Reduction in Carotid Arterial Wall Thickness Using Lovastatin and Dietary Therapy

Controlled clinical trials to assess the effects of therapy for atherosclerosis have generally used either clinical cardiovascular events or serial arterial imaging as end points. Serial coronary angiographic assessment of lesion change has been the most commonly used end point in arterial imaging trials [1]. Evidence now indicates that high-resolution B-mode ultrasonographic measurement of the combined thickness of the carotid arterial intima-media complex is a reliable end point for lipid-lowering interventional trials [1]. Unlike angiographic imaging procedures that focus on changes in the lumen, noninvasive imaging of the arterial wall 1) directly quantifies the response of early atherosclerotic changes to risk-factor modification, 2) can be done as often as necessary in symptomatic or asymptomatic patients of any age, and 3) carries negligible risk [2]. In addition, the measurement of distal common carotid arterial far wall intima-media thickness (carotid arterial intima-media thickness) by automated computerized edge detection varies less than angiographic measurement of the coronary arteries; therefore, a much smaller sample size is needed to show a therapeutic effect [1, 3-5]. The Cholesterol Lowering Atherosclerosis Study (CLAS), a randomized, placebo-controlled, serial arterial imaging clinical trial [6], was designed to comprehensively survey the effects of colestipol-niacin plus dietary therapy on the progression of atherosclerosis in the coronary [7-9], femoral [10], and carotid arterial beds [3, 4]. That study was the first randomized, controlled clinical trial to provide evidence for a drug-induced reduction in carotid arterial intima-media thickness using B-mode ultrasonographic imaging as an outcome measurement of change in early, preintrusive atherosclerosis (the stage at which atherosclerosis is limited to the arterial wall and does not intrude into the arterial lumen) [3, 4]. The serial ultrasonographic results showed the powerful advantage of modeling sequential change over time [4]. The Monitored Atherosclerosis Regression Study (MARS) was a randomized, double-blind, placebocontrolled, serial arterial imaging clinical trial in which carotid arterial intima-media thickness was measured every 6 months with B-mode ultrasonography. Similar in design to CLAS, MARS was an independent clinical trial that showed that lovastatin plus dietary therapy reduced the progression of lesions of the coronary arteries as determined by both visual assessment of coronary artery change and quantitative coronary angiography [11, 12]. The ultrasonographic methods and findings from CLAS were replicated in a subgroup of 30 patients from MARS who, like the CLAS participants, were nonsmoking men who had had coronary arterial bypass grafts [1]. In this paper, we extend the findings of these studies to the complete cohort of patients from MARS with serial carotid arterial ultrasonographic measurements. We report evidence of benefit from lovastatin plus dietary therapy on early, preintrusive atherosclerosis through the measurement of change in the carotid arterial intima-media thickness in 188 patients evaluated every 6 months for as long as 4 years. We also relate these changes in early atherosclerosis to clinical, lipid, lipoprotein, and apolipoprotein risk factors for atherosclerosis. Methods Design of the Monitored Atherosclerosis Regression Study This study was a randomized, double-blind, placebo-controlled serial coronary angiographic and carotid arterial ultrasonographic imaging clinical trial that has been described previously [12]. Two hundred seventy patients (91% male, smokers and nonsmokers, 37 to 67 years of age) with total serum cholesterol levels between 4.92 and 7.64 mmol/L (190 mg/dL to 295 mg/dL) were randomly assigned to receive either lovastatin, 80 mg/d, or placebo. Two hundred fifteen patients were at the University of Southern California, and 55 were at the University of Wisconsin. The lovastatin group and the placebo group had identical dietary goals: no more than 250 mg of cholesterol per day, 27% of energy as total fat calories, 7% of energy as saturated fat calories, 10% of energy as monounsaturated fat calories, and 10% of energy as polyunsaturated fat calories. All patients had angiographically confirmed coronary artery disease in at least two segments with at least a 50% stenosis. Coronary angiography was done under standardized conditions at baseline and at 2 and 4 years after randomization; the 2-year results have been reported previously [11]. The 215 patients at the University of Southern California also had B-mode carotid ultrasonography, the results of which were used to measure carotid arterial intima-media thickness at baseline and every 6 months throughout the trial. Of these 215 patients, 196 (91%) had angiography at 2 years; 58% of the patients (77 receiving lovastatin and 48 receiving placebo) agreed to participate in an optional, double-blind, 2-year extension of the original randomized trial. An independent External Data and Safety Monitoring Committee recommended discontinuing the 2-year extension because of the observed treatment benefit in the per-patient reduction in stenosis for high-grade lesions as determined by quantitative coronary angiography, and in a panel assessment of overall change in coronary artery lesions [11]. Sixty-nine (32%) of the 215 patients at the University of Southern California had angiography at 4 years. Patients in the lovastatin group who participated in the 2-year extension had on-trial lipid levels, coronary artery end points, and rates of change in carotid arterial intima-media thickness similar to those of the treated patients who chose not to participate in the extension. Patients in the placebo group who chose not to participate in the extension had poorer coronary outcome at 2 years than did the patients who participated (P = 0.02), but both participants and nonparticipants had similar on-trial lipid levels and equivalent rates of change in carotid arterial intima-media thickness (0.02 0.04 mm/y). Lipid, Lipoprotein, and Apolipoprotein Levels Total serum cholesterol and total serum triglyceride levels were measured using an enzymatic method under the Standardization Program of the National Centers for Disease Control and Prevention; patients fasted 8 hours before samples for analysis were collected. High-density lipoprotein (HDL) cholesterol levels were measured after apolipoprotein B-containing lipoproteins were precipitated in whole plasma with heparin manganese chloride. Low-density lipoprotein (LDL) cholesterol levels were estimated using the Friedewald equation [12, 13]. Plasma apolipoprotein A-I, B, C-III, and E levels [14-17] were measured using electroimmunoassay at the Oklahoma Medical Research Foundation. Apolipoprotein C-III levels were measured in whole plasma as well as in heparin manganese supernatants (apolipoprotein C-III-HS) and heparin manganese precipitates (apolipoprotein C-III-HP) as previously described [18]. Apolipoprotein C-III-HS approximates apolipoprotein C-III in HDL, and apolipoprotein C-III-HP approximates apolipoprotein C-III contained in very-low-density lipoprotein (VLDL) plus LDL. Lipid, lipoprotein, and apolipoprotein levels were measured at baseline, and then lipid levels were measured every 2 months and lipoprotein and apolipoprotein levels were measured every 4 months throughout the study. Ultrasonography and Image Analysis The methods for ultrasonography and image analysis for carotid arterial intima-media thickness and their reproducibility have been previously described [3, 5]. B-mode scanning was done with a Diasonics CV400 ultrasound system with a 7.5-MHz probe (Diasonics, Milpitas, California). Longitudinal views of the near wall and the far wall of the right distal common carotid artery were recorded with the minimum gain needed to clearly visualize structures. An image analyst, blinded to treatment assignment, measured the distal common carotid arterial far wall intima-media thickness by automated computerized edge detection using a 386/33 PC computer equipped with a Data Translation DT 2862 image-processing board (Data Translation, Marboro, Massachusetts). The automated computerized edge-finding algorithm results in closely spaced measurements of intima-media thickness, approximately 100 to 120 points/cm, from which the average intima-media thickness is determined. The distance between the echoes arising from the blood-intima interface and from the media-adventitia interface was taken as the measure of the intima-media thickness complex [19]. The computerized edge-detection method has been described previously [5]. Statistical Analysis We compared the clinical measures and lipid, lipoprotein, and apolipoprotein levels at baseline for patients with and those without carotid arterial ultrasonographic end point data. For patients with carotid arterial ultrasonographic end point data, we compared carotid arterial intima-media thickness at baseline, and clinical measures and lipid, lipoprotein, and apolipoprotein levels at baseline and on-trial in the treatment group with those in the placebo group. On-trial levels of these variables were computed as averages of all measurements made during the trial, weighted by the time interval between measurements. Significance testing was done using the Student t-test for independent samples, with the two-sided significance level set at 0.05. For each patient, we fit a least-squares regression line relating carotid arterial intima-media thickness to the time since baseline ultrasonography to estimate the annual rate of progression of carotid arterial intima-media thickness. Thus, the unit of analysis was a patient-specific annualized progression rate of intima-media thickness; mean progression rates were then compared between treatment groups. Because carotid arterial intima-media thickness differed significantly at baseline between the treatment groups, we