Secondary Logo

Journal Logo

CLINICAL SCIENCE

Substantial decline in heavily treated therapy-experienced persons with HIV with limited antiretroviral treatment options

Bajema, Kristina L.a; Nance, Robin M.a; Delaney, Joseph A.C.a; Eaton, Ellenb; Davy-Mendez, Thibautc; Karris, Maile Y.d; Moore, Richard D.e; Eron, Joseph J.c; Rodriguez, Benignof; Mayer, Kenneth H.g; Geng, Elvinh; Garris, Cindyi; Saag, Michael S.b; Crane, Heidi M.a; Kitahata, Mari M.a; Centers for AIDS Research Clinical Network of Integrated Systems (CNICS)

Author Information
doi: 10.1097/QAD.0000000000002679

Abstract

Introduction

Antiretroviral therapy (ART) is highly effective in controlling HIV viremia, decreasing disease morbidity and mortality in persons with HIV (PWH) [1,2], and preventing HIV transmission [3]. However, the development of antiretroviral drug resistance has the potential to limit these therapeutic benefits [4,5]. Historically, a high burden of antiretroviral drug resistance mutations developed in heavily treatment-experienced (HTE) persons through sequential addition of active drugs, incomplete adherence and exposure to lower potency, less tolerable regimens [6–8], which limited treatment options (LTOs) [9] and posed a significant challenge to disease control. Genotypic resistance testing is recommended before ART initiation and in the setting of virologic failure to guide therapeutic decision making [9].

The population of HTE PWH has evolved throughout the modern ART era with the introduction of more potent antiretroviral drugs with higher barrier to resistance, less frequent administration, coformulation, reduced pill burden, and fewer side effects [10]. Information regarding changing prevalence and predictors of HTE PWH, in particular in the setting of contemporary ART, is limited.

Previous studies have used varying approaches to define HTE PWH, often relying on virologic failure and antiretroviral treatment history in the absence of genotypic resistance data, resulting in inconsistent findings [11–15]. Studies that have used genotypic resistance data to more accurately define HTE PWH were largely conducted in the early ART era [6,8,16–21] prior to the availability of the integrase strand transfer inhibitor (INSTI) class which has been shown to be highly effective in achieving virologic control in treatment-experienced PWH [22–24]. In addition, analyses restricted to PWH undergoing resistance testing [25–27] have been shown to overestimate HTE prevalence [16,28]. Studies of select subpopulations of PWH and examination of the most recent genotypic test rather than cumulative resistance [17,28] have resulted in conflicting estimates of prevalence as well as outcomes related to virologic suppression, disease progression, and mortality among HTE PWH [5,12,14,29–31].

Furthermore, prior studies have focused on the number of antiretroviral drugs or classes to which a PWH is resistant [32], rather than how many active drugs they have available, which is the key to achieving virologic suppression [33]. As new drugs and antiretroviral classes become available, the prevalence of PWH with LTOs may decrease. We conducted this study to examine trends in LTO throughout the modern ART era from 2000 to 2017. We also determined predictors of and clinical outcomes among PWH with LTO defined by cumulative genotypic resistance data and the number of active antiretroviral drugs available.

Methods

Data source

The Center for AIDS Research (CFAR) Network of Integrated Clinical Systems (CNICS) is a dynamic prospective clinical cohort of adult PWH receiving care at eight participating academic sites distributed across the United States. Demographic and clinical characteristics of PWH in the CNICS cohort are similar to the overall population of PWH in the United States [34]. Comprehensive clinical data collected through electronic medical records and other institutional data systems undergo rigorous quality assessment, are harmonized in a central repository, and are updated on a quarterly basis [35]. The CNICS Data Management Core at the University of Washington works closely with investigators, clinicians, and data teams at each site to ensure comprehensive capture of antiretroviral drugs and genotypic resistance tests that are processed using the Stanford HIV Drug Resistance Database [36]. Seven of eight CNICS sites collect genotypic resistance data. Institutional review boards at each site approved the cohort protocol.

Study population

We studied all ART-experienced PWH aged 18 or older in care at the seven CNICS sites (Case Western Reserve University; Fenway Community Health Center of Harvard University; Johns Hopkins University; University of Alabama at Birmingham; University of California, San Diego; University of North Carolina; University of Washington) with available resistance data between 1 January 2000 and 31 December 2017. We defined ART as a multidrug regimen including at least one drug from the following core classes: nonnucleoside reverse transcriptase inhibitors (NNRTIs), protease inhibitors, or INSTIs. Chemokine coreceptor antagonists and fusion inhibitors were infrequently prescribed and therefore not considered as additional core classes. Participants entered the study in the year of their first CNICS visit between 2000 and 2017 at which they were receiving ART. Drug resistance was ascertained within four major classes: nucleoside reverse transcriptase inhibitors (NRTIs), NNRTIs, protease inhibitors, and INSTIs. A given drug was considered active if scored by the Stanford algorithm as susceptible (<10 points) and inactive if scored as having potential low-level, intermediate-level, or high-level resistance (≥10 points) [36]. For each individual, resistance mutations were carried forward to assess cumulative antiretroviral drug resistance, including mutations detected in genotypic testing performed prior to ART initiation.

Heavily treatment-experienced persons with limited treatment options

We defined individuals with LTOs as having only two active antiretroviral classes available in which there were a limited number of active drugs or one or less antiretroviral class available. Three antiretroviral classes were available at the beginning of the study period and were considered limited if there were two or fewer active NRTIs or protease inhibitors, but as only two NNRTIs were available between 2000 and 2007, this class was considered limited if there were one or fewer active drugs available. With the introduction of the first INSTI in 2007, this new antiretroviral class was considered available with one active drug. The pool of drugs available in each class per calendar year was defined according to the Food and Drug Administration approval dates (Supplemental Table 1, https://links.lww.com/QAD/B826) [37].

Statistical analysis

We examined the annual prevalence as of December 31 of a given year of PWH with LTO among all ART-experienced PWH in care, defined as having had a clinical visit in that year, regardless of HIV RNA level (viral load). Given the approval of new antiretroviral drugs over time within all classes and the introduction of a new antiretroviral class (i.e. INSTI), an individual may contribute to the prevalence of LTO in one calendar year and be classified as non-LTO the following year when a new active drug becomes available.

We used multivariable Cox proportional hazards models to examine time from study entry (baseline) to first occurrence of LTO. The primary variable of interest was entry year in 3-year calendar periods. Other variables were measured as of baseline and included age, sex, race/ethnicity, CNICS site, lowest CD4+ cell count, maximum HIV viral load, ART naive at CNICS entry, and mono-NRTI or dual-NRTI treatment prior to ART initiation. We used multiple imputation with chained equations to address missing lowest CD4+ cell count and maximum viral load values. Participants were followed from study entry until incident LTO, loss to follow-up (12 months without a clinic visit) or administrative censoring (31 December 2017), whichever came first.

To determine whether PWH who were ever classified as having LTO achieved viral suppression over time, we calculated the annual percentage of undetectable HIV viral load (<400 copies/ml) tests stratified by LTO status, such that once designated LTO a person remained classified as LTO going forward. We also accounted for loss to follow-up by incorporating inverse probability of censoring weights. In addition, we examined the number of antiretroviral drugs received and the distribution of antiretroviral drug resistance for the study population at the end of follow-up by LTO status.

Sensitivity analyses

Although genotypic tests likely capture the majority of PWH who have developed any drug resistance, it is possible that tests were not performed in all instances where resistance had occurred. Previous work has shown that the degree of this type of underascertainment is low [38]. To address this potential bias, we conducted a sensitivity analysis using multiple imputation where LTO status was imputed for PWH without genotypic resistance testing who experienced virologic failure (defined as a single HIV viral load >400 copies/ml on ART) followed by any antiretroviral switch within 3 months. We also conducted a sensitivity analysis classifying a given drug with potential low-level resistance (<15 Stanford score) as active [36].

In addition, we evaluated the predictive value of virologic failure and antiretroviral treatment history, as used in prior studies that lacked genotypic resistance data, to identify PWH with LTO [24,30,39,40]. We divided our data into a training set used to develop the LTO prediction model, which included six of the seven CNICS sites, and a test set, which included the remaining CNICS site, to evaluate how well the model predicts the outcome. We used a Bayesian Model Averaging [41,42] approach to identify the optimal model for predicting LTO in 2016 in the development training set of PWH and evaluated potential predictors as of 31 December 2016 including age, sex, race/ethnicity, study entry year, lowest and latest CD4+ cell count, latest viral load, number of antiretroviral drugs and classes received, and number of virologic failures (defined as a single HIV viral load >400 copies/ml on ART) with any antiretroviral switch within 3 months. We also examined alternative definitions for virologic failure requiring two sequential viral loads more than 400 and more than 1000 copies/ml, respectively, followed by any antiretroviral switch within 3 months. Variables with more than 50% chance of being in the best fitting model were included in a logistic prediction model to estimate their association with risk of LTO in the training set. We assessed the performance of this prediction model via the area under the curve (AUC) of the receiver operating characteristics curve in the test set. We examined the ability of virologic failure with antiretroviral switch as well as the number of antiretroviral drugs received to accurately identify PWH with LTO among participants in the test set. Statistical models were fit using Stata version 14 (StataCorp, College Station, Texas, USA).

Results

There were 27 133 ART-experienced PWH in care between 2000 and 2017. Genotypic resistance testing was performed after ART initiation in 8961 PWH, averaging 2.0 tests per individual, totaling 17 803 genotypic tests. The number of ART-experienced PWH in care in a given year increased annually from 3941 in 2000 to over 13 500 in 2017; half of the entire study population was in care at the end of the 18-year study period. As shown in Table 1, 916 PWH were classified as ever having LTO, the majority of whom were male (85%), white (49%), men who had sex with men as a risk factor for HIV acquisition (54%), with median lowest CD4+ cell count 71 (interquartile range [IQR] 15–182) cells/μl, median age 41 years at study entry, and median follow-up 4 years (IQR 2–7). Almost half (45%) had received mono-NRTI or dual-NRTI treatment prior to initiation of ART.

Table 1
Table 1:
Distribution of demographic and clinical characteristics among antiretroviral therapy-experienced persons with HIV by limited treatment options status, 2000–2017.

As shown in Fig. 1, the annual prevalence of PWH with LTO was 5.2–7.5% in 2000–2006 (514 of 6857 in care in 2004), declined significantly to 1.8% in 2007 (151 of 8438 in care in 2007), and decreased to less than 1% in 2012 (107 of 13 350 in care in 2014) through 2017.

Fig. 1
Fig. 1:
Annual prevalence of persons with HIV with limited treatment options among antiretroviral therapy-experienced persons in care by year (2000–2017).

In multivariable analysis, PWH entering the study in 2009–2011 had an 80% lower risk of LTO compared with those entering in 2006–2008 [adjusted hazard ratio (aHR) 0.20; 95% confidence interval (CI): 0.09–0.42], and risk of LTO remained significantly lower in all subsequent calendar periods (Table 2). Lower baseline CD4+ cell count and higher baseline maximum viral load were significantly associated with greater risk of LTO (aHR per 100 higher CD4+ cells/μl 0.82; 95% CI: 0.78–0.87, aHR per 10-fold higher HIV viral load copies/ml 1.37; 95% CI: 1.26–1.49) as were increasing age and male sex (aHR per 10 additional years 1.13; 95% CI: 1.05–1.22, aHR female sex 0.72; 95% CI: 0.59–0.87). Participants who had previously received treatment with mono-NRTIs or dual-NRTIs were more than twice as likely to have LTO compared with those who had not (aHR 2.47; 95% CI: 2.14–2.83).

Table 2
Table 2:
Adjusted hazard ratios for persons with HIV with limited treatment options according to calendar period and baseline demographic and clinical characteristics, N = 27 133a.

On average, 90% of PWH in care in a given year had at least one HIV viral load test in that year, including 92% of PWH with LTO, throughout the study period. As shown in Fig. 2, fewer than 30% of HIV viral load tests among persons with LTO were undetectable in 2001 compared with more than 50% of tests among PWH who did not have LTO. The proportion of undetectable viral load tests among PWH ever classified as having LTO increased to over 80% in 2011 comparable with persons who never had LTO. Results with and without accounting for loss to follow-up did not differ.

Fig. 2
Fig. 2:
Percentage of undetectable HIV viral load tests by year among antiretroviral-experienced persons with HIV by limited treatment option status (2000–2017).

At the end of follow-up, PWH with LTO had received twice the number of antiretroviral drugs as PWH who never had LTO (median 11 [IQR 9–13] versus 5 [3–7]) (Table 3). Further, among persons with any antiretroviral drug resistance, PWH with LTO were resistant to three times the number of antiretroviral drugs compared with PWH who never had LTO (median 16 [IQR 13–19] versus 5 [3–8]). Among all PWH, as well as those with LTO, the most common antiretroviral resistance was to drugs in the NRTI and NNRTI classes. Notably, 54% of PWH with LTO had no active NNRTIs and no more than one NRTI drug available at some point in time.

Table 3
Table 3:
Distribution of antiretrovirals received and antiretroviral resistance among antiretroviral therapy-experienced persons with HIV at the end of follow-up by limited treatment options status.

Results of sensitivity analyses imputing LTO for PWH without genotypic resistance testing who experienced virologic failure and when scoring inactive drugs as at least 15 points did not differ from the main analysis (data not shown). In the Bayesian Model Averaging approaches to evaluate the value of antiretroviral treatment history and virologic failure to predict LTO status, the only significant predictor of LTO in the best fitting model in the training set was the number of antiretroviral drugs received and this was associated with an increased risk of LTO of nearly 70% per additional drug (odds ratio per drug 1.68; 95% CI: 1.58–1.78). However, when the model was evaluated in the test set, the number of antiretroviral drugs received did not predict LTO status despite excellent model fit (AUC 0.92; 95% CI: 0.86–0.98). Compared with LTO defined by genotypic resistance, identifying potential LTO by receipt of at least 14 antiretroviral drugs had 42% sensitivity, 99% specificity, and a positive predictive value (PPV) of 20%, whereas receipt of at least nine antiretroviral drugs had 68% sensitivity, 92% specificity, and a PPV of 5%. Furthermore, compared with LTO defined by genotypic resistance, potential LTO identified by virologic failure with antiretroviral switch had 29% sensitivity, 83% specificity, and a PPV of only 1%. The sensitivity and PPV were also very low for alternative definitions of virologic failure at 21–22% sensitivity and 2% specificity, respectively.

Discussion

In this large multicenter study of over 27 000 ART-experienced PWH in care spanning nearly two decades, the prevalence of PWH with LTO declined significantly from 7.5% in 2004 to less than 2% in 2007 after the introduction of a new antiretroviral class, and since 2012 has remained less than 1% throughout the current ART era. The availability of new drugs in all antiretroviral classes (i.e. NRTI, NNRTI, protease inhibitor, INSTI) contributed to the decreasing prevalence of LTO in PWH throughout the study period. To our knowledge, this study is the first to report the prevalence of PWH with LTO in the most recent time period utilizing longitudinal genotypic resistance data. After accounting for population differences, by 2009 PWH were 80% less likely to have LTO than in previous periods. Most importantly, over the past decade the proportion of PWH in the cohort who ever had LTO and subsequently achieved viral suppression increased dramatically to greater than 80%, which was equivalent to persons who never had LTO. This was likely due to the availability of the INSTI class and NNRTIs and protease inhibitors active against resistant HIV variants [43]. These results support growing evidence of the effectiveness of contemporary ART regimens in achieving virologic control in treatment-experienced PWH [22–24].

As expected, PWH with lower CD4+ cell count and higher HIV viral load, older persons, and those previously treated with mono-NRTIs or dual-NRTIs were significantly more likely to have LTO. We also found that resistance in the NRTI and NNRTI classes was far more common than in the protease inhibitor class. Over half of PWH with LTO had no more than one NRTI and no active NNRTIs at some time, highlighting treatment challenges posed by limited availability of drugs in these classes needed to construct active three-drug regimens for HTE PWH. Thus, the historical focus on HTE defined as PWH with triple class resistance did not adequately capture the clinical significance of reverse transcriptase inhibitor resistance in limiting treatment options [6,18,19,44,45]. Treatment options may be further limited for some patients due to drug intolerance or drug–drug interactions, which were not evaluated in our study.

Studies employing approaches to identify HTE PWH with LTO based on virologic failure or number of prior antiretroviral drug switches, common surrogates for genotypic resistance, are limited by incomplete antiretroviral drug history and lack of information on treatment adherence [11]. We examined the performance of alternative measures of HTE PWH with LTO using three definitions of virologic failure. Irrespective of the definition, virologic failure with antiretroviral switch failed to accurately identify PWH with LTO, but rather resulted in a large proportion of false positive cases as demonstrated by poor PPV. Similarly, the number of antiretroviral drugs received failed to identify PWH with LTO due in part to the low prevalence of PWH with LTO in the contemporary ART era. These findings demonstrate that alternative approaches to identify PWH with LTO in the absence of genotypic resistance data, may have limited clinical utility.

Accurately capturing an individual's resistance profile including archived resistance requires cumulative genotypic test data [38]. Thus prevalence estimates based on the most recent genotypic test [17,18,46–48] or in the setting of limited study follow-up [49] can underestimate true prevalence of PWH with LTO. Studies that are restricted to PWH who had resistance testing are known to overestimate prevalence because they examine a select subgroup not representative of the population of PWH on ART in clinical care [16,27]. Furthermore, analyses based on the number of antiretroviral drugs to which a PWH is resistant rather than how many active drugs they have available, fail to account for decreasing prevalence of LTO as new drugs are introduced. As new treatment approaches including dual drug regimens are utilized, the definition of LTO must continue to evolve while remaining clinically relevant to providers treating HTE PWH.

Strengths of our large multicenter study include comprehensive clinical data on over 27 000 ART-experienced PWH in routine clinical care across the United States with extensive follow-up and robust cumulative genotypic resistance data throughout the 18-year study period that includes a decade of INSTI use. Results of sensitivity analyses imputing LTO status suggest lack of genotypic resistance testing in our cohort was not a factor. Whereas previous studies reported variable outcomes with regard to virologic control [14,28,29,50,51], we show unequivocally that PWH with LTO have benefitted from the introduction of modern, potent antiretroviral drugs in all classes and have been virally suppressed over the past decade to the same extent as persons who never had LTO. The geographic, racial/ethnic, and clinical diversity of our cohort greatly strengthens the generalizability of our findings to PWH in care in the United States.

Results from this large and diverse HIV-infected population demonstrate a dramatic decline in PWH with LTO and a marked increase in virologic control with the introduction of more potent antiretroviral drugs and classes throughout the contemporary ART era. In addition to early and sustained access to ART, new treatment options will be important to support continued improvement in HIV outcomes and prevention of HIV transmission [1,3,52].

Acknowledgements

We dedicate this work to Dr. Benigno Rodriguez, outstanding researcher, trusted colleague, and friend who always put patients first. We miss his wisdom, insights, and the care he exhibited to all of his patients and friends.

Author contributions: K.L.B., R.M.N., J.A.C.D., E.E., T.D.-M., M.Y.K., R.D.M., J.J.E., B.R., K.H.M., E.G., M.S.S., H.C., and M.M.K. contributed to study conception, design, and performance. R.M.N. and J.A.C.D. analyzed the data. K.L.B., R.M.N., J.A.C.D., H.C., and M.M.K. wrote the first draft of the article. All authors were involved in data interpretation, review, writing, and approved the final version of the article.

The project received support from ViiV Healthcare and NIH including CNICS (NIAID grant R24AI067039), and CFARs at the University of Alabama at Birmingham (P30AI027767), University of Washington (P30AI027757), University of California, San Diego (P30AI036214), University of California, San Francisco (P30AI027763), Case Western Reserve University (P30AI036219), Johns Hopkins University (P30AI094189, U01DA036935), Fenway Health/Harvard (P30AI060354), and University of North Carolina Chapel Hill (P30AI50410). No funders had access to primary data.

Conflicts of interest

E.E. has received research support to UAB on her behalf from the Gilead HIV Research Scholarship and ViiV Healthcare. M.Y.K. has received funding to her institution from ViiV Healthcare and Gilead Sciences. J.J.E. is an ad-hoc consultant to Merck, Gilead Sciences, Janssen and ViiV Healthcare, and the University of North Carolina receives funding for clinical trials from Gilead, Janssen and ViiV Healthcare for which he is the site principal investigator. B.R. has received honoraria from Gilead Sciences and ViiV Healthcare that are not related to his participation in this article. K.H.M. has received unrestricted research grants from Gilead Sciences and Merck to study antiretrovirals for prevention and from Janssen to study HIV vaccines and is on Scientific Advisory Boards for Gilead and Merck. C.G. is an employee of ViiV Healthcare, sponsor of this study, and owns stock in GSK, parent company of ViiV Healthcare. M.S.S. has received grant support paid to his institution from ViiV Healthcare and Gilead Sciences. H.M.C. has received grant support paid to her institution from ViiV Healthcare. K.L.B., R.M.N., J.A.C.D., T.D.-M., R.D.M., E.G., and M.M.K. have no conflicts of interest to report.

Presented in part at the 10th International AIDS Society (IAS) Conference on HIV Science, Mexico City, July 2019.

References

1. Kitahata MM, Gange SJ, Abraham AG, Merriman B, Saag MS, Justice AC, et al. Effect of early versus deferred antiretroviral therapy for HIV on survival. N Engl J Med 2009; 360:1815–1826.
2. Lundgren JD, Babiker AG, Gordin F, Emery S, Grund B, Sharma S, et al. Insight Start Study Group. Initiation of antiretroviral therapy in early asymptomatic HIV infection. N Engl J Med 2015; 373:795–807.
3. Cohen MS, Chen YQ, McCauley M, Gamble T, Hosseinipour MC, Kumarasamy N, et al. Antiretroviral therapy for the prevention of HIV-1 transmission. N Engl J Med 2016; 375:830–839.
4. Hogg RS, Bangsberg DR, Lima VD, Alexander C, Bonner S, Yip B, et al. Emergence of drug resistance is associated with an increased risk of death among patients first starting HAART. PLoS Med 2006; 3:e356.
5. Lucas GM, Gallant JE, Moore RD. Relationship between drug resistance and HIV-1 disease progression or death in patients undergoing resistance testing. AIDS 2004; 18:1539–1548.
6. Napravnik S, Keys JR, Quinlivan EB, Wohl DA, Mikeal OV, Eron JJ Jr. Triple-class antiretroviral drug resistance: risk and predictors among HIV-1-infected patients. AIDS 2007; 21:825–834.
7. Di Giambenedetto S, Prosperi M, Fanti I, Bruzzone B, Paolucci S, Penco G, et al. Update on emergence of HIV-1 resistance to antiretroviral drug classes in an Italian national database: 2007–2009. Clin Microbiol Infect 2011; 17:1352–1355.
8. Lima VD, Harrigan PR, Senecal M, Yip B, Druyts E, Hogg RS, et al. Epidemiology of antiretroviral multiclass resistance. Am J Epidemiol 2010; 172:460–468.
9. Panel on Antiretroviral Guidelines for Adults and Adolescents. Guidelines for the use of antiretroviral agents in adults and adolescents living with HIV. In: Department of Health and Human Services; 2009.
10. Sheth AN, Ofotokun I, Buchacz K, Armon C, Chmiel JS, Hart RL, et al. Antiretroviral regimen durability and success in treatment-naive and treatment-experienced patients by year of treatment initiation, United States, 1996–2011. J Acquir Immune Defic Syndr 2016; 71:47–56.
11. Lohse N, Obel N, Kronborg G, Laursen A, Pedersen C, Larsen CS, et al. Declining risk of triple-class antiretroviral drug failure in Danish HIV-infected individuals. AIDS 2005; 19:815–822.
12. Mocroft A, Ledergerber B, Viard JP, Staszewski S, Murphy M, Chiesi A, et al. Time to virological failure of 3 classes of antiretrovirals after initiation of highly active antiretroviral therapy: results from the EuroSIDA Study Group. J Infect Dis 2004; 190:1947–1956.
13. Nakagawa F, Lodwick R, Costagliola D, van Sighem A, Torti C, Podzamczer D, et al. The Pursuing Later Treatment Option II (PLATO II) Project Team for the Collaboration of Observational HIV Epidemiological Research Europe (COHERE) Group. Calendar time trends in the incidence and prevalence of triple-class virologic failure in antiretroviral drug-experienced people with HIV in Europe. J Acquir Immune Defic Syndr 2012; 59:294–299.
14. Phillips AN, Leen C, Wilson A, Anderson J, Dunn D, Schwenk A, et al. Risk of extensive virological failure to the three original antiretroviral drug classes over long-term follow-up from the start of therapy in patients with HIV infection: an observational cohort study. Lancet 2007; 370:1923–1928.
15. Lodwick R, Costagliola D, Reiss P, Torti C, Teira R, Dorrucci M, et al. Triple-class virologic failure in HIV-infected patients undergoing antiretroviral therapy for up to 10 years. Arch Intern Med 2010; 170:410–419.
16. Abraham AG, Lau B, Deeks S, Moore RD, Zhang J, Eron J, et al. Missing data on the estimation of the prevalence of accumulated human immunodeficiency virus drug resistance in patients treated with antiretroviral drugs in North America. Am J Epidemiol 2011; 174:727–735.
17. Pillay D, Green H, Matthias R, Dunn D, Phillips A, Sabin C, et al. Estimating HIV-1 drug resistance in antiretroviral-treated individuals in the United Kingdom. J Infect Dis 2005; 192:967–973.
18. De Luca A, Dunn D, Zazzi M, Camacho R, Torti C, Fanti I, et al. Declining prevalence of HIV-1 drug resistance in antiretroviral treatment-exposed individuals in Western Europe. J Infect Dis 2013; 207:1216–1220.
19. Tamalet C, Fantini J, Tourres C, Yahi N. Resistance of HIV-1 to multiple antiretroviral drugs in France: a 6-year survey (1997–2002) based on an analysis of over 7000 genotypes. AIDS 2003; 17:2383–2388.
20. Audelin AM, Lohse N, Obel N, Gerstoft J, Jorgensen LB. The incidence rate of HIV type-1 drug resistance in patients on antiretroviral therapy: a nationwide population-based Danish cohort study 1999–2005. Antivir Ther 2009; 14:995–1000.
21. von Wyl V, Yerly S, Burgisser P, Klimkait T, Battegay M, Bernasconi E, et al. Long-term trends of HIV type 1 drug resistance prevalence among antiretroviral treatment-experienced patients in Switzerland. Clin Infect Dis 2009; 48:979–987.
22. Cahn P, Pozniak AL, Mingrone H, Shuldyakov A, Brites C, Andrade-Villanueva JF, et al. Dolutegravir versus raltegravir in antiretroviral-experienced, integrase-inhibitor-naive adults with HIV: week 48 results from the randomised, double-blind, noninferiority SAILING study. Lancet 2013; 382:700–708.
23. Castagna A, Maggiolo F, Penco G, Wright D, Mills A, Grossberg R, et al. Dolutegravir in antiretroviral-experienced patients with raltegravir- and/or elvitegravir-resistant HIV-1: 24-week results of the phase III VIKING-3 study. J Infect Dis 2014; 210:354–362.
24. Nance RM, Vannappagari V, Smith K, Calingaert B, Johannes C, Saltus C, et al. Virologic failure among people living with HIV initiating dolutegravir-based versus other recommended regimens in real-world clinical care settings. J Acquir Immune Defic Syndr 2019; 81:572–577.
25. Paquet AC, Solberg OD, Napolitano LA, Volpe JM, Walworth C, Whitcomb JM, et al. A decade of HIV-1 drug resistance in the United States: trends and characteristics in a large protease/reverse transcriptase and co-receptor tropism database from 2003 to 2012. Antivir Ther 2014; 19:435–441.
26. Sista P, Rinehart A, Wasikowski B, Winters B, Pattery T, Bacheler L. Nine-year trends in clinically relevant reduced susceptibility of HIV-1 to antiretrovirals. J Clin Virol 2009; 44:190–194.
27. Hurt CB, Sebastian J, Hicks CB, Eron JJ. Resistance to HIV integrase strand transfer inhibitors among clinical specimens in the United States, 2009–2012. Clin Infect Dis 2014; 58:423–431.
28. Gill VC, Lynch T, Ramazani S, Krentz HB. Reporting on the prevalence of antiretroviral drug resistance in a regional HIV population over 20 years: a word of caution. Antivir Ther 2017; 22:277–286.
29. Costagliola D, Lodwick R, Ledergerber B, Torti C, van Sighem A, Podzamczer D, et al. The Pursuing Later Treatment Option II (PLATO II) Project Team for the Collaboration of Observational HIV Epidemiological Research Europe (COHERE) Group. Trends in virological and clinical outcomes in individuals with HIV-1 infection and virological failure of drugs from three antiretroviral drug classes: a cohort study. Lancet Infect Dis 2012; 12:119–127.
30. Cozzi-Lepri A, Phillips AN, Clotet B, Mocroft A, Ruiz L, Kirk O, et al. Detection of HIV drug resistance during antiretroviral treatment and clinical progression in a large European cohort study. AIDS 2008; 22:2187–2198.
31. Mauro Z, Federica F, Patrizia L, Francesca CS, Valerio T, Paola TM, et al. Continuous evidence of fast HIV disease progression related to class-wide resistance to antiretroviral drugs: a 6 year follow-up analysis of a large observational database. AIDS 2007; 21:1824–1826.
32. Scherrer AU, von Wyl V, Yang WL, Kouyos RD, Boni J, Yerly S, et al. Emergence of acquired HIV-1 drug resistance almost stopped in Switzerland: a 15-year prospective cohort analysis. Clin Infect Dis 2016; 62:1310–1317.
33. Gandhi RT, Tashima KT, Smeaton LM, Vu V, Ritz J, Andrade A, et al. Long-term outcomes in a large randomized trial of HIV-1 salvage therapy: 96-week results of AIDS Clinical Trials Group A5241 (OPTIONS). J Infect Dis 2020; 221:1407–1415.
34. Centers for Disease Control and Prevention. Estimated HIV incidence and prevalence in the United States, 2010–2016. Division of HIV/AIDS Prevention, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention (CDC), U.S. Department of Health and Human Services, Atlanta, Georgia. HIV Surveillance Supplemental Report 2019 Web site. 2019. [Accessed 13 February 2020]
35. Kitahata MM, Rodriguez B, Haubrich R, Boswell S, Mathews WC, Lederman MM, et al. Cohort profile: the Centers for AIDS Research Network of Integrated Clinical Systems. Int J Epidemiol 2008; 37:948–955.
36. Stanford University HIV Drug Resistance Database. 2018. https://hivdb.stanford.edu [Accessed 29 November 2018]
37. AIDSinfo. FDA Approval of HIV Medicines. US Department of Health and Human Services. https://aidsinfo.nih.gov/understanding-hiv-aids/infographics/25/fda-approval-of-hiv-medicines. 2018. [Accessed 29 November 2018]
38. Davy-Mendez T, Eron JJ, Brunet L, Zakharova O, Dennis AM, Napravnik S. New antiretroviral agent use affects prevalence of HIV drug resistance in clinical care populations. AIDS 2018; 32:2593–2603.
39. von Wyl V, Yerly S, Boni J, Burgisser P, Klimkait T, Battegay M, et al. Emergence of HIV-1 drug resistance in previously untreated patients initiating combination antiretroviral treatment: a comparison of different regimen types. Arch Intern Med 2007; 167:1782–1790.
40. Swenson LC, Min JE, Woods CK, Cai E, Li JZ, Montaner JS, et al. HIV drug resistance detected during low-level viraemia is associated with subsequent virologic failure. AIDS 2014; 28:1125–1134.
41. Hoeting JA, Madigan D, Raftery AE, Volinsky CT. Bayesian model averaging: a tutorial. Stat Sci 1999; 14:382–401.
42. Viallefont V, Raftery AE, Richardson S. Variable selection and Bayesian model averaging in case–control studies. Stat Med 2001; 20:3215–3230.
43. Yazdanpanah Y, Fagard C, Descamps D, Taburet AM, Colin C, Roquebert B, et al. High rate of virologic suppression with raltegravir plus etravirine and darunavir/ritonavir among treatment-experienced patients infected with multidrug-resistant HIV: results of the ANRS 139 TRIO trial. Clin Infect Dis 2009; 49:1441–1449.
44. Buchacz K, Baker R, Ward DJ, Palella FJ, Chmiel JS, Young B, et al. Trends in decline of antiretroviral resistance among ARV-experienced patients in the HIV Outpatient Study: 1999–2008. AIDS Res Treat 2012; 2012:230290.
45. Grover D, Copas A, Green H, Edwards SG, Dunn DT, Sabin C, et al. What is the risk of mortality following diagnosis of multidrug-resistant HIV-1?. J Antimicrob Chemother 2008; 61:705–713.
46. Gill VS, Lima VD, Zhang W, Wynhoven B, Yip B, Hogg RS, et al. Improved virological outcomes in British Columbia concomitant with decreasing incidence of HIV type 1 drug resistance detection. Clin Infect Dis 2010; 50:98–105.
47. Di Giambenedetto S, Bracciale L, Colafigli M, Cattani P, Pinnetti C, Bacarelli A, et al. Declining prevalence of HIV-1 drug resistance in treatment-failing patients: a clinical cohort study. Antivir Ther 2007; 12:835–839.
48. Di Giambenedetto S, Zazzi M, Corsi P, Gonnelli A, Di Pietro M, Giacometti A, et al. Evolution and predictors of HIV type-1 drug resistance in patients failing combination antiretroviral therapy in Italy. Antivir Ther 2009; 14:359–369.
49. Masquelier B, Costagliola D, Schmuck A, Cottalorda J, Schneider V, Izopet J, et al. Prevalence of complete resistance to at least two classes of antiretroviral drugs in treated HIV-1-infected patients: a French nationwide study. J Med Virol 2005; 76:441–446.
50. Jones R, Nelson M, Bower M, Powles T, Mandalia S, Gazzard B, et al. Triple-class antiretroviral agent resistance in a large cohort: prevalence and clinical outcomes. Arch Intern Med 2008; 168:1926–1927.
51. Tozzi V, Zaccarelli M, Bonfigli S, Lorenzini P, Liuzzi G, Trotta MP, et al. Drug-class-wide resistance to antiretrovirals in HIV-infected patients failing therapy: prevalence, risk factors and virological outcome. Antivir Ther 2006; 11:553–560.
52. Marschner IC, Collier AC, Coombs RW, D’Aquila RT, DeGruttola V, Fischl MA, et al. Use of changes in plasma levels of human immunodeficiency virus type 1 RNA to assess the clinical benefit of antiretroviral therapy. J Infect Dis 1998; 177:40–47.
Keywords:

antiretroviral drug resistance; antiretroviral therapy experienced; heavily treatment experienced; HIV; limited treatment options

Supplemental Digital Content

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.