Vascular Events During Follow-up in Patients With Aortic Arch Atherosclerosis
Background and Purpose An association between aortic arch atherosclerosis and vascular events has been demonstrated. However, few data exist regarding follow-up evaluation of this disease.
Methods In this study, 183 patients with the diagnosis of aortic arch atherosclerosis were prospectively followed up. This diagnosis was made during an echocardiographic cross-sectional study. In 136 patients, raised plaques with thickness <5 mm had been shown to exist, and in 47 patients complex plaques with thickness ≥5 mm or plaques with mobile components had been demonstrated on the initial transesophageal echocardiography.
Results During a mean follow-up period of 16±7 months, vascular events with a presumed embolic origin occurred in 15 patients. The incidence was 4.1 per 100 person-years in patients with raised plaques compared with 13.7 per 100 person-years in the group with complex plaques. The Kaplan-Meier survival analysis revealed a significantly higher rate of vascular events in patients who were found to have complex plaques (P<.01). In the Cox proportional hazards analysis, the finding of complex plaques (relative risk [RR], 4.3; 95% confidence interval [CI], 1.5 to 12.0; P=.006), coronary artery disease (RR, 4.0; 95% CI, 1.2 to 13.1; P=.02), and a history of previous embolism (RR, 4.0; 95% CI, 1.1 to 14.4; P=.03) were independent predictors of vascular events.
Conclusions Patients with the finding of protruding plaques or plaques with mobile components have a high risk of subsequent vascular events.
It has been shown that atherosclerosis of the aortic arch is a risk factor for systemic embolism.1 2 3 4 Recently published echocardiographic and pathological studies revealed that the relative risk arising from advanced atherosclerosis of the aortic arch is as important as that of established sources of embolism including atrial fibrillation, left atrial thrombi, and large-artery atherosclerosis.5 6
Few studies exist regarding follow-up evaluation of patients with atherosclerotic changes of the aorta indicating an elevated risk for subsequent embolic events.7 8 Therefore, patients who were found to have moderate to severe aortic arch atherosclerosis detected by transesophageal echocardiography were prospectively followed up in the present study.
Subjects and Methods
Patients in whom atherosclerosis of the aortic arch with raised plaques had been detected by transesophageal echocardiography in a previous echocardiographic cross-sectional study5 were enrolled in the present study. The initial echocardiographic examination was performed between November 1991 and November 1993 and included patients who had been referred to the echocardiographic laboratory by their attending physicians with standard indications.5 Enrollment criteria were met by 196 patients, 13 (6.6%) of whom were lost to follow-up. The study group comprised 183 patients (107 men and 76 women) with a mean age of 69±10 years, who were followed up over a mean period of 16±7 months. Follow-up information was obtained from the patients, their attending physicians, or hospital charts, when the patients had required inpatient treatment during follow-up.
The following echocardiographic variables were collected at the enrollment examination: atherosclerosis of the aortic arch was classified as moderate when raised plaques extending <5 mm into the aortic lumen were detected or as complex when plaques ≥5 mm in thickness or plaques with mobile components irrespective of the thickness were found.5 Plaques located at the junction of the ascending aorta and the arch were assigned to the aortic arch because the innominate artery cannot be adequately visualized by transesophageal echocardiography.
Cardiac abnormalities included atrial fibrillation, cardiac thrombi, atrial septal aneurysm,9 left atrial spontaneous contrast,10 aortic valve calcification, mitral annular calcification, and infective endocarditis. Left ventricular dysfunction was defined as marked regional or generalized wall motion abnormalities documented by either transthoracic or transesophageal echocardiography.
The clinical variables included in the analysis were age, sex, systemic hypertension (defined as either diastolic pressure >95 mm Hg or systolic pressure >160 mm Hg or ongoing antihypertensive therapy), diabetes mellitus (defined as a fasting glucose level >10 mmol/L or ongoing antidiabetic therapy), hypercholesterolemia (defined as a fasting total cholesterol level >6.5 mmol/L), cigarette smoking, history of coronary artery disease, and history of embolic events before enrollment.
Ultrasound examination of the carotid arteries had been performed for all patients who had suffered a cerebral infarction but for only some asymptomatic patients. Therefore, these findings were not included in the multivariate analysis.
During follow-up, new vascular events with a presumed embolic origin were recorded, including (1) stroke (CT or MRI evidence of cerebral infarction or typical neurological symptoms); (2) visceral embolism (confirmed by surgery or autopsy; acute impairment of renal function with no other etiology); and (3) peripheral embolism (surgically proven or sudden onset of typical symptoms) or a typical blue toe syndrome.11 12 Transient ischemic attacks were registered and analyzed separately. They were not included in the multivariate analysis.
The χ2 test was used for comparison of noncontinuous variables, and ANOVA was used for comparison of means. The incidence of vascular events in the groups was compared by means of a Kaplan-Meier survival analysis and a log-rank test. In a multivariate approach, the Cox proportional hazards method was used to assess the contribution of clinical and echocardiographic variables to the development of vascular events and death during follow-up. Variables included in the Cox model were age, sex, aortic arch plaque (thickness <5 mm, thickness or ≥5 mm or mobile components), atrial fibrillation, left atrial spontaneous contrast, thrombus, infective endocarditis, atrial septal aneurysm, aortic valve calcification, mitral annular calcification, left ventricular dysfunction, history of embolic events, systemic hypertension, diabetes, hypercholesterolemia, coronary artery disease, antiplatelet therapy, and anticoagulant therapy. We constructed the model using the forward stepwise method with removal testing based on the probability of the likelihood-ratio statistic based on conditional parameter estimates (probability was .05 for entry and .10 for removal). Relative risks were calculated with 95% confidence intervals. Statistical significance was defined as P<.05. We analyzed the data using the SPSS for Windows statistical package (version 5.0.2).
At the enrollment examination, the extent of aortic arch atherosclerosis was defined as raised plaques in 136 patients (74%) and as complex plaques in 47 patients (26%), including 30 patients with protruding lesions and 17 patients with plaque-related mobile masses. No patient was found to have severe atherosclerotic changes in the ascending aorta. Baseline characteristics of the patients according to the grade of atherosclerosis are provided in Table 1⇓.
During a follow-up period of 241 patient-years, vascular events with a presumed embolic origin occurred in 15 patients (Table 2⇓). The incidence was 4.1 per 100 person-years in patients with raised plaques compared with 13.7 per 100 person-years in those with complex plaques (Table 3⇓). Significant difference could be demonstrated by Kaplan-Meier analysis in the survival curves of the two groups of patients (P<.01; Figure).⇓ Independent predictors of subsequent embolic events in the Cox proportional hazards analysis were the finding of complex plaques in the aortic arch, coronary artery disease, and a history of previous embolic events (Table 4⇓). Among patients with new events during follow-up, other potential sources of embolism, including carotid stenosis, were present in 6 of 7 patients with raised plaques compared with 1 of 8 patients with complex lesions (P<.005). In 9 of 12 patients, stroke was confirmed by CT or pathological examination.
Two additional events occurred during the follow-up period that were sequelae of invasive procedures, including catheterization of the aorta and cardiac surgery, and another two transient ischemic events also occurred. If these four events are included in the analysis, event rates increase to 4.6 per 100 person-years in patients with raised plaques and 18.9 per 100 person-years in patients with complex aortic arch atherosclerosis. Table 3⇑ also provides incidences of events separately for patients who had been symptomatic before enrollment and those who had been asymptomatic. No information about vascular events could be obtained in the case of 4 patients who died during the course of the study.
Fifty-two patients (28%) of the study group died during follow-up. The presumed cause of death was reported to be cardiovascular in 23 patients (44%), the sequelae of embolic events in 10 patients (19%), malignant tumors in 2 patients (4%), the sequelae of surgery or diagnostic procedures in 4 patients (8%), pulmonary embolus in 2 patients (4%), sepsis or severe infection in 4 patients (8%), complications of renal failure in 2 patients (4%), and undetermined in 5 patients (10%).
The predictive value of clinical and echocardiographic variables for death during follow-up was calculated by the Cox proportional hazards method. Analysis showed a significant predictive value for age, left ventricular dysfunction, diabetes mellitus, and aortic valve calcification (Table 5⇓).
Various studies have demonstrated an association between complex atherosclerosis of the aorta and systemic embolism.1 2 3 4 5 13 The relative risk was mainly correlated with the thickness of the plaques and the presence of mobile components associated with plaques.5 13 Most studies have used a cross-sectional design, enrolling patients in whom vascular events had already occurred at the time of enrollment. Few data have existed until now regarding the prospective evaluation of patients with a confirmed diagnosis of aortic arch atherosclerosis. Therefore, we conducted the present follow-up study of patients with moderate to complex atherosclerotic lesions detected by transesophageal echocardiography. We found that patients who had complex plaques with either thickness ≥5 mm or mobile components in the aortic arch developed significantly more vascular events during follow-up than patients who were found to have raised plaques with thickness <5 mm.
The incidence of events with a presumed embolic origin rose from 4.1 in patients with moderate atherosclerosis to 13.7 per 100 person-years in patients with complex atherosclerosis of the aortic arch. Additionally, symptomatic patients with moderate plaques had a significantly greater number of other potential sources of embolism than symptomatic patients with complex lesions. In most of the latter patients, events would have been classified as events of unknown cause without the recognition of aortic arch atherosclerosis. For patients with complex atherosclerosis who had been symptomatic before enrollment, the incidence of new events was 15.9 per 100 person-years. The combined incidence of recurrent stroke and peripheral embolism recently found by the French Study Group for patients with a plaque thickness ≥4 mm was 16.2 per 100 person-years, which compares well with our data.8 Tunick and coworkers7 observed 19 vascular events among 14 of 42 study patients with protruding atheromas of the thoracic aorta during a mean follow-up of 13 months, which yielded an event rate of approximately 30%. The patients in the latter study population were highly selected, which reinforces the evidence of a causative link between aortic arch atherosclerosis and vascular events but does not take into account the influence of other variables. All studies consistently show that aortic arch atherosclerosis carries a significant risk for vascular events during follow-up evaluation.
A different conclusion was reached by Mitchell and coworkers14 who, during a mean period of 7 months, followed up 5 patients in whom plaques with mobile components had been detected by intraoperative transesophageal echocardiography. The authors recorded only one event of a transient left foot claudication during the follow-up period and inferred a benign course for the aortic lesions studied. However, several objections to this conclusion have been raised.15 The patients were recruited during cardiac surgery and thus presumably represent a positive selection with respect to the patients' general condition and age. The aortic lesions were detected in the descending aorta in 4 of the 5 patients. Either this segment appears to have a lower embolic potential or the proportion of clinically silent events is higher. Accordingly, pathological studies have demonstrated that visceral embolism is often not diagnosed.6 If one interpreted the transient limb ischemia as an embolic event, an annual event rate of almost 12% could be calculated from the aforementioned follow-up observation.
In a multivariate analysis, the predictive value of aortic arch plaques for the development of events was adjusted for the presence of clinical and echocardiographic variables including cardiac sources of embolism, risk factors, and therapy. The presence of complex atherosclerosis was an independent predictor of vascular events with a presumed embolic origin other than coronary artery disease and a history of previous embolism in the Cox proportional hazards analysis.
A substantial proportion of patients died during the course of the study. The major causes of death were coronary artery disease and sequelae of embolic events. Significant predictors of death during follow-up as calculated by Cox proportional hazards analysis were left ventricular dysfunction at enrollment, diabetes mellitus, age, and aortic valve calcification. The predictive value of aortic valve calcification is understood as a general marker for atherosclerosis, since no patient with aortic stenosis who had not undergone surgery was in the study group.
The present study has certain limitations. The patients were recruited from the echocardiographic laboratory and therefore represent a selected population. Although in our hospital all patients who sustain embolic events are usually referred for transesophageal echocardiography except those who are critically ill, recruitment bias has to be presumed. However, a large number of clinical studies chose this kind of access and presented findings comparable to those of pathological studies1 6 and clinical studies that examined consecutive patients with cerebral infarctions.13 One must presume that the documented risks are higher than in the general population, since the enrolled patients were found to have a high morbidity and mortality.
In conclusion, the present findings indicate that complex aortic arch atherosclerosis, including protruding plaques and plaque-related mobile masses, is associated with a high risk of subsequent vascular events.
We gratefully acknowledge the statistical advice of Dr Thomas Kohlmann.
- Received February 22, 1996.
- Revision received September 27, 1996.
- Accepted October 1, 1996.
- Copyright © 1997 by American Heart Association
Tunick PA, Perez JL, Kronzon I. Protruding atheromas in the thoracic aorta and systemic embolization. Ann Intern Med. 1991;115:423-427.
Jones EF, Kalman JM, Calafiore P, Tonkin AM, Donnan GA. Proximal aortic atheroma: an independent risk factor for cerebral ischemia. Stroke. 1996;26:218-224.
Mitusch R, Stierle U, Tepe C, Kummer-Kloess D, Kessler C, Sheikhzadeh A. Systemic embolism in aortic arch atheromatosis. Eur Heart J. 1994;15:1373-1380.