|Home | About | Journals | Submit | Contact Us | Français|
To determine the effect of statins and hormone replacement therapy on submaximal exercise induced coronary artery blood flow in postmenopausal women without a history of coronary artery disease.
Hormone replacement or statin therapy in early postmenopausal women without coronary artery disease have been shown to enhance arterial endothelial function; we hypothesized that these agents would improve submaximal exercise induced coronary artery blood flow.
Sixty-four postmenopausal women, aged 50–65 years without documented coronary artery disease, were randomized in a double blinded, cross-over fashion to receive 8 weeks of hormone replacement therapy versus placebo, with or without 80 mg/day of atorvastatin. Prior to receipt of any therapy and after each treatment period, each woman underwent measures of coronary artery blood flow at rest and stress.
The combination of hormone replacement therapy and atorvastatin increased submaximal exercise induced coronary artery blood flow (p=0.04). In the subgroups of women compliant with treatment, resting coronary artery blood flow increased in those receiving hormone replacement therapy (p=0.03) or statin therapy (p=0.02).
In postmenopausal women aged 50–65 years without documented coronary artery disease, rest and submaximal exercise induced coronary artery blood flow improve after receipt of high dose atorvastatin and conjugated estrogen therapy.
For older postmenopausal women, particularly those with existing coronary artery disease (CAD), hormone replacement therapy (HRT) does not reduce the risk of future cardiac events,1–3 nor favorably enhance arterial endothelial function.4 However, in middle aged women without known CAD that begin HRT early after menopause, results from several studies have shown that HRT may reduce the incidence of future cardiac events, and improve endothelial function.5–8 Perhaps HRT may exert benefits on arterial function depending on a woman’s age or the presence of atherosclerosis.
Three-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors (statins) have been shown to enhance arterial endothelial function in postmenopausal women.9 Statins exert other beneficial effects on the cardiovascular system through reduction of inflammation and serum low-density lipoprotein (LDL-C).9 Statin therapy may have an additional benefit in women receiving HRT: in a post hoc analysis of the Heart and Estrogen/Progestin Replacement Study (HERS) trial, women receiving statins exhibited an attenuated risk of cardiovascular events, including myocardial infarction (MI), when compared to women receiving HRT without statin therapy.9
Since both HRT and statins have been associated with lower rates of MI, and these therapies favorably influence arterial endothelial function in women without known CAD, we hypothesized that both HRT and statins may also favorably influence coronary artery blood flow (CABF) and the response of CABF to low-moderate levels of stress exercise. Accordingly, this study was performed to assess the effects of HRT and/or statin therapy on resting and submaximal exercise induced CABF in postmenopausal women without documented CAD.
The study was approved by the Institutional Review Board at the Wake Forest University School of Medicine, and all participants provided informed consent. This randomized, double-blinded, placebo controlled trial (awarded through NIH M01RR07122) began recruitment in March 1, 2000. The trial was designed to enroll women in a 2×3 factorial design testing the independent and joint effects of 80 mg per day of atorvastatin (ATORV) and two HRT treatments (Figure 1). The women were randomized to receive ATORV or placebo pills with equal probability. The effect of the HRT treatments were assessed using a cross-over design in which participants were assigned to receive three 8 week treatment periods of (1) placebo, (2) conjugated equine estrogen 0.625 mg (CEE), or (3) CEE 0.625 mg+medroxyprogesterone acetate 2.5 mg (CEE+MPA). The order of these 3 treatments was established using a randomized block (on time) design. Each patient was randomly assigned to one of the 6 possible orders for the receipt of the 3 hormone/placebo treatments with equal probability. Each treatment period was preceded by a 6-week period of no therapy (Figure 1). At the end of each treatment period, 10 mg of MPA was given on 10 successive days to induce menses and thus abort any induction of endometrial hyperplasia.
Women eligible for this study included those aged 50 to 65 years with a serum follicular stimulating hormone level >40 MIU/ml and no natural menses for at least 1 year, or a documented oophorectomy and a serum estrone level <25 pg/ml. Prior to study participation, women that previously used HRT could enroll in the study if they abstained from HRT 8 weeks before study enrollment. Eligible participants had no historical, physical exam, or electrocardiogram (ECG) evidence of CAD.
Participants were recruited by obtaining mailing lists to addresses of all women between the age of 50 to 65 years within a 15 zip code (approximately 20 mile) radius of downtown Winston-Salem, a 290,000 population city in western North Carolina.
Women were excluded from participation if they had a history of current or prior breast or endometrial carcinoma, arterial or venous thrombus formation, symptomatic gallstone disease, fasting triglycerides >400 mg/dl, active liver disease with a serum glutamic oxalacetic transaminase (SGOT) level >40 U/L, class III or IV New York Heart Association heart failure, Mobitz II or III atrioventricular block, severe systemic hypertension, ventricular tachycardia, unstable angina, CAD (including history of angina, coronary artery revascularization, MI, or Q wave on 12-lead ECG), obstructive hypertrophic cardiomyopathy, atrial fibrillation with rapid ventricular response, moderate or severe aortic stenosis, severe obstructive or reactive airway disease, or a contraindication to magnetic resonance imaging (MRI) scanning (such as claustrophobia, pacemakers, defibrillators or another implanted electronic devices).
Before beginning the study, and after each 8-week treatment period, serum lipids and rest/stress MRI measurements of CABF were assessed. The SGOT and total creatine kinase (CK) levels were measured 14 weeks after entering the study. To confirm compliance with each treatment period, pill counts and serum testing of estrone and estradione (E2) were performed. Participants were considered compliant if (1) pill counts demonstrated the patient took 85% of the prescribed pills (placebo or treatment), (2) their estrone level after receiving HRT was ≥25 pg/ml, and (3) their estrone level after a period of treatment without HRT was ≤25 pg/ml.
The CABF was measured noninvasively with phase-contrast MRI according to previously published techniques.10,11 Magnetic resonance imaging studies were performed on a 1.5-Tesla, GE CV/I whole-body imaging system (General Electric Medical Systems, Waukesha, WI) with participants positioned supine. A phased array cardiac surface coil positioned centrally over the chest was used as a radiofrequency receiver. Electrocardiographic monitoring leads and a brachial blood pressure cuff were applied for monitoring heart rate and systemic blood pressure throughout the exam.
The left anterior descending (LAD) coronary artery was imaged according to previously published techniques.12 Once the LAD was visualized, we acquired our flow measurement 2.5 cm distal to the origin of the LAD at its bifurcation from the left main coronary artery. Using previously published, reproducibility techniques,10 we utilized methods to locate the same position in the LAD throughout all of the repeat measurements acquired in the study. Magnetic resonance scan parameters included an 8-mm-thick slice, a 256×256 matrix, a 20 cm field of view, a through-plane velocity encoding of 150 cm/s, a 45-degree flip angle, a 13.8 ms repetition time, and a 6.7 ms echo time. Segmented k-space and view sharing were used to obtain 7 frames per RR interval. Coronary artery blood flow was obtained for each slice position by multiplying the cross-sectional area of the vessel by the average of the flow velocity of blood in that vessel.11,13
After 3 flow measurements at rest were obtained (these values were averaged to obtain baseline flow), the participant was withdrawn from the bore to perform submaximal exercise on a nonferromagnetic electronically braked bike (Lode, The Netherlands).10 To maintain the position of the subject’s torso for repeat artery imaging, the bike was mounted on the end of the MRI table, and the participant pedaled in a supine position. Also, according to previously published techniques, the subjects were marked at the apex of their suprasternal notch to ensure repositioning after exercise.10 After exercising 10 watts for 1 minute, 25 watts for 3 minutes, and 35 to 55 watts for 3 minutes, the subject was advanced back into the scanner, and images and flow measurements were once again taken in the original slice positions used to obtain the flow measurements at rest. A single stress flow measurement was accomplished within 40 seconds of exercise cessation. The reproducibility of this technique has been established previously in 11 women undergoing repeated measures separated by 8 weeks. Repeated stress-rest changes in CABF in the LAD were highly correlated (r=0.86), and using this technique 24 subjects are needed per group to detect a 20% difference in stress-rest difference in flow at 80% power.
Baseline population characteristics for categorical variables are presented as percentages, and continuous variables are presented as mean±standard deviation. Comparisons of baseline measures between those randomized to ATORV for categorical variables were tested using Fisher’s Exact test and continuous variables were tested using two-sided unpaired Student t-tests. Preliminary analyses considered the skewness and equal variances between treatment groups of the outcome variables. All means, standard deviations, and/or errors are presented with original units. Because of positively skewed distributions, tests for statistical significance and p-values for the outcome variables triglyceride, estrone, E2, and resting and stress CABF were analyzed using a logarithmic transformation.
All analyses of outcome measures were performed using SAS PROC MIXED procedures for performing repeated measures two-way analysis of covariance with the baseline value of the outcome measure as the covariate and terms in the model for period effect, lipid lowering (statin) drug effect, hormone effect, and possible lipid drug and hormone interaction. Hormone replacement therapy was the repeat factor with an unstructured covariance matrix specified. The SGOT and total CK levels were not measured at baseline, so they were not included as a covariate for the outcome measures. The data measured at the end of each HRT treatment are expressed as least square means±standard error. Except for comparisons between specific combination therapies for the total cholesterol to high-density lipoprotein (HDL-C) ratio, all other p-values referred to comparisons with placebo.
The primary analysis for this randomized clinical trial was based on the intention to treat randomized assignment. Seven subjects did not take their assigned medication (ATORV or HRT). For these subjects, a compliance with therapy analysis was performed in which the follow-up period outcome measures in which medical noncompliance occurred were set to missing.
The sample size estimate for this study was determined from a pilot cross-over study on the effects of CEE on vascular reactivity that observed a 26% relative increase in activity due to CEE. This proposed trial addressing the effect of CEE and CEE+MPA was originally designed to have 80% power to detect half the estimated relative effect (13%) between these 2 HRT therapy groups. Based on estimates of variation and correlations from the pilot study, we estimated that a sample of 124 evaluable subjects was needed to achieve the stated goal. In two previous trials we conducted with a similar follow-up period, we had a 13% drop-out rate. Allowing for a 15% drop-out rate, we planned to randomize 146 subjects to ensure participation of at least 124 evaluable subjects.
Midway through enrollment, the trial was stopped due to results from the Women’s Health Initiative (WHI) study indicating an increase in cardiac events, strokes, and pulmonary thromboembolism upon receipt of estrogen in combination with progestin.3 After 20 months of review by the National Institute of Health, the Institutional Review Board of the home institution, and the data safety monitoring board affiliated with the trial, it was determined to restart the study with a 2-period cross-over (placebo versus daily oral CEE 0.625mg) and the randomization of ATORV or placebo. The CEE+MPA treatment period was removed (gray boxes in Figure 1). As a result, the enrollment period of the study was reduced from 36 months to 16 months, and 64 participants rather than 146, were enrolled into the study. This sample size would still provide 95% power to detect to the overall 26% relative effect estimated in the pilot study between the HRT groups versus the placebo and provide 80% power to detect an effect that is 80% of the originally estimated effect of CEE alone.
Sixty-four postmenopausal women aged 51 to 65 years were enrolled and randomized to receive treatment. Of these, 8 participants did not continue with the study due to scheduling conflicts (n=3), or due to the development of “hot flashes” after discontinuing HRT prior to study entry (n=5). As a result, 56 individuals underwent serial assessments of serum lipids, and of these 47 received MRI; 9 individuals did not wish to participate in the MRI imaging. Demographic data regarding the participants are provided in Table 1.
Before suspension of study enrollment due to the WHI study results, 14 participants received the combination of CEE 0.625 mg and MPA 2.5 mg that underwent MRI measures of CABF. No serious adverse events occurred in the study population. Nine participants experienced adverse events. Seven occurred in those receiving ATORV (4 cases with CK elevation in the range of 174 U/L to 231 U/L, and 3 with abdominal or musculoskeletal pain). Two adverse events occurred in participants receiving HRT without ATORV (1 with vaginal bleeding and 1 with an elevation of CK to 237 U/L). In those with adverse events, therapy was discontinued for 2 to 6 weeks until the CK and SGOT levels returned to baseline or vaginal bleeding stopped. The individual with vaginal bleeding was evaluated by a gynecologist.
Hemodynamic data pertaining to exercise are shown in Table 2. Although some of the resting and submaximal exercise induced measures of CABF trended higher than values obtained with placebo, only the submaximal exercise induced measurement of CABF in participants receiving ATORV and CEE was higher than placebo (83±5 ml/min verses 66±6 ml/min, p<0.04 respectively) in the intention to treat analysis (Table 3). Serum lipids for the participants are shown in Table 4. In addition, the combination of HRT and ATORV lowered serum LDL-C (p<0.001), raised serum HDL-C (p<0.001), and modified the total cholesterol to HDL-C ratio (p<0.001).
A secondary analysis was performed based on participants that were compliant with therapy. The rest and stress induced change in CABF in the LAD were respectively 43 ml/min and 61 ml/min in those receiving placebo; 53 ml/min and 75 ml/min in those receiving CEE (p=0.03); 55 ml/min and 78 ml/min in those receiving ATORV (p=0.02); and 50 ml/min and 83 ml/min in those receiving ATORV+CEE (p=0.13). In those compliant with therapy, submaximal exercise induced change in CABF increased in those receiving ATORV (p=0.047) or HRT +ATORV (p=0.016) compared to placebo. When stratified by age (50–59 years and 60–65 years), the effects of HRT on resting a stress blood flow were neither statistically nor clinically different between the two strata.
There are 3 findings associated with this study. First, in postmenopausal women aged 50 to 65 years without a history of CAD, the combination of high dose oral ATORV and CEE improves submaximal exercise induced CABF. Second, in postmenopausal women aged 50 to 65 years without documented CAD who comply with the administration of 80 mg of ATORV or 0.625 mg of CEE per day, resting CABF improves by 24% to 28% (p=0.02 to 0.03). Also, after submaximal exercise, the stress-rest difference in CABF increases by 28% after receipt of ATORV (p=0.047), 22% after receipt of CEE, or 35% after receipt of ATORV+CEE (p=0.016). The combination of CEE and ATORV reduced serum LDL-C, raised serum HDL-C, and reduced serum triglycerides (p=0.001 for all).
Statin therapy reduces cardiovascular events,14–17 atherogenesis and inflammation, and enhances arterial endothelial function.14–19 These effects are mediated by enhancing the production of nitrix oxide and/or reducing the formation of oxygen derived free radicals.20–23 In the current study, we studied the effects of statin therapy on submaximal exercise induced CABF because it is this level of exercise that most individuals perform activities of daily living (such as grocery shopping or walking to the bus stop).10 With our sample size, we were able to capitalize on the precision of MRI for detecting changes in CABF after submaximal exercise. Using this noninvasive, relatively novel technique we were able to measure LAD CABF at the same location in the artery repeatedly over 14 week intervals throughout the study without exposing women to ionizing radiation, contrast material, or vasoactive sedation procedures that could perturb resting or submaximal exercise induced CABF. Importantly, our data indicate that 80 mg per day of ATORV increases CABF in the LAD at rest and after submaximal exercise (p<0.05) relative to placebo in postmenopausal women aged 50 to 65 years without a history of CAD.
Although HRT enhances peripheral arterial endothelial function,24 data from the WHI indicate that HRT involving progestins increases the incidence of MI, stroke, and pulmonary embolism. Importantly however, in a subgroup analysis of the WHI in women without documented CAD aged <60 years that were <10 years since the onset of menopause, fewer major cardiac events and coronary revascularization procedures were observed in women taking HRT versus placebo.25 To this end, our study addressed the effects of HRT on submaximal exercised induced CABF in postmenopausal women <65 years in age without documented CAD. In those women taking 85% of their prescribed medicine regimen of CEE 0.625 mg per day, or CEE+ATORV 80 mg per day, CABF at rest and after submaximal exercise improved relative to placebo. These data indicate that CEE, particularly when combined with relatively high doses of ATORV, improves submaximal exercise induced CABF, a level of activity for which many of these women perform activities of daily living. This may provide insight into potential mechanisms of reduced coronary events demonstrated in other studies.
This study was not designed to assess the relationship between CABF and serum lipid levels. However, the data from our study provide new insight into the combined effect of high dose ATORV and CEE on serum lipid profiles in older women. Though HRT modifies serum HDL-C levels,26,27 and lowers LDL-C, CEE does not lower LDL-P; therefore it may not be associated with the same benefits as a statin that lowers LDL-C.28,29 In postmenopausal women without CAD, we found that 80 mg per day of ATORV significantly decreased total cholesterol, LDL-C, triglyceride level, and the cholesterol to HDL-C ratio (p<0.001), but did not significantly change the serum HDL-C level (p=0.25). Daily oral CEE 0.625 mg significantly decreased serum LDL-C (p<0.01), the total cholesterol to HDL-C ratio (p<0.001), but increased the serum HDL-C (p<0.01) and triglyceride (p<0.01) levels (Figure 3). Many studies, including one large randomized controlled trial,27 have documented that estrogen therapy in postmenopausal women decreases serum total cholesterol and LDL-C levels, while increasing serum HDL-C and triglyceride levels. In those receiving high dose ATORV and CEE in this study, total serum cholesterol, LDL-C, and triglyceride levels, as well as the total cholesterol to HDL-C ratio were reduced (p<0.001). In addition, the serum HDL-C level was increased (p<0.001). This occurred with a decrease in serum triglycerides.
Our study has some limitations. First, we were unable to achieve the targeted enrollment goals due to mid-study suspension of patient recruitment as a result of the findings from the WHI indicating adverse cardiac events associated with estrogen/progestin HRT. Because a potential benefit may exist for early postmenopausal women without CAD using estrogen alone, it was felt that the study enrollment could be resumed in a safe manner; no participants experienced serious adverse events related to the administration of HRT in this study. Second, because we did not reach our target enrollment, several of the intention to treat analyses were underpowered. A larger number of subjects enrolled and studied could provide additional results and conclusions. Several of our results apply to those women that reliably took >85% of the prescribed therapies studied in this trial rather than the complete randomized sample. As a result, there may be a greater chance for residual confounding that could lead to a biased assessment of the effects of our single therapy drug (HRT or ATORV) interventions. Third, most women in this study were Caucasian; we are uncertain of results in women of other race. Fourth, we did not test multiple doses of therapy (HRT or ATORV), treat for periods longer than 8 weeks, obtain data on duration of prior receipt of HRT, or have the ability to comfirm an absence of CAD with invasive testing. For this reason, we have no data regarding these circumstances, and thus other studies would be needed to address these important questions. Finally, we studied CABF in the LAD coronary artery. We selected the LAD coronary artery because it supplies 40% of the blood flow to the left ventricular myocardium, and disease processes affecting the LAD coronary artery are associated with adverse outcomes. We are uncertain of the influence of these therapies on submaximal exercise induced blood flow in the right or left circumflex coronary arteries.
In conclusion, these data suggest that in relatively early (aged 50 to 65 years) postmenopausal women without documented CAD, atorvastatin or conjugated estrogen replacement therapy enhances submaximal exercise induced CABF in the LAD coronary artery.
Research supported in part by NIH General Clinical Research Center Grant NIH M01RR07122. Dr. Hundley receives support from NIH ROI HL76438, R33 CA121296, and P3O AG21332
Source of Funding
Research supported in part by National Institutes of Health grants NIH 3M01RR07122, P3OAG21332, R33CA121296, and RO1HL76438.