Aortic distensibility in patients with isolated hypercholesterolaemia, coronary artery disease, or cardiac transplant

270

Aortic

distensibility in patients with isolated hypercholesterolaemia, coronary artery disease, or cardiac transplant

The stiffness of the thoracic aorta can be assessed non-invasively. If aortic stiffness can be shown to be related to coronary heart disease, perhaps it can be used to identify which patients with hypercholesterolaemia are most likely to have atheromatous changes and thus to be selected for intensive cholesterol-lowering treatment. Hence the distensibility of the transverse aortic arch was measured by echocardiography of the aortic arch in four groups of patients—symptom-free patients with normal serum cholesterol; symptom-free patients with raised serum cholesterol; patients with coronary heart disease (all with raised serum cholesterol), and post-heart-transplant patients. In all groups distensibility fell with age. The regression slope was steeper (p<0·05) for patients with known coronary disease than for either of the disease-free groups, and among cardiac transplant recipients there was also a segregation of distensibility values between those with and without atheroma in their native hearts. The results indicate that aortic distensibility might be an indicator of coronary heart disease and that it might be useful in identifying which symptom-free subjects with modest hypercholesterolaemia should be treated aggressively.

Introduction

Hypercholesterolaemia is a major precursor of atheromatous arterial disease. Atheromatous changes may vary from overt atheroma to any of the less advanced lesions such as intimal thickening, increased collagen deposition, and foam-cell infiltration. Furthermore, the changes are usually not confined to one part of the vascular tree. Since the changes can be expected to alter passive mechanical properties, it should be possible to assess "atheroma status" from non-invasive assessment of the mechanical properties of accessible arteries--eg, the assessment of the thoracic aorta by use of ultrasoundi-3 or by magnetic resonance imaging.4 Indexing the aortic expansion that occurs during systole for the prevailing arterial pressure permits an assessment of aortic stiffness in terms of the distensibility observed. Non-invasive assessments of aortic distensibility seem to correlate with invasive evaluations.5 An initial study of putatively disease-free subjects with no cardiovascular risk factors has shown that aortic stiffness increased with advancing age.3 Such a finding is consistent with the age-related reduction in arterial elastin and increase in collagen content.6,7 Here we report our study of whether the aortic stiffness increase with age is amplified in the presence of coronary disease and what changes occur in patients with

without overt coronary disease other cardiovascular risk factors.

hypercholesterolaemia

or

Subjects and methods All subjects gave their informed consent for the study, which was the Institutional Ethics Review Committee. The subjects were grouped as follows: (i) 24 symptom-free subjects (11male) with no known vascular abnormality or disease who were long-term non-smokers, normotensive, and whose total serum cholesterol was < 5-5 mmoljl, recruited from our cardiovascular risk-factor assessment clinic (table i). Mean total cholesterol for this group was 4 38 (SEM 0-12)

approved by

mmol/1. (ii) 30 symptom-free subjects (15 male) with no clinically identified vascular disease who were normotensive and long-term nonsmokers but whose total cholesterol was > 5 mmol/1, who had been referred to a specialist lipid clinic after initial evaluation at our risk-factor assessment clinic (table I). Mean total cholesterol for this group was 8-84 (SEM 0-21) mmol/l. (iii) 16 subjects (all male) with known severe coronary artery disease (> 50% narrowing of more than one coronary artery on coronary angiography, or previous myocardial infarction proven by electrocardiographic change or enzyme rise) who had been referred for risk-factor assessment, and who were included irrespective of their smoking history, prevailing blood pressure, or serum cholesterol concentration. (iv) 23 subjects who had undergone cardiac transplantation for end-stage cardiac failure and who had subsequently been referred for risk-factor assessment. 12 (all men) had been transplanted for coronary disease and 11(4 women) for idiopathic cardiomyopathy. Echocardiography of the transverse aortic arch was done in all subjects with a Hewlett Packard 77020A echo-doppler system. The suprastemal view was used to obtain a 2D directed M mode cut perpendicularly through the transverse aortic arch. Brachial artery pressures (systolic and phase V diastolic) were determined by sphygmomanometry before and after the echocardiographic study and the mean of the pressures was used in subsequent calculations. Echocardiograms of non-transplant subjects were read by two cardiologists unaware of the age or status of each patient. Echocardiograms from transplant recipients were analysed by an additional cardiologist unaware of the cause of the pre-transplant heart failure. Videotape recordings were analysed with electric calipers and the diastolic (Dd) and maximum systolic (Ds) aortic diameters for several (between 3 and 8) cardiac cycles were measured. Mean values of these measurements were used in subsequent calculations. Aortic stiffness was measured in terms of the beta índex.8 This index is given as: beta = In

(Ps/Pd) - (Ds-Dd)/Dd,

Ps = systolic arterial pressure, Pd = diastolic arterial pressure, Ds = maximum aortic diameter during systole, and Dd aortic diameter at end diastole. where =

ADDRESSES Alfred and Baker Medical Unit, Alfred Hospital, Melbourne, Australia (A. M. Dart, MRCP, F. Lacombe, MD, J. K. Yeoh, MRCP, J. D Cameron MEngSc, G. L Jennings, MD, E Laufer, FRACP); and Heart and Lung Replacement Services, Alfred Hospital, Melbourne Australia (D. S Esmore, FRCS). Correspondence to: Dr A Dart, Alfred Hospital and Baker Medical Research Institute, Commercial Road, Prahan, Victoria, Australia 3181.

271

TABLE I-ANTHROPOMETRIC AND BIOCHEMICAL DATA FOR SYMPTOM-FREE SUBJECTS

Data given as mean (SEM). *p < 0 01 for the comparison between high and low cholesterol groups, with age as covanate in analysis of variance. Other differences were not statistically significant

TABLE II-SLOPE AND CONSTANT FROM LINEAR REGRESSION

Equation

analyses. Results Effects of isolated hypercholesterolaemia Both symptom-free subjects with low (normal) cholesterol and subjects with isolated hypercholesterolaemia but no clinical evidence of coronary artery disease showed an age-dependent increase in aortic stiffness with beta (0-16 x age) - 2-0 for the two groups combined (n = 54). Where age, sex (male or female), and cholesterol concentration were entered as independent variables in a multiple regression analysis, age (p < 0’01) and cholesterol concentration (p=0’04) were found to be significant determinants of beta with beta =l-3+(0-15x age)=

beta

=

(b x age) ± c

concentration). Regression with respect subjects (males and females combined) were classified as high or low (with respect to serum (0’31

to

Group data are given as mean (SEM). Interobserver variability in beta estimation was determined by comparing findings made independently by two observers analysing eleven recordings from the normal (low cholesterol) group. There were no statistically significant differences between their estimates, mean value for beta of 6.8 (SEM 1-1) being obtained by one, and 7-2 (1-4) by the other (p=0’6, paired t test). The mean percentage difference between paired observations was 6-5 (10)%. Results were analysed by use of SPSS.PC+ V statistical software.9 Regression diagnostic procedures justified the use of untransfonned beta values in the

was

x

cholesterol

age when

cholesterol >5-5

mmol/1

or

<5-5

mmol/1)

were

beta = (0-12 x age) - 0-5 and beta - (0-21 x age) - 3-6, respectively (fig 1, table II). The slopes of these regression equations were significantly different by analysis of variance (F=4’21; 1 vs50DF,p<0’05).

Subjects with coronary artery disease These subjects tended to be older than those without such disease and all had serum cholesterol > 5-5 mmol/l. In addition all were men.1 subject who had no detectable aortic expansion during systole was not included in further analysis because of an infinite value for beta. Unlike subjects with hypercholesterolaemia but no known coronary artery disease, the remainder of these subjects had higher values of beta than did the normal cholesterol group. Entry of age and presence or absence of coronary disease into a multiple regression analysis for hypercholesterolaemic men (n 30) revealed age (p 0-01) and presence of coronary disease (p = 003) to be significant determinants of beta. For hypercholesterolaemic men without coronary disease beta = (0-075 x age) + 05, whereas in the presence of coronary disease beta = (0-436 x age) - 10-8 (fig 2). The slopes of these regression equations were significantly different (F 424; DF 1 vs 26; p < 0-05). Regression slopes (b) and constants (c) for the change in beta with age for =

=

=

Fig 1-Relation between beta (dimensionless) and age for subjects putatively free from coronary artery disease is shown.

Regression lines and statistics determined for combined male and female data. Regression equations were beta = (021 x age) - 3 6 for the low cholesterol and beta = (0,12 x age) - 0-5 for the high cholesterol group. The slopes of these regression equations were significantly different by analysis of variance (p < 05).

for between 2-Relation beta and Fig age hypercholesterolaemic (total cholesterol> 5.5 mmol) men with (n = 15) and without (n = 15) coronary disease. Beta was significantly dependent on both age (p = 0.01 ) and presence of coronary artery disease (p=003).

272

symptom-free subjects and those with known coronary disease are shown in table II. In further analyses the slopes of the regression equations for subjects with coronary disease were also compared with those for all symptom-free men (n = 26, b==0-16) and all symptom-free subjects (n = 54, b 0-16). Analysis of variance showed the slope for subjects =

with coronary disease to be steeper than those for either of these two other groups (F = 4.24;1 vs 26 DF, p < 0-05 and F 7.3; 1 vs 65 DF, p < 0.01 respectively). =

Subjects with cardiac transplantation No difficulty was experienced in obtaining satisfactory M mode cuts through the transverse aortic arch which was distal to the suture line between the endogenous and transplanted ascending aorta. Histological validation of the presence

these

or

absence of coronary disease

available in all for non-ischaemic

was

patients. 3 subjects transplanted cardiomyopathy were found on histologial examination to have evidence of mild, unsuspected coronary atheroma. These 3 subjects were treated separately in further analysis. The remaining subjects were found to have either extensive coronary disease or completely normal coronary arteries. Since hypercholesterolaemia is a striking feature in many transplant recipients10 the post-transplant cholesterol value was thought not to accurately reflect prevailing cholesterol status during previous years and was not therefore entered into the analysis. Beta values from transplant recipients with and without previous coronary disease are shown in comparison with the 95% prediction intervals (for the prediction of beta from age) for non-coronary disease, non-transplant subjects (ie, combined high and low cholesterol (fig 3). 6 of the 8 patients whose explanted hearts revealed no coronary atheroma had beta values within the 95% prediction intervals for the symptom-free subjects, whereas only 5 of the 15 patients with atheroma had beta values within these limits. The difference between the actual beta value and the value estimated from the regression line for symptom-free subjects was determined for each transplant recipient. For patients without coronary atheroma (n = 8) the mean

3-Regression line beta= (0.16 x age)- 2.0 and 95% prediction intervals for the prediction of beta by age for all subjects (male and female) without coronary disease or transplantation.

Fig

The beta values of transplant recipients are shown for those transplanted for ischaemic cardiomyopathy (n=12), non ischaemic cardiomyopathy (n=8), and non ischaemic cardiomyopathy but subsequently shown to have coronary atheroma (n = 3)

difference from the regression line was 3.15 (SEM 096), whereas for subjects with coronary artheroma it was 1113 (2-56). These values were signifiantly different (p<0’01, unpairedt test). Exclusion of the 3 subjects transplanted for idiopathic cardiomyopathy but subsequently shown to have atheroma had no real effect on this difference (11 ’05 [3.11] vs 3.15 [0-96], p=0-03). All transplant recipients were receiving cyclosporin, prednisolone, and azathioprine. Antihypertensive treatment was similar for those whose explanted heart revealed the presence or absence of coronary artheroma. Of those with atheroma, 40% were receiving angiotensin converting enzyme inhibitors, 40% diuretics, and 20% vasodilators, whereas of those without atheroma the corresponding proportions were 50%, 60%, and 30%,

respectively. 11 subjects

had beta values within the 95% prediction intervals and 8 of these were receiving antihypertensive therapy. 12 subjects had beta values outside these confidence intervals, and 10 of those were receiving antihypertensive therapy. 4 of the 11 subjects with "normal" beta values and 3 of the 12 with raised values were on

multiple drug therapy. Discussion Our fmdings confirm that the stiffness of the thoracic rises sharply with advancing age. This age dependency is present irrespective of the presence of coronary disease or the prevailing serum concentration of cholesterol and can be accounted for by changes in arterial wall elastin and collagen content.6,7,11 Similar age-dependent changes have been shown for other human arteries.12,13 An unexpected finding was that the slope relating beta index to age was less steep for symptom-free subjects with hypercholesterolaemia than for their counterparts with normal serum cholesterol, although separation between groups was not complete. Confirmation is needed for the finding that aortic distensibility is increased in hypercholesterolaemic subjects, but several explanations can be offered for the apparent paradox that subjects with a raised cardiovascular risk factor show a reduced effect of ageing on the vasculature. One possibility is that subjects, particularly those aged over 50, with hypercholesterolaemia who are free of clinically evident atheromatous disease are constitutionally resistant to the damaging effects of a high serum cholesterol. Total serum cholesterol seems to be a less powerful predictor of the onset of new coronary disease in the elderly.14 The selection criteria for this study would automatically have excluded subjects with high cholesterol and arterial disease and biased the high cholesterol group in favour of those with pristine arteries. No subject in either symptom-free group was on vasoactive medication or was a smoker, and blood pressures in the two groups were similar, so differences in other major cardiovascular risk factors between the groups could not be the explanation for this paradox. It is possible that there were other differences between the two disease-free groups as a result of lifestyle changes associated with the presence of hypercholesterolaemia. Thus hypercholesterolaemic subjects may have been taking more exercise, and effects of level of fitness on aortic stiffness have been reported.4 Similarly, the two groups may have been taking different diets, although there is no evidence for this. Long-term differences in lifestyle between those with a high or low serum cholesterol are, however, unlikely since all the hypercholesterolaemic subjects were investigated soon after their initial referral to a lipid clinic, and such referrals aorta

273

were

made usually within months of the initial cholesterol

measurement.

Subjects with coronary artery disease showed a greater increase in stiffness with age than did disease-free subjects. The paucity of men with coronary artery disease in the age-range 40-55 years does not seem to have invalidated this conclusion since several of the transplant recipients were of this age and had raised beta values. The slope (b) for all men with coronary disease was 0-37. Our findings indicate that non-invasive assessment of the thoracic aorta can help distinguish between patients with and those without coronary disease. A similar conclusion has been reported in other invasive and non-invasive studies of arterial distensibility .1,4,5,13,15 Separation between groups is, however, not complete and the stiffness of the aorta does not uniquely discriminate between these subject groups. 78% of transplant recipients were receiving antihypertensive drugs at the time of their aortic study. There were, however, no significant differences in total percentages treated, or in percentages treated with particular classes of antihypertensive drug, between those with and without coronary artheroma. The mechanisms leading to hypertension after cardiac transplantation are unclear.16

They seem to be related particularly to cyclosporin therapy 16 and possibly to hypervolaemia in these patients.17 An additional possibility is that cardiac/vasculature mismatch may contribute to the raised pressures.16 Thus replacement of a failing with a normally contracting heart in a circulation with high systemic vascular resistance and low compliance could be expected to lead to hypertension when the cardiac abnormalities were corrected. Systemic vascular resistance does remain raised in many of these subjects16,18 although some studies show increased values in only a minority of transplant recipients. 17 Cardiac/vasculature mismatch could still be operating in subjects without increased resistance if vascular compliance were inappropriately low. However, our findings do not support this concept, at least to the extent that regional aortic distensibility is a measure of compliance. It is important to note that no investigations were done to confirm whether any of the supposedly disease-free subjects did, in fact, have subclinical coronary disease. Thus there were 4 hypercholesterolaemic females with high beta values and some or all may have had evidence of coronary disease if investigated. It is informative, in this regard, to note the findings in transplant recipients who were transplanted for non-ischaemic cardiomyopathy. 2 of these subjets had high beta values but were known from histological assessment to have coronary atheroma. In conclusion, our results suggest that a formal prospective evaluation of aortic stiffness measurement in the assessment of subjects considered at risk of developing coronary disease is indicated. If our results are confirmed the discrimination provided by aortic stiffness determination could be of considerable value in predicting which subjects with hypercholesterolaemia should be aggressively treated/9 since it is clear that not all acquire vascular disease. This study was funded by the National Health and Medical Research Council, and the Victorian Health Promotion Foundation.

REFERENCES 1. Bamdt R, Shah A. Ultrasonic measurements of aortic strain and stiffness: a non-invasive method to predict arteriosclerosis in vivo. Am J Cardiol 1982; 49: 997 (abstr).

2. Isnard RN, Pannier BM, Laurent S, London GM, Diebold B, Safar ME. Pulsatile diameter and elastic modulus of the aortic arch in essential hypertension: a non invasive study. J Am Coll Cardiol 1989; 13: 399-405. 3. Lacombe F, Dart A, Laufer E, Jennings G. A comparison of

Echo-Doppler indices of aortic elastic properties. Eur Heart J 1990; 11: 124 (abstr). 4. Mohiaddin RH, Underwood SR, Bogren HG, et al. Regional aortic compliance studied by magnetic resonance imaging: the effects of age, training and coronary artery disease. Br Heart J 1989; 62: 90-96. 5. Stefanadis C, Stratos C, Boudoulas H, Kouroukalis C, Toutouzos P. Distensibility of the ascending aorta: comparison of invasive and non-invasive techniques in healthy men and in men with coronary artery disease. Eur Heart J 1990; 11: 990-96. 6. Learoyd BM, Taylor MG. Alterations with age in the viscoelastic properties of human arterial wall. Circ Res 1966; 18: 278-92. 7. Roach MR, Burton AC. The effect of age on the elasticity of human iliac arteries. Can J Biochem Physiol 1959; 37: 557-70. 8. Hayashi K, Sato M, Handa HK, Moritake K. Biomechanical study of the constitutive laws of vascular walls. Exp Mechan 1974; 14: 440-44. 9. Norusis MJ. Base manual/Advanced statistics. SPSS/PC + Chicago, SPSS Inc, 1986/8. 10. Stamler JS, Vaughan DE, Rudd A, et al. Frequency of hypercholesterolemia after cardiac transplantation. Am J Cardiol 1988; 62: 1268-72. 11.

Neijjar I, Peiraggi MT, Thiers JC, Bouissou H. Age-related changes in the elastic tissue of the human thoracic aorta. Atherosclerosis 1990; 80:

199-208. 12. Kawaskai T,

Sasayama S, Yagi S, Asakawa T, Hirai T. Non invasive of the age related changes in stiffness of major branches of the human arteries. Cardiovasc Res 1987; 21: 678-87. 13. Hiari T, Sasayama S, Kawasaki T, Yagi S. Stiffness of systemic arteries in patients with myocardial infarction. Circulation 1989; 80: 78-86. 14. Demke MA, Grundy SM. Hypercholesterolaemia in elderly persons: resolving the treatment dilemma. Ann Intern Med 1990; 112: 780-92. 15. Stefanadis C, Wooley CF, Bush CA, Kolibash AJ, Boudoulas H. Aortic distensibility abnormalities in coronary artery disease. Am J Cardiol 1987; 59: 1300-04. 16. Starling RC, Cody RJ. Cardiac transplant hypertension. Am J Cardiol assessment

1990; 65: 106-11. 17. Corcos T, Tamburino C, Leger P, et al. Early and late hemodynamic evaluation after cardiac transplantation. J Am Coll Cardiol 1988; 11: 264-69. 18. Olivari MT, Antolick A, Ring WS. Arterial hypertension in heart transplant recipients treated with triple-drug immunosuppressive therapy. J Heart Transplant 1989; 8: 34-39. 19. Leaf A. Management of hypercholesterolaemia. Are preventive interventions advisable? N Engl J Med 1989; 321: 680-83.

From The Lancet Script scribbling Trivial as such a matter may seem, it is well to bear in mind that a prescription should be carefully read over before being delivered up; for somehow it does usually occur that, if a chance for misunderstanding be afforded, error is sure to arise. Hence the importance of medical men, and especially physicians, writing their prescriptions in some such good round-hand as the authorities of the Foreign Office compel their clerks to adopt. It is perfectly true that early professional education, the practice of scribbling notes of lectures and extracts from books, tends to spoil an originally fme hand. But a little training and care will soon obviate this defect, and doctor, chemist, and patient will in the end derive advantage. As a general rule, it may be laid down that the most illegible hands are those in which the letters incline most to the right. This method of writing is a comparatively recent and useless innovation. The old upright hand of the seventeenth century, such as we find in old receipt-books and diaries, might well serve for a model of clearness, and the more nearly the calligraphy approaches to this original style the more legible it becomes. every part of

(April 28, 1866)