Embolic strokes of undetermined source (ESUS) represent a subset of cryptogenic strokes that are suspected to have an occult embolic source. The risk of stroke recurrence in patients with ESUS varies between 1.9%/year and 19.0%/year depending on the prevalence of vascular risk factors. Part of the elevated recurrence rate is due to the inability to identify high-risk treatable causes such as cardiac thrombi as those found in the left atrial appendage (LAA), left atrium (LA), left ventricle (LV), valves, or aortic arch. The most frequently used diagnostic method in clinical practice to detect cardioaortic thrombi is transesophageal echocardiography (TEE). However, the relatively low availability, higher cost, and invasive nature of TEE limit its large-scale usability. In most stroke centers, patients presenting with an acute ischemic stroke or TIA undergo a tomography (CT) angiography (CTA) of the neck and intracranial vessels. This standard of care CTA (sCTA) classically includes the aortic arch, the higher portion of the ascending/descending aorta, and the rostral portion of the cardiac chambers, but does not involve the LAA, LV, or cardiac valves. A recent study performed among 300 patients with an acute ischemic stroke showed an overall LAA thrombus detection of 6.6% and 15% in patients with AF by extending the CTA 6 cm below the carina. This is an extraordinarily high prevalence of LAA compared to 0.5% to 4.8% of intracardiac thrombi identified on TEE in most previous studies. The major limitation of previous CTA and TEE studies is their observational design, so the differing prevalence of LAA thrombi could be explained by dissimilar population characteristics or selection bias. Based on the methodological limitation of prior studies and the promising role of extended CTAs (eCTA), a randomized controlled trial comparing eCTA + standard of care stroke workup vs. sCTA + standard of care stroke workup is needed.

Between 16% and 25% of patients with ischemic strokes do not have an identifiable cause after a full stroke workup, and their strokes are classified as cryptogenic. Embolic strokes of undetermined source (ESUS) represent a subset of cryptogenic strokes that are suspected to have an occult embolic source. The risk of stroke recurrence in patients with ESUS varies between 1.9%/year and 19.0%/year depending on the prevalence of vascular risk factors. Part of the elevated recurrence rate is due to the inability to identify high-risk treatable causes such as cardioaortic thrombi as those found in the left atrial appendage (LAA), left atrium (LA), left ventricle (LV), valves, or aortic arch.

The most frequently used diagnostic method in clinical practice to detect cardioaortic thrombi is transthoracic echocardiography (TTE). However, the sensitivity of TTE for detecting LV thrombi is as low as 35% without contrast and 64% with proper ultrasound contrast. Furthermore, the sensitivity for LAA thrombi is zero, given that the LAA cannot be visualized on a TTE. Transesophageal echocardiography (TEE) is a better alternative in terms of sensitivity for LV thrombus detection. However, the relatively low availability, higher cost, and invasive nature of TEE limit its large-scale usability.

In most stroke centers, patients presenting with an acute ischemic stroke or TIA undergo a computer tomography (CT) angiography (CTA) of the neck and intracranial vessels. This standard of care CTA (sCTA) classically includes the aortic arch and the higher portion of the ascending/descending aorta, but does not involve the LAA, LV, or cardiac valves. A recent study performed among 300 patients with an acute ischemic stroke showed an overall LAA thrombus detection of 6.6% and 15% in patients with atrial fibrillation (AF) by extending the CTA 6 cm below the carina. This is an extraordinarily high prevalence of LAA compared to 0.5% to 4.8% of intracardiac thrombi identified on TEE in most previous studies. The major limitation of previous CTA and TEE studies is their observational design, so the differing prevalence of LAA thrombi could be explained by dissimilar population characteristics or selection bias. Based on the methodological limitation of prior studies and the promising role of extended CTAs (eCTA), a randomized controlled trial comparing eCTA + standard of care stroke workup vs. sCTA + standard of care stroke workup is needed.

Demonstrating that performing an eCTA can significantly increase the detection of LAA thrombi compared to sCTA has three important implications with the potential to improve clinical practice, patients' outcomes, and clinical guidelines. Furthermore, an eCTA could become part of the standard of care workup of patients presenting to the Emergency Department with a hyperacute ischemic stroke. Proving that eCTA increases the detection of cardio-aortic thrombi in stroke patients in a randomized controlled trial (highest level of evidence) may result in the recommendation of this approach in future guidelines. The increased detection of LAA thrombi may increase the use of oral anticoagulants, which have proven efficacy for the prevention of recurrent strokes in patients with cardiac thrombi, ultimately resulting in fewer stroke recurrences.10 However, proving the latter concept would require a larger randomized clinical trial with stroke recurrence as the primary efficacy endpoint. The LAA is the most frequent source of thromboembolism in patients with AF11 and LAA thrombi are associated with increased detection of AF on cardiac rhythm monitoring, implying that finding an LAA thrombus may help improve the selection of patients who could benefit from prolonged cardiac monitoring after stroke.

Inclusion Criteria:

• All adult patients with a suspected cerebrovascular event who are evaluated in the Emergency Department or the Urgent Stroke Prevention Clinic at University Hospital, London Health Sciences Centre, London, Ontario, Canada.
• A confirmed diagnosis of stroke or TIA is not mandatory for patient enrollment and randomization. However, patients without a cerebrovascular event will be further excluded from the analysis of efficacy and safety outcomes. Only patients with an adjudicated ischemic stroke or TIA will be included in the latter analyses. The same patient may be included more than once in the study in case of presenting to the hospital with a suspected stroke on different dates. These patients will be excluded only if they had an outcome event (cardioaortic thrombus) previously adjudicated in the study. The reason for including patients more than once is that the cause of the stroke or the stroke mechanism can change with time (e.g. patient with post-stroke MI developing a new LV thrombus, a patient who had a stroke due to AF who is now on anticoagulants and presents with a new contralateral stroke due to severe carotid artery stenosis).

Exclusion Criteria:

• Allergy to iodinated contrast agents
• Pregnancy
• Lack of a peripheral vein access for intravenous contrast administration
• Any contraindication for the clinical use of a CTA for hyperacute stroke care (e.g., end-stage renal disease that contraindicates a CTA), and active or past cancer of the head, neck, or chest)
• Patients with known or newly diagnosed AF will not be excluded.