Virologic failures on initial boosted-PI regimen infrequently possess low-level variants with major PI resistance mutations by ultra-deep sequencing.
ABSTRACT It is unknown whether HIV-positive patients experiencing virologic failure (VF) on boosted-PI (PI/r) regimens without drug resistant mutations (DRM) by standard genotyping harbor low-level PI resistant variants. CASTLE compared the efficacy of atazanavir/ritonavir (ATV/r) with lopinavir/ritonavir (LPV/r), each in combination with TVD in ARV-naïve subjects.
To determine if VF on an initial PI/r-based regimen possess low-level resistant variants that may affect a subsequent PI-containing regimen.
Patients experiencing VF on a Tenofovir/Emtricitabine+PI/r regimen were evaluated by ultra deep sequencing (UDS) for mutations classified/weighted by Stanford HIVdb. Samples were evaluated for variants to 0.4% levels. 36 VF subjects were evaluated by UDS; 24 had UDS for PI and RT DRMs. Of these 24, 19 (79.2%) had any DRM by UDS. The most common UDS-detected DRM were NRTI in 18 subjects: M184V/I (11), TAMs(7) & K65R(4); PI DRMs were detected in 9 subjects: M46I/V(5), F53L(2), I50V(1), D30N(1), and N88S(1). The remaining 12 subjects, all with VLs<10,000, had protease gene UDS, and 4 had low-level PI DRMs: F53L(2), L76V(1), I54S(1), G73S(1). Overall, 3/36(8.3%) subjects had DRMs identified with Stanford-HIVdb weights >12 for ATV or LPV: N88S (at 0.43% level-mutational load 1,828) in 1 subject on ATV; I50V (0.44%-mutational load 110) and L76V (0.52%-mutational load 20) in 1 subject each, both on LPV. All VF samples remained phenotypically susceptible to the treatment PI/r.
Among persons experiencing VF without PI DRMs with standard genotyping on an initial PI/r regimen, low-level variants possessing major PI DRMs were present in a minority of cases, occurred in isolation, and did not result in phenotypic resistance. NRTI DRMs were detected in a high proportion of subjects. These data suggest that PIs may remain effective in subjects experiencing VF on a PI/r-based regimen when PI DRMs are not detected by standard or UDS genotyping.
- [Show abstract] [Hide abstract]
ABSTRACT: Technologic advances in HIV sequencing have revolutionized the study of antiretroviral drug resistance and are increasingly moving from the laboratory to clinical practice. These techniques are able to detect HIV drug resistance mutations present at low frequencies not detectable by current HIV genotyping assays. For a number of commonly used antiretroviral medications, such as non-nucleoside reverse transcriptase inhibitors, the detection of these drug-resistant minority variants significantly increases the risk of treatment failure. The level of evidence, though, is insufficient to determine the impact of HIV minority variants for several other classes of antiretroviral medications. Clinicians should be aware of the novel technologies that are moving into routine clinical use and the clinical implications of HIV minority variants. Additional studies are needed to determine the optimal platform for clinical application of these new technologies and to provide guidance to clinicians on the type and frequency of clinically important HIV minority variants.Clinical Infectious Diseases 02/2013; · 9.37 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Clinical relevance of low-frequency HIV-1 variants carrying drug resistance associated mutations (DRMs) is still unclear. We aimed to study the prevalence of low-frequency DRMs, detected by Ultra-Deep Sequencing (UDS) before antiretroviral therapy (ART) and at virological failure (VF), in HIV-1 infected patients experiencing VF on first-line ART. Twenty-nine ART-naive patients followed up in the ANRS-CO3 Aquitaine Cohort, having initiated ART between 2000 and 2009 and experiencing VF (2 plasma viral loads (VL) >500 copies/ml or one VL >1000 copies/ml) were included. Reverse transcriptase and protease DRMs were identified using Sanger sequencing (SS) and UDS at baseline (before ART initiation) and VF. Additional low-frequency variants with PI-, NNRTI- and NRTI-DRMs were found by UDS at baseline and VF, significantly increasing the number of detected DRMs by 1.35 fold (p<0.0001) compared to SS. These low-frequency DRMs modified ARV susceptibility predictions to the prescribed treatment for 1 patient at baseline, in whom low-frequency DRM was found at high frequency at VF, and 6 patients at VF. DRMs found at VF were rarely detected as low-frequency DRMs prior to treatment. The rare low-frequency NNRTI- and NRTI-DRMs detected at baseline that correlated with the prescribed treatment were most often found at high-frequency at VF. Low frequency DRMs detected before ART initiation and at VF in patients experiencing VF on first-line ART can increase the overall burden of resistance to PI, NRTI and NNRTI.PLoS ONE 01/2014; 9(1):e86771. · 3.73 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Given the extreme variability of the human immunodeficiency virus (HIV) and its ability to replicate as complex viral populations, HIV variants with reduced susceptibility to antiretroviral drugs or with specific coreceptor tropism (CCR5 and/or CXCR4) may be present as minority members of the viral quasispecies. The sensitivity of current HIV genotypic or phenotypic assays is limited, and thus, these tests usually fail to detect low-abundance viral variants. Next-generation (deep) sequencing (NGS) produces an enormous amount of information that allows the detection of minority HIV variants at levels unimaginable using standard Sanger sequencing. NGS technologies continue to evolve, opening new and more affordable opportunities to implement this methodology in clinical laboratories, and HIV is not an exception. The ample use of a battery of more effective antiretroviral drugs, together with careful patient monitoring based on HIV resistance testing, has resulted in HIV-infected patients whose disease is usually well-controlled. The vast majority of adherent patients without detectable resistance become virologically suppressed; however, a subset of these patients with undetectable resistance by standard methods may fail antiretroviral therapy, perhaps due to the presence of minority HIV-resistant variants. Novel NGS-based HIV assays with increased sensitivity for identifying low-level drug resistance and/or coreceptor tropism may play an important role in the success of antiretroviral treatments.Current Infectious Disease Reports 04/2014; 16(4):401.
Virologic Failures on Initial Boosted-PI Regimen
Infrequently Possess Low-Level Variants with Major PI
Resistance Mutations by Ultra-Deep Sequencing
Max Lataillade1,2*, Jennifer Chiarella2, Rong Yang1, Michelle DeGrosky1, Jonathan Uy3, Daniel Seekins3,
Birgitte Simen4, Elizabeth St. John4, Elizabeth Moreno4, Michael Kozal2
1Research and Development, Bristol-Myers Squibb, Wallingford, Connecticut, United States of America, 2Yale University School of Medicine and Veterans Affairs Health
Care Systems, New Haven, Connecticut, United States of America, 3Research and Development, Bristol-Myers Squibb, Plainsboro, New Jersey, United States of America,
4454 Life Sciences-Roche Co., Branford, Connecticut, United States of America
Background: It is unknown whether HIV-positive patients experiencing virologic failure (VF) on boosted-PI (PI/r) regimens
without drug resistant mutations (DRM) by standard genotyping harbor low-level PI resistant variants. CASTLE compared
the efficacy of atazanavir/ritonavir (ATV/r) with lopinavir/ritonavir (LPV/r), each in combination with TVD in ARV-naı ¨ve
Objective: To determine if VF on an initial PI/r-based regimen possess low-level resistant variants that may affect a
subsequent PI-containing regimen.
Methods/Results: Patients experiencing VF on a Tenofovir/Emtricitabine+PI/r regimen were evaluated by ultra deep
sequencing (UDS) for mutations classified/weighted by Stanford HIVdb. Samples were evaluated for variants to 0.4% levels.
36 VF subjects were evaluated by UDS; 24 had UDS for PI and RT DRMs. Of these 24, 19 (79.2%) had any DRM by UDS. The
most common UDS-detected DRM were NRTI in 18 subjects: M184V/I (11), TAMs(7) & K65R(4); PI DRMs were detected in 9
subjects: M46I/V(5), F53L(2), I50V(1), D30N(1), and N88S(1). The remaining 12 subjects, all with VLs,10,000, had protease
gene UDS, and 4 had low-level PI DRMs: F53L(2), L76V(1), I54S(1), G73S(1). Overall, 3/36(8.3%) subjects had DRMs identified
with Stanford-HIVdb weights .12 for ATV or LPV: N88S (at 0.43% level-mutational load 1,828) in 1 subject on ATV; I50V
(0.44%-mutational load 110) and L76V (0.52%-mutational load 20) in 1 subject each, both on LPV. All VF samples remained
phenotypically susceptible to the treatment PI/r.
Conclusion: Among persons experiencing VF without PI DRMs with standard genotyping on an initial PI/r regimen, low-
level variants possessing major PI DRMs were present in a minority of cases, occurred in isolation, and did not result in
phenotypic resistance. NRTI DRMs were detected in a high proportion of subjects. These data suggest that PIs may remain
effective in subjects experiencing VF on a PI/r-based regimen when PI DRMs are not detected by standard or UDS
Citation: Lataillade M, Chiarella J, Yang R, DeGrosky M, Uy J, et al. (2012) Virologic Failures on Initial Boosted-PI Regimen Infrequently Possess Low-Level Variants
with Major PI Resistance Mutations by Ultra-Deep Sequencing. PLoS ONE 7(2): e30118. doi:10.1371/journal.pone.0030118
Editor: Alan Landay, Rush University, United States of America
Received July 15, 2011; Accepted December 9, 2011; Published February 15, 2012
This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for
any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Funding: Funds for this study were provided by Bristol-Myers Squibb, Global Clinical Research. The funders participated in study design and data collection.
Competing Interests: The authors have read the journal’s policy and have the following conflicts: Max Lataillade, Rong Yang, Michelle DeGrosky, Daniel Seekins,
Jonathan Uy, and Donnie McGrath are employees of Bristol Myers Squibb, who funded this study. Yale University receives grant support from Merck, Pfizer, Gilead,
Abbott and Bristol-Myers Squibb for studies that MJK serves as the principal investigator. Michael J. Kozal receives royalties from a patent owned by Stanford
University for some HIV diagnostic tests. Elizabeth P. St. John and Birgitte B. Simen are employees of 454 Life Sciences-Roche Co. There are no products in
development or marketed products to declare. This does not alter the authors’ adherence to all the PLoS ONE policies on sharing data and materials. The four
patent numbers are: 5,631,128; 5,650,268; 5,856,086; and US RE38,352E.
* E-mail: Max.lataillade@BMS.com
HIV treatment guidelines recommend combination antiretro-
viral (ARV) regimens for treatment naive patients consisting of a
nucleoside/nucleotide reverse transcriptase inhibitor (N(t)RTI)-
based backbone along with either a non-nucleoside reverse
transcriptase inhibitor (NNRTI), an integrase inhibitor (INI) or a
ritonavir (RTV)-boosted protease inhibitor (PI) . At the end of
2009, 584,000 HIV positive patients under care were receiving
ARVs in the USA . Approximately 40% of patients were
initiated on an N(t)RTI+RTV-boosted PI (PI/r) as a first line
The most common drug resistance mutation pattern by
standard HIV genotyping in patients experiencing virologic failure
on an initial regimen of Tenofovir/Emtricitabine (TDF/FTC) plus
a ritonavir boosted protease inhibitor (PI/r) is a solitary reverse
transcriptase (RT) M184V mutation or no resistance mutations
[3–7]. It is not known if HIV patients on a PI/r regimen
experiencing virologic failure without evidence of PI resistance
mutations by standard genotyping harbor low-level PI-resistant
PLoS ONE | www.plosone.org1February 2012 | Volume 7 | Issue 2 | e30118
variants that are below the detection levels of standard genotyping
methods. Further, if low-level resistant variants exist in these
patients, could they affect the use of a subsequent PI/r containing
The objective of this study was to determine if patients with
virologic failure on an initial PI/r-based regimen without
resistance by standard genotyping possess low-level drug resistant
variants that could affect the use of a subsequent PI/r containing
regimen. To investigate this question, ultra deep sequencing was
performed on the virologic failure specimens from the CASTLE
study [5–7]. CASTLE compared the efficacy of atazanavir/
ritonavir (ATV/r) with lopinavir/ritonavir (LPV/r), each in
combination with TDF/FTC in ARV-naı ¨ve patients through 96
weeks of treatment [5–7].
All available Week 48 and 96 CASTLE virologic failure
specimens from patients without PI resistance mutations by
standard genotype were evaluated by ultra deep sequencing
(454-Life Sciences/Roche, Branford, CT) for N(t)RTI+PI/r
mutations (see Figure 1a) and classified by Stanford HIVdb
algorithm . Virologic failure specimens with HIV viral loads
.1,000 copies/mL allowed for ultra deep sequencing to be
attempted. Ultra deep sequencing was performed on both the
protease (PR) and RT genes for virologic failure specimens with
HIV viral load .10,000 c/mL. Virologic failure specimens with
HIV viral loads ,10,000 c/mL had ultra deep sequencing
performed only for the PR gene. All drug resistance mutations
and polymorphisms were evaluated by Stanford HIV db
algorithm; specific PI mutations with a value .12 for the PI used
in the study (ATV and LPV) were considered significant resistance
mutations. Ultra deep sequencing was performed as described in
previous studies [9–10]. Ultra deep sequencing was performed to
0.4% variant detection levels depending upon the sample HIV
viral load. Sampling of HIV RNA from a plasma sample follows
Poisson distribution and is subject to the stochastic effects of
sampling variation [9,11,12]. An estimated mutational load (ML)
was calculated by percent of variant detected x sample viral load
(Roche Amplicor Assay). Historical standard genotypes and
phenotypes were available on all CASTLE specimens for baseline
and virologic failure samples (Monogram Biosciences, South San
Francisco, CA). Susceptibility cut-off used for TDF, FTC, ATV/
RTV and LPV/RTV in CASTLE were clinical cut-offs (fold
change-FC) from Monogram Biosciences (TDF.1.4, FTC/
3TC.3.5, ATV/RTV.5.2, LPV/RTV.9).
All IRB approvals for this study were already obtained under
the parent CASTLE study. All subjects had given informed
consent. The parent CASTLE study is registered with Clinical-
Trials.gov, number NCT00272779.
Of 78 patients with virologic failure at Weeks 48 and/or 96
from the CASTLE study without PI resistance mutations by
standard genotype, 57 were available for ultra deep sequencing
(Figure 1a). Of the 57 virologic failure specimens, 36 had a viral
load .1,000 copies/mL, which allowed for ultra deep sequencing
to be attempted. Ultra deep sequencing was performed on both
the PR and RT genes for all 24 (66.7%) specimens with HIV viral
load .10,000 c/mL. The remaining 12 (33.3%) virologic failure
specimens with HIV viral loads ,10,000 c/mL had ultra deep
sequencing performed only for the PR gene (Figure 1a). For the 24
samples with ultra deep sequencing for both PR and RT, 5/24
(20.8%) were wild type at all codon positions (no resistance
mutations). Nineteen of 24 (79.2%) had any PI, N(t)RTI and/or
NNRTI resistance mutations; 18/24 (75%) had N(t)RTI muta-
tions: M184V/I (11), TAMs(7) and K65R (4), and 9/24 (37.5%)
had PI mutations: M46I/V(5), F53L(2), I50V(1), D30N(1) &
N88S(1) (Figure 1b). Of the 12 specimens that had ultra deep
sequencing of only the protease gene, 4/12 (33.3%) had low level
PI mutations: F53L(2), L76V(1), I54S(1) and G73S(1) (Figure 1b).
Overall, 3/36 (8.3%) patients had PI mutations identified with
Stanford-HIVdb weights .12 for ATV or LPV: 1 patient on ATV
with N88S (0.43%-mutation load 1,828) and 2 patients on LPV:
one with I50V (0.44%-mutation load 110) and the remaining
patient with L76V (0.52%-mutation load 20). The mutations
identified by UDS in each specimen and the estimated mutational
loads with the corresponding phenotypes are listed in Table S1
and Table S2. All samples with low-level variants with PI
mutations remained phenotypically susceptible to PIs (Table S1
and Table S2-Historical phenotypes). Out of 11 patients with
M184V/I at virologic failure, 4 with M184V developed
phenotypic resistance to FTC/3TC (variant range 95.93% to
100%; phenotypic range 57.54 to 75.63). Of the seven patients
with a M184V/I without phenotypic resistance to 3TC/FTC, the
variant levels ranged from 0.42% to 31.1%. Four patients had a
K65R at virologic failure (variant range 0.52% to 1.28%), and
none developed phenotypic resistance to TDF (Table S1).
The results from this study demonstrated that among patients
on an initial PI/r based regimen experiencing virologic failure
without PI resistance mutations by standard genotyping, low-level
variants possessing major PI mutations were present in a minority
of cases, occurred in isolation, and did not result in phenotypic PI
resistance. Only 3 patients (8.3%) had a PI mutation with a
Stanford algorithm weight .12 for the PIs used in the study (ATV
or LPV). However, they were at very low levels with low
mutational loads. It should be noted that none of the low-level
variants with a major PI mutation were at levels $1% of the viral
population, a level that has been shown to be clinically significant
for NNRTIs based regimens . However, the PI mutations
identified in these specimens were known to be associated with
drug resistance to the corresponding PI. Of note, N(t)RTI
resistance mutations were detected in a high proportion (75%) of
virologic failure specimens and with greater frequency than PI
mutations, which may partly explain the viral breakthrough in
these patients. Interestingly, not all patients with M184V/I
developed phenotypic resistance to FTC/3TC. Eleven of 24
patients had low or high level M184V/I variants; 7 had M184V/I
at $1% of the viral population, however only 4 of the 7 developed
phenotypic resistance to FTC/3TC (all had M184V variant levels
.95%). None of the patients with K65R (all at ,2%) developed
phenotypic resistance to TDF (See Table S1).
These findings underscore the limitation of stand-alone
phenotypic susceptibility results and emphasize the importance
of complementary and/or more sensitive genotyping techniques
when evaluating for resistance . In addition, these data further
suggest that the impact of drug resistance mutations on therapy is
multi-factorial and should take into account the proportion of the
variant present, the mutational load, and the specific mutation
As previously reported by our group, low-level variants
possessing PI mutations are infrequently identified in treatment-
naı ¨ve persons and did not significantly affect the efficacy of PI/r
based regimens , a finding in contrast to low-level NNRTI
resistance mutations and NNRTI-based regimens . In this
Virologic Failures on Initial Boosted-PI Regimen
PLoS ONE | www.plosone.org2 February 2012 | Volume 7 | Issue 2 | e30118
study, we have shown that low-level PI drug resistant variants are
also infrequently detected by ultra deep sequencing when patients
experience virologic failure on TDF/FTC+PI/r without PI
mutations detected by standard genotype.
The results of this study may have implications for first line and
subsequent ARV sequencing strategies. According to current
guidelines, second line ARV regimens for subjects with resistance
to NNRTI or PI/r based regimen should consist of at least 2,
preferably 3 new and fully susceptible ARVs . This decision
should be based on drug history, genotypic and phenotypic data,
and the mechanism of action of the new drugs. Of note, drug
potency and viral susceptibility are as important as the number of
new drugs prescribed . As shown in this study, patients who
failed a PI/r regimen in CASTLE without PI resistance by
standard genotype while remaining phenotypically susceptible
infrequently had low level variants resistant to the PI/r used or to
any other relevant PI/r. This data suggest that boosted PIs may
still provide good safety and efficacy and could possibly be re-used
in later lines of therapy with a different ARV backbone if no PI
mutations are present by ultra deep sequencing and as long as
adherence is addressed. However, this hypothesis would need to
be formally tested in clinical studies.
Our study had several limitations. Given the requirement for
other predetermined study analyses we were only able to recover
Figure 1. Description and UDS results for virologic failure samples. a: Virologic failure samples for UDS. UDS: Ultra Deep Sequencing; VF:
Virologic Failure. 78 subjects had virologic failure at week 48 and/or 96. *21 samples were either exhausted or could not be located. 57 patients with
virologic failure without PI resistance had samples for UDS. 21 patients failed with HIV RNA,1,000 copies and UDS could not be performed. 36
unique patients had UDS data. b: UDS results for 36 Virologic Failures. VF: Virologic Failure; PR: Protease; RT: Reverse Transcriptase. UDS: Ultra Deep
Sequencing; DRMs: Drug Resistance Mutations. 36 patients without PI resistance mutations at VF by standard genotype were evaluated by UDS. 36
patients with VF had PR evaluated by UDS. 24/36 samples had HIV VL.10,000 c/ml. 9/24 had PI DRMs at low levels. 12/36 samples had HIV
VL,10,000 c/ml. 4/12 had PI DRMs at low levels. Only 3/36 patients with VF had PI mutations with HIVdb weight .12 for ATV or LPV. 24/36 patients
with HIV VL.10,000 c/ml had RT evaluated by UDS. 18/24 samples had NRTI DRMs.
Virologic Failures on Initial Boosted-PI Regimen
PLoS ONE | www.plosone.org3February 2012 | Volume 7 | Issue 2 | e30118
specimens for 57 of 78 Week 48 and/or 96 CASTLE virologic
failures. Further, we were unable to perform ultra deep sequencing
on 21 virologic failure specimens that had an HIV viral load
#1,000 c/mL. Full length reverse transcriptase ultra deep
sequencing for samples with HIV VL.10,000 c/mL was limited
to 24 samples. Given stochastic effects of RNA sampling for the
samples with HIV VL,10,000 c/mL, the reported variants may
or may not represent the proportion of variants in plasma. Thus,
we are unable to confidently analyze all specimens to low levels
which may have lead to an undetected selection bias.
In conclusion, among patients on an initial PI/r based regimen
experiencing virologic failure without PI mutations by standard
genotyping, low-level variants possessing major PI mutations by
ultra deep sequencing were infrequently identified and if PI
mutations were present they were found in isolation and did not
result in phenotypic PI resistance. Only 3 patients (8.3%) had a PI
mutation with a Stanford algorithm weight .12 for ATV or LPV,
however they were at very low levels (,1%) with low mutational
loads. NRTI mutations were detected in a high proportion of
patients with virologic failure and with greater frequency than PI
mutations. These data suggest that PIs may remain effective in
patients experiencing virologic failure on a PI/r-based regimen,
and may be re-used in subsequent ARV regimens with new
optimized background therapy when PI resistance are not detected
by standard or ultra deep sequencing. As the HIV population
grows older, clinicians will need to conserve agents and use potent
and relatively safer ARVs for as long as possible.
These data were presented in part at the 2010 International
HIV & Hepatitis Virus Drug Resistance Conference. Dubrovnik,
Croatia, Abstract #102.
24/36 VF subjects. Ultra deep sequencing, mutational load,
and phenotypic susceptibility results for 24 samples from subjects
experiencing virologic failure on TDF+FTC+PI/r. # UDS for
two virologic failure (VF) timepoints – both VF specimens had a
M184V mutation. NRTI DRMs: M184V and K65R are bolded
in black. PI DRMs with a Stanford HIVdb weight .5 are bolded
in black with an asterix *. Please note that if a sample had different
HIV variants with the same mutation they are listed separately e.g.
Subject #2 had different HIV variants with a M184V at 3.92%
and M184V at 94.94%.
UDS data for RT and PR and phenotype for
VF subjects. Ultra deep sequencing and mutational load for PIs
only, and phenotypic susceptibility results for 12 samples from
subjects experiencing virologic failure on TDF+FTC+PI/r. PI
DRMs with a Stanford HIVdb weight .5 are bolded.
UDS data for PR only and phenotype for 12/36
Conceived and designed the experiments: ML MK. Performed the
experiments: JC ML MK ESJ EM. Analyzed the data: ML MK JC RY
BS. Contributed reagents/materials/analysis tools: MDG JU DS RY.
Wrote the paper: ML MK JC. Manuscript review: JC JU DS RY.
1. Panel on Antiretroviral Guidelines for Adult and Adolescents (2008) Guidelines
for the use of antiretroviral agents in HIV-1 infected adults and adolescents.
Department of Health and Human Services. January 29, 2008; 1–128.
pdf. Accessed 2010 Aug 22.
2. Synovate HIV Market Research (2009) all rights reserved.
3. MacArthur RD, Novak RM, Peng G, Xiang Y, Huppler Hullsiek K, et al. (2006)
Long-Term Clinical and Immunologic Outcomes Are Similar in HIV-Infected
Persons Randomized to NNRTI versus PI versus NNRTI+PI-based Antiretro-
viral Regimens as Initial Therapy: Results of the CPCRA 058 FIRST Study.
LANCET 368: 2125–2135.
4. Kozal MJ, KHuppler Hullsiek, Macarthur RD, Berg-Wolf M, Peng G, et al.
(2007) The Incidence of HIV Drug Resistance and Its Impact on Progression of
HIV Disease Among Antiretroviral Naı ¨ve Participants Started on Three
Different Antiretroviral Therapy Strategies. HIV ClinTrials 6(6): 357–370.
5. Molina JM, Andrade-Villanueva J, Echevarria J, Ploenchan C, Corral J, et al.
(2008) Once-daily atazanavir/ritonavir versus twice-daily lopinavir/ritonavir,
each in combination with tenofovir and emtricitabine, for management of
antiretroviral-naive HIV-1-infected patients: 48 week efficacy and safety results
of the CASTLE study. LANCET 372(9639): 646–655.
6. Molina JM, Andrade-Villanueva J, Echevarria J, Chetchotisakd P, Corral J,
et al. (2008) Atazanavir/Ritonavir vs. Lopinavir/Ritonavir in Antiretroviral-
Naı ¨ve HIV-1–Infected Patients: CASTLE 96 Week Efficacy and Safety, 48th
Annual ICAAC/IDSA 46th Annual Meeting, Washington, DC.
7. Lataillade M, Yang R, Mancini MD, McGrath D, et al. (2008) Impact of HIV
viral diversity and baseline resistance on treatment outcomes and the emergence
of resistance: The CASTLE study 48-week results. Journal of the International
AIDS Society - 11(suppl 1): P180.
8. Rhee SY, Gonzales MJ, Shafer RW (2003) Human immunodeficiency virus
reverse transcriptase and protease sequence database. Nucleic Acids Res - 31:
298–303. Available: http://hivdb.stanford.edu.
9. Simen BB, Simons JF, Hullsiek KH, Novak RM, Macarthur RD, et al. (2009)
Low Abundance Drug Resistant Viral Variants in Chronically HIV-infected
Antiretroviral-Naı ¨ve Patients Significantly Impact Treatment. Journal of
Infectious Diseases 199(5): 693–701.
10. Le T, Chiarella J, Simen BB, Hanczaruk B, Egholm M, et al. (2009) Low-
abundance HIV drug-resistant viral variants in treatment-experienced persons
correlate with historical antiretroviral use. PLoS One- 4(6): e6079.
11. Stenman J, Lintula S, Rissanen O (2003) Quantitative detection of low-copy-
number mRNAs differing at single nucleotide positions. Biotechniques 34:
12. Shafer RW (2009) Low-Abundance Drug-Resistant HIV-1 Variants: Finding
Significance in an Era of Abundant Diagnostic and Therapeutic Options.
Journal of Infectious Diseases: 199: 610–612.
13. Underwood MR, Ross LL, Irlbeck DM, Gerondelis P, Rouse E, et al. (2009)
Sensitivity of phenotypic susceptibility analyses for nonthymidine nucleoside
analogues conferred by K65R or M184V in mixtures with wild-type HIV-1.
J Infect Dis 199(1): 84–8.
14. Lataillade M, Chiarella J, Yang R, Schnittman S, Wirtz M, et al. (2010)
Prevalence and Clinical Significance of HIV Drug Resistance Mutations by
Ultra-deep Sequencing in Antiretroviral-naı ¨ve Subjects in the CASTLE Study.
PLoS ONE 5(6): e10952.
Virologic Failures on Initial Boosted-PI Regimen
PLoS ONE | www.plosone.org4 February 2012 | Volume 7 | Issue 2 | e30118