No part of this publication may be reproduced or photocopying. without the prior written permission. of the publisher

Transcription

1 AIDS Reviews 2006;8: Clinical Implications of HIV Drug Resistance to Nucleoside and Nucleotide Reverse Transcriptase Inhibitors Anna Maria Geretti Royal Free Hospital and Royal Free & University College Medical School, London, UK Abstract 210 This review focuses on issues pertinent to the epidemiology and clinical interpretation of resistance to nucleos(t)ide reverse transcriptase inhibitors. Nucleoside reverse transcriptase inhibitor resistance mutations, and especially thymidine analog mutations, remain the most common form of resistance detected in drug-naive patients. At the same time, improved treatment strategies, changes in prescribing policies, and prompt management of treatment failure are changing the prevalence and patterns of nucleoside reverse transcriptase inhibitor resistance in treatment-experienced patients. The clinical interpretation of nucleoside reverse transcriptase inhibitor resistance is being increasingly refined, helped by improved understanding of resistance pathways, development of sophisticated methods of analysis of genotypic resistance patterns, and introduction of clinically relevant cutoffs. Correlation of genotypic and phenotypic resistance data to clinical outcomes is essential to allow appropriate interpretation. In some cases, phenotypic data, either obtained directly by phenotypic tests or extrapolated from genotypic results, provide the most immediate predictors of virologic response. In other cases, genotypic analyses identify mutations that impact on responses without showing a marked effect on the phenotype, by either acting as sentinel markers for the presence of resistance undetectable by standard methods, or by lowering the genetic barrier to the evolution of resistance. The potential benefits of nucleoside reverse transcriptase inhibitor resistance, through hypersusceptibility and fitness effects, are also increasingly understood and exploited in clinical practice. Although progress has been significant, there remain many challenges. It is often questioned whether genotypic scores and clinical cutoffs obtained by various methods and from frequently small datasets can be reliably extrapolated to the general population of treated patients. At the same time, there is a need to define the role of reverse transcriptase mutations that are identified by statistical analyses as being associated with nucleoside reverse transcriptase inhibitor exposure, but have unknown effects on virus phenotype and clinical outcome. Novel mechanisms have also been proposed to play a role in nucleoside reverse transcriptase inhibitor resistance, including changes in RNaseH that are not targeted by routine testing at present. One additional, currently unresolved issue is the clinical relevance of minority resistant species and the feasibility of introducing ultra-sensitive resistance tests in routine diagnostic settings. The most appropriate viral-load cutoff for performing resistance tests and the reliability of results obtained at low copy numbers are similarly controversial. In spite of these limitations, resistance testing with appropriate interpretation provides an important guide to successful treatment outcomes and necessary support to the introduction of new treatment strategies. (AIDS Reviews 2006;8:210-20) Corresponding author: Anna Maria Geretti, a.geretti@medsch.ucl.ac.uk Key words Genotype. Phenotype. Fitness. Hypersusceptibility. Transmission. Correspondence to: Anna Maria Geretti Dept of Virology Royal Free Hospital Pond Street London NW3 2QG, Permanyer UK Publications a.geretti@medsch.ucl.ac.uk

2 Anna Maria Geretti: Interpreting NRTI Resistance Table 1. NRTI resistance mutations Mutation group Thymidine analog Codons M41L D67N/G K70R L210W T215Y/F K219Q/E/N M184V/I E44D/A + V118I loop K65R T69D/N L74V/I V75T/M/A Y115F Multi-nucleoside resistance Accessory, non polymorphic Accessory, polymorphic T69 insertion Q151M complex (Q151M A62V V75I F77L F116Y) K43E/N/Q E203D/K H208Y D218E K20R V35M T39A K122E G196E Introduction The HIV-1 reverse transcriptase (RT) is a heterodimer composed of two related subunits, p51 and p66, encoded by the pol gene. The p51 subunit spans the first 440 amino acids of p66, but has no enzymatic activity and acts as a structural scaffold for the enzymatically active p66. The p66 subunit consists of 560 amino acids and folds into two subdomains, polymerase (440 amino acids) and RNaseH (120 amino acids) linked by the connection subdomain. The shape of the polymerase domain has been likened to a human hand, with three main subdomains referred to as fingers, palm, and thumb. The polymerase active site is in the palm subdomain (amino acid positions 110, 185, and 186). Most nucleoside reverse transcriptase inhibitor (NRTI)-resistance mutations are in the fingers and palm subdomains (amino acid positions 1-235) (Table 1). Because NRTI lack the 3 OH found on normal deoxynucleoside triphosphates (dntp), they act as chain terminators when incorporated into the growing viral DNA chain. Resistance to NRTI can occur by two mechanisms: a) enhanced ability to discriminate between the NRTI and the natural substrate prior to incorporation into the viral DNA. This is the mechanism of resistance to lamivudine (3TC) mediated by the M184V mutation; b) enhanced excision of the incorporated NRTI. This mechanism, known as primer unblocking or pyrophosphorolysis, mediates resistance to zidovudine (ZDV) associated with the thymidine analog mutations (TAM). In addition to the major resistance mutations, several accessory mutations have been identified in RT by statistical analyses of large genotypic databases of NRTI-experienced patients. In subtype-b, 12 mutations have been found to be positively associated with NRTI treatment (Table 1), and many cluster with major NRTI-resistance mutations 1. These accessory mutations, some of which also occur as natural polymorphism, appear to play a role in drug resistance, but the mechanisms of action and whether they augment resistance or compensate for reduced fitness in the presence of major resistance mutations remain currently uncertain. Mutations have also been detected in RNaseH following in vitro passage with ZDV, and the changes have been shown to contribute resistance to ZDV, especially in the presence of TAM 2,3. In addition, RNaseH mutations have been found to occur more commonly in NRTI-experienced patients than in NRTI-naive patients 4. The RNaseH is not targeted by current genotypic assays, and further data on the clinical significance of the changes observed are awaited to inform clinical practice. As is the case with accessory RT mutations, if changes in RNaseH only occur in the context of major NRTI mutations, it remains to be proven that including these mutations would significantly alter the clinical interpretation of resistance results. Epidemiology of NRTI resistance NRTI resistance remains the most common form of resistance detected in treatment-experienced cohorts, although a downward trend has been observed in recent years. In the Permanyer Publications UK, the prevalence of NRTI-resistance 2010 mutations among treated persons declined from over 60% in the late 1990s to 211

3 AIDS Reviews 2006; RT PR Prevalence (%) approximately 50% in Improved treatment strategies, low tolerance of detectable viral load, and prompt management of treatment failure are reducing the accumulation of NRTI resistance in treatment-experienced persons. Thus, in European cohorts, the prevalence of persons with 3 TAM declined from 45 to 30% between 1999 and , and lower rates of resistance to ZDV, 3TC, stavudine (d4t), abacavir (ABC), and didanosine (ddi) were observed in the period when compared to Traditionally, the most common NRTI-resistance mutations detected in treatment-experienced patients have been M184V/I and TAM, reflecting the extensive use of 3TC, ZDV, and d4t, but also the use of mono and dual NRTI therapy in the pre-haart era and the suboptimal use of therapy in the early HAART era. In 2003, an analysis of the Virologic Database of over 16,000 samples showed that 44% had M184V 8. Among TAM, the prevalence was 31% for T215Y/F, 25% for M41L, 20% for D67N and K219Q/E/N, 16% for K70R, and 15% for L210W. The multicenter European Capture study confirmed that M184V and the TAM M41L, D67N, and T215Y/F remain the most frequent NRTI mutations in treated patients surveyed in Similar findings have been reported from Spanish 6 and Italian 9 cohorts. A recent study reported resistance data on 100 consecutive patients seen between 2004 and 2006 in 0 M184V K103N D67N T215Y/F M41L K70R K219Q/E L210W Y181C L90M I84V M46I I54V Figure 1. Prevalence of major resistance mutations among 100 persons with viral load > 50 and 5000 copies/ml while on NRTI-based HAART. RT: reverse transcriptase; PR: protease (adapted from reference 10). London and showing a viral load > 50 copies/ml but < or equal to 5000 copies/ml (median 1260, range copies/ml) while on HAART. Patients were receiving one or more NRTI in combination with either nonnucleoside RT inhibitors (NNRTI, n = 22) or ritonavir-boosted protease inhibitors (PI/r, n = 78). In testing performed for routine clinical care, the overall prevalence of resistance was 45% for the NRTI, 36% for the NNRTI, and 7% for the PI. The most common NRTI-resistance mutations were M184V and TAM (Fig. 1) 10. NRTI-resistance mutations remain also the most prevalent form of transmitted drug resistance detected in treatmentnaive persons 11. Temporal trends vary across different cohorts. Among persons with recently acquired infection in New York City, the prevalence of NRTI resistance increased from 11.8% in to 16.1% in Other studies show stable or declining prevalence, which may partly reflect changes in the survey methodology. Nonetheless, a true decline in transmission of NRTI resistance may also be occurring where the proportion of potential transmitters of resistance is decreasing over time 5. The most commonly reported mutations in persons with transmitted resistance are TAM, and the mutational patterns described frequently involve a single mutation, although multiple TAM are also observed (Table 2) 11. In studies of drug-naive persons in the UK 13, Germany 14, and Brazil 15, 1.3-3% showed a single TAM, % showed two, and % showed three. Several mutations detected in treatment-experienced persons are less common in recently infected patients, suggesting a hierarchy in transmissibility. In RT, M184V/I and T215Y/F may have a relative transmission disadvantage when compared to K219QE 9. In contrast, revertants of the T215Y/F mutation (T215rev, e.g. T215S/C/D/E/N/L), one of the most commonly reported forms of transmitted drug resistance 11, show similar prevalence in treatment-experienced persons and recently infected patients, suggesting efficient transmission 9. The T215rev may be responsible for the primary infection, although reversion can also occur after transmission of T215Y/F, due to the improved fitness of T215rev relative to T215Y/F and growth advantage over the resistant mutants. Over a median 15 months (range 10-23) of follow-up, reversion (by standard genotypic testing) to either other codons or wild-type is commonly seen with T215Y/F, K70R, and M184V, less commonly with D67N, T215S, and K219N, and uncommonly for M41L, T69D/N, L210W, T215L/C/E, and K219Q 16. The findings contribute to explain the mutational patterns observed in chronically infected persons and should be kept in mind when assessing the potential clinical significance of resistance mutants detected in drug-naive patients with established infection. Persistence of more extensive resistance within proviral DNA and minority species in plasma has been documented, with potential long-lasting effects on treatment outcomes 17.

4 Anna Maria Geretti: Interpreting NRTI Resistance Table 2. Resistance mutations detected in persons newly diagnosed with HIV-1 infection in a large clinical centre in London in Patient gender Risk group Ethnicity (CoB) Subtype Resistance mutations RT Protease M1 MSM W (UK) B None V82L M2 MSM W (UK) B K103N None M3 MSM MR (UK) B K103N None M4 MSM W (UK) B None L90M M5 MSM W (UK) B M41L T215S None M6 HTS MR (UK) B M41L T215S None M7 MSM W (UK) B K103N None M8 MSM W (UK) B M41L/M None M9 MSM W (UK) B T215D None M10 MSM W (UK) B M41L/M T215S/T None M11 MSM BC (UK) B D67N T69N T215V K219Q L33F G48V I54T G73S V82A L90M M12 MSM W (UK) B M41L L210F T215D None M13 MSM W (UK) B K219Q/R None M14 HTS A (Pakistan) Cpx f None M46L F1 HTS BA (South Africa) C K103N None M15 HTS BA (UK) B D67N T69D/N 219Q None M16 MSM W (UK) B D67N T69N K219Q None M: male; F: female; MSM: men who have sex with men; HTS: heterosexual; CoB: country of birth; W: white; MR.: mixed race; BC: black-caribbean; A: Asian; BA: black-african. Adapted from reference 13. NRTI resistance pathways The NRTI resistance pathways observed during first-line therapy with common NRTI backbones include M184V/I, with or without TAM, K65R, and L74V (Table 3). The overall prevalence of treatment-emergent NRTI resistance has been low in recent clinical trials For instance, in the Gilead 934 study, over 96 weeks only one of 29 patients experiencing failure of first-line therapy with Combivir /efavirenz (EFV) developed treatment-emergent TAM, and only 31% had evidence of 3TC resistance 18. It should be assumed, however, that analyses within clinical trials provide a minimal estimate of the risk of resistance, as close monitoring ensures that virologic rebound is detected and managed early. In addition, resistance testing is generally performed on the intent-to-treat population and the analysis includes treatment failure due to reasons other than virologic failure. This helps explain the substantial proportion of patients who have no detectable resistance to 3TC or EFV, drugs with a low genetic barrier to resistance. In routine clinical practice, the level of NRTI resistance detectable at the time of treatment failure can be more substantial than that seen in clinical trials. In a Mexican study of patients starting first-line therapy in , 3 TAM were detected in 32, 32, and 19% of patients failing NRTI therapy in combination with NNRTI, unboosted PI, or PI/r, respectively 24. The introduction of PI/r in therapy has led to an apparent reduction in the prevalence of NRTI resistance at the time of treatment failure. In trials of firstline therapy with NRTI backbones including 3TC, the prevalence of M184V/I among genotyped patients ranged between 13 and 50% in those receiving PI/r (lopinavir, fosamprenavir, atazanavir), and 55 to 88% among those receiving an unboosted PI (nelfinavir, fosamprenavir, atazanavir) 23,

5 AIDS Reviews 2006;8 Table 3. NRTI resistance patterns observed in recent clinical trials of drug-naive persons starting first-line therapy Study (week) NRTI Third drug Genotypes NRTI mutations GS934* (96) 18 Combivir EFV 29 M184V TAM Prevalence (%) Third-drug resistance (%) 62 GS934* (96) 18 TDF FTC EFV 14 M184V GS903 (48) 19 TDF 3TC EFV 29 M184V K65R ABT418 (96) 20 TDF FTC LPV/r 23 M184V 17 0 CNA30024 (48) 21 ABC 3TC EFV 13 M184V L74V CNA30021 (48) 22 ABC 3TC EFV 38 M184V L74V K65R CNA30021 (48) 22 ABC 3TC EFV 31 M184V L74V K65R SOLO (48) 23 ABC 3TC FPV/r 32 M184V 12 0 EFV: efavirenz; TAM: thymidine analog mutations; TDF: tenofovir; FTC: emtricitabine; 3TC: lamivudine; ABC: abacavir; LPV: lopinavir; r: ritonavir; FPV: fosamprenavir. *Patients with baseline NNRTI resistance (n = 11) excluded. Patients with baseline NNRTI resistance (n = 3) included. Patients with baseline resistance (n = 7) included. Patients with baseline resistance (n = 7) excluded. 214 The presence of transmitted resistance mutations is an important determinant of the evolution of resistance during treatment failure. Patients with baseline resistance are in some studies excluded from the analyses (Table 3). Within the CNA30021 study 22, the proportion of persons with the ABC mutation L74V was substantially higher if persons with baseline NRTI or NNRTI resistance were included in the analysis. These observations support the importance of baseline resistance testing prior to commencing HAART as a guide for tailoring the regimen, reducing the risk of virologic failure, and preventing the rapid accumulation of resistance and crossresistance should treatment failure occur. The decline in the prevalence of TAM among treatmentexperienced patients has been accompanied by an increase in the prevalence of the previously rare mutation K65R, which is selected by tenofovir (TDF) as well as ABC and ddi. The prevalence of the mutation increased sharply after TDF was licensed in 2001, but has subsequently stabilized 28. These trends reflect the initial use of TDF in highly treatment-experienced patients and the subsequent use within poorly active triple-nrti regimens such as TDF/3TC/ABC. Recent clinical trials using TDF and emtricitabine (FTC) in combination with either EFV or ritonavir-boosted lopinavir have shown a low risk for the emergence of K65R 18. Data from routine clinical practice are awaited. Meanwhile, data have been presented from in vitro studies and animal models suggesting interesting interactions between TDF and FTC resistance pathways. Whereas in vitro passage of HIV-1 with ABC/3TC selects M184V, TDF/FTC has been shown to select K65R 29. In addition, during in vitro passage of SIVmac251, M184V is maintained by 3TC but deselected in favor of K65R by TDF/3TC 30. Similarly, macaques who have developed M184V maintain the mutation if treated with FTC, but deselect the mutation in favor of K65R if treated with TDF or TDF/FTC 30. These interactions may result from multiple mechanisms, including M184V-mediated restored susceptibility to TDF in the presence of K65R, increased intracellular levels of FTC triphosphate caused by TDF, and the fitness effects of both M184V and K65R. From a clinical perspective, it is unclear whether simultaneous pressure with TDF/3TC or TDF/FTC would be as effective as 3TC or FTC alone in maintaining the M184V mutation. Effects of common NRTI-resistance mutations Thymidine analog mutations The TAM are selected by ZDV and d4t and accumulate in Permanyer Publications step-wise fashion. Their emergence 2010 is delayed by the concomitant use of 3TC and, conversely, is rapid with ZDV/ddI

6 Anna Maria Geretti: Interpreting NRTI Resistance and ddi/d4t. The TAM have resistance effects for all available NRTI, although their impact is greater for ZDV, d4t, ddi, ABC, and TDF than for 3TC. Overall, the greater the number of TAM, the greater the degree of NRTI resistance and cross-resistance. The presence of 4 TAM typically causes > 100-fold decreased susceptibility to ZDV, five- to sevenfold decreased susceptibility to ABC, and two- to fivefold decreased susceptibility to d4t, ddi and TDF. Not all TAM have the same impact on NRTI resistance however. Pathway-1 TAM, including M41L, L210Y and T215Y, have greater resistance effects than pathway-2 TAM, including K70R, T215F and L219W. The D67N mutation has significant resistance effects and may appear in combination with either mutation pathway. In the Jaguar study, which investigated the effects of intensifying a failing regimen with either ddi or placebo, the strongest association with virologic response to ddi was observed with a set of six mutations including M41L, T69D, L74V, L210W, T215F/Y, and K219E/Q 31. Patients with no mutations experienced approximately 1 log 10 copies/ml reduction in viral load, whereas responses were reduced in those with three or more mutations. Similarly, in the Gilead trials of TDF intensification, responses were impaired (albeit not completely abolished) in patients with 3 TAM, including those at positions 41, 67, 210, or Recent studies have confirmed that D67N/H plays an important role in TDF resistance 33. Mutation M184V/I The M184V mutation is selected by 3TC, FTC, and ABC and confers high-level resistance to 3TC and FTC. The M184I results from a change from A-G (methionine) to A-A (isoleucine) and usually develops before M184V (A-G to G-G, valine) in patients receiving 3TC because RT is more prone to G-A than to A-G substitutions. Although M184I also causes highlevel resistance to 3TC, the enzymatic efficiency of M184I is less than that of M184V, and as a result of higher relative fitness, M184V nearly always replaces M184I. Mutation M184V also confers low-level phenotypic resistance to ABC and ddi, but in isolation does not compromise virologic responses to the drugs. However, M184V in combination with 3 TAM or with mutations at positions 65, 74, or 115 causes significant resistance to ABC. In contrast, adding ddi to a treatment regimen in the presence of M184V and varying numbers of TAM leads to a median viral-load reduction of 0.6 log 10 copies/ml 31. The mutation has been extensively studied and several important effects have been demonstrated 34,35, including a reduction in viral replication rate and progressive cdna synthesis, influence on RNaseH function 36, and increase of RT fidelity. The M184V also delays the emergence of TAM when combined with ZDV, and causes increased susceptibility to ZDV, d4t, and TDF through inhibition of primer unblocking. These properties are generally described in terms of reduced viral fitness and re-sensitization effects. Viruses carrying the combination of mutations are less likely to become fully sensitized to ZDV when M184V is present. Those carrying the cluster of mutations are more likely to become fully sensitive to ZDV when M184V is present. Continuing 3TC in patients with M184V The first evidence of a beneficial effect of continuing 3TC in the presence of high-level resistance to the drug was provided by the NUCA3001 study published in Patients receiving 3TC monotherapy showed initially a > 1 log 10 copies/ml drop in viral load, followed by virologic rebound after week four as resistance to 3TC emerged. Despite the rebound, however, during 52 weeks of follow-up the viral load remained 0.5 log 10 copies/ml lower than at baseline. The Caesar study demonstrated the clinical benefit of adding 3TC in patients receiving ZDV monotherapy. Over 52 weeks, the addition of 3TC significantly slowed the progression of HIV disease and improved survival 38. Subsequent studies have yielded a large body of evidence addressing the mechanisms that may explain partial virus suppression despite high-level phenotypic resistance, including residual antiviral activity, impaired viral fitness, and enhanced HIV-specific immunity 34. More recently, the E-184 study confirmed the benefit of continuing 3TC in patients with M184V 39. Patients on 3TC-containing therapy with M184V, viral load > 1000 copies/ml, and CD4+ > 500 cells/mm 3 were randomized to either continue 3TC or undergo a complete treatment interruption. Over 48 weeks there was a lower risk of clinical or immunologic failure (defined as CD4+ < 350 cells/mm 3 ) in patients on 3TC compared to the treatment-interruption group. In addition, the viral load was ± 0.5 log 10 copies/ml lower in the 3TC group than in the treatment-interruption group. Patients on 3TC were shown to maintain M184V as well as other mutations, possibly linked to M184V on the same viral genome, whereas a greater degree of reversion of resistance was seen in the treatmentinterruption group. The greater level of resistance correlated with an overall lower replicative capacity measured in the 3TC group, which may have contributed to the beneficial virologic and immunologic effects. The Plato study has similarly demonstrated the benefit of continuing therapy in patients with resistance by showing that at a given viral-load level, treated patients with multidrug resistance had smaller CD4+ declines than untreated patients 40. Partial-treatment-interruption studies, where only one class of drugs is discontinued, have demonstrated the benefit of continuing NRTI therapy in the presence of NRTI resistance 41. A viral load rebound of log 10 copies/ml has been de- 215

7 AIDS Reviews 2006;8 216 scribed after NRTI interruption in patients with NRTI resistance continuing other therapy, a change far greater than that observed after discontinuation of PI, or enfuvirtide 42. Nonetheless, continuing NRTI therapy during ongoing virus replication may eventually lead to increasing levels of resistance and the emergence of compensatory changes that restore virus fitness. Several of the accessory mutations described for the NRTI appear late during the evolution of resistance, following the emergence of major resistance mutations, suggesting a possible compensatory role 1,43. For example, the H208Y mutation in RT is rare in drug-naive persons and occurs in just 4.5% of treatment-experienced persons with subtype-b. The mutation shows a strong association with NRTI experience, including exposure to 3TC, and is most prevalent in genotypes harboring multiple NRTI mutations including M184V, M41L, D67N, L210W, and T215Y 43. In agreement with these observations, treatment failure in the presence of the accessory mutations K43E, K122E, and H208Y has been found to be significantly associated with higher viremia and lower CD4+ count 1. NRTI Incorporation Stability Net susceptibility ZDV q pp Increased d4t q p Intermediate ddi qq Reduced ABC q p Intermediate TDF qq p Reduced 3TC q Reduced FTC q Reduced Figure 2. Effects of K65R on binding/incorporation of NRTI and the stability/retention of incorporated NRTI. ZDV: zidovudine; d4t: stavudine; ddi: didanosine; ABC: abacavir: TDF: tenofovir: 3TC: lamivudine; FTC: emtricitabine. ment failure while on first-line therapy with d4t/3tc/efv 19. The mutation has resistance effects for TDF, ABC, ddi, 3TC, FTC, and at least by enzymatic assay, d4t 45. The emergence of K65R is antagonized by concomitant therapy with ZDV and the presence of TAM 46. Mutation K65R is found uncommonly Resistance to FTC with TAM, and the combination of K65R+T215Y in particular is rarely found on the same viral genome, unless they coexist Emtricitabine is structurally similar to 3TC, and M184V is with Q151M. There are important reciprocal interactions between these mutations: K65R reduces TAM-mediated primer the most common mutation selected by the drug both in vitro and in vivo. Phenotypic data from the Monogram Bioscience unblocking, whereas TAM reduce K65R-mediated substrate database recently showed that the phenotypic resistance effects, measured as mean fold-changes in IC 50 discrimination 47. Like M184V, K65R has been shown to reduce replication capacity. The clinical significance of this, were not significantly different for 3TC and FTC in the presence of K65R observation remains uncertain at present. (8.3 vs. 7.0), L74V/I (1.7 vs. 1.5), and Q151M (2.5 vs. 3.1) 44. The K65R has complex, contrasting effects on NRTI incorporation and excision, and the net impact on resistance de- In contrast, phenotypic resistance was significantly higher (p < ) in samples with 2-3 TAM including those at pends on the strength of the individual effect 47,48 (Fig. 2). positions 41, 210, and 215 (2.1 vs. 2.5), 2-3 TAM including Overall, the strong effect of K65R in reducing the excision of those at positions 67, 70, and 219 (3.2 vs. 4.1), any 3-4 TAM incorporated ZDV results in increased susceptibility to the (3.6 vs. 4.9), and any 5-6 TAM with (6.7 vs. 10.3) or without drug. The effect has been invoked to explain the virologic (4.5 vs. 7.1) E44D/V118I. The level of phenotypic resistance success reported in a small group of patients who had developed K65R on various combination regimens, and whose was also significantly higher in the presence of the T69 insertion, with TAM allowed (10.2 vs. 21.7). The TAM cause lowlevel 3TC resistance, but do not appear to compromise the regimen 49. The patients had developed K65R, L74V, Y115F, therapy was intensified by the addition of ZDV to the failing drug activity. The clinical significance of the higher levels of and M184V while on ABC/3TC/ddI or ABC/ddI/TDF, and K65R, phenotypic FTC resistance is No unclear, part as fold-changes of this can publication Y181C, and G190S may while on be TDF/3TC/NVP. All achieved and only be interpreted in the context of adequately determined, maintained an undetectable viral load after addition of ZDV. drug-specific, clinical cutoffs. Although further studies are needed to extend these observations, the antagonistic and hypersusceptibility effects may Mutation K65R be exploited to enhance NRTI activity in drug-experienced patients with resistance, and to antagonize the evolution of The K65R mutation is located in the loop formed by codons NRTI resistance. The antagonism should not be regarded as (between the ß2 and ß3 strands in the fingers of the region publisher absolute, however. In the DART study, among 300 patients of the RT), which makes important contacts with the incoming receiving first-line therapy with TDF/3TC/ZDV for 48 weeks, dntp. The mutation Permanyer is selected by TDF, ddi, and ABC, Publications and resistance analysis in 20 patients 2010 with viral load > 1000 copies/ml showed four with M184V alone, with M184V has was also detected in 4% of persons experiencing treat- plus

8 Anna Maria Geretti: Interpreting NRTI Resistance Table 4. Median fold change in IC 50 observed with clinical isolated containing K65R or L74V alone or in combination with M184V (PhenoSense assay) Drug M184V n = 1720 L74V n = 22 K65R n = 82 L74V M184V n = 74 K65R M184V n = 54 ABC TDF TC > > 300 > 300 ZDV d4t ddi ABC: abacavir; TDF: tenofovir; 3TC: lamivudine; ZDV: zidovudine; d4t: stavudine; ddi: didanosine. Adapted from reference 56. TAM (mean 2.4, range 1-4), one with TAM alone, and two with the VircoType (Virco, Belgium) proposed clinical cutoffs for K65R in combination with T215Y or Y115F. One other patient TDF include a lower cutoff of 0.9 (90% CI: ) and an who switched from ZDV to d4t also developed K65R 50. upper cutoff of 2.1 (90% CI: ), with values in-between indicating intermediate levels of resistance and drug activity 54. Mutation L74V The PhenoSenseGT (Monogram Biosciences, USA) uses a lower cutoff of 1.4 and an upper cutoff of Thus, the two The L74V mutation is selected by ABC and ddi. The mutation assays are not interchangeable in terms of predicting levels of has also been shown to occur in 4% of patients receiving resistance to the drug. Along the same lines, the ABC cutoffs d4t/3tc/efv. The mutation has significant resistance effects for are 0.8 and 1.9 in the VircoType, and 4.5 and 6.5 in the PhenoSenseGT, whereas those for ddi are closer, at 0.9 and 2.6 ABC and ddi. On a background of TAM, it increases resistance to ABC, ddi, 3TC, and d4t, but reduces resistance to ZDV and in the VircoType and 1.3 and 2.2 in the PhenoSenseGT. TDF. The L74V inhibits TAM-mediated primer unblocking, showing the greatest effect in the presence of mutations at position significant improvement relative to the use of technical or Clinical cutoffs provide a useful guide and represent a 67, 70, and 219, but smaller effects with mutations at positions biological cutoffs. Nonetheless, the activity of any drug in a 41, 210, and 215, or in the presence of six TAM 51,52. There is given regimen is influenced by the presence of other active a certain antagonism between L74V and K65R, and the mutations are found uncommonly on the same viral genome, prob- extrapolate the activity of a single drug. Clinical cutoffs also drugs, and complex statistical adjustments are required to ably reflecting a strong effect on viral replicative capacity. necessarily reflect the clinical isolates of the population from Recently, L74I has been identified as an alternative mutation which they have been derived and their particular treatment pathway to L74V, associated with ABC and EFV use 53. history. Furthermore, in the case of most NRTI, resistance should be seen as a continuum and the introduction of cutoffs Clinical interpretation of common in the continuum is by necessity artificial. It is important to NRTI-resistance patterns appreciate that phenotypic results obtained with different No part of this publication clinical isolates are may spread be around a median, with some values falling above and others below the proposed clinical Phenotypic data cutoff. Clinical isolates with phenotypic values just above or The interpretation of NRTI-resistance patterns may make use below a given cutoff are not necessarily fully resistant or fully of phenotypic data, either obtained by phenotypic testing susceptible. extrapolated from genotypic results. The interpretation of phenotypic results is not always straightforward. Relevant cutoffs require careful interpretation. Using the PhenoSenseGT as- Despite the availability of clinical cutoffs, phenotypic results are required that translate into a measure of clinical of the activity. publisher say, for example, the median level of phenotypic resistance Clinical cutoffs are becoming increasingly available, frequently varies according to whether K65R and L74V are present alone presented as a lower Permanyer cutoff at which responses start to decline, Publications or with M184V (Table 4) 56. Relative 2010 to K65R alone, the presence of M184V increases the median level of phenotypic and a higher cutoff at which responses are virtually lost. Thus, re- 217

9 AIDS Reviews 2006;8 218 sistance to ABC and ddi, but reduces phenotypic resistance to TDF, to a level just below the proposed clinical cutoff of 1.4. There is good evidence from the Gilead TDF-intensification studies that the presence of M184V at baseline has a small (estimated viral load response 0.13 log 10 copies/ml,) but significant (p = ) beneficial effect on virologic responses to TDF 32. Conversely, the six patients who had K65R at baseline showed virtually no response to the addition of TDF. The numbers were too small to asses the effect of M184V in combination with K65R. Thus, whether the resensitization effects seen in vitro with M184V in the presence of K65R have clinical relevance is currently unknown. As a result, the reported phenotypic data for the combination of K65R/M184V must be interpreted with caution. Along the same line, the M184V mutation is consistently seen to increase the level of phenotypic resistance to ddi in vitro. Yet, data from the Jaguar ddi intensification trial showed that in highly treatment-experienced patients, the median change in viral load (log 10 copies/ml) after four weeks was 0.15 in eight patients without M184V (vs in six patients on placebo) and 0.60 in 93 patients with M184V (vs in 46 patients on placebo) 31. Thus, these data seem to indicate that responses to ddi are in fact improved by the presence of M184V. Genotypic data Like clinical cutoffs, genotypic scores for a given drug may differ, depending on the treatment history of the population from whom they are derived and the methodology of analysis. Genotypic scores necessarily reflect older treatment strategies and require continuous adjustment as new data emerge. This evolution should be seen as an intrinsic feature of the interpretation of resistance, but can be disconcerting for the treating physician. In some cases, genotypic analysis can provide insights into the clinical significance of resistance that are not evident by phenotypic analysis. For example, analysis of clinical databases can identify mutations in the genotypic score that predict reduced virologic responses by acting as a proxy for the presence of significant resistance. This is seen with T215rev, which reduce virologic responses despite their lack of significant effects on the virus phenotype, by either signaling the presence of more extensive NRTI resistance in minority plasma species and proviral DNA, or by accelerating the emergence of the T215Y/F mutation 57. As a further example, multiple genotypic studies have shown that the RT mutation L74V predicts reduced response to TDF. In the Gilead TDF-intensification studies, predictors of reduced response to TDF included the presence at baseline of K65R, 3 TAM (among 41, 67, 210, and 215), and L74V 31. The estimated viral load (log 10 copies/ml) was for each TAM, for K65R, and for L74V. In a different analysis of a treatment-experienced cohort, the mutations identified as part of the TDF score included M41L, E44D, D67N, T69D/N/S, L74V, L210W, and T215Y/F 58. However, the L74V mutation does not confer phenotypic resistance to TDF. In fact, L74V increases susceptibility to TDF on a background of TAM 59. One key indicator of the possible mechanism of the observed genotypic association is that the presence of L74V at baseline increased the risk of developing K65R in the TDFintensification studies 60. The model proposed is that in treatment-experienced patients who have received therapy with ABC or ddi and show L74V as the dominant mutation, K65R will also be selected and will be present at low levels within the minority quasispecies. Introduction of TDF and ongoing virus replication will then rapidly induce the emergence of K65R as dominant, leading to virologic failure. In clinical practice, the impact of minority species is likely to depend on multiple factors, including prevalence of the mutant within the quasispecies, coexistence with other mutations on the same genome, effect on fitness, whether the patient has experienced one or multiple treatment failures, the specific drug affected, the activity of other drugs in the regimen, and the level of adherence. Thus, the significance of minority K65R may be greater in patients who are highly treatment-experienced and receive TDF in the context of a poorly suppressive regimen. In contrast, the impact is likely to be less significant after failure of first-line therapy with rapid switch to a well designed second-line regimen. There is some evidence in support of this hypothesis. In four patients who had experienced failure of first-line therapy with ABC/3TC in combination with either EFV, nelfinavir, or fosamprenavir, standard genotypic testing showed the presence of M184V and/or L74V 61. Using single-genome analysis with a sensitivity of % for the detection of minority species, K65R was found at failure (week 56) in one of the four patients at a prevalence of 0.2%. A second patient had K65R already detectable at baseline with a prevalence of 0.1%, which was not enriched at failure (week 20). Conclusions Improved treatment strategies, changes in prescribing policies, and prompt management of treatment failure are changing the prevalence and patterns of NRTI resistance in treatment-experienced patients. Prevalence of TAM is declining in many cohorts, and M184V is currently the most commonly reported NRTI mutation in clinical trials of first-line therapy. The mechanisms of NRTI resistance are increasingly well defined, and complex interactions are being characterized between emergence and evolution of resistance

10 Anna Maria Geretti: Interpreting NRTI Resistance and virus replication. Several NRTI mutations, most notably M184V and K65R, have strong effects on virus replicative capacity in vitro, and there is good clinical evidence, especially for 3TC, for a virologic and immunologic benefit of continuing NRTI therapy in the presence of resistance. In addition to fitness effects, NRTI mutations can interact strongly with each other, leading to reciprocal antagonisms and hypersusceptibility effects. In vitro, the strongest effects are those mediated by M184V on susceptibility to ZDV, d4t, and TDF. In addition, K65R and L74V can also cause hypersusceptibility to ZDV. There is limited clinical evidence that these hypersusceptibility effects improve virologic responses, although small case series and analysis of single-drug intensification studies provide support to the concept. Thus, the clinical interpretation of NRTI resistance must take into account recognized effects of genotypic changes on the virus phenotype, phenotypic cutoffs that impact on responses and the way they have been derived, genotypic scores, the good evidence of residual NRTI activity despite resistance, and the potential fitness and hypersusceptibility effects of resistance mutations. Clinical translation of in vitro effects is not always straightforward. In addition, further evidence is required to guide interpretation of the possible impact of minority NRTI-resistant species. Meanwhile, testing for baseline resistance and switching early after failure are the best strategies to limit emergence of NRTI resistance and cross-resistance. References 1. Svicher V, Sing T, Santoro M, et al. Involvement of novel HIV-1 reverse transcriptase mutations in the regulation of resistance to nucleoside inhibitors. J Virol 2006;80: Brehm J, Koontz D, Pathak D, Sluis-Cremer N, Mellors J. AZT selects mutations in the connection (A371V) and RNaseH (Q509L) domains of RT that increase AZT resistance in combination with TAM and reduce susceptibility to other NRTI. Program and abstracts XV International HIV Drug Resistance Workshop; June 13-17, 2006; Sitges, Spain [abstract 125]. 3. Nikolenko G, Frankenberry K, Palmer S, et al. The HIV-1 reverse transcriptase connection domain from treatment-experienced patients contributes to AZT resistance. Program and abstracts XV International HIV Drug Resistance Workshop; June 13-17, 2006; Sitges, Spain [abstract 126]. 4. Marcelin A, Roquebert B, Malet I, et al. Relationship between mutations in HIV-1 RNaseH domain and NRTI-resistance mutations in experienced patients. Antivir Ther 2005;10:S Health Protection Agency. HIV Drug Resistance in the United Kingdom: data to end of CDR Weekly 2006;6:HIV/STIs. 6. Garcia-Leon M, de Mendoza C, Corral A, Gonzales-Lahoz J, Soriano V. Evolution of drug-resistance mutations in antiretroviral-experienced individuals seen between Program and abstracts 4 th European HIV Drug Resistance Workshop; March 29-31, 2006; Monte Carlo, Monaco [abstract 15]. 7. Van de Vijver D, Wensing A, Asjo B, et al. Patterns of predicted drug susceptibility and its change over time among 2000 isolates across Europe: the CAPTURE study. Program and abstracts 4 th European HIV Drug Resistance Workshop; March 29-31, 2006; Monte Carlo, Monaco [abstract 4]. 8. McColl D, Parkin N, Miller M. Characterization of patient-derived HIV-1 isolates containing the L74V or K65R mutations in reverse transcriptase. Program and abstracts 44 th ICAAC. October 30-November 2, 2004; Washington, DC [abstract H-178]. 9. Corvasce S, Violin M, Romano L, et al. Evidence of differential selection of HIV-1 variants carrying drug-resistant mutations in seroconverters. Antivir Ther 2006;11: Booth L, Garcia-Diaz A, Nebbia G, Youle M, Johnson M, Geretti A. The prevalence of antiretroviral drug resistance in patients with low-level detectable viremia while on HAART. Antivir Ther 2006;11:S Geretti A. Epidemiology of antiretroviral drug resistance in drugnaive persons. Curr Opin Infect Dis 2007 (in press). 12. Shet A, Berry L, Mohri H, et al. Tracking the prevalence of transmitted antiretroviral drug-resistant HIV-1: a decade of experience. J Acquir Immune Defic Syndr 2006;41: Booth C, Garcia-Diaz A, Youle M, Johnson M, Phillips A, Geretti A. Prevalence and predictors of antiretroviral drug resistance in newly diagnosed HIV-1 infection. J Antimicrob Chemother (in press). 14. Oette M, Kaiser R, Daumer M, et al. Primary HIV drug resistance and efficacy of first-line antiretroviral therapy guided by resistance testing. J Acquir Immune Defic Syndr 2006;41: Maia Teixeira S, Bastos F, Hacker M, et al. Trends in drug-resistance mutations in antiretroviral-naive intravenous drug users of Rio de Janeiro. J Med Virol 2006;78: Bezemer D, de Ronde A, Prins M, et al. Evolution of transmitted HIV-1 with drug-resistance mutations in the absence of therapy: effects on CD4+ T-cell count and HIV-1 RNA load. Antivir Ther 2006;11: Ghosn J, Pellegrin I, Goujard C, et al. HIV-1 resistant strains acquired at the time of primary infection massively fuel the cellular reservoir and persist for lengthy periods of time. AIDS 2006;20: Gallant J, Pozniak A, DeJesus E, et al. Efficacy and safety of tenofovir DF (TDF), emtricitabine (FTC) and efavirenz (EFV) compared to fixed dose zidovudine/lamivudine (CBV) and EFV through 96 weeks in antiretroviral treatment-naive patients. Program and abstracts XVI International AIDS Conference. August 13-18, 2006; Toronto, Canada [abstract TUPE0064]. 19. Miller M, Margot N, McColl D, Tran S, Coakley D, Cheng A. Genotypic and phenotypic characterization of virologic failure through 48 weeks among treatment-naive patients taking tenofovir DF or stavudine in combination with lamivudine and efavirenz. Program and abstracts 6 th International Congress on Drug Therapy in HIV infection. November 17-21, 2002; Glasgow, UK [abstract P205]. 20. Molina J, Wilkin A, Domingo P, et al. Once-daily vs. twice-daily lopinavir/ritonavir in antiretroviral-naive patients: 96-week results. Program and abstracts 3 rd IAS Conference on HIV Pathogenesis and Treatment, July 24-27, 2005; Rio de Janeiro, Brazil [abstract WePe12.3C12]. 21. Dejesus E, Herrera G, Teofilo E, et al. Efficacy and safety of abacavir versus zidovudine in antiretroviral therapy-naive adults with HIV-1 infection (Study CNA30024). Program and abstracts 43 rd ICAAC. September 14-17, 2003; Chicago, IL [abstract H-446]. 22. Gazzard B, DeJesus E, Cahn P, et al. Abacavir once daily plus lamivudine once daily in combination with efavirenz once daily is well-tolerated and effective in the treatment of antiretroviral therapy naive adults with HIV-1 infection (ZODIAC Study: CNA30021). Program and abstracts 43 rd ICAAC. September 14-17, 2003; Chicago, IL [abstract H-1722b]. 23. MacManus S, Yates P, Elston R, White S, Richards N, Snowden W. GW433908/ritonavir once daily in antiretroviral therapy naive HIVinfected patients: absence of protease resistance at 48 weeks. AIDS 2004;18: Zurita Macias Valadez L, Rodriguez-Diaz R, Fuentes-Romero L, Andrade-Toscano E, Soto-Ramirez L. Development of TAM after failure of first ARV treatment in Mexico. Program and abstracts 4 th European HIV Drug Resistance Workshop; March 29-31, 2006; Monte Carlo, Monaco [abstract 45]. 25. Kempf D, King M, Bernstein B, et al. Incidence of resistance in a double-blind study comparing lopinavir/ritonavir plus stavudine and lamivudine to nelfinavir plus stavudine and lamivudine. J Infect Dis 2004;189:

11 AIDS Reviews 2006;8 26. Hicks C, King M, Gulick R, White A, et al. Long-term safety and durable antiretroviral activity of lopinavir/ritonavir in treatment-naive patients: 4 year follow-up study. AIDS 2004;18: Malan N, Krantz E, David N, et al. Efficacy and safety of atazanavirbased therapy in antiretroviral naive HIV-1 infected subjects, both with and without ritonavir: 48-week results from AI Program and abstracts 13 th CROI; February 5-8, 2006; Denver, CO [abstract 107LB]. 28. Van Houtte M, Staes M, Geretti A, Pattery T, Bacheler L. Tenofovir drug resistance and mutational pathways associated with the RT mutation K70E in HIV-1. Program and abstracts 4 th European HIV Drug Resistance Workshop; March 29-31, 2006; Monte Carlo, Monaco [abstract 94]. 29. Stone C, Craig C. Identification of a new HIV-1 in vitro mutation to tenofovir and mutation patterns observed to the drug combinations abacavir/tenofovir/3tc, abacavir/3tc and tenofovir/ftc. Program and abstracts 3 rd European HIV Drug Resistance Workshop; March 30-April 1, 2005; Athens, Greece [abstract 98]. 30. Murry J, Higgins J, Matthews T, et al. Reversion of the M184V mutation in simian immunodeficiency virus reverse transcriptase is selected by tenofovir, even in the presence of lamivudine. J Virol 2003;77: Marcelin A, Flandre P, Pavie J, et al. Clinically relevant genotype interpretation of resistance to didanosine. Antimicrob Agents Chemother 2005;49: Waters J, Margot N, McColl D, Zhong L, Miller M. K65R, L74V and thymidine analog mutations in HIV-1 RT associated with reduced response to tenofovir DF in ART-experienced patients. Program and abstracts 13 th CROI; February 5-8, 2006; Denver, CO [abstract 633]. 33. Stone C, Lanier E, Craig C. In vitro selection of a novel reverse transcriptase mutation, D67H, by tenofovir/zidovudine in wild-type virus and of D67N/H by TFV alone in a background of thymidine analog mutations. Program and abstracts 4 th European HIV Drug Resistance Workshop; March 29-31, 2006; Monte Carlo, Monaco [abstract 98]. 34. Wainberg M. The impact of the M184V substitution on drug resistance and viral fitness. Expert Rev Anti Infect Ther 2004;2: Diallo K, Gotte M, Wainberg M. Molecular impact of the M184V mutation in HIV-1 reverse transcriptase. Antimicrob Agents Chemother 2003;47: Liang F, Miranda L, Kuritzkes D. The Impact of the M184V substitution in HIV-1 reverse transcriptase on DNA 3 -end and RNA 5 -enddirected ribonuclease H (RNaseH) activity. Program and abstracts 13 th CROI; February 5-8, 2006; Denver, CO [abstract 600]. 37. Eron J, Benoit S, Jemsek J, et al. Treatment with lamivudine, zidovudine, or both in HIV-positive patients with CD4+ cells/ mm 3. N Engl J Med 1995;333: The CAESAR Trial Study Group. Randomized trial of addition of lamivudine or lamivudine plus loviride to zidovudine-containing regimens for patients with HIV-1 infection: the CAESAR trial. Lancet 1997;349: Castagna A, Danise A, Menzo S, et al. Lamivudine monotherapy in HIV-1-infected patients harboring a lamivudine-resistant virus: a randomized pilot study (E-184V study). AIDS 2006;20: Ledergerber B, Lundgren J, Walker A, et al. Predictors of trend in CD4-positive T-cell count and mortality among HIV-1-infected individuals with virologic failure to all three antiretroviral-drug classes. Lancet 2004;364: Deeks S, Hoh R, Neilands T, et al. Interruption of treatment with individual therapeutic drug classes in adults with multidrug-resistant HIV-1 infection. J Infect Dis 2005;192: Deeks S, Lu J, Hoh R, et al. Interruption of enfuvirtide in patients with enfuvirtide resistance. Program and abstracts 12 th CROI; February 22-25, 2005; Boston, MA [abstract 680]. 43. Geretti A, Sabin C, Dunn D. The H208Y mutation in reverse transcriptase is significantly associated with NRTI exposure, M184V and TAM mutation pathways. Antivir Ther 2006;11:S Ross L, Parkin N, Gerondelis P, et al. Differential impact of thymidine analog mutations on emtricitabine and lamivudine susceptibility. J Acquir Immune Defic Syndr 2006;43: Garcia-Lerma J, MacInnes H, Bennett D. A novel genetic pathway of HIV-1 resistance to stavudine mediated by the K65R mutation. J Virol 2003;77: Parikh U, Bacheler L, Koontz D, Mellors J. The K65R mutation in HIV- 1 reverse transcriptase exhibits bidirectional phenotypic antagonism with thymidine analog mutations. J Virol 2006; 80: White K, Chen J, Feng J, et al. The K65R reverse transcriptase mutation in HIV-1 reverses the excision phenotype of zidovudine resistance mutations. Antivir Ther 2006;11: White K, Margot N, Ly J, et al. A combination of decreased NRTI incorporation and decreased excision determines the resistance profile of HIV-1 K65R RT. AIDS 2005;19: Staszewski S, Dauer B, Locher L, Moesch M, Gute P, Stuermer M. Intensification of a failing regimen with zidovudine may cause sustained virologic suppression in the presence of the K65R mutation. Program and abstracts 13 th CROI; February 5-8, 2006; Denver, CO [abstract 635]. 50. DART Virology Group and Trial Team. Virologic response to a triple nucleoside/nucleotide analog regimen over 48 weeks in HIV-1-infected adults in Africa. AIDS 2006;20: Frankel F, Marchand B, Turner D, Gotte M, Wainberg M. Impaired rescue of chain-terminated DNA synthesis associated with the L74V mutation in HIV-1 reverse transcriptase. Antimicrob Agents Chemother 2005;49: Miranda L, Gotte M, Liang F, Kuritzkes D. The L74V mutation in HIV- 1 reverse transcriptase counteracts enhanced excision of zidovudine monophosphate associated with thymidine analog resistance mutations. Antimicrob Agents Chemother 2005;49: Wirden M, Roquebert B, Derache A, et al. Risk factors for selection of the L74I reverse transcriptase mutation in HIV-1-infected patients. Antimicrob Agents Chemother 2006;50: ccoffs&product=vircotype [accessed 03/12/2006] [accessed 03/12/2006]. 56. Underwood M, St Clair M, Ross L, Gerondelis P, Parkin N, Lanier R. Cross-resistance of Clinical Samples with K65R, L74V, and M184V Mutations. Program and abstracts 12 th CROI; February 22-25, 2005; Boston, MA [abstract 714]. 57. Violin M, Cozzi-Lepri A, Velleca R, et al. Risk of failure of patients with 215 HIV-1 revertants starting their first thymidine analog-containing HAART. AIDS 2004;18: Masquelier B, Tamalet C, Montes B, et al. Genotypic determinants of the virologic response to tenofovir disoproxil fumarate in NRTIexperienced patients. Antivir Ther 2004;9: Trivedi V, Han X, Montes-Walters M, Filizzola M, Ferguson M, O Brien W. Effect of L74V or M184V, with multiple thymidine analogassociated mutations, on reverse transcriptase inhibitor phenotypic sensitivity. Program and abstracts 12 th CROI; February 22-25, 2005; Boston, MA [abstract 703]. 60. Svarovoskaia E, Margot N, Bae A, Waters J, Borroto-Esoda K, Miller M. Allele-specific PCR shows low-level K65R in treatment experienced patients with L74V in the absence of TAM. Antivir Ther 2006;11:S Descamps D, Delarue S, Ait-Khaled M, Craig C, Collin G, Brun-Vézinet F. Rare selection of the K65R mutation in antiretroviral-naive patients failing a first-line abacavir/lamivudine containing HAART regimen. Program and abstracts 4 th European HIV Drug Resistance Workshop; March 29-31, 2006; Monte Carlo, Monaco [abstract 64]. 220