The treatment options for patients infected with highly drug-resistant HIV markedly improved with the introduction of new antiretroviral compounds, such as fusion inhibitors, second-generation nonnucleoside reverse transcriptase inhibitors (NNRTIs), or new boosted protease inhibitors (PIs), CCR5 antagonists, and integrase inhibitors.1-7 To date, knowledge about the optimal combination of these compounds in salvage therapy is lacking. Nucleoside reverse transcriptase inhibitors (NRTIs) are often co-administered in salvage therapy, even if genotypic resistance tests (GRTs) indicate high-level resistance. A therapeutic benefit is assumed because of the possible residual activity of these NRTIs and the maintenance of a resistant virus with reduced replicative capacity.8-12 On the other hand, costs, drug-drug interactions, tolerability, and toxicity of these additional NRTIs have to be taken into account. NRTIs can cause mitochondrial dysfunction by inhibiting the DNA γ-polymerase resulting in plasma hyperlactataemia and variable clinical syndromes, such as lipoathrophy and peripheral neuropathy.13-21
The clinical benefit of NRTIs with decreased activity due to drug resistance mutations to date has not been properly assessed. The number of antiretroviral compounds has increased, and additional drug classes have become available, making NRTIs potentially expendable in salvage therapy.
Here, we focused on salvage regimens including raltegravir (RAL) because this drug is now frequently used in Switzerland to treat patients with highly resistant viruses.22 Using data from the highly representative Swiss HIV Cohort Study (SHCS),23,24 we report on the genotypic activity and composition of salvage therapies with RAL and the effect of partially active or inactive NRTIs on the viral suppression rate.
Data and Patient Selection
Data from the SHCS were included for our analysis (up to June 30, 2010). The SHCS is a nationwide clinic-based cohort study with continuous enrolment and at least semiannual study visits (www.shcs.ch).24 It has been approved by ethical committees of all participating institutions, and written informed consent has been obtained from participants. The SHCS drug resistance database contains all HIV resistance tests performed by the 4 authorized laboratories in Switzerland using commercial assays (Viroseq Vs. 1 PE Biosystems, Rotkreuz, Switzerland; Virsoseq Vs. 2, Abbott AG, Baar, Switzerland; vircoTYPE HIV-1 Assay, Virco Lab, Mechelen, Belgium) and in-house methods.25 Sequences are stored in SmartGene's (Zug, Switzerland) Integrated Database Network System version 3.5.8).26
To analyze the effect of partially active or inactive NRTIs in salvage therapy, the SHCS was screened for patients who started a regimen containing RAL. Inclusion criteria were a GRTs on antiretroviral therapy (ART) before the RAL start and baseline HIV-1 RNA >500 copies per milliliter. Patients receiving more than 2 NRTIs were excluded from the study due to the small number of cases (n = 12). For further analyses, patients receiving 0 or 1 NRTI were considered as one group and compared with patients receiving 2 NRTIs beside RAL. This classification turned out to be appropriate because patients treated with 0 and 1 NRTI had similar characteristics, and results did not differ markedly when analyzing these 2 groups separately (not shown).
Baseline Characteristics and Estimated Activity of Available Treatment Options
Patient characteristics were compared between patients receiving <2 NRTIs and 2 NRTIs with Fisher exact test (categorical variables) and Wilcoxon rank-sum test (continuous variables). The baseline was set at the date of RAL start. The self-reported adherence was categorized in 2 groups: patients who never missed a drug and patients who missed ≥1 drug in the 4 weeks preceding the study visit.27 To assess the availability of active antiretroviral compounds, results from Stanford interpretation algorithm (version 6.0.8) were mapped to a genotypic sensitivity score (GSS) for all approved drugs except enfuvirtide (T20), maraviroc (MAR), and RAL. The 5 resistance categories from the Stanford algorithm were regrouped as follows: viruses with a GSS less than 15 were considered as fully susceptible (GSS = 1), those with a GSS between 15 and 59 were considered to have intermediate resistance (GSS = 0.5), and those with a GSS greater than 59 were considered to be fully resistant (GSS = 0). If T20 and MAR have not previously been included in a failing regimen, they were considered fully susceptible because transmission of HIV with resistance to T20 is very rare and coreceptor tropism testing was always performed before MAR prescription (Trofile assay, Monogram Biosciences, San Francisco, CA).28
The effect of NRTIs in salvage therapies with RAL was assessed at week 24. The viral suppression rate (HIV-1 RNA < 50 copies/mL) was analyzed, and different approaches were implemented as follows: an intent-to-treat analysis was performed with 2 different methods dealing with missing information, last observation carried forward (LOCF) and missing equal failure (m = f), and a per-protocol analysis. For the per-protocol analysis, only patients who did not change, stop, or interrupt treatment until week 24 and who had a viral load measurement between week 18 and week 30 were included.
Logistic regressions were performed and adjusted for ethnicity, age, sex, the GSS of the treatment (without NRTIs), number of drug classes, HIV-1 RNA, and CD4 cell count before RAL treatment start. In the present study, confounding by indication must be addressed because many factors, for example, number of drug classes in the background regimen, GSS of available drugs, or adherence, may influence not only the suppression rate but also the number of NRTIs physicians chose for the salvage therapy. A solution to overcome a selection bias is to perform a marginal structural model.29,30 Weights were defined as the inverse of the probability for receiving <2 NRTIs as estimated by multivariable logistic regression including the following possible confounders: sex, adherence, age, transmission category, ethnicity, MAR, etravirine (ETV), or darunavir (DRV) in the background treatment, GSS of available NRTIs, GSS of PIs, and NNRTIs in the salvage therapy, CD4 nadir, baseline HIV-1 RNA, year of treatment, and whether the patient was ever treated with mono/dual NRTI therapy. This method creates a pseudopopulation, in which the probability for receiving <2 or 2 NRTIs is unrelated to baseline factors which are also prognostic for the treatment response. Multicollinearity was checked, and a variance inflation factor <3 was tolerated for regression models. To check whether single observations had a disproportionately large impact on our results due to the weighting, the analysis was repeated 1000 times on bootstrapped data sets.
To confirm results, an additional analysis was performed assessing time to viral suppression with a Cox regression model. The same covariables were included as in the logistic regression described above, and the same procedure was followed to calculate the weights. Patients were included when they had at least 1 HIV-1 RNA measured, and they were censored when they changed, stopped, or interrupted therapy.
Statistical analyses were performed with Stata 11 SE (StataCorp, College Station, TX), all confidence intervals (CIs) are 95% CI, and the level of significance was set at P = 0.05.
Study Population and Baseline Characteristics
A total of 142 patients who had a viral load >500 copies per milliliter, a GRT performed before RAL treatment start, and follow-up HIV-1 RNA measurements were considered for analysis. Patients who received more than 2 NRTIs were excluded from further analysis (11 with 3 NRTIS, 1 with 4 NRTIs). Patients who received no NRTI (n = 38, 26.8%) or 1 NRTI (n = 38, 26.8%) were handled as one group and compared with patients receiving 2 NRTIs (n = 54, 38.0%).
Most baseline characteristics were similar between patients with <2 NRTIs and 2 NRTIs (Table 1), but patients with 2 NRTIs were younger, had more often baseline HIV-1 RNA >100,000 copies per milliliter and tended to have started the first ART later. The self-reported adherence during the 4 weeks proceeding the study visit before RAL start was similar: 72.2% (2 NRTIs group) and 71.1% (<2 NRTIs group). Additional factors that might be a sign of nonadherence were tested as follows: the number of therapies patients started, the number of treatment interruptions (cessation of ART and resumption at a later date), psychiatric treatment in the past, alcohol abuse, and current intravenous drug use or smoking. All these characteristics were similar among groups (data not shown).
The median number of NRTI, NNRTI, and PI mutations (International AIDS Society [IAS-USA] drug resistance mutations printed in bold)31 was comparable between patients with <2 and 2 NRTIs, 9 [interquartile range (IQR): 5-13] and 10 (7-14.5) (P = 0.100), respectively. Of note, the median number of NRTI mutations was lower in the 2 NRTIs group, 4 (1-5) compared with 5 (3-5) in the <2 NRTIs group (P = 0.040). The number of major PI and NNRTI mutations was not significantly different.
Available Treatment Options Based on Genotypic Data
Some treatment options differed between patients with 2 and <2 NRTIs and were important reasons to perform a weighted logistic regression. The GSS of the best 2 NRTIs was <1, 1, and >1 in 29.6%, 20.4%, and 50.0% of patients with 2 NRTIs, slightly higher than in the <2 NRTIs group (43.4%, 26.3%, and 30.3%, P = 0.075). As shown in Table 2, the NRTI with the highest estimated activity was tenofovir (TDF), with only 3.7% (2 NRTIs group) and 11.8% (<2 NRTIs group) being fully resistant. In contrast, full resistance against the following NRTI groups was common in patients with 2 and <2 NRTIs: zidovudine/stavudine (37.0% and 59.2%), emtricitabine (FTC)/lamivudine (3TC) (81.5 and 85.5%), and abacavir (ABC)/didanosine (DDI) (37.0% and 52.6%). Resistance against the new NNRTI ETV was rare 1.9% (2 NRTIs group) and 5.3% (<2 NRTI group). In contrast, full resistance against first-line NNRTIs (EFV and NVP) was common in patients with 2 and <2 NRTIs, 63.0% and 68.4%, respectively. DRV was the best PI with an estimated full activity in 73.9% of the cases, followed by tipranavir (48.5%), lopinavir (43.9%), saquinavir (39.2%), and indinavir (38.5%).
Composition of the Salvage Therapy With RAL
The composition of the non-NRTI treatment patients received together with RAL differed markedly between groups. The number of non-NRTI drugs beside RAL was lower in the 2 NRTIs group. The percentage of patients with ≤1, 2, and 3 drugs was 68.5%, 27.8%, and 3.7% compared with 30.3%, 56.6%, and 13.2% (P < 0.001). Also the number of drug classes beside NRTIs and RAL was lower in the 2 NRTIs group [median IQR: 1 (1-2) compared with the <2 NRTI group: median (IQR): 2 (1-2), P Wilcoxon < 0.001].
As shown in Figure 1, most patients with 2 and <2 NRTIs additionally received a boosted PI, 66.7% and 80.3% (P = 0.102). Most patients with a boosted PI received DRV (74.3%). NNRTIs also were often co-administered, in 44.4% (2 NRTIs group) and 68.4% (<2 NRTIs group) of cases (P = 0.007). Patients with a NNRTI most often had ETV (88.2%). MAR and T20 were rarely administered (3.7% and 18.4%, P = 0.014; 3.7% and 6.6%, P = 0.699).
In the 2 NRTIs group, the predominant NRTI combination was 3TC/FTC and TDF (38/54, 70.4%), followed by ABC and 3TC (14.8%), ABC and TDF (11.1%), TDF and DDI, and 3TC and zidovudine (each 1.9%). Patients with 1 NRTI most often received 3TC (16/38, 42.1%) or TDF (15/38, 39.5%). Three patients received ABC (7.9%), 2 DDI (5.3%), and 1 each AZT and stavudine (2.6%).
Estimated Genotypic Activity of the Salvage Therapy
The GSS of all non-NRTI drugs in the salvage therapy was lower in the 2 NRTIs group with a median GSS of 2 (1.5-2.5) compared with 2.5 (2-3) (P < 0.001). However, when also considering the GSS of NRTIs, the overall GSS of the treatment tended to be higher in the 2 NRTIs group 3 (2.5-3.0) compared with 2.5 (2-3, P = 0.059).
The contribution to the GSS of each NRTI was similar in the <2 NRTIs and the 2 NRTIs group, the GSS was <0.5, 0.5, and >0.5 in 46.3%, 31.5%, and 22.2% compared with 47.4%, 31.6%, and 21.1% (P = 1.000) cases. Most patients (47/63, 74.6%) receiving 3TC/FTC had viral strains carrying the M184I/V mutations, in the <2 and 2 NRTIs group 81.3% and 76.6%, respectively.
The described differences in salvage therapy composition and in particular the higher number of drug classes included in the <2 NRTIs group are possibly interfering with our aim to measure the effect of partially active or inactive NRTIs. For this purpose, a marginal structural model was performed. The model creates a “pseudopopulation” in which group differences in salvage treatment composition are balanced.
With the LOCF approach, the crude percentages of patients who achieved viral suppression at week 24 were 72.2% and 71.0% for patients with 2 and <2 NRTIs, respectively. The median (IQR) week of measurement was 24 (20-27) and similar between patients with 2 and <2 NRTIs [24.1 (19.9-26.6) compared with 23.9 (20.1-27.5)]. About 2.6% (2 NRTIs group) and 5.6% (<2 NRTIs group) had no RNA measurement performed. A similar number of patients stopped, interrupted, or changed treatment before week 24, 22.2% (12 of 54) and 29.0% (22 of 76) in the 2 and <2 NRTIs group, respectively (P = 0.425). Toxicity was the reason for the change among 16.7% (2 of 12) and 36.4% (8 of 22) of the cases (P = 0.430). As shown in Figure 2, multivariable logistic regressions showed that patients treated with <2 NRTIs compared with 2 NRTIs had a decreased chance to achieve viral suppression [multivariable odds ratio (OR): 0.59, P = 0.269; weighted multivariable OR: 0.34, P = 0.027] (Table 3). The robustness of the results was tested with a bootstrap analysis (1000 replications), it yielded a similar result [mean multivariable weighted OR: 0.41 (fifth and 95th percentiles: 0.11-0.98)], suggesting that the observed differences were not hinging on a few specific observations in our dataset but were broadly consistent.
The crude percentage of patients with suppressed viral load was slightly lower when using the m = f approach (2 NRTIs: 64.8%, <2 NRTIs: 59.2%). The beneficial effect of 2 NRTIs was confirmed in multivariable (OR: 0.47, P = 0.099), and in weighted multivariable models (OR: 0.33, P = 0.027) (Fig. 2). Also the per-protocol analysis confirmed results. Eighty-three patients were included who did not change, stop, or interrupt treatment until week 24 and who had a viral load measurement performed within the given time frame (multivariable OR: 0.54, P = 0.337, weighted multivariable OR: 0.19, P = 0.023).
Different sensitivity analyses were performed to verify the results. Because the higher GSS of NRTIs in the 2 NRTIs group might partially explain the results, a subanalysis including only patients with a cumulative NRTI GSS ≤0.5 in the regimen was performed (n = 93 ITT LOCF, n = 60 per protocol). It was confirmed that additional NRTIs with low activity are beneficial for virological outcome (ITT LOCF: weighted multivariable OR: 0.13, 95% CI: 0.03 to 0.55; per protocol: OR: 0.06, 95% CI: 0.01 to 0.38). Because the last GRT was not always performed immediately before RAL start, the estimation of the GSS might be imprecise. Therefore, a logistic regressions including exclusively patients who had a GRT on the last failing regimen was performed [ITT LOCF (n = 104): weighted multivariable OR: 0.36, 95% CI: 0.11 to 1.18, per protocol (n = 65): OR: 0.07, 95% CI: 0.01 to 0.46].
As an additional analysis, time to viral suppression was studied. Patients were included if they had at least 1 RNA measurement performed before treatment change, stop, or interruption (n = 109). The frequency of RNA measurements after RAL start was similar between groups (data not shown). Compared with patients with 2 NRTIs, patients with <2 NRTIs had a longer time to viral suppression. The hazard ratio with the multivariable and weighted multivariable regression was 0.63 (95% CI: 0.39 to 1.03, P = 0.064) and 0.54 (95% CI: 0.37 to 0.80, P = 0.002), respectively.
The availability of second-line antiretroviral agents and the introduction of new drug classes increased the options for salvage treatment markedly and raised the question of the optimal combination of compounds. Particularly, the role of genotypic partially or completely inactive NRTIs in such situations is unknown. In our study, we saw that NRTIs were often replaced by other drug classes, such as second-line NNRTIs, boosted PIs, T20, or MAR. Importantly, patients who received 2 inactive or partially active NRTIs were more likely to achieve viral suppression at week 24. These NRTIs might have a residual antiretroviral activity or select viruses with reduced replicative capacity, which might be favorable to achieve viral suppression.9 The presence of 2 NRTIs increased the chance to achieve viral suppression three times. Single NRTI did not significantly increase the chance to achieve viral suppression but tended to show an additional benefit compared with NRTI-sparing regimen (data not shown). Also the time to viral suppression was faster when 2 NRTIs were given. A short time to suppression might be beneficial because it may decrease the chance to accumulate resistance associated mutations in the very early phase of therapy. These findings were consistent and were confirmed with different approaches and sensitivity analyses.
The use of NRTIs in salvage therapies has several potential advantages. In contrast to new compounds, NRTIs are well studied after 20 years of use: Their long-term toxicities are well characterized, and the potential for drug-drug interactions is low. Costs are much lower compared with newer antiretroviral compounds, which are particularly relevant for developing countries and will become more important in the future when generic antiretroviral agents will be available.
Previous studies showed that NRTI-sparing regimens suppress viremia in treatment-naive and treatment-experienced patients but increase the probability to select for drug resistance mutations. They reduce the frequency of lipoatrophy, but other adverse events occurred when combining remaining drug classes, (eg, PIs and NNRTIs).32-36
Preliminary results from another study addressing the effect of inactive NRTIs in salvage therapy are in contradiction with our findings. However, in contrast to our study, no adherence data were available, no weighting was performed, the number of patients with <2 NRTIs was very small (27 compared with 76 in our study) and it was not differentiated between patients receiving salvage treatment with RAL, MAR, or ETV.37
We used marginal structural models to overcome confounding by indication. The model performed in this study simulated a hypothetical randomized controlled trial in which patients were randomly assigned to receive a treatment with <2 or 2 NRTIs. As in any observational studies, it is impossible to exclude unmeasured confounding. In particular, we cannot fully exclude that there were additional factors, which led physicians to choose a treatment with <2 NRTIs, but which were also associated with a worse treatment outcome. However, in absence of a randomized controlled trial, observational studies represent the best available evidence. To analyze long-term effects and NRTI-related toxicities, large cohort collaborations will be needed.
The possibility to maintain NRTI in the salvage regimen despite the presence of major drug resistance mutation is of high relevance because the drug pipeline of new antiretroviral agents starts to decline and on a global scale resistance will continue to accumulate.38-41
To summarize, our study demonstrated that partially active or inactive NRTIs showed a beneficial effect on the short-term virological outcome in patients receiving RAL. Therefore, our study supports the strategy to administer 2 NRTIs in salvage therapy with RAL even if inactive or only partially active according to GSS. The negative impact on viral fitness by maintaining drug resistance mutations and the residual activity of NRTIs must not be underestimated. However, the benefit of these NRTIs should be balanced with potential complications because complex antiretroviral regimens can be associated with increased toxicity or poor adherence. Further studies and collaborations are needed to support our findings and to analyze the long-term benefit of partially active or inactive NRTIs.
We thank the patients who participate in the SHCS; the physicians and study nurses for excellent patient care; the resistance laboratories for high quality genotypic drug resistance testing; SmartGene, Zug, Switzerland for technical support; Brigitte Remy, Martin Rickenbach, MD, F. Schöni-Affolter and Yannick Vallet from the SHCS Data Center in Lausanne for the data management and Marie-Christine Francioli for administrative assistance.
1. Cooper DA, Steigbigel RT, Gatell JM, et al. Subgroup and resistance analyses of raltegravir for resistant HIV-1 infection. N Engl J Med
2. Steigbigel RT, Cooper DA, Kumar PN, et al. Raltegravir with optimized background therapy for resistant HIV-1 infection. N Engl J Med
3. Gulick RM, Lalezari J, Goodrich J, et al. Maraviroc for previously treated patients with R5 HIV-1 infection. N Engl J Med
4. Lazzarin A, Clotet B, Cooper D, et al. Efficacy of enfuvirtide in patients infected with drug-resistant HIV-1 in Europe and Australia. N Engl J Med
5. Lalezari JP, Henry K, O'Hearn M, et al. Enfuvirtide, an HIV-1 fusion inhibitor, for drug-resistant HIV infection in North and South America. N Engl J Med
6. Lazzarin A, Campbell T, Clotet B, et al. Efficacy and safety of TMC125 (etravirine) in treatment-experienced HIV-1-infected patients in DUET-2: 24-week results from a randomised, double-blind, placebo-controlled trial. Lancet
7. Madruga JV, Cahn P, Grinsztejn B, et al. Efficacy and safety of TMC125 (etravirine) in treatment-experienced HIV-1-infected patients in DUET-1: 24-week results from a randomised, double-blind, placebo-controlled trial. Lancet
8. Deeks SG, Hoh R, Neilands TB, et al. Interruption of treatment with individual therapeutic drug classes in adults with multidrug-resistant HIV-1 infection. J Infect Dis
9. Wainberg MA. The impact of the M184V substitution on drug resistance and viral fitness. Expert Rev Anti Infect Ther
10. Arnaudo E, Dalakas M, Shanske S, et al. Depletion of muscle mitochondrial DNA in AIDS patients with zidovudine-induced myopathy. Lancet
11. Dalakas MC, Illa I, Pezeshkpour GH, et al. Mitochondrial myopathy caused by long-term zidovudine therapy. N Engl J Med
12. Lewis W, Day BJ, Copeland WC. Mitochondrial toxicity of NRTI antiviral drugs: an integrated cellular perspective. Nat Rev Drug Discov
13. Bonnet F, Bonarek M, Morlat P, et al. Risk factors for lactic acidosis in HIV-infected patients treated with nucleoside reverse-transcriptase inhibitors: a case-control study. Clin Infect Dis
14. Cossarizza A, Moyle G. Antiretroviral nucleoside and nucleotide analogues and mitochondria. AIDS
15. Cote HC, Brumme ZL, Craib KJ, et al. Changes in mitochondrial DNA as a marker of nucleoside toxicity in HIV-infected patients. N Engl J Med
16. Lonergan JT, Behling C, Pfander H, et al. Hyperlactatemia and hepatic abnormalities in 10 human immunodeficiency virus-infected patients receiving nucleoside analogue combination regimens. Clin Infect Dis
17. Carr A, Miller J, Law M, et al. A syndrome of lipoatrophy, lactic acidaemia and liver dysfunction associated with HIV nucleoside analogue therapy: contribution to protease inhibitor-related lipodystrophy syndrome. AIDS
18. Mallal SA, John M, Moore CB, et al. Contribution of nucleoside analogue reverse transcriptase inhibitors to subcutaneous fat wasting in patients with HIV infection. AIDS
19. Kovari H, Ledergerber B, Peter U, et al. Association of noncirrhotic portal hypertension in HIV-infected persons and antiretroviral therapy with didanosine: a nested case-control study. Clin Infect Dis
20. Imhof A, Ledergerber B, Gunthard HF, et al. Risk factors for and outcome of hyperlactatemia in HIV-infected persons: is there a need for routine lactate monitoring? Clin Infect Dis
21. Taburet AM, Piketty C, Chazallon C, et al. Interactions between atazanavir-ritonavir and tenofovir in heavily pretreated human immunodeficiency virus-infected patients. Antimicrob Agents Chemother
22. Scherrer AU, von Wyl V, Fux CA, et al. Implementation of raltegravir in routine clinical practice: selection criteria for choosing this drug, virologic response rates, and characteristics of failures. J Acquir Immune Defic Syndr
23. Ledergerber B, Egger M, Opravil M, et al. Clinical progression and virological failure on highly active antiretroviral therapy in HIV-1 patients: a prospective cohort study. Swiss HIV Cohort Study. Lancet
24. Schoeni-Affolter F, Ledergerber B, Rickenbach M, et al. Cohort profile: the swiss HIV cohort study. Int J Epidemiol
25. Yerly S, Vora S, Rizzardi P, et al. Acute HIV infection: impact on the spread of HIV and transmission of drug resistance. AIDS
26. von Wyl V, Yerly S, Boni J, et al. Emergence of HIV-1 drug resistance in previously untreated patients initiating combination antiretroviral treatment: a comparison of different regimen types. Arch Intern Med
27. Glass TR, De Geest S, Weber R, et al. Correlates of self-reported nonadherence to antiretroviral therapy in HIV-infected patients: the Swiss HIV Cohort Study. J Acquir Immune Defic Syndr
28. Peuchant O, Capdepont S, Ragnaud JM, et al. Primary resistance to enfuvirtide (T20) in recently HIV-1 infected, antiretroviral-naive patients from the ANRS Aquitaine Cohort. Antivir Ther
29. Hernan MA, Brumback B, Robins JM. Marginal structural models to estimate the causal effect of zidovudine on the survival of HIV-positive men. Epidemiology
30. Robins JM, Hernan MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology
31. Johnson VA, Brun-Vezinet F, Clotet B, et al. Update of the drug resistance mutations in HIV-1: December 2009. Top HIV Med
32. Calmy A, Petoumenos K, Lewden C, et al. Combination antiretroviral therapy without a nucleoside reverse transcriptase inhibitor: experience from 334 patients in three cohorts. HIV Med
33. Dwyer DE, Workman C, Hales G, et al. Enfuvirtide in HIV-1-infected individuals changing therapy to a nucleoside reverse transcriptase inhibitor sparing regimen: the ALLIANCE Study. Antivir Ther
34. Fischl MA, Collier AC, Mukherjee AL, et al. Randomized open-label trial of two simplified, class-sparing regimens following a first suppressive three or four-drug regimen. AIDS
35. Joly V, Yeni P. Nucleoside analogue-sparing strategy for the treatment of chronic HIV infection: potential interest and clinical experience. Antivir Ther
36. Riddler SA, Haubrich R, DiRienzo AG, et al. Class-sparing regimens for initial treatment of HIV-1 infection. N Engl J Med
37. Trottier B, Thomas R, Nguyen V, et al. Should inactive nucleoside/tide reverse transcriptase inhibitors (NRTIs) still be used in salvage regimens, with new classes/generations of antiretrovirals in three-class-experienced, multi-drug resistant patients? Paper presented at: 5th IAS Conference on HIV Pathogenesis, Treatment and Prevention; 2009; Cape Town, South Africa.
38. Lodwick R, Costagliola D, Reiss P, et al. Triple-class virologic failure in HIV-infected patients undergoing antiretroviral therapy for up to 10 years. Arch Intern Med
39. Hosseinipour MC, van Oosterhout JJ, Weigel R, et al. The public health approach to identify antiretroviral therapy failure: high-level nucleoside reverse transcriptase inhibitor resistance among Malawians failing first-line antiretroviral therapy. AIDS
40. Haidara A, Chamberland A, Sylla M, et al. High level of primary drug resistance in Mali. HIV Med
41. Gupta RK, Hill A, Sawyer AW, et al. Virological monitoring and resistance to first-line highly active antiretroviral therapy in adults infected with HIV-1 treated under WHO guidelines: a systematic review and meta-analysis. Lancet Infect Dis
APPENDIX I: MEMBERS OF THE SHCS
The members of the SHCS are Battegay M, Bernasconi E, Böni J, Bucher HC, Bürgisser P, Calmy A, Cattacin S, Cavassini M, Dubs R, Egger M, Elzi L, Fischer M, Flepp M, Fontana A, Francioli P (President of the SHCS), Furrer H (Chairman of the Clinical and Laboratory Committee), Fux CA, Gorgievski M, Günthard H (Chairman of the Scientific Board), Hirsch HH, Hirschel B, Hösli I, Kahlert C, Kaiser L, Karrer U, Kind C, Klimkait T, Ledergerber B, Martinetti G, Müller N, Nadal D, Paccaud F, Pantaleo G, Rauch A, Regenass S, Rickenbach M (Head of Data Center), Rudin C (Chairman of the Mother & Child Substudy), Schmid P, Schultze D, Schüpbach J, Speck R, de Tejada BM, Taffé P, Telenti A, Trkola A, Vernazza P, Weber R, Yerly S.