P2469The yield of transoesophageal echocardiography in embolic stroke of unknown source

Abstract Purpose Transthoracic echocardiography (TTE) and transoesophageal (TOE) are the main diagnostic tools for the detection of potential cardiac and aortic sources of embolism (PCSE). TOE is superior to TTE. However, its therapeutic impact is questioned. TOE is no longer recommended in ESUS (embolic stroke of unknown source), which is thought to be caused by undiagnosed atrial fibrillation (AF). Recent studies however report an about 10% change in therapeutic management due to TOE. We aimed to assess prospectively, in patients admitted to our stoke unit and diagnosed ESUS, the rate of treatment change induced by TOE. Methods Patients diagnosed with acute (≤7 days) ESUS were included to undergo TOE. ON TTE, PCSE, left ventricular ejection fraction (LVEF, %), left atrial area (LA-A, cm2), indexed left atrial volume (LAV, ml/m2) were recorded. On TOE, PCSE, left atrial appendage area (LAA-A, cm2), left atrial appendage emptying velocity (LAA-EV, cm/s) were recorded. PCSE on TTE and TOE were classified according to EAE guidelines. Results Between October 2016 and May 2018, 1322 TTEs were performed in patients admitted to our newly labelled stroke unit. ESUS was diagnosed in 152 patients, who underwent TOE. Mean delay (95% CI) between stroke onset and TTE was 2.2 days (1.9–2.5). Mean (95% CI) age was 61 years (58–63). 56/152 were female (37%). On TTE, mean (95% CI) LVEF was 62% (61–64), mean LA-A was 19 cm2 (19–20) and mean LA-V was 32 ml/m2 (31–34) ml/m2. In 38 patients, a major PCSE was identified by TTE. In 4 cases, there were 2 potential major PCSE and in one case 3. In 79 patients, at least one minor PCSE was identified by TTE, among which 35 PFO. On TEE, mean (95% CI) LAA-A and LAA-EV were: 5.2 cm2 (4.9–5.5) % and 76 cm/s (72–80) cm/s. In 50 patients, a major PCSE was identified by TOE. Eleven patients had 2 major PCSE. In 115 patients, at least one minor PCSE was identified by TOE, among which 34 PFO. According to EAE guidelines, a treatment change induced by TOE occurred in 12 patients. Anticoagulation in 5 patients (aortic arch thrombus in 2, thrombus of the proximal descending aorta in 1, LAA thrombus in 1, and thrombotic non-infectious vegetations in 2). Surgery in one patient (LAA fibroelastoma), antibiotics in two patients with infectious endocarditis, One patient had aortic fibroelastoma not seen on TTE. Finally, two patients with mobile aortic arch debris had dual antiplatelet therapy (DAT). In two other patients, TOE showed a mitral fibroel