Introduction
Successful, durable control of HIV replication with antiretroviral (ARV) drugs requires high levels of medication adherence [1]. Surveys of HIV-positive patients receiving highly active antiretroviral therapy (HAART) indicate a preference for ARV regimens that can be taken once daily, involve few pills, and do not have specific dietary requirements [2]. A meta-analysis of virologic outcome data from clinical trials of various HAART regimens found a significant correlation between lower pill burden and treatment efficacy [3]. The expanded availability of once-daily ARV agents allows the clinician to construct regimens that are minimally intrusive to a patient's lifestyle and potentially facilitate greater adherence to therapy. Whereas there are now a number of ARV agents with the pharmacokinetic properties to allow once daily dosing, the available clinical evidence to support the use of particular once-daily regimens remains limited [4]. In this study we set out to compare the safety, virological efficacy and adherence of two once-daily thymidine-sparing ARV regimens with different dosing requirements in treatment naive subjects: lamivudine (3TC)/didanosine (ddI)/efavirenz (EFV) taken fasted versus tenofovir disoproxil fumarate (TDF)/ddI/EFV taken with food.
Methods
This was a single-centre, randomized, open-label study in ARV-naive, HIV-1-infected adults. Subjects were randomized 1: 1 to commence either 3TC (300 mg)/ddI (400 mg)/EFV 600 mg) taken fasted (3TC group) or TDF (300 mg)/ddI (250 mg)/EFV (600 mg) taken with food (TDF group). The dose of ddI was adjusted to 200 mg for body weight below 60 kg and both regimens were taken prior to bedtime.
Patients
ARV-naive, HIV-1-infected adults requiring ARV treatment, as assessed by their physician were included. There were no exclusion criteria regarding CD4+ cell count or plasma HIV-1 RNA. Patients were excluded if they had an active or ongoing opportunistic infection or if they were being treated with chemotherapy for a malignancy. Approval for the study was obtained from the local ethics committee (Riverside Research Ethics Committee) and all subjects gave written informed consent to participate in the study.
Primary endpoint
The primary endpoint of the trial was the difference in viral load at week 48 as measured by time-weighted change in averages from baseline. An unplanned analysis of the data was performed at week 12.
Efficacy and safety assessments
Following screening procedures, participating subjects returned (≤ 28 days later) for baseline assessments and randomization.
At each trial visit (week 4, 12, 24, 36, and 48 following ARV therapy initiation), subjects were to be assessed for HIV-related events, ARV drug-related adverse events, and any changes to concomitant medication. A physical examination was also to be performed and blood drawn to measure CD4+ cell counts and plasma HIV-RNA using Chiron Quantiplex (lower limit of detection 50 copies/ml; Chiron Corp., Emeryville, California, USA). In addition, clinical chemistry and hematology assessments were to be performed throughout the study. Genotypic analysis using the Virco TYPE HIV-1 assay were retrospectively performed on stored samples drawn at screening, baseline and at all subsequent time points should the subject have a HIV-RNA greater than 1000 copies/ml.
Adherence measures
Medication Event Monitoring System (MEMScap; Aardex Ltd, Zug, Switzerland) monitoring, self reported adherence questionnaires and food/therapy diaries were to be recorded in all subjects between weeks 0-4, 24-28 and 44-48.
Therapeutic drug monitoring
Samples for therapeutic drug monitoring (TDM) of EFV were to be collected at weeks 4 and 12 in all subjects. Blood was drawn between 10 and 15 h following the previous EFV dose. Should this not be possible blood was drawn at any time following the EFV dose and the 12-h concentration projected using a population-based linear regression method. EFV plasma concentrations were measured by a fully validated method using high-performance liquid chromatography (UV detection) at the TDM Service of the University of Liverpool, Liverpool, UK. Variability in EFV plasma concentrations between subjects was expressed as the percentage coefficient of variation (CV: SD/mean).
Statistical methods
The planned sample size was 100 subjects; this provided statistical power for the protocol to determine non-inferiority between the two treatment groups. The primary endpoint of the study was to be the difference in viral load at week 48, as measured by time-weighted change in averages from baseline. Treatment failure was defined as either: a less than 1 log10 drop in HIV-1 RNA by week 4, a rebound in viral load from nadir greater than 0.5 log10, or the emergence of genotypic resistance.
The unplanned interim analysis at week 12 used an intention to treat method where missing results equals failure and the last result available was carried forward. This comparative analysis evaluated mean plasma HIV-1 RNA expressed in log10 and median change from baseline of CD4+ cell count at weeks 4 and 12. Both indices have been presented with 95% confidence intervals (CI). Between-group comparisons of qualitative data were assessed using χ2 test and where appropriate this was adjusted for small numbers in cells using Yate's correction. EFV concentrations were categorized into equally spaced categories. Linear regression was used to test for association between log10 virological response between week 4 and week 12 and categorized EFV concentrations. A multivariate regression method was used to adjust for gender, ethnicity and treatment arms. All P values presented are two tailed.
Results
Baseline characteristics
Seventy-seven subjects were enrolled into the trial prior to the unplanned interim analysis that resulted in recruitment being suspended. Thirty-six subjects were randomized to the 3TC group and 41 to the TDF group. The baseline characteristics of the two groups were comparable with regard to gender, age, disease stage and laboratory indices (Table 1). Mean baseline viral load (95% CI) were 5.13 log10 (4.94-5.32) for the 3TC group and 4.99 log10 (4.78-5.22) for the TDF group, median (range) CD4+ cell counts was 158 × 106 cells/l (8-572) for the 3TC group and 174 × 106 cells/l (20-351) for the TDF group. A history of CDC Class C (AIDS) event (eight Kaposi's sarcoma, five Pneumocystis carinii pneumonia) was noted in 22% of the 3TC group and 12% of the TDF group. Fifteen of the 41 subjects in the TDF group and 18 of the 36 in the 3TC group had a baseline CD4 cell count less than 200 × 106 cells/l and viral load greater than 100 000 copies/ml.
Efficacy analyses
The interim analysis was performed after the investigators were notified of concerns regarding the efficacy of the TDF/ddI/EFV regimen in initial therapy raised from a small Spanish study [5] and the observation of four treatment failures occurring in the TDF group (three who failed to suppress HIV-1 RNA by ≥ 1 log10 by week 4 and the fourth rebounding from virological nadir at week 12). Mean (95% CI) viral load at weeks 4 and 12 for the 3TC group were 2.67 log10 (2.47-2.87) and 1.83 log10 (1.74-1.92), and for the TDF group 2.75 log10 (2.45-3.05) and 2.28 log10 (1.96-2.6), showing significant difference in viral response between the two groups at week 12 (P = 0.013). Median (range) increase from baseline in CD4 cell count at week 12 was comparable between the two groups, 117 × 106 cells/l (87-147) in the 3TC group and 108 × 106 cells/l (81-136) in the TDF group (Fig. 1).
Following this intention-to treat analysis further recruitment to the study was suspended. Subjects in the 3TC group continued as per the original trial protocol, and those subjects in the TDF group were promptly recalled to clinic. All of the subjects in the TDF group then had a genotype performed and empirically switched therapy to zidovudine/3TC (as Combivir) twice daily plus atazanavir/ritonavir 300/100 mg once daily. A fifth subject in the TDF group who had achieved a viral load reduction greater than 1 log10 at week 4 was subsequently found to have developed resistance mutations at week 8 when the ARV regimen was empirically changed.
By week 12 more patients in the TDF group had experienced virological failure; five of 41 (12.2%) compared with the 3TC group none of 36 (0%) (P < 0.05). The five patients who failed treatment in the TDF group shared similar baseline characteristics: CD4+ cell count less than 200 × 106 cells/l (ranging from 24 to 195), HIV-1 RNA greater than 100 000 copies/ml (ranging from 355 351 to 4 764 400), HIV-1 wild-type and HIV-1 clade B. The first resistance mutations observed to emerge in each of the patients who failed therapy in the TDF group were characteristic of non-nucleoside reverse transcriptase inhibitor (NNRTI) resistance [6]. Overall, the most prevalent reverse transcriptase (RT) mutation observed was 103N/T (three subjects), although a range of other NNRTI-associated mutations were also seen: 108I (two subjects), 190E/S (two subjects), 188L (two subjects), 100I (two subjects), 179D (two subjects), and 225H (one subject). Of the three subjects who developed NNRTI-associated mutations by week 4, one switched therapy soon after this time-point whereas two continued on TDF/ddI/EFV until week 12. These latter two subjects developed additional nucleoside reverse transcriptase inhibitor (NRTI)-associated mutations (shown by the genotype test performed at week 12), including 65R (one subject) and 74V (one subject) (Table 2).
Safety analyses
By week 12 no subjects in either group had discontinued the study due to a serious adverse event, withdrawn consent, or were lost to follow-up.
The only adverse event of note reported by week 12 was EFV-related rash, this occurred at an expected incidence with no significant difference between the two treatment groups [seven of 36 (19.4%) in the 3TC group and six of 41 (14.6%) in the TDF group] and did not result in the discontinuation of EFV in any subjects. No grade 3 or 4 abnormalities in clinical chemistry, lipids or haematology laboratory tests were observed up to week 12.
Adherence measures
MEMScap monitoring, adherence questionnaires and treatment diaries between baseline and week 4 confirmed a high level of adherence to treatment in both groups. MEMScap monitoring from baseline to week 4 demonstrating greater than 99% of doses were taken in both treatment groups (99.1% in the 3TC group and 99.9% in the TDF group). Adherence in all five individuals with virological failure in the TDF group was 100% by MEMScaps
Therapeutic drug monitoring
Samples from 65 patients (nine females) were analysed (n = 130). Median (range) EFV concentrations were 1569 (354-11611) and 1705 (466-13351) ng/ml at week 4 and 12, respectively. EFV concentrations were significantly higher in women than in men at weeks 4 and 12 [median (range) EFV concentrations in women 5294 (1049-11611) and 3369 (936-13351), respectively (P = 0.015 and 0.022). Among those subjects who were not considered treatment failures, 17 had at least one EFV concentration lower than the suggested minimum effective concentration (MEC) of 1000 ng/ml [7], while of the five failures (TDF group), three had concentrations lower than the MEC. The CV of the EFV concentration was greater than 90% at both weeks. Subjects with concentrations higher than 1100 ng/ml at week 4 and 12 were more likely to reach an undetectable viral load at week 12 (P < 0.001) and showed a greater decrease in plasma HIV-RNA between baseline and week 12 (P = 0.009), and this was consistent after adjusting for sex, ethnicity, and treatment group. No significant difference in EFV concentrations was observed between the two treatment groups or when comparing subjects who had a baseline viral load greater than 100 000 copies/ml and CD4 cell count lower than 200 × 106 cells/l with subjects who did not match these baseline characteristics.
Discussion
This study indicates that the combination of TDF, ddI, and EFV as initial therapy has diminished virological efficacy, particularly in those subjects with baseline CD4 cell count less than 200 × 106 cells/l and HIV-1 RNA greater than 100 000 copies/ml. Virological failure of this regimen occurred early after initiation of treatment and could not be attributed to problems with baseline resistance, efavirenz exposure or subject adherence.
HIV therapy is evolving toward the use of antiretroviral combinations that are compact, simplified, and with lesser toxicities, both short- and long-term. Patients strongly favour once daily therapy, as they must reconcile the adherence demands necessary to maintain virologic suppression with the realities of their daily lives [2]. Compact regimens involving triple nucleoside/nucleotide [N(t)RTI] combinations have been demonstrated to result in a sub-optimal virologic response in ARV-naive subjects [8-11]. Moreover, it has been shown in a recent study performed in 36 ARV-naive subjects that treatment with TDF, ddI and EFV resulted in a high early virological failure rate associated with the development of resistance mutations [5].
Despite TDF and ddI being recommended as an alternative N/NtRTI backbone in the International AIDS Society (IAS) guidelines for the treatment of HIV-1 infection in ARV-naive adults [4] our findings suggest that these agents do not provide sufficient antiviral activity when combined with EFV. The reasons for TDF/ddI being a sub-optimal N/NtRTI backbone are not clear. In the studies involving triple N/NtRTI combinations the most likely reason for virological failure would appear to be the selection of a single mutation, K65R which results in reduced sensitivity to all of the drugs in the regimen [8-11]. The reason for the failure of TDF/ddI when combined with EFV in the present study is less clear as the first mutations to appear were those associated with the NNRTI component and only further treatment seemed to select the K65R mutation in one of our subjects. Moreover, a retrospective cohort analysis of 14 subjects initiating therapy with TDF/ddI combined with either EFV or NVP has reported a virological failure rate of 50% with K65R observed in two out of seven failing patients at week 12 and four out of seven at week 24 [12]. It is possible that K65R was a common minority species in all of our subjects who failed ddI/TDF/EFV; to investigate this hypothesis further we are currently undertaking single gene probe clonal analysis of the available samples.
Our analysis of EFV plasma concentrations showed that having higher values leads to a more rapid reduction in viral load. However, no difference in EFV plasma concentrations was shown between the two treatment groups, with few subjects who responded to treatment (six in the 3TC and 11 in the TDF group) and three out of five who failed ddI/TDF/EFV having concentrations lower than the suggested MEC at least once. Interestingly, EFV concentrations were significantly higher in women, who all achieved an undetectable viral load at week 12. Despite confirming the relationship between plasma concentrations and virological response [13], the importance of concentrations higher than the suggested cut-off [7], and the frequent occurrence of low drug plasma concentrations (approximately 30%) [14], the information obtained by TDM alone could not explain the virological failures in the TDF group. However, it could be hypothesized that combinations including N/NtRTI backbones with low genetic barrier to resistance may be more vulnerable if they are combined with a third agent characterized by a wide inter-individual variability in plasma concentrations (90%).
Furthermore, whether a possible explanation for the higher rate of virological failure in the TDF group is a pharmacokinetic interaction between TDF and ddI is unclear. It is thought that this interaction is dependent upon the inhibition by TDF of purine nucleoside pyrophosphorylase (PNP)-dependent ddI degradation [15,16]. For this reason the dose of ddI is decreased when it is co-administered with TDF [17]. However a remarkable inter-subject variability in ddI plasma concentrations has been observed in subjects when ddI and TDF are co-administered independent of ddI dosage [18]. Planned analysis of the intra-cellular triphosphate concentrations of TDF and ddI may also provide an explanation of the poor virological response observed in the TDF group of our study. In vitro data has indicated no intracellular interaction between these two agents although such data may not always be an accurate indicator of what happens in vivo [19]. The antiviral activity of the two drugs when combined may be limited by the fact that they share a common pathway of activity, TDF and ddI both being adenosine analogues that inhibit HIV replication by causing premature chain termination during HIV-1 RNA reverse transcription [20].
Treatment with TDF/ddI-based regimens has recently been reported to be associated with paradoxical reductions or lack of increase in CD4+ cell counts despite subjects having undetectable viral loads. Paradoxically these CD4+ cell count drops were being observed most often after more than 6 months of treatment with a TDF/ddI-based regimen and when standard dose ddI (400 mg) was co-administered with TDF [21]. We did not observe any paradoxical drops in CD4+ cell count in subjects treated for up to 12 weeks with a TDF/ddI-based regimen in our study. A higher rate of pancreatitis has also been reported in subjects treated with TDF/ddI-based regimens in comparison with those treated with regimens including ddI without TDF or TDF without ddI [22].
Whatever the cause of the poor virological response observed in this study, it is clear that TDF/ddI should be used with caution as a N/NtRTI backbone. Clinicians should therefore be cautious when co-administering ddI/TDF with EFV in treatment-naive subjects with high baseline viral loads and low CD4 cell counts. TDF/ddI is also frequently prescribed in experienced patients based on the results of resistance testing [23]. Whether ddI/TDF-based regimens are sufficiently efficacious and safe for the treatment of experienced patients requires further study and careful follow up of those patients already prescribed these combinations.
Acknowledgements
Sponsorship: This study was financially supported by a grant kindly provided by Bristol Meyers Squibb, UK.
References
1. Carmona A, Knobel H, Guelar A, Grau S, Mateu-de Antonio J, Lopez-Colomes JL, et al. Factors influencing survival in HIV infected patients treated with HAART. Presented at: XIIIth International AIDS Conference. Durban, July 2000 [abstract TuOrB417].
2. Moyle G. The assessing patient's preferred treatments (APPT-1) study. Int J STD AIDS 2003; 14(Suppl 1):34-36.
3. Bartlett JA, DeMasi R, Quinn J, Moxham C, Rousseau F. Overview of the effectiveness of triple combination therapy in antiretroviral naive HIV-1 infected adults. AIDS 2001; 15:1369-1377.
4. Yeni PG, Hammer SM, Hirsch MS, Saag MS, Schechter M, Carpenter CCJ, et al. Treatment for adult HIV infection: 2004 recommendations of the International AIDS Society- USA panel. JAMA 2004; 292:251-265.
5. Podzamczer D, Ferrer E, Gatell JM, Niubo J, Dalmau D, Leon A, et al. Early virologic failure and with a combination of tenofovir, didanosine and efavirenz. Antivir Ther 2005; 10:171-177.
6. Johnson VA, Brun-Vezinet F, Clotet B, Conway B, D'Aquila RT, Demeter LM, et al. Update of the drug resistance mutations in HIV-1: 2004. Topics in HIV Med 2004; 12:119-124.
7. Back D, Blaschke T, Boucher C, Burger D, Fletcher C, Flexner C,
et al.
Optimizing TDM in HIV clinical care: a practical guide to performing therapeutic drug monitoring (TDM) for antiretroviral agents. Version 1.0. 2003. Available at:
http://www.hivpharmacology.com/.
8. Gerstoft J, Kirk O, Obel N, Pedersen C, Mathiesen L, Nielsen H, et al. Low efficacy and high frequency of adverse events in a randomized trial of the triple nucleoside regimen abacavir, stavudine and didanosine. AIDS 2003; 17:2045-2052.
9. Gallant JE, Rodriguez AZ, Weinberg W, Young B, Berger D, Lim ML, et al. Early non-response to tenofovir (TDF) + abacavir (ABC) and lamivudine (3TC) in a randomized trial compared to efavirenz (EFV) + ABC and 3TC: ESS30009 unplanned interim analysis. Presented at: 43rd Interscience Conference on Antimicrobial Agents and Chemotherapy. Chicago, September 2003 [abstract H-1722a].
10. Jemsek J, Hutcherson P, Harper E. Poor virologic responses and early emergence of resistance in treatment naïve, HIV-1 infected patients receiving a once daily triple nucleoside regimen of didanosine, lamivudine and tenofovir. Presented at: 11th Conference on Retroviruses and Opportunistic Infections. San Francisco, February 2004 [Abstract 51].
11. Gulick RM, Ribaudo HJ, Shimuka CM, Lustgarten S, Meyer WA, Klingman K, et al. Triple-nucleoside regimens versus efavirenz-containing regimens for the initial treatment of HIV-1 infection. N Engl J Med 2004; 350:1850-1861.
12. Leon A, Martinez E, Mallolas J, Laguno M, Blanco JL, Pumarola T, et al. Early virological failure in treatment-naive HIV-infected adults receiving didanosine and tenofovir plus efavirenz or nevirapine. AIDS 2005; 19:213-215.
13. Marzolini C, Telenti A, Decosterd LA, Greub G, Biollaz J, Buclin T. Efavirenz plasma levels can predict treatment failure and central nervous system side effects in HIV-1-infected patients. AIDS 2001; 15:71-75.
14. Alexander CS, Asselin JJ, Ting LS, Montaner JS, Hogg RS, Yip B, et al. Antiretroviral concentrations in untimed plasma samples predict therapy outcome in a population with advanced disease. J Infect Dis 2003; 188:541-548.
15. Ray A, Olson L, Fridland A. Role of purine nucleoside phosphorylase interactions between 2′,3′-dideoxyinosine and allopurinol, ganciclovir or tenofovir. Antimicrob Agents Chemother 2004; 48:1089-1095.
16. Kakuda T, Anderson P, Becker S. CD4 cell decline with didanosine and tenofovir and failure of triple nucleoside/nucleotide regimens may be related. AIDS 2004; 18:2442-2444.
17. Kearney B, Isaacson E, Sayre J. Didanosine and tenofovir DF drug-drug interaction: assessment of didanosine dose reduction. Presented at: 10th Conference on Retroviruses and Opportunistic Infections. Boston, February 2003 [abstract 533].
18. Rendon A, Barrios A, Gonzalez de Requena D, Maida I, Labarga P, Martin-Carbonero L, et al. Didanosine plasma levels in patients taking tenofovir concomitantly: high interindividual variability but association with hyperlactatemia. HIV Med 2004; 5:233 [abstract 81].
19. Robbins BL, Wilcox CK, Fridland A, Rodman JH. Metabolism of tenofovir and didanosine in quiescent or stimulated human peripheral blood mononuclear cells. Pharmacother 2003; 23:695-701.
20. Samarasinghe Y, Boffito M, Di Perri G. Nucleoside/nucleotide reverse transcriptase inhibitor drug interactions. J HIV Ther 2004; 9:79-86.
21. Negredo E, Molto J, Burger D, Viciana P, Ribera E, Paredes R, et al. Unexpected CD4 cell-count decline in patients receiving didanosine and tenofovir-based regimens despite undetectable viral load. AIDS 2004; 18:459-463.
22. Martinez E, Milinkovic A, De Lazzari E, Ravasi G, Blanco JL, Larrousse M, et al. Pancreatic toxic effects associated with co-adminstration of didansoine and tenofovir in HIV-infected adults. Lancet 2004; 364:8-10.
23. McColl DJ, Miller MD. The use of tenofovir disoproxil fumarate for the treatment of nucleoside-resistant HIV-1. J Antimicrob Chemother 2003; 51:219-223.
© 2005 Lippincott Williams & Wilkins, Inc.