Comparison of Progression of Coronary Calcium in Postmenopausal Women on Versus Not on Estrogen/Progestin Therapy
Article Outline
The prophylactic effect of postmenopausal hormone replacement therapy on coronary atherosclerosis remains controversial. We, therefore, examined the influence of combined estrogen/progestin therapy on the progression of coronary calcium as a marker of coronary atherosclerosis. We determined the extent of coronary calcium in 277 women (age 57 ± 6 years, time after menopause 3.9 ± 2.4 years, group I) at the beginning of hormone replacement therapy using multislice computed tomography. For quantification, we calculated the volume score. After an observation period of 3 years, we determined the progression of coronary calcium in a second scan. The results were compared with those from an age- and risk factor-adjusted group of postmenopausal women without hormone substitution (group II). No significant difference was found in the volume score (59 ± 95 vs 58 ± 88) or risk factor distribution between the 2 groups on study entry. In 56 women of group I and 52 women of group II, coronary calcium could be excluded on the initial scan (p = NS). After a mean observation time of 38.5 ± 4.9 months, we observed no significant difference between the 2 groups regarding an increase in volume score (17 ± 24 vs 19 ± 27, p = NS) or the fraction of women with an increase in volume score (82.2% vs 84.2%). In conclusion, a reduced progression of coronary calcium in postmenopausal women on combined estrogen/progestin therapy could not be observed compared with a matched group of women without hormone substitution.
Multislice computed tomography allows the visualization of coronary calcium, which correlates with the extent of coronary atherosclerosis.1, 2 As shown on histopathologic studies, the amount of coronary calcium also reflects the extent of noncalcified plaques.3, 4 Therefore, the quantification of coronary calcium using the volume score method allows the noninvasive monitoring of the progression of coronary atherosclerosis with high sensitivity and reproducibility.1 In the present prospective study, we used this modality to determine the progress of coronary atherosclerosis in asymptomatic postmenopausal women on combined estrogen and progestin therapy and compared the results with those of a matched group of women without hormone replacement therapy (HRT).
Methods
The local clinical institutional review board approved the research protocol, which complied with the Declaration of Helsinki. We included 281 white women (age 56.8 ± 5.8 years) who intended to start HRT because of postmenopausal discomfort (physical dysregulation, fatigue, sleep disturbance) in this study from 2000 to 2001 (group I). On study entry, all women were asymptomatic and showed no signs of coronary artery disease on electrocardiography, stress electrocardiography, and echocardiography. For all women, cardiovascular risk factors were assessed by personal interview and screening of medical records. In addition, the arterial blood pressure and low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride, and fasting blood glucose levels were measured in our hospital. We determined the 10-year risk of coronary events as described by the Adult Treatment Panel III.5
After giving written informed consent, all women underwent multislice computed tomographic scanning for coronary calcium. The extent of coronary calcium was quantified using the volume score (see the following). To determine the influence solely of HRT on the progression of coronary calcium, we selected a matched comparison group (group II) of 281 women with an equal body mass index, age, risk factor, and volume score distribution. These women were derived from a collective of women sent to our clinic for cardiovascular examination. Also, the women of the comparison group showed no signs of coronary artery disease on electrocardiography, stress electrocardiography, or echocardiography.
All women in group I received combined HRT with 0.625 mg of conjugated equine estrogen plus 2.5 mg of medroxyprogesteron acetate. The study participants who were receiving HRT had previously undergone a 3-month washout period before study entry. The mean time after menopause was 3.9 ± 2.4 years in group I and 3.7 ± 2.4 years in group II, with a minimum time after menopause of 2.0 years. The uterus was intact in all women. No woman had a history of cancer, especially breast cancer. All women had a regular platelet count and hematocrit. The study patients were interviewed after a mean observation time of 38.5 ± 4.9 months (range 33.6 to 41.4) and were invited to undergo a second scan to determine the progression of coronary calcium. The end point of this study was the progression of coronary calcium as quantified by the volume score. In addition, cardiovascular events were assessed using a standardized interview. In the case of hospital admission or further cardiologic examinations, the patient’s medical records were reviewed for complaints of chest discomfort, dyspnea, myocardial infarction, coronary revascularization, and stroke.
The multislice computed tomographic studies were performed with a Siemens Sensation 4 scanner. For calcium scanning, a standardized protocol was applied. Four axial slices with 4 × 2.5-mm collimation and 0.5-second rotation time were acquired simultaneously (120 kV, 133 mAS), allowing coverage of the whole heart in 18 ± 3 seconds in 1 breath-hold. Images were reconstructed with the Adaptive Cardio Volume algorithm in diastole with a 3-mm slice width (increment 1.5 mm) and ≤125-ms temporal resolution, dependent on heart rate. To achieve optimum image quality, retrospective electrocardiographic gating was used. With this technique, the optimal reconstruction position for every coronary artery with least-motion artifacts can be achieved. Using electrocardiographically controlled tube output modulation, the total radiation exposure was calculated as <2 mSv for women.6
The scans were read by experienced readers (AB, AK) using the volumetric quantification algorithm with continuous isotropic interpolation as proposed by Callister et al.7 This scoring method was chosen because of its improved reproducibility compared with the Agatston score (interscan variability <10%) mainly evoked by the reduction of partial volume errors using fast multislice computed tomographic volume coverage and an overlapping image reconstruction (3-mm slide width, 1.5-mm increment) algorithm.
Statistical analyses were performed using the Statistical Package of Social Sciences software package, version 10.0 (SPPS, Chicago, Illinois). The volume score is expressed as the mean ± SD, except as indicated. To compare the score values in the different risk groups, we used the Wilcoxon signed rank test for unpaired data. To determine the significance of score changes within each group, we used the Wilcoxon signed rank test for paired data. Comparisons of the mean relative changes in volume score between groups were made with the 2-sample t test. A p value of <0.05 was considered statistically significant. Absolute changes in the volume score were set in relation to the initial scores, and the Pearson’s correlation coefficient was calculated to demonstrate the correlation between the height of the initial score and the score increase. Because we did not find a normal distribution, all scores were logarithmically transformed.
Results
The mean observation period was 38.4 ± 4.9 months (range 33.6 to 41.4). During that period, 4 women ceased HRT because of gynecologic disorders and were excluded from the study. A second scan could not be performed in these women. Thus, 277 women completed the study. Table 1 lists the baseline characteristics of the 2 groups. No significant difference was found between the 2 groups regarding cardiovascular risk factor distribution or the number of risk factors per person. The proportion of women with hyperlipidemia who received treatment with statins was equal in the 2 groups, as was the mean low-density lipoprotein cholesterol and high-density lipoprotein cholesterol levels. Also, a significant change in low-density lipoprotein and high-density lipoprotein cholesterol levels during our study were not observed in the 2 groups. No significant difference was found in the average hemoglobin A1c fraction in women with diabetes mellitus between the 2 groups (8.7 ± 3.5% vs 8.9 ± 3.7%, p = NS; Table 1).
Table 1. Baseline characteristics on study entry
| Variable | HRT | |
|---|---|---|
| Yes (n = 277) | No (n = 277) | |
| Age (yrs) | 56.6 ±8.9 | 56.1±9.3 |
| Body mass index (kg/m2) | 27.9±7.1 | 27.3±7.8 |
| Time after menopause (yrs) | 3.9±2.4 | 3.7±2.4 |
| Initial volume score | 59±65 | 58±61 |
| Women with score 0 | 56 | 52 |
| No cardiovascular risk factors | 66(24%) | 80(29%) |
| Arterial hypertension | 130(47%) | 138(50%) |
| Hyperlipidemia (total cholesterol >240 mg/dl) | 102(37%) | 97(35%) |
| Women on statin therapy | 63(23%) | 68(25%) |
| LDL level (mg/dl) on study entry | 157±102 | 149±112 |
| LDL level (mg/dl) at study end | 146±108 | 144±109 |
| HDL level (mg/dl) on study entry | 44±31 | 47±34 |
| HDL level (mg/dl) at study end | 43±32 | 47±36 |
| Diabetes mellitus | 53(19%) | 60(22%) |
| HbA1c fraction (%) | 8.7.±3.5 | 8.9±3.7 |
| Familiar risk profile | 150(54%) | 147(53%) |
| Smokers | 72(26%) | 66(24%) |
| Average number of risk factors per women | 2.4 | 2.5 |
| ATP III risk score | 13.1±4.6 | 12.6±5.8 |
No significant difference was found between the mean volume score in groups I and II (59 ± 65 vs 57 ± 61, p = NS) on study entry in all age groups. In 55 women of group I and 52 women of group II, coronary calcium could be initially excluded. In the 2 groups, the mean volume score increased significantly with age (Table 2). The volume score was significantly higher in women in groups I and II with the cardiovascular risk factors arterial hypertension (66 ± 49, range 0 to 214), hyperlipidemia (82 ± 68, range 0 to 310), and diabetes mellitus (94 ± 87, range 0 to 580) compared with the women without these risk factors (28 ± 22, range 0 to 164, p <0.05). In all risk groups, the wide range of scores was conspicuous. Although in all risk groups, women without calcium could be found, in the group without cardiovascular risk factors, 34 women (6.1%) had a calcium score >75th percentile.
Table 2. Score distribution of women with and without hormone replacement therapy (HRT), divided into different age groups, in initial and final scan
| Variable | HRT | |||||
|---|---|---|---|---|---|---|
| Yes (n = 80) | No (n = 80) | Yes (n = 148) | No (n = 148) | Yes (n = 49) | No (n = 49) | |
| Age (yrs) | 50–55 | 55–60 | 60–65 | |||
| Initial score | 30±41 | 29±37 | 60±102 | 62±90 | 86±123 | 89±123 |
| Final score | 40±50 | 42±53 | 81±104 | 84±119 | 112±100 | 118±132 |
| Absolute increase | 10±14 | 13±15 | 21±20 | 22±21 | 26±24 | 29±28 |
| Relative increase (%) | 33±45 | 44±47 | 35±32 | 35±33 | 41±37 | 32±31 |
| Women with score 0 on initial scan (n) | 20 | 20 | 26 | 23 | 9 | 9 |
| Women with score 0 on both scans (n) | 16 | 17 | 22 | 20 | 7 | 8 |
| Women with score reduction (n) | 3 | 5 | 0 | 1 | 2 | 2 |
| Average score reduction | 9±5 | 13±8 | — | 13 | 4±3 | 4±2 |
In groups I and II, a significant increase in the mean volume score was observed after the observation period of 38.4 ± 4.9 months: in group I, from 59 ± 65 to 76 ± 79 and in group II, from 57 ± 61 to 76 ± 71 (Figure 1). In groups I and II, 82.4% and 84.2% of women, respectively, had an increase in the volume score. A score reduction was observed in 5 women (1.8%) in group I and 8 women (2.8%, p = NS) in group II. No significant difference was found in the absolute or relative increase in the volume score between women with and without HRT during the observation period. This was true for all age groups (Table 2). Figure 2 depicts the absolute change of volume scores in relation to the initial score values. The increase in volume showed a strong correlation to the initial score (correlation coefficient r = 0.84 for group I and 0.83 for group II). A significant difference in score changes between women of group I and group II could not be observed.
Figure 1.
Mean volume score of women with HRT (group I, n = 277) (black bars) and without HRT (group II, n = 277) (gray bars) according to initial scan and final scan after an observation period of 33.4 ± 5.1 months. *p <0.05.
Figure 2.
Difference between initial and final volume score in dependence of initial score for women with HRT (group I, n = 277, correlation coefficient r = 0.84) (black circles) and women without HRT (group II, n = 277, correlation coefficient r = 0.83) (white squares).
The rate of women with initial exclusion of coronary calcium who developed coronary calcium during the observation period was equal in the 2 groups (10 women in group I vs 7 women in group II, p = NS).
During the observation period, 12 women in group I and 10 women in group II had myocardial infarctions, and 18 women in group I and 20 women in group II underwent coronary revascularization. Stroke occurred in 4 patients in group I and 4 patients in group II. The cardiovascular events were registered at 18.5 ± 7.6 months (range 9.5 to 27.6) in group I and 17.7 ± 6.9 months (range 8.7 to 25.1) in group II after study entry. No significant difference was found in event rates between the 2 groups (Table 3). Cardiovascular events were registered at 18.5 ± 7.6 months (range 9.5 to 27.6) in group I and 17.7 ± 6.9 months (range 8.7 to 25.1) in group II after study entry (p = NS).
Table 3. Cardiovascular events (absolute number and annualized rate) of women with and without hormone replacement therapy (HRT) during observation period of 33.4 ± 5.1 months
| Variable | HRT | |||
|---|---|---|---|---|
| Yes | No | |||
| n | Annualized rate (%) | n | Annualized rate (%) | |
| Coronary revascularization | 18 | 2.2 | 20 | 2.5 |
| Myocardial infarction | 12 | 1.5 | 10 | 1.2 |
| Cardiac death | 5 | 0.6 | 7 | 0.9 |
| Stroke | 4 | 0.5 | 4 | 0.5 |
| Noncardiac death | 2 | 0.2 | 1 | 0.1 |
We observed no cardiovascular events in patients with exclusion of coronary calcium. Patients in groups I and II with cardiovascular events had a volume score >75th age- and gender-adjusted percentile. The mean initial score in these patients was significantly higher than the score in women without cardiovascular events (201 ± 159 vs 79 ± 63, p <0.01). Apart from those women who died of myocardial infarction or noncardiac causes, we performed a second scan in all women with cardiovascular events after an observation time of 36.7 ± 4.8 months. These women had a significantly higher increase in volume score compared with the event-free collective (44 ± 17 vs 19 ± 12, p <0.05).
Discussion
To observe the influence solely of HRT on coronary atherosclerosis, we compared women treated with HRT with a collective of women matched regarding the distribution of cardiovascular risk factors. In addition, and in contrast to previous studies, the comparison group consisted of women with initially comparable volume scores. Therefore, a comparative level of coronary atherosclerosis in both groups on study entry was ensured.
As expected, the mean scores in groups I and II increased with patient age and in the presence of cardiovascular risk factors. Thus, the wide range of scores within each risk group indicates that coronary calcium reflects not only the presence of cardiovascular risk factors, but the individual extent of coronary atherosclerosis actually induced by the underlying cardiovascular risk factors. The score values and score distribution according to age and risk factors in both groups were comparable to those in a population of 35,246 adults published by Hoff et al.8 The volume scores of postmenopausal women in the present study were significantly lower than the scores of men of the same age as determined in previous studies.8, 9
In 82.4% of group I and 84.2% of group II, an increase in the volume score was observed. The mean progression in both groups was comparable to the changes observed in previous studies.10, 11, 12 A significant difference in the absolute or relative progression between groups I and II could not be found in any age group. Thus, an influence of HRT on the progression of coronary calcium was not observed. Rather, the absolute increase in coronary calcium was related to the initial volume score, indicating that women with accelerated atherosclerosis at study entry continued to develop coronary calcium more rapidly than women with an initial volume score below average. These findings are consistent with those of a study by Budoff et al.13 A significant difference in the progression of coronary calcium between women with and without HRT was also not detected.13
Also, no significant difference was found between groups I and II in the fraction of women with an initial exclusion of coronary calcium who developed coronary calcium during the study. Thus, a prophylactic effect of HRT on women without initial coronary calcium cannot be assumed. These findings are in contrast, in particular, to a histopathologic study by Christian et al.14 They reported a significantly lower content of coronary calcium in postmenopausal women on HRT compared with women without therapy.14 Thus, a slower progression of coronary calcium could have been expected. In addition to the positive effects on lipoprotein metabolism and platelet activation, some clinical studies have reported a reduction of cardiovascular events in women on HRT.15, 16, 17 However, in most studies, a generally healthier lifestyle for the women on HRT has been reported, and an incomplete adjustment of risk factor distribution in the comparison groups limited the significance of several studies.18 These very facts could lead to a reduced calcium score as reported by McLaughlin et al19 in these women compared with a matched group regarding conventional cardiovascular risk factors. To avoid this bias, we selected a comparison group with an initially equal volume score, representing the extent of coronary atherosclerotic activity, and performed a prospective study focusing on the progression of coronary calcium.
A reduced progression of coronary calcium on HRT in postmenopausal women was not observed in our study. Although large studies have shown a close correlation between coronary calcium and concomitant coronary artery disease and future cardiovascular events,10 the observation of coronary calcium alone needs to be supplemented by studies evaluating the influence of HRT on cardiovascular event rates, because combined estrogen progestin therapy might possess beneficial effects induced by vascular mechanisms apart from atherosclerosis. Because of the limited population and observation period, the present study was not able to provide information on the effect of HRT on cardiovascular events. As an additional limitation, the present study was not a double-blind study.
References
- . Electron beam computed tomographic coronary calcium score cutpoints and severity of associated angiographic lumen stenosis. J Am Coll Cardiol. 1997;29:1542–1548
- . Correlation of coronary calcification and angiographically documented stenoses in patients with suspected coronary artery disease: results of 1764 patients. J Am Coll Cardiol. 2001;37:451–457
- . Coronary artery calcium area by electron beam computed tomography and coronary atherosclerotic plaque area: a histopathologic correlative study. Circulation. 1995;92:2157–2162
- . Coronary artery calcification: assessment with electron beam computed tomography and histomorphometric correlation. Radiology. 1994;192:619–623
- . Executive summary of the third report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). JAMA. 2001;285:2486–2497
- . Determination of coronary calcium with multi-slice spiral computed tomography: a comparative study with electron-beam CT. Int J Cardiovasc Imaging. 2002;18:295–303
- . Coronary artery disease: improved reproducibility of calcium scoring with an electron-beam CT volumetric method. Radiology. 1998;208:807–814
- . Conventional coronary artery disease risk factors and coronary artery calcium detected by electron beam tomography in 30,908 healthy individuals. Ann Epidemiol. 2003;13:163–169
- . Age and gender distributions of coronary artery calcium detected by electron beam tomography in 35,246 adults. Am J Cardiol. 2001;87:1335–1339
- . Progression of coronary artery calcium and risk of first myocardial infarction in patients receiving cholesterol-lowering therapy. Arterioscler Thromb Vasc Biol. 2004;24:1–7
- . Effect of HMG-CoA reductase inhibitors on coronary artery disease as assessed by electron-beam computed tomography. N Engl J Med. 1998;339:1972–1978
- . Rates of progression of coronary calcium by electron beam tomography. Am J Cardiol. 2000;86:8–11
- . Effects of hormone replacement on progression of coronary calcium as measured by electron beam computed tomography. J Womens Health. 2005;14:389–396
- . Estrogen status correlates with the calcium content of coronary atherosclerotic plaques in women. J Clin Endocrinol Metab. 2002;87:1062–1067
- . Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. JAMA. 1998;280:605–613
- . Effects of estrogen replacement on the progression of coronary-artery atherosclerosis. N Engl J Med. 2000;343:522–529
- . The protective effects of estrogen on the cardiovascular system. N Engl J Med. 2001;345:34–40
- . Understanding the divergent data on postmenopausal hormone therapy. N Engl J Med. 2003;348:645–650
- . Relation between hormone replacement therapy in women and coronary artery disease estimated by electron beam tomography. Am Heart J. 1997;124:1115–1119
PII: S0002-9149(06)02065-0
doi:10.1016/j.amjcard.2006.08.040
© 2007 Elsevier Inc. All rights reserved.
