Abstract Number: Esoc2026a2094 Defining Clot Homogeneity in Acute Ischemic Stroke (AIs): A Data-Driven Threshold

Abstract Background and aims Clot composition in AIS may influence thrombectomy outcome, but there is no clear method to classify clots by distribution of components and heterogeneity of composition. A clearer definition is needed to better understand how clot homogeneity and heterogeneity relate to clinical characteristics. Methods 269 AIS clots from the RESTORE registry were stained with Martius Scarlet Blue. Two histopathologists independently classified the clots as Homogeneous Red Blood Cell (RBC) -rich, Homogeneous Fibrin/Platelet-rich, or Mixed based on the appearance and distribution of dominant components. By optimising accuracy, sensitivity and specificity of the percentage composition, we identified a data-driven cut-off to separate the three homogeneity groups. This threshold was then applied to the RESTORE stroke clot registry of 3, 576 clots to study associations with stroke etiology, clot size, occlusion pattern, and NIHSS score on admission. Results Using a threshold of 75%, stroke etiology differed significantly across clot homogeneity groups (χ2 (3, 1. 071) =31. 96, p0. 0001), with RBC-rich clots more frequently associated with large artery atherosclerosis and fibrin/platelet-rich clots with cardioembolism. Mixed clots were associated with larger clot size (H (2) = 86. 20, p0. 0001) and a higher frequency of occlusions involving more than one vascular segment (p0. 001). No association was observed between clot homogeneity groups and vascular territory (anterior vs posterior, p = 0. 40), or NIHSS score on admission (H (2) = 1. 875, p = 0. 39). Conclusions An empirical cut-off may provide a simple and reliable classification of clot homogeneity with clinical relevance. Future work will explore further statistical approaches for robust threshold definition. Conflict of interest KD group is funded by Research Ireland 13/RC/2073₂ and Industry partner J&J Medtech