ABSTRACT NUMBER: ESOC2026A1184 WHITE MATTER HETEROGENEITY ASSESSED ON NON-CONTRAST CT IN STROKE PATIENTS UNDERGOING REPERFUSION TREATMENT

Abstract Background and aims Magnetic resonance imaging markers of white matter damage are associated with poor functional outcomes after ischemic stroke, but their use in the acute setting is limited because most patients undergo computed tomography (CT). We aimed to establish a novel CT-based marker, white matter heterogeneity (WHCT), and to evaluate its prognostic value after ischemic stroke. Methods We included 302 stroke patients with large vessel occlusion undergoing thrombectomy from the German Stroke Registry (mean age 76y, 56% women, median NIHSS 14, 53% received thrombolysis). White matter was automatically segmented on admission non-contrast CT using a convolutional neural network (SynthSeg). WHCT was computed as the average absolute Hounsfield unit difference between all voxel pairs within white matter contralateral to the occluded vessel. Associations of WHCT with 90-day functional outcome and hemorrhagic transformation at 24h were tested using logistic regression, adjusting for age, sex, baseline NIHSS score, ASPECTS, thrombolysis and mTICI score. Associations with cardiovascular risk factors were explored in univariate analyses. Results Higher WHCT was independently associated with worse functional outcome (mRS shift, mRS>2), but not with increased risk of hemorrhagic transformation (adjusted odds ratio per 1 Hounsfield unit increase [95%-CI]: 1.60 [1.22–2.11], 2.42 [1.44–4.05] and 1.31 [0.93–1.86], respectively). Elevated blood pressure and atrial fibrillation were significantly associated with higher WHCT, while dyslipidemia, diabetes, and smoking were not. Conclusions Increased WHCT is associated with poor outcomes after ischemic stroke and vascular risk factors. Further validation against MRI-based markers is needed to clarify its relationship to white matter and small vessel pathology. Conflict of interest Helge Kniep: consults for Eppdata GmbH, is shareholder of Eppdata GmbH. Götz Thomalla: received consultancy fees from Acandis, Boehringer Ingelheim, Bayer, and Portola, and fees as lecturer from Acandis, Alexion, Amarin, Bayer, Boehringer-Ingelheim, BMS/Pfizer, Daiichii Sankyo and Portola. He serves in the board of the TEA Stroke Study and of ESO. Jens Fiehler: is consultant for Cerenovus, Medtronic, Microvention, Penumbra, Phenox, Roche and Tonbridge. He serves on the advisory board of Stryker and Phenox. He is a stock holder of Tegus Medical, Eppdata and Vastrax. He serves as Associate Editor at JNIS. All other others declare no conflict of interest. Figure 1 - belongs to Methods Figure 2 - belongs to Results