Broad Nucleoside Reverse-Transcriptase Inhibitor Cross-Resistance in Human Immunodeficiency Virus Type 1 Clinical Isolates

Transcription

1 MAJOR ARTICLE Broad Nucleoside Reverse-Transcriptase Inhibitor Cross-Resistance in Human Immunodeficiency Virus Type 1 Clinical Isolates Jeannette M. Whitcomb, Neil T. Parkin, Colombe Chappey, Nicholas S. Hellmann, a and Christos J. Petropoulos ViroLogic, South San Francisco, California. Nucleoside reverse-transcriptase inhibitors (NRTIs) are important components of most antiretroviral combination treatment regimens. Using a large collection of clinical isolates, we characterized patterns of crossresistance among all NRTIs. Drugs were grouped by the effect of the M184V mutation: susceptibility to group 1 drugs (zidovudine, stavudine, tenofovir, and adefovir) increased when M184V was present, whereas susceptibility to group 2 drugs (didanosine, zalcitabine, abacavir, and lamivudine) decreased. Significant crossresistance was observed among all NRTIs and was most notable when samples with or without M184V were analyzed separately. An increasing number of thymidine-analogue mutations (TAMs) was associated with a progressive reduction in drug susceptibility for all NRTIs. The modulating effect of M184I/V on drug susceptibility was present regardless of the number of TAMs. The broad range of susceptibility observed for viruses containing the same number of TAMs indicates that the genetic correlates of NRTI resistance remain to be fully elucidated. The introduction of highly active antiretroviral therapy (HAART) for the treatment of human immunodeficiency virus (HIV) type 1 infected individuals has greatly reduced the morbidity and mortality associated with this disease [1]. However, HIV has the ability to develop resistance to all available antiretroviral (ARV) agents [2]. As a consequence, an ongoing challenge for clinicians is the appropriate selection and sequencing of treatment regimens to achieve long-term virus suppression. Thus, the selection of active drugs after the failure of prior regimens is a critical issue for optimal patient management. Nucleoside reverse-transcriptase inhibitors (NRTIs) Received 21 February 2003; accepted 22 April 2003; electronically published 16 September Presented in part: 9th Conference on Retroviruses and Opportunistic Infections, Seattle, February 2002 (abstract 569); 1st IAS Conference on HIV Pathogenesis and Treatment, Buenos Aires, 8 11 July 2002 (abstract 592). All coauthors are employed at and are shareholders of ViroLogic, Inc., which is a commercial supplier of phenotypic and genotypic drug resistance diagnostic services. a Present affiliation: Roche Molecular Systems, Pleasanton, California. Reprints or correspondence: Dr. Christos Petropoulos, ViroLogic, Inc., 345 Oyster Point Blvd., South San Francisco, CA (cpetropoulos@virologic.com). The Journal of Infectious Diseases 2003; 188: by the Infectious Diseases Society of America. All rights reserved /2003/ $15.00 were the first ARV drugs to be approved for the treatment of HIV, with zidovudine (ZDV) receiving Food and Drug Administration (FDA) approval in Although 7 protease inhibitors and 3 nonnucleoside reverse-transcriptase inhibitors (NNRTIs) are now available for use in treatment regimens, NRTIs continue to be the cornerstone of HAART [3]. Members of the NRTI class that are currently available in the United States include abacavir (ABC), didanosine (ddi), lamivudine (3TC), stavudine (d4t), zalcitabine (ddc), ZDV, emitricitabine, and tenofovir (TDF), a nucleotide reverse-transcriptase (RT) inhibitor. A number of investigational NRTI drugs are currently in development. RT mutations frequently identified in viruses from patients who have received ZDV (M41L, D67N, K70R, L210W, T215Y or F, and K219Q) were thought initially to be associated only with ZDV resistance [4 7]. However, the response to other NRTIs after ZDV failure is often muted [8], and ZDV mutations emerge during therapy with d4t or ddi in the absence of ZDV [9]. As a result, ZDV mutations are now commonly included among the thymidine analogue mutations (TAMs), to indicate their broader impact on resistance to the NRTI class. The magnitude of reduced ZDV 992 JID 2003:188 (1 October) Whitcomb et al.

2 susceptibility conferred by TAMs can be modulated by mutations selected for by other drugs. First described for the 3TC resistance mutation M184V, a total of 3 NRTI mutations (K65R, L74V, and M184V) and 2 NNRTI mutations (L100I and Y181C) are now recognized to increase ZDV susceptibility by suppressing resistance or resensitizing HIV to ZDV in the presence of 1 TAMs [10 17]. The availability of highly sensitive phenotypic assays has resulted in an appreciation of the extent of cross-resistance among NRTIs. For example, a recent study that used the PhenoSense HIV assay [18] examined the clinical relevance of d4t crossresistance in patients who had received ZDV [19]. In that study, patients who had received ZDV monotherapy for 13 years and then switched to d4t were more likely to have a virological response if the baseline susceptibility to d4t was not 11.4-fold higher than that of the susceptible wild-type reference virus. Cross-resistance among NRTI drugs has been found to be associated with certain multidrug resistance (MDR) mutations in RT, such as Q151M and T69 insertion mutations [20 24]. However, these MDR mutations are seen relatively infrequently [25, 26] and cannot account for the much higher rates of crossresistance among NRTIs that is seen in clinical settings. The optimum use and sequencing of NRTI drugs is dependent on a better understanding of cross-resistance among these drugs. The purpose of the present study was to characterize the extent, patterns, and etiology of cross-resistance among NRTI drugs in the absence of previously defined MDR mutations, using a large database of clinical samples submitted for routine phenotypic and genotypic testing. MATERIALS AND METHODS Phenotypic drug susceptibility was determined using the PhenoSense HIV assay [18], and the results were expressed as the fold change (FC), calculated as the IC 50 of the patient virus sample divided by the IC 50 of the NL4-3 reference strain tested in the same assay batch. Bivariate scatter plots of log-transformed FC data were generated for each pair of NRTIs. In each plot, samples were divided into 2 groups on the basis of the absence or presence of a mutation at position 184 in RT (I or V). RT genotypes were determined using GeneSeq HIV (ViroLogic), which uses the same resistance test vector as that in the PhenoSense assay. Deduced amino-acid sequences of patient viruses were compared with a reference virus strain (NL4-3). Nucleotide sequences were determined using dideoxy nucleotide dye terminator chemistry and resolved on ABI 3700 automated sequencers. The GeneSeq HIV assay can detect most minor virus populations in plasma samples when they are present at 10% 20% of the total virus population (GeneSeq HIV validation data; ViroLogic; data not shown). A database containing NRTI susceptibilities from 5932 clinical HIV-1 samples with matched phenotypic and genotypic measurements was evaluated. All available phenotypic data for 7 of the 8 FDA-approved NRTIs and adefovir (ADV) were included. Samples were excluded from the analyses if the isolates lacked any nucleoside analogue resistance-associated mutation, contained MDR mutations (Q151 complex or T69 insertions), or genotypic mixtures at the following positions in RT: 41, 67, 70, 184, 210, 215, and 219. When the IC 50 of the patient virus exceeded the highest drug concentration tested, the FC was assigned a fixed value that was determined by dividing the highest drug concentration tested by the IC 50 of the wild-type reference virus. TAMs were defined as M41L, D67N, K70R, L210W, T215F or Y, and K219E, H, N, Q, or R. On the basis of existing data, M184I and V were considered to have equivalent effects on NRTI susceptibility [18, 27]. The M184I variant was present in 2% of samples with any mutation at position 184. Provided that sufficient clinical outcome data are available, a threshold (or cutoff) can be derived that defines the drug susceptibility value above which the virological response to treatment with that drug significantly declines. Clinical cutoffs have been defined for 5 of 8 FDA-approved NRTIs: 4.5-fold for ABC [28], 1.7-fold for ddi and d4t [29 32], 1.4-fold for TDF [33], and 3.5-fold for 3TC [34]. Cutoffs for ZDV, ddc, and ADV were 2.5, 1.7, and 1.4-fold, respectively, and were derived from assay reproducibility data [35]. All cutoffs were derived using the PhenoSenseHIV assay and may not be directly applicable to other phenotypic assays. Pairwise comparisons of NRTI susceptibilities as continuous variables were done for each drug using the nonparametric Spearman s rank test or linear correlation using Statview software (version 5.0; SAS). RESULTS Patterns of cross-resistance. Susceptibilities of patient viruses to each NRTI were evaluated using phenotypic assays. Results were expressed as the FC in IC 50 compared with the NL4-3 reference. Bivariate scatter plots of log-transformed FC data were generated for each pair of NRTIs and are shown in figures 1 and 2. The scatter plots of NRTI pairs exhibiting weaker correlations (figure 2) suggested the existence of 2 distinct populations that, if evaluated separately, might exhibit stronger correlations. The subset of NRTI pairs that included 3TC displayed exaggerated separations and suggested that the M184V mutation might distinguish the 2 apparent populations. To test this hypothesis, we reexamined the correlations for each NRTI pair using subpopulations segregated on the basis of the presence or absence of the M184I or V mutations. This analysis revealed 2 groups of NRTIs: group 1 included ZDV, d4t, TFV, and ADV, which showed increased susceptibility in the presence of M184I/V; and group 2 included 3TC, ddi, ddc, and ABC, which showed reduced susceptibility in the presence Broad NRTI Cross-Resistance in HIV-1 JID 2003:188 (1 October) 993

3 Figure 1. Scatter plots displaying phenotypic cross-resistance (log 10 -transformed fold change [FC]) among pairs of nucleoside reverse-transcriptase inhibitors (NRTIs): within-group comparisons are shown. Samples with multiple NRTI resistance mutations (Q151M or T69 insertions), mixtures at M184, or thymidine-analogue mutations were excluded. Samples with M184I or V are in black; those lacking such mutations are in gray. Points at the extreme high end of the lamivudine (3TC) or abacavir (ABC) scale represent samples with high-level resistance that were assigned a fixed FC value that reflects the highest measurable level of resistance in the assay (see Materials and Methods). The oval in the zidovudine (ZDV) tenofovir (TDF) plot indicates a group of samples containing the K65R mutation. Group 1 NRTIs: ZDV, stavudine (d4t), TDF, and adefovir (ADV); group 2: 3TC, ABC, didanosine (ddi), and zalcitabine (ddc). of M184I/V. NRTI pairs within the same group all exhibited strong correlations that were independent of M184I/V (figure 1), as shown by the relatively small differences in the Spearman s rank r value among all samples, compared with those that did not have M184I/V mutations (table 1). NRTI pairs from different groups all exhibited weaker correlations if M184I/V was not considered (figure 2), and the Spearman s rank r value increased dramatically in samples without M184I/ V (table 1). For the within-group comparisons (figure 1), with the exception of pairs involving 3TC, the data points fell along a single diagonal line with a high correlation coefficient; samples containing or lacking M184I/V clustered at either end of the line. This is most striking for the ZDV:TDF and ADV:TDF plots, whereas somewhat more scatter was seen for pairs such as d4t:zdv and TDF:d4T. Conversely, for between-group comparisons, 2 separate but parallel linear clusters of samples were usually observed that were defined by the presence or absence of the M184I/V mutation (figure 2). Samples with 184 mutations tended to cluster along a shifted diagonal line parallel to the samples without 994 JID 2003:188 (1 October) Whitcomb et al.

4 Figure 2. Scatter plots displaying phenotypic cross-resistance (log 10 -transformed fold change) among pairs of nucleoside reverse-transcriptase inhibitors (NRTIs): between-group comparisons are shown. Samples with multiple NRTI resistance mutations (Q151M or T69 insertions), mixtures at M184, or thymidine-analogue mutations were excluded. Samples with M184I or V are in black; those lacking such mutations are in gray. Points at the extreme high end of the lamivudine (3TC) or abacavir (ABC) scale represent samples with high-level resistance that were assigned a fixed FC value that reflects the highest measurable level of resistance in the assay (see Materials and Methods). Group 1 NRTIs: zidovudine (ZDV), stavudine (d4t), tenofovir (TDF), and adefovir (ADV); group 2: 3TC, ABC, didanosine (ddi), and zalcitabine (ddc). 184 mutations, as exemplified by the ABC:d4T or ddi:tdf plots. Because the M184I/V mutation causes very high-level resistance to 3TC, a more dramatic separation of the samples with and without 184 mutations was observed in the 3TC scatter plots. However, strong correlations ( r p ) in sus- ceptibility between 3TC and other NRTIs was readily apparent from the samples without 184 mutations (table 1). Although the susceptibilities among all NRTIs were correlated, the range of reduced susceptibility varied dramatically among NRTIs. Although most of the scatter plots revealed linear relationships between the log-transformed FC data for each pair of NRTIs, there were a few subtle, nonlinear patterns that could be discerned. For example, in the ABC:ddI and ABC:ddC plots, the cluster of samples with M184I/V appeared to lie on a diagonal line that was slightly shifted, compared with samples without M184 mutations (figure 1), which demonstrates the greater impact of the 184 mutation on ABC susceptibility than on the other drugs. Also, in the TDF:ZDV plot, a small cluster of samples with decreased TDF susceptibility but increased ZDV susceptibility can be seen. These samples all have the K65R mutation (oval in figure 1). Similarly, clusters of samples in plots involving ZDV or TDF with ddi or ABC could be seen Broad NRTI Cross-Resistance in HIV-1 JID 2003:188 (1 October) 995

5 Table 1. Correlation coefficients (Spearman s r) of nucleoside reverse-transcriptase inhibitor susceptibility. Comparison, drug 1 and drug 2 All 184M 184I or V Difference (r 184M allr) Within-group ABC and 3TC NA 0.04 ABC and ddc ABC and ddi ddc and 3TC NA 0.12 ddi and 3TC NA 0.04 ddi and ddc TDF and ADV TDF and d4t ZDV and d4t ZDV and TDF Between-group d4t and 3TC NA 0.49 d4t and ABC d4t and ddc d4t and ddi TDF and 3TC NA 0.66 TDF and ABC TDF and ddc TDF and ddi ZDV and 3TC NA 0.44 ZDV and ABC ZDV and ddc ZDV and ddi NOTE. All lamivudine (3TC) fold-change values in samples with M184I or V were ABC, abacavir; ADV, adefovir; d4t, stavudine; ddc, zalcitabine; ddi, didanosine; NA, not applicable; TDF, tenofovir; ZDV, zidovudine. (figure 2) that tended to contain other mutations such as L74I/ V (not shown). Correlation of NRTI susceptibilities. To more accurately quantitate the correlation among NRTI susceptibilities and the degree of cross-resistance between drugs, Spearman s rank correlation coefficient was determined for all pairs of NRTIs (table 1). The results for all samples indicated that significant positive correlations ( P!.0001) existed for all pairwise comparisons except TDF:3TC; however, within-group correlations (r p ) were all stronger than between-group correlations ( r p ). The strongest correlation within group 1 was TDF:ADV ( r p 0.95), and that within group 2 was ddi:abc ( r p 0.90). Linear correlation coefficients were highly consis- tent with Spearman s rank test results (data not shown). In subset analyses of viruses without M184I/V, all pairwise comparisons, both within and between groups, showed strong positive correlations ( r p ; all P!.0001). Notably, be- tween-group correlations within samples without M184I/V improved significantly compared with all samples (mean r change, +0.34), whereas within-group correlations were unchanged (mean r change, 0.08). The TDF:ADV comparison had the highest Spearman rank correlation coefficient in 184 wild-type samples ( r p 0.90), and the ZDV:TDF, ABC:3TC, and d4t: ABC comparisons had linear correlation coefficients in the same samples (table 1 and data not shown). Correlations were evident between susceptibilities to ZDV and all other NRTIs in M184 wild-type samples, with all r values 10.5 except that for ZDV:ddC. Effect of TAMs and M184V on NRTI susceptibility distributions. Because strong correlations were observed between ZDV susceptibility and that of the other NRTIs, the impact of TAMs on susceptibility to each NRTI was assessed by comparing the NRTI susceptibility of virus populations containing increasing numbers of TAMs (figure 3). The effect of the M184I/V mutation on susceptibility to NRTIs in the context of various numbers of TAMs was also examined. Invariably, increasing numbers of TAMs were correlated with decreased susceptibility to all NRTIs. Given equal numbers of TAMs, virus populations containing the M184I/V mutation were more susceptible to group 1 NRTIs than virus populations without M184I/ V. Viruses with NRTI hypersusceptibility ( FC! 0.4) for all group 1 NRTIs were observed in M184I/V virus populations without TAMs. In some cases, NRTI hypersusceptibility was observed in M184I/V viruses that had up to 2 TAMs (e.g., TDF; figure 3). Virus populations containing the M184I/V mutation were always less susceptible to group 2 NRTIs than virus populations without M184I/V that had the same number of TAMs. The effect of TAMs on 3TC susceptibility in M184I/V viruses could not be evaluated under these assay conditions (i.e., 3TC concentrations). The likelihood of a virus to exhibit reduced susceptibility to a given NRTI was evaluated by tabulating the percentage of samples over the cutoff for each NRTI within genotypic groups, defined by the presence or absence of M184I or V or TAMs (table 2). Viruses with the M184I/V mutation were less likely to have an FC that exceeded the cutoffs of group 1 NRTIs. Viruses with the M184I/V mutation were much more likely to have an FC that exceeded the cutoffs of group 2 NRTIs. DISCUSSION We describe the existence of broad cross-resistance among all members of the NRTI class. Two distinct groups of NRTIs were identified on the basis of the impact of the M184I or V mutation on drug susceptibility. The etiology of the broad NRTI crossresistance is clearly associated with, but not completely explained by, mutations in RT that are typically considered to be associated with thymidine nucleoside analogue drugs (ZDV and d4t). The extent of NRTI cross-resistance has been underappreciated for at least 3 reasons. First, phenotypic assays with sufficient accuracy and precision to measure low-level reductions 996 JID 2003:188 (1 October) Whitcomb et al.

6 Figure 3. Box plots showing nucleoside reverse-transcriptase inhibitors susceptibility and the effects of thymidine-analogue mutations (TAMs; mutations at positions 41, 67, 70, 210, 215, and 219; see Materials and Methods) and M184I/V are shown. For each box, the center line represents the median fold change (FC), the box boundaries define the 25th and 75th percentiles, and the whiskers define the 10th and 90th percentiles. Light gray boxes, M184 wild type; dark gray boxes, M184I/V; thick dotted lines, current clinical (stavudine [d4t], tenofovir [TDF], didanosine [ddi], abacavir [ABC], and lamivudine [3TC]) or assay (zidovudine [ZDV]) cutoffs. in susceptibility for some NRTIs such as ddi, d4t, and TDF have only recently become available. Because subtle (!2-fold) reductions in drug susceptibility are associated with the treatment failure of these drugs, less-precise phenotypic assays were unable to fully characterize the clinically relevant levels of drug susceptibility. Second, inadequate sample numbers hindered prior evaluations of cross-resistance. The large data set used in the present study provided adequate power for cross-resistance analyses and was generated by a single laboratory by use of well-controlled phenotypic and genotypic assays. In addition, samples that contain mixtures of wild-type and mutant amino acids at certain resistance-associated positions were eliminated, further minimizing noise in the data. Mixtures of wild-type and resistant viruses can dilute the level of resistance observed and act as an important cause of phenotype-genotype discordance [36, 37]. Finally, the ability of the M184I/V mutation to increase susceptibility to some NRTIs and decrease susceptibility to others masks correlations between many NRTI pairs and must be taken into account in order to reveal the relationships between NRTIs of different groups. Reductions in drug susceptibility beyond certain thresholds (the clinical cutoff) are associated with an increased likelihood of a poor virological response during treatment with that drug [28, 29, 33, 34, 38]. Thus, the findings of the present study may have important implications for the sequencing of NRTIs in treatment regimens. NRTIs remain a cornerstone of most antiretroviral treatment regimens and must be used prudently to sustain virus suppression. For NRTIs with defined clinical cutoffs, such as TDF (1.4-fold), ABC (4.5-fold), ddi and d4t (1.7-fold), and 3TC (3.5-fold), it is clear that clinically relevant Broad NRTI Cross-Resistance in HIV-1 JID 2003:188 (1 October) 997

7 Table 2. Percentage of samples with or without M184I or V that are over the fold-change (FC) cutoff for each nucleoside reverse-transcriptase inhibitor. 184 TAMs Samples over the FC cutoff, % ZDV d4t TDF ADV 3TC ddi ddc ABC M I or V M IorV NOTE. FC cutoffs were 2.5 for zidovudine (ZDV); 1.7 for didanosine (ddi), stavudine (d4t), and zalcitabine (ddc); 1.4 for adefovir (ADV) and tenofovir (TDF); 3.5 for lamivudine (3TC), and 4.5 for abacavir (ABC). TAMs, thymidineanalogue mutations. reductions in susceptibility can result when TAMs accumulate during prior NRTI exposure. Clinical data support the notion that cross-resistance to a previously unused NRTI can occur after exposure to other NRTIs [9, 19, 28, 33, 34]. The clinical relevance of increases in susceptibility to the group 1 NRTIs when M184I or V is present is not fully understood. However, the presence of M184I/V, a 3TC resistance mutation, has been reported to preserve the activity of ZDV even though multiple TAMs are also present, which suggests that the continuation of 3TC therapy may provide a clinical benefit during ZDV therapy [39]. TAMs emerge more slowly in patients receiving ZDV/3TC combination therapy than in patients receiving ZDV alone [40]. Similar data have also suggested that virological responses to TDF therapy are greater in the presence of the M184V mutation [41]. Thus, for optimal NRTI sequencing, it is important to consider both the phenotypic susceptibility (to determine presence or absence of NRTI resistance) and the genotype (to determine the presence of detrimental or potentially beneficial mutations) when deciding whether to change a drug regimen in the face of a rebound in virus load. The presence of other mutations that can also increase ZDV susceptibility (such as K65R, L74V, L100I, and Y181C) is also important to assess, because these mutations may have similar resensitization effects on other group 1 NRTIs [42]. The inability to evaluate RT sequences beyond aa 305 with the PhenoSense assay is a potential limitation of our study, because alterations in NRTI susceptibility by mutations beyond aa 305 would be missed. However, when mutations in this region, which have been reported to be involved in ZDV-3TC resistance [43], were introduced into the NL4-3 reference vector, alone or in combination with various TAMs or M184V, significant changes in NRTI susceptibility were not seen [27]. Mutations at positions 44 and 118 have been proposed to be partly responsible for low-level resistance to 3TC [44]. Although these mutations have been found at higher frequency in samples that display NRTI cross-resistance, such samples also have higher numbers of TAMs, and many cross-resistant samples lack these mutations (authors unpublished observations), which makes it difficult to discern the effect of 44/118 from those of increased numbers of TAMs. Similarly, the effects of mutations at positions 208, 211, and 214 [45 47] cannot be distinguished from those of multiple TAMs, although they may play a subtle modulatory role. Our observations have important implications for the selection of ARV therapy regimens that contain NRTIs and indicate that cross-resistance among NRTIs is far more common than has been previously appreciated. The loss of activity to a single NRTI, especially ZDV or d4t, can have significant implications for the activity of many or all NRTIs. These results provide an explanation for the findings from many clinical trials that the failure of the first NRTI-containing regimen is associated with a suboptimal response to subsequent NRTI regimens. Phenotypic assays can detect clinically relevant cross-resistance that results from TAMs or other mutations in RT in cases where genotypic correlates are not well defined. Acknowledgments This work would not have been possible without the dedication and attention to quality of the ViroLogic Clinical Reference Laboratory staff during phenotypic and genotypic testing. We are also grateful to Howard Cash and GeneCodes for the provision of the customized sequence analysis software used in the GeneSeq assay. References 1. Palella FJ, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. N Engl J Med 1998; 338: Richman D. Antiretroviral drug resistance: mechanisms, pathogenesis, clinical significance. Adv Exp Med Biol 1996; 394: Yeni PG, Hammer SM, Carpenter CC, et al. Antiretroviral treatment for adult HIV infection in 2002: updated recommendations of the International AIDS Society USA Panel. JAMA 2002; 288: Larder BA, Kemp SD. Multiple mutations in HIV-1 reverse transcriptase confer high-level resistance to zidovudine (AZT). Science 1989; 246: Kellam P, Boucher CAB, Larder B. Fifth mutation in human immunodeficiency virus type 1 reverse transcriptase contributes to the development of high-level resistance to zidovudine. Proc Natl Acad Sci USA 1992; 89: Hooker DJ, Tachedjian G, Solomon AE, et al. An in vivo mutation from leucine to tryptophan at position 210 in human immunodeficiency virus type 1 reverse transcriptase contributes to high-level resistance to 3 -azido-3 -deoxythimidine. J Virol 1996; 70: Harrigan PR, Konghorn I, Bloor S, et al. Significance of amino acid variation at human immunodeficiency virus type 1 reverse transcriptase residue 210 for zidovudine susceptibility. J Virol 1996; 70: Miller V, Larder BA. Mutational patterns in the HIV genome and crossresistance following nucleoside and nucleotide analogue drug exposure. Antivir Ther 2001; 6: Coakley EP, Gillis JM, Hammer SM. Phenotypic and genotypic resistance patterns of HIV-1 isolates derived from individuals treated with didanosine and stavudine. AIDS 2000; 14:F JID 2003:188 (1 October) Whitcomb et al.

8 10. Larder BA. 3 -Azido-3 -deoxythymidine resistance suppressed by a mutation conferring human immunodeficiency virus type 1 resistance to nonnucleoside reverse transcriptase inhibitors. Antimicrob Agents Chemother 1992; 36: Tisdale M, Kemp SD, Parry NR, Larder BA. Rapid in vitro selection of human immunodeficiency virus type 1 resistant to 3 -thiacytidine inhibitors due to a mutation in the YMDD region of reverse transcriptase. Proc Natl Acad Sci USA 1993; 90: Wainberg MA, Miller MD, Quan Y, et al. In vitro selection and characterization of HIV-1 with reduced susceptibility to PMPA. Antivir Ther 1999; 4: Byrnes VW, Emini EA, Schleif WA, et al. Susceptibilities of human immunodeficiency virus type 1 enzyme and viral variants expressing multiple resistance engendering amino acid substitutions to reverse transcriptase inhibitors. Antimicrob Agents Chemother 1994; 38: Bazmi HZ, Hammond JL, Cavalcanti SC, Chu CK, Schinazi RF, Mellors JW. In vitro selection of mutations in the human immunodeficiency virus type 1 reverse transcriptase that decrease susceptibility to ( )- b-d-dioxolane-guanosine and suppress resistance to 3 -azido-3 -deoxythymidine. Antimicrob Agents Chemother 2000; 44: Boucher CAB, Cammack N, Schipper P, et al. High-level resistance to ( )enantiomeric 2 -deoxy-3 - thiacytidine in vitro is due to one amino acid substitution in the catalytic site of human immunodeficiency virus type 1 reverse transcriptase. Antimicrob Agents Chemother 1993; 37: Nijhuis M, Schuurman R, de Jong D, et al. Lamivudine-resistanthuman immunodeficiency virus type 1 variants (184V) require multiple amino acid changes to become co-resistant to zidovudine in vivo. J Infect Dis 1997; 176: Schmit JC, Martinez-Picado J, Ruiz L, et al. Evolution of HIV drug resistance in zidovudine/zalcitabine- and zidovudine/didanosine-experienced patients receiving lamivudine-containing combination therapy. Antivir Ther 1998; 3: Petropoulos CJ, Parkin NT, Limoli KL, et al. A novel phenotypic drug susceptibility assay for human immunodeficiency virus type 1. Antimicrob Agents Chemother 2000; 44: Shulman NS, Hughes MD, Winters MA, et al. Subtle decreases in stavudine phenotypic susceptibility predict poor virologic response to stavudine monotherapy in zidovudine-experienced patients. J Acquir Immune Defic Syndr 2002; 31: Schmit JC, Cogniaux J, Hermans P, et al. Multiple drug resistance to nucleoside analogues and nonnucleoside reverse transcriptase inhibitors in an efficiently replicating human immunodeficiency virus type 1 patient strain. J Infect Dis 1996; 174: de Jong J, Goudsmit J, Lukashov V, et al. Insertion of two amino acids combined with changes in reverse transcriptase containing tyrosine-215 of HIV-1 resistant to multiple nucleoside analogs. AIDS 1999; 13: Larder B, Bloor S, Kemp S, et al. A family of insertion mutations between codons 67 and 70 of human immunodeficiency virus type 1 reverse transcriptase confer multinucleoside analog resistance. Antimicrob Agents Chemother 1999; 43: Winters M, Coolley K, Girard Y, et al. A 6 base pair insert in the reverse transcriptase gene of human immunodeficiency virus type 1 confers to multiple nucleoside inhibitors. J Clin Invest 1998; 102: Shafer RW, Kozal MJ, Winters MA, et al. Combination therapy with zidovudine and didanosine selects for drug-resistant human immunodeficiency virus type 1 strains with unique patterns of pol gene mutations. J Infect Dis 1994; 169: Masquelier B, Race E, Tamalet C, et al. Genotypic and phenotypic resistance patterns of human immunodeficiency virus type 1 variants with insertions or deletions in the reverse transcriptase (RT): multicenter study of patients treated with RT inhibitors. Antimicrob Agents Chemother 2001; 45: Van Vaerenbergh K, Van Laethem K, Albert J, et al. Prevalence and characteristics of multinucleoside-resistant human immunodeficiency virus type 1 among European patients receiving combinations of nucleoside analogues. Antimicrob Agents Chemother 2000; 44: Tian H, Whitcomb JM, Limoli K, et al. Zidovudine-lamivudine (AZT- 3TC) co-resistance is preceded by a transient period of zidovudine hypersensitivity. Antivir Ther 1998; 3(Suppl 1): Lanier ER, Hellmann N, Scott J. Determination of a clinically relevant phenotypic resistance cutoff for abacavir using the PhenoSense assay [abstract 254]. In: Program and abstracts of the 8th Conference on Retroviruses and Opportunistic Infections (Chicago). Alexandria, VA: Foundation for Retrovirology and Human Health, Haubrich R, Keiser P, Kemper C, et al. CCTG 575: a randomized, prospective study of phenotype testing versus standard of care for patients failing antiretroviral therapy. Antivir Ther 2001; 6(Suppl 1): Haubrich R, Keiser P, Kemper C, et al. CCTG 575: a randomized, prospective study of phenotype testing (PHENO) versus standard of care (SOC) for patients failing antiretroviral therapy (ARV) [abstract 127]. In: Program and abstracts of the 1st IAS Conference on HIV Pathogenesis and Treatment (Buenos Aires). Stockholm: International AIDS Society, Katzenstein DA, Bosch RJ, Hellmann N, Wang N, Bacheler L, Albrecht MA. Phenotypic susceptibility and virological outcome in nucleosideexperienced patients receiving three or four antiretroviral drugs. AIDS 2003; 17: Shulman NS, Hughes MD, Winters MA, et al. Subtle decreases in stavudine phenotypic susceptibility predict poor virologic response to stavudine monotherapy in zidovudine-experienced patients. J Acquir Immune Defic Syndr 2002; 31: Lu B, Hellmann NS, Bates M, Dawson K, Rooney J, Miller MD. Determination of clinical cut-offs for reduced response to tenofovir DF therapy in antiretroviral-experienced patients. Antivir Ther 2002; 7(Suppl 1):S Skowron G, Whitcomb J, Wesley M, et al. Viral load response to the addition of lamivudine correlates with phenotypic susceptibility to lamivudine and the presence of T215Y/F in the absence of M184V. Antivir Ther 1999; 4(Suppl 1): Hellmann N, Johnson P, Petropoulos C. Validation of the performance characteristics of a novel, rapid phenotypic drug susceptibility assay, PhenoSenseHIV. Antivir Ther 1999; 4(Suppl 1): Parkin N, Chappey C, Maroldo L, Bates M, Hellmann NS, Petropoulos CJ. Phenotypic and genotypic HIV-1 drug resistance assays provide complementary information. J Acquir Immune Defic Syndr 2002; 31: Parkin NT, Chappey C, Petropoulos CJ. Improving lopinavir genotype algorithm through phenotype correlations: novel mutation patterns and amprenavir cross-resistance. AIDS 2003; 17: Kempf DJ, Isaacson JD, King MS, et al. Analysis of the virologic response with respect to baseline viral phenotype and genotype in protease inhibitor experienced HIV-1 infected patients receiving lopinavir/ritonavir therapy. Antivir Ther 2002; 7: Kuritzkes DR, Quinn JB, Benoit SL, et al. Drug resistance and virologic response in NUCA 3001, a randomized trial of lamivudine (3TC) versus zidovudine (ZDV) versus ZDV plus 3TC in previously untreated patients. AIDS 1996; 10: Larder BA, Kemp SD, Harrigan PR. Potential mechanism for sustained antiretroviral efficacy of AZT-3TC combination therapy. Science 1995; 269: Margot NA, Isaacson E, McGowan I, Cheng AK, Schooley RT, Miller MD. Genotypic and phenotypic analyses of HIV-1 in antiretroviralexperienced patients treated with tenofovir DF. AIDS 2002; 16: Parkin NT, Chappey C, Petropoulos CJ, Hellmann N. HIV-1 reserve transcriptase mutations that suppress zidovudine resistance also increase in vitro susceptibility to tenofovir but not stavudine. Antirvir Ther 2003; 8:S Kemp SD, Shi C, Bloor S, Harrigan PR, Mellors JW, Larder BA. A novel polymorphism at codon 333 of human immunodeficiency virus type 1 reverse transcriptase can facilitate dual resistance to zidovudine and L-2,3 -dideoxy-3 -thiacytidine. J Virol 1998; 72: Hertogs K, Bloor S, De Vroey V, et al. A novel human immunodefi- Broad NRTI Cross-Resistance in HIV-1 JID 2003:188 (1 October) 999

9 ciency virus type 1 reverse transcriptase mutational pattern confers phenotypic lamivudine resistance in the absence of mutation 184V. Antimicrob Agents Chemother 2000; 44: Romano L, Venturi G, Bloor S, et al. Broad nucleoside-analogueresistance implications for human immunodeficiency virus type 1 reverse-transcriptase mutations at codons 44 and 118. J Infect Dis 2002; 185: Sturmer M, Staszewski S, Doerr H-W, Larder B, Bloor S, Hertogs K. Correlation of phenotypic zidovudine resistance with mutational patterns in the reverse transcriptase of human immunodeficiency virus type 1: interpretation of established mutations and characterization of new polymorphisms at codons 208, 211, and 214. Antimicrob Agents Chemother 2003; 47: Stoeckli TC, MaWhinney S, Uy J, et al. Phenotypic and genotypic analysis of biologically cloned human immunodeficiency virus type 1 isolates from patients treated with zidovudine and lamivudine. Antimicrob Agents Chemother 2002; 46: JID 2003:188 (1 October) Whitcomb et al.