Background and Purpose--Findings on transesophageal echocardiography (TEE) after ischemic stroke predict recurrent embolic events and prompt therapy; however, the additive predictive power of TEE findings on long-term mortality is unknown. Our goal was to study the impact of TEE findings on all cause mortality in ischemic stroke patients referred for TEE.

Methods--We reviewed 245 consecutive patients who underwent TEE for ischemic stroke of undetermined origin (2000 to 2003). Long-term survival was assessed using the Social Security Death Index.

Results--In a mean follow-up period of 3.0 (1.4 to 4.8) years, death occurred in 19.2% of patients. TEE findings included patent foramen ovale (18.8%), left atrium/left ventricle thrombus (2.4%), spontaneous echo contrast (3.7%), atrial septal aneurysm (3.3%), valve vegetation/mass/tumor (7.8%), complex aortic atheroma ([CAA]; 14.7%), and the composite of any cardiac source of embolus (39.2%). A total atherosclerotic burden (TAB) score was also recorded. On Cox hazard regression analysis, measures of aortic atherosclerosis (CAA [hazard ratio (HR), 2.7; 95% CI, 1.4 to 5.3] or TAB score [HR, 1.4; 95% CI, 1.2 to 1.6]) were independent predictors of death, whereas other TEE findings were not.

Conclusion--In patients with ischemic stroke of undetermined origin referred for TEE, measures of aortic atherosclerosis, including CAA, represent the only TEE findings that predict long-term mortality after all other clinical factors are considered. Further study is needed to determine whether treatments for CAA effect long-term survival in patients with ischemic stroke.