Transesophageal echocardiography in patients with cryptogenic ischemic stroke: a systematic review.

Background: The clinical utility of routine transesophageal echocardiography (TEE) for patients with unexplained ischemic stroke is controversial. We performed a systematic review to determine the frequency of detection of new cardiac findings in patients with cryptogenic ischemic stroke (IS) undergoing transesophageal echocardiography (TEE).

Methods: Systematic review and meta-analysis of cohort studies of consecutive patients with "cryptogenic" IS undergoing TEE after routine etiologic workup. Patients were categorized into 2 groups: A (< 55 years) and B (≥ 55 years). Outcomes included proportion of patients with new TEE-detected cardiac findings and proportion of patients commenced on oral anticoagulation after TEE.

Results: Twenty-seven studies were included (n = 5,653). We identified significant heterogeneity among studies and report a range of prevalence rates and I2 statistic as our primary analysis. Prevalence of individual cardiac findings on TEE varied significantly among studies; patent foramen ovale (A: 12.0%-57.8%, I2 = 89.9%; B: 3.9%-43.5%, I2 = 86.7%), atrial septal aneurysm (A: 0-48.9%, I2 = 91.9%; B: 3.5%-25.0%, I2 = 84.5%), left atrial thrombus (A: 0-10.9%, I2 = 61.1%; B: 0-21.2%, I2 = 91.7%), spontaneous echo contrast (A: 0-11.9%, I2 = 57.2%; B: 0-21.3%, I2 = 89.8%), and aortic atheroma (A: 0-9.6%, I2 = 53.8%; B: 2.8%-44.4%, I2 = 89.7%). Definitions of common findings were not provided for many studies. Five studies (n = 591) reported on the proportion of patients who were commenced on anticoagulant therapy after TEE (range 0-30.7%).

Conclusions: Routine TEE in patients with cryptogenic IS identifies cardiac findings in a large proportion. However, there is marked interstudy variation in the definition and prevalence of common findings. Transesophageal echocardiography-detected findings prompted the introduction of anticoagulant therapy in up to one-third of patients. However, these were mostly not for established guideline-based indications based on randomized controlled trial evidence. It is unclear if routine use of TEE in patients with cryptogenic IS is indicated.