Effects of Lovastatin and Warfarin on Early Carotid Atherosclerosis
Background—Few clinical trials have documented the efficacy of preventive treatment in asymptomatic women.
Methods and Results—Lovastatin and minidose warfarin were evaluated in a factorially designed, placebo-controlled, randomized trial. The primary outcome was 3-year change in the mean maximum intimal-medial thickness of the carotid arteries as measured by B-mode ultrasonography. Participants (n=919) were randomized to 1 of 4 treatment groups: lovastatin alone, warfarin alone, lovastatin+warfarin combination, or a double-placebo group. Eligible participants were asymptomatic for cardiovascular disease, with evidence of early carotid atherosclerosis and moderately elevated LDL cholesterol level. Almost half (n=445) of the participants were women. To avoid confounding, 117 women taking estrogen were excluded from analysis. Both sexes experienced reductions in disease progression with lovastatin; there was no evidence of an overall sex×treatment interaction (P=0.72). When estimates of the sex-specific results were examined post hoc, women experienced disease regression to the greatest extent with the lovastatin+warfarin combination (P=0.02), although the women on lovastatin alone also had a reduction in progression (P=0.09). Men experienced the greatest reduction with lovastatin alone (P=0.02), although there is a suggestion that warfarin may also reduce progression to some extent.
Conclusions—Lovastatin is beneficial in reducing disease progression in women and men. Warfarin has no effect in women, although it may reduce progression in men. In men, warfarin does not add to the benefit of lovastatin and has no advantage over lovastatin alone.
The Asymptomatic Carotid Artery Progression Study (ACAPS) previously reported that lovastatin reversed progression of carotid atherosclerosis and reduced the risk of major cardiovascular events and mortality.1 This report extends those findings by presenting the disease progression results by sex for lovastatin and warfarin individually and for the lovastatin/warfarin combination. In contrast to most previously published cardiovascular clinical trials, almost half of the 919 randomized participants in ACAPS were women.
ACAPS was a randomized, double-masked, placebo-controlled, multicenter clinical trial conducted in asymptomatic individuals to determine the effects of lovastatin and warfarin on the progression of intimal-medial thickening in the carotid arteries and on the rate of cardiovascular events.1 2 The study population was 919 women and men 40 to 79 years old with evidence of early carotid atherosclerosis (as determined by B-mode ultrasonography) and LDL cholesterol levels between the 60th and 90th percentile for middle-aged Americans.3 Almost half of the randomized participants were women (445 of 919, or 48%). Individuals with a history of myocardial infarction, stroke, or angina were excluded. A factorial design was used, and participants were randomly assigned to 1 of 4 drug combination groups: active lovastatin/warfarin placebo (LP, or lovastatin alone), lovastatin placebo/active warfarin (PW, or warfarin alone), active lovastatin/active warfarin (LW, or the combination of lovastatin+warfarin), and lovastatin placebo/warfarin placebo (PP, or double placebo). Participants were recruited from 4 clinical centers.
The primary objective was to assess the effect of treatment on the mean 3-year progression of a summary ultrasound measure of carotid atherosclerosis, the mean of the maximum intimal-medial thickness (IMT) measurements from the 12 walls, near and far, of the common carotid, the bifurcation, and the internal carotid arteries on both sides of the neck. Participants had up to 9 ultrasound B-mode examinations during the trial: 2 at baseline, 1 every 6 months for the next 30 months, and 2 at the final clinic visit.1 2 4
After randomization, the initial dose of lovastatin was 20 mg/d. For participants randomized to lovastatin, the dose was either increased at the 4.5-month follow-up visit to 40 mg/d (if the LDL cholesterol was >110 mg/dL) or decreased to 10 mg/d (if the LDL cholesterol was <90 mg/dL). The warfarin dose was maintained at 1 mg/d throughout the trial. All participants were originally to be followed up for 3 years. Because of delays in the initiation and completion of recruitment, the last cohort of 300 enrollees was followed up for only 33 months.
ACAPS has previously reported that the use of noncontraceptive estrogen (compared with nonuse) was associated with IMT regression (−9 μm/y) in postmenopausal women randomly assigned to the double-placebo group.5 The magnitude of this change was roughly equivalent to the overall regression associated with lovastatin use.1 To avoid confounding, the present analyses exclude the 117 women on estrogen replacement therapy (ERT) at baseline.
Because of an observed treatment-group interaction,1 we model treatment as a 4-level factor (PP versus LP versus PW versus LW) rather than working with margins of the original factorial design. In addition, examination of the sex-specific data suggested a nonlinear relationship for women on lovastatin alone between the mean maximum IMT and time. Therefore, in the analyses conducted for this study, we included time as a 5-level categorical factor (6, 12, 18, 24, and 30 to 36 months) and used mixed models6 7 to estimate both the between-subject (treatment) and within-subject (time) effects on mean maximum IMT. Annualized progression rates were calculated on the basis of fitted contrasts as the difference between treatment group means at baseline and at the end of follow-up. Clinical center, reader, and baseline age and LDL cholesterol level were routinely used as covariates in the progression models.
Sex-Specific Baseline Characteristics
The overall baseline comparability of the treatment groups has been demonstrated previously.1 Because this article focuses on the 802 participants not on ERT, the age-adjusted treatment group comparability of the 4 treatment groups was reexamined for the population excluding the 117 women on ERT. Although women on ERT represent 13% of all randomized participants (and 26% of all randomized women), the treatment groups were still comparable after these women were excluded. The only characteristic that was differentially distributed among the 4 treatment groups (P<0.05) was pack-years of cigarette smoking (averaging 15.9 for the PP group, 23.3 for the LP group, 19.3 for the PW group, and 16.1 for the LW group).
Eighty-five percent of the women had completed menopause, and the women had a mean of 3.1 pregnancies (median, 3; range, 0 to 13). No woman was on an oral contraceptive at baseline.
Follow-Up and Changes in IMT
The mean follow-up time among the 802 participants analyzed was 34.1 months and was nearly identical for women and men. Twelve percent of the participants were not taking study medication at the last visit; this proportion was essentially the same across treatments and between sexes. During follow-up, 82% of women and men took at least 80% of their prescribed study medications, based on pill count. Of the participants assigned to active lovastatin who were still on their medication at the 30-month clinic visit, 7% were on 10 mg/d, 43% were on 20 mg/d, and 50% were on 40 mg/d. Women were more likely to be on 40 mg/d: 58% of the 130 women taking active lovastatin were on 40 mg/d, compared with 45% of the 216 men (P=0.04). This sex difference was maintained when the data were further stratified by whether the baseline LDL cholesterol was <160 or ≥160 mg/dL.
The Table⇓ presents 3-year fitted IMT progression rates as estimated by the mixed-model approach. The Figure⇓ plots the adjusted cross-sectional IMT values over time, with the semiannual values averaged (6 and 12, 18 and 24, and 30 and 36 months) to reduce random noise. The top panel of the Table⇓ presents the overall treatment group results of the 919 randomized participants, and it is noted that the trends are consistent with those previously reported from the linear model approach.1 The second panel of the Table⇓ and the top panel of the Figure⇓ present the results after women on ERT were excluded. Again, the trends are consistent with the overall treatment effects among all randomized participants: lovastatin, either alone (P=0.003) or in combination with warfarin (P=0.01), was associated with disease regression. Overall, the differences among the 4 treatment groups appeared to be comparable between women and men (P=0.72), although the statistical power for this sex×treatment interaction is not optimal.
The data were next stratified by sex (bottom 2 panels of both the Table⇑ and Figure⇑). Whereas women experienced net average reductions in both the lovastatin alone (LP) and combination (LW) groups, only women assigned to the combination group had a statistically significant reduction in progression compared with the double-placebo (PP) group (from +5.7 μm/y for PP to −10.4 μm/y for LW, P=0.02). Women on lovastatin alone had a moderate reduction in disease progression (−6.6 μm/yr LP), although it did not reach statistical significance (P=0.09). Warfarin alone had no effect on progression (P=0.67). In addition, the effect of lovastatin alone on IMT was noted to be curvilinear in the Figure⇑: women in the LP group had an initial increase in IMT of 50 μm at 6 months after randomization (not shown in graph), followed by a steady subsequent decrease, with the 12-month value still 35 μm greater than the baseline value. However, at no point during follow-up did the difference between the LP and the PP groups ever reach statistical significance at the 0.05 level.
In contrast, whereas men experienced reductions in IMT in all 3 active treatment groups, only men on lovastatin alone (LP) had a nominal statistically significant reduction in progression compared with the double-placebo (PP) group (from +0.5 μm/y for PP to −15.1 μm/y for LP, P=0.02). Also, men on warfarin alone had a moderate reduction in disease progression (−9.8 μm/y PW), although it did not reach statistical significance (P=0.11) and did not add to the effect of lovastatin (−8.7 μm/y LW, P=0.14). Note that the difference in the rate of progression between the sexes in the PP groups, although large (+5.7 μm/y for women versus +0.5 μm/y for men), is not statistically significant (P=0.72).
Additional models adjusting for baseline covariates produced treatment group probability values that were nearly identical to those from the minimally adjusted analyses. Also, among the women, the observed treatment effects were not affected by history of menopause or the number of pregnancies.
A beneficial effect of lipid lowering on disease progression in women has also been noted in 2 angiographic trials of patients with established disease, the Canadian Coronary Atherosclerosis Intervention Trial (CCAIT)8 and the University of California San Francisco Specialized Center of Research in Arteriosclerosis (SCOR) Intervention Trial.9 The CCAIT investigators reported that among 54 women with 2-year paired coronary angiograms, disease progression (defined as a decrease in the minimal diameter in at least 1 coronary vessel by at least 0.4 mm) occurred in 28% of the women on lovastatin, compared with 59% of the women on placebo (P=0.03).8 Although there was no significant sex×treatment interaction, the comparable rates for the 245 CCAIT men were 45% and 56% (P=0.13). The SCOR investigators, studying patients with familial hypercholesterolemia, reported that among 41 women with 26-month paired angiograms, the change in the area of stenosis was −2.06% (denoting regression) for women on intensive lipid-lowering therapy (combination of colestipol, niacin, and lovastatin), compared with +1.07% for control women (P=0.05). Again, although there was no significant sex×treatment interaction, the comparable figures for the 31 SCOR men were −0.88% and +0.41% (P=0.42).9
Whereas CCAIT and the SCOR Trial were conducted in patients with established disease, ACAPS extends the sex-specific findings from those trials to asymptomatic women by demonstrating that the progression of atherosclerosis can be slowed or reversed in such women, specifically through the lipid-lowering therapy with lovastatin. This effect is comparable to the effect observed in men. Warfarin appears to provide no benefit to women in slowing atherosclerosis and no clear advantage to men over lipid-lowering therapy.
This study was supported by grants (R01-HL-38194) from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Md; Merck, Sharp and Dohme Research Laboratories, West Point, Pa; and DuPont Pharmaceuticals, Wilmington, Del. Drugs were supplied by Merck, Sharp and Dohme Research Laboratories (lovastatin), DuPont Pharmaceuticals (warfarin), and Sterling Drug Co, New York, NY (aspirin).
- Received December 16, 1998.
- Revision received March 24, 1999.
- Accepted April 17, 1999.
- Copyright © 1999 by American Heart Association
Furberg CD, Adams HP, Applegate WB, Byington RP, Espeland MA, Hartwell T, Hunninghake DB, Lefkowitz DS, Probstfield J, Riley WA, Young B, for the Asymptomatic Carotid Artery Progression Study (ACAPS) Research Group. Effect of lovastatin on early carotid atherosclerosis and cardiovascular events. Circulation. 1994;90:1679–1687.
Heiss G, Tamir I, Davis CE, Tyroler HA, Rifkind BM, Schonfeld G, Jacobs D, Frantz ID. Lipoprotein-cholesterol distributions in selected North American populations: the Lipid Research Clinics Program Prevalence Study. Circulation. 1980;61:302–315.
Riley WA, Barnes RW, Applegate WB, Dempsey R, Hartwell T, Davis VG, Bond MG, Furberg CD. Reproducibility of noninvasive ultrasonic measurement of carotid atherosclerosis. Stroke. 1992;23:1062–1068.
Espeland MA, Applegate W, Furberg CD, Lefkowitz D, Rice L, Hunninghake D, for the ACAPS Investigators. Estrogen replacement therapy and progression of intimal-medial thickness in the carotid arteries of postmenopausal women. Am J Epidemiol. 1995;142:1011–1019.
SAS Institute Inc. SAS/STAT® Software: Changes and Enhancements Through Release 6.12. Cary, NC: SAS Institute; 1997.
Waters D, Higginson L, Gladstone P, Boccussi S, Cook T, Lesperance J, for the CCAIT Study Group. Effects of cholesterol lowering on the progression of coronary atherosclerosis in women: a Canadian Coronary Atherosclerosis Intervention Trial (CCAIT) substudy. Circulation. 1995;92:2404–2410.