Skip Navigation

Improved Interpretation of Genotypic Changes in the HIV-1 Reverse Transcriptase Coding Region That Determine the Virological Response to Didanosine

  1. Dr. Andrea De Luca1,
  2. Simona Di Giambenedetto1,
  3. Maria Paola Trotta2,
  4. Manuela Colafigli1,
  5. Mattia Prosperi3,
  6. Lidia Ruiz4,
  7. John Baxter5,
  8. Philippe Clevenbergh6,
  9. Roberto Cauda1,
  10. Carlo-Federico Perno2 and
  11. Andrea Antinori2
  1. 1 Institute of Clinical Infectious Diseases, Catholic University, Rome, Italy
  2. 2 National Institute of Infectious Diseases, “Lazzaro Spallanzani,”, Rome, Italy
  3. 3 Department of Computer Science and Automation, University Roma Tre, Rome, Italy
  4. 4 Hospital Universitari Germans Trias I Pujol, Badalona, Spain
  5. 5 Cooper University Hospital/University of Medicine and Dentistry of New Jersey-Robert Wood Johnson Medical School, Camden, New Jersey
  6. 6 Maladies Infectieuses Service de Medecine Interne A, Hopital Lariboisiere, Paris, France
  1. Reprints or correspondence: Dr. Andrea De Luca Istituto di Clinica delle Malattie Infettive Policlinico Universitario “Agostino Gemelli” Universita Cattolica del Sacro Cuore Largo F. Vito 1 00168 Roma Italy (andrea.deluca{at}rm.unicatt.it).

  1. Presented in part: 3rd European HIV Drug Resistance Workshop, Athens, 30 March-1 April 2005 (abstract 9.15).

 
Next Section

Abstract

Background. Consensus on the interpretation of mutations in the human immunodeficiency virus (HIV)-1 reverse transcriptase (RT) gene that predict the response to didanosine treatment is needed.

Methods. Baseline HIV-1 RT genotypes and 12-week virological outcomes for patients undergoing didanosine-containing salvage regimens were extracted from prospective studies. Existing didanosine genotypic-resistance interpretation rules were validated in the entire-patient data set. Mutations were given weighted positive or negative scores according to their coefficient of correlation with virological response in a derivation set. The score resulting from the algebraic sum of the mutations was then validated in an independent data set.

Results. A total of 485 patients were analyzed. The didanosine-resistance scores derived from the Jaguar and Gesca studies predicted virological outcome. The best correlation with response was found with the derived score (M41L × 2) + E44D/A/G + T69D/S/N/A + (L210W × 2) + T215Y or revertants + L228H/R — D123E/N/G/S, by use of which viruses were categorized as being susceptible (score ⩽0), as having intermediate resistance (1–3), and as being resistant (⩾4) to didanosine. In the validation set, the adjusted mean difference in 12-week virological response was +0.34 log10 copies/mL (95% confidence interval, +0.11 to +0.57; P = .004) per higher resistance category. Correlation with virological response constantly outperformed that obtained with the previous interpretation.

Conclusion. The improved genotypic-resistance interpretation score can be applied to better guide the use of didanosine in treatment-experienced individuals.

The control of HIV-1 replication is the major determinant of the clinical efficacy of highly active antiretroviral therapy (HAART) [1]. Resistance to antiretroviral drugs is both cause and consequence of the failure to suppress plasma viral load and can be associated with clinical progression [2–4]. Results from prospective studies showing a small but beneficial impact of using HIV-1 drug-resistance assays to guide treatment choice in patients experiencing HAART failure has led to the universal recommendation of using resistance testing in patients with therapeutic failure [3, 5–12].

Nevertheless, a major difficulty with resistance genotyping assays is the interpretation of results. Evaluation of the predictive value of virological response by the interpretation of genotypic resistance to individual drugs has shown major discordance and less predictive value in the interpretation of genotypic resistance to abacavir and didanosine [13, 14]. For example, the contribution of the reverse transcriptase (RT) M184V mutation, typically selected by lamivudine and emtricitabine, to didanosine resistance is unclear, because it confers reduced phenotypic susceptibility in vitro [15] but shows no influence on—and sometimes association with—a favorable virological outcome to didanosine therapy in vivo [16, 17]. Recently, 2 algebraic mutation scores have been derived by correlation analyses between viral RT genotypes and treatment response, one from the Jaguar study, a didanosine 4-week add-on clinical study [16,18], and a second from the Gesca study [19]. These scores, though, have either been cross-validated on the same data set used for their derivation by use of resampling methods [16, 20] or did not undergo cross-validation at all. In the present study, we undertook validation of existing didanosine genotypic-resistance interpretation rules and developed a new didanosine score from a large set of patients experiencing treatment failure who were starting didanosine-based rescue therapy.

Previous SectionNext Section

Patients, Materials, and Methods

Study patients. HIV-1-infected patients with previous failure of HAART (defined as the concomitant use of ⩾3 antiretrovirals) undergoing didanosine-containing salvage regimens were extracted from those enrolled in 4 randomized studies (ARGENTA, VIRADAPT, GART, and Havana) [5–8] and a prospective cohort study (the INMI cohort) [4]. Only patients with baseline genotype and viral load determined <8 weeks before starting the didanosine-based regimen, with follow-up viral load determined at 12 weeks (allowing a range between 8 and 16 weeks), and with no changes in the regimen up to the virological follow-up date were selected. HIV-1 RNA values at baseline and at 12 weeks as well as the previous and the new treatment regimens were retrieved. All patients provided written informed consent for participation in the respective studies.

Genotypic-resistance testing and interpretation. Genotypic-resistance assays were those used in the different studies. Because different assays were used, the RT sequence varied in length; therefore, the minimum available sequence encompassing codon 20–238 was analyzed. The genotypic mutations from consensus B HIV-1 were uploaded via the Stanford database for the VIRADAPT, GART, and Havana studies (courtesy of R. Shafer, Stanford University); the ARGENTA data were already available in house in FASTA format, and the INMI cohort data were obtained through strings of amino acid substitutions from consensus B. Sequences containing a mixture of wild-type and mutant residues at single positions were considered to have a mutation at that codon.

Didanosine resistance was interpreted using the rules from the Agence Nationale de Recherche de la SIDA (ANRS) AC11 algorithm (version 14; available at: http://www.hivfrenchresistance.org), those from the Stanford HIVdb (version 4.2.1; available at http://hivdb.stanford.edu), those from the Jaguar algebraic score [16], and those from the Gesca score [19]. Resistance to the other drugs used in the therapy (backbone therapy) was interpreted using both the Stanford and the ANRS AC11 algorithm.

For each drug, susceptibility (including Stanford's potential low level resistance) was scored as 1, partial resistance (ANRS AC11's possible resistance and Stanford's low-level resistance and intermediate resistance) as 0.5, and resistance as 0. The Jaguar and Gesca algebraic scores were analyzed as continuous variables (table 1). For the backbone therapy, a genotypic susceptibility score (GSS) was calculated by summing the scores for the individual drugs used in the salvage regimen, as reported elsewhere [13].

Derivation and validation of the new didanosine score. The database was randomly split 2:1 into a derivation and a validation set, following an automated procedure included in the SPSS statistical software package (version 13.0; SPSS). In the derivation set, RT substitutions showing an association with a significance of P ⩽ .2 with respect to the virological response (magnitude of change in 12-week HIV-1 RNA level from baseline) by the Mann-Whitney U test were given different combinations of weighted or unweighted positive or negative scores according to the correlation coefficient found by linear regression. Different scores resulting from the algebraic sum of the weighted mutations were then derived, and that with the best linear correlation with virological response was selected. The best-fitting score was then translated in categorized levels of resistance. The robustness ofthe correlation between the score and virological response was finally tested on the validation set and compared with that of the Jaguar mutation score and of the Gesca score.

Other statistical analyses. HIV-1 RNA values were log10transformed before calculations. Whether various scores and variables were associated with the magnitude of the virological response was analyzed by use of simple and multiple linear regression models. Association between variables and categorical outcomes (e.g., reaching HIV-1 RNA levels <500 copies/mL at week 12) was analyzed by univariable and multivariable logistic regression. All analyses were performed using SPSS (version 13.0) and Statistica (version 6.0; Statsoft) software packages.

Previous SectionNext Section

Results

Baseline characteristics, treatment regimens, and virological responses of patients. We analyzed 485 patients; 315 were included in the derivation set and 170 in the validation set. Fifty-one patients (10.5%) were from ARGENTA, 25 (5.2%) from VIRADAPT, 63 (13.0%) from GART, 105 (21.6%) from Havana, and 241 (49.7%) from the INMI cohort. Patients had experienced a median of 7 antiretroviral drugs (interquartile range, 5–11 antiretroviral drugs), including didanosine in 76% of patients; 3% had experienced 1 class, 55% had experienced 2 classes, and 42% had experienced 3 classes of antiretrovirals.

At baseline, the mean HIV-1 RNA level was 4.32 log10 copies/mL (SD, 0.74 log10 copies/mL), with no difference between the derivation and the validation set. Apart from didanosine, the rescue regimens contained the following drugs: zidovudine in 6.6%, lamivudine in 14.6%, stavudine in 59.0%, abacavir in 21.4%, nevirapine in 19.4%, efavirenz in 28.7%, nelfinavir in 24.5%, ritonavir at therapeutic doses in 25.2%, indinavir in 16.5%, saquinavir in 14.4%, lopinavir in 11.3%, and any ritonavir-boosted protease inhibitor apart from lopinavir in 22.7%. No significant differences in drug use were observed between the derivation and validation set. The mean ± SD GSS for the backbone therapy was 1.48 ± 0.86 and 1.50 ± 0.97 by the ANRS AC11 and Stanford interpretations, respectively (P = .27, paired t test), with no differences between the derivation and validation sets.

At 12 weeks, the mean change from baseline in HIV-1 RNA level was -1.06 log10 copies/mL (SD, 1.18 log10 copies/mL), and 45.4% of patients achieved a viral load <500 copies/mL. Again, no significant differences between the derivation and validation sets were observed.

Baseline HIV-1 RT genotyping results. The frequency of individual RT amino acid substitutions (relative to consensus B) present in ⩾5% of patients is illustrated in figure 1. The most frequently mutated nucleoside RT inhibitor (NRTI) resistance-associated positions according to the International AIDS Society—USA list [22] are illustrated in figure 1A, whereas figure 1B shows the mutation frequency at all of the other codons. The frequency of other RT mutations present in <5% of the sample but potentially relevant in terms of didanosine resistance was as follows: K65R (1%), insertion in 69 (0.2%), L74V/I (3.5%), F116Y (1.9%), and Q151M (2.7%).

Figure 1.
View larger version:
Figure 1.

Prevalence of HIV-1 isolates with substitutions in reverse transcriptase (relative to consensus B virus) in patients used in the derivation set (black bars, n = 315) and those used in the validation set (white bars, n = 170). With the exception of 215 revertant mutations (rev; amino acids C/D/E/G/I/N/S/V), only substitutions with a prevalence of >5% are shown. A, Mutations from the International AIDS Society-USA list [21]. B, Other substitutions from consensus B. Nos. with letters indicate reverse transcriptase codons with respective substituted amino acid residues. Unspecified residues indicate all nonsynonymous substitutions from consensus B.

With the exception of A98G/S, which was significantly more represented in the validation set than in the derivation set (P = .027), no other difference in the frequency of an RT mutation was observed between the different sets.

Validation of existing didanosine genotypic-resistance interpretation rules. We first evaluated the predictors of virological outcomes in the enitre set of 485 patients. Crude analysis of the association between variables and (1) the 12-week change from baseline HIV-1 RNA level and (2) HIV-1 RNA level <500 copies/mL is summarized in table 2. Higher baseline viral load and the use of nevirapine or nelfinavir in the salvage regimen predicted worse virological responses, whereas the use of efavirenz, lopinavir, and a higher GSS in the backbone therapy calculated using either the Stanford or the ANRS AC11 interpretation were significantly associated with better virological outcomes. Didanosine-susceptibility interpretation using the Stanford and ANRS AC11 rules did not predict a change in viral load, although the ANRS AC11 interpretation was associated with the odds of reaching a HIV-1 RNA level <500 copies/mL. The scores derived from the Jaguar study (median, +1; range, -2 to +4) and from the Gesca study (median, 0; range, -2 to + 3) were significantly related to the HIV-1 RNA response. In a multiple linear regression model adjusting for baseline HIV-1 RNA level, GSS of backbone therapy (ANRS AC11 interpretation), and nevirapine, efavirenz, nelfinavir, and lopinavir use, the Jaguar score independently predicted the 12-week virological response (mean HIV-1 RNA change per higher point in score, +0.10 log10 copies/mL [95% confidence interval {CI}, + 0.02 to +0.18]; P = .021). After adjustment for the same variables, the Gesca score was also independently associated with the virological response (mean HIV-1 RNA change per higher point in score, +0.11 log10 copies/mL [95% CI, +0.04 to +0.18]; P = .003). In the multivariable logistic regression adjusting for the covariates listed above and including study cohort, the Jaguar score and the Gesca score were independently predictive of reaching HIV-1 RNA levels <500 copies/mL at week 12 (adjusted OR per higher point in Jaguar score, 0.75 [95% CI, 0.64 to 0.88]; P <.001) (adjusted OR per higher point in Gesca score, 0.78 [95% CI, 0.68 to 0.90]; P = .001).

Figure 2.
View larger version:
Figure 2.

Virological response, expressed as the magnitude of plasma viral load inhibition at 12 weeks, according to the didanosine-resistance score in the derivation set (A; n = 315) and in the validation set (B; n = 170). Central lines indicate medians, boxes indicate upper and lower quartiles, and whiskers indicate 5th and 95th percentiles.

Table 1.
View larger version:
Table 1.

Available rules of interpretation for the evaluation of genotypic resistance to didanosine.

Table 2.
View larger version:
Table 2.

Predictors of 12-week virological outcomes: univariable analyses of the entire set of patients (n= 485).

Derivation of a new didanosine genotypic-resistance interpretation score from virological outcomes. In the randomly selected derivation set of patients (n = 315), the following mutations present in >1% of the sample were associated, at the P ⩽ .2 level, with a reduced 12-week change from baseline viral load: M41L, E44D/A/G, T69D/S/N/A, L210W, T215Y or revertants, and L228H/R (table 3). D123E/N/G/S was, on the other hand, associated with improved virological responses. All other analyzed residues did not correlate with the magnitude of the virological response; these included K65R (P = .79), K70R (P = .59), L74V/I (P = .56), M184V/I (P = .29), T215F (P = .87), and K219Q/E (P = .83). These mutations were confirmed by logistic regression using the percentage of patients with a week12 viral load <500 copies/mL as outcome (data not shown). Mutations associated with response were used for a stepwise construction ofdifferent scores, first using simple sums of the mutations associated with a worse response (score 1 in table 4), then weighting the mutations with the strongest association on the basis of their correlation coefficient (score 2 in table 4), and finally using algebraic sums with negative values for the mutations associated with better virological outcomes (score 3 in table 4). The score showing the strongest association with the 12-week viral load change was (M41L × 2) + E44D/A/G + T69D/S/N/A + (L210W × 2) + T215Y or revertants + L228H/R — D123E/N/G/S. This was then used to establish categorical classifications: the classification showing the strongest correlation with the 12-week virological response was one where viruses with scores ⩽0 were categorized as being susceptible, those with scores between 1 and 3 were categorized as having intermediate resistance, and those with scores ⩾4 were categorized as being resistant (table 4). Subjects infected with a virus interpreted as being susceptible to didanosine exhibited a mean change from baseline HIV-1 RNA level of -1.29 log10 copies/mL (55% <500 copies/mL), subjects with a virus categorized as having intermediate resistance showed a mean change of -1.00 log10 copies/mL (48% <500 copies/mL), and individuals with a virus categorized as being resistant exhibited a mean change of -0.72 log10 copies/mL (26% <500 copies/mL) (figure 2A).

Table 3.
View larger version:
Table 3.

Substitutions in HIV-1 reverse transcriptase (RT) associated (P ⩽ .2 by Mann-Whitney U test) with the magnitude of the 12-week change from the baseline viral load.

Table 4.
View larger version:
Table 4.

Construction of different didanosine scores using HIV-1 reverse transcriptase mutations associated with virological outcomes and their correlation with the 12-week virological response in the derivation set.

Validation of the didanosine-resistance interpretation score and comparison with existing rules. The categorical didanosine score derived from the validation set was then validated on the independent data set (n = 170). There was a significant correlation between an increase in resistance category and viral load changeat12weeks (R = 0.28; P <.0001). Subjects with avirus showing predicted susceptibility to didanosine exhibited a mean change from baseline HIV-1 RNA level of -1.38 log10 copies/mL (64% <500 copies/mL), subjects with a virus showing predicted intermediate resistance exhibited a mean change of -0.99 log10copies/mL (39% <500 copies/mL), and subjects with a virus showing predicted resistance exhibited a mean change of -0.63 log10 copies/mL (17% <500 copies/mL) (figure 2B).

In multiple linear regression analysis (adjusting for baseline viral load; the GSS of backbone therapy [ANRS AC11 interpretation]; nevirapine, efavirenz, nelfinavir, and lopinavir use; and study cohort), the score was independently predictive of virological response (mean change in 12-week HIV-1 RNA level from baseline for each increase in resistance category, +0.34 log10copies/mL [95% CI, +0.11 to +0.57]; P = .004). When the Jaguar score was included in the model, it was not associated with virological response (P = .53), whereas the new didanosine categorical score remained independently predictive (mean change in HIV-1 RNA level from baseline per increase in resistance category, + 0.42 log10 copies/mL [95% CI, +0.08 to +0.76]; P = .016). When the Jaguar score was replaced by the Gesca score in the same model, neither showed independent prediction of the viral load outcome, although the new didanosine score showed a stronger correlation than the Gesca score (R = 0.17 and P = .22 vs. R = 0.09 and P = .45). In a multi-variable logistic regression model adjusting for the same variables, the new didanosine score was independently predictive of HIV-1 RNA level <500 copies/mL (table 5). When the Gesca score was included in the same model, the new categorized didanosine score was independently predictive of HIV-1 RNA level <500 copies/mL at 12 weeks (adjusted OR, 0.41 [95% CI, 0.17 to 0.99]; P = .048), whereas the Gesca score failed to predict the outcome (P = .72). Similarly, when the Jaguar score was included instead, the new didanosine score remained independently predictive of HIV-1 RNA level <500 copies/mL (adjusted OR, 0.41 [95% CI, 0.18 to 0.94]; P = .036), whereas the Jaguar score was not (P = .71).

Table 5.
View larger version:
Table 5.

Multivariable logistic regression model showing the independent association between the new didanosine score and the 12-week virological response in the validation data set (n = 170).

Previous SectionNext Section

Discussion

The HIV-1 RT genotypic determinants of didanosine resistance are still incompletely defined. In the present study, we have shown that the scores from the Jaguar and the Gesca study were independently predictive of virological response in a large set of almost 500 patients from different clinical settings undergoing a didanosine-containing regimen after experiencing HAART failure. When both scores were included in the same multivariable model, neither seemed to predict the outcomes independently from the other.

Subsequently, we developed a new algebraic didanosine-resistance score, one that showed better predictive value for the virological response than the Gesca and Jaguar scores, as reflected by its higher correlation coefficient for continuous and dichotomous virological outcomes as well as by its independent predictive value when combined with the Jaguar or Gesca scores in the same multivariable models.

There were several notable differences in the developed score compared with the previous ones. First, RT residues M41L and L210W, 2 typical type 1 thymidine analogue (TA) mutations (TAMs), received a higher resistance weight because of a stronger correlation with reduced virological response. This result further confirms that this cluster of mutations is typically associated with more extensive cross-resistance to NRTIs. Very similarly, the presence of at least 3 mutations including either M41L or L210W was associated with a reduced response to tenofovir, another adenine analogue [22]; moreover, type 1 TAMs were shown to be the strongest predictor of a reduced response to didanosine in the Gesca study [19]. We have previously shown that L210W is the most characteristic marker of type 1 TAMs [23]. This substitution often appears as the final one in a specific resistance pathway [24] and is very rarely seen in combination with K70R [23]. The T215 revertants played a role in conferring reduced virological response to didanosine. This was, in part, observed in a previous study [19] and shows that 215 revertants, being typically derived from the T215Y residue, should be considered mutations that confer resistance not only to thimidine analogues [25] but also to didanosine and, potentially, to other NRTIs. Second, the type 2 TAMs K70R, T215F, and K219Q/E did not appear to confer any negative or positive effect on virological responses. Previous studies have found that K70R and K219Qare associated with better virological responses [16, 19]. It is possible that in previous studies these mutations reflected a negative counterpart of type 1 TAMs [23] or viruses with lower replicative fitness [26] that may be associated with more favorable virological outcomes in the short term only. Our observational study was based on a larger set of patients and analyzed longer-term effects. Probably because of sample size limitation, the Jaguar and Gesca studies found both residues F and Y at codon 215 to be associated with reduced response to didanosine. In contrast—and in agreement with its strong association with type 2 TAMs [23, 27]—we did not find that residue F had any influence on response.

Contrary to the present findings, a previous study showed a consistent association between M184V/I mutations and better response to didanosine [16]. This discrepancy might be explained by differences in the viral populations, treatment regimens used, and duration of follow-up. The Jaguar study had a very limited follow-up (4 weeks), and an initial beneficial effect, which might depend on the lower viral fitness associated with M184V, might be lost over a longer follow-up period [28]. In exploratory analyses, we tried to investigate whether the effect of M184V/I could be observed in specific subsets of patients; however, we did not find any association in subsets of patients concomitantly treated with lamivudine or in those undergoing functional didanosine monotherapy (GSS of backbone therapy, 0), and an association with an improved response was found in only one of the studies included in the present analysis (ARGENTA) (data not shown). Given the lack of confirmation of the positive role played by M184V/I and considering the evidence from previous studies [15,17], it seems reasonable to judge substitutions at this position as neutral with regard to the therapeutic efficacy of didanosine.

In contrast with previous observations [16, 29], L74V/I was not related to a decreased response to didanosine. The prevalence of this mutation was lower than that observed in previous studies (3.5% compared, for example, to 10% in the didanosine arm of the Jaguar study) [16]; thus, we hypothesize that its effect may have not been observed because of insufficient statistical power or that it could have been masked by concomitant treatments or by the stronger effect of type 1 TAMs.

Finally, we observed an influence on virological response of polymorphisms at 2 RT codons: L228H/R, associated with worse virological responses, and D123E/N/G/S, associated with better virological responses. The L228H/R residues are more frequent in viruses from NRTI-experienced individuals than in those from antiretroviral-naive subjects [30, 31]. In line with this observation, an analysis of RT polymorphisms in the Jaguar study was recently published indicating an association between changes at codon 211 and 228 and a reduced response to didanosine [32]. Although the role of the polymorphism at codon 211 was not confirmed in our study, the consistent association between mutations at position 228 and response to didanosine in independent data sets underscores its clinical importance. The polymorphisms at RT codon 123 are very common in both treatment-naive and -experienced patients [31]; the association between these polymorphisms and improved virological response found in the present study might be due to a negative effect on viral fitness or to a resensitization/hypersusceptibility effect with regard to didanosine. Testing these hypotheses-as well as the relative contribution to drug susceptibility of these mutations—requires specific in vitro site-directed mutagenesis studies.

A potential limitation of the present study is related to the fact that most of the patients concomitantly received other new drugs in association, such that the pure effect of didanosine on the virological response was confounded. Nonetheless, in several multivariable models, we adjusted the interpretation of didanosine activity for all variables found to be potential confounders of the virological response. Furthermore, most of the patients studied received the buffered formulation of didanosine, because they were enrolled before the enteric-coated drug became available. Although some in the cohorts most recently recruiting patients may have received the newer formulation, this information was not available for the study. In any case, in vivo studies have shown bioequivalence between the 2 formulations, and effects in this study were adjusted for study cohort [33]. Another weakness of the present study is that the data set is relatively old, given the rapid evolution of the use of drugs. Notwithstanding, TAMs are still the most frequently detected NRTI-resistance mutations in patients experiencing treatment failure [34]. The situation may be particularly frequent and of relevance for resource-limited settings in which first-line regimens always contain TAs and didanosine is a recommended second-line option [35]. Nonetheless, because the analysis was predominantly based on subtype B viruses (only 4 [<1%] of 485 viruses belonged to subtypes other than B), these findings should be applied only to patients carrying subtype B viruses.

In conclusion, we have developed and validated a refined weighted algebraic score that consistently outperformed the previous scores in predicting virological response in an independent set of patients. The present study underscores the relevance of analyzing therapeutic-outcome data sets for defining the role played by viral mutations in influencing the therapeutic response and the importance of selecting independent data sets for the validation of developed drug-resistance interpretation rules [36, 37]. The refined interpretation of viral mutations influencing the response to didanosine-containing regimens may help to guide a more rational use of this drug in patients previously experiencing the failure of NRTI-containing ART.

Previous SectionNext Section

Acknowledgment

We are grateful to Alessandro Cozzi-Lepri for his competent statistical advice.

Previous SectionNext Section

Footnotes

  • Potential conflicts of interest: A.D.L. has been a member of advisory boards or has received speakers' honoraria from GlaxoSmithKline (GSK), Abbott Virology, Boehringer Ingelheim (BI), and Bayer Health Care Diagnostics; R.C. has been a member of advisory boards or has received speakers' honoraria from GSK, Gilead, Abbott Virology, BI, and Merck; C.F.P. has been a recipient of grants from or has performed ad hoc consultations for GSK, Bristol-Myers Squibb (BMS), Gilead, Roche, Abbott Virology, Merck, Janssen, and BI. A.A. has been a member of advisory boards or has received speakers' honoraria from GSK, BMS, Gilead, Abbott Virology, Roche, and BI.

  • Financial support: V and VI Programma Nazionale AIDS, Istituto Superiore di Sanita, Ministero della Sanita, Rome, Italy (grants 30F.17, 30F.18, and 30G.8 to A.D.L.).

  • Received February 21, 2007.
  • Accepted May 21, 2007.
Previous Section
 

References

  1. 1.↵
    1. Demeter LM,
    2. Hughes MD,
    3. Coombs,
    4. et al
    . Predictors of virologic and clinical outcomes in HIV-1-infected patients receiving concurrent treatment with indinavir, zidovudine and lamivudine. AIDS Clinical Trials Group Protocol 320. Ann Intern Med 2001;135:954-64.
    Abstract/FREE Full Text
  2. 2.↵
    1. Parkin NT,
    2. Deeks SG,
    3. Wrin MT,
    4. et al
    . Loss of antiretroviral drug susceptibility at low viral load during early virological failure in treatment-experienced patients. AIDS 2000;14:2877-87.
    CrossRefMedlineWeb of Science
  3. 3.↵
    1. Hirsch MS,
    2. Brun-Vezinet F,
    3. Clotet B,
    4. et al
    . Antiretroviral drug resistance testing in adults infected with human immunodeficiency virus type 1: 2003 recommendations of an International AIDS Society-USA panel. Clin Infect Dis 2003;37:113-28.
    Abstract/FREE Full Text
  4. 4.↵
    1. Zaccarelli M,
    2. Tozzi V,
    3. Lorenzini P,
    4. et al
    . Multiple drug class-wide resistance associated with poorer survival after treatment failure in a cohort of HIV-infected patients. AIDS 2005;19:1081-9.
    MedlineWeb of Science
  5. 5.↵
    1. Durant J,
    2. Clevenbergh P,
    3. Halfon P,
    4. et al
    . Drug-resistance genotyping in HIV-1 therapy: the VIRADAPT randomised controlled study. Lancet 1999;353:2195-9.
    CrossRefMedlineWeb of Science
  6. 6.
    1. Baxter JD,
    2. Mayers BL,
    3. Wentworth DN,
    4. et al
    . A randomized study of antiretroviral management based on plasma genotypic resistance testing in patients failing therapy. AIDS 2000;14:F83-93.
    CrossRefMedlineWeb of Science
  7. 7.
    1. Tural C,
    2. Ruiz L,
    3. Holtzer C,
    4. et al
    . Clinical utility of HIV-1 genotyping and expert advice: the Havana trial. AIDS 2002;16:209-18.
    CrossRefMedlineWeb of Science
  8. 8.↵
    1. Cingolani A,
    2. Antinori A,
    3. Rizzo MG,
    4. et al
    . Usefulness of monitoring HIV drug resistance and adherence in individuals failing highly active anti-retroviral therapy: a randomized study (ARGENTA). AIDS 2002;16:369-79.
    CrossRefMedlineWeb of Science
  9. 9.
    1. Cohen CJ,
    2. Hunt S,
    3. Sension M,
    4. et al
    . A randomized trial assessing the impact of phenotypic resistance testing on antiretroviral therapy. AIDS 2002;16:579-88.
    CrossRefMedlineWeb of Science
  10. 10.
    1. Meynard JL,
    2. Vray M,
    3. Morand-Joubert L,
    4. et al
    . Phenotypic or genotypic resistance testing for choosing antiretroviral therapy after treatment failure: a randomized trial. AIDS 2002;16:727-36.
    CrossRefMedlineWeb of Science
  11. 11.
    1. Vandamme AM,
    2. Sonnerborg A,
    3. Ait-Khaled M,
    4. et al
    . Updated European recommendations for the clinical use of HIV drug resistance testing. Antivir Ther 2004;9:829-48.
    MedlineWeb of Science
  12. 12.↵
    Panel on Clinical Practices for Treatment of HIV Infection—Department of Health and Human Services. Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. 2006 Oct 10 [20 October 2007]. (Available at: http://AIDSinfo.nih.gov).
  13. 13.↵
    1. De Luca A,
    2. Cingolani A,
    3. Di Giambenedetto S,
    4. et al
    . Variable prediction of antiretroviral treatment outcome by different systems for interpreting genotypic human immunodeficiency virus type 1 drug resistance. J Infect Dis 2003;187:1934-43.
    Abstract/FREE Full Text
  14. 14.↵
    1. Ravela J,
    2. Betts BJ,
    3. Brun-Vezinet F,
    4. et al
    . HIV-1 protease and reverse transcriptase mutation patterns responsible for discordances between genotypic drug resistance interpretation algorithms. J Acquir Immune Defic Syndr 2003;33:8-14.
    MedlineWeb of Science
  15. 15.↵
    1. Gu Z,
    2. Gao Q,
    3. Li X,
    4. Parniak MA,
    5. Wainberg MA
    . Novel mutation in the human immunodeficiency virus type 1 reverse transcriptase gene that encodes cross-resistance to 2′,3′-dideoxyinosine and 2',3'-dideoxycytidine. J Virol 1992;66:7128-35.
    Abstract/FREE Full Text
  16. 16.↵
    1. Marcelin AG,
    2. Flandre P,
    3. Pavie J,
    4. et al
    . Clinically relevant genotype interpretation of resistance to didanosine. Antimicrob Agents Chemother 2005;49:1739-44.
    Abstract/FREE Full Text
  17. 17.↵
    1. Sproat M,
    2. Pozniak AL,
    3. Peeters M,
    4. et al
    . The influence of the M184V mutation in HIV-1 reverse transcriptase on the virological outcome of highly active antiretroviral therapy regimens with or without didanosine. Antivir Ther 2005;10:357-61.
    MedlineWeb of Science
  18. 18.↵
    1. Molina JM,
    2. Marcelin AG,
    3. Pavie J,
    4. et al
    . Didanosine in HIV-1-infected patients experiencing failure of antiretroviral therapy: a randomized placebo-controlled trial. J Infect Dis 2005;191:840-7.
    Abstract/FREE Full Text
  19. 19.↵
    1. Capdepont S,
    2. Aurillac-Lavignolle V,
    3. Faure M,
    4. et al
    . An additive/subtractive genotypic score as a determinant of the virological response to didanosine in HIV-1 infected patients. J Clin Virol 2006;36:36-42.
    CrossRefMedlineWeb of Science
  20. 20.↵
    1. Flandre P,
    2. Marcelin AG,
    3. Pavie J,
    4. et al
    . Comparison of tests and procedures to build clinically relevant genotypic scores: application to the Jaguar study. Antivir Ther 2005;10:479-87.
    MedlineWeb of Science
  21. 21.↵
    1. Johnson VA,
    2. Brun-Vezinet F,
    3. Clotet B,
    4. et al
    . Update of the drug resistance mutations in HIV-1: fall 2005. Top HIV Med 2005;13:125-31.
    Medline
  22. 22.↵
    1. Miller MD,
    2. Margot N,
    3. Lu B,
    4. et al
    . Genotypic and phenotypic predictors of the magnitude of response to tenofovir disoproxil fumarate treatment in antiretroviral-experienced patients. J Infect Dis 2004;189:837-46.
    Abstract/FREE Full Text
  23. 23.↵
    1. De Luca A,
    2. Di Giambenedetto S,
    3. Romano L,
    4. et al
    . Frequency and treatment-related predictors of thymidine-analogue mutation patterns in HIV-1 isolates after unsuccessful antiretroviral therapy. J Infect Dis 2006;193:1219-22.
    Abstract/FREE Full Text
  24. 24.↵
    1. Harrigan PR,
    2. Kinghorn I,
    3. Bloor S,
    4. et al
    . Significance ofamino acid variation at human immunodeficiency virus type 1 reverse transcriptase residue 210 for zidovudine susceptibility. J Virol 1996;70:5930-4.
    Abstract/FREE Full Text
  25. 25.↵
    1. Violin M,
    2. Cozzi-Lepri A,
    3. Velleca R,
    4. et al
    . Risk of failure in patients with 215 HIV-1 revertants starting their first thymidine analog-containing highly active antiretroviral therapy. AIDS 2004;18:227-35.
    CrossRefMedlineWeb of Science
  26. 26.↵
    1. Hu Z,
    2. Giguel F,
    3. Hatano H,
    4. Reid P,
    5. Lu J,
    6. Kuritzkes DR
    . Fitness comparison of thymidine analog resistance pathways in human immunodeficiency virus type 1. J Virol 2006;80:7020-7.
    Abstract/FREE Full Text
  27. 27.↵
    1. Marcelin AG,
    2. Delaugerre C,
    3. Wirden M,
    4. et al
    . Thymidine analogue reverse transcriptase inhibitors resistance mutations profiles and association to other nucleoside reverse transcriptase inhibitors resistance mutations observed in the context of virological failure. J Med Virol 2004;72:162-5.
    CrossRefMedlineWeb of Science
  28. 28.↵
    1. Castagna A,
    2. Danise A,
    3. Menzo S,
    4. et al
    . Lamivudine monotherapyin HIV-1-infected patients harbouring a lamivudine-resistant virus: a randomized pilot study (E-184V study). AIDS 2006;20:795-803.
    MedlineWeb of Science
  29. 29.↵
    1. St Clair MH,
    2. Martin JL,
    3. Tudor-Williams G.
    Resistance to ddI and sensitivity to AZT induced by a mutation in HIV-1 reverse transcriptase. Science 1991;253:1557-9.
    Abstract/FREE Full Text
  30. 30.↵
    1. Gonzales MJ,
    2. Wu TD,
    3. Taylor J,
    4. et al
    . Extended spectrum of HIV-1 reverse transcriptase mutations in patients receiving multiple nucleoside analog inhibitors. AIDS 2003;17:791-9.
    CrossRefMedlineWeb of Science
  31. 31.↵
    1. Ceccherini-Silberstein F,
    2. Gago F,
    3. Santoro M,
    4. et al
    . High sequence conservation of human immunodeficiency virus type 1 reverse transcriptase under drug pressure despite the continuous appearance of mutations. J Virol 2005;79:10718-29.
    Abstract/FREE Full Text
  32. 32.↵
    1. Marcelin AG,
    2. Flandre P,
    3. Furco A,
    4. Wirden M,
    5. Molina JM,
    6. Calvez V
    . Impact of HIV-1 reverse transcriptase polymorphism at codons 211 and 228 on virological response to didanosine. Antivir Ther 2006;11:693-9.
    MedlineWeb of Science
  33. 33.↵
    1. Damle BD,
    2. Kaul S,
    3. Behr D,
    4. Knupp C
    . Bioequivalence of two formulations of didanosine, encapsulated enteric-coated beads and buffered tablet, in healthy volunteers and HIV-infected subjects. J Clin Pharmacol 2002;42:791-7.
    Abstract
  34. 34.↵
    1. Zazzi M,
    2. Corsi P,
    3. Gonnelli A,
    4. et al
    . Evolution of the prevalence of HIV drug resistance patterns of 9 years of HAART and its relation with changes in the failing treatment regimens: an analysis of a large Italian database [poster 124]. XV International HIV Drug Resistance Workshop (Sitges, Spain). Antivir Ther 2006;11:S137-S137.
    Web of Science
  35. 35.↵
    World Health Organization. Antiretroviral therapy for HIV infection in adults and adolescents: recommendations for a public health approach, 2006 revision. (Available at: http://www.who.int/hiv/pub/guidelines/artadultguidelines.pdf.).
  36. 36.↵
    1. Baxter JD,
    2. Schapiro JM,
    3. Boucher CA,
    4. et al
    . Genotypic changes in human immunodeficiency virus type 1 protease associated with reduced susceptibility and virologic response to the protease inhibitor tipranavir. J Virol 2006;80:10794-801.
    Abstract/FREE Full Text
  37. 37.↵
    1. Costagliola D,
    2. Cozzi-Lepri A,
    3. Assuomou KL,
    4. Cheng B
    ; on behalf of the Standardization and Clinical Relevance of HIV Drug Resistance Testing Project from the Forum for Collaborative HIV Research. Initiatives for developing and comparing genotype interpretation systems: an updated analysis of validation of existing rules-based algorithm for abacavir and ddI evaluated on virologic response [abstract 84]. XV International HIV Drug Resistance Workshop (Sitges, Spain; 13–17 June 2006). Antivir Ther 2006;11:S94-S94.
    Web of Science
| Table of Contents

This Article

  1. J Infect Dis. (2007) 196 (11): 1645-1653. doi: 10.1086/522231
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 205 (5)
  1. Journal of Infectious Diseases
  1. Alert me to new issues

Most