deep Sequencing in Virologic Failure Subjects on Therapy from Hunan Province, China. 2016 Feb; 11(2): e0149215.C
harest H, Coualla-Bell F, Cantin R, Murphy DG, Brenner B, Moisi D, et al. A Significant Reduction in the Frequency of HIV-1 Drug Resistance in Québec from 2001 to 2011 is Associated with a Decrease in the Monitored Viral Load. 2014 Oct; 9(10): e109420.20
C
hkhartishvili N, Sharvadze L, Dvali N, Karchava M, Rukhadze N, Lomtadze M, et al.Virologic Outcomes of Second-line Antiretroviral Therapy in Eastern European Country of Georgia. 2014 Jul; 11: 18.
C
hung-Chih Lai, Wen-Chun Liu, Chi-Tai Fang, Yang Jyh-Yuan, Chang Lan-Hsin, Wu Pei- Ying, et al. Transmitted Drug Resistance of HIV-1 Strains Among Individuals Attending Voluntary Counselling and Testing in Taiwan. 2016 Jan; 71(1): 226-34.C
hung M.H, Silverman R, Beck I.A, Yatich N, Dross S, McKernan-Mullin J, et al.Increasing HIV-1 Pretreatment Drug Resistance Among Antiretroviral Naïve Adults Initiating Treatment Between 2006 and 2014 in Nairobi, Kenya. 2016 Jun; 30(10): 1680-2.
C
orado AL, Bello G, Leão R.A, Granja F, Naveca F.G. HIV-1 Genetic Diversity andAntiretroviral Drug Resistance Among Individuals from Roraima state, Northern Brazil. 2017 Mar; 12(3): e0173894.
D
agnra Y.A, Vidal N, Mensah A, Patassi A, Aho K, Salou M, et al. High Prevalence of HIV- 1 Drug Resistance Among Patients on First-line Antiretroviral Treatment in Lomé, Togo.2011 Jun; 14: 30.
D
iallo Km Zheng DP, Rottinghaus K.E, Bassey O, Yang C. Viral Genetic Diversity and Polymorphisms in a Cohort of HIV-1-Infected Patients Eligible for Initiation of Antiretroviral Therapy in Abujan, Nigeria. 2015 May; 31(5): 564-575.E
scoto-Delgadillo M, Vázquez-Valls E, Ramírez-Rodríquez M, Corona-Nakamura A, Amaya-Tapia G, Quintero-Pérez N, et al. Drug-Resistance Mutations in Antiretroviral-Naïve Patients with Established HIV-1 Infection in Mexico. 2005 Nov; 6(6): 403-409.H
auser A, Hofmann A, Hanke K, Bremer V, Bartmeyer B, Kucherer C, et al. Nation Molecular Surveillance of Recently Acquired HIV Infections in Germany, 2013 to 2014.2017 Jan; 22(2): 30436.
H
ofstra L.M, Sánchesz Rivas E, Nijuis M, Bank A.E.L, Wilkinson E, Kelly K, et al. High Rates of Transmission of Drug-resistant HIV in Aruba Resulting in Reduced Susceptibility to the WHO Recommended First-line Regimen in Nearly Half of Newly Diagnosed HIV-infected Patients. 2017 Apr; 64(8): 1092-1097.
K
arlsson A, Björkman P, Bratt G, Ekvall H, Gisslén M, Sönnerborg A, et al. Low Prevalence of Transmitted Drug Resistance in Patients Newly Diagnosed with HIV-1 Infection inSweden 2003-2010.
21
K
asang C, Kalluvya S, Majinge C, Stich A, Bodem J, Kongola G, et al. HIV Drug Resistance (HIVDR) in Antiretroviral Therapy-Naïve Patients in Tanzania Not Eligible for WHOThreshold HIVDR Survey is Dramatically High. 2011 Aug; 6(8): e23091.
K
assaye G.S, Grossman Z, Balamane M, Johnston-White B, Liu C, Kumar P, et al.Transmitted HIV Drug Resistance is High and Longstanding in Metropolitan Washington, DC. 2016 Sep; 63(6): 836-843.
K
iertiburanakul S, Pinsai S, Chantratita W, Pasomsub E, Leechawengwongs M, Thipmontree W, et al. Prevalence of Primary HIV Drug Resistance in Thailand Detected by Short Reverse Transcriptase Genotypic Resistance Assay. 2016 Feb: 11(2): e0147945.L
avu E, Kave E, Mosoro E, Markby J, Aleksic E, Gare J, et al. High Levels of Transmitted HIV Drug Resistance in a Study in Papua New Guinea. 2017 Feb; 12(2): e0170265.M
amadou s, Hanki Y, Ali Maazou A.R, Aoula B, Diallo S. Genetic Deversity and Drug Resistance Mutations in HIV-1 from Untreated Patients in Niamey, Niger. 2011 Nov; 2011:797463.
N
ishizawa M, Hattori J, Shiino T, Matano T, Heneine W, Johnson A.J, et al. Highly- Sensitive Allele-Specific PCR Testing Identifies a Greater Prevalence of Transmitted HIV Drug Resistance in Japan. 2013; 8(12): e83150.L
iégeois F, Vella C, Eymard-Duvernay S, Sica J, Makosso L, Mouinga-Ondémé A, et al.Virological Failure Rates and HIV-1 Drug Resistance Patterns in Patients on First-line Antiretroviral Treatment in Semirural and Rural Gabon. 2012 Nov; 15(2): 1785.
P
ark M, Kee MK, Rhee JE, Park JY, Soon Kim S, Kang C, et al. The Trend of Transmitted Dug Resistance in Newly Diagnosed Antiretroviral-naïve HIV/AIDS Patients During 1999- 2012 in South Korea. 2016 Aug; 81: 53-57.P
atil T.R, Gupta M.R, Sen S, Tripathy P.S, Chaturbhuj N.D, Hingankar K.N, et al.Emergence of Drug Resistance in Human Immunodeficiency Virus Type 1 Infected Patients from Pune, India, at the End of 12 Months of First Line Antiretroviral Therapy Initiation.
2014 Apr; 2014: 674906.
P
anichsillapakit T, Smith M.D, Wertheim O.J, Richman D.D, Little J.S, Mehta R.S.Prevalence of Transmitted HIV Drug Resistance Among Recently Infected Persons in San Diego, California 1996-2013. 2016 Feb; 71(2): 228-236.
22
P
arczewski M, Leszcyszyn-Pynka M, Witak-Jedra M, Maciejewska K, Rymer W, Szymczak A, et al. Transmitted HIV Drug Resistance in Antiretroviral-treatment-naïve Patients from Poland Differs by Transmission Category and Subtype. 2015 Jan; 70(1): 233-42.P
érez L, Kourí V, Alemán Y, Abrahantes Y, Correa C, Argonés C, et al. Antiretroviral Drug Resistance in HIV-1 Therapy-naïve Patients in Cuba. 2013 Jun; 16: 144-150.P
huong Thao V, Minh Quang V, Day N.J, Tran Chinh N, Shikuma M.C, Farrar J, et al. High Prevalence of PI Resistance in Patients Failing Second-line ART in Vietnam. 2016 Mar;71(3): 762-774.
Ross L, Lim M.L, Liao Q, Wine B, Rodriguez A.E, Weinberg W, et al. Prevalence of Antiretroviral Drug Resistance and Resistance-associated Mutations in Antiretroviral therapy-naïve HIV-infected Individuals from 40 United States cities. 2007 Feb; 8(1): 1-8.
R
usine J, Asiimwe-Kateera B, Van de Wijgert J, Boer R.K, Mukantwali E, Karita E, et al.Low Primary and Secondary HIV Drug-Resistance after 12 Monthes of Antiretroviral Therapy in Human Immunodeficiency Virus Type 1 (HIV-1)-Infected Individuals from Kigali, Rwanda. 2013 Aug; 8(8): e64345.
S
eu L, Mulenga B.L, Siwingwa M, Sikazwe I, Lambwe N, Guffey B.M, et al.Characterization of HIV Drug Resistance Mutations Among Patients Failing First-line Antiretroviral Therapy from a Tertiary Referral Center in Lusaka, Zambia. 2015 Jul; 87(7):
1149-1157.
S
teegen K, Carmona S, Bronze M, Papathanasopoulos A.M, Van Zyl G, Goedhals D, et al.Moderate Levels of Pre-Treatment HIV-1 Antiretroviral Drug Resistance Detected in the First South African National Survey. 2016 Dec; 11(12): e0166305.
T
emereanca A, Ene L, Mehta S, Manolescu L, Duiculescu D, Ruta S. Transmitted HIV Drug Resistance in Treatment-naïve Romanian Patients. 2013 Jul; 85(7): 1139-1147.T
ostevin A, White E, Dunn D, Croxford S, Delpech V, Williams I, et al. Recent Trends and Patterns in HIV-1 Transmitted Drug Resistance in the United Kingdom. 2017 Mar; 18(3):204-213.
V
ega Y, Delgado E, Fernández-García A, Cuevas T.M, Thomson M.M, Montero V, et al.Epidemiological Surveillance of HIV-1 Transmitted Drug Resistance in Spain in 2004-2012:
relevance of transmission clusters in the propagation of resistance mutations. 2015 May;
10(5): e0125699.
23
Table: primary data
Author Year Country Continent Region
Economic
status Participants HIVDR%
Liégeois et al. 2012 Gabon Africa Central Africa Middle-income 141 56,7%
Rusine et al. 2013 Rwanda Africa Central Africa Low-income 158 5,7%
Mamadou et al. 2011 Niger Africa Western Africa Low-income 96 8,3%
Dagnra et al. 2011 Togo Africa Western Africa Low-income 188 24,5%
Diallo et al. 2015 Nigeria Africa Western Africa Middle-income 271 28,0%
Steegen et al. 2016 South Africa Africa Southern Africa Middle-income 277 9,0%
Chung et al. 2014 Kenya Africa Eastern Africa Middle-income 386 6,0%
Seu et al. 2015 Zambia Africa Eastern Africa Middle-income 68 98,5%
Kasang et al. 2011 Tanzania Africa Eastern Africa Low-income 120 73,3%
Andargachew et al. 2015 Etihopia Africa Eastern Africa Low-income 127 4,7%
Patil et al 2014 India Asia South-East Asia Middle-income 19 52,6%
Baesi et al. 2014 Iran Asia South-East Asia Middle-income 62 58,1%
Chung-Chih et al. 2015 Taiwan Asia South-East Asia Middle-income 440 11,1%
Kiertiburanakul et al. 2016 Thailand Asia South-East Asia Middle-income 265 7,9%
Nishizawa et al. 2013 Japan Asia Western Pacific High-income 149 27,5%
Lavu et al. 2017
Papua New
Guinea Asia Western Pacific Middle-income 123 13,0%
Park et al. 2016 South Korea Asia Western Pacific High-income 8176 11,4%
Phunong Thao et al. 2015 Vietnam Asia Western Pacific Middle-income 173 97,1%
Chen et al. 2016 China Asia Western Pacific Middle-income 29 62,1%
Chkhartishvili et al. 2014 Georgia Europe Europe Middle-income 84 85,7%
Parczewski et al. 2015 Poland Europe Europe High-income 833 9,0%
Temereanva et al. 2013 Romania Europe Europe Middle-income 61 14,8%
Hauser et al. 2017 Germany Europe Europe High-income 809 10,8%
Vega et al. 2015 Spain Europe Europe High-income 1614 9,98%
Tostevin et al. 2017 United Kingdom Europe Europe High-income 16425 7,45%
Karlsson et al. 2012 Sweden Europe Europe High-income 1463 5,6%
Ross et al. 2007 United States
North
America Americas High-income 317 9,8%
Brooks et al. 2013 Canada
North
America Americas High-income 155 12,3%
Kassaye et al. 2016 Unites States
North
America Americas High-income 497 22,7%
Burchell et al. 2013 Canada
North
America Americas High-income 330 13,6%
Charest et al. 2014 Canada
North
America Americas High-income 4105 13,1%
Panichsillapakit et al. 2016 United States
North
America Americas High-income 496 13,5%
Avila-Ríos et al. 2016 Nicaragua Latin America Americas Middle-income 283 19,4%
24
Hofstra et al. 2016 Aruba Latin America Americas High-income 104 32,7%
Pérez et al. 2012 Cuba Latin America Americas Middle-income 401 7,2%
Avila-Ríos et al. 2015 Honduras Latin America Americas Middle-income 365 11,5%
Corado et al. 2017 Brazil Latin America Americas Middle-income 73 23,3%
Bissio et al. 2017 Argentina Latin America Americas Middle-income 239 13,0%
Escoto-Delgadillo et al. 2005 Mexico Latin America Americas Middle-income 96 15,6%
Barrow et al. 2013 Jamaica Latin America Americas Middle-income 79 25,3%
