Figures
Abstract
Background
Good adherence to antiretroviral therapy (ART) is critical for successful HIV treatment. However, some patients remain virologically suppressed despite suboptimal adherence. We hypothesized that this could result from host genetic factors influencing drug levels.
Methods
Eligible individuals were Caucasians treated with efavirenz (EFV) and/or boosted lopinavir (LPV/r) with self-reported poor adherence, defined as missing doses of ART at least weekly for more than 6 months. Participants were genotyped for single nucleotide polymorphisms (SNPs) in candidate genes previously reported to decrease EFV (rs3745274, rs35303484, rs35979566 in CYP2B6) and LPV/r clearance (rs4149056 in SLCO1B1, rs6945984 in CYP3A, rs717620 in ABCC2). Viral suppression was defined as having HIV-1 RNA <400 copies/ml throughout the study period.
Results
From January 2003 until May 2009, 37 individuals on EFV (28 suppressed and 9 not suppressed) and 69 on LPV/r (38 suppressed and 31 not suppressed) were eligible. The poor adherence period was a median of 32 weeks with 18.9% of EFV and 20.3% of LPV/r patients reporting missed doses on a daily basis. The tested SNPs were not determinant for viral suppression. Reporting missing >1 dose/week was associated with a lower probability of viral suppression compared to missing 1 dose/week (EFV: odds ratio (OR) 0.11, 95% confidence interval (CI): 0.01–0.99; LPV/r: OR 0.29, 95% CI: 0.09–0.94). In both groups, the probability of remaining suppressed increased with the duration of continuous suppression prior to the poor adherence period (EFV: OR 3.40, 95% CI: 0.62–18.75; LPV/r: OR 5.65, 95% CI: 1.82–17.56).
Citation: Glass TR, Rotger M, Telenti A, Decosterd L, Csajka C, Bucher HC, et al. (2012) Determinants of Sustained Viral Suppression in HIV-Infected Patients with Self-Reported Poor Adherence to Antiretroviral Therapy. PLoS ONE 7(1): e29186. https://doi.org/10.1371/journal.pone.0029186
Editor: Cristian Apetrei, University of Pittsburgh Center for Vaccine Research, United States of America
Received: September 22, 2011; Accepted: November 22, 2011; Published: January 3, 2012
Copyright: © 2012 Glass et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This study has been financed in the framework of the Swiss HIV Cohort Study (SHCS project 619) which is in turn supported by the Swiss National Foundation (SNF 33CS30_134277). CM has been supported by a grant from the Stiftung Forschung Infektionskrankheiten (project 35), University Hospital Basel. TRG and HCB are supported by Santésuisse and Gottfried and Julia Bangerter-Rhyner-Foundation. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have read the journal's policy and have the following conflicts: HCB has received travel grants, honoraria and unrestricted research grants from GlaxoSmithKline, Bristol-Myers-Squibb, Merck Sharp & Dohme, Gilead, Janssen, Roche, Abbott, Tibotec, Boehringer-Ingelheim and ViiV Healthcare. HFG has been an adviser and/or consultant for the following companies: GlaxoSmithKline, Abbott, Novartis, Boehringer Ingelheim, Roche, Tib tec and Bristol-Myers Squibb, and has received unrestricted research and educational grants from Roche, Abbott, Bristol-Myers Squibb, GlaxoSmithKline, Tibotec and Merck Sharp & Dohme (integral financial support went to the institution). EB received honoraria or travel grants from the following pharmaceutical companies: Abbott, Bristol-Myers Squibb, Boehringer Ingelheim, Gilead, ViiV Healthcare, Merck Sharp & Dohme, Janssen. MB has received speaker's honoraria from Boehringer Ingelheim, Bristol-Myers Squibb, Gilead Sciences, GlaxoSmithKline, Merck Sharp & Dohme and Tibotec and has received unrestricted research grants from Boehringer Ingelheim, Gilead Sciences, GlaxoSmithKline and serves as a consultant for Boehringer Ingelheim Switzerland. TRG, MR, AT, LD, CC, MR, DN, BH, GW, and CM declare no competing interests. This does not alter the authors' adherence to all the PLoS ONE policies on sharing data and materials.
Introduction
High level of adherence to antiretroviral therapy (ART) is needed to achieve and maintain virological suppression, prevent drug resistance and improve survival in HIV-infected individuals [1]. The minimal level of adherence was initially established as 95% of drug intake and was based on data from non-boosted protease inhibitors (PI) based regimens [1]. More recent studies have indicated that the optimal level of adherence may differ according to the antiretroviral class [2], [3], and should be greater than 95% for non-boosted PI or boosted PI and at least 80% for non-nucleoside reverse transcriptase inhibitors (NNRTI) based regimens [4].
Although nearly perfect adherence to ART is critical for successful HIV treatment, some patients remain suppressed despite poor adherence. The ability to achieve and sustain virological suppression despite suboptimal adherence could result from host genetic and/or pharmacological factors. In fact, the administration of standard doses of most antiretroviral drugs results in a large inter-individual variability in plasma drug concentrations [5], [6]. The reasons for this variability are multifactorial and may involve, besides adherence, factors such as drug-drug interactions, co-morbidities, ethnicity or weight differences, and genetics. In the recent years, genetic polymorphisms in genes coding for drug metabolizing enzymes or transporters have been shown to influence the pharmacokinetics of antiretroviral agents [7]. For instance, a correlation with efavirenz disposition was demonstrated for a 516G>T polymorphism in CYP2B6 (CYP2B6*6). Individuals with the CYP2B6 516TT (homozygous loss of function) genotype had higher efavirenz levels compared to carriers of 516GT (heterozygous loss of function) or 516CC (homozygous reference) genotypes [8]. More recently, the contribution of CYP2B6 genetic variants to efavirenz disposition was assessed in a pharmacogenetic- pharmacokinetic population analysis, and interestingly, explained half of the inter-individual variability observed with this drug [9]. All together, these factors could explain that some individuals may have higher drug concentrations and may be able to maintain effective levels even after missing doses. Intrinsic drug characteristics such as half-life (long half-life) or drug potency are additional factors that may also contribute to a continuous drug pressure despite suboptimal adherence.
We used the longitudinal nature of the Swiss HIV Cohort Study (SHCS) to select individuals with long periods of self-reported poor adherence to HIV therapy and to investigate the genetic and pharmacological characteristics in those remaining suppressed compared to those not suppressed.
Methods
Participants
This study included HIV-infected patients enrolled in the SHCS, a nationwide prospective cohort study enrolling HIV-infected individuals aged 18 years or older who are followed-up in HIV clinics (www.SHCS.ch) [10]. Eligible patients were Caucasians, reported poor adherence on at least 2 consecutive follow-up visits more than 3 months apart between January 2003 until May 2009, had at least 2 HIV-1 RNA values during the study period, and were on regimens containing efavirenz (EFV) and/or boosted lopinavir (LPV/r) for more than 24 weeks (see figure 1). The study period was defined for each participant as the first until the last reported date of poor adherence. Participants involved in a drug trial with planned treatment interruptions were not eligible for this study.
Ethics statement
The ethics committees of all participating centers approved the genetic project and the permission for genetic analyses was approved by the Institutional Review Boards/Ethics committees of the University Hospitals of Basel, Lausanne, Zurich, Bern, Geneva, St Gallen and Lugano. The participants gave written informed consent for genetic testing.
Genetic analyses
Genotyping of single nucleotide polymorphisms (SNPs) in candidate genes previously reported to decrease EFV (rs3745274, marker of CYP2B6*6; rs35303484, marker of CYP2B6*11 and rs35979566, marker of CYP2B6*15) [9], [11] and LPV/r clearance (rs4149056, marker of SLCO1B1*5; rs6945984 in CYP3A and rs717620 in ABCC2) [12] was performed using commercially available TaqMan allelic discrimination assays (Applied Biosystems, Foster City, California, USA).
For each drug, the influence of candidate alleles was tested both independently and jointly in a genetic score. Scores for each SNP were defined as 0 = Hom-Ref (homozygous reference) allele, 1 = Het-LOF (heterozygous loss of function) allele or Hom-LOF (homozygous loss of function) allele. For LPV/r, score groups were subsequently classified as: <2 LOF alleles or ≥2 LOF alleles for the combination of the 3 SNPs as done previously [12].
Drug levels
EFV and LPV/r drug levels were measured during therapeutic drug monitoring only in case of clinical indication and therefore were either not available or regularly documented in all patients. Due to the limited number of drug levels measured within the poor adherence period, we considered all EFV or LPV/r levels available in a given individual. Drug levels are expected to be consistently higher in carriers of genetic variants thus making extrapolation a reasonable approach. Drug levels were quantified by liquid chromatography coupled to tandem mass spectrometry in accordance with previously validated methods [13], [14].
Since plasma levels were measured at random time after last dose intake, observed EFV and LPV/r concentrations were categorized according to percentiles to allow for comparison between individuals. Concentration-time percentile curves (10th, 25th, 50th, 75th and 90th percentiles) were derived using simulations based on previously published population pharmacokinetic models [12], [15] and performed using NONMEM (version VI, Icon Development Solutions, Ellicott City, Maryland, USA). For individuals with multiple drug measurements, the average percentile was calculated.
Adherence monitoring and definition
Adherence was measured by participants' self-report of non-adherence. During the SHCS follow-up visits, the participants were systematically asked two questions on adherence by their clinicians: (i) How often did you miss a dose of your medication in the last month? Daily, more than once a week, once a week, once every two weeks, once a month, never; and (ii) Did you miss more than one dose in a row? Yes, no. In this study, poor adherence was defined as reporting missing doses of medication either “daily”, “more than once a week”, or “once a week” on at least two consecutive reports. The poor adherence criteria were based on a previous analysis of the SHCS that showed a significantly higher risk of virological rebound in virally suppressed individuals reporting missing doses of ART≥once a week in the last month compared to individuals reporting never missing a dose of ART [16].
Outcome and covariate definitions
The primary endpoint of the study was virological suppression defined as having all HIV-RNA<400 copies/ml during the study period.
Covariates that were potential confounders of the relationship between host genetic or pharmacological factors and viral suppression were considered for inclusion in the analysis: demographic-related factors (age, gender), patient-related factors (body mass index, current intravenous drug use, co-infection with hepatitis B or C as defined below), condition-related factors (time living with HIV, time on ART, CD4 cell count), and treatment related-factors (prior mono or dual therapy, prior exposure to ART, duration of viral suppression prior to study period, nucleoside backbone, co-medications).
Hepatitis B virus infection was considered active if HBs antigen or HBV DNA were positive. Hepatitis C virus infection was considered active if anti-HCV antibody and HCV RNA were positive and healed if HCV serology was positive and HCV RNA negative. Elevation of liver enzymes (alanine aminostransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP)) were graded according to the AIDS adverse event grading table.
Statistical methods
Logistic regression models were utilized to estimate the effect of host genetic factors and drug levels on viral suppression. Potential confounders of the relationship between host genetic and pharmacological factors were considered for inclusion in the multivariable model although the small sample size limited the number of variables. All analyses were carried out separately for EFV and LPV/r. Estimates of the association between the predictors and outcome are presented with odds ratios (OR) and 95% confidence intervals (CI).
In sensitivity analyses, we also considered an alternative definition for viral suppression - all HIV-1 RNA<50 copies/ml.
All analyses were done with SAS v 9.2 (SAS Institute, Cary, North Carolina) and Stata 10.1 (StataCorp, College Station, Texas).
Results
Study population
Over the 6.5 year study period, 648 poor adherence periods from 542 individuals were recorded. The patient selection process is depicted in Figure 1. In the end, 37 individuals on EFV and 69 on LPV/r were included. The median length of poor adherence in both groups was 32 weeks with 18.9% of EFV and 20.3% of LPV/r patients indicating they missed a dose of ART on a daily basis. Demographics of the two populations are described in Table 1.
In the EFV group, 28/37 (75.7%) maintained a viral load <400 copies/ml throughout the period of poor adherence. In the LPV group, only 38/69 (55.1%) of patients remained suppressed throughout the study period.
Effect of genetic factors on viral suppression
In the group of patients receiving EFV, genetic variants were found in CYP2B6 only for rs3745274 with an allele frequency of 25%. In the group of patients receiving LPV/r, genetic variants were found in SLCO1B1 (rs4149056), CYP3A (rs6945984) and ABCC2 (rs717620) with allele frequencies of 19%, 9% and 22%, respectively (Table 2). Overall, the observed allele frequencies were in good agreement with publicly available SNPs database (www.ncbi.nlm.nih.gov/SNP).
Univariable and multivariable logistic regression analyses for EFV failed to detect an association between 516G>T polymorphism in CYP2B6 and the likelihood of remaining suppressed (Table 3).
Univariable and multivariable logistic regression analyses for LPV/r failed to detect an association between each independent SNP and viral suppression (p>0.15) with the exception of the univariable result for SLCO1B1*5 (OR 0.37, 95% CI: 0.13–1.06, p = 0.06). When analyzing the joint effect of the 3 SNPs in a genetic score, the univariable logistic regression showed a trend between a decreased likelihood of viral suppression in carriers of genetic variations (p = 0.08) however the trend did not remain in the multivariable analysis.
Effect of drug levels on viral suppression
Drug concentrations were only available in 30% of EFV and 50% of LPV/r patients. For most patients, the intra-individual variability in drug levels was very pronounced further supporting the poor adherence pattern in this selected population. No association were found between higher drug levels (expressed as average percentile value) and presence of genetic variants in EFV patients (p = 0.36) or combined genetic variants in LPV/r patients (p = 0.26). The analysis of independent genetic variants in LPV/r patients showed a correlation between carriers of SLCO1B1*5 and higher drug levels (average percentile value in carriers vs non carriers: 62 vs 42, p = 0.03) but not for carriers of variants in CYP3A (p = 0.63) and ABCC2 (p = 0.61).
No association was found between higher drug levels and the likelihood of remaining suppressed in EFV (p = 0.54) and LPV/r patients (p = 0.12). However, single or repetitive exposure to low drug levels (i.e. ≤10th percentile) was associated with a lower likelihood of viral suppression in LPV/r patients (OR 4.40, 95% CI: 1.04–18.60, p = 0.04). Due to the small number of patients with available drug concentrations, drug levels were not included in the multivariable analyses.
Effect of adherence on viral suppression
Despite all patients having reported poor adherence, differences in the likelihood of viral suppression could still be seen between the adherence levels (Tables 3 and 4). Individuals on LPV/r missing >1 dose per week were significantly less likely to experience viral suppression than those missing 1 dose a week (OR 0.06, 95% CI: 0.01–0.55, p = 0.01). For EFV patients, a similar trend was found but this was not statistically significant (OR 0.11, 95% CI: 0.01–2.03, p = 0.14).
Predictors of virological suppression
In univariable models of EFV patients, only low adherence was significantly associated with decreased odds of viral suppression (p = 0.05) (Table 3). However, in multivariable models, the effect was attenuated and was no longer significant (p = 0.14).
In univariable models of LPV/r patients, having low adherence levels, past or current IDU, co-medication, mild elevation of ALP (>1.25–2.5 above the upper limit of normal), or longer time since HIV diagnosis was significantly correlated with a decreased likelihood of remaining suppressed during the period of poor adherence. Prior suppression on the current regimen and higher baseline CD4 cell count were associated with an increased likelihood of remaining suppressed despite poor adherence (Table 4). In multivariable models, only low adherence and prior suppression remained significant whereas older age was associated with an increased likelihood of remaining suppressed.
These findings remained consistent when defining viral suppression as RNA<50 copies/ml.
Discussion
Nearly perfect adherence to antiretroviral therapy is widely accepted as the primary determinant for achieving and maintaining viral response. However, less is known about the determinants of virological suppression observed paradoxically in some individuals with long periods of self-reported poor adherence to ART. We hypothesized that individuals carrying genetic variants associated with a slow metabolizing profile would maintain sufficient viral inhibitory drug levels even after missing doses.
In our study population, we failed to show a correlation between 516G>T polymorphism in CYP2B6 and the likelihood of remaining suppressed in patients with self-reported suboptimal adherence to EFV based regimens. The population pharmacokinetic estimates of EFV half-lives have been reported to be 23, 27 and 48 hours for individuals with the CYP2B6 516 GG, GT and TT genotypes, respectively [17]. Our study population was mainly characterized by carriers of the CYP2B6 516 GG (53%) and 516 GT genotypes (44%) whereas only one patient had the 516 TT genotype (Table 2). Thus, our results suggest that the modest increase in EFV half-life (17%) associated with the CYP2B6 516 GT genotype might not be strong enough to compensate for inconsistent adherence by maintaining effective levels until the subsequent dose.
In the multivariable analyses, no correlation was found between independent or combined genetic variants and the likelihood of remaining suppressed in patients with suboptimal adherence to LPV/r based regimens. Although there was a trend to increased drug levels with the SLCO1B1*5 allele, this did not translate in protection from virological failure. One should note that while the genetic determinants of EFV blood levels are well understood [9]), the contribution of genetics to the inter-individual variability of LPV/r pharmacokinetics is modest (less than 5%) [12].
In our population, the likelihood of remaining suppressed was rather related to the degree of adherence. The estimated minimal level of adherence to maintain viral suppression was 86% for EFV and 93% for LPV/r (mainly administered twice daily in our population) as indicated by the increased risk of viral failure in patients missing more than 1 dose a week compared to those missing 1 dose a week (Tables 3 and 4). The lower adherence threshold required for EFV compared to LPV/r was supported by the difference in observed rate of viral suppression in that 76% of the patients receiving EFV versus 55% treated with LPV/r remained suppressed throughout the study period. Our results are consistent with previous studies showing that viral suppression can be reached in some but not all patients at a level of adherence averaging 80% for NNRTI-based regimens [3], [4], [18], [19] and >90% for PI-based regimens [2], [3], [4], [18], [20]. The higher adherence threshold observed in our study might reflect the fact that we used self-reported adherence, which overestimates adherence by 10–20% compared to electronic monitoring [21], [22]. In our study, the degree of adherence rather than consecutive missed doses was a stronger predictor for viral suppression in EFV and LPV/r models. EFV forgiveness of non-adherence has been reported to be likely attributable to its extended half-life [3], [4], [19], whereas LPV/r has a high genetic barrier to resistance which may enhance its forgiveness [23], [24].
Besides the degree of adherence, we found that the likelihood of remaining suppressed was also dependent on the dosing schedule for LPV/r. In our population, LPV/r was administered twice daily (BID) in 64 patients and once daily (QD) in 5 patients. Interestingly, the failure rate was higher in those on a QD regimen with 80% (4/5) of the patients on the LPV/r QD regimen compared to 42% (27/64) on the LPV/r BID regimen not maintaining viral suppression. This observation is compatible with the findings of a recent study comparing the pharmacokinetics of LPV/r BID and QD following drug cessation [25]. Although no participants receiving LPV/r BID had drug levels below the minimal effective concentration (MEC) at 12 hours, 44% of participants receiving LPV/r QD were below the MEC at 24 hours. The higher occurrence of virological failure observed in some clinical trials with the LPV/r QD regimen seems to be related to the fast decay of LPV/r and the achievement of sub-therapeutic concentrations [25].
Interestingly, our data showed that the probability of remaining suppressed increased with the duration of continuous suppression prior to the poor adherence period in both groups, which is in agreement with previous data [26], [27]. It has been postulated that initial treatment during high viral burden likely required higher levels of adherence for full viral suppression than later in chronic treatment when viral burden was less [26], [27]. Furthermore, it has been reported that the latent reservoir and its ability to reactivate upon stimulation with cognate antigen decreases with duration of treatment [28]. Thus, individuals may be more vulnerable to suboptimal adherence shortly after achieving viral suppression but may be able to tolerate missed doses after long-term viral suppression. It should be noted that the probability of remaining suppressed might also be related to inter-individual differences in immune response to HIV virus. Factors such as thymus output, co-infections, residual viral production, age, gender, genetics of immune system, nadir of CD4 cell are known to impact the immune response to HIV [29]. This assumption is supported by the observation that in the univariable model of LPV/r patients, higher baseline CD4 cell count was associated with an increased likelihood of remaining suppressed.
Limitations
Some limitations of our study should be acknowledged. The stringent selection criteria of our study population (severe non adherence and being steadily on EFV and/or LPV/r) gave us a small sample size and limited power to explore associations in our data. In addition, the limited number of patients with documented drug levels did not allow satisfactory exploration of the role of drug concentration as a predictor of viral suppression. Although, due to the high level of non-adherence, this relationship may have been difficult to quantify precisely.
Conclusions
The investigated genetic variants did not play a significant role in the sustained viral suppression of individuals with poor adherence. The risk of virological failure decreased with longer duration of viral suppression in this selected population. Patients varied considerably in their prior ART exposure and, although we attempted to adjust for this and the possibility of resistance, we cannot exclude the potential impact of these factors.
Acknowledgments
The members of the Swiss HIV Cohort Study are J. Barth, M. Battegay, E. Bernasconi, J. Böni, H.C. Bucher, C. Burton-Jeangros, A. Calmy, M. Cavassini, C. Cellerai, M. Egger, L. Elzi, J. Fehr, J. Fellay, M. Flepp, P. Francioli (President of the SHCS), H. Furrer (Chairman of the Clinical and Laboratory Committee), C. Fux, M. Gorgievski, H. Günthard (Chairman of the Scientific Board), D. Haerry (Deputy of “Positive Council”), B. Hasse, H.H. Hirsch, B. Hirschel, I. Hösli, Ch. Kahlert, L. Kaiser, O. Keiser, C. Kind, Th. Klimkait, H. Kovari, B. Ledergerber, G. Martinetti, B. Martinez, K. Metzner, N. Müller, D. Nadal, G. Pantaleo, A. Rauch, S. Regenass, M. Rickenbach (Head of Data Center), C. Rudin (Chairman of the Mother & Child Substudy), P. Schmid, D. Schultze, F. Schöni-Affolter, J. Schüpbach, R. Speck, P. Taffé, P. Tarr, A. Telenti, A. Trkola, P. Vernazza, R. Weber, S. Yerly.
This work has been presented at the 6th International Conference on HIV treatment and prevention adherence. May 22–24, 2011, Miami, Florida, USA, abstract 69975.
Author Contributions
Conceived and designed the experiments: TRG AT MB CM. Performed the experiments: TRG M. Rotger LD CC CM. Analyzed the data: TRG CM. Contributed reagents/materials/analysis tools: M. Rotger AT LD CC HCB HFG M. Rickenbach DN BH EB GW MB. Wrote the paper: TRG CM. Critical revision of the manuscript and approval: TRG M. Rotger AT LD CC HCB HFG M. Rickenbach DN BH EB GW MB CM.
References
- 1. Paterson DL, Swindells S, Mohr J, Brester M, Vergis EN, et al. (2000) Adherence to protease inhibitor therapy and outcomes in patients with HIV infection. Ann Intern Med 133: 21–30.
- 2. Bangsberg DR (2006) Less than 95% adherence to nonnucleoside reverse-transcriptase inhibitor therapy can lead to viral suppression. Clin Infect Dis 43: 939–941.
- 3. Maggiolo F, Ravasio L, Ripamonti D, Gregis G, Quizan G, et al. (2005) Similar adherence rates favor different virologic outcomes for patients treated with nonnucleoside analogues or protease inhibitors. Clin Infect Dis 40: 158–163.
- 4. Lima VD, Harrigan R, Murray M, Moore DM, Wood E, et al. (2008) Differential impact of adherence on long-term treatment response among naive HIV-infected individuals. AIDS 22: 2371–2380.
- 5. Marzolini C, Buclin T, Decosterd LA, Biollaz J, Telenti A (2001) Nelfinavir plasma levels under twice-daily and three-times-daily regimens: high interpatient and low intrapatient variability. Ther Drug Monit 23: 394–398.
- 6. Marzolini C, Telenti A, Decosterd LA, Greub G, Biollaz J, et al. (2001) Efavirenz plasma levels can predict treatment failure and central nervous system side effects in HIV-1-infected patients. AIDS 15: 71–75.
- 7. Telenti A, Zanger UM (2008) Pharmacogenetics of anti-HIV drugs. Annu Rev Pharmacol Toxicol 48: 227–256.
- 8. Haas DW, Ribaudo HJ, Kim RB, Tierney C, Wilkinson GR, et al. (2004) Pharmacogenetics of efavirenz and central nervous system side effects: an Adult AIDS Clinical Trials Group study. AIDS 18: 2391–2400.
- 9. Arab-Alameddine M, Di Iulio J, Buclin T, Rotger M, Lubomirov R, et al. (2009) Pharmacogenetics-based population pharmacokinetic analysis of efavirenz in HIV-1-infected individuals. Clin Pharmacol Ther 85: 485–494.
- 10. Schoeni-Affolter F, Ledergerber B, Rickenbach M, Rudin C, Günthard HF, et al. (2010) Cohort profile: the Swiss HIV Cohort study. Int J Epidemiol 39: 1179–1189.
- 11. Rotger M, Tegude H, Colombo S, Cavassini M, Furrer H, et al. (2007) Predictive value of known and novel alleles of CYP2B6 for efavirenz plasma concentrations in HIV-infected individuals. Clin Pharmacol Ther 81: 557–566.
- 12. Lubomirov R, di Iulio J, Fayet A, Colombo S, Martinez R, et al. (2010) ADME pharmacogenetics: investigation of the pharmacokinetics of the antiretroviral agent lopinavir coformulated with ritonavir. Pharmacogenet Genomics 20: 217–230.
- 13. Colombo S, Beguin A, Telenti A, Biollaz J, Buclin T, et al. (2005) Intracellular measurements of anti-HIV drugs indinavir, amprenavir, saquinavir, ritonavir, nelfinavir, lopinavir, atazanavir, efavirenz and nevirapine in peripheral blood mononuclear cells by liquid chromatography coupled to tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 819: 259–276.
- 14. Fayet A, Beguin A, de Tejada BM, Colombo S, Cavassini M, et al. (2008) Determination of unbound antiretroviral drug concentrations by a modified ultrafiltration method reveals high variability in the free fraction. Ther Drug Monit 30: 511–522.
- 15. Csajka C, Marzolini C, Fattinger K, Decosterd L, Fellay J, et al. (2003) Population pharmacokinetics and effects of efavirenz in patients with human immunodeficiency virus infection. Clin Pharmacol Ther 73: 20–30.
- 16. Glass TR, De Geest S, Hirschel B, Battegay M, Furrer H, et al. (2008) Self-reported non-adherence to antiretroviral therapy repeatedly assessed by two questions predicts treatment failure in virologically suppressed patients. Antivir Ther 13: 77–85.
- 17. Ribaudo HJ, Haas DW, Tierney C, Kim RB, Wilkinson GR, et al. (2006) Pharmacogenetics of plasma efavirenz exposure after treatment discontinuation: an Adult AIDS Clinical Trials Group Study. Clin Infect Dis 42: 401–407.
- 18. Martin M, Del Cacho E, Codina C, Tuset M, de Lazzari E, et al. (2008) Relationship between adherence level, type of the antiretroviral regimen, and plasma HIV type 1 RNA viral load: a prospective cohort study. AIDS Res Hum Retroviruses 24: 1263–1268.
- 19. Nachega JB, Hislop M, Dowdy DW, Chaisson R, Regensberg L, et al. (2007) Adherence to nonnucleoside reverse transcriptase inhibitor-based HIV therapy and virologic outcomes. Ann Intern Med 146: 564–573.
- 20. Shuter J, Sarlo JA, Kanmaz TJ, Rode RA, Zingman BS (2007) HIV-infected patients receiving lopinavir/ritonavir-based antiretroviral therapy achieve high rates of virologic suppression despite adherence rates less than 95%. J Acquir Immune Defic Syndr 45: 4–8.
- 21. Pearson CR, Simoni JM, Hoff P, Kurth AE, Martin DP (2007) Assessing antiretroviral adherence via electronic drug monitoring and self-report: an examination of key methodological issues. AIDS Behav 11: 161–173.
- 22. Wagner G, Miller LG (2004) Is the influence of social desirability on patients' self-reported adherence overrated? J Acquir Immune Defic Syndr 35: 203–204.
- 23. King MS, Brun SC, Kempf DJ (2005) Relationship between adherence and the development of resistance in antiretroviral-naive, HIV-1-infected patients receiving lopinavir/ritonavir or nelfinavir. Journal of Infectious Diseases 191: 2046–2052.
- 24. von Wyl V, Yerly S, Boni J, Burgisser P, Klimkait T, et al. (2007) Emergence of HIV-1 drug resistance in previously untreated patients initiating combination antiretroviral treatment: a comparison of different regimen types. Arch Intern Med 167: 1782–1790.
- 25. Boffito M, Else L, Back D, Taylor J, Khoo S, et al. (2008) Pharmacokinetics of atazanavir/ritonavir once daily and lopinavir/ritonavir twice and once daily over 72 h following drug cessation. Antivir Ther 13: 901–907.
- 26. Lima VD, Bangsberg DR, Harrigan PR, Deeks SG, Yip B, et al. (2010) Risk of viral failure declines with duration of suppression on highly active antiretroviral therapy irrespective of adherence level. J Acquir Immune Defic Syndr 55: 460–465.
- 27. Rosenblum M, Deeks SG, van der Laan M, Bangsberg DR (2009) The risk of virologic failure decreases with duration of HIV suppression at greater than 50% adherence to antiretroviral therapy. PLoS One 4: e7196.
- 28. Strain MC, Gunthard HF, Havlir DV, Ignacio CC, Smith DM, et al. (2003) Heterogeneous clearance rates of long-lived lymphocytes infected with HIV: intrinsic stability predicts lifelong persistence. Proc Natl Acad Sci U S A 100: 4819–4824.
- 29. Corbeau P, Reynes J (2011) Immune reconstitution under antiretroviral therapy: the new challenge in HIV-1 infection. Blood 117: 5582–5590).