The US Department of Health and Human Services (DHHS) has recommended that regimens be simplified by reducing the number of pills and dosing frequency and minimizing drug interactions and side effects to improve the rate of adherence.1 Fixed-dose combinations (FDCs) that help to address these needs by reducing pill burden and dosing frequency represent an important advance in HIV-1 treatment.
Abacavir (ABC; Ziagen, GlaxoSmithKline) is a nucleoside reverse transcriptase inhibitor (NRTI) that has been in use since 1998. The original dosing guidelines were for 300 mg of ABC administered twice daily; however, recent studies have confirmed that ABC can be effectively administered at a dose of 600 mg once daily, and once-daily use was approved by the US Food and Drug Administration (FDA) in 2004. Another NRTI, lamivudine (3TC; Epivir, GlaxoSmithKline), was approved by the FDA in 1995 for use at 150 mg administered twice daily, and a 300-mg formulation for once-daily use was approved in 2000.
ABC and 3TC have been prescribed extensively as separate agents. More recently, ABC and 3TC have been prescribed together as a dual-nucleoside backbone for protease inhibitor (PI)- and nonnucleoside reverse transcriptase inhibitor (NNRTI)-based highly active antiretroviral therapy (HAART) regimens. More than 1000 patients were treated with the NNRTI efavirenz in combination with ABC and 3TC in studies CNA30021 and CNA30024.2,3 Approximately 900 patients were evaluated in the fosamprenavir pivotal studies NEAT and SOLO, which administered ABC and 3TC in combination with fosamprenavir or nelfinavir.4,5 These studies confirm the clinical utility of an NRTI backbone composed of ABC and 3TC in combination with PIs and NNRTIs.
The FDC tablet of ABC + 3TC, as a component of combination antiretroviral therapy (ART), has the benefit of simplifying a regimen already used in clinical practice and may provide further adherence and efficacy advantages. The ABC + 3TC FDC is manufactured by GlaxoSmithKline and sold in the United States under the trade name Epzicom (EPZ) and in other regions as Kivexa.
ESS30008 was a phase 3, randomized, open-label, 48-week study comparing an FDC tablet of ABC + 3TC administered once daily with the separate components of ABC and 3TC administered twice daily as part of a regimen containing a PI or NNRTI. Eligible subjects were 18 years of age or older, HIV-1 infected, and taking an initial antiretroviral regimen composed of twice-daily ABC and 3TC in combination with a PI or NNRTI for at least 6 months before screening. Virologic suppression (defined as HIV-1 RNA level <400 copies/mL) must have been demonstrated for at least the previous 3 months before and at screening. CD4+ counts were required to be ≥50 cells/mm3 at screening. Exclusion criteria included a history of a Centers for Disease Control and Prevention (CDC) category C event within 45 days of screening, a pregnant or breast-feeding woman, clinically relevant pancreatitis or hepatitis within 6 months of screening, and anticipated inability to complete the 48-week study. Subjects were ineligible if they had any of the following laboratory results within 28 days of baseline: hemoglobin <10 g/dL (male) or <9 g/dL (female), absolute neutrophil count <1000 cells/mm3, platelet count <75,000 cells/mm3, aspartate aminotransferase (AST)/alanine aminotransferase (ALT) >5 times the upper limit of normal (ULN), serum pancreatic amylase >1.5 times the ULN, or serum creatinine >1.5 times the ULN.
This was a multicenter study conducted at 62 sites in the United States, Panama, Costa Rica, and Puerto Rico. The study was approved by institutional review boards (IRBs) at participating sites, and all participants provided written informed consent.
Subjects were randomized 1:1 to continue taking the separate components ABC and 3TC twice daily or to switch to EPZ once daily, in combination with the entry PI or NNRTI. Randomization was stratified based on PI or NNRTI use at entry. Within-class substitutions were allowed for the PI or NNRTI; however, no substitutions were permitted for randomized treatment (ABC, 3TC, or EPZ).
Assessments and Outcomes
A screening visit was conducted within 28 days of study entry. Study visits occurred at entry (baseline) and at weeks 4, 8, 12, 16, 20, 24, 32, 40, and 48. Clinical and laboratory assessments (clinical chemistry, hematology, and plasma HIV-1 RNA level) were conducted at all visits. Pregnancy tests were performed before entry and every 24 weeks. CD4+ and CD8+ cell counts were performed at screening; baseline; and weeks 12, 24, and 48. Plasma for genotypic drug resistance testing was collected at baseline and at any confirmatory visits for virologic failure. Pill counts to assess adherence to randomized therapy were conducted at all postbaseline visits.
All laboratory testing was performed centrally (Quest Laboratories, Van Nuys, CA). Genotypic analyses were performed by GlaxoSmithKline Clinical Virology (Research Triangle Park, NC) using the ViroSeq HIV-1 Genotyping system.
The primary efficacy end point was the proportion of responders (defined as subjects who did not meet the protocol definition of virologic failure) at week 48 (intent-to-treat [ITT] analysis). Virologic failure was defined as confirmed plasma HIV-1 RNA level ≥1265 copies/mL (0.5 log10-copies/mL increase from 400 copies/mL). All subjects with confirmed virologic failure were withdrawn from the study. The primary safety end point was the frequency of treatment-limiting adverse events (AEs), grade 2 through 4 AEs, and serious AEs.
Secondary end points included the proportion of subjects with HIV-1 RNA levels <400 and <50 copies/mL, change in CD4+ cell count from baseline, time to clinical disease progression or death, medication adherence, and development of genotypic resistance at virologic failure.
Efficacy and safety analyses included all subjects (ITT). A sample size of 240 (120 subjects per treatment group) provided at least 80% power to establish the noninferiority of EPZ administered once daily compared with ABC + 3TC administered twice daily for the primary analysis comparing the proportion of responders (nonvirologic failures) through week 48 between treatment groups. The sample size was based on an ITT analysis in which patients who dropped out or were lost to follow-up were counted as having failed treatment. For the primary efficacy analysis, noninferiority was established if the lower 95% confidence boundary (equivalent to a 2-sided 90% confidence interval [CI]) for the difference in proportions (EPZ once daily − ABC + 3TC twice daily) exceeded −12%.
For the virologic efficacy analyses, the ITT missing equals failure (M = F) data set was used, in which all missing assessments were considered failures. In the primary analysis (proportion of nonvirologic failures), missing assessments were considered to represent >1265 copies/mL. For the proportions analyses, the ITT M = F and as-treated (AT) data sets were used. In the AT analysis, missing assessments were considered unevaluable.
For the safety analyses, the ITT observed data set was used, which does not include missing values. AEs were graded by the investigator according to the Division of AIDS toxicity table (1993) and coded by an AE dictionary (MedDRA).
Adherence was calculated as the ratio of the total number of doses taken to the total number of doses that should have been taken. The numerator of the ratio was calculated as the number of tablets dispensed minus the number of tablets returned. For each subject, the denominator was equal to the number of days between study start and week 48 (or premature discontinuation) multiplied by the number of doses that should have been taken each day. Maximum adherence rates were truncated at 100%.
Genotypic summaries were based on differences in the plasma virus sequence from the molecular wild-type strain HXB2. Reverse transcriptase (RT) and protease (PRO) mutations associated with the development of resistance to ART as listed by the International AIDS Society (IAS) Drug Resistance Mutation Group were used in the analyses.6
All analyses were performed using SAS version 8 (SAS Institute, Cary, NC) on a system of UNIX computers.
Two hundred sixty subjects were enrolled between September 2002 and May 2003, and all were exposed to study medication. Table 1 summarizes the demographics and baseline characteristics, which were well matched between the treatment groups. Overall, the study population was predominantly male (82%) and racially diverse, with approximately equal proportions of black (27%), white (38%), and Hispanic (34%) subjects. At study entry, the median HIV-1 RNA level was below the limit of detection (50 copies/mL) and the median CD4+ count was 554 cells/mm3. Median time on a regimen containing ABC and 3TC before study entry was 22 months. Table 2 summarizes the background antiretroviral agents taken by subjects during study. Most subjects were taking an NNRTI, and the most commonly used background antiretroviral agents were efavirenz (62%), fosamprenavir (17%), and nelfinavir (14%).
Subject disposition through week 48 is shown in Table 3. In summary, 92% of subjects in the EPZ once-daily group and 90% of subjects in ABC + 3TC twice-daily group completed the study.
In the primary efficacy analysis, 95% (123 of 130) of subjects treated with EPZ administered once daily were responders (nonvirologic failures) compared with 93% (121 of 130) of subjects treated with the separate components twice daily (ITT M = F). The EPZ once-daily minus ABC + 3TC twice-daily difference estimate was 1.5, with the 90% CI (−3.4, 6.4) establishing noninferiority of EPZ administered once daily. An exploratory analysis using a stricter definition of virologic failure, HIV-1 RNA level ≥50 copies/mL, confirmed the noninferiority of EPZ administered once daily compared with ABC + 3TC administered twice daily (90% CI: −1.4, 12.2).
The proportions of subjects with an HIV-1 RNA level <400 and <50 copies/mL through week 48 are shown in Figure 1. At week 48, 81% and 82% of subjects had an HIV-1 RNA level <50 copies/mL in the EPZ once-daily and ABC + 3TC twice-daily treatment groups, respectively (ITT M = F). In the AT population, 100% (109 of 109) of the EPZ once-daily group and 99% (108 of 109) of the ABC + 3TC twice-daily group had an HIV-1 RNA level <50 copies/mL at week 48. There was not a significant difference in response based on background ART strata (PI or NNRTI).
Immunologic responses were stable. Median CD4+ cell counts were high at baseline (EPZ once daily: 565 cells/mm3; ABC + 3TC twice daily: 549 cells/mm3), and median changes from baseline at week 48 were negligible (−28 and −24 cells/mm3 for EPZ once daily and ABC + 3TC twice daily, respectively; ITT observed). The slight reduction in median CD4+ cell counts was driven by subjects who entered with CD4+ counts greater than 500 cells/mm3 and by intrasubject and assay variability. There was not a consistent decline in CD4+ cell counts within individual subjects.
The overall incidence of AEs as well as the incidence of grade 2 through 4 AEs was comparable between treatment groups. No single grade 2 through 4 AE was reported by ≥5% of study participants; the most common AEs were headache (5% EPZ, 4% ABC + 3TC), nasopharyngitis (4% EPZ, 5% ABC + 3TC), upper respiratory infection (4% EPZ, 3% ABC + 3TC), depression (2% EPZ, 4% ABC + 3TC), sinusitis (3% EPZ, 2% ABC + 3TC), and diarrhea (3% EPZ, 2% ABC + 3TC). Although 2 subjects in the ABC + 3TC twice-daily group discontinued the study because of AEs, those events were not related to treatment. Serious AEs were reported by 7 subjects (5 in EPZ once-daily group and 2 in ABC + 3TC twice-daily group); however, none were treatment related. No hypersensitivity reactions to ABC were reported in either treatment group. Two pregnancies (1 in each treatment arm) were reported for study subjects: 1 subject discontinued the study, and the other pregnancy occurred after the treatment phase of the study.
Treatment-emergent grade 3 to 4 laboratory abnormalities were infrequent and comparable between treatment groups. Grade 3 to 4 cholesterol was reported by 6% of EPZ once-daily subjects and 5% of ABC + 3TC twice-daily subjects. The only other grade 3 to 4 laboratory abnormality reported by ≥5% of subjects was elevated triglycerides, which was reported in 4% and 7% of the EPZ once-daily and ABC + 3TC twice-daily subjects, respectively. Concurrent medications to treat lipid elevations were permitted, and the prevalence of lipid-reducing medications was comparable between groups: 18% in the EPZ once-daily group and 21% in the ABC + 3TC twice-daily group.
Over the 48-week study, median adherence with the ABC and 3TC components of the regimen was 93% in both treatment groups as well as 93% within the PI and NNRTI strata of each group. A larger proportion of subjects in the EPZ once-daily group (39%) achieved ≥95% adherence compared with the ABC + 3TC twice-daily group (31%).
Of the 260 randomized subjects, 6 (2.3%) were identified as meeting criteria for virologic failure (2 subjects in the EPZ once-daily group, 4 subjects in the ABC + 3TC twice-daily group). All 6 subjects had documented nonadherence to ART before virologic failure. Five subjects had baseline genotypes, and all 6 had on-treatment genotypes at the virologic failure time point available for comparison (Fig. 2).
There was no evidence of ABC- and/or 3TC-related mutations at baseline in 5 of 6 subjects who met virologic failure criteria (baseline genotype not available for 1 subject). Of the subjects meeting virologic failure criteria, 4 acquired ABC- and/or 3TC-associated mutations while on randomized treatment (3 subjects with M184V and 2 subjects with K65R). Neither subject with K65R at failure had evidence of K65R at baseline. Both subjects had documented viral suppression on their previous ABC + 3TC twice-daily regimen and entered the study with an HIV-1 RNA level <400 copies/mL. One subject misdosed the study drug and took dual therapy (900 mg of ABC with 300 mg of 3TC) for the first 12 weeks. The other subject reported nonadherence to study medications.
The durability of a nucleoside backbone composed of ABC and 3TC was demonstrated in both treatment groups. The median time on ABC and 3TC at the time of study entry was 22 months; thus, at the end of this 48-week study, subjects were approaching a median of 3 years on therapy. At week 48, rates of virologic suppression to <50 copies/mL were high: 81% for EPZ once daily and 82% for ABC + 3TC twice daily (ITT M = F). EPZ administered once daily was established as not inferior to continued treatment with ABC + 3TC administered twice daily at week 48. Virologic failure was rare in both treatment groups and could be attributed to nonadherence with randomized or background therapy in all cases.
The incidence of grade 2 through 4 AEs was low and comparable between treatment groups. In this population of ABC-experienced subjects, no cases of hypersensitivity were reported. Additionally, no subject discontinued because of a treatment-related AE, and no treatment-related serious AEs or deaths occurred on study.
Although median adherence was 93% in both treatment groups, the proportion of subjects achieving the level of adherence shown to optimize virologic control (≥95%) was higher in the EPZ once-daily group (39%) than in the ABC + 3TC twice-daily group (31%). The adherence benefit is likely to be even greater in the real world, where FDC products not only reduce pill burden but reduce the number of prescriptions and the number of copayments. Replacing 2 medications with 1 eliminates the potential for differential adherence to the components and therefore avoids the possibility of the prescription refill schedules becoming desynchronized. Such real-world adherence benefits are suggested by 2 database studies in which patients receiving the FDC product Combivir (3TC and zidovudine; GlaxoSmithKline) were 2 to 3 times more likely to achieve optimal adherence than patients given the component medications.7,8 Additionally, although the proportion of subjects with ≥95% adherence was 31% to 39%, a much higher proportion of subjects (81% [ITT M = F]) maintained a plasma HIV-1 RNA level <50 copies/mL. These results suggest an acceptable forgiveness for the administration of ABC + 3TC twice daily and EPZ once daily in this patient population.
Of the 260 enrolled subjects, 6 (2.3%) met virologic failure criteria. Four subjects acquired ABC- and/or 3TC-associated mutations (3 subjects with M184V and 2 subjects with K65R). Although the M184V mutation has been commonly seen with 3TC use, the K65R mutation is rarely acquired in treatment regimens, including the ABC + 3TC combination. In both subjects with the K65R mutation at failure, incidences of drug misdosing and nonadherence were noted.
Patients controlled on twice-daily therapy with individual products can be switched to a once-daily FDC without a reduction in virologic control or an increase in toxicity. The FDC tablet of ABC + 3TC provides an effective, well-tolerated, and convenient dual-nucleoside backbone option for creating PI- or NNRTI-containing regimens.
The authors thank the study participants, clinical investigators, study coordinators, and GlaxoSmithKline staff who made this study successful. Participating investigators were J. Adams, J. Bartlett, S. Becker, N. Bellos, R. Bolan, J. Brand, A. Burnside, P. Cook, P. Daly, E. DeJesus, D. Donnell, R. Dretler, M. Eyster, M. Fischl, J. Flamm, G. Frechette, J. Gathe Jr, J. Giron, E. Godofsky, H. Grossman, F. Haas, G. Herrera, F. Kramer, P. Kumar, A. LaMarca, J. Lang, C. Lucasti, G. McKinley, P. McLeroth, A. Mestre, L. Miller, M. Mogyoros, R. Myers, R. Nahass, C. Newman, D. Parks, G. Richmond, A. Rodriguez, J. Rodriguez, P. Salvato, M. Sands, J. Santana, K. Sathasivam, S. Schneider, R. Schwartz, M. Sension, G. Sepulveda, G. Simon, N. Sosa, R. Steigbigel, C. Steinhart, D. Sweet, M. Thompson, J. Torres, C. Walworth, D. Ward, W. Weinberg, B. Yangco, and B. Young.