Kuopio Atherosclerosis Prevention Study (KAPS)
A Population-Based Primary Preventive Trial of the Effect of LDL Lowering on Atherosclerotic Progression in Carotid and Femoral Arteries
Background The atherosclerotic progression–reducing effect of LDL cholesterol (LDL-C) lowering has been established in subjects with severe atherosclerotic disease but not in persons with elevated LDL cholesterols without severe atherosclerosis. KAPS (Kuopio Atherosclerosis Prevention Study) is the first population-based trial in the primary prevention of carotid and femoral atherosclerosis.
Methods and Results The eligibility requirements were serum LDL-C ≥4.0 mmol/L and total cholesterol <7.5 mmol/L. Out of a geographically defined population, 447 men aged 44 to 65 years (mean, 57) were randomized to pravastatin (40 mg/d) or placebo for 3 years. Less than 10% of the subjects had prior myocardial infarction. Thirty-nine men discontinued study medication; however, efficacy data were available for 424 men. The primary outcome was the rate of carotid atherosclerotic progression, measured as the linear slope over annual ultrasound examinations in the average of the maximum carotid intima-media thickness (IMT) of the far wall of up to four arterial segments (the right and left distal common carotid artery and the right and left carotid bulb). For the carotid arteries, at the overall mean baseline IMT of 1.66 mm, the rate of progression of carotid atherosclerosis was 45% (95% CI, 16 to 69%) less in the pravastatin (0.017 mm/y) than the placebo (0.031 mm/y) group (P=.005). In the common carotid artery there was a treatment effect of 66% (95% CI, 30 to 95%; pravastatin 0.010 mm/y; placebo 0.029 mm/y; P<.002) at the overall mean baseline IMT of 1.35 mm. A treatment effect of 30% (95% CI, −1% to 54%) was found for the carotid bulb (pravastatin, 0.028; placebo, 0.040; P=.056) at the overall mean baseline IMT of 2.0 mm. The treatment effect was larger in subjects with higher baseline IMT values, in smokers and in those with low plasma vitamin E levels. There was no significant treatment effect on atherosclerotic progression in the femoral arteries.
Conclusions These data establish the antiatherogenic effect of LDL-C lowering by pravastatin in hypercholesterolemic men in a primary prevention setting and suggest a greater effect in smokers than in nonsmokers.
Previous clinical trials based on both angiographic and ultrasonographic imaging of arteries have established the antiatherogenic effect of LDL cholesterol (LDL-C) lowering treatment in persons with advanced atherosclerotic disease.1 Not much is known about the preventive effects of LDL-C lowering in subjects without established atherosclerotic vascular disease and with moderately elevated cholesterol levels. All previous studies to date have selected subjects on the basis of high cholesterol levels and the presence of atherosclerosis at baseline.2 3 4 5 6 7 Ultrasonographic imaging provides a reliable and relatively simple way to noninvasively evaluate the progression of atherosclerosis in carotid and femoral arteries.8 9 10 11 The KAPS study was designed to evaluate the effect of LDL-C lowering by pravastatin on the progression of carotid and femoral atherosclerosis in all eligible men from an observational population-based study, the Kuopio Ischaemic Heart Disease Risk Factor Study (KIHD).12 The subjects were hypercholesterolemic men, not selected on the basis of either their cardiovascular disease history or atherosclerotic findings at entry.13
KAPS was a randomized, double masked, placebo-controlled, single center study to evaluate the effect of LDL-C lowering with a hydroxy methylglutaryl coenzyme A (HMG CoA) reductase inhibitor, pravastatin, on the arterial wall of the carotid and the femoral arteries as evaluated by B-mode ultrasound. Subjects were stratified according to smoking status and presence of atherosclerotic lesions at baseline. The study consisted of a 2 month placebo lead-in period and a 3-year double masked placebo controlled treatment period. The primary outcome was the rate of carotid atherosclerotic progression, measured as the linear slope over annual ultrasound examinations in the average of maximum carotid IMT of the far wall of up to four arterial segments (left and right common carotid artery and left and right carotid bulb). Secondary outcomes included the rate of atherosclerotic progression in the far walls of the common carotid artery, bulb and femoral artery individually, and the combined outcome of the carotid and femoral arteries. The study protocol was approved by an international policy advisory board and by the Research Ethical Committee of the University of Kuopio. The advisory board also acted as the safety committee. Two independent monitors ascertained that the study was conducted according to good clinical practice standards.
The present study was carried out in subjects drawn from an observational population study, KIHD. KIHD is an ongoing study to investigate and identify risk factors for coronary heart disease, extracoronary atherosclerosis, and related outcomes.12 The study sample consisted of a randomly selected cohort of men, aged either 42, 48, 54, or 60 years at the entry examination, living in the city of Kuopio and seven neighboring rural communities. In the baseline examinations, carried out between 1984 and 1989, a total of 2682 men (82.9% of those invited and eligible) participated. The details of the examination have been described earlier.12
All KIHD participants with serum LDL-C levels of 4.25 mmol/L or more and body mass index of 32 kg/m2 or less (n=1194) were invited to LDL-C rescreening in 1989 or 1990. If at rescreening a subject had serum LDL-C >4.25 mmol/L, serum total cholesterol <8.0 mmol/L, body mass index <32 kg/m2, and liver enzymes (alanine aminotransferase [ALAT] and aspartate aminotransferase [ASAT]) not exceeding 1.5-fold the laboratory upper normal limit, he was eligible for participation in KAPS. Of the 987 rescreening participants, 606 men were eligible and invited to participate in KAPS. Of those invited, 21 men refused to participate, 11 had severe illness, and 17 were already on LDL-C lowering medication that was considered necessary. Five hundred and fifty-seven (557) eligible men were enrolled in KAPS, started on placebo and given dietary advice to lower LDL-C. After the 2-month placebo lead-in/dietary advice period, 447 men whose serum LDL-C remained >4.0 mmol/L and total serum cholesterol was <7.5 mmol/L were randomized. Lipid measurements were repeated at randomization. Before entry, all participants signed a written informed consent. All subjects were entered into the double-masked phase between January and September 1990 and they completed the trial during January to September 1993.
Randomization, Blinding, Double-Masked Treatment, and Follow-up
The participants were randomized either to pravastatin 40 mg once daily at bedtime or to placebo. The randomization scheme was generated by a KAPS biostatistician and the double-masked treatment units were prepared at the Bristol-Myers Squibb Pharmaceutical Research Institute, Moreton, UK, which also provided the drug supplies. Placebo and pravastatin tablets looked identical. Randomization was stratified to obtain equal distribution over the treatment groups and to enable statistical tests of effect modification. Regular smokers (at least 10 cigarettes/d) and nonsmokers (for the purpose of stratified randomization defined as less than 10 cigarettes/d) and subjects with and without atherosclerotic lesions at their baseline ultrasound examination were randomized separately. To assure the masking of the investigators and other staff, the lipid values were kept in a data register, to which there was no access for investigators other than the chief lipid chemist (KN). The subjects visited the study center (Research Institute of Public Health, University of Kuopio) at 3-month intervals.
Ultrasonographic Assessment of Atherosclerosis
High-resolution B-mode ultrasonography was used to image carotid and femoral arteries. The protocol involved the scanning of the right and left common carotid artery and the area of the carotid sinus (bulb) as high up as possible and the right and left common femoral artery including the femoral bifurcation. Three fixed angles of interrogation were used: anterolateral, lateral, and posterolateral. Images were focused on the posterior (far) wall. Ultrasonographic examinations were performed with the subject in the supine position after a rest of 15 minutes. The ultrasound system used was the Biosound Phase 2 scanner equipped with a 10-MHz annular array probe. On the basis of wedge phantom studies, the precision in the measurement of distances between interfaces from video recordings was of the order of 0.03 mm.14 The calibration of distance measurements was checked every 2 weeks against an RMI 414B tissue phantom. All scannings were done with a single ultrasound system. Four ultrasound technicians carried out the scannings. They were trained for 6 to 12 months before the study. The PCVISION Plus Frame Grabber digitizer board (Imaging Technology Inc), installed in an IBM PC 80386 microcomputer, was used to digitize B-scan frames. Image-Measure morphometry software (Microscience Inc) was used to measure distances. The site of the most advanced atherosclerotic lesion in each arterial segment and the projection showing the greatest distance between the lumen-intima interface and the media-adventitia interface IMT at baseline was located on the basis of real-time IMT measurements. The scannings were recorded by a single Panasonic AG-7330E super-VHS PAL VCR. The maximal IMT was measured from digitized frozen longitudinal (sometimes confirmed in cross-sectional) images from the video recordings. The IMT of the posterior wall was measured as the distance from the leading edge of the first echogenic (bright) line to the leading edge of the second echogenic line, as explained earlier in detail.11 Near wall measurements were not done because of their greater measurement variability.15
One measurement of IMT was carried out of both the right and left artery, separately in three arterial segments: (1) distal common carotid artery below the carotid bulb (below the locally dilated part), (2) carotid bulb area (the locally dilated part), and (3) common femoral artery, each at the site of the greatest IMT at baseline. All measurements were always done consecutively in the same session for each subject after the subject had completed the study. The “paired” reading procedure of baseline and follow-up scannings was chosen to ensure that the IMT measurements were made in the same location and angle for the baseline and all follow-up studies for each subject. All IMT measurements were carried out by a single observer (A.M.). Data on the intraobserver and interobserver variability of common carotid IMT measurements are presented elsewhere.14 16 For the purpose of the present study, a subsample of 63 baseline videotapes was reread by the same observer (A.M.). The Pearson’s correlation coefficient between the original and repeat IMT measurements was .89 for common carotid arteries, .79 for carotid bulbs, and .90 for femoral arteries.
Serum LDL-C was precipitated using PVS (polyvinyl sulfate, Boehringer Mannheim) and calculated as the difference between total and supernatant cholesterol. Serum HDL cholesterol (HDL-C) concentrations were measured after precipitation with magnesium chloride dextran sulphate.17 Serum total cholesterol and triglycerides, and plasma apolipoprotein B were measured with an autoanalyzer (Kone Specific, Kone Ltd). Cholesterol concentrations were determined enzymatically (Kone Diagnostics). The between-batch coefficient of variation (CV) was below 2.2%. Serum LDL-C concentration was measured 2 weeks before baseline (the value determining the eligibility for the study), and at baseline and each annual follow-up.
For triglycerides an enzymatic colorimetric method (Boehringer Mannheim) was used and for apolipoprotein B an immunoturbidimetric method from Kone Ltd. The between-batch CV was below 2.1% for triglycerides and below 4.5% for apolipoprotein B. Apolipoprotein(a) [Apo(a)] was measured at baseline from frozen EDTA plasma and at 36 months from frozen serum. A radioimmunoassay with an antibody against Apo(a) in the lipoprotein(a) particle was used (Pharmacia). The between-batch CV was below 6.5%. Plasma fibrinogen was determined with a clotting method (KC4, Amelung GmbH).18 The between-batch CV of fibrinogen measurement was below 5.0%. Plasma α-tocopherol concentration was determined by a high-performance liquid chromatographic method.19 Medical history and the current number of cigarettes, cigars, and pipefuls of tobacco smoked daily, the duration of regular smoking in years was assessed in an interview both at baseline and at each annual follow-up examination.
Baseline characteristics of treatment groups were compared by one-way ANOVA for continuous variables and χ2 tests for categorical variables. The analysis of ultrasound data included all randomized subjects with any follow-up ultrasound measurements regardless of their compliance with treatment. The ultrasound data were analyzed by a two-way ANCOVA of sas statistical software, version 6. As the slopes of regression of IMT change on baseline IMT were different in the four groups according to treatment and smoking status, models allowing for unequal slopes were used to adequately describe atherosclerotic progression. The dependent variable was the individual subject’s rate of progression obtained by least squares regression of the mean IMT values on follow-up time. The model included terms for treatment, smoking, treatment by smoking interaction, baseline mean IMT as the covariate, and baseline mean IMT by treatment by smoking interaction. Although subjects were stratified by smoking status and presence of atherosclerotic lesions at baseline, less than 5% of all subjects were in the category of having no carotid atherosclerosis at baseline. Therefore, baseline mean IMT was included as a covariate in the above model. Ninety-five percent confidence intervals for a reduction in atherosclerotic progression rate were estimated using the Fieller’s theorem for a ratio.20 No interim analyses were planned or done for the primary efficacy variables. The percent changes in lipid levels from baseline to the mean of annual follow-up visits were analyzed by a two-way ANCOVA of the logarithms of the mean posttreatment to pretreatment ratios. Point estimates (and 95% confidence intervals) for mean percent changes within treatment groups were obtained by exponentiating the adjusted means (and 95% confidence limits) obtained from the ANCOVA. All tests were two-sided at .05 level of significance.
Subject Enrollment and Compliance
A total of 447 subjects were randomized in the double-masked phase. During the study 39 subjects discontinued study medication, 16 in the pravastatin group and 23 in the placebo group. Of these discontinuations, 20 were due to adverse events (pravastatin 8, placebo 12); six subjects died, three in each group; five subjects discontinued at their own request (pravastatin 3, placebo 2); two subjects in the placebo group were lost to follow-up; four subjects, two in each group, were discontinued because of poor compliance with the protocol, and two subjects in the placebo group were discontinued because they received prohibited lipid-lowering medication. Of the 39 subjects that discontinued study medication, 16 had at least one follow-up visit with evaluable ultrasound measurements and thus ultrasound examinations from 424 subjects were included in the statistical analyses. On the basis of the count of returned tablets, compliance with study drug was 93% in the placebo group and 92% in the pravastatin group.
The mean age of the subjects was 57.3 years (SD, 4.3; range, 44 to 65) in the pravastatin group and 57.5 years (SD, 4.4; range, 44 to 63) in the placebo group. Baseline serum LDL-C levels (last assessment before randomization) were 4.9 mmol/L in the pravastatin group and 4.9 mmol/L in the placebo group. Other baseline characteristics are presented in Table 1⇓. There were no statistically significant differences in any of the baseline characteristics measured between the subjects who received pravastatin and those receiving placebo.
Effects on Lipids and Other Laboratory Analyses
The effects of pravastatin on lipids and other laboratory parameters are presented in Table 2⇓. For the effects of pravastatin on lipid levels, the average levels for the duration of the study were used; for the effects on apolipoproteins, fibrinogen, and α-tocopherol levels, the baseline and end of study values were used. Serum total cholesterol and LDL-C levels declined in the pravastatin treated group by 21.0 (95% CI: −22.0, −19.6) and 27.4 (95% CI: −29.0, −25.6) percent from baseline, whereas there was no significant change in the placebo group (P<.001 between groups). Pravastatin also significantly decreased triglycerides by 7.6% (95% CI: −12.0, −3.1) (between groups, P<.001).
For HDL-C, although the pravastatin group did not change statistically significantly from baseline, the placebo group showed a significant reduction from baseline and there was a significant between-group difference (P<.001).
In the pravastatin group plasma apolipoprotein B levels decreased by 19.5% (95% CI: −22.0, −17.1), whereas there was no change in the placebo group (between groups, P<.001). There was no statistically significant difference in the effects on fibrinogen levels; in both groups fibrinogen levels increased statistically significantly (P<.001): 10.2% (95% CI: 7.8, 12.6) in the placebo group and 8.2% (95% CI: 5.8, 10.6) in the pravastatin group. No changes in apo(a) were observed in either treatment group.
Plasma α-tocopherol, divided by LDL-C, the major carrier for α-tocopherol, showed a statistically significantly greater increase in the pravastatin (55.9%, 95% CI: 50.2, 61.8) than in the placebo group (28.0%, 95% CI: 23.3, 32.8; P<.001, between groups).
Effects on Atherosclerotic Progression
In the placebo group the annual rate of progression was dependent on the baseline IMT: the higher the baseline IMT the greater the rate of progression, whereas there was no relationship between baseline IMT and rate of progression in the pravastatin group. Since the linear relationship between atherosclerotic progression and baseline IMT was different for each combination of treatment and smoking status (P<.007 for baseline by treatment-by-smoking interaction) estimated rates of progression for smokers and nonsmokers are presented separately in the pravastatin and placebo groups as a function of mean IMT at baseline for the primary outcome of the common carotid artery and bulb combined (Fig 1⇓). The pravastatin group showed nearly constant progression rates regardless of baseline IMT and smoking status. In contrast, the placebo group showed increasing progression rates with increasing baseline IMTs for both smokers and nonsmokers. Placebo smokers progressed faster than placebo nonsmokers; the treatment effect was therefore greater in smokers than in nonsmokers.
Fig 2⇓ shows the treatment effect and 95% confidence intervals on the estimated annual rate of change at the overall mean baseline IMTs for the combined outcome of the common carotid artery and the bulb, and the secondary outcomes: the common carotid artery, bulb and femoral artery individually, and the combined outcome of all segments. The results show that compared to placebo, progression of atherosclerosis was slower in the pravastatin group for all segments. The treatment effect was greatest in the common carotid artery.
Table 3⇓ presents estimated (based on a regression model) annual progression rates and the treatment effects for all outcomes at the overall mean IMTs at baseline and at arbitrarily selected IMT values above the baseline mean, to show the increasing treatment effect at higher baseline wall thicknesses.
For the carotid artery segments (ie, common carotid artery and bulb) the annual rate of progression in the pravastatin group (0.017 mm/y) was significantly different (P=.005) from the placebo group in which there was progression of 0.031 mm/y at the overall mean baseline IMT of 1.66 mm. This represents a 45% (95% CI: 16, 69) reduction in atherosclerotic progression. Similar results were seen for the common carotid and carotid bulb segments examined separately and for the combined outcome of all carotid and femoral arterial segments. In the common carotid artery segment there was a statistically significant treatment effect of 66% (95% CI: 30, 95; P=.002) at the overall mean baseline IMT (1.35 mm). The treatment effect in the carotid bulb was marginally nonsignificant (30%, 95% CI: −1, 54; P=.056) at the overall mean baseline IMT (2.00 mm). The femoral artery segment showed similar progression rates in both groups. However, when all carotid and femoral segments were combined, there was a statistically significant treatment effect of 32% (95% CI: 7, 53; P=.020) at the overall mean baseline IMT (1.81 mm).
For higher baseline IMT values the treatment effect was greater and was highly significant for the common carotid artery and marginally nonsignificant for the bulb. For the femoral artery a similar relationship existed, although the treatment effect did not reach statistical significance for any of the investigated baseline IMT levels.
The effects of baseline plasma α-tocopherol levels on atherosclerotic progression were also investigated for each of the treatment and smoking combinations by means of regression analyses. As noted above, in placebo smokers progression was faster with increasing baseline IMT values (P=.035). However, at any given baseline IMT value, progression of atherosclerosis was also faster with decreasing baseline α-tocopherol levels (P=.057). Consequently, the treatment effect was greater (63%) in the participants with decreased (below the median, <28 μmol/L) plasma α-tocopherol levels at baseline than in those with higher α-tocopherols (14%), when allowing for smoking and baseline IMT in two separate models (z=2.08, P<.05 for difference).
Table 4⇓ presents the clinical cardiovascular events that occurred during the trial. The number of cardiovascular events in the pravastatin group was lower than in the placebo group, particularly for myocardial infarctions, although the difference was not statistically significant. Of the four strokes in the placebo group, one was fatal. Of the noncardiovascular deaths, one pravastatin-treated subject died of metastatic lung cancer. One placebo subject discontinued study medication for a cholecystectomy and died of a postoperative pancreatitis.
Eight (4%) of the 224 pravastatin-treated subjects and 12 (5%) of the placebo-treated subjects discontinued study medication due to adverse events, mostly gastrointestinal complaints (pravastatin, 3; placebo, 7). Other reasons for discontinuation were, one of each: elevated liver enzymes, pneumonia, eczema, nerve pain, and stroke in the pravastatin group, and prostate cancer, chest pain, depression, and stroke (two cases) in the placebo group.
The most common adverse events were musculoskeletal pain (pravastatin, 22.8%; placebo, 20.2%), abdominal pain (pravastatin, 11.2%; placebo, 9.4%), and cough (pravastatin, 8.9%; placebo, 8.5%). Four (1.8%) pravastatin-treated and three (1.3%) placebo-treated subjects had elevations of ALAT over three times the upper limit of normal on at least one occasion (one of the pravastatin-treated subjects was discontinued from therapy). Four percent of the pravastatin-treated subjects and 5% of the placebo-treated subjects had elevations of creatine kinase (CK) over 4 times the pretreatment value on at least one occasion. None of these elevations were associated with myopathic syndromes. Analyses of the adverse events showed no statistically significant differences between the treatment groups.
The present trial based on all eligible men from an observational population-based study confirms previous observations in selected clinical samples concerning the effects of LDL-C lowering on progression of atherosclerosis. In the pravastatin group, for both smokers and nonsmokers, the rates of progression were nearly constant for all baseline IMT values, indicating that atherosclerotic progression had been attenuated to a lower level that was independent of the amount of atherosclerosis present at baseline or smoking status. Progression of atherosclerosis was faster in placebo smokers compared to placebo nonsmokers. The treatment effect of pravastatin was therefore greater in smokers than in nonsmokers. A greater treatment effect with LDL-C lowering in smokers has not been observed before and remains to be confirmed in further studies. There are at least two possible explanations for the greater atherosclerosis preventive effect of pravastatin treatment in smokers than in nonsmokers. First, in the same study area, in prospective population studies, a synergistic association has been observed of elevated serum cholesterol levels and smoking both for the risk of myocardial infarction21 and for the progression of carotid atherosclerosis.22 If the atherogenic effect of high LDL-C levels is greater in smokers than in nonsmokers, then also the lowering of elevated LDL-C levels should have a greater antiatherogenic effect in smokers. Second, the effect of LDL-C lowering may simply be greater in persons who already have atherosclerotic lesions with lipid deposition. In this trial, the treatment effect on the primary study outcome was very small if the baseline carotid wall thickness (IMT) was below 1.6 mm. In the majority of previous clinical trials concerning the effect of LDL-C lowering on atherosclerotic progression, the treatment effect has been greater in subjects with advanced atherosclerosis in the beginning of the study.1
A novel finding was that especially in smokers, atherosclerotic progression was faster with lower baseline plasma vitamin E levels independent of the baseline IMT value. This might indicate that particularly in smokers the antioxidative effects of vitamin E can attenuate atherosclerotic progression. In nonsmokers, or after LDL-C reduction with pravastatin, the protective effects of vitamin E might be less. Even though based on a post-hoc interaction analysis, the greater antiatherogenic effect of LDL-C lowering in men with low vitamin E status is consistent with the suggestion that the atherogenicity of elevated LDL-C levels is mediated to an important extent through lipid peroxidation.23 24
Few trials have been previously published in which ultrasonographically assessed atherosclerotic progression was the primary outcome.25 26 In a two-by-two factorial trial, Furberg and coworkers gave 10 to 40 mg/d of lovastatin, another HMG-CoA reductase inhibitor, or warfarin (1 mg/d) or both or placebo to 919 men or women, who had mild to moderate hypercholesterolemia and early carotid atherosclerosis. Daily aspirin (81 mg/d) was recommended for all subjects. The average annual change in the mean (over 12 sites) maximum IMT was 0.006 mm (SE, 0.003 mm) in the double placebo group, −0.009 mm (SE, 0.003 mm) in the active lovastatin-warfarin placebo group, and −0.003 mm (SE, 0.003) in the lovastatin-warfarin group, averaging −0.006 mm for both lovastatin-treated groups. Because of aspirin treatment and consequently a very small annual increase in carotid IMT, the absolute changes in treated groups cannot be directly compared with our study. However, the absolute net treatment effect of lovastatin was a reduction of 0.012 mm in the mean maximum carotid IMT. The respective treatment effect in our study was similar, 0.014 mm smaller annual progression for both carotid segments combined and 0.019 mm/y treatment effect in the common carotid arteries.
In the Pravastatin, Lipids, and Atherosclerosis in the Carotids (“PLAC-II”) trial, 151 coronary patients were randomized either to 20 to 40 mg/d of pravastatin or to placebo for 3 years.26 In this secondary preventive trial the atherosclerotic progression, as measured by the aggregate mean maximum carotid IMT, was reduced by 12% by the treatment (0.068 mm/y in the placebo and 0.059 mm/y in pravastatin group), the absolute treatment effect being 0.009 mm annually. As in our study, the treatment effect was greater in the common carotid arteries (35% or 0.016 mm/y, P=.03) than in the carotid bifurcations (13% or 0.014 mm/y, P=.44). In the internal carotid arteries, there was no treatment effect.
Our findings indicate that the effect of LDL-C lowering is greater in the carotid than in the femoral arteries. The observation is consistent with our earlier finding based on cross-sectional data.11 27 The smaller treatment effect in the carotid bulb than in the common carotid arteries may be explained partly by the larger random measurement variability in the bulb than in the straight part of the common carotid artery. Also, atherosclerotic progression in the carotid bulb and the femoral bifurcation, in which the blood flow is turbulent, may be determined more by hemodynamic, rheologic, and local mechanical factors than by lipids.11 27 On the basis of this trial, we recommend the use of the straight part of the common carotid artery as the site for ultrasonographic imaging in future studies in which atherosclerotic progression will be assessed by B-mode ultrasonography.
It has previously been demonstrated that the wall thickness and the severity of atherosclerosis in the common carotid arteries are strong predictors of the extent of coronary atherosclerosis and coronary events.11 28 29 30 Carotid atherosclerosis also appears to have a similar risk factor profile as coronary atherosclerosis.8 22 31 All these findings suggest that carotid wall thickness is associated with coronary atherosclerosis and can be used as its surrogate in clinical trials and population studies.
In this study a favorable effect of pravastatin was observed on cardiovascular events. Although the differences in events were not statistically significant, the reductions in event rates were similar as those reported with other pravastatin studies in subjects with evidence of atherosclerotic disease or with multiple risk factors for coronary artery disease.32 33
In conclusion, the findings of the present study confirm the preventive effect of LDL-C lowering on the progression of carotid atherosclerosis in persons free of advanced atherosclerotic disease and establish the antiatherogenic effect of pravastatin. They also suggest that the benefit of LDL-C lowering may be greater in smokers and possibly in men with low vitamin E status, hypotheses to be tested in further trials. Finally, our data are inconclusive with regard to the effects of LDL-C lowering on the progression of femoral atherosclerosis.
This study was supported by grants from the Academy of Finland and the Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ. The authors thank Margot Mellies, MD, and Mark McGovern, MD, for their contribution to the study protocol, Arja Malkki, registered radiology technician, for carrying out the ultrasonographic measurements, Arto Haapanen, MD, PhD, for consulting in the ultrasound quality control studies, Kimmo Ronkainen, MPh, for data management and analyses, Richard Kronmal, PhD, for advice in statistical analysis, Ulrika Persson and Karin Käll for monitoring the study, public health nurses Anne Airaksinen, Hannele Kastarinen, Annikki Konttinen, Leena Leskinen, and Raili Weeman for subject management, Timo Lakka, MD, PhD, for participating in the clinical work, and Erkki Voutilainen, MD, PhD, lipidologist, for clinical consulting.
- Received February 1, 1995.
- Revision received April 19, 1995.
- Accepted May 3, 1995.
- Copyright © 1995 by American Heart Association
Blankenhorn D, Hodis HN. Arterial imaging and atherosclerosis reversal. Arterioscler Thromb. 1994;14:177-192.
Levy RI, Brensike JF, Epstein SE, et al. The influence of changes in lipid values induced by cholestyramine and diet on progression of coronary artery disease: results of the NHLBI Type II Coronary Intervention Study. Circulation. 1984;69:325-337.
Salonen R, Seppänen K, Rauramaa R, Salonen JT. Prevalence of carotid atherosclerosis and serum cholesterol levels in eastern Finland. Arteriosclerosis. 1988;8:788-792.
Crouse JR, Harpold GH, Kahl FR, Toole JF, McKinney WM. Evaluation of a scoring system for extracranial carotid atherosclerosis extend with B-mode ultrasound. Stroke. 1986;17:270-275.
Salonen JT, Salonen R. Ultrasound B-mode imaging in observational studies of atherosclerotic progression. Circulation. 1993;87(suppl II):II-56-II-65.
Salonen JT, Salonen R, Nyyssönen K, Korpela H, Voutilainen E, Ylitalo P. Kuopio Atherosclerosis Prevention Study. CVD Epidemiology Newsletter. American Heart Association 1992, Winter 1991/1992, No. 47:6.
Salonen JT, Korpela H, Nyyssönen K, Salonen R. Precision and reproducibility of ultrasonographic measurement of progression of common carotid artery atherosclerosis. Lancet. 1993;341:1158-1159.
Salonen R, Salonen JT. Intima-media changes in a population study: KIHD. In: Boccalon H, ed. Vascular Medicine. New York, NY: Elsevier Science Publishers BV; 1993.
Finley PR, Schifman RB, Williams RJ, Lichti DA. Cholesterol in high-density lipoprotein: use of Mg2+/dextran sulfate in its enzymic measurement. Clin Chem. 1978:24:931-933.
Finney DJ. Statistical Method in Biological Assay. London, England: Griffin; 1978:80-82.
Salonen JT, Puska P, Kottke TE. Smoking, blood pressure and serum cholesterol as risk factors of acute myocardial infarction and death among men in Eastern Finland. Eur Heart J. 1981;2:365-373.
Furberg CD, Adams HP, Applegate WB, et al. Effect of lovastatin on early carotid atherosclerosis and cardiovascular events. Circulation. 1994;90:1679-1687.
Salonen JT, Salonen R. Ultrasonographically assessed carotid morphology and the risk of coronary heart disease. Arterioscler Thromb. 1991;11:1245-1249.
Solberg LA, Strong JP. Risk factors and atherosclerotic lesions: a review of autopsy studies. Arteriosclerosis. 1983;3:187-198.
Craven TE, Ryu JE, Espeland MA, et al. Evaluation of the associations between carotid artery atherosclerosis and coronary artery stenosis: a case control study. Circulation. 1990;82:1230-1242.