Supplementary Material for: A Novel Evaluation for Predicting Aortic Complicated Lesions Using Calcification on Chest X-ray

<p><b><i>Background:</i></b> The aorta is a significant source of cerebral thromboembolisms. Aortic complicated lesions (ACLs) are key findings on transesophageal echocardiography (TEE) for assessing aortic sources of emboli to the brain. TEE is sometimes avoided due to its invasiveness. However, few reports have examined alternative methods for predicting ACLs. We investigated relationships between aortic arch calcification (AAC) on chest X-ray and ACLs. <b><i>Methods:</i></b> Participants comprised 300 patients with acute ischemic stroke or transient ischemic attack who underwent TEE for the evaluation of the aortic arch and heart. A postero-anterior plain chest X-ray in the recumbent position was evaluated on admission for each patient. AAC was evaluated using 4 grades (0-3) and “AAC thickness” defined as the distance from the inner margin of the most distant AAC to the outer margin of the aortic vessel wall. ACLs were defined by intima-media thickness (IMT) ≥4.0 mm or presence of ulcerated or mobile plaques. Carotid maximum IMT on ultrasonography was also evaluated. Comparison of the diagnostic ability to predict ACL was performed between AAC grades and AAC thickness or AAC thickness and carotid maximum IMT using the Delong method. <b><i>Results:</i></b> ACLs were identified in 71 patients (23.7%), including ACLs with ulcerated plaques in 24 (8.0%) and ACLs with mobile plaques in 9 (3.0%). Plaque thickness was greater in higher AAC grades or higher quartiles of AAC thickness (<i>p</i> for trend <0.001 each). The Cochran-Armitage test showed that both higher AAC grade and higher quartile of AAC thickness were significantly associated with the presence of ACLs, as well as the presence of ulcerated or mobile plaques (<i>p</i> for trend < 0.001 each). Receiver-operating characteristic (ROC) analysis showed optimal cut-off values for AAC thickness of 5.6 mm for ACLs and 6.0 mm for ulcerated or mobile plaques. Multivariate logistic regression revealed a higher grade of AAC (grades 2-3) and AAC thickness (≥6 mm) as significantly associated with ACLs and ulcerated or mobile plaques (<i>p</i> < 0.001 each). ROC curve comparisons showed that AAC thickness offered a better marker of ACLs than AAC grade (<i>p</i> = 0.019), although no significant difference was evident between AAC thickness and carotid maximum IMT (<i>p</i> = 0.567). <b><i>Conclusions:</i></b> AAC on chest X-ray, evaluated by both AAC grade and AAC thickness to the outer aortic vessel wall, was significantly associated with ACLs on TEE. AAC thickness was suggested as more useful than AAC grade and equivalent to carotid IMT in predicting ACLs.</p>