Right–Left Propensity and Lesion Patterns Between Cardiogenic and Aortogenic Cerebral Embolisms
Background and Purpose—Based on thrombus location and nature and anatomic features of aorta and cerebral arteries, we hypothesized that cardiogenic embolisms (CE) and aortogenic embolisms (AE) might have different right–left propensity and lesion patterns.
Methods—We retrospectively reviewed patients with acute ischemic stroke with high-risk CE sources or moderate-or-severe aortic atherosclerotic plaques on transesophageal echocardiography. Lesion side and patterns on diffusion-weighted imaging were compared between CE and AE.
Results—CE was identified in 123 and AE in 63. In multivariate analysis, right-sided lesions and corticosubcortical infarcts were independently associated with CE, and left-sided lesions and pial infarcts were independently associated with AE.
Conclusions—CE and AE have different radiological characteristics, as shown by the right–left propensity and lesions patterns of cerebral infarcts.
The majority of cryptogenic strokes have infarct patterns compatible with embolic features, suggesting they were caused by cardiogenic (CE) or aortogenic (AE) embolisms. Although several studies have focused on the lesion patterns in cardiogenic and aortogenic infarctions,1,2 their radiological differences remain largely unknown.
The direction of cerebral embolization may be affected by anatomic and rheological characteristics of the aortic segment and the location of thrombi (http://stroke.ahajournals.org).3 In addition, CE and AE have distinct pathological mechanisms of thrombus formation in terms of fibrin-rich and platelet-rich thrombi.4 Thus, we hypothesized that right–left propensity and lesion patterns of cerebral infarcts might differ between CE and AE.
We retrospectively assessed all patients with stroke admitted to the Stroke Center at Asan Medical Center, Seoul, South Korea, between November 1, 2002, and October 30, 2008. Patients were screened if they had an acute symptomatic stroke identified on diffusion-weighted imaging and if transesophageal echocardiography was performed during admission to include those with definitive isolated high-risk CE or AE. CE was defined if there were high-risk cardioembolic sources according to SSS-TOAST classification.5 AE was defined if there was a moderate or severe degree of atherosclerotic plaques (≥4 mm thick) with or without complex plaques (eg, mobile debris, ulcerating plaques) in the ascending aorta or aortic arch.6 Patients with low- or uncertain risk of CE, a mild degree of aortic atherosclerosis, or combined CE and AE were excluded from this study. The study was approved by the Institutional Review Board of our center. Patient informed consent was not required due to the retrospective design.
Lesion side and patterns on diffusion-weighted imaging were determined blinded to stroke etiology. Right-sided lesions were defined as those in the territories of right innominate artery and right posterior inferior cerebellar artery, whereas left-sided lesions were defined as those in the territories of left common carotid artery and left posterior inferior cerebellar artery. We excluded patients with infarcts in the territories of anterior inferior cerebellar, basilar, and posterior cerebral arteries, because these lesions could occur by embolization through either side of vertebral arteries. In addition, anatomic variations (common trunk of anterior cerebral arteries or fetal-type posterior cerebral artery) were considered when determining the side of the index lesions. Diffusion-weighted imaging lesions in each patient were classified into 1 of 8 patterns: 4 single and 4 multiple patterns (Table).
Clinical and imaging characteristics were compared between the CE and AE groups using Student t test, Mann–Whitney U test, or χ2 or Fisher exact tests as appropriate. Multiple logistic regression analysis was used to determine independent factors of CE and AE.
Of a total of 545 patients who had acute symptomatic stroke and underwent transesophageal echocardiography, 186 patients met the eligibility criteria of this study: 97 (52.2%) were men, and the median age was 66 years (range, 20 to 89 years). CE was diagnosed in 123 patients and AE in 63. The most common etiology of CE was atrial fibrillation (n=96 [60.4%]) followed by rheumatic valve disease (n=21 [13.2%]). Of the 63 patients with AE, 38 (60.3%) had a severe and 25 (39.7%) had a moderate degree of aortic atherosclerosis. Patients with AE were older (P=0.003) and had diabetes mellitus (P=0.012) and hypercholesterolemia (P=0.012) more frequently than patients with CE.
Patients with CE were more likely to have right-sided lesions, whereas patients with AE more frequently had left-sided lesions. Of the lesion patterns, corticosubcortical infarcts were more frequently associated with CE and pial infarcts were more often associated with AE (Table; Figure). The frequency of multiple lesions was similar between CE (46.3%) and AE (50.8%).
In multivariate analysis, right-sided lesions (OR, 2.74; 95% CI, 1.34 to 5.58), corticosubcortical infarcts (OR, 2.35; 95% CI, 1.09 to 5.09), and female (OR, 2.15; 95% CI, 1.03 to 4.47) were independently associated with CE. In contrast, left-sided lesions (OR, 2.77; 95% CI, 1.39 to 5.53), pial infarcts (OR, 3.0; 95% CI, 1.05 to 8.62), age (OR, 1.07; 95% CI, 1.03 to 1.11), and diabetes (OR, 2.53; 95% CI, 1.10 to 5.79) were independently associated with AE.
To our knowledge, this study is the first to compare the radiological characteristics of CE and AE.
Several mechanisms may explain the differing right–left propensity of cerebral infarcts between CE and AE. First, the innominate artery supplying the right brain has the largest caliber and heads upward and parallel to the direction of ascending aorta, whereas the left carotid artery arises perpendicular to the aortic arch. Consequently, cardiogenic emboli may flow more frequently into the right hemisphere. Second, aortic plaques are more commonly observed in the distal aortic arch than in the ascending aorta,3 and thereby aortic plaques and superimposed thrombus may easily obliterate left carotid or vertebral arteries, resulting in left-sided strokes. Nevertheless, we found that approximately one third of patients with AE had right-sided lesions. Emboli from aortic plaques may move backward into the right hemisphere because the aorta has normally turbulent flows in the diastolic stage.7
We also found that CE was associated with large corticosubcortical infarcts, whereas AE was associated with pial infarcts. AE is thought to be due to plaque rupture and to the embolization of platelet-rich small thrombi similar to the stroke mechanism in internal carotid artery stenosis.1 In contrast, CE is thought to be due primarily to the stasis-related formation of thrombi within the cardiac chamber and the subsequent dislodgement of large-sized thrombi.
This study had several limitations. The study design was retrospective. Although our center has a standardized stroke work-up protocol, transesophageal echocardiography was performed in selected patients. Transesophageal echocardiography may underestimate atherosclerosis in the more proximal aortic arch and ascending aorta owing to imaging in the far field and interference by air in the trachea and bronchus.
Source of Funding
This study was supported by a grant of the Korea Healthcare Technology R&D Project, Ministry for Health, Welfare and Family Affairs, Republic of Korea (A080201).
The online-only Data Supplement is available at http://stroke.ahajournals.org/cgi/content/full/STROKEAHA.110.616573/DC1.
- Received February 7, 2011.
- Accepted February 23, 2011.
- © 2011 American Heart Association, Inc.
- Lee DK,
- Kim JS,
- Kwon SU,
- Yoo SH,
- Kang DW
- Kronzon I,
- Tunick PA