Supplemental Material (Online-Only)

Supplemental material (online-only)

Patients and diagnostics

Routine diagnostics in all patients comprised brain CT and/or MRI, extra-/intracranial angiography (MRA or CTA) or conventional angiography if indicated, ultrasound of extra-/intracranial arteries, transthoracic echocardiography (TTE), 12 lead- and Holter-ECG, routine laboratory tests and additionally factor V Leiden mutation, decreased antithrombin III, protein C and S, and increased antiphospholipid antibody titer in patients <60 years of age. TEE examination was offered to all patients regardless if etiology of stroke was determined or undetermined after routine diagnostics. Patients were excluded if they were in a bad clinical condition or not able to consent (n=92), refused study participation (n=92), or if TEE examination was not possible due to contraindications (n=23). In 10 patients TEE examination was not completed due to patient non-compliance or difficulties when inserting the probe, and 13 patients were excluded retrospectively because the diagnosis of ischemic stroke was uncertain at discharge.Severity of stroke on admission was assessed using the National Institutes of Health Stroke Scale (NIHSS) 1. Cerebral MRI and/or CT images were analyzed regarding the localization of the ischemic lesion by an experienced neuroradiologist who was blinded to patient data. Stroke etiology was classified after completed workup (including TEE) using the modified TOAST criteria 2.

Diagnostic workup in stroke patients

Median stroke severity according to the NIHSS was 2 (interquartile range=0-4). Routine brain MRI including diffusion weighted imaging (DWI) was performed in 342 (70.5%) and routine brain CT in 304 (62.7%) patients on admission. Visible acute ischemic lesions were found in the area of the posterior circulation in 76 (15.7%) patients, of the left hemisphere in 60 (12.4%), and the right hemisphere in 77 (15.9%). Infarcts in multiple territories were found in 26 (5.4%) patients. 166 (34.2%) lesions were classified as embolic. 246 (50.7%) patients had no detectable acute lesion on brain imaging. 56 (11.5%) patients had atherosclerotic stenosis ≥50% in an intra-/ or extracranial artery supplying the ischemic field diagnosed by ultrasound examinations using NASCET criteria 3. 59 (12.1%) patients were diagnosed with atrial fibrillation by 12-lead- and/or Holter-ECG. Complex plaques were detected by TEE in the ascending aorta and/or proximal aortic arch in 90 (18.4%) subjects. Stroke etiology was: (1) large-artery atherosclerosis in 127 (26.2%) patients, (2) cardioembolism in 113 (23.3%) patients, (3) small-vessel disease in 19 (3.9%) patients, (4) stroke of other determined etiology in 12 (2.5%) patients, and (5) undetermined in (5a) 40 (8.2%) patients with multiple probable sources, (5b) 165 (34.0%) patients with cryptogenic etiology, and (5c) 9 (1.9%) patients with insufficient stroke workup.

Transesophageal echocardiography protocol

Blood pressure was measured at the right arm and heart rate was documented every 5 minutes during the examination. After assessment of the heart and cardiac valves, the thoracic aorta was examined. The proximal aorta (ascending aorta and proximal aortic arch) was defined as the area between the aortic valve and the outlet of the left subclavian artery (LSA). We defined the proximal DAo as the area between the outlet of the LSA and the transition zone between the arch curvature and the straight part of the DAo. The distal DAo was defined as the straight descending part of the thoracic aorta. First, intima-media-thickness was assessed during systole on a freeze frame in the ascending aorta, proximal arch, proximal DAo, and distal DAo. Then, complex plaques (defined as ≥4-mm thick, ulcerated (≥2-mm deep and wide), or containing mobile thrombi) of the thoracic aorta were detected. Subsequently, 2D pulsed-wave Doppler ultrasound was used to measure antegrade and retrograde blood flow velocities in longitudinal section at the proximal DAo using a 20-mm sample-volume at 90-140° Doppler angle. Color-guided pulsed wave spectral Doppler velocities were obtained as parallel to the flow as possible by rotation and angulation of the TEE probe (~30° from the transducer and ~60° from axial flow in the descending aorta). Velocity-time-integrals (i.e. the area under the velocity-time curve, see Fig.2) in antegrade (VTIantegrade) and retrograde (VTIretrograde) directions and the ratio between both (VTIratio=VTIretrograde/VTIantegrade) were calculated offline by a blinded reader in order to estimate retrograde blood flow. In case of difficulties in velocity time curve delineation (e.g., in the presence of arrhythmia), another two experienced echocardiographers were involved for selection of the appropriate velocity time curve in consensus reading. Furthermore, the aortic distensibility coefficient was calculated from systolic and diastolic short-axis images of the descending thoracic aorta as 2×[(Dmax/Dmin)/PP], where Dmax is the maximum-minimum diameter, Dmin is the minimum diameter and PP is calculated from systolic minus diastolic blood pressure. The trailing-edge to leading-edge technique was used to measure the aortic luminal area by circular regions of interest placed between the calipers to outline the aortic lumen. Aortic strain was assessed as (χSA–χDA)/χDA, where χSA is the 3-heart-beat average of the aortic luminal area in systole and χDA the corresponding average area in diastole.

Prevalence of complex atheroma in the proximal descending aorta

Complex plaques were found in the proximal DAo in 79 (16.0%) patients of which 20 (25.3%) had stroke due to large-artery athereosclerosis, 16 (20.3%) had cardioembolic stroke, 5 (6.3%) stroke due to small-vessel disease, 26 (32.9%) had multiple etiologies, and 12 patients (15.2%) cryptogenic stroke. 26 of 67 (38.8%) patients with determined stroke etiology (one or multiple) and complex Dao plaques had visible acute stroke with embolic patterns compared to 3 of 12 (25.0%) patients with cryptogenic stroke and complex plaques (p=0.52). We did not detect a difference in the prevalence of posterior and left-hemispheric stroke between non-cardioembolic stroke patients with DAo plaques 12/51 (23.5%) and cardioembolic stroke patients without DAo plaques 18/102 (17.7%) (p=0.84).

References

1. Lyden P, Lu M, Jackson C, Marler J, Kothari R, Brott T, et al. Underlying structure of the National Institutes of Health Stroke Scale: results of a factor analysis. NINDS tPA Stroke Trial Investigators. Stroke. 1999;30(11):2347–54.

2. Adams HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke. 1993;24(1):35–41.

3. Ferguson GG, Eliasziw M, Barr HW, Clagett GP, Barnes RW, Wallace MC, et al. The North American Symptomatic Carotid Endarterectomy Trial : surgical results in 1415 patients. Vol. 30, Stroke. 1999;30(9):1751-8.