Author + information
- Received January 15, 2010
- Revision received February 26, 2010
- Accepted March 12, 2010
- Published online June 15, 2010.
- Todd C. Villines, MD*,* (, )
- Eric J. Stanek, PharmD†,
- Patrick J. Devine, MD*,
- Mark Turco, MD‡,
- Michael Miller, MD§,
- Neil J. Weissman, MD∥,
- Len Griffen, MD¶ and
- Allen J. Taylor, MD*,∥
- ↵*Reprint requests and correspondence:
Dr. Todd C. Villines, Cardiology Service, Walter Reed Army Medical Center, 6900 Georgia Avenue, NW, Building 2, Room 4A-34A, Washington, DC 20307
Objectives This report describes the final results of the ARBITER 6-HALTS (Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol 6–HDL and LDL Treatment Strategies in Atherosclerosis) trial.
Background The ARBITER 6-HALTS trial was terminated early on the basis of a pre-specified interim analysis showing superiority of niacin over ezetimibe on change in carotid intima-media thickness (CIMT). After termination, an additional 107 subjects completed a close-out assessment.
Methods Patients with coronary heart disease (CHD) or CHD equivalent with low-density lipoprotein cholesterol <100 mg/dl and high-density lipoprotein cholesterol <50 mg/dl for men or 55 mg/dl for women while receiving stable statin treatment were randomly assigned to ezetimibe (10 mg/day) or extended-release niacin (target dose, 2,000 mg/day). The primary end point was change in mean CIMT, analyzed according to a last observation carried forward method. The relationships of study medication adherence, dosage, and cumulative exposure (product of adherence, dose, and time) with change in CIMT were explored.
Results Results in 315 patients included 208 with 14-month follow-up and 107 after mean treatment of 7 ± 3 months. Niacin (n = 154) resulted in significant reduction (regression) in mean CIMT (−0.0102 ± 0.0026 mm; p < 0.001) and maximal CIMT (−0.0124 ± 0.0036 mm; p = 0.001), whereas ezetimibe (n = 161) did not reduce mean CIMT (−0.0016 ± 0.0024 mm; p = 0.88) or maximal CIMT (−0.0005 ± 0.0029 mm; p = 0.88) compared with baseline. There was a significant difference between ezetimibe and niacin treatment groups on mean changes in CIMT, favoring niacin, for both mean CIMT (p = 0.016) and maximal CIMT (p = 0.01). Increased cumulative drug exposure was related to regression of CIMT with niacin, and progression of CIMT with ezetimibe.
Conclusions Niacin induces regression of CIMT and is superior to ezetimibe for patients taking statins. (Comparative Study of the Effect of Ezetimibe Versus Extended-Release Niacin on Atherosclerosis; NCT00397657)
Despite significant reductions in major cardiovascular events, lipid abnormalities and residual risk in patients receiving statin monotherapy are prevalent (1). We conducted a comparative-efficacy trial examining the addition of ezetimibe or extended-release niacin (ERN) on carotid intima-media thickness (CIMT) in high-risk patients on stable, chronic statin monotherapy (2). Results within the initial 208 patients completing the 14-month trial demonstrated superiority of ERN over ezetimibe in the change in CIMT, and the trial was stopped early in a pre-specified interim analysis (3). Herein, we extend our initial report with results on final CIMT imaging among all patients enrolled in the ARBITER 6-HALTS (Arterial Biology for the Investigation of the Treatment Effects of Reducing Cholesterol 6–HDL and LDL Treatment Strategies) trial. These results include an analysis of the last-observed CIMT assessment for the full trial sample, and the effect on CIMT of study medication adherence, dose optimization, and treatment duration.
The rationale, design, and primary results of the ARBITER 6-HALTS study have been described (2,3). Briefly, inclusion criteria were as follows: 1) patients at least 30 years of age with known atherosclerotic cardiovascular disease or coronary heart disease (CHD) equivalent; 2) currently taking statin monotherapy at a consistent dose; 3) low-density lipoprotein cholesterol (LDL-C) <100 mg/dl (2.6 mmol/l); and 4) high-density lipoprotein cholesterol (HDL-C) <50 mg/dl in men or <55 mg/dl in women (1.3 or 1.4 mmol/l, respectively), documented on a lipid panel within 3 months of enrollment.
Eligible patients were randomly assigned to open-label therapy with either ezetimibe (10 mg/day) or ERN (target dose 2,000 mg/day). The ERN was initiated at 500 mg/day taken at bedtime, and increased by 500 mg every other week to the maximum tolerated dose. There were no protocol-directed changes in statin medications or dosage throughout the study. Adherence to study medication was determined through tablet counts. The primary end point was the between-group difference in the change in mean CIMT after 14 months.
B-mode ultrasonography of the carotid arteries
Carotid ultrasound examinations were performed at baseline, 8 months, and 14 months. We have previously described the protocol for CIMT assessment utilized in this study (2,3). In summary, mean and maximal diastolic CIMT of the distal 1 cm of the far wall of the right and left common carotid arteries was measured by a single, blinded, trained observer utilizing an automated border-detection algorithm. All images were obtained in duplicate, and no scans were excluded on the basis of image quality.
Between-group data for continuous variables were evaluated using a ttest for independent variables or a Mann-Whitney Utest, as appropriate. Categorical variables were evaluated using the chi-square test. Within-group comparisons of continuous variables were performed using a paired ttest or Wilcoxon signed-rank test, as appropriate. A 2-sided p value ≤0.05 was considered statistically significant. The SPSS for Windows version 16 software (SPSS Inc., Chicago, Illinois) was used for all statistical analyses. On the basis of a minimum sample size of 150 per group, the trial had an 80% power to detect a difference of CIMT change between agents of 0.02 ± 0.06 mm/year (α = 0.05).
As previously described (2,3), the study design pre-specified the performance of a blinded, interim analysis according to the conservative method of O'Brien and Fleming (4) with an alpha spending function, to be conducted after 180 subjects (60% of the planned sample size) had completed the trial. After the interim analysis (March 2009), an independent data advisory committee (June 4, 2009) evaluated the primary end point data without knowledge of treatment assignment, and based upon the trial findings, unanimously recommended that the trial should be terminated.
After study termination on June 4, 2009, all actively enrolled patients were contacted and returned for final collection of clinical variables, laboratory data, and blinded CIMT assessment. Among the 363 patients initially enrolled in the trial, 208 patients had completed the entire 14 months of the study period at the time of their final visit (111 ezetimibe, 97 ERN), and 44 had left the study. Final ultrasound examinations could not be obtained in 4 additional subjects after termination of the study (total of 48 participants dropped out), leaving 315 patients for this analysis. In 107 of the 315 patients analyzed, final achieved lipid values and CIMT measurements were performed after study termination at a mean treatment duration of 7 ± 3 months, and these values are included in the primary end point (a last observation carried forward analysis).
The baseline characteristics of the 363 patients enrolled in the trial were similar between the 2 treatment groups (Table 1).A majority of the patients were male (80%), hypertensive (87%), age 65 ± 11 years, and had taken a statin (atorvastatin or simvastatin by 95%) at a mean dose of 42 mg for 6 ± 5 years. There was no difference with respect to age, baseline lipid values, and CIMT between the 48 participants who did not complete the trial and the remainder of the trial participants. In addition, there was no significant difference among baseline covariates, including demographics, blood pressure, lipid levels, baseline therapies, or baseline CIMT between patients who completed the entire 14-month trial (n = 208) and patients who completed the trial after study termination (n = 107). Baseline and final lipid and biomarker values in the 315 patients who completed the trial are shown in Table 2.Significant reductions in baseline LDL-C and triglycerides were seen with both ezetimibe and ERN. As compared with ERN, patients treated with ezetimibe achieved significantly lower total cholesterol, LDL-C, and HDL-C, and had higher triglyceride values. There was no difference in baseline or final fasting glucose values between the study groups. There was no significant difference between groups in clinically directed changes in the statin drug or dose during the study. Mean study drug adherence rates over the duration of the study exceeded 80% in each arm, and were significantly higher with ezetimibe as compared with ERN (87.5 ± 15.3% vs. 82.1 ± 17.2%, respectively; p = 0.005).
Primary end point
The primary end point was assessed among all randomized subjects (n = 315) who completed a final CIMT measurement after either 14 months (n = 208) or after <14 months of treatment (n = 107) among subjects who had not yet completed the study at the time of its termination (Table 3).Treatment with niacin (n = 154) resulted in significant reduction (regression) in mean CIMT (−0.0102 ± 0.0026 mm; p < 0.001) and maximal CIMT (−0.0124 ± 0.0036 mm; p = 0.001) compared to baseline. Treatment with ezetimibe (n = 161) had no effect on mean CIMT (−0.0016 ± 0.0024 mm; p = 0.88) or maximal CIMT (−0.0005 ± 0.0029 mm; p = 0.88) compared to baseline. There was a significant difference between the ezetimibe and niacin treatment groups on the change in CIMT, favoring niacin for both mean CIMT (p = 0.016) and maximal CIMT (p = 0.01) among the 315 patients completing the trial. Imputing baseline CIMT as the final CIMT (last observation of baseline CIMT carried forward) in the 48 patients who did not complete the trial resulted in no change in the primary outcome.
Impact of cumulative drug exposure
Based on differences in study medication adherence, ERN dosage achieved and treatment duration (due to early trial termination) between trial participants, we performed an exploratory analysis to assess the impact of these variables on change in CIMT among niacin and ezetimibe treatment groups. The product of study drug adherence, dosage, and treatment duration was calculated to estimate cumulative exposure to study drug as an integrated measure of drug effect (5). The relationships between increasing cumulative drug exposure (lowest, quartile 1 to highest, quartile 4) and change in mean CIMT for all subjects using the method of last observation carried forward are shown in Table 4and Figure 1.Increased cumulative drug exposure resulted in regression of CIMT with niacin, and progression of CIMT with ezetimibe. Specifically, comparing participants treated optimally (best-case comparison), defined as those with the highest quartile of study drug adherence and those reaching the target 2,000 mg/day in the niacin group (quartile 4) for the entire 14 months of the study, patients treated with ERN had significant reduction (regression) of mean CIMT from baseline (−0.0128 ± 0.0078 mm), whereas patients treated with ezetimibe experienced CIMT progression (0.0067 ± 0.0059 mm).
We recently demonstrated that ERN leads to significant regression of CIMT and was superior to ezetimibe when added to chronic statin therapy among 208 high-risk patients who had completed 14 months of randomized, open-label, parallel group treatment in the ARBITER 6-HALTS trial (3). This comparative-efficacy trial was stopped before all enrolled patients completing the intended 14 months of drug treatment after a planned interim analysis clearly demonstrated the superiority of ERN over ezetimibe in the effect on CIMT. In the current analysis, we extend our original observations by examining the effect of the study treatments on CIMT within a last observation carried forward and therapy optimization analysis.
Consistent with the results seen for patients completing 14 months of treatment, among the entire study group (n = 315) including 208 subjects treated for the full study duration and 107 subjects treated for a mean of 7 months, therapy with ERN led to significant regression of baseline CIMT among statin-treated patients with an LDL-C <100 mg/dl and an HDL-C <50 or 55 mg/dl. Treatment with ezetimibe did not significantly change CIMT. Comparatively, there was a significant difference between the effect of niacin and ezetimibe on changes in CIMT, favoring niacin. These results strengthen the findings from the interim analysis through the inclusion of patients treated for shorter times (last observation carried forward analysis), which might have been expected to reduce the magnitude of effect between the 2 treatment strategies originally observed. This is a noteworthy possibility within the trial design of ARBITER 6-HALTS with an 8-week titration period for ERN to the target dose of 2,000 mg/day, as opposed to ezetimibe, which was initiated at the maximal clinical dose. Final results confirm the difference in the primary outcome and preserved magnitude of CIMT regression among patients taking ERN. In addition, the magnitude of effect of ERN on mean CIMT, producing regression (−0.0142 ± 0.0041 mm), over 14 months of treatment in the relatively high risk ARBITER 6-HALTS study population is noteworthy in comparison with other contemporaneous lipid-lowering trials using similar CIMT methodology. In the METEOR (Measuring Effects of Intima-Media Thickness: An Evaluation of Rosuvastatin) trial, treatment for 24 months with high-potency statin monotherapy (rosuvastatin 40 mg; mean final LDL 78 mg/dl) in a low-risk population resulted in a change in mean common CIMT of 0.0004 mm (95% confidence interval: −0.0011 to 0.0019 mm), as compared to placebo 0.0088 mm (95% confidence interval: 0.0064 to 0.0112 mm) (6).
The observation that the CIMT response was related to niacin adherence, dose, and increased treatment duration, calculated as cumulative drug exposure, is consistent with a drug effect as shown in prior studies demonstrating a favorable impact of niacin on clinical events and atherosclerosis (7–9). A prior study has shown regression of CIMT when ERN adherence rates are high, even when the dose was <2,000 mg/day (8,9). Additional studies examining the optimal and minimal effective dose of ERN on atherosclerosis, and ultimately clinical events, are warranted.
The relationship between cumulative drug exposure and the CIMT effect of ERN supports an expected, direct relationship between increasing intensity of drug exposure (through a composite of dose, adherence, and time) and its effect on atherosclerosis. In contrast, findings with ezetimibe showing an unexpected inverse relationship between intensity of drug exposure and CIMT draw further attention to a growing body of evidence on the diverse effects of ezetimibe on cholesterol transport mechanisms, such as interference with the pivotal HDL receptors SRB-1, and ABCA1 (10,11). Although the net health impact of ezetimibe's described off-target receptor effects are yet to be fully understood, this evolving science clearly indicates that the pharmacologic effects of ezetimibe extend beyond the simple inhibition of enteric cholesterol absorption, and reduction of LDL-C.
A limitation of this study is the use of CIMT as a surrogate for clinical end points. While the preponderance of studies demonstrate the validity of CIMT as a surrogate for cardiovascular events (3,12–14), the ultimate net health impact of therapeutics requires clinical end point trials. Additionally, our analysis evaluating cumulative drug exposure is post-hoc and exploratory in nature. Sample size is limited within ERN subjects who did not achieve the 2,000 mg/day target dose, and such subjects also were less adherent to the therapy. Prior studies in highly adherent subjects treated for 1 to 2 years at the 1,000 mg/day dose showed regression of CIMT (8,9). Dose-ranging studies would be useful to further elucidate dose response effects of ERN on CIMT.
Final results from the ARBITER 6-HALTS trial confirm the superiority of extended-release niacin over ezetimibe for the end point of change in CIMT and the ability of niacin to induce CIMT regression. Increased cumulative drug exposure was related to regression of CIMT with niacin, and progression of CIMT with ezetimibe.
Abbott Pharmaceuticals funded the study by an unrestricted, investigator-initiated research grant administered by the Henry M. Jackson Foundationfor the Advancement of Military Medicine, Rockville, Maryland. Dr. Villines has received lecture honoraria from Novartis Pharmaceuticals. Dr. Stanek is an employee, with significant equity holdings, of Medco Health Solutions Inc. Dr. Miller has received speaking honoraria from Merck. Dr. Turco received consulting fees and lecture honoraria from Abbott Cardiovascular. Dr. Taylor receives lecture honoraria from Abbott that are donated to charity. The opinions or assertions herein are the private views of the authors and are not to be construed as reflecting the views of the Department of the Army or the Department of Defense.
- Abbreviations and Acronyms
- coronary heart disease
- confidence interval
- carotid intima-media thickness
- extended-release niacin
- high-density lipoprotein cholesterol
- highly-sensitive C-reactive protein
- interquartile range
- low-density lipoprotein cholesterol
- Received January 15, 2010.
- Revision received February 26, 2010.
- Accepted March 12, 2010.
- American College of Cardiology Foundation
- Taylor A.J.,
- Sullenberger L.E.,
- Lee H.J.,
- Lee J.K.,
- Grace K.A.