Reliability of aortic MRI at 3 Tesla in patients with acute cryptogenic stroke

Objectives: To evaluate a new three dimensional (3D) MRI protocol for the reliable detection of aortic high risk plaques compared with transoesophageal echocardiogra-phy (TOE) and to test the reliability of additional MRI in stroke of undetermined aetiology. Methods: 74 acute stroke patients were examined by both TOE and MRI at 3 Tesla with special regard to aortic high risk plaques (ie, ≥4 mm, superimposed thrombi). ECG synchronised pre- and post-contrast T1 weighted 3D imaging (spatial resolution ∼1 mm³) covering the thoracic aorta was employed. In plaques ≥3 mm, additional two dimensional T2 imaging and time resolved (CINE) imaging sequences were performed. Aetiology of brain ischaemia was classified according to modified TOAST (Trial of Org 10172 in Acute Stroke Treatment) criteria. Aortic high risk embolic sources detected by MRI in patients with cryptogenic stroke were evaluated. Results: Differences in maximum aortic wall thickness for TOE and MRI were not statistically significant for all aortic segments. The overall number of high risk plaques detected by MRI (n = 74) was substantially higher compared with TOE (n = 47). Most noticeably, MRI identified aortic high risk pathologies in 8/26 (30.8%) patients with cryptogenic stroke after standard diagnostics, including TOE (n = 2: dissection or thrombus; n = 6: plaques ≥4 mm). Conclusions: Our results demonstrate the feasibility of this 3D MRI protocol for the reliable detection of aortic high risk plaques in acute stroke patients. Because of improved visualisation of the aortic arch and the detection of additional embolic sources not seen by standard diagnostics, this novel technique may become a valuable tool for future patients with cryptogenic stroke.