1

1. DRUG RESISTANCE: DEFINITIONS AND CONCEPTS

• José Ramón Santos • Marc Noguera-Julian • Maria Casadellà • Josep Maria Llibre • Bonaventura Clotet • Roger Paredes

1.1. INTRODUCTION

DefinitionAntiretroviral drug resistance is the ability of HIV to replicate in the presence of

antiretroviral drug concentrations that suppress viral replication of non-resistant virus.Most antiretroviral drugs are competitive inhibitors, i.e. they inhibit a viral enzyme

(protease, PR; reverse transcriptase, RT; integrase, IN) by competing with its natural substrate for attachment to the enzyme’s catalytic site. Resistance to competitive inhibitors is a function of viral susceptibility and the drug levels achieved in the target cells. Higher drug levels can suppress partially resistant viruses, with resistance occurring in a continuum. Viral susceptibility to competitive inhibitors is expressed as the drug concentration able to inhibit virus growth in vitro to 50% (50% inhibitory concentration, IC50) or 90% (IC90), relative to a wildtype reference virus.1

The small-molecule CCR5 antagonists, are allosteric inhibitors of the human CCR5 transmembrane protein, one of the two co-receptors required for HIV entry into target cells. Due to their allosteric inhibition, decrease in viral susceptibility to CCR5 antagonists is reflected by progressive decreases in the percent of maximal inhibition rather than by shifts in IC50. Further increases in drug concentrations once drug resistance is established do not achieve further virological suppression.

PathogenesisHIV has a quasispecies distribution.1 Soon after infection with a relatively homogeneous

viral population, viral replication ensues at an extraordinary rate: 109-12 new virions are generated every day. Because HIV’s RT lacks proofreading ability, 10-3 to 10-4 mutations (one or two per genome) are spontaneously generated per replication cycle.2, 3 Given HIV’s high replication rate, any single mutant and some dual mutants could be generated per day. Most mutations are deleterious and drive mutant viruses to extinction. Others, have neutral or beneficial effects on HIV’s replicative capacity and remain incorporated in the quasispecies.

Variants in the virus quasispecies may have different fitness in different environments.4 The variant with better ability to replicate in the absence of therapy, the wildtype (WT) variant, predominates before therapy initiation. Mutants with a fitness advantage in the presence of therapy remain at very low levels in the absence of treatment. However, they can outcompete the WT within days after therapy initiation if viral replication is not averted. Secondary mutations often accumulate in the presence of continued viral replication; they compensate the potential fitness losses derived from primary resistance mutations and increase cross-class resistance.

2

❖ Factors involved in the emergence and evolution of drug-resistance

• Rapid turnover of HIV-1 (half-life free virus <2 hours).• Large amounts of daily virus production (1010) • High error rate of the reverse transcriptase (RT) (~1:104).• Recombination (1 recombination event per every 3-8 mutations)• Incomplete suppression of viral replication in subjects under therapy (suboptimal

therapies, low adherence, malabsorption, etc…).• Genetic barrier of antiretroviral agents contained in a regimen.• Magnitude of resistance conferred by mutations• Viral Fitness• Interaction between resistance mutations and pathways• Adherence in the context of different treatments

Risk of resistance by antiviral activity of therapyThe likelihood of developing antiretroviral resistance depends on the relative potency

of the antiretroviral regimen and the degree of ongoing replication in the presence of therapy.5,6,7 (Figure 1) A regimen with small antiviral potency creates minimal selective pressure on the virus and leads to slow resistance evolution, even if replication persists. A more potent regimen that is unable to suppress viral replication leads to increased selective pressure over the virus, which rapidly accumulates resistance. Finally, a highly potent regimen that decreases viral replication to minimal levels is associated with slow resistance accumulation, despite the potent selective pressure exerted on the virus.

Figure 1. Relation between antiviral drug activity and emergence of resistance. ✕

3

❖ Adherence and antiretroviral resistance

• Each antiretroviral therapeutic class has a unique adherence-resistancerelationship (Figure 2)8.• NNRTI-treated individuals rarely develop resistance at high levels of adherence due to the virological effectiveness of these regimens. NNRTI resistance develops rapidly at moderate to low levels of resistance due to the low ‘fitness’ costs associated with single mutations. • Unboosted PI-treated individuals may develop resistance at high levels of adherence because residual viral replication is often seen in such patients. PI resistance is uncommon at low levels of adherence because of the significant fitness costs associated with these mutations. • Resistance to a ritonavir-boosted PI is only possible in a narrow range of adherence where there is sufficient drug around to present for mutations that reduce fitness while still allowing residual viral replication.

6

Figure 2. Relationship between medication adherence and the risk of developing PI or NNRTI drug resistance.8

Cross-resistance, hypersusceptibility and replication capacity Given the molecular structure similarities within compounds of the same antiretroviral

family and their interaction with similar target sites, the emergence of resistance to one drug may extend to the other drugs of the same family.

On the other hand, some mutations conferring high-level resistance to one agent may increase viral susceptibility to another compound, resulting in a so-called “hypersusceptible” virus to the second agent.

4

In addition, resistance-conferring mutations may decrease replication capacity in com-parison with the WT virus. The clinical correlates of replication capacity measurements, however, remain unclear.

Minority drug-resistant variants Resistant HIV can be transmitted from person-to-person. In addition, the high turnover

and production of genetic variants ensure that every possible variant containing one resistance mutation and many variants with two resistance mutations can be spontaneously generated in treatment-naïve individuals. Resistant variants present at a frequency of less than 15-20% of the viral population unlikely to be detected by standard drug resistance assays.

Several studies suggest that pre-existing minority NNRTI-resistant variants increase the risk of virological failure to first-line NNRTI-including regimens 3 to 6-fold.9-11 In addition, its detection by means of ultrasensitive HIV-1 genotyping tests has demonstrated to be a useful tool to improve the GSS-based prediction of virological response in treatment-experienced patients.12 Studies are ongoing to establish clinically meaningful thresholds that discriminate outcomes with high sensitivity and specificity. The clinical role of minority resistant variants in antiretroviral regimens with higher genetic barrier is also under evaluation.

❖ Two major mechanisms are involved in HIV resistance to NRTI and NNRTI: 1) impairment of the incorporation of the analogue into DNA and 2) removal of the prematurely terminated DNA chain.13, 14

• Steric exclusion of the analogue incorporation: Mutations M184V, K65R and the Q151M complex promote resistance by selectively impairing the ability of reverse transcriptase to incorporate an analogue into DNA.

• Removal of the analogue from the terminated DNA chain: This mechanism is associated with the thymidine analogue mutations. TAMs induce removal of the nucleoside analogue from the 3’ end of the terminated DNA chain. This process involves an ATP- or pyrophosphate-mediated attack to the phosphodiester bond linking the nucleoside analogue to the DNA chain. Entry of ATP and pyrophosphate, a by-product of DNA polymerization, is facilitated by the structure of a reverse transcriptase expressing TAMs. However, such entry is significantly decreased in the presence of the M184V mutation, what explains the difficulty for TAMs to emerge in the presence of M184V.

5

❖ Implications of drug resistance

• Loss of treatment efficacy• Increase in complexity and decrease in tolerability of therapy• Increased risk of virological failure to subsequent ART • Increased cost of HIV management• Increase in susceptibility to some drugs • Some authors have suggested increased mortality, but that is uncertain

❖ Drug resistance can be inferred from genotypic or phenotypic assays.

• Genotypic resistance: presence of mutations in the HIV genome, which are known to be associated with phenotypic resistance to one or more drugs.

• Phenotypic resistance: Increased ability of a virus to replicate in the presence of drugs, relative to a wild type reference strain.

❖ There are two ways of measuring drug resistance:

Genotypic testing: Comparison of the genomic nucleotide composition of a patient’s virus with that of a wild-type reference strain. Mutations are coded as: “wildtype aminoacid” -gene codon where the mutation occurs – “mutant aminoacid”.

6

Phenotypic testing: Assessment of the susceptibility of a virus to antiretroviral drugs in a virus replication assay. Results can be expressed as:

* Competitive inhibitors

IC50, IC90, IC95: Concentration (in µ g/ml or µ M) of drug needed to inhibit the virus growth in vitro by 50%, 90% or 95%, respectively.

X-Fold rise IC50 from the patient isolate or =Fold change IC50 from a wild type laboratory strain

* Allosteric inhibitors (CCR5 antagonists)PMI: Percent Maximal Inhibition: plateau in drug concentrations <95% relative to

the maximal inhibition of a wildtype virus

Drug Concentration (log scale) Drug Concentration (log scale)

Perc

ent I

nhib

ition

Perc

ent I

nhib

ition

Figure 3. Inhibition curves of competitive and non-competitive inhibitors. The left panel shows a typical inhibition curve of a susceptible virus (solid line) with a typical

competitive inhibitor (e.g., a protease inhibitor). The IC50 value of the resistant virus (dotted line) is shifted to the right (arrow). The right panel shows an example of a noncompetitive inhibitor (e.g., a chemokine receptor 5 antagonist). The susceptible virus (solid line) shows a typical inhibition curve, but in this case, the resistant virus (dotted line) reaches a plateau. The maximum achievable percent inhibition is shifted downward (arrow), but the curve does not shift to the right; hence, the IC50 value remains unchanged. Source: Hirsch et al. Antiretroviral Drug Resistance Testing in Adult HIV-1 Infection: 2008 Recommendations of an International AIDS Society-USA Panel. Clin. Infect Dis. 2008;47:266–85.

7

GENETIC BARRIER

❖ The number of mutations required for reducing the drug antiviral activity. Genetic barrier may be classified as follows:

• Low: Loss of antiviral activity by the appearance or selection of a single mutation.

• High: Loss of antiviral activity after the appearance or selection of multiple mutations.

→ However, a high level of resistance (FC) is NOT NECESSARILY required for treatment failure.

Figure 4. Genetic barrier

8

1.2. RESISTANCE ASSAYS. ADVANTAGES AND LIMITATIONS

The development of genotypic and phenotypic tests has helped to guide the therapeutic management in HIV-infected patients. Multiple prospective and retrospective studies have demonstrated the clinical benefit of resistance testing. However, the interpretat ion of genotypic and phenotypic tests is complex and requires expertise.

COMMERCIALLY AVAILABLE GENOTYPIC TESTS

Region amplified

(aminoacids)

Minimum VL (c/mL) required

Comments

GenoSure PRIme (Monogram Biosciences)

PR (complete), RT (1-400), IN (complete)

500 CLIA-certified

ViroSeqTM HIV-1 Genotyping System (Abbot Molecular)

PR (1-99) RT (1-335)

500 FDA-cleared (16 and 96 capillary systems)CLIA-certified Hands-free gel electrophoresis & sequencing

At present, there are more than 100 mutations involved in the development of antiretroviral resistance. The construction of tables containing the mutations that confer resistance to the several drugs currently in use helps the interpretation of genotyping results.

9

GENOTYPIC INTERPRETATION RULES

There are different rules or algorithms available to infer the phenotypic susceptibility of the virus to the different antiretroviral drugs based on its genotype. Although each rule shows slightly distinct features, their overall performance is equivalent.

Algorithms and rules for inferring HIV susceptibility to ARVs from its genotype 15-22

Interpretation Systema Source Levelsb Access

HIVdbStanford, USA

ExpertsRule-based

S/PL/LL/IR/HR http://hivdb.stanford.edu/

Rega (HIV-1&HIV-2)Leuven, Belgium

ExpertsRule-based

S/I/R with drug-GSS weighting factors

http://rega.kuleuven.be/cev/avd/files/software/rega_algorithm/Rega_HIV1_Rules_v9.1.0.pdf

ResRIS (in Spanish);Spanish AIDS Network

ExpertsRule-based

S/I/R www.retic-ris.net

GenoSure MG (Monogram Bioscience)

Database (>100000 G/P)

S/R/RP www.monogrambio.com

HIV-GRADE 12/2008;Germany

ExpertsRule-based

S/I/LS/R www.hiv-grade.de

Geno2pheno, v. 3.4;Arevir, Germany

Database (>1000 G/P)

S/I/RQuantitative

www.geno2pheno.org/index.php

EuResist 1.0EuResist Network GEIE

Database (>65000 TCE plus additional features

Quantitative. Probability for short-term response with specific drug combinations

www.euresist.org

ViroSeqTM v2.8,Abbott/Celera

ExpertsScores -based

S/P/R no access or information on line availablewww.abbottmolecular.comwww.celera.com

10

Interpretation Systema Source Levelsb Access

ANRS(HIV-1&HIV-2);France

ExpertsRule-based

S/I/R www.hivfrenchresistance.org/

a S: susceptible; PL: possible low level resistance; LL: low level resistance; IR or I: intermediate resistance; HR: high level resistance; R: resistance; PM: possible multi-NRTI resistance, P: possible resistance, HM: high level of multi-NRTI resistance, H: high level of resistance; LS: low susceptibility; RP: resistance possibleb TCE: treatment change episodes.

11

COMMERCIALLY AVAILABLE PHENOTYPIC ASSAYS

Region amplified (aminoacids)

Minimum VL (c/mL) required

Comments

PhenoScriptTM (Viralliance)

PR (1-99) + p2/p7/p1/p6 gag cleavage sitesRT (1-400)ENV (gp160)

500 Separate amplifications and transfections for PR, RT and ENV.Includes gag cleavage sites relevant for PI susceptibility.Works with all group M subtypes.

PhenoSense™ HIV (Monogram Biosciences)

PR (1-99) + p7/p1/p6 gag cleavage sitesRT (1-305)

500 Continuous amplicon.Includes gag cleavage sites relevant for PI susceptibility.Works with all M group subtypes.Provides estimates of replication capacity (RC) relative to a wild-type control.

PhenoSenseTM

Entry

(Monogram Biosciences)

ENV (gp160) 500 Research Use Only (RUO).Assesses resistance to enfuvirtide (T-20) and CCR5 antagonists.The susceptibility cutoff for enfuvirtide is at the 99th percentile of the distribution of 220 enfuvirtide phenotypes in a naive reference population (baseline isolates from the TORO1 and TORO2 clinical trials).Susceptibility to CCR5 antagonists reported as percent maximal inhibition.

PhenoSenseTM Integrase(Monogram Biosciences)

RT (C-terminal)IN (1-288)

500 First commercial assay to assess phenotypic susceptibility to IN inhibitors.Provides estimates of replicationcapacity (RC) relative to a wild-type control.

All current phenotyping technologies include slight variations of essentially the same procedure: the generation of a recombinant virus by PCR-amplifying the HIV genomic region of interest from patient’s plasma, and inserting it into the backbone of a laboratory

12

clone of HIV from which this region of the genome has been removed, either by cloning or by in vitro recombination. Large quantities of replication-competent recombinant virus are thus produced and transfected into susceptible cells. Replication of the recombinant virus at different drug concentrations is monitored by expression of a reporter gene and is compared with replication of a reference HIV strain. The drug concentration that inhibits 50% of viral replication (i.e., the median inhibitory concentration [IC] 50) is calculated, and the ratio of the IC50 of test and reference viruses is reported as the fold-increase in IC50 (i.e., fold resistance).

In the PhenoSenseTM and PhenoscriptTM assays, data are analyzed by plotting percent inhibition of luciferase activity or beta-galactosidase production, respectively, versus log10 concentration of drug. The drug susceptibility curve is used to calculate the concentration of drug required to inhibit viral replication by 50% (IC50) or 90% (IC90). The IC50 of the patient’s virus is compared to the IC50 of a drug-sensitive reference virus control to calculate fold change in susceptibility.

Reduced drug susceptibility is indicated by a shift in the patient inhibition curve toward higher drug concentrations (to the right).

In the case of CCR5 antagonist susceptibility, reduced susceptibility is indicated by a reduction in the percent maximal inhibition (PMI) (“plateau” effect) relative to the wildtype control. This reduction in PMI is characteristic of the development of resistance to all non-competitive inhibitors including allosteric inhibitors (Figure 3)

Automated, recombinant phenotypic assays (PhenoSense) are commercially available with results available in 2–3 weeks. Viral phenotypic susceptibility can also be measured using “in-house” methods, which are usually restricted to few specialized laboratories.

13

Principles of commercial phenotypic assays

Assay Characteristics

PhenoSenseTM

(Monogram Biosciences)

The genes of interested are amplified from HIV sequence pools and incorporated into an indicator gene viral vector (IGVV) by conventional cloning methods using ApaI and PinAI restriction sites to construct a resistance test vector (RTV). Host cells are co-transfected with RTV DNA and a plasmid that expresses the envelope protein of amphotropic murine leukemia virus (MLV). Following transfection, virus particles are harvested and used to infect fresh target cells. The completion of a single cycle of viral replication results in the production of luciferase. Serial dilutions of PIs are added at the transfection step and RT inhibitors at the infection step. For the measurement of susceptibility to entru inhibitors (EI), indicator cells expressing CCR5 or CXCR4 co-receptors are treated with serial dilutions of drugs and infected with recombinant viruses harvested from the producer cellDrug susceptibility is measured by comparing the luciferase activity in the presence and absence of drugs. Susceptible viruses result in decreased levels of luciferase activity in the presence of drugs, whereas viruses with reduced susceptibility produce comparable levels to the wildtype control.

PhenoScriptTM

(Viralliance)The Phenoscript is based on a single cycle of in vitro replication and measures viral capacity of replication in the presence of drugs. Plasma is obtained from the patient’s blood sample viral RNA is extracted and three regions – gag-protease (GP), reverse-transcriptase (RT) and envelope (ENV) – are separately amplified to test PIs, RTIs and EIs respectively. Each PCR product is then separately co-transfected into producer cells along with the corresponding PHENOSCRIPT™ plasmid. For the PI and RTI assays, the single cycle of infection is ensured by the deletion of the envelope encoding region of the HIV plasmid. The envelope of the recombinant virus is provided by the G protein of the Vesicular Stomitis virus (VSV-G protein), for which the genetic information is carried on a separate plasmid. Serial dilutions of PIs are added at the transfection step and RT inhibitors at the infection step. For the measurement of susceptibility to EIs, indicator cells expressing CCR5 or CXCR4 co-receptors are treated with serial dilutions of drugs and infected with recombinant viruses harvested from the producer cell. The reporter cells used contain a LacZ gene under control of the HIV LTR. Once cells are infected, b-galacosidase is produced, the amount of which is detected using a CPRG based colorimetric assay and measured by Optical density.

14

OTHER NON COMMERCIAL PHENOTYPIC ASSAYS

Technology

Plaque reduction23

Recombinant viruses24

PBMC assays25} Laboratory specific

or home brew assay

❖ Interpretation of phenotypic tests26-28

The interpretation of phenotypic tests is limited because of the scarce data available about the correlation between the degree of in vitro resistance and the in vivo activity of a certain drug. For the interpretation of the phenotypic resistance/ susceptibility results, it is important to know the reproducibility of the technique used (technical cutoff), the range of IC50 required to inhibit wild-type virus replication (biological cutoff) and the clinical significance of the different decreases in drug susceptibility (clinical cutoff).

TECHNICAL BIOLOGICAL CLINICAL

• Based on assay reproducibility• Not drug-specific

• Based on upper limit of susceptibility range observed in panel of wild-type isolates• Drug specific

• Based on observations of virologic response (change in viral load) in treated patients• Drug-specific and specific to the treated population from which they are derived (unless widely cross-validated)

Monogram biosciences has defined clinical cut-offs for most of drugs, that correlate more closely with the antiretroviral activity of a certain drug. In some instances, such cut- offs reveal higher rates of resistance than previously described. In the case of resistance to drugs such as didanosine or stavudine, for example, low-level resistance measured in vitro translates in significant reductions in treatment responses in vivo. Such cut-offs reveal a higher incidence of resistance than previously described.

15

❖ Biological and clinical cut-off values of the different phenotypic tests

PHENOSENSETM

BIOLOGICAL CUT-OFF a CLINICAL CUT-OFFS (LOWER-UPPER) b, c

AZT 1.9

3TC 3.5

DDI 1.3-2.2

D4T 1.7

ABC 4.5-6.5

FTC 3.5

TDF 1.4-4

NVP 4.5

EFV 3

ETR 2.9-10

RPV 2.5

IDV

IDV/r 10

NFV 3.6

SQV 1.7

SQV/r 2.3-12

FPV 2

FPV/r 4-11

LPV/r 9-55

ATV 2.2

ATV/r 5.2

TPV/r 2-8

DRV/r 10-90

RAL 2.2

16

PHENOSENSETM

BIOLOGICAL CUT-OFF1 CLINICAL CUT-OFFS (LOWER-UPPER) 2,3

EVG 3.5

DTG 4-13

N/A: not available

a Biological cut-off values (BCO) are separating HIV-1 strains with a normal range of susceptibility from viral strains with reduced levels of susceptibility. The BCO for the Antivirogram® were set at the 97.5th percentile of the fold change values determined by in vitro phenotypic testing on wild-type viruses. For the Phenosense Assay, BCOs are defined as the fold change value below which reside 99% of tested wild-type isolates. With Geno2Pheno [resistance], cutoffs can be chosen before submitting the sequence.

b Lower clinical cutoff denotes the fold change which was the best discriminator of reduced clinical response using drug-specific clinical outcome data. Reduced response was defined by the clinical endpoint for the specific clinical cohort analyzed for each cutoff value. Upper clinical cutoff denotes the fold change above which a clinical response is unlikely (<.5 log reduction in HIV RNA) and which was determined using the same drug-specific clinical cohort data as for the lower clinical cutoff

c CCO1 (lower cut-off): predicted fold-change associated with a 20% loss of the wildtype virologic reponse due to viral resistance. CCO2 (upper cut-off): predicted fold-change associated with a 80% loss of he wildtype virologic reponse due to viral resistance

17

COMPARISONS BETWEEN GENOTYPIC AND PHENOTYPIC ASSAYS

Genotypic assays

Advantages

• Relatively simple to perform.• Widely available.• May detect mutations prior to apparent effects on phenotype.• Allow the detection of reversal mutations (e.g. 215 A/C/D/S) as “signatures”of past drug resistance.• Take into account the genetic barrier towards high level resistance.• Quick turnaround time and cost effective

Disadvantages

• Insensitive for minor variants.• Indirect measure of drug susceptibility• Interpretation requires prior knowledge of the genetic determinants of resistance.• Mutational interactions can not be predicted.

Phenotypic Assays

Advantages

• Provides quantitative resistance infor- mation, including assessment of hyper- susceptibility and partial susceptibility• Provides information on resistance to new drugs, for which genotypic corre- lates of resistance not established.• Provides information on resistance in non-type B infection, for which genoty- pic correlates of resistance not well established.

Disadvantages

• Require proper clinical cut-offs• Less sensitive than genotypes for wild type/mutant mixtures• Insensitive for minor variants• Time-consuming and more costly• The complexity of the assays limits its availability to a small number of laboratories

Discordances between phenotypic and genotypic methods are related to:

• Genotypic mixtures (they occur in up to 5% of the genes involved in antiretroviralresistance).• Incomplete understanding of genotypic correlates of resistance • Variability in the effects of specific mutations on phenotypic susceptibility tospecific drugs • Transitional mutations.• Antagonistic mutations.• Complex pattern of mutations.• Mutational interactions between and within genes.

18

❖ Other aspects to be considered for resistance testing

Sensitivity of resistance assays

The utility of resistance testing is partially limited by the sensitivity of amplification of the genetic material from plasma samples containing a low number of viral RNA copies and by the difficulty of detecting minor mutant virus that may be present in the virus populations. Point mutation assays have shown a better sensitivity for detecting minor populations of resistant viruses (for some codons up to 10%)29 compared to automatic sequencing (>20-30%) or phenotypic assays (>20-50%). The ability to detect minor populations in a mixture of virus depends on both the specific mutation and the laboratory performing the test.

HIV subtypes

During the expansion of HIV among humans, HIV group M has diversified in several subtypes which differ from one to another between 10% and 30% of the genomes.30 In Europe and North America, the majority of isolates belong to subtype B, however, the majority of isolates from the rest of the world belong to non B subtypes. The information about non-B HIV drug susceptibility is scarce.

The genetic variability between subtypes can affect the results of genotypic tests given that the primers used in RT-PCR, PCR and sequencing may hybridize less efficiently in non-B subtypes.

The ability to detect HIV-1 non-B subtypes is still not well known in both genotypic and phenotypic assays. However, some studies have reported amplification and successful resistance analysis for all group M (A-H) subtypes.31 The great diversity appearing in the sequence of areas of interest and the possible coinfection of several subtypes requires a continuous investigation in the development of amplification and sequencing techniques.

19

Resistance Websites

http://hivinsite.ucsf.edu/http://www.iasusa.org/resistance_mutationshttp://home.ncifcrf.gov/hivdrphttp://www.hiv.lanl.gov/content/sequence/RESDB/http://hivdb.stanford.edu/http://www.hivfrenchresistance.org/http://regaweb.med.kuleuven.be/software/rega_algorithm/http://www.who.int/hiv/topics/drugresistance/en/index.htmlhttp://www.youtube.com/watch?v=TvNOmwRh0I0http://www.thebody.com/index/treat/resistance.htmlhttp://www.thebodypro.com/index.html?ic=3001http://www.clinicaloptions.com/HIV/Topics/Resistance.aspxhttp://www.monogramvirology.com/hiv-tests/resistance-testing/http://www.janssendiagnostics.com/hiv-resistance/vircotype-hiv-1

References

1. Hirsch MS, et al. Clin Infect Dis 2008; 47: 266-85.2. Domingo E, et al. Annu Rev Microbiol 1997; 51: 151-78.3. Coffin JM. Science 1995; 267: 483-489.4. Quiñones-Mateu ME, et al. Drug Resist Updat 2002; 5: 224-233.5. Molla A, et al. Nat Med 1996; 2: 760-666. Kuritzkes DR, et al. AIDS 1996; 10: 975-817. Pillay D, et al. BMJ 1998; 317: 660-628. Bangsberg DR et al. J Antimicrob Chemother 2004; 53: 696-999. Paredes R, et al. J Infect Dis 2010; 201: 662-72.10. Simen BB, et al. J Infect Dis 2009; 199: 693-70111. Johnson J, et al. PloS Med 2008; 5: e15812. Pou C, et al. Clin Infect Dis 2014; 59: 578-88.13. Sarafinos SG, et al. The EMBO Journal 2002; 21: 6614-24.14. Boyer PL ,et al. J. Virol 2001; 75: 4832-42.15. Shafer RW, et al. Nat Med 2000; 6: 1290-92.16. Rhee SY, et al. Nucl Acids Res 2003; 31: 298-303.17. Van Laethem K, et al. Antivir Ther 2002; 7: 123-918. Leitner T, et al. HIV Sequence Compendium 2003. Los Alamos, NM: Los Alamos National

Laboratory; 2003.19. Johnson VA, et al. Top Antivir Med 2014; 22: 642-50.20. Mazzotta F, et al. J Acquir Immune Defic Syndr 2003; 32: 268-80.21. Race E, et al. AIDS 1999; 13: 2061-68.22. Torti C, et al. Clin Infect Dis 2005; 40: 1828-36.

20

23. Larder BA, et al. Antimicrob Agents Chemother 1990; 34: 436-41.24. Kellam P and Larder BA. Antimicrob Agents Chemother 1994; 38: 23-30.25. Japour AJ, et al. Antimicrob Agents Chemother 1993; 37: 1095-101.26. Harrigan PR, et al. AIDS 2001; 15: 1671-727. Hellmann N, et al. Antivir Ther 1999; 4(Suppl. 1): 34-3528. Vingerhoets J, et al. Antivir Ther 1999; 4(Suppl. 1): 5329. Nissley DV, et al. J Clin Microbiol 2005; 43: 5696-704.30. Korber B, et al. Science 2000; 288: 1789-96.31. Schuurman R, et al. J Clin Microb 1999; 37: 2291-96.