- Email: [email protected]
Aortic atherosclerosis and stroke
Aortic Atherosclerosis and Stroke Dennis P. Briley, MB, BS, Zaki Lababidi, MD, Anthony J. Bowdler, MD, M. Krishnamurthy, MB, BS, Silvestre P. Cansino, MD, and Robert C. Touchon, MD
Methods: Studies using transesophageal echocardiography (TEE) suggest aortic atherosclerosis may be a risk factor for stroke, particularly stroke of undetermined mechanism, but controls in prior studies were not balanced for vascular risk factors. We used TEE to evaluate aortic atherosclerosis in 60 patients with stroke compared with a high-risk control population of 46 subjects. We also examined the possible association of plasma viscosity and fibrinogen levels to aortic atherosclerosis. Results: The mean maximal plaque thickness (MMPT) was similar for the control (2.8 + 3.6 mm) and the stroke group (3.3 +-3.5 mm), but varied with stroke mechanism. The MMPT was similar in stroke of undetermined and atheroscterotic mechanism [3.5 -+ 4 mm (n = 25) and 4.2 -+ 4.3 rnm (n = 16), respectively], significantly greater than in stroke of other mechanisms (1.7 +- 1.2 ram, P < .05, n = 19). Patients with stroke of undetermined mechanism were four times more likely (95% confidence interval [CI] 1.2-12) to have plaques ->5 mm compared with controls. Ulcerated plaque was associated with plaque thickness (P < .001) and plasma viscosity (P < .001). Conclusions: Aortic atherosclerosis is associated with stroke of undetermined cause suggesting atherosclerosis is a cause of stroke of undetermined etiology. Plaque ulceration was associated with the thickness of aortic plaque and plasma viscosity. Key Words:Stroke--Aortic atherosclerosis--Transesophageal echocardiography--Plasma viscosity--Fibrinogen.
The possibility that atheromatous plaques in the aortic arch m a y lead to cerebral ischemia has been highlighted by A m a r e n c o et al, based on an autopsy study. 1 With the a d v e n t of transesophageal echocardiography (TEE), it has become increasingly feasible to image relevant portions of the aorta. 2 Several studies utilizing TEE have suggested that aortic atheromatous plaques are especially associated with stroke of u n d e t e r m i n e d cause. 3-5Aortic atherosclerosis m a y be a risk factor for recurrent strokeP Furthermore, three studies indicated that aortic plaque is an independent risk factor for stroke, although the choice of controls
From the Department of Medicine, Huntington VAMC and Marshall University School of Medicine, Huntington, WV. Supported by the Department of Veterans Affairs and Marshall University. Received August 12,1996; accepted September 11, 1996. Address reprint requests to Dennis P. Briley, MB, BS, Chief, Neurology Section, Department of Medicine, Marshall University School of Medicine, 180l Sixth Avenue, Huntington, WV 25755. Copyright © 1997 by National Stroke Association
1052-3057/97/0603-0003503.00/0
causes some uncertainty, s,7,8 In the TEE studies of Amarenco et al and Di Tulho et al, patients referred for TEE for reasons unrelated to stroke or aortic atherosclerosis were used as the control population, s,7 The s t u d y of Jones et al used a c o m m u n i t y based age-matched control population, which, although age-matched, was not risk factor matched. 8 We therefore c o m p a r e d TEE findings in patients with stroke to findings in a population without history of stroke, but considered at high risk. Both fibrinogen and plasma viscosity are risk factors for myocardial infarction, stroke, and peripheral vascular disease. 9-14 Accordingly, w e also investigated w h e t h e r plasma viscosity, fibrinogen, and other blood proteins are associated with aortic atherosclerosis detected by TEE in our control subjects.
Methods We studied two groups of patients, a control and a stroke group. The controls were patients enrolled in a TEE s t u d y to evaluate cardiac risk factors for cerebral embo-
Journal of Stroke and Cerebrovascular Diseases, Vol. 6, No. 3, 1997: pp 125-129
125
126
D.P. BRILEY ET AL
lism. This study was a p p r o v e d by the Huntington VAMC institutional review board, a n d all patients gave informed consent. The entry criterion for this group was either atrial fibrillation or left ventricular dysfunction as measured by an ejection fraction less than 45%. All patients were interviewed and examined by a neurologist to exclude a prior symptomatic cerebral ischemic episode. The second group, the "stroke" group, comprised patients referred for TEE after a cerebral ischemic event, either transient or an infarct. A vascular risk factor history was obtained for all patients, including hypertension, diabetes, smoking history, lipid history, and cardiac history. For the stroke group, probable stroke mechanism was classified using the criteria of A d a m s et aD 5 Stroke of u n d e t e r m i n e d etiology was diagnosed based on a negative diagnostic evaluation. All transesophageal echo studies were performed by a cardiologist using a 5 MHz biplane transesophageal transducer and a 128 XP/10 Acuson machine, (Acuson c o m p u t e d sonography, Acuson corporation, CA). The transesophageal probe was m a n i p u l a t e d to obtain adequate views, and all studies were recorded on video tape for further analysis. Each segment of the aorta was examined for atherosclerotic plaques by an observer blinded to clinical data. We measured plaque thickness perpendicular to the vessel wall using software from I m a g e v u e (Nova Microsonic, Nova Microsonic Inc., Carmel, IN). The maximal plaque thickness of the most severe lesion in the proximal (ascending and aortic arch, proximal to the takeoff of the left subclavian artery) and of the distal aorta (descending, distal to the takeoff of the left subclavian artery) were measured. A qualitative assessment was m a d e for ulceration; an ulcer was considered to be present if there was m a r k e d irregularity of the surface with the appearance of a crater. In the control group, measurements were m a d e of plasma, serum, whole blood viscosity, fibrinogen, and blood proteins b y serum protein electrophoresis. Viscosity was measured at 37°C on a Wells Brookfield viscometer. Fibrinogen levels were m e a s u r e d using a modified thrombin clotting time.
Table 1. Demographics of the study population
Factor
Control group
Stroke group
Number Age Diabetes Hypertension Never smoked Ex-smoker Current Smoker Atrial fibrillation CHF
46 66 ± I0 26% (12) 59% (27) 13% (6) 52% (24) 35% (16) 61% (27) 61% (27)
60 64 ± 9 35% (21) 58% (35) 7% (4) 30% (18) 63% (38) 7% (4) 7% (4)
P value
ns ns ns
P < .02 P < .00l P < .001
Aortic atherosclerosis was analyzed as both a continuous variable for the measured thickness of the plaque, and as a dichotomous variable for both ulceration, and for plaque thickness as either ~ 5 m m or <5 mm. Statistical analysis was performed on a personal computer using STATA software. 16
Results The control group contained 46 subjects, and there were 60 patients in the stroke group. Table 1 shows the demograpt~cs for the two groups. The two groups were balanced for age, diabetes, and hypertension, but not for smoking or heart disease. There was a preponderance of current smokers in the stroke group. As expected, all patients in the control g r o u p had either atrial fibrillation or congestive heart failure. Table 2 displays the relationships between aortic atherosclerosis and stroke. Although the mean maximal plaque thickness (MMPT) was similar in the stroke and control group for both the distal and proximal aorta, there was a trend for more patients in the stroke g r o u p to have plaques ->5 m m c o m p a r e d with the control group. Plaque thickness was significantly related to stroke mechanism. More patients with stroke of atherothrombotic or undetermined mechanism had plaques ---5 m m c o m p a r e d with either the control group or c o m p a r e d with stroke of other
Table 2. Plaque sever#)., and ulceration with stroke broken down by stroke mechanisn~
Control (n = 46) Stroke group (aggregate) Atherothrombotic (n = 16) Undetermined (n = 25) Cardioembolic (n = 5) L a c u n a r ( n = 12) Other (n = 2)
mmpt (ram)
Proximal mmpt (ram)
Distal mmpt (mm)
Plaque >5 mm
Ulceration
2.8 _-z-3.6 3.3 --- 3.5 4.2 _ 4.3 3.5 --- 4 1.6 --- 1.5 1.8± 1.1 1
1.6 ± 2.0 1.8 --- 3 2.4 ± 2.3 2.3 ± 3.9 0.8 ± I 0 . 8 ± 1.1 0.5 --+ 0.7
2.6 __. 3 2.9 + 3.6 3.5 - 4.4 3.4 _ 4 1.6 ± 1.5 1.6± 1.2 I
13% (6) 25% (15)t 38% (6)* 36% (9)* 0 0 0
9% (4) 5% (3) t3% (2) 1% (4) 0 0 0
*P < .05 compared with controls; t P = .13 compared with controls.
AORTIC ATHEROSCLEROSIS AND STROKE
127
Table 3. Comparison of stroke risk factors by stroke mechanism Atherothrombotic
Undetermined
Other
Control
Factor
N = 16
N=25
N = 19
N=46
Age (years) Diabetes Hypertension Non-smoker Current smoker Cholesterol (mmol/L) Fibrinogen (g/L)
63 ± 8 25% (4) 69% (11) 31% (5) 69% (11) 5.77 ± 0.96* 4.07 ± 1.27
65 ± 8 20% (5) 48% (12) 36% (9) 64% (16) 5.51 ± 1.29" 4.12 ± 1.14
65 ± 10 63% (12)I" 63% (12) 43% (8) 58% (11) 5.46 ± 0.91 3.70 - 0.90
66 - 10 26% (12) 59% (27) 65% (30) 35% (16) 4.99 --+ 1.03 3.80 --+ 1.18
*P < .02 for mean cholesterol level in atherothrombotic and undetermined group combined compared with control. t P = .01 for presence of diabetes in stroke of other etiology compared with stroke of atherothrombotic/undetermined etiology or compared with controls.
etiology. The o d d s ratio (OR) for the presence of plaque ->5 m m in patients with stroke of u n d e t e r m i n e d mechanism c o m p a r e d with controls was 4 (95% CI 1.2-12). The OR was also 4 (95% CI 1.1-15) for atherothrombotic stroke c o m p a r e d with controls. The MMPT was similar in both the atherothrombotic and u n d e t e r m i n e d stroke patients and was significantly greater c o m p a r e d with patients with stroke of other etiology (3.8 ± 4 m m versus 1.7 --- 1.2 ram, P < .05), b u t not significantly greater than the controls. Although our s t u d y was unbalanced for smoking, the OR for plaque ->5 m m in nonsmokers with atherothrombotic a n d u n d e t e r m i n e d stroke was 4 (95% CI 0.8-15), and in smokers was also 4 (95% CI 0.8-~) compared with controls. In addition, logistic regression found both plaque ->5 m m and smoking were i n d e p e n d e n t l y associated with a t h e r o t h r o m b o t i c / u n d e t e r m i n e d stroke combined c o m p a r e d with controls (OR for plaque ->5 m m 4, 95% CI 1.3-12, P = .02; OR for current smoking 4, 95% CI 1.4-9, P = .006). The risk factor profiles for patients with stroke of atherothrombotic m e c h a n i s m a n d stroke of undetermined mechanism, s h o w n in Table 3, were similar. Mean cholesterol levels were higher in patients with stroke of atherothrombotic and u n d e t e r m i n e d mechanism compared with controls; there was also a trend toward higher fibrinogen levels in these two groups. The prevalence of diabetes was greater in patients with stroke of other mechanism (predominantly cardioembolic and small vessel disease). The degree of atherosclerosis in the distal aorta was more severe than the proximal aorta (P < .0002), although the two were significantly correlated (R 2 = .53, P < .0001). Mobile lesions, p r e s u m e d to be thrombi, were seen in two subjects only, one control, and one patient with stroke of u n d e t e r m i n e d etiology, respectively. Data on carotid stenosis were available for 93% (56/60) of the stroke group. There was a trend to more severe carotid stenosis in patients with plaques ->5 m m in our stroke patients, so that 43% (6/14) h a d carotid stenosis >50% and plaques
->5 r a m c o m p a r e d to 24% (10/42) of patients with plaque < 5 ram, (P = .19). Ulcerated plaques were detected in 7 patients. There were no obvious differences in standard vascular risk factors between patients with ulcers and those without. Table 4 shows that subjects with ulcerated plaques had increased p l a s m a and serum viscosity, along with a trend to increased fibrinogen levels. Patients w i t h ulcerated plaques h a d m a r k e d l y increased plaque thickness. Multiple linear regression analysis f o u n d both plaque thickness (P < .001) and plasma viscosity (P < .001) were i n d e p e n d e n t variables associated with the presence of ulceration (R 2 = .64, P < .0001). N o correlation between plaque thickness to age, cholesterol, p l a s m a viscosity or fibrinogen was detected. There was a slight correlation of
Table 4. Risk factors for ulcerated plaque
Factor
Ulcerated
No Ulcer
N = 7
N = 99
Age (years) 68 ± 11 65 - 9 Hypertension 43% (3) 60% (59) Diabetes 29% (2) 31% (31) Never smoker 0 10% (10) Ex-smoker 43% (3) 39% (39) Current smoker 57% (4) 60% (59) MMPT (mm) 11.3 ± 5.5 2.4 - 2.5 Fibrinogen (g/L) 4.46 ± 0.61 3.73 ± 1.21 Plasma viscosity (mPas)* 2.3 ± 0.3 1.6 ± 0.2 Serum viscosity (mPas)* 1.5 +-_0.3 1.3 ± 0.1 Whole blood viscosity at 23s -I(mPas)* 6 . 6 ± 1.4 7 . 9 ± 1.5 Whole blood viscosity at 230 s-I (mPas)* 4.2 ± 0.7 4.9 ± 0.6 *Data from control subjects only.
P value ns ns ns
ns P < .0001 P = .08 P < .0001 P < .02 P=.I P = .1
128 plaque thickness to alpha 2 globulins measured on standard serum protein electrophoresis (R 2 = .1, P = .02).
Discussion Our study confirms the reported association between aortic atherosclerosis with stroke of undetermined mechanism in older patients. We were unable to substantiate an association between aortic atherosclerosis and all-cause stroke. In addition to the smaller numbers in our study, several factors contribute to the lack of association between aortic atherosclerosis and all-cause stroke. One is the choice of controls. In our study, the control population comprised patients at risk for cardioembolic stroke based on the presence of left ventricular dysfunction or atrial fibrillation, usually as a consequence of cardiac ischemia. Thus, our control group was possibly biased toward the presence of aortic atherosclerosis. Nevertheless, although our study was not balanced for smoking, our control population was otherwise balanced for risk factors compared to the stroke group, and is, to our knowledge, the most closely balanced control population with respect to cerebrovascular risk factors, which has been reported to date. A second issue is that the multiple causes of stroke confound the ability to demonstrate a relationship between aortic atherosclerosis and all-cause stroke. Despite our stroke group representing a referral population, we had reasonable proportions of each of the common stroke mechanisms. In our study, patients with stroke on the basis of either cardioembolism or small vessel disease had significantly less atherosclerosis compared to patients with atherosclerotic or undetermined stroke. Thus, we believe our results validate the reports of investigators demonstrating an association between aortic atherosclerosis and stroke of undetermined etiology.3-s Our study provides other, indirect, evidence supporting the association between aortic atherosclerosis and stroke of undetermined etiology. The risk factor profiles of the two groups of patients were similar, both in terms of conventional vascular risk factors, as well as for fibrinogen levels. The degree of aortic atherosclerosis in patients with stroke of undetermined mechanism was similar to those with stroke of atherothrombotic mechanism. It is interesting that our threshold for detection of this association is plaque thickness ->5 mm. This figure is very similar to that reported by others, suggesting that there may be a "threshold" effect.5-8 We believe, therefore, that the cause of stroke in many patients in which this is undetermined is atherosclerotic disease. It remains unclear whether aortic atherosclerosis is directly causative of stroke. In our study, the association was strongest for the maximal degree of atherosclerosis despite location, ie, either the distal or proximal aorta. This suggests that the detection of aortic atherosclerosis is a nonspecific marker of atherosclerosis rather than the specific cause of stroke, unless retrograde embolization is
D.P. BRILEY ET AL
hypothesized. We would therefore suggest treatment strategies for these patients be based on the treatment of atherosclerosis. The recent successes with lipid lowering as exemplified by the 4S trial suggest this therapy may be of use in these patients17; this approach was successful in preventing progression in carotid plaques, m,19Treatment is indicated because patients with stroke and severe aortic atherosclerosis have a high rate of recurrent stroke and other thromboembolic events. 6,2° We found increased levels of fibrinogen in patients with stroke of atherosclerotic or undetermined cause. Both fibrinogen and elevated plasma viscosity are predictors of vascular events during the subsequent 3 to 5 years. 1°'14 The association of increased plasma viscosity with ulcerated plaque was striking. This observation was in our control patients, and thus cannot be attributed to an acute phase response. Fibrinogen is considered to be the primary determinant of plasma viscosity, although other factors can contribute to viscosity.21 In this regard, the association between aortic atherosclerosis and c~2 globulins is of interest, and has been reported previously for coronary heart disease. 22The large variability in the assay for fibrinogen may explain our inability to show an association between ulceration and fibrinogen. Recent studies have highlighted the role of inflammation in atherosclerosis, particularly with plaque disruption. 23,24 We suspect that increased plasma viscosity may represent a consequence of the inflammatory response associated with plaque ulceration. The ability of fibrinogen and plasma viscosity to predict future vascular events may reflect their association with plaque ulceration and plaques with higher degrees of inflammation. In summary, our study indicates that atherosclerosis is the probable cause of stroke in many patients with stroke of undetermined cause. Aortic atherosclerosis detected by TEE may be a surrogate marker for atherosclerosis, rather than directly causative of stroke. Increased plasma viscosity and possibly fibrinogen should alert clinicians to the possibility of plaque ulceration. We suggest that the treatment of these patients should include risk factor control, including lipid lowering agents according to current NCEP guidelines for patients with a recent coronary event. 2s
Acknowledgements: The authors thank Jeyenne Holley and Vickie Harbour for their high quality support in the performance of the transesophageal echocardiograms.
References 1. Amarenco P, Duyckaerts C, Tzourio C, H4nin D, Bousser M-G, Hauw J-J. The prevalence of ulcerated plaques in the aortic arch in patients with stroke. N Engl J Med 1992;326:221-5. 2. Karalis DG, Chandrasekaran K, Victor MF, Ross JJ, Mintz GS. Recognition and embolic potential of intraaortic atherosclerotic debris. J Am Coll Cardiol 1991;17:73-8.
AORTIC ATHEROSCLEROSIS AND STROKE
3. Stone DA, Hawke MW, LaMonte M, Kittner SJ, Acosta J, Corretti M, et al. Ulcerated atherosclerotic plaques in the thoracic aorta are associated with cryptogenic stroke: a multiplane transesophageal echocardiographic study. Am Heart J 1995;130:105-8. 4. Labovitz AJ, Camp A, Castello R, Martin TJ, Ofili EO, Rickmeyer N, et al. Usefulness of transesophageal echocardiography in unexplained cerebral ischemia. Am J Cardiol 1993;72:1448-52. 5. Di Tullio MR, Sacco RL, Gersony D, Nayak H, Weslow RG, Kargman DE, et al. Aortic atheromas and acute ischemic stroke: a transesophageal echocardiographic study in an ethnically mixed population. Neurology 1996;46:1560-66. 6. The French Study of Aortic Plaques in Stroke Group. Atherosclerotic disease of the aortic arch as a risk factor for recurrent ischemic stroke. N Engl J Med 1996;334: 1216-21. 7. Amarenco P, Cohen A, Tzourio C, Bertrand B, Hommel M, Besson G, et al. Atherosclerotic disease of the aortic arch and the risk of ischemic stroke. N Engl J Med 1994;331:1474-79. 8. Jones EF, Kalman JM, Calafiore P, Tonkin AM, Donnan GA. Proximal aortic atheroma: an independent risk factor for cerebral ischemia. Stroke 1995;26:218-24. 9. CouU BM, Beamer N, de Garmo P, Sexton G, Nordt F, Knox R, et al. Chronic blood hyperviscosity in subjects with acute stroke, transient ischemic attack, and risk factors for stroke. Stroke 1991;22:162-68. 10. Yarnell JWG, Baker IA, Sweetnam PM, Bainton D, O'Brien JR, Whithead PJ, et al. Fibrinogen, viscosity, and white blood cell count are major risk factors for ischemic heart disease. Circulation 1991;83:836-44. 11. Resch KL, Ernst E, Matrai A, Paulsen HF. Fibrinogen and viscosity as risk factors for subsequent cardiovascular events in stroke survivors. Ann Intern Med 1992;117: 371-75. 12. Ernst E, Resch KL. Fibrinogen as a cardiovascular risk factor: a meta-analysis and review of the literature. Ann Intern Med 1993;118:956-63. 13. Kannel WB, Wolf PA, Castelli WP, D'Agostino RB. Fibrinogen and risk of cardiovascular disease: the Framingham study. JAMA 1987;258:1183-86. 14. Thompson SG, Kienast J, Pyke SDM, Haverkate F, van de Loo JCW. The European Concerted Action on Thrombosis and Disabilities Angina Pectoris Group. Hemostatic
129
15.
16. 17.
18.
19.
20.
21. 22. 23.
24.
25.
factors and the risk of myocardial infarction or sudden death in patients with angina pectoris. N Engl J Med 1995;332:635-41. Adams HP, Bendixen BH, KappeUe LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke: definitions for use in a multicenter clinical trial. Stroke 1993;24:35-41. Stata Corporation. Stata reference manual; Release 3.1, 6th ed., College Station, TX: Strata Corporation, 1993. Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383-89. Furberg CD, Adams HP, Applegate WB, Byington RP, Espeland MA, Hartwell T, et al for The Asymptomatic Carotid Artery Progression Study (ACAPS) Research Group. Effect of lovastatin on early carotid atherosclerosis and cardiovascular events. Circulation 1994;90: 1679-87. Crouse JR III, Byington RP, Bond MG, Espeland .4., Craven TE, Sprinkle JW, et al. Pravastatin, lipids and atherosclerosis in the carotid arteries (PLAC-II). Am J Cardiol 1995;75:455-59. Tunick PA, Rosenzweig BP, Katz ES, Freedberg RS, Perez JL, Kronzon I. High risk for vascular events in patients with protuding aortic atherornas: a prospective study. J Am Coll Cardiol 1994;23:1085-90. Lowe GDO. Blood viscosity, lipoproteins, and cardiovascular risk. Circulation 1992;85:2329-331. Ducimetiere P, Warnet JM, Richard JL. Raised serum alpha-2 globulins as an independent risk factor of coronary heart disease. J Chron Dis 1976;29:423-29. van der Wal AC, Becker AE, van der Loos CM, Das PK. Site of intimal rupture or erosion of thrombosed coronary atherosclerotic plaques is characterized by an inflammatory process irrespective of the dominant plaque morphology. Circulation 1994;89:36-44. Fuster V, Badimon L, Badimon JJ, Chesebro JH. The pathogenesis of coronary artery disease and the acute coronary syndromes. N Engl J Med 1992;326:242-50 (part 1);310-18 (part 2). National Cholesterol Education Program. Second report of the expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult treatment panel II). Circulation1994;89:1329-445.
Methods: Studies using transesophageal echocardiography (TEE) suggest aortic atherosclerosis may be a risk factor for stroke, particularly stroke of undetermined mechanism, but controls in prior studies were not balanced for vascular risk factors. We used TEE to evaluate aortic atherosclerosis in 60 patients with stroke compared with a high-risk control population of 46 subjects. We also examined the possible association of plasma viscosity and fibrinogen levels to aortic atherosclerosis. Results: The mean maximal plaque thickness (MMPT) was similar for the control (2.8 + 3.6 mm) and the stroke group (3.3 +-3.5 mm), but varied with stroke mechanism. The MMPT was similar in stroke of undetermined and atheroscterotic mechanism [3.5 -+ 4 mm (n = 25) and 4.2 -+ 4.3 rnm (n = 16), respectively], significantly greater than in stroke of other mechanisms (1.7 +- 1.2 ram, P < .05, n = 19). Patients with stroke of undetermined mechanism were four times more likely (95% confidence interval [CI] 1.2-12) to have plaques ->5 mm compared with controls. Ulcerated plaque was associated with plaque thickness (P < .001) and plasma viscosity (P < .001). Conclusions: Aortic atherosclerosis is associated with stroke of undetermined cause suggesting atherosclerosis is a cause of stroke of undetermined etiology. Plaque ulceration was associated with the thickness of aortic plaque and plasma viscosity. Key Words:Stroke--Aortic atherosclerosis--Transesophageal echocardiography--Plasma viscosity--Fibrinogen.
The possibility that atheromatous plaques in the aortic arch m a y lead to cerebral ischemia has been highlighted by A m a r e n c o et al, based on an autopsy study. 1 With the a d v e n t of transesophageal echocardiography (TEE), it has become increasingly feasible to image relevant portions of the aorta. 2 Several studies utilizing TEE have suggested that aortic atheromatous plaques are especially associated with stroke of u n d e t e r m i n e d cause. 3-5Aortic atherosclerosis m a y be a risk factor for recurrent strokeP Furthermore, three studies indicated that aortic plaque is an independent risk factor for stroke, although the choice of controls
From the Department of Medicine, Huntington VAMC and Marshall University School of Medicine, Huntington, WV. Supported by the Department of Veterans Affairs and Marshall University. Received August 12,1996; accepted September 11, 1996. Address reprint requests to Dennis P. Briley, MB, BS, Chief, Neurology Section, Department of Medicine, Marshall University School of Medicine, 180l Sixth Avenue, Huntington, WV 25755. Copyright © 1997 by National Stroke Association
1052-3057/97/0603-0003503.00/0
causes some uncertainty, s,7,8 In the TEE studies of Amarenco et al and Di Tulho et al, patients referred for TEE for reasons unrelated to stroke or aortic atherosclerosis were used as the control population, s,7 The s t u d y of Jones et al used a c o m m u n i t y based age-matched control population, which, although age-matched, was not risk factor matched. 8 We therefore c o m p a r e d TEE findings in patients with stroke to findings in a population without history of stroke, but considered at high risk. Both fibrinogen and plasma viscosity are risk factors for myocardial infarction, stroke, and peripheral vascular disease. 9-14 Accordingly, w e also investigated w h e t h e r plasma viscosity, fibrinogen, and other blood proteins are associated with aortic atherosclerosis detected by TEE in our control subjects.
Methods We studied two groups of patients, a control and a stroke group. The controls were patients enrolled in a TEE s t u d y to evaluate cardiac risk factors for cerebral embo-
Journal of Stroke and Cerebrovascular Diseases, Vol. 6, No. 3, 1997: pp 125-129
125
126
D.P. BRILEY ET AL
lism. This study was a p p r o v e d by the Huntington VAMC institutional review board, a n d all patients gave informed consent. The entry criterion for this group was either atrial fibrillation or left ventricular dysfunction as measured by an ejection fraction less than 45%. All patients were interviewed and examined by a neurologist to exclude a prior symptomatic cerebral ischemic episode. The second group, the "stroke" group, comprised patients referred for TEE after a cerebral ischemic event, either transient or an infarct. A vascular risk factor history was obtained for all patients, including hypertension, diabetes, smoking history, lipid history, and cardiac history. For the stroke group, probable stroke mechanism was classified using the criteria of A d a m s et aD 5 Stroke of u n d e t e r m i n e d etiology was diagnosed based on a negative diagnostic evaluation. All transesophageal echo studies were performed by a cardiologist using a 5 MHz biplane transesophageal transducer and a 128 XP/10 Acuson machine, (Acuson c o m p u t e d sonography, Acuson corporation, CA). The transesophageal probe was m a n i p u l a t e d to obtain adequate views, and all studies were recorded on video tape for further analysis. Each segment of the aorta was examined for atherosclerotic plaques by an observer blinded to clinical data. We measured plaque thickness perpendicular to the vessel wall using software from I m a g e v u e (Nova Microsonic, Nova Microsonic Inc., Carmel, IN). The maximal plaque thickness of the most severe lesion in the proximal (ascending and aortic arch, proximal to the takeoff of the left subclavian artery) and of the distal aorta (descending, distal to the takeoff of the left subclavian artery) were measured. A qualitative assessment was m a d e for ulceration; an ulcer was considered to be present if there was m a r k e d irregularity of the surface with the appearance of a crater. In the control group, measurements were m a d e of plasma, serum, whole blood viscosity, fibrinogen, and blood proteins b y serum protein electrophoresis. Viscosity was measured at 37°C on a Wells Brookfield viscometer. Fibrinogen levels were m e a s u r e d using a modified thrombin clotting time.
Table 1. Demographics of the study population
Factor
Control group
Stroke group
Number Age Diabetes Hypertension Never smoked Ex-smoker Current Smoker Atrial fibrillation CHF
46 66 ± I0 26% (12) 59% (27) 13% (6) 52% (24) 35% (16) 61% (27) 61% (27)
60 64 ± 9 35% (21) 58% (35) 7% (4) 30% (18) 63% (38) 7% (4) 7% (4)
P value
ns ns ns
P < .02 P < .00l P < .001
Aortic atherosclerosis was analyzed as both a continuous variable for the measured thickness of the plaque, and as a dichotomous variable for both ulceration, and for plaque thickness as either ~ 5 m m or <5 mm. Statistical analysis was performed on a personal computer using STATA software. 16
Results The control group contained 46 subjects, and there were 60 patients in the stroke group. Table 1 shows the demograpt~cs for the two groups. The two groups were balanced for age, diabetes, and hypertension, but not for smoking or heart disease. There was a preponderance of current smokers in the stroke group. As expected, all patients in the control g r o u p had either atrial fibrillation or congestive heart failure. Table 2 displays the relationships between aortic atherosclerosis and stroke. Although the mean maximal plaque thickness (MMPT) was similar in the stroke and control group for both the distal and proximal aorta, there was a trend for more patients in the stroke g r o u p to have plaques ->5 m m c o m p a r e d with the control group. Plaque thickness was significantly related to stroke mechanism. More patients with stroke of atherothrombotic or undetermined mechanism had plaques ---5 m m c o m p a r e d with either the control group or c o m p a r e d with stroke of other
Table 2. Plaque sever#)., and ulceration with stroke broken down by stroke mechanisn~
Control (n = 46) Stroke group (aggregate) Atherothrombotic (n = 16) Undetermined (n = 25) Cardioembolic (n = 5) L a c u n a r ( n = 12) Other (n = 2)
mmpt (ram)
Proximal mmpt (ram)
Distal mmpt (mm)
Plaque >5 mm
Ulceration
2.8 _-z-3.6 3.3 --- 3.5 4.2 _ 4.3 3.5 --- 4 1.6 --- 1.5 1.8± 1.1 1
1.6 ± 2.0 1.8 --- 3 2.4 ± 2.3 2.3 ± 3.9 0.8 ± I 0 . 8 ± 1.1 0.5 --+ 0.7
2.6 __. 3 2.9 + 3.6 3.5 - 4.4 3.4 _ 4 1.6 ± 1.5 1.6± 1.2 I
13% (6) 25% (15)t 38% (6)* 36% (9)* 0 0 0
9% (4) 5% (3) t3% (2) 1% (4) 0 0 0
*P < .05 compared with controls; t P = .13 compared with controls.
AORTIC ATHEROSCLEROSIS AND STROKE
127
Table 3. Comparison of stroke risk factors by stroke mechanism Atherothrombotic
Undetermined
Other
Control
Factor
N = 16
N=25
N = 19
N=46
Age (years) Diabetes Hypertension Non-smoker Current smoker Cholesterol (mmol/L) Fibrinogen (g/L)
63 ± 8 25% (4) 69% (11) 31% (5) 69% (11) 5.77 ± 0.96* 4.07 ± 1.27
65 ± 8 20% (5) 48% (12) 36% (9) 64% (16) 5.51 ± 1.29" 4.12 ± 1.14
65 ± 10 63% (12)I" 63% (12) 43% (8) 58% (11) 5.46 ± 0.91 3.70 - 0.90
66 - 10 26% (12) 59% (27) 65% (30) 35% (16) 4.99 --+ 1.03 3.80 --+ 1.18
*P < .02 for mean cholesterol level in atherothrombotic and undetermined group combined compared with control. t P = .01 for presence of diabetes in stroke of other etiology compared with stroke of atherothrombotic/undetermined etiology or compared with controls.
etiology. The o d d s ratio (OR) for the presence of plaque ->5 m m in patients with stroke of u n d e t e r m i n e d mechanism c o m p a r e d with controls was 4 (95% CI 1.2-12). The OR was also 4 (95% CI 1.1-15) for atherothrombotic stroke c o m p a r e d with controls. The MMPT was similar in both the atherothrombotic and u n d e t e r m i n e d stroke patients and was significantly greater c o m p a r e d with patients with stroke of other etiology (3.8 ± 4 m m versus 1.7 --- 1.2 ram, P < .05), b u t not significantly greater than the controls. Although our s t u d y was unbalanced for smoking, the OR for plaque ->5 m m in nonsmokers with atherothrombotic a n d u n d e t e r m i n e d stroke was 4 (95% CI 0.8-15), and in smokers was also 4 (95% CI 0.8-~) compared with controls. In addition, logistic regression found both plaque ->5 m m and smoking were i n d e p e n d e n t l y associated with a t h e r o t h r o m b o t i c / u n d e t e r m i n e d stroke combined c o m p a r e d with controls (OR for plaque ->5 m m 4, 95% CI 1.3-12, P = .02; OR for current smoking 4, 95% CI 1.4-9, P = .006). The risk factor profiles for patients with stroke of atherothrombotic m e c h a n i s m a n d stroke of undetermined mechanism, s h o w n in Table 3, were similar. Mean cholesterol levels were higher in patients with stroke of atherothrombotic and u n d e t e r m i n e d mechanism compared with controls; there was also a trend toward higher fibrinogen levels in these two groups. The prevalence of diabetes was greater in patients with stroke of other mechanism (predominantly cardioembolic and small vessel disease). The degree of atherosclerosis in the distal aorta was more severe than the proximal aorta (P < .0002), although the two were significantly correlated (R 2 = .53, P < .0001). Mobile lesions, p r e s u m e d to be thrombi, were seen in two subjects only, one control, and one patient with stroke of u n d e t e r m i n e d etiology, respectively. Data on carotid stenosis were available for 93% (56/60) of the stroke group. There was a trend to more severe carotid stenosis in patients with plaques ->5 m m in our stroke patients, so that 43% (6/14) h a d carotid stenosis >50% and plaques
->5 r a m c o m p a r e d to 24% (10/42) of patients with plaque < 5 ram, (P = .19). Ulcerated plaques were detected in 7 patients. There were no obvious differences in standard vascular risk factors between patients with ulcers and those without. Table 4 shows that subjects with ulcerated plaques had increased p l a s m a and serum viscosity, along with a trend to increased fibrinogen levels. Patients w i t h ulcerated plaques h a d m a r k e d l y increased plaque thickness. Multiple linear regression analysis f o u n d both plaque thickness (P < .001) and plasma viscosity (P < .001) were i n d e p e n d e n t variables associated with the presence of ulceration (R 2 = .64, P < .0001). N o correlation between plaque thickness to age, cholesterol, p l a s m a viscosity or fibrinogen was detected. There was a slight correlation of
Table 4. Risk factors for ulcerated plaque
Factor
Ulcerated
No Ulcer
N = 7
N = 99
Age (years) 68 ± 11 65 - 9 Hypertension 43% (3) 60% (59) Diabetes 29% (2) 31% (31) Never smoker 0 10% (10) Ex-smoker 43% (3) 39% (39) Current smoker 57% (4) 60% (59) MMPT (mm) 11.3 ± 5.5 2.4 - 2.5 Fibrinogen (g/L) 4.46 ± 0.61 3.73 ± 1.21 Plasma viscosity (mPas)* 2.3 ± 0.3 1.6 ± 0.2 Serum viscosity (mPas)* 1.5 +-_0.3 1.3 ± 0.1 Whole blood viscosity at 23s -I(mPas)* 6 . 6 ± 1.4 7 . 9 ± 1.5 Whole blood viscosity at 230 s-I (mPas)* 4.2 ± 0.7 4.9 ± 0.6 *Data from control subjects only.
P value ns ns ns
ns P < .0001 P = .08 P < .0001 P < .02 P=.I P = .1
128 plaque thickness to alpha 2 globulins measured on standard serum protein electrophoresis (R 2 = .1, P = .02).
Discussion Our study confirms the reported association between aortic atherosclerosis with stroke of undetermined mechanism in older patients. We were unable to substantiate an association between aortic atherosclerosis and all-cause stroke. In addition to the smaller numbers in our study, several factors contribute to the lack of association between aortic atherosclerosis and all-cause stroke. One is the choice of controls. In our study, the control population comprised patients at risk for cardioembolic stroke based on the presence of left ventricular dysfunction or atrial fibrillation, usually as a consequence of cardiac ischemia. Thus, our control group was possibly biased toward the presence of aortic atherosclerosis. Nevertheless, although our study was not balanced for smoking, our control population was otherwise balanced for risk factors compared to the stroke group, and is, to our knowledge, the most closely balanced control population with respect to cerebrovascular risk factors, which has been reported to date. A second issue is that the multiple causes of stroke confound the ability to demonstrate a relationship between aortic atherosclerosis and all-cause stroke. Despite our stroke group representing a referral population, we had reasonable proportions of each of the common stroke mechanisms. In our study, patients with stroke on the basis of either cardioembolism or small vessel disease had significantly less atherosclerosis compared to patients with atherosclerotic or undetermined stroke. Thus, we believe our results validate the reports of investigators demonstrating an association between aortic atherosclerosis and stroke of undetermined etiology.3-s Our study provides other, indirect, evidence supporting the association between aortic atherosclerosis and stroke of undetermined etiology. The risk factor profiles of the two groups of patients were similar, both in terms of conventional vascular risk factors, as well as for fibrinogen levels. The degree of aortic atherosclerosis in patients with stroke of undetermined mechanism was similar to those with stroke of atherothrombotic mechanism. It is interesting that our threshold for detection of this association is plaque thickness ->5 mm. This figure is very similar to that reported by others, suggesting that there may be a "threshold" effect.5-8 We believe, therefore, that the cause of stroke in many patients in which this is undetermined is atherosclerotic disease. It remains unclear whether aortic atherosclerosis is directly causative of stroke. In our study, the association was strongest for the maximal degree of atherosclerosis despite location, ie, either the distal or proximal aorta. This suggests that the detection of aortic atherosclerosis is a nonspecific marker of atherosclerosis rather than the specific cause of stroke, unless retrograde embolization is
D.P. BRILEY ET AL
hypothesized. We would therefore suggest treatment strategies for these patients be based on the treatment of atherosclerosis. The recent successes with lipid lowering as exemplified by the 4S trial suggest this therapy may be of use in these patients17; this approach was successful in preventing progression in carotid plaques, m,19Treatment is indicated because patients with stroke and severe aortic atherosclerosis have a high rate of recurrent stroke and other thromboembolic events. 6,2° We found increased levels of fibrinogen in patients with stroke of atherosclerotic or undetermined cause. Both fibrinogen and elevated plasma viscosity are predictors of vascular events during the subsequent 3 to 5 years. 1°'14 The association of increased plasma viscosity with ulcerated plaque was striking. This observation was in our control patients, and thus cannot be attributed to an acute phase response. Fibrinogen is considered to be the primary determinant of plasma viscosity, although other factors can contribute to viscosity.21 In this regard, the association between aortic atherosclerosis and c~2 globulins is of interest, and has been reported previously for coronary heart disease. 22The large variability in the assay for fibrinogen may explain our inability to show an association between ulceration and fibrinogen. Recent studies have highlighted the role of inflammation in atherosclerosis, particularly with plaque disruption. 23,24 We suspect that increased plasma viscosity may represent a consequence of the inflammatory response associated with plaque ulceration. The ability of fibrinogen and plasma viscosity to predict future vascular events may reflect their association with plaque ulceration and plaques with higher degrees of inflammation. In summary, our study indicates that atherosclerosis is the probable cause of stroke in many patients with stroke of undetermined cause. Aortic atherosclerosis detected by TEE may be a surrogate marker for atherosclerosis, rather than directly causative of stroke. Increased plasma viscosity and possibly fibrinogen should alert clinicians to the possibility of plaque ulceration. We suggest that the treatment of these patients should include risk factor control, including lipid lowering agents according to current NCEP guidelines for patients with a recent coronary event. 2s
Acknowledgements: The authors thank Jeyenne Holley and Vickie Harbour for their high quality support in the performance of the transesophageal echocardiograms.
References 1. Amarenco P, Duyckaerts C, Tzourio C, H4nin D, Bousser M-G, Hauw J-J. The prevalence of ulcerated plaques in the aortic arch in patients with stroke. N Engl J Med 1992;326:221-5. 2. Karalis DG, Chandrasekaran K, Victor MF, Ross JJ, Mintz GS. Recognition and embolic potential of intraaortic atherosclerotic debris. J Am Coll Cardiol 1991;17:73-8.
AORTIC ATHEROSCLEROSIS AND STROKE
3. Stone DA, Hawke MW, LaMonte M, Kittner SJ, Acosta J, Corretti M, et al. Ulcerated atherosclerotic plaques in the thoracic aorta are associated with cryptogenic stroke: a multiplane transesophageal echocardiographic study. Am Heart J 1995;130:105-8. 4. Labovitz AJ, Camp A, Castello R, Martin TJ, Ofili EO, Rickmeyer N, et al. Usefulness of transesophageal echocardiography in unexplained cerebral ischemia. Am J Cardiol 1993;72:1448-52. 5. Di Tullio MR, Sacco RL, Gersony D, Nayak H, Weslow RG, Kargman DE, et al. Aortic atheromas and acute ischemic stroke: a transesophageal echocardiographic study in an ethnically mixed population. Neurology 1996;46:1560-66. 6. The French Study of Aortic Plaques in Stroke Group. Atherosclerotic disease of the aortic arch as a risk factor for recurrent ischemic stroke. N Engl J Med 1996;334: 1216-21. 7. Amarenco P, Cohen A, Tzourio C, Bertrand B, Hommel M, Besson G, et al. Atherosclerotic disease of the aortic arch and the risk of ischemic stroke. N Engl J Med 1994;331:1474-79. 8. Jones EF, Kalman JM, Calafiore P, Tonkin AM, Donnan GA. Proximal aortic atheroma: an independent risk factor for cerebral ischemia. Stroke 1995;26:218-24. 9. CouU BM, Beamer N, de Garmo P, Sexton G, Nordt F, Knox R, et al. Chronic blood hyperviscosity in subjects with acute stroke, transient ischemic attack, and risk factors for stroke. Stroke 1991;22:162-68. 10. Yarnell JWG, Baker IA, Sweetnam PM, Bainton D, O'Brien JR, Whithead PJ, et al. Fibrinogen, viscosity, and white blood cell count are major risk factors for ischemic heart disease. Circulation 1991;83:836-44. 11. Resch KL, Ernst E, Matrai A, Paulsen HF. Fibrinogen and viscosity as risk factors for subsequent cardiovascular events in stroke survivors. Ann Intern Med 1992;117: 371-75. 12. Ernst E, Resch KL. Fibrinogen as a cardiovascular risk factor: a meta-analysis and review of the literature. Ann Intern Med 1993;118:956-63. 13. Kannel WB, Wolf PA, Castelli WP, D'Agostino RB. Fibrinogen and risk of cardiovascular disease: the Framingham study. JAMA 1987;258:1183-86. 14. Thompson SG, Kienast J, Pyke SDM, Haverkate F, van de Loo JCW. The European Concerted Action on Thrombosis and Disabilities Angina Pectoris Group. Hemostatic
129
15.
16. 17.
18.
19.
20.
21. 22. 23.
24.
25.
factors and the risk of myocardial infarction or sudden death in patients with angina pectoris. N Engl J Med 1995;332:635-41. Adams HP, Bendixen BH, KappeUe LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke: definitions for use in a multicenter clinical trial. Stroke 1993;24:35-41. Stata Corporation. Stata reference manual; Release 3.1, 6th ed., College Station, TX: Strata Corporation, 1993. Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383-89. Furberg CD, Adams HP, Applegate WB, Byington RP, Espeland MA, Hartwell T, et al for The Asymptomatic Carotid Artery Progression Study (ACAPS) Research Group. Effect of lovastatin on early carotid atherosclerosis and cardiovascular events. Circulation 1994;90: 1679-87. Crouse JR III, Byington RP, Bond MG, Espeland .4., Craven TE, Sprinkle JW, et al. Pravastatin, lipids and atherosclerosis in the carotid arteries (PLAC-II). Am J Cardiol 1995;75:455-59. Tunick PA, Rosenzweig BP, Katz ES, Freedberg RS, Perez JL, Kronzon I. High risk for vascular events in patients with protuding aortic atherornas: a prospective study. J Am Coll Cardiol 1994;23:1085-90. Lowe GDO. Blood viscosity, lipoproteins, and cardiovascular risk. Circulation 1992;85:2329-331. Ducimetiere P, Warnet JM, Richard JL. Raised serum alpha-2 globulins as an independent risk factor of coronary heart disease. J Chron Dis 1976;29:423-29. van der Wal AC, Becker AE, van der Loos CM, Das PK. Site of intimal rupture or erosion of thrombosed coronary atherosclerotic plaques is characterized by an inflammatory process irrespective of the dominant plaque morphology. Circulation 1994;89:36-44. Fuster V, Badimon L, Badimon JJ, Chesebro JH. The pathogenesis of coronary artery disease and the acute coronary syndromes. N Engl J Med 1992;326:242-50 (part 1);310-18 (part 2). National Cholesterol Education Program. Second report of the expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult treatment panel II). Circulation1994;89:1329-445.