Regression of Atherosclerosis in Female Monkeys
Abstract The objective of this study was to determine the structural and functional changes that occur in the artery wall in response to plasma lipid lowering and hormone replacement in surgically postmenopausal monkeys with established coronary artery atherosclerosis. Eighty-eight surgically postmenopausal cynomolgus monkeys were fed an atherogenic diet for 24 months and were then allocated into 4 groups: group 1 (n=20), a baseline necropsy group; group 2 (n=25), a lipid-lowering diet only; group 3 (n=22), lipid lowering plus conjugated equine estrogen treatment equivalent to 0.625 mg/d for a woman; and group 4 (n=21), lipid lowering plus conjugated equine estrogen and medroxyprogesterone acetate treatment (equivalent to 2.5 mg/d for a woman). Treatment was for 30 months. Histomorphometric analysis of perfusion-fixed coronary arteries revealed that plaque size did not change significantly in any of the groups compared with group 1 (P>.20). Plasma lipid lowering permitted coronary artery remodeling to occur (coronary artery and lumen size doubled compared with group 1) (P<.05); however, hormone therapy did not augment remodeling. Quantitative angiographic analysis of coronary artery reactivity revealed that lipid lowering improved dilator responses to acetylcholine by 22±4% (P=.01) but not to nitroglycerin (P=.23). Hormone replacement did not further affect vascular reactivity to the agonists tested (P>.4), but addition of medroxyprogesterone acetate diminished the beneficial effects of conjugated estrogens on coronary flow reserve (P=.03). In summary, the major arterial sequelae of lipid lowering in female monkeys were artery and lumen enlargement and improved reactivity of large epicardial coronary arteries. Addition of hormone replacement to the dietary modification did not further augment these improvements, except for the dilator capacity of the coronary microcirculation.
- Received March 7, 1995.
- Accepted April 21, 1995.
The new National Cholesterol Education Program guidelines emphasize dietary and pharmacological interventions for lowering LDL cholesterol (LDL-C) concentrations in patients with established coronary heart disease (CHD).1 The guidelines also endorse the use of estrogen replacement therapy as an alternative or supplement to conventional lipid-lowering interventions to reduce the risk of CHD in postmenopausal women.
One of the primary goals of these therapies is to produce regression of atherosclerotic lesions in coronary arteries. Traditionally, regression was used to describe the reduction in lesion size after lipid-lowering therapy. More recently, regression has been used as a term that relates to the structural, chemical, and functional improvements that occur in arteries after treatment.2
Previous studies of regression of diet-induced atherosclerosis in nonhuman primates have focused primarily on male animals and plaque size only.3 4 5 Recently, in studies of male cynomolgus monkeys, we considered not only plaque size but artery/lumen size and coronary artery reactivity.6 In the study reported here, we focused on surgically postmenopausal cynomolgus monkeys with preexisting atherosclerosis and considered not only plaque size but artery/lumen size and function.
Estrogen replacement therapy reduces the risk of CHD among postmenopausal women by as much as 50%.7 8 9 Mechanisms that may be involved include the beneficial effects of estrogen on plasma lipid concentrations,10 inhibition of progression of coronary artery atherosclerosis,11 and improvement of dilator responses of atherosclerotic coronary arteries.12 13 However, it remains unknown whether estrogen replacement, or the practice of adding a progestin to estrogen replacement, modifies the effects of estrogen on the structural, chemical, or functional changes in arteries that may occur during lipid lowering. Therefore, a study was undertaken to define the arterial sequelae of lipid-lowering therapy in surgically postmenopausal monkeys and then determine whether the addition of hormone replacement therapy modulated the effects of lipid lowering on lesion regression.
The subjects of this study were 120 feral adult female cynomolgus monkeys (Macaca fascicularis) imported from Indonesia (Charles River Primates). They ranged in age from 5 to 13 years, as estimated from dentition, and were not pregnant. After the second month of a 3-month quarantine, the monkeys began consuming a moderately atherogenic diet containing 43% of calories from fat and 0.44 mg cholesterol per kilocalorie (see Table 1⇓). The atherogenic diet was fed for 2 years. Animals were housed in social groups of 4 to 8 monkeys each. Bilateral ovariectomies were done on all animals at the start of the atherosclerosis induction period. All monkeys were housed and treated in accordance with all state and federal regulations and with the approval of the Institutional Animal Care and Use Committee.
At the end of the atherosclerosis induction phase, the animals were randomized into four groups (Fig 1⇓): a baseline necropsy group (n=20; group 1), a group fed the plasma lipid-lowering diet (see Table 1⇑) (n=28; group 2), a group fed the plasma lipid-lowering diet plus conjugated equine estrogens (n=28; group 3), and a group fed the plasma lipid-lowering diet plus conjugated equine estrogens and medroxyprogesterone acetate (n=27; group 4). The treatment phase was for 30 months.
Seventeen animals died during the atherosclerosis induction period, and 15 animals died during the treatment period, all from causes unrelated to the experimental manipulations (mostly from trauma and gastrointestinal disorders). This resulted in 103 animals available for randomization and 88 available for analysis: 20 animals in group 1, 25 in group 2, 22 in group 3, and 21 in group 4.
Allocation of animals to groups was determined by a stratified randomization scheme using as stratification variables (1) the ratio of total plasma cholesterol (TPC) to HDL cholesterol (HDL-C) during the atherogenic diet period, (2) most recent bone density measurement (8 months before onset of treatment), and (3) time since ovariectomy.
For 8 months of the 30-month treatment period, groups 3 and 4 received 7.2 μg of conjugated equine estrogens (Premarin, Wyeth-Ayerst) per monkey per day. For 22 of the 30 months, we increased the dose of conjugated equine estrogens to 166 μg per monkey per day to be equivalent to women receiving 0.625 mg/d. Throughout the 30-month treatment phase, group 4 received 650 μg per monkey per day of medroxyprogesterone acetate (Cycrin, Wyeth-Ayerst) to be equivalent to a woman’s dose of 2.5 mg/d. The hormones were administered twice daily in the diet.
An additional 10 adult female monkeys that were not part of this trial population were used to determine normal coronary flow reserve in atherosclerotic, hypercholesterolemic monkeys. All were ovariectomized and had been fed an atherogenic diet (similar to the atherosclerosis trial diet) for at least 2 years. This extra group of atherosclerotic monkeys was needed because the Doppler equipment used for coronary flow experiments was not purchased until after the baseline necropsy group (from the main trial) had been studied.
Plasma Lipids and Lipoproteins
During the atherosclerosis induction phase, quarterly measurements were made of the animals’ plasma lipid and lipoprotein concentrations (TPC, HDL-C, and triglycerides) and characteristics (HDL subfractions and LDL-C as determined by molecular weight). LDL-C concentrations were determined indirectly by subtracting HDL-C from TPC. Therefore, the term “LDL-C” is used throughout the manuscript, but actually represents LDL-C plus VLDL cholesterol. The details of the methods used to measure these components have been published previously.14
Apolipoproteins (apo) A-I and A-II15 and lipoprotein (a) [Lp(a)]16 were measured periodically according to previously published methods. HDL subclasses were measured once during month 18 of the trial.
All hormone measurements were done at the Comparative Endocrinology Laboratory of the Yerkes Regional Primate Center of Emory University (Atlanta, Ga) by Dr Mark Wilson. During the atherosclerosis induction phase, estradiol and progesterone concentrations were determined to confirm the completeness of ovariectomy. During the trial, plasma concentrations of 17β-estradiol and medroxyprogesterone acetate were determined from serum obtained 4 hours after the hormone-containing diets were given (ie, at peak concentration). The measurements were made at regular intervals during the treatment phase. The assay for 17β-estradiol was carried out by a previously published method17 that is a modification of a commercially available kit (Diagnostic Products Corp). The assay has a sensitivity of 5 to 7 pg/mL.
Medroxyprogesterone acetate also was measured by radioimmunoassay. Before assay, samples were extracted with diethyl ether, yielding a recovery of 91.2%. The antibody was a rabbit anti-medroxyprogesterone acetate serum (American Biochemicals), and the assay had a minimum detection limit of 14 pg/mL.
Systolic and diastolic blood pressures were measured at 6-month intervals throughout the trial while monkeys were sedated, by methods described previously.18 19 Eight to 18 minutes after sedation, three measurements of systolic and diastolic pressures were taken with a Dinamap portable adult/pediatric and neonatal vital signs monitor (model 8100) that uses the oscillometric technique. The average of the three measurements is reported.
Vascular Responses of Large Epicardial Coronary Arteries
Vascular responses of large epicardial coronary arteries were measured just before euthanasia and necropsy. Monkeys were anesthetized with ketamine hydrochloride (10 to 15 mg/kg body wt IM) and butorphanol (0.025 mg/kg body wt IM). Supplemental doses of both agents were given as required to maintain light anesthesia, and the animals were allowed to breathe spontaneously. These anesthetics were chosen because they do not interfere significantly with hemodynamics (blood pressure and heart rate) and respiration rate. The monkeys were warmed with a heating pad, and blood gases were maintained within normal limits throughout the experiment. A catheter was inserted into the right brachial artery for measurement of blood pressure and heart rate. A custom-designed 3F (tapered to 1.5F) catheter was inserted into the right femoral artery and advanced into the left main coronary artery under fluoroscopic guidance. Blood pressure was monitored from the tip of the coronary catheter to exclude damping and significant obstruction of blood flow.
With an infusion pump (Harvard Apparatus), serial 2.5-minute intracoronary infusions were made in the following sequence: (1) 5% dextrose in water (control); (2) acetylcholine (10−8, 10−7, and 10−6 mol/L) (estimated final concentration in the coronary artery, assuming left coronary blood flow of 10 mL/min20 ); (3) another control; and (4) nitroglycerin (15 μg/min). This dose of nitroglycerin was chosen because it produces maximal dilation in arteries.20 A 3F pigtail catheter was then placed in the left ventricle via the left common carotid artery, and angiograms were repeated during the control period (≈10 minutes after the catheter was placed) and during intraventricular infusion of adenosine (5 μm · kg−1 · min−1). Immediately after each infusion, cineangiographic images were obtained in the 30° right anterior oblique projection at 60 frames per second. Images were taken during a hand injection of 2 mL of nonionic contrast solution (Omnipaque, Squibb) into the left main coronary artery. This projection was used to provide optimal visualization of the circumflex coronary artery, which is the largest branch of the left coronary artery in most cynomolgus monkeys. Approximately 10 minutes elapsed between drug infusions.
Quantitative Coronary Angiography
Quantitative coronary angiography (QCA) was done in the Bowman Gray Cardiology Image Analysis Laboratory. A single frame from baseline and following each infusion was selected for analysis on the basis of clarity of the image of the proximal 2 to 3 cm of the circumflex coronary artery. Criteria for clarity included maximal opacification, no overlapping structures, and minimal motion artifact. Care was taken to select all frames from a single monkey from the same time in the cardiac cycle (end diastole). Each frame was optimally magnified by use of a cine-video projector (SME-3500, Sony Corp of America) and digitized to a 480×384×10-bit gray-scale image by use of a frame grabber (4 meg, Epix Inc) installed in a 486 personal computer. The mean diameters of the vessel segments of interest were measured by previously validated QCA methods (QCA Plus, Sanders Data Systems).21 When possible, specific anatomic landmarks were used to ensure that the same portion of the vessel was analyzed after each infusion. Each film was analyzed identically on two separate occasions by an operator who was unaware of the initial results. For the purposes of analysis, the average of the two measurements at baseline and after each infusion were used. Estimates of the precision of the QCA methods as applied to the monkey angiograms also were derived from the repeated-measures data.
To determine the accuracy and precision of the QCA methods used at the Comparative Medicine Clinical Research Center, a number of additional analyses were done. Images of a Plexiglas phantom with five precision-drilled holes ranging in diameter from 0.73 to 4.79 mm were obtained under radiographic conditions (kVp and Ma) similar to those used in monkey angiography. The mean (±SD) error was 0.070±0.161 mm, and the mean absolute error was 0.117 mm. The multiple correlation coefficient for the correlation between measured and true values was R2=.99. For the analysis of the monkey angiograms, 176 measurements were made on two separate occasions. The correlation between repeated measures was .98, and the mean difference (±SD) between measures was −0.002±0.028 mm. The mean absolute difference was 0.04 mm, and the coefficient of variation was 2.4%.
Coronary Artery Flow Velocity
Coronary flow velocity measurements were done with a flexible, steerable Doppler coronary artery flow guide wire (175 cm long, 0.018 in. in diameter). This wire has a 0.016-in. 12-MHz piezoelectric transducer mounted on its tip and is coupled to a 12-MHz pulsed Doppler ultrasound velocimeter (Flowmap, Cardiometrics, Inc), which consists of a real-time spectral analysis system with scrolling gray-scale display. The Doppler system also can compute a variety of on-line spectral variables, including instantaneous peak velocity and average peak velocity (APV). All data were recorded on a video recorder and video page printer. Doppler measurements were done after the angiography experiment. A catheter was advanced from the right carotid artery to the left ventricle for infusion of adenosine. The flow wire was inserted through the coronary artery catheter and advanced to the middle third of the left circumflex coronary artery under fluoroscopic guidance. APV was monitored for 10 minutes to ensure that the Doppler wire was not altering flow in the coronary artery. Additionally, electrocardiographic output was monitored in lead II to ensure that the myocardium had not become ischemic and there were no arrhythmias. After the baseline APV measurements were obtained, adenosine (5 μg · kg−1 · min−1) was infused into the left ventricle for a total of 2 minutes, and the measurements were repeated. Adenosine was infused into the ventricle because it is metabolized quickly. Flow reserve was calculated as the ratio of the APV obtained during infusion of adenosine divided by the APV at baseline.
This system has been validated to measure coronary flow velocity in in vivo animal experiments.22 Detailed specifications of in vivo and in vitro studies examining repetition frequency, burst length beam divergence, pulse duration, and sampling delay have been published previously.22 23 Additionally, correlation studies have been done comparing this Doppler system with electromagnetic flow measurements (r=.9).22 In five untreated monkeys from group 2, repeat measures were made, which determined that APV did not change over the course of the experiment (each experiment lasted for approximately 90 minutes). The correlation between measures was r=.98. Therefore, it is unlikely that time of anesthesia, presence of the Doppler flow wire, length of experiment, or lack of repeatability affected experimental results.
After the vascular reactivity studies, monkeys in the main trial were transferred immediately to the necropsy laboratory and anesthetized to a surgical plane with sodium pentobarbital (a method consistent with the Report of the Panel on Euthanasia of the American Veterinary Medical Association). The cardiovascular system was flushed with normal saline, followed by perfusion with 10% neutral buffered formalin (NBF) at a pressure of 100 mm Hg for 1 hour. The heart was removed, and the coronary arteries were perfused with 10% NBF at a pressure of 100 mm Hg for 1 hour, after which the heart was immersed in 10% NBF. The aorta was dissected free and immersion-fixed in 10% NBF.
Five serial tissue blocks for histological sectioning were cut perpendicular to the long axis of the left anterior descending, left circumflex, and right coronary arteries. One histological section was made from each block and stained with Verhoeff–van Gieson’s stain. These sections were projected, and the cross-sectional areas of intimal lesion (fatty streak and/or plaque) and areas within the internal elastic lamina (IEL) were measured with a digitizer. Atherosclerosis extent was expressed as the cross-sectional area of lesion in square millimeters. Artery size was expressed as the area within the IEL, and lumen area was determined by subtracting the intimal area from the area within the IEL. We examined treatment effects on plaque areas measured at necropsy by tissue block (section) of each coronary artery using repeated-measures analysis of variance (ANOVA) with tissue block as the repeated factor. There were no significant differential effects of treatment by artery block (P=.41). In the remainder of the analyses, we used the mean plaque areas of all 15 blocks of the three coronary arteries.
Aortic Cholesterol Measurements
At necropsy, the abdominal aorta was carefully cleaned of adventitial tissue, opened longitudinally along the posterior surface, and sectioned into segments for determination of chemical composition. The segments were approximately 1 cm2 and weighed approximately 60 mg. Wet weights were obtained from tissue that was blotted to remove surface liquid. The wet tissues were delipidated with 20 volumes (vol/wt) chloroform:methanol (3:1 vol/vol). Tissue cholesterol content was determined by the method of Rudel and Morris.24
Reported values are mean±SEM. Variables not meeting homogeneity of variance estimates by Levene’s test were subjected to logarithmic transformation. Comparisons of baseline levels of measures among the treatment groups were done by ANOVA. Tests for changes in plasma lipid concentrations over time were done by repeated-measures ANOVA. Comparisons among treatment groups of plasma lipid concentrations and intimal area were done by repeated-measures analysis of covariance using baseline prerandomization concentrations of HDL-C and triglycerides and LDL molecular weight as the covariates. Post hoc analysis of data was done by multiple comparison tests with Bonferroni-adjusted significance levels. Statistical significance was set at 95%. The coronary flow reserve values obtained from the extra group of 10 monkeys were not compared statistically with those from the main trial because the extra monkeys were not part of the original study design (ie, were not randomized at the same time). Coronary flow reserve values from the extra 10 monkeys were used for reference purposes only to give additional information about coronary flow reserve in atherosclerotic ovariectomized cynomolgus monkeys.
Plasma Hormones, Body Weight, and Blood Pressure
Plasma estradiol and medroxyprogesterone concentrations were at the bottom of the detectable range of the assays (5.0 and 25 pg/mL, respectively) in monkeys after ovariectomy and in monkeys not receiving hormone replacement therapy. Monkeys treated with conjugated equine estrogens (groups 3 and 4) had mean plasma concentrations of 167.1±10 and 160.9±14 pg/mL, respectively. Mean plasma concentrations of medroxyprogesterone acetate were 116±5 pg/mL in group 4 monkeys. Body weights were similar among treatment groups before treatment (range, 3.27 to 3.36 kg). During treatment, group 3 monkeys (conjugated equine estrogens only) had significantly lower mean body weights than groups 2 and 4, after adjustment for pretreatment body weights (group 2, 3.99 kg; group 3, 3.59 kg; group 4, 4.20 kg). Systolic blood pressures were generally about 100 mm Hg and diastolic blood pressures about 55 mm Hg in all groups, and they remained within this range throughout the treatment period.
Plasma Lipid and Lipoprotein Concentrations
Plasma lipid concentrations are summarized in Table 2⇓. There were no significant differences in TPC concentrations among the treatment groups at the initiation of the treatment phase, nor were there group differences during the treatment phase.
The HDL-C concentrations of the groups during the atherosclerosis progression phase were similar. During the treatment phase, there were significant differences among groups during months 14 through 25 of the trial (adjusted mean concentrations of 1.89, 1.06, and 1.45 mmol/L for groups 2 through 4, respectively) (P<.0001; Table 2⇑).
The two hormone treatment groups (groups 3 and 4) showed a major effect of treatment on HDL subclasses. The main change was a reduction in HDL2b concentrations (group 2, 36% of total HDL; groups 3 and 4, 14%; P<.0001). The reductions in HDL2b in the hormone treatment groups were offset primarily by increases in HDL3a and HDL3b subclasses (3a, 7.5% of total HDL in group 2, 13.7% in group 3, and 21% in group 4, P=.002; 3b, 6.7% of total HDL in group 2, 18.0% in group 3, and 17.3% in group 4, P<.0001).
The ratios of TPC to HDL-C were similar among the three groups during the atherosclerosis progression phase, all being about 24. During the treatment phase, there were marked reductions in the ratios of all groups as well as significant differences between groups (P<.0001; Table 2⇑).
Significant differences in plasma triglyceride concentrations were seen among the groups during months 14 through 25 of the trial, with concentrations being highest in the conjugated equine estrogen group (group 3) (P<.0001).
The treatment effects on plasma apo A-I concentrations closely paralleled the changes observed in HDL-C (Table 3⇓). The two hormone-treated groups had lower plasma apo A-I concentrations than the diet-only group (group 2), and the conjugated equine estrogen group (group 3) was lower than the combined treatment group (group 4). There were no significant treatment effects on plasma concentrations of apo A-II (Table 3⇓). In addition, Lp(a) concentrations were not different between treatment groups (Table 3⇓).
LDL molecular weights during the progression and treatment phases also are shown in Table 3⇑. LDL molecular weight was not one of our randomization variables. Due to chance, there were significant differences among the groups before the treatment phase (P=.004). After adjustment for the pretreatment values, groups 3 and 4 had significantly lower LDL molecular weights than group 2 (P=.002). However, there was no significant difference between groups 3 and 4 (P=.15).
Because LDL molecular weights were different between groups before treatment and because that variable is highly associated with coronary artery atherogenesis, we adjusted the coronary artery plaque size data for pretreatment differences as well as for differences in plasma triglyceride and HDL-C concentrations (see Table 4⇓). There was little or no indication of coronary artery atherosclerotic plaque progression among the animals treated with lipid lowering alone (group 2). Coronary artery plaque sizes of groups 3 and 4 were not different. Furthermore, these relationships persisted even when unadjusted values for plaque size were compared.
Coronary Artery Lumen Size
Fig 2⇓ is a schematic illustration of coronary artery sizes, plaque sizes, and lumen areas of the baseline necropsy group and the three treatment groups. During the treatment phase, there were marked increases in overall artery size as well as lumen size (P=.01 versus baseline). In both the left circumflex and left anterior descending coronary arteries, there was a tendency toward more remodeling among the animals of group 4 compared with groups 2 and 3.
Abdominal Aorta Chemical Composition
After 2 years of an atherogenic diet, the abdominal aortas of the baseline group had high amounts of total arterial cholesterol (≈14 μg cholesterol/mg wet wt) compared with normal arteries (1 to 2 μg cholesterol/mg wet wt). The three groups fed the lipid-lowering diet for 2 years had significantly less total aortic cholesterol content relative to the baseline group (6 μg cholesterol/mg wet wt; P<.05 versus group 1). There was no effect of hormone replacement therapy on total, free, or esterified cholesterol content in the abdominal aorta.
Reactivity of Large Epicardial Coronary Arteries
The mean coronary artery diameters before infusion of agonists (during the first intracoronary infusion of 5% dextrose in water) were 1.23±0.21 mm in the baseline necropsy group (group 1), 1.18±0.25 mm in group 2, 1.24±0.17 mm in group 3, and 1.11±0.20 mm in group 4 (F=0.8, P=.25). There was virtually no change in coronary artery diameters in response to acetylcholine at 10−8 mol/L, but there was a statistically significant reduction in coronary artery diameter in response to 10−7 mol/L acetylcholine. Coronary arteries of group 1 (baseline) monkeys constricted (compared with diameter measured at first control) in response to intracoronary infusion of 10−6 mol/L acetylcholine (Fig 3⇓). There were no differences in response to 10−6 mol/L acetylcholine among the three treatment groups (P>.4). Coronary arteries in all groups of monkeys dilated (compared with second control) in response to nitroglycerin (P>.4; Fig 4⇓). There was virtually no change in diameter of the proximal left circumflex coronary artery during infusion of adenosine (P>.4).
Coronary Flow Reserve
Coronary flow reserve was greater in monkeys treated with both lipid lowering and estrogen replacement than lipid lowering alone (P=.02; Fig 5⇓). However, addition of a progestin to the estrogen regimen diminished the effect of estrogen on coronary flow reserve (P=.03). Coronary flow reserve in the atherosclerotic monkeys that were not part of the trial was 1.5±0.6.
The major result of the study is that dietary lowering of plasma cholesterol concentrations induced both structural and functional improvements in the atherosclerotic arteries of surgically postmenopausal female monkeys. However, addition of estrogen replacement (with or without an added progestin) offered very little additional improvement in artery “regression.”
Regression in Females: Comparison With Regression in Males: Plasma Lipids
Similar to what has been shown in male monkeys,25 reducing the dietary consumption of cholesterol by females reduces plasma concentrations of LDL-C and increases plasma concentrations of HDL-C. There is a corresponding increase in plasma apo A-I and apo A-II concentrations. Additionally, the LDL molecular weight is predictably lower after plasma lowering of cholesterol. All of these changes are consistent with those expected to reduce the progression of coronary artery atherosclerosis and reduce the incidence of CHD. Therefore, both males and females appear to have similar plasma lipid changes in response to lowering dietary cholesterol intake.
The effect of estrogen treatment on the plasma lipids and lipoproteins of women has been studied extensively and is the subject of a number of recent reviews.26 27 28 In 1991, Lobo29 summarized the effect of a daily 0.625-mg dose of conjugated equine estrogens on plasma lipids and lipoproteins in women. He reported decreases of 5% to 10% in TPC, LDL-C, and apo B concentrations and increases from 15% to 28% in plasma triglyceride, VLDL, HDL-C, HDL subclasses 2 and 3, and apo A-I concentrations. These results from observational studies have recently been confirmed in the report from a randomized trial (PEPI: Postmenopausal Estrogen/Progestin Intervention).30 The PEPI results are reported as change from baseline, and conjugated equine estrogen treatment (0.625 mg/d) resulted in a significantly greater lowering of LDL-C and significantly greater increase in HDL-C concentrations compared with placebo.
How the addition of medroxyprogesterone acetate to a regimen of conjugated equine estrogens affects cardiovascular risk factors also has been reported from the PEPI trial. Women who received conjugated equine estrogens with medroxyprogesterone acetate either cyclically (10 mg/d for 12 d/mo) or continuously (2.5 mg/d) had significantly greater increases in HDL-C than women given placebo and significantly lower HDL-C increases than women receiving conjugated equine estrogens only. There was no significant difference in HDL-C changes between the two groups given medroxyprogesterone acetate, although blood sample collection times relative to medroxyprogesterone administration were not reported for the cyclic group.
In an analysis of available information in 1993, Lobo26 concluded that medroxyprogesterone acetate may cause reductions in HDL-C and even increases in LDL-C; however, the balance between the dose of estrogen and the dose of progestin is important in determining the magnitude of such changes. He reported that when a 5-mg dose of medroxyprogesterone acetate was given to women, reductions of about 6% in HDL-C, about 11% in the HDL2 subclass, and 5% in apo A-I were seen. Our observations on the plasma lipids and lipoproteins of hormone-treated monkeys are different from those in women. We have no current metabolic explanation for the HDL-C decreases we observed in groups 3 and 4, nor can we explain the different effects in monkeys and women relative to estrogen influences on apo A-I and the HDL subclasses. The fact that both estrogen treatments lowered the plasma HDL-C concentrations should not detract from the relevance of the atherosclerosis observations. In studies using the cynomolgus macaque model treated with oral contraceptive steroids, ethinyl estradiol protected against the development of atherosclerosis despite rather marked lowering of plasma HDL-C concentrations.31 Similarly, in studies of surgically postmenopausal monkeys given Silastic implants of estradiol, the progression of coronary artery atherosclerosis was reduced by half despite no significant differences in the plasma HDL-C concentrations.11 In that study and others, we showed that only about 20% of the beneficial effects of estrogen on coronary artery atherosclerosis relate to plasma lipoprotein changes; the remaining 80% we called the “residual effects” of estrogen, which presumably operate at the level of the coronary artery.32
We found no effect of hormone treatment and a lipid-lowering diet on plasma Lp(a) concentrations. In women, the literature regarding the effects of estrogen and progestin treatment on plasma Lp(a) concentrations is conflicting.33 34 35 36 The Atherosclerosis Risk in Communities cross-sectional study33 reported a significant difference between postmenopausal estrogen users and nonusers, but there was no difference between those treated with estrogen only and estrogen plus progestin. Lobo et al34 reported that plasma concentrations of Lp(a) are influenced primarily by genetic factors and that estrogen treatment had only a minor influence on its hepatic synthesis. Soma et al35 evaluated the plasma concentrations of Lp(a) of 55 postmenopausal women not treated or treated with 1.25 mg conjugated equine estrogens plus 10 mg medroxyprogesterone acetate. They reported Lp(a) reductions of about 50% in the treated group, with a return to pretreatment concentrations after cessation of therapy. Farish et al36 reported that treatment of 18 surgically postmenopausal women with unopposed estrogen had no large or consistent effect on plasma Lp(a) concentrations, leaving open the question whether the difference between their study and that of Soma et al35 related to dosage or the use of medroxyprogesterone acetate.
Among nonhuman primates, there is a major male/female difference in LDL particle size, with premenopausal monkeys having significantly smaller LDL particles than males.37 Further, among females, small LDL particles are related to plasma estradiol concentrations. Following surgical menopause, LDL molecular weights increase from 3.23 to 3.35 g/μmol (P<.005) to become indistinguishable from those of males. Lower LDL molecular weights and less LDL accumulation were found in the coronary arteries of ovarian hormone–replaced monkeys.38 The association between LDL molecular weight and coronary artery atherosclerosis extent in monkeys is strong.39 In the present study, both of the hormone-treated groups had significantly reduced LDL molecular weights, while the addition of medroxyprogesterone acetate tended to attenuate the reduction.
The implications of these LDL molecular weight changes for women are clearer now than in the past. Campos et al40 reported different effects of two doses of conjugated equine estrogens (without medroxyprogesterone acetate) in postmenopausal women, depending on their LDL subclasses: the proportion of large LDL particles decreased, whereas the proportion of smaller, intermediate-size LDL particles increased.
Coronary Artery Atherosclerosis
There were no effects of hormone replacement therapy on plaque size during lowering of plasma cholesterol. This was unexpected, since hormone replacement therapy with subcutaneous estradiol plus progesterone is known to inhibit progression of atherosclerosis.12 However, these results are in line with our finding that lipid lowering itself did not decrease plaque size. There was a trend toward larger lumens and overall artery size in monkeys that received combined treatment (group 4). We cannot tell whether this would have been statistically significant if more animals had been in the treatment groups or whether results would have been different if other estrogens or progesterones had been used. Therefore, it remains unclear whether hormone replacement therapy aids in the remodeling of coronary arteries during regression.
Coronary Artery Reactivity
There was no further effect of hormone replacement therapy on reactivity of large epicardial coronary arteries beyond that measured during lipid lowering. These findings are in contrast with those that show favorable effects of estrogen on reactivity of atherosclerotic coronary arteries.12 13 As mentioned previously, lipid lowering reduced the amount of cholesterol in the arteries. This cholesterol is a source of superoxide anions, which may inhibit nitric oxide–mediated dilation.41 Whatever effect estrogen may have on coronary artery reactivity of “regressed” arteries may be minimal compared with the effect of reducing the oxidant stress in the artery.
There is little information regarding the effects of progestins on coronary vasomotion. One report indicated that pharmacological doses of progesterone improved endothelium-independent dilation of nonatherosclerotic coronary arteries of rabbits,42 although it may be inappropriate to compare the effects of progesterone and synthetic progestins on coronary vasomotion. We reported recently that medroxyprogesterone acetate diminishes the beneficial effects of conjugated equine estrogens on endothelium-mediated dilation of coronary arteries of ovariectomized female monkeys.43 At present, it remains undetermined why a progestin would affect large-artery reactivity in atherosclerotic but not “regressed” coronary arteries. Results of the present experiment extend those of previous experiments by showing that addition of a progestin to an estrogen replacement regimen does not affect responsivity of large epicardial coronary arteries after dietary lowering of plasma cholesterol.
Modulation of Vasomotion of Large Atherosclerotic Epicardial Coronary Arteries
Atherosclerosis impairs endothelium-mediated dilation of arteries at various arterial sites.44 45 The combination of acetylcholine and nitroglycerin does not definitively indicate a mechanism by which lipid lowering affects vascular reactivity. Ohara et al41 showed that superoxide anion production is increased in atherosclerotic arteries. They speculated that the source of superoxide anions may be related to the presence of macrophages, oxidized lipid, and the parent cell types in the artery wall (eg, endothelial cells, smooth muscle cells). The superoxide anions from these sources could inactivate nitric oxide en route to the smooth muscle cells. During lipid lowering, both free and esterified cholesterol are removed from the arteries.46 47 We speculate that lipid lowering improved endothelium-induced dilation by reducing superoxide anion concentrations derived from the cells and associated cholesterol in the atherosclerotic lesions, thereby increasing the availability of nitric oxide. However, we cannot rule out other possibilities, such as upregulating endothelial cell muscarinic-receptor activity or enhancing the nitric oxide sympathetic pathway, as explanations for the effects of lipid lowering on endothelium-induced dilation.
Coronary Flow Reserve
Studies by Reiss et al48 indicated that acute estrogen treatment improves coronary flow reserve among postmenopausal women. Results of our studies are consistent with those of Reiss et al but also show that chronic estrogen treatment improves coronary flow reserve after dietary lipid lowering.
Additional monkeys were studied to determine the coronary flow reserve in atherosclerotic female monkeys. Comparisons of flow reserve values in groups 2 through 4 with those of the additional monkeys indicate that coronary flow reserve is highest in nonatherosclerotic monkeys, lowest in atherosclerotic monkeys, and slightly increased (compared with atherosclerotic monkeys) in monkeys undergoing dietary lowering of plasma cholesterol. Statistical analysis between additional groups of monkeys and trial monkeys is not appropriate, because these monkeys were not all part of the original trial population. The coronary flow reserve values in monkeys are lower than those reported previously in dogs22 and human beings. It is unclear whether this represents a species difference or some impairment of flow by the Doppler wire in the small coronary arteries of monkeys. The latter possibility seems unlikely, however, since coronary flow velocity, blood pressure, and heart rate were unaffected by the presence of the flow wire.
Since the diameter of large epicardial coronary arteries did not change dramatically in response to infusion of adenosine, the increase in coronary flow reserve in estrogen-treated animals is most likely due to dilation of downstream, resistance-size coronary arteries. Therefore, our results suggest that estrogen augmented dilation of smaller, resistance-size coronary arteries even after dietary lowering of plasma cholesterol. It is unclear why estrogen would improve vascular responses only in smaller coronary arteries after lipid lowering. One possibility is that estrogen modulates reactivity of large epicardial and resistance-size coronary arteries through different mechanisms, although these mechanisms were not examined in this experiment. Although atherosclerosis does not develop in smaller, intramyocardial resistance-size arteries, they are exposed to the same dyslipoproteinemia as the larger arteries and may, in fact, be more sensitive to changes in plasma lipids. Regardless of the mechanism(s), this finding may help explain, in part, how estrogen treatment reduces the clinical symptoms of microvascular angina in women.49
Interestingly, addition of a progestin to the estrogen regimen diminished the beneficial effects of estrogen on coronary flow reserve. This is consistent with results of a recent report that medroxyprogesterone acetate diminished the beneficial effects of conjugated estrogens on endothelium-mediated dilation of large epicardial coronary arteries.50 Progestins usually are added to an estrogen replacement regimen to reduce the risk of endometrial cancer associated with unopposed estrogen replacement. Therefore, this finding may have important implications about the risk/benefit ratio of progestin therapy. It is unclear why two closely related steroid hormones (an estrogen and a progestin) would have opposite effects on coronary vasomotion. Effects of progestins on coronary artery reactivity need to be explored to better understand how their risk/benefit ratio relates to CHD in postmenopausal women.
Lowering plasma cholesterol concentrations of atherosclerotic monkeys has favorable effects on both the structure and function of coronary arteries. These include remodeling of coronary arteries, removal of cholesterol from the abdominal aorta, and improved dilator responses of large epicardial coronary arteries to acetylcholine.
Estrogen replacement reduces the risk of coronary events in women with existing coronary artery disease. Addition of hormone replacement to the lipid-lowering regimen in the present study of monkeys had very little additional benefit on artery structure and function. These data raise interesting questions about the mechanism(s) by which estrogen reduces the risk of CHD in women with existing atherosclerosis. If estrogen does not cause significant reductions in plaque size, it may promote coronary artery remodeling and stabilize atherosclerotic plaques, making them less susceptible to rupture and thrombosis.
This work was supported in part by grants HL-38964 and PO1-HL-45666, National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Md; and by a Grant-in-Aid (92007730, to Dr Williams) from the American Heart Association. The authors thank Karen Potvin Klein for editing the manuscript and Jamie Fox, Debbie Golden, Vickie Hardy, and Maryanne Post for technical assistance.
National Cholesterol Education Program Expert Panel. Second Report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel II). Circulation. 1994;89:1329-1445.
Brown BG, Zhao X-Q, Sacco DE, Albers JJ. Lipid lowering and plaque regression: new insights into prevention of plaque disruption and clinical events in coronary disease. Circulation. 1993;87:1781-1791.
Armstrong ML, Megan MJ. Arterial fibrous proteins in cynomolgus monkeys after atherogenic and regression diets. Circ Res. 1975;36:256-261.
Clarkson TB, Bond MG, Bullock BC, Marzetta CA. A study of atherosclerosis regression in Macaca mulatta, IV: changes in coronary arteries from animals with atherosclerosis induced for 19 months and then regressed for 24 or 48 months at plasma cholesterol concentrations of 300 or 200 mg/dl. Exp Mol Pathol. 1981;34:345-368.
Williams JK, Vita JA, Selwyn AP, Manuck SB, Kaplan JR. Psychosocial factors impair vascular responses of coronary arteries. Circulation. 1991;84:2146-2153.
Knopp RH. The effects of postmenopausal estrogen therapy on the incidence of arteriosclerotic vascular disease. Obstet Gynecol. 1988;72(suppl):23S-30S.
Sullivan JM, Vander Zwaag R, Lemp GF, Hughes JP, Maddock V, Kroetz FW, Ramanathan KB, Mirvis DM. Postmenopausal estrogen use and coronary atherosclerosis. Ann Intern Med. 1988;108:358-363.
Adams MR, Kaplan JR, Manuck SB, Koritnik DR, Parks JS, Wolfe MS, Clarkson TB. Inhibition of coronary artery atherosclerosis by 17-beta estradiol in ovariectomized monkeys: lack of effect of added progesterone. Arteriosclerosis. 1990;10:1051-1057.
Williams JK, Adams MR, Klopfenstein HS. Estrogen modulates responses of atherosclerotic coronary arteries. Circulation. 1990;81:1680-1687.
Parks JS, Pelkey SJ, Babiak J, Clarkson TB. Contraceptive steroid effects on lipids and lipoproteins in cynomolgus monkeys. Arteriosclerosis. 1989;9:261-268.
Koritnik DR, Rudel LL. Measurement of apolipoprotein A-I concentration in nonhuman primate serum by enzyme-linked immunosorbent assay (ELISA). J Lipid Res. 1989;24:200-207.
Wagner JD, Adams MR, Schwenke DC, Clarkson TB. Oral contraceptive treatment decreases arterial low density lipoprotein degradation in female cynomolgus monkeys. Circ Res. 1993;72:1300-1307.
Castro MI, Rose J, Green W, Lehner N, Peterson D, Taub D. Ketamine-HCl as a suitable anesthetic for endocrine, metabolic, and cardiovascular studies in Macaca fascicularis. Proc Soc Exp Biol Med. 1981;168:389-394.
Corbett WT, Schey HM, Lehner NDM, Greene AW. Standardized method for recording blood pressure in anesthetized Macaca fascicularis. Lab Anim. 1981;15:37-40.
Marcus ML, Armstrong ML, Heistad DD, Eastham CL, Mark AL. Comparison of three methods of evaluating coronary obstructive lesions: postmortem arteriography, pathologic examination and measurement of regional myocardial perfusion during maximal vasodilation. Am J Cardiol. 1982;49:1699-1706.
Doucette JW, Corl PD, Payne HM, Flynn AE, Goto M, Nassi M, Segal J. Validation of a Doppler guide wire for intravascular measurement of coronary flow velocity. Circulation. 1992;85:1899-1911.
Rudel LL, Morris MD. Determination of cholesterol using o-phthaldehyde. J Lipid Res. 1973;14:364-366.
Lobo RA. Hormones, hormone replacement therapy, and heart disease. In: Douglas PS, ed. Cardiovascular Health and Disease in Women. Philadelphia, Pa: WB Saunders Co; 1993:153-173.
LaRosa JC. Lipoproteins and lipid disorders. In: Douglas PS, ed. Cardiovascular Health and Disease in Women. Philadelphia, Pa: WB Saunders Co; 1993:175-189.
Barrett-Connor E, Bush TL. Estrogen replacement and coronary heart disease. In: Douglas PS, ed. Heart Disease in Women. Philadelphia, Pa: FA Davis; 1989:159-172.
Clarkson TB, Williams JK, Adams MR, Wagner JD, Klein KP. Experimental effects of estrogens and progestins on the coronary artery wall. In: Wenger NK, Speroff L, Packard B, eds. Cardiovascular Health and Disease in Women. Proceedings of an NHLBI Conference. Greenwich, Conn: LeJacq Communications; 1993:169-174.
Farish E, Rolton HA, Barnes JF, Walsh DJ, Spowart KJM, Hart DM. Lipoprotein(a) and postmenopausal oestrogen. Acta Endocrinol. 1993;129:225-228.
Rudel LL, Pitts LL, Nelson CA. Characterization of plasma low density lipoproteins of nonhuman primates fed dietary cholesterol. J Lipid Res. 1977;18:211-222.
Wagner JD, Clarkson TB, St Clair RW, Schwenke DC, Shively CA, Adams MR. Estrogen and progesterone replacement therapy reduces low density lipoprotein accumulation in the coronary arteries of surgically postmenopausal cynomolgus monkeys. J Clin Invest. 1991;88:1995-2002.
Ohara Y, Peterson TE, Harrison DG. Hypercholesterolemia increases endothelial superoxide anion production. J Clin Invest. 1993;91:2546-2551.
Williams JK, Shively CA, Clarkson TB. Determinants of coronary artery reactivity in premenopausal female cynomolgus monkeys with diet-induced atherosclerosis. Circulation. 1994;90:983-987.
Heistad DD, Armstrong ML, Marcus ML, Piegors DJ, Mark AL. Augmented responses to vasoconstrictor stimuli in hypercholesterolemic and atherosclerotic monkeys. Circ Res. 1984;54:711-718.
Armstrong ML, Warner E, Connor W. Regression of coronary atheromatosis in rhesus monkeys. Circ Res. 1970;27:59-67.
Clarkson TB, Bond MG, Bullock BC, Marzetta CA. A study of atherosclerosis regression in Macaca mulatta. Exp Mol Pathol. 1981;34:345-368.
Reiss SE, Gloth ST, Blumenthal RS, Zacur H, Resar JR, Gerstenblith G, Brinker JA. Ethinyl estradiol acutely attenuates abnormal coronary vasomotor responses to acetylcholine in postmenopausal women. Circulation. 1994;89:52-60.
Williams JK, Honoré EK, Washburn SA, Clarkson TB. Effects of hormonal replacement therapy on reactivity of atherosclerotic coronary arteries in cynomolgus monkeys. J Am Coll Cardiol. 1994;27:1757-1761.