Development of a predictive scale for cardioembolic stroke using extracted thrombi and angiographic findings

Abstract

We analyzed the histopathological findings of the clots obtained from patients with acute ischemic stroke by mechanical thrombectomy. We then developed a clinical scoring system for predicting pathogenic causes in patients with undetermined ischemic stroke using these histopathological and the angiographic findings during endovascular treatment. Only cases with the occlusion of the intracranial internal carotid artery or the proximal part of the middle cerebral artery were included in this study. Histopathologic findings of clots were compared and analyzed using the Trial of Org 10,172 in Acute Stroke Treatment (TOAST; large artery atherosclerosis, cardioembolic, and undetermined groups) and angiographic occlusion type (AOT; branching-site occlusion and truncal-type occlusion groups) classification systems. Fifty-two patients had enough clots extracted by mechanical thrombectomy for full histopathologic examination. There was no significant within-group difference in the fraction of components in the thrombi for either the TOAST or AOT system; however, the platelet distribution patterns were different. The large artery atherosclerotic group and truncal-type occlusion group had mostly peripheral patterns, whereas the cardioembolic group, undetermined group and branching-site occlusion group had mostly clustering patterns (p = 0.02 in TOAST classification; p = 0.007 in AOT classification). Patients with scores of 3 or 4 on our new scale had a sensitivity of 93.5% and a specificity of 100% for cardioembolic stroke. The BOCS2 scale, developed using a combination of the TOAST and AOT classification systems, may be helpful as an adjunctive diagnostic tool for identifying cases caused by cardiogenic embolism in patients with undetermined ischemic stroke.