Treatable potential cardiac sources of embolism in patients with cerebral ischemic events: a selective transesophageal echocardiographic study.

Objectives: To characterize cardiac sources of emboli detected by transesophageal echocardiography (TEE) in patients without recognizable cause of transient ischemic attack (TIA) and/or ischemic stroke and TIA.

Methods: We examined a prospective registry that included all patients with TIA and/or stroke evaluated by TEE between July 2000 and August 2001 at our medical center. Ischemic events were classified according to clinical and neuroimaging findings as cortical, lacunar, or vertebrobasilar circulation. Demographic, clinical, neuroimaging, and echocardiographic characteristics were studied.

Results: TEE was performed in 237 patients without recognized cause of TIA and/or stroke, of which 105 (44%) events were cortical, 35 (15%) were lacunar, 32 (14%) were vertebrobasilar, and 65 (27%) were in multiple distributions. Mean age of patients was 59 [+ or -] 14 years (range, 21 to 93 years); 119 (51%) were men. Potential treatable cardioembolic sources were detected in 146 (61%) patients: patent foramen ovale with right-to-left shunt (n = 59), left atrial clot (n = 6), left atrial appendage clot (n = 8), and severe thoracic aortic atherosclerotic plaque disease (plaque thickness >4 mm) (n = 79 patients [33%], 56 had an ulcerated plaque and 4 had mobile plaque). Patient age and topography of the ischemic event did not correlate with TEE-defined cardioembolic sources.

Conclusions: TEE identified high frequencies of potential treatable cardioembolic sources in patients with ischemic events. TEE should be considered in all patients who do not have identified cause of TIA and/or stroke for early treatment and prevention of recurrent events.

Key Words: atrial thrombus, cardioembolic sources, patent foramen ovale, thoracic aortic plaque, transesophageal echocardiography, transient ischemic attack

**********

The detection of a potential source of thromboembolism after a cerebral ischemic event continues to be a challenge in the care of stroke patients, yet is of paramount importance for the prevention of subsequent episodes of ischemia. Each year, more than 750,000 Americans will have a first or recurrent stroke. (1) Approximately one third of all stroke survivors will have another stroke within 5 years. The recurrence rate of stroke is highest during the first month. Prompt identification of treatable cardiac sources of a transient ischemic attack (TIA) or minor stroke may prevent a disabling major stroke. In cases of acute stroke, uncovering the source of embolism is as important as the urgent treatment of stroke.

The advent of the omni-plane transesophageal echocardiography (TEE) ultrasound probe has allowed for improvement in the quality of acoustic images, and a higher rate of detection of cardioembolic sources in patients with acute stroke. (1,2) TEE is considered superior to transthoracic echocardiography in detecting potential sources of cardiac emboli such as patent foramen ovale (PFO), atrial septal aneurysm, left atrial echo contrast, and left atrial clot. (3-9) In a study conducted by Ozeren et al, (10) TEE revealed more specific findings in every stroke subtype (cardioembolic, atherothrombotic, and lacunar) than transthoracic echocardiography (74.0 versus 10.2%, P < 0.001). Recently, thoracic aortic atheroma, proven to be a potential source of embolism (11-16) and a poor prognosis, (2) has been identified by the use of TEE in numerous patients with ischemic stroke.

Despite a growing body of literature favoring TEE as the ideal test for evaluating cardiac sources of embolism, it is not clear which patient group should be selected for TEE for better identification of potential cardiac sources. Performing TEE to reveal potential cardiac sources is justified only if these sources are treatable, meaning that TEE data will provide objective information to the clinicians for patient treatment. The objective of our study was to determine the frequencies of treatable potential cardiac sources of embolism in patients with ischemic events who do not have recognized cause of TIA and/or stroke. Therefore, adequate treatment could be initiated for the prevention of recurrent episodes.

Materials and Methods

A prospective database was established and maintained for all consecutive patients at our university-affiliated medical center who were without identified cause of TIA and/or stroke, and who underwent TEE within 2 weeks of symptom onset for the evaluation of possible cardiac sources of embolism. Patients from the database who underwent TEE between July 2000 and August 2001 were analyzed for this study. Patients with ischemic events related to carotid stenosis, carotid dissection, or intracoronary occlusive disease did not receive TEE evaluation. Patients with an abnormal coagulation profile attributed to the ischemic event did not receive TEE evaluation either. The demographic and clinical characteristics of these patients, including cerebrovascular risk factors (previous stroke, coronary artery disease, hypertension, diabetes mellitus, hyperlipidemia, and cigarette smoking) were documented along with the topographic location of ischemic events. We defined treatable cardiac sources as PFO with right-to-left shunt, left atrial clot, left atrial appendage clot, and severe aortic atheroma with or without mobile plaque, because available medical treatments are found to be effective for their resolution and prevention of recurrent events. We classified cerebral ischemic events as cortical, lacunar, and vertebrobasilar, according to the National Institute of Neurologic Disorders and Stroke classification scheme. (17) This scheme is based on clinical symptoms and neuroimaging findings shown by magnetic resonance imaging (MRI) and/or computed tomography (CT). Patients who had electrocardiographic evidence of atrial fibrillation, or had undergone prosthetic valve placement, were excluded from the study.

Informed consent was obtained before TEE was performed. Patients were kept without food for at least 6 hours before the test. Both a local anesthetic agent (lidocaine spray) and conscious sedation (intravenous midazolam and meperidine) were used for the procedure. The patient's blood pressure, heart rate, respiratory rate, and oxygen saturation were kept within normal range during the procedure.

Complete TEE with basal, mid-esophageal, and transgastric views was performed with Agilent Sonos 5500 and Agilent Sonos 1000 imaging ultrasound systems, each with a 5-MHz omni-plane transducer (Agilent, Andover, MA). Two experienced echocardiographers who had knowledge of all aspects of potential sources of emboli performed and interpreted TEE. Diagnosis on TEE findings was established upon agreement between the two readers. The [kappa] statistic for TEE findings ranged from 0.4 to 0.5 between the two independent raters.

The interatrial septum, left atrium, and left atrial appendage were evaluated at angles ranging from 30 to 90 degrees for possible identification of left atrial appendage pathology, PFO, and atrial septal aneurysm. Using two-dimensional echocardiographic Doppler examination, movement of the interatrial septum was established for identification of atrial septal aneurysms. Color echocardiographic Doppler examination was used to document flow through the interatrial septum. Agitated saline (9 mL saline and 1 mL air) was administered intravenously for detection of a right-to-left intracardiac shunt through the interatrial septum. In the absence of spontaneous passage of bubbles through the interatrial septum with agitated saline injection, the Valsalva maneuver was also used to increase the sensitivity of detecting a right-to-left shunt.

We defined atrial septal aneurysm as movement of the interatrial septum more than 15 mm from its resting position. (18) Using two-dimensional echocardiographic Doppler, the echo dropout between septum primum and septum secundum of the interatrial septum was established for anatomic identification of PFO. (19) PFO was defined as either color flow across the interatrial septum documented by echocardiographic color Doppler examination and/or by observing bubbles crossing through the interatrial septum during agitated saline injection. PFO with right-to-left shunt was defined as the passage of bubbles from the right atrium through the interatrial septum to the left atrium, occurring either spontaneously or after use of the Valsalva maneuver. The presence of an echo density in the left atrial cavity consistent with stagnant flow was defined as left atrial smoke.

Left ventricular hypertrophy was defined as mild (ventricular wall thickness, >1.5 to <2.0 cm), moderate (wall thickness, 2.0 to 2.5 cm), and severe (wall thickness, >2.5 cm). (19) Global systolic dysfunction of the left ventricle was classified as mild (ejection fraction 40 to 49%), moderate (30 to 39%), and severe (<30%).

All segments of the thoracic aorta, including the descending, arch, and ascending segments, were evaluated for thoracic aortic atheroma by using the transverse and longitudinal axes of the omni-plane probe (0 and 90 degrees). When an atheroma was clearly visualized, a still-frame, two-dimensional image was captured for measurements. Thoracic aortic atheromas were classified according to thickness, extent of involvement of the aorta, and morphologic characteristics of the plaque. Plaque thickness up to 2 mm was considered mild; thickness from 2 to 4 mm was considered moderate; and more than 4 mm was considered severe. Plaque involving an entire segment was classified as generalized, whereas plaque localized to a segment of the aorta was focal.

Statistical Analysis

A Student t test of independent samples was performed to compare continuous variables, and either a Pearson [chi square] test or a Fisher exact test was used to compare categorical data. Logistic regression was performed to estimate the association of various risk factors with the presence of potential cardioembolic sources. For multivariable logistic analysis, the variables were selected by using forward stepwise regression. [alpha] levels of P < 0.05 were considered significant. All analyses were carried out with the use of STATA statistical software (version 7.0; Intercooled, College Station, TX). The [kappa] statistics for TEE findings were also evaluated between the two independent raters.

Results

Between July 2000 and August 2001, 237 consecutive patients with cerebral ischemic events underwent TEE evaluation for detection of possible cardiac sources of emboli. The mean age of these patients was 59 [+ or -] 14 years (range, 21 to 93 years); 119 (51%) were men. Hypertension was present in 53% (125 of 237) of the patients, hyperlipidemia in 45% (107 of 237), coronary artery disease in 16% (39 of 237), diabetes mellitus in 21% (50 of 237), and active or previous smoking in 36% (85 of 237). A history of stroke was present in 15% of patients (36 of 237). Ischemic events were cortical in 105 (44%), subcortical in 35 (15%), vertebrobasilar distribution in 32 (14%), or in multiple locations in 65 patients (27%) (Table 1).

Treatable potential cardiac sources of emboli were identified in 146 (61%) patients, and these are PFO with right-to-left shunt, left atrial clot, left atrial appendage clot, and severe thoracic aortic atherosclerotic plaque disease (Table 2). PFO with right-to-left shunt with agitated saline injection was detected in 59 (25%) patients. Left atrial clot was present in 6 (3%) patients and left atrial appendage clot was present in 8 (3%) patients. Ascending aortic plaques were present in 31 (13%) cases, arch and descending aortic plaques were present in 107 (45%) and 128 (54%) cases, respectively. Severe thoracic aortic atherosclerotic plaque disease was present in 79 (33%) patients; ulceration of the plaque was present in 56 of these patients, and a mobile segment in the plaque was present in 4.

In univariate analysis, neither patient age, sex, nor specific topographic location of the ischemic event was associated with TEE identification of a treatable potential cardioembolic source.

Discussion

A high rate of potential treatable cardioembolic sources was detected in our series, which is comparable with previously published series that have included thoracic aortic atheroma as a potential source of emboli. (10,12,16) Other studies report lower frequencies of cardioembolism by TEE when thoracic aortic atheroma is not included as a potential source (45%). (3,7) In our study, 33% of the patients had a significant thoracic aortic atheroma; excluding this lowers the frequency of detection of treatable potential cardioembolic sources to 28%, which is comparable to the study of Albers et al. (3) In that study, TEE revealed at least one potential source of cardioembolism in 46% of patients with clinical risk factors for cardioembolism or unexplained stroke. Thoracic aortic atheroma is only detectable by TEE. Therefore, performing a TEE during the acute period of a cerebral ischemic event not only identifies a treatable potential source of embolism from the heart, but also detects another potential treatable source of embolism from the ascending thoracic aorta. (20,21)

In the study conducted by Ozeren et al, (10) TEE revealed more specific findings than transthoracic echocardiography for all stroke subtypes (74.0% versus 10.2%, P < 0.001). Similar results were reported by Paemelaere et al. (22) Using TEE, they identified a possible embolic source in 65.9% of patients with cardiovascular risk factors, as compared with 29.7% in patients without cardiovascular risk factors. Our patient group also had a high frequency of cerebrovascular risk factors, including hypertension, hyperlipidemia, coronary artery disease, diabetes mellitus, and smoking, which may positively correlate with the high frequencies of potential sources of cardiac emboli that we observed.

Despite the demonstrated superiority of TEE in the detection of potential cardiac sources of embolism, controversy remains as to the selection of stroke patients for TEE evaluation with respect to age and location of stroke. Some authors postulate that younger patients (<55 years) with cerebral ischemic events are more likely to have patent foramen ovale and atrial septal aneurysm (23,24); others theorized that older patients ([greater than or equal to]60 years) with ischemic events are likely to have thoracic aortic atheromas. (25) Lacunar stroke has long been assumed to be noncardioembolic, thus evaluation with TEE has been presumed unnecessary until recently, when the yield of TEE was found to be substantial in all stroke subtypes. (3) The yield of TEE in our study involving men and women 21 to 93 years of age was also high across all ischemic subtypes.

Identification of potential sources of emboli in patients with cerebral ischemic events is important for the prevention of recurrent cerebral and systemic embolic events by means of appropriate medical treatment. (20,21,26-31) Anticoagulation is considered to be the main treatment of left atrial and left atrial appendage thrombus or clot. (26-28) In addition to medical therapy, percutaneous transcatheter closure of PFO is considered an alternative treatment option for PFO with right-to-left shunt. (29-31) In a recent study, (20) statin was found to have a protective effect against recurrent cerebral embolic events (P = 0.0001) in patients with severe thoracic aortic atheroma (TAA) (Table 3 and Table 4). In that study, absolute risk reduction was 17%, relative risk reduction was 59%, and the number needed to treat was 6. The mechanism by which statin was protective against recurrent embolic events was postulated to be plaque stabilization and plaque reduction. (32) Statin is not only effective in regression of plaques in symptomatic patients, but is also effective in the regression of plaques in asymptomatic patients. (33) In another study, (21) warfarin-treated mobile patients with TAA had fewer embolic events than patients treated with aspirin (5% versus 45%). In the same study, stroke occurred in 27% of patients, and none of these patients were treated with warfarin. In a recent article, TAA treatment options for the prevention of systemic and cerebral embolism have been described. (20) Combination therapy with statin and antiplatelets for the treatment of TAA to prevent recurrent embolic events is undergoing evaluation in various studies.

The major limitation of our study is that we did not have a control group. Other limitations are that we excluded patients with ischemic events related to carotid stenosis, carotid dissection, intracoronary occlusive diseases, and coagulation abnormalities. Inclusion of the above excluded patients might have affected our frequencies of TAA, their characteristics, and its association with cardiovascular risk factors. Further study is required to answer all our unsolved questions.

Conclusion

Treatable potential cardiac sources of emboli could be identified in 61% of patients who do not have any identified cause of TIA and/or stroke. These sources were patent foramen ovale with right-to-left shunt (25%), left atrial clot (3%), left atrial appendage clot (3%), severe thoracic aortic atheroma (33%), and mobile aortic plaque (2%). Most of these embolic sources are medically treatable, and subsequent ischemic events could therefore be prevented. We conclude that TEE should be considered for all patients who do not have recognizable causes of TIA and/or stroke, regardless of age and the topographic location of ischemic events.

Table 1. Demographics and characteristics of patients who underwent
transesophageal echocardiography

Variables                            Values*

Number of patients                   237
Mean age (yr)                         59 [+ or -] 14
Male                                 119 (50%)
Previous stroke                       36 (15%)
Coronary artery disease               39 (16%)
Hypertension                         125 (53%)
Hyperlipidemia                       107 (45%)
Diabetes mellitus                     50 (21%)
Active or former smoker               85 (36%)
Cortical ischemic events             105 (44%)
Subcortical ischemic events           35 (15%)
Vertebrobasilar artery distribution   32 (14%)
Multiple locations                    65 (27%)

*Frequencies or mean [+ or -] SD.

Table 2. Types of potential treatable cardioembolic sources detected by
transesophageal echocardiography

                                               Values
Variables                                      (frequency)

At least one cardiac source of embolism        146 (61%)
Left atrial clot                                 6 (3%)
Left atrial appendage clot                       8 (3%)
Patent foramen ovale with right-to-left shunt   59 (25%)
Severe aortic atheroma                          79 (33%)
Ulcerated aortic plaque                         56 (24%)
Mobile aortic plaque                             4 (2%)

Table 3. Classifications of thoracic aortic atheroma according to
severity or thickness of plaques

Variables        Values

Mild plaque      13/237 (5%)
Moderate plaque  49/237 (19%)
Severe plaque    79/237 (33%)

Table 4. Classifications of thoracic aortic atheroma according to
locations and distributions

Variables                  Values

Ascending aortic plaques    31/237 (13%)
Aortic arch plaques        107/237 (45%)
Descending aortic plaques  128/237 (54%)
Generalized plaques        109/237 (45%)
Focal plaques               32/237 (13%)

Accepted February 17, 2004.

References

1. Broderick J, Brott T, Kothari R, et al. The Greater Cincinnati/Northern Kentucky Stroke Study: preliminary first-ever and total incidence rates of stroke among blacks. Stroke 1998;29:415-421.

2. Kissela B, Broderick J, Woo D, et al. Greater Cincinnati/Northern Kentucky Stroke Study: volume of first-ever ischemic stroke among blacks in a population-based study. Stroke 2001;32:1285-1290.

3. Albers GW, Comess KA, DeRook FA, et al. Transesophageal echocardiographic findings in stroke subtypes. Stroke 1994;25:23-28.

4. Archer SL, James KE, Kvernen LR, et al. Role of transesophageal echocardiography in the detection of left atrial thrombus in patients with chronic nonrheumatic atrial fibrillation. Am Heart J 1995;130:287-295.

5. Asinger RW, Herzog CA, Dick CD. Echocardiography in the evaluation of cardiac sources of emboli: the role of transthoracic echocardiography. Echocardiography 1993;10:373-396.

6. Bussiere JP, Bonnet D, Renard JL, et al. Contribution of transesophageal echocardiography in the investigation of the atrium in systemic embolism. Ann Med Interne 1992;143:5-10.

7. Pearson AC. Transthoracic echocardiography versus transesophageal echocardiography in detecting cardiac sources of embolism. Echocardiography 1993;10:397-403.

8. Pollick C, Taylor D. Assessment of left atrial appendage function by transesophageal echocardiography: implications for the development of thrombus. Circulation 1991;84:223-231.

9. Vigna C, Russo A, De Rito V, et al. Frequency of left atrial thrombi by transesophageal echocardiography in idiopathic and in ischemic dilated cardiomyopathy. Am J Cardiol 1992;70:1500-1501.

10. Ozeren A, Acarturk E, Sarica Y, et al. Transesophageal echocardiography in various ischemic stroke subtypes. Jpn Heart J 1997;38:199-206.

11. Di Tullio MR, Sacco RL, Savoia MT, et al. Aortic atheroma morphology and the risk of ischemic stroke in a multiethnic population. Am Heart J 2000;139:329-336.

12. Mahagney A, Sharif D, Weller B, et al. Diagnosis of cerebral embolism by transesophageal echocardiography. Harefuah 1998;134:256-259.

13. Rundek T, Di Tullio MR, Sciacca RR, et al. Association between large aortic arch atheromas and high-intensity transient signals in elderly stroke patients. Stroke 1999;30:2683-2686.

14. Sen S, Wu K, McNamara R, et al. Distribution, severity and risk factors for aortic atherosclerosis in cerebral ischemia. Cerebrovasc Dis 2000;10:102-109.

15. Stern A, Tunick PA, Culliford AT, et al. Protruding aortic arch atheromas: risk of stroke during heart surgery with and without aortic arch endarterectomy. Am Heart J 1999;138:746-752.

16. Ferrari E, Vidal R, Chevallier T, et al. Atherosclerosis of the thoracic aorta and aortic debris as a marker of poor prognosis: benefit of oral anticoagulants. J Am Coll Cardiol 1999;33:1317-1322.

17. Special report from the National Institute of Neurological Disorders and Stroke: classification of cerebrovascular diseases III. Stroke 1990;21:637-676.

18. Hanley PC, Tajik AJ, Hynes JK, et al. Diagnosis and classification of atrial septal aneurysm by two-dimensional echocardiography: report of 80 consecutive cases. J Am Coll Cardiol 1985;6:1370-1382.

19. Marshall AC, Lock JE. Structural and compliant anatomy of the patent foramen ovale in patients undergoing transcatheter closure. Am Heart J 2000;140:303-307.

20. Tunick PA, Kronzon I. Embolism from the aorta: atheroemboli and thromboemboli. Curr Treat Options Cardiovasc Med 2001;3:181-186.

21. Dressler FA, Craig WR, Castello R, et al. Mobile aortic atheroma and systemic emboli: efficacy of anticoagulation and influence of plaque morphology on recurrent stroke. J Am Coll Cardiol 1998;31:134-138.

22. Paemelaere JM, Sirinelli A, Dreyfus X, et al. Transesophageal echography and systemic ischemic incidences: 235 cases. Rev Neurol 1996;152:27-31.

23. Chant H, McCollum C. Stroke in young adults: the role of paradoxical embolism. Thromb Haemost 2001;85:22-29.

24. Overell JR, Bone I, Lees KR. Interatrial septal abnormalities and stroke: a meta-analysis of case-control studies. Neurology 2000;55:1172-1179.

25. Autore C, Cartoni D, Piccininno M: Multiplane transesophageal echocardiography and stroke. Am J Cardiol 1998;81:79G-81G.

26. Jaber WA, Prior DL, Thamilarasan M, et al. Efficacy of anticoagulation in resolving left atrial and atrial appendage thrombi: a transesophageal echocardiographic study. Am Heart J 2000;140:150-156.

27. Sun SS, Hsieh JF, Tsai SC, et al. Monitoring the effect of anticoagulations on left atrial thrombi in patients with rheumatic heart disease: assessment with 111 In-oxide-labelled platelet heart scintigraphy and transesophageal echocardiography. Nucl Med Commun 2000;21:627-30.

28. Labeque JN, Laffort P, Lafiitte S, et al. Treatment of left atrial thrombosis by low-molecular-weight heparin: a preliminary study of 6 cases. Arch Mal Coeur Vaiss 2000;93:1528-1533.

29. Koenig P, Cao QL, Heitschmidt M, et al. Role of intracardiac echocardiographic guidance in transcatheter closure of atrial septal defects and patient foramen ovale using the Amplatzer device. J Interv Cardiol 2003;16:51-62.

30. Onorato E, Melzi G, Casilli R, et al. Patent foramen ovale with paradoxical embolism: mid-term results of transcatheter closure in 256 patients. J Interv Cardiol 2003;16:43-50.

31. Bialkowski J, Kusa J, Szkttnik M, et al. Percutaneous catheter closure of atrial septal defect: short-term and mid-term results. Rev Esp Cardiol 2003;56:383-388.

32. Herrara CJ, Frazin LJ, Dau PC, et al. Atherosclerotic plaque evolution in the descending thoracic aorta in familial hypercholesterolemic patients: a transesophageal echo study. Arterioscler Thromb 1994;14:1723-1729.

33. Corti R, Fuster V, Fayad ZA, et al. Lipid lowering by simvastatin induces regression of human atherosclerotic lesions: two years' follow-up by high-resolution noninvasive magnetic resonance imaging. Circulation 2002;106:2884-2887.

RELATED ARTICLE: Key Points

* Detection of potential sources of thromboembolism following a cerebral ischemic event continues to be a challenge in the care of stroke patients, yet is of paramount importance for the prevention of subsequent episodes.

* TEE identifies high frequencies of potential treatable cardioembolic sources in patients with ischemic events.

* TEE should be considered in all patients without an identified cause of TIA and/or stroke for early treatment and prevention of recurrent events.

Abutaher M. Yahia, MD, Aasma B. Shaukat MD, MPH, Jawad F. Kirmani, MD, Andrew Xavier, MD, Nancy G. Manalio, RN, and Adnan I. Qureshi, MD

From the Departments of Neurology and Neurosciences and Medicine, University of Medicine and Dentistry, New Jersey Medical School, Newark, NJ.

Reprint requests to Dr Abutaher M. Yahia, Michigan State University, Department of Neurology, 138 Service Road, A217, East Lansing, MI 48824. Email: [e-mail]