Share this article on:

Connection Domain Mutations in Treatment-Experienced Patients in the OPTIMA Trial

Dau, Birgitt MD*; Ayers, Dieter MSc†; Singer, Joel PhD†; Harrigan, P Richard PhD‡; Brown, Sheldon MD§; Kyriakides, Tassos PhD‖; Cameron, D William MD¶; Angus, Brian MD#; Holodniy, Mark MD***

JAIDS Journal of Acquired Immune Deficiency Syndromes: 1 June 2010 - Volume 54 - Issue 2 - pp 160-166
doi: 10.1097/QAI.0b013e3181cbd235
Clinical Science

Objectives: To determine the frequency of mutations in the connection domain (CD) of HIV reverse transcriptase in treatment-experienced patients in the Options in Management with Antiretrovirals trial, their impact on susceptibility to antiretroviral (ARV) drugs, and their impact on virologic outcomes.

Methods: Baseline plasma ARV genotypes and inferred resistance phenotypes were obtained. Frequencies of E312Q, Y318F, G333D, G333E, G335C, G335D, N348I, A360I, A360V, V365I, A371V, A376S, and E399G were compared with a treatment-naive population. The association of CD mutations with inferred IC50 fold changes to nucleos(t)ide reverse transcriptase inhibitors was evaluated. Univariate and multivariate analyses examined the association of CD mutations with a >1 log10 per milliliter decrease in HIV viral load after 24 weeks on a new ARV regimen.

Results: Higher CD mutation rates were seen in Options in Management with Antiretrovirals patients (n = 345) compared with a treatment-naive population. CD mutations were associated with increased inferred IC50 fold changes to abacavir, stavudine, tenofovir, and zidovudine. On univariate analysis, A371V was associated with lack of virologic response, as was having any CD mutation on multivariate analysis.

Conclusions: CD mutations are frequent in treatment-experienced populations. They are associated with reduced susceptibility to some nucleos(t)ide reverse transcriptase inhibitors and with a diminished response to ARV therapy.

From the *AIDS Research Center, Veterans Affairs Palo Alto Health Care System, Palo Alto, CA; †Canadian Institutes for Health Research, Canadian HIV Trials Network, Vancouver, BC, Canada; ‡British Columbia Centre for Excellence in HIV/AIDS, Vancouver, British Columbia, Canada; §Division of Infectious Diseases, Bronx Veterans Affairs Medical Center, New York, NY; ∥Cooperative Studies Program Coordinating Center, Veterans Affairs, West Haven, CT; ¶Division of Infectious Disease, Department of Medicine, Ottawa General Hospital, Ottawa, Ontario, Canada; #Department of Medicine, University of Oxford, Oxford, United Kingdom; and **Department of Medicine, Stanford University, Division of Infectious Diseases and Geographic Medicine, Stanford, CA.

Received for publication July 25, 2009; accepted November 2, 2009.

The Options in Management with Antiretrovirals study was funded in part by the Cooperative Studies Program of the US Department of Veterans Affairs Office of Research and Development, the Canadian Institutes for Health Research, the Canadian HIV Trials Network, and the UK Medical Research Council. This substudy was funded in part by a Veterans Affairs grant to M.H. and by the Canadian HIV Trials Network.

Correspondence to: Birgitt Dau, MD, 3801 Miranda Ave. (132), VA Palo Alto Health Care System, Palo Alto, CA 94304 (e-mail:

Back to Top | Article Outline


Resistance testing of the HIV-1 reverse transcriptase (RT) gene has often included approximately the first 250 amino acids of the RT enzyme and has excluded the C-terminal RT domain, comprising amino acids 289-560.1,2 The connection domain (CD) is located between codons 316 and 437 of HIV-1 RT.2-4 The CD connects the DNA polymerase (codons 1-315) and the RNase H (codons 438-560) domains. Its role in antiretroviral (ARV) drug resistance was recently reviewed.5 The RT enzyme converts viral single-stranded RNA into double-stranded proviral DNA. First, the DNA polymerase creates a DNA strand from the RNA template, and then RNase H degrades the template RNA strand from the DNA:RNA hybrid.6 Both the RNase H and CD have contact with the nucleic acid in the active site.7 Triphosphorylated nucleos(t)ide reverse transcriptase inhibitors (NRTIs) replace the natural nucleoside triphosphates in the active site and are incorporated into the primer chain causing chain termination.8 Excision of the NRTI from the DNA chain allows chain elongation before RNA template degradation by RNase H.9 There is evidence that mutations in RNase H slow its activity, thereby increasing resistance to NRTIs.10,11 Mutations in the CD also decrease RNase H activity.1,3,7,12,13 Thymidine-associated mutations (TAMs) increase nucleotide excision of the chain-terminating NRTI and are synergistic with RNase H mutations in causing resistance to NRTIs in vitro.11,14 TAMs in association with CD mutations are able to further increase NRTI [zidovudine, (ZDV)] excision.1-3,13,14

Many CD mutations are associated with decreases in the in vitro ARV susceptibility of NRTIs and nonnucleos(t)ide reverse transcriptase inhibitors (NNRTIs), both alone and in combination with TAMs or K103N; including ZDV, lamivudine (3TC), abacavir (ABC), nevirapine (NVP), and efavirenz (EFV).1,2,11,15-17 Because of shared selection pressure or effects on viral fitness, some CD mutations may appear in conjunction with primary RT mutations, though these associations are not yet fully defined.2 Some clinical data have emerged: in a large clinical database, the Y318F substitution was associated with a 10-fold decrease in NNRTI susceptibility in clinical isolates and was associated with a history of delavirdine (DLV) and NVP use.16 In a retrospective clinical study, the appearance of N348I was associated with a concurrent increase in HIV viral load.2 ARV drug exposure has recently been linked to the development of CD mutations; in a study from Brazil, the presence of A371V was associated with a history of ZDV exposure.18 Thus, although there is in vitro evidence and limited clinical evidence to suggest that CD mutations may affect NRTI and NNRTI susceptibility and virologic response, the clinical impact of these mutations on ARV susceptibility and virologic outcome is unknown.19

We therefore studied the prevalence of CD mutations in ARV treatment-experienced patients, their effect on ARV susceptibility, and whether they affected virologic outcome in the Options in Management with Antiretrovirals (OPTIMA) trial.

Back to Top | Article Outline


The OPTIMA study is a randomized, prospective, multicenter strategy trial in patients with multidrug resistance who have failed at least 2 ARV regimens.20,21 Patients were randomized to an ARV drug-free period (ARDFP) for 3 months or not (no ARDFP) and to treatment by either standard antiretroviral therapy (ART) (≤4 ARV drugs) or Mega-ART (≥5 ARV drugs). Primary outcomes were time to a new or recurrent AIDS event or death. Secondary outcomes included changes in CD4 count and HIV-1 viral load. The minimum follow-up was 1 year. HIV-1 protease (PR) and RT (codons 1-400) gene sequences and virtual phenotypes (VircoType, Virco) were analyzed from baseline plasma samples from subjects randomized in the OPTIMA trial. Sequences were submitted to Genbank and assigned accession numbers GU581657-GU582076. HIV subtype was determined from the genotype. “Virtual” or estimated phenotypic susceptibility scores (vPSS) were calculated by adding the score for each drug in the patient's initial on-study ARV regimen. A score of 0 indicates no activity (fold change, FC > clinical cutoff 2 (CCO2) or FC > cut-off if no clinical cut-off available), 0.5 indicates partial activity (FC >CCO1 and <CCO2), and 1 indicates full activity (<CCO1 or FC <cut-off). Virco virtual phenotypes define CCO1 as the baseline fold change associated with a 20% loss of the wild-type virologic response due to viral resistance, whereas CCO2 is the fold change associated with 80% loss of the wild-type virologic response.22-24 Clinical cutoffs are updated by Virco routinely; clinical cutoffs updated in January 2008 were used to determine the vPSS scores for each drug. Virologic response was defined as an HIV viral load reduction of >1 log10 per milliliter after 24 weeks of ARV treatment.

Baseline key drug resistance mutations in both the RT and PR genes were determined using the International AIDS Society-USA Drug Resistance Mutations 2008 update.19 The CD mutations selected for analysis were those with in vitro data demonstrating reduced susceptibility to NRTIs or NNRTIs. Included were E312Q (part of the RT C-terminal domain), Y318F, G333D, G333E, G335C, G335D, N348I, A360I, A360V, V365I, A371V, A376S, and E399G. The prevalence of these CD mutations in the treatment-experienced OPTIMA cohort was compared with the prevalence of these mutations in treatment-naive patients with subtype B HIV-1 in the Stanford HIV Drug Resistance Database25 using a Fisher exact test.

The association of CD mutations with primary RT mutations was analyzed using Fisher exact tests. Median fold changes in IC50 to NRTIs and NNRTIs were determined for those patients with and without any of the 13 CD mutations (E312Q-which adjoins the CD, Y318F, G333D, G333E, G335C, G335D, N348I, A360I, A360V, V365I, A371V, A376S, and E399G) identified as causing ARV resistance in vitro. The fold changes in these 2 groups were compared with Wilcoxon rank sum tests. The NRTIs were chosen for further analysis to examine which CD mutations were associated with increases in fold changes. After controlling for CD4 count and HIV-1 viral load, analysis of variance was used to determine whether individual CD mutations were associated with an increase in fold change to any of the NRTIs. Of note, we did not correct for multiple comparisons because these analyses are exploratory and aimed to help determine directions for future study; confirmatory studies will be required.

Logistic regressions were then used to analyze the effect of individual CD and RT mutations on virologic response. CD mutations were analyzed individually using univariate analysis and also in multivariate models that included baseline absolute CD4 count, baseline viral load, treatment arm, and primary RT mutations. Multivariate models were run in a step-wise fashion to determine which factors had the greatest effect on virologic response. Because it is unlikely the effect of individual mutations would be seen on a multivariate analysis that included the vPSS, which is correlated with individual mutations, multivariate models were run both with and without the vPSS. We also compared the virologic outcomes of all samples that included any of the 13 CD mutations of interest with those that contained none. This was done using a step-wise multivariate model that included baseline CD4 count and HIV-1 viral load and individual RT mutations.

Back to Top | Article Outline


Three hundred sixty-eight subjects were randomized, 98% were male, with a mean age of 49 years. At baseline, mean CD4 was 130 cells per cubic millimeter3 and mean viral load was 4.71 log10 copies per milliliter.21 There was extensive prior ARV use.26 Ninety-six percent of patients had used at least 3 NRTIs (median 5). Ninety-seven percent had been exposed to at least 1 NNRTI (median 1), and 63% had been exposed to at least 3 PIs (median of 3). Comparing standard vs. mega-ART and ARDFP versus no-ARDFP in an intent-to-treat analysis, there was no significant difference in the time to the primary outcome for AIDS or death [ARDFP/no ARDFP P = 0.8416; standard/mega P = 0.7909 (from a proportional hazards regression with both treatments included)]. There were also no significant differences in CD4 count or HIV viral load changes between the treatment arms.21

Baseline genotypes and follow-up laboratory values were available for 345 of the 368 randomized subjects. The overwhelming majority of subjects had HIV-1 subtype B (96.8%); subtype C comprised 1.4%, and each of subtypes A, AG, CD, D, and G were present in only 1 or 2 patients each. RT resistance mutations were prevalent; 333 of 345 patients (96.5%) had at least 1 RT mutation. M41L, M184V, and K103N were found in at least half of all patients (Table 1).26 The median number of RT mutations per patient was 7, and the median number of PR gene mutations was 7. The proportion of patients with RT mutations in the ARDFP and no ARDFP group were similar: 160 of 166 and 173 of 179, respectively (P = 0.8942), as in the ART versus mega-ART arms: 174 of 182 and 159 of 163, respectively (P = 0.3258).

Median virtual phenotype IC50 fold change results per drug are presented in Table 2. The median HIV IC50 fold change to most ARV drugs was greater than the CCO1 for susceptibility. Within the NRTI and NNRTI classes, the virtual phenotype results demonstrated median fold changes greater than CCO2 for ABC, DLV, EFV, emtricitabine, 3TC, NVP, and ZDV and fold changes between the CCO1 and CCO2 for the other NRTIs listed in Table 2 except zalcitabine. Baseline virtual phenotypes demonstrated full resistance, with IC50 fold changes above CCO2, to an average of 7.1 drugs in the ARDFP arm, 7.6 in the no ARDFP arm, 7.3 in the ART arm, and 7.4 in the mega-ART arm (ARDFP/no ARDFP P = 0.2203; standard/mega P = 0.7844).

Between 863 and 1925 HIV genotypes from ARV-naive patients in the Stanford HIV Database25 were available for comparison with OPTIMA data for each CD mutation of interest (Table 3). Of the CD mutations analyzed, 5 were present in greater than 5% of the baseline OPTIMA genotypes: G333E, N348I, V365I, A371V, and A376S. Of these, G333E, N348I, A371V, and A376S were present in more than 10% of baseline genotypes (Table 3). Eight mutations were significantly more prevalent in our treatment-experienced population than in the ARV-naive population, including Y318F, G333E, G335D, N348I, V365I, A371V, A376S, and E399G (Table 3). The number of patients with CD mutations in the ARDFP and no ARDFP group were 103 of 166 and 114 of 179, respectively (P = 0.7529), and in the standard versus Mega-ART arms were 117 of 182 and 100 of 163, respectively (P = 0.5730).

Each of the 5 most frequent CD mutations (G333E, N348I, V365I, A371V, and A376S) was significantly associated with the presence of certain TAMS (L210W, T215F, T215Y, K219E, and K219Q), V118I, or M184V but not with other primary RT mutations (Table 4). The fold change for each NNRTI drug was unchanged in the presence of CD mutations. The NRTI drugs ABC, 3TC, stavudine, TDF, and ZDV had significant increases in the IC50 fold change when any of the 13 CD mutations were present (Table 5). We further examined which of the CD mutations were driving the increase in fold change for each of these drugs using analysis of variance. After controlling for baseline CD4 count and HIV-1 viral load, Y318F and A371V were the only CD mutations significantly associated with the increase in fold change to ABC, while A371V was the only CD mutation associated with increase in fold change to TDF and ZDV (Table 6).

Finally, we examined whether any of the 5 most prevalent CD mutations were associated with a lack of virologic response (<1 log10/mL decrease after 24 weeks on ARV therapy). On univariate analysis, only the A371V mutation was associated with a lack of virologic response (P = 0.047) (data not shown); the vPSS was also associated with virologic response (P = 0.002) (data not shown). On multivariate analysis which included the baseline CD4, baseline viral load, effect of the drug free period, effect of standard versus mega-HAART, and individual primary RT and CD mutations, only the baseline CD4 and baseline HIV-1 viral load were associated with virologic response (Table 7). On step-wise multivariate analysis using a binary variable of presence or absence of the 13 CD mutations, baseline CD4, baseline HIV-1 viral load, and individual primary RT mutations, the presence of CD mutations made it less likely a patient would have a virologic response to therapy, along with the baseline CD4 and HIV-1 viral load (Table 8).

Back to Top | Article Outline


We found a high prevalence of primary ARV resistance mutations in this group of highly treatment-experienced patients; each patient had an average of 7 RT and 7 PR mutations. CD mutations were far more frequent in this ARV treatment-experienced population than in a comparative untreated population. CD mutations in the OPTIMA cohort are certainly more prevalent than we have reported because population-based sequencing would only be expected to detect mutations with a prevalence of at least 20% in the total viral population. Clonal analysis performed on these samples would provide more accurate information on the prevalence of CD mutations. The linkage of CD mutations with primary RT mutations could also be investigated with clonal analysis.

CD mutations were associated with increased inferred IC50 fold changes for NRTIs in the OPTIMA cohort. Despite in vitro data showing an effect of Y318F and N348I on NNRTI susceptibility, we did not show an effect of CD mutations on fold changes to the NNRTIs DLV, EFV, or NVP. Median NNRTI, 3TC, and emtricitabine mean IC50 fold changes were well above the CCO2 without CD mutations. This is a result of primary RT mutations, including K103N, Y181C/I, and M184V which were present in high numbers in our population and likely overshadowed the effect of CD mutations. The fact that these viruses had already lost their NRTI and NNRTI susceptibility limited our ability to demonstrate in vivo effects. No median IC50 fold change to any ARV drug shifted clinical categories (from partial activity to no activity) as a result of CD mutations. However, the IC50 fold changes to ABC, 3TC, stavudine, TDF, and ZDV all increased significantly in the presence of CD mutations, which substantiates in vitro data.1,3,14,17 The CD mutation underlying most of the shift in IC50 fold change associated with CD mutations was A371V. This corroborates in vitro data, which shows an increase in the fold change to ZDV, ABC, 3TC, and TDF when A371V is present in association with TAMs and Q509L.14

CD mutations are associated with and predict the presence of primary RT mutations, likely as a result of shared selection pressure (ie, shared ARV treatment history), or possibly because of a functional dependency (ie, compensatory mutations). Though CD mutations clearly have an additive effect on primary RT mutations in vitro, their effect on outcomes in clinical practice remains incompletely characterized. This relates to the difficulty of isolating the effect of a single mutation from a complicated background of mutations and the fact that many significant mutations colocate in the same viral population. Our multivariate analysis found the RT mutation V118I as a predictor of virologic response when the vPSS was not included but did not find an effect of individual CD mutations. V118I contributes to resistance in the presence of TAMs, appears late in the course of treatment, and is probably a proxy for highly mutated virus.19,27,28 V118I was significantly correlated with A371V and A376S in our cohort, which supports the assertion that it tracks with other RT mutations. The fact that no other primary RT mutations, including all those that are known to contribute far more to ARV drug resistance than V118I, were associated with virologic outcome on multivariate analysis demonstrates the difficulty of identifying the effect of individual mutations with this type of analysis.

The presence of any CD mutation reduced the likelihood of virologic response, where individual RT mutations did not. This may represent the importance of CD mutations in the ability of highly treatment-experienced patients to respond to ARV therapy or may represent the subset of patients with the most accumulated resistance mutations overall. Information that might direct providers in ARV selection would be the decreased susceptibility to ABC, tenofovir, and ZDV in HIV-1 carrying the A371V mutation. More information is needed on the effect of this mutation in patient response to ARV therapy. With a larger sample size including more CD mutations, the power to detect the association of virologic outcome with CD mutations would be greater.

It must be kept in mind that multiple analyses were done to look at this data from the different perspectives presented here. Given that this is an exploratory analysis, corrections for multiple comparisons were not done, and therefore all P values presented must be interpreted in this light.

The RNase H domain was not included in our analysis because we did not analyze the RT gene beyond codon 400. However, mutations in the RNase H domain may also have clinical relevance. The epidemiology of mutations at the RNase H domain and their association with CD mutations are important areas for future research, as is the evolution of CD and RNase H mutations over time given varying selection pressures.

In conclusion, CD mutations are highly prevalent in treatment-experienced patients, and our data suggests, even in the face of extensive background NRTI resistance, that the presence of CD mutations may further decrease the susceptibility to some NRTIs. However, there remains insufficient data to recommend sequencing of the entire HIV-1 RT gene for guidance in making ARV treatment decisions. Given that NRTI treatment options are limited in this type of patient population, decisions are often made to continue NRTIs despite significant reduction in or loss of NRTI susceptibility. The clinical significance in ARV susceptibility loss contributed by CD mutations requires further study.

Back to Top | Article Outline


We wish to thank Virco for providing virtual phenotype results for OPTIMA.

Back to Top | Article Outline


1. Nikolenko GN, Delviks-Frankenberry KA, Palmer S, et al. Mutations in the connection domain of HIV-1 reverse transcriptase increase 3'-azido-3'-deoxythymidine resistance. Proc Natl Acad Sci U S A. 2007;104:317-322.
2. Yap SH, Sheen CW, Fahey J, et al. N348I in the connection domain of HIV-1 reverse transcriptase confers zidovudine and nevirapine resistance. PLoS Med. 2007;4:e335.
3. Ehteshami M, Beilhartz GL, Scarth BJ, et al. Connection domain mutations N348I and A360V in HIV-1 reverse transcriptase enhance resistance to 3'-azido-3'-deoxythymidine through both RNase H-dependent and -independent mechanisms. J Biol Chem. 2008;283:22222-22232.
4. Kohlstaedt LA, Wang J, Friedman JM, et al. Crystal structure at 3.5 A resolution of HIV-1 reverse transcriptase complexed with an inhibitor. Science. 1992;256:1783-1790.
5. Gotte M. Should we include connection domain mutations of HIV-1 reverse transcriptase in HIV resistance testing. PLoS Med. 2007;4:e346.
6. Schultz SJ, Champoux JJ. RNase H activity: structure, specificity, and function in reverse transcription. Virus Res. 2008;134:86-103.
7. Julias JG, McWilliams MJ, Sarafianos SG, et al. Mutation of amino acids in the connection domain of human immunodeficiency virus type 1 reverse transcriptase that contact the template-primer affects RNase H activity. J Virol. 2003;77:8548-8554.
8. El Safadi Y, Vivet-Boudou V, Marquet R. HIV-1 reverse transcriptase inhibitors. Appl Microbiol Biotechnol. 2007;75:723-737.
9. Arion D, Kaushik N, McCormick S, et al. Phenotypic mechanism of HIV-1 resistance to 3'-azido-3'-deoxythymidine (AZT): increased polymerization processivity and enhanced sensitivity to pyrophosphate of the mutant viral reverse transcriptase. Biochemistry. 1998;37:15908-15917.
10. Brehm JH, Mellors JW, Sluis-Cremer N. Mechanism by which a glutamine to leucine substitution at residue 509 in the ribonuclease H domain of HIV-1 reverse transcriptase confers zidovudine resistance. Biochemistry. 2008;47:14020-14027.
11. Nikolenko GN, Palmer S, Maldarelli F, et al. Mechanism for nucleoside analog-mediated abrogation of HIV-1 replication: balance between RNase H activity and nucleotide excision. Proc Natl Acad Sci U S A. 2005;102:2093-2098.
12. Delviks-Frankenberry KA, Nikolenko GN, Barr R, et al. Mutations in human immunodeficiency virus type 1 RNase H primer grip enhance 3'-azido-3'-deoxythymidine resistance. J Virol. 2007;81:6837-6845.
13. Delviks-Frankenberry KA, Nikolenko GN, Boyer PL, et al. HIV-1 reverse transcriptase connection subdomain mutations reduce template RNA degradation and enhance AZT excision. Proc Natl Acad Sci U S A. 2008;105:10943-10948.
14. Brehm JH, Koontz D, Meteer JD, et al. Selection of mutations in the connection and RNase H domains of human immunodeficiency virus type 1 reverse transcriptase that increase resistance to 3'-azido-3'-dideoxythymidine. J Virol. 2007;81:7852-7859.
15. Figueiredo A, Tyssen D, Kuiper M, et al. The E399G mutation in the connection domain of the HIV-1 reverse transcriptase potentiates resistance to efavirenz. Presented at: 4th IAS Conference on HIV Pathogenesis, Treatment and Prevention incorporating the 19th ASHM Conference. July 22-25, 2007, Sydney, Australia. Abstract MOPEA060.
16. Harrigan PR, Salim M, Stammers DK, et al. A mutation in the 3' region of the human immunodeficiency virus type 1 reverse transcriptase (Y318F) associated with nonnucleoside reverse transcriptase inhibitor resistance. J Virol. 2002;76:6836-6840.
17. Kemp SD, Shi C, Bloor S, et al. A novel polymorphism at codon 333 of human immunodeficiency virus type 1 reverse transcriptase can facilitate dual resistance to zidovudine and L-2',3'-dideoxy-3'-thiacytidine. J Virol. 1998;72:5093-5098.
18. Santos AF, Lengruber RB, Soares EA, et al. Conservation patterns of HIV-1 RT connection and RNase H domains: identification of new mutations in NRTI-treated patients. PLoS ONE. 2008;3:e1781.
19. Johnson VA, Brun-Vezinet F, Clotet B, et al. Update of the drug resistance mutations in HIV-1: Spring 2008. Top HIV Med. 2008;16:62-68.
20. Kyriakides TC, Babiker A, Singer J, et al. An open-label randomized clinical trial of novel therapeutic strategies for HIV-infected patients in whom ART has failed: rationale and design of the OPTIMA Trial. Control Clin Trials. 2003;24:481-500.
21. Brown S, and The OPTIMA Study Team. Clinical outcomes from OPTIMA: a randomized controlled trial of antiretroviral treatment interruption or intensification in advance multi-drug resistant HIV infection. Presented at: AIDS 2008-XVII International AIDS Conference, August 3-8, 2008, Mexico City, Mexico. Abstract no. LBPE1145.
22. Winters B, Montaner J, Harrigan PR, et al. Determination of clinically relevant cutoffs for HIV-1 phenotypic resistance estimates through a combined analysis of clinical trial and cohort data. J Acquir Immune Defic Syndr. 2008;48:26-34.
23. Winters B, Van Craenenbroeck E, Van der Borght K, et al. Clinical cut-offs for HIV-1 phenotypic resistance estimates: update based on recent pivotal clinical trial data and a revised approach to viral mixtures. J Virol Methods. 2009;162:101-108.
25. Stanford University HIV Drug Resistance Database. Available at: Accessed February 25, 2009.
26. Holodniy M, Singer J, Ayers D, et al. Baseline Antiretroviral Resistance Profile and Correlation with Clinical Events in the OPTIMA Trial. Presented at: 4th IAS Conference on HIV Pathogenesis, Treatment and Prevention; July 22-25, 2007; Sydney, Australia. Abstract TUPEB044.
27. Gianotti N, Galli L, Boeri E, et al. The 118I reverse transcriptase mutation is the only independent genotypic predictor of virologic failure to a stavudine-containing salvage therapy in HIV-1-infected patients. J Acquir Immune Defic Syndr. 2006;41:447-452.
28. Johnson VA, Brun-Vezinet F, Clotet B, et al. Update of the drug resistance mutations in HIV-1. Top HIV Med. 2008;16:138-145.

antiretroviral; clinical trial; HIV; resistance; veterans

© 2010 Lippincott Williams & Wilkins, Inc.