Abstract DP192: Association of Endovascular Thrombectomy With Mortality and Functional Outcomes in Large Ischemic Core: Analysis of Linked RAPID AI and Florida Stroke Registry data

Introduction: Six randomized clinical trials have shown that endovascular thrombectomy (EVT) reduces mortality and improves clinical outcomes in patients with large-core anterior circulation ischemic stroke. Whether these benefits extend to routine practice outside strict trial criteria remains uncertain. This study evaluates in-hospital outcomes among EVT-treated patients with large ischemic cores using linked data from the Florida Stroke Registry (FSR) and RapidAI imaging. Methods: From January 2023 to December 2024, linked FSR–RapidAI data were available for 11,275 cases of acute ischemic stroke (AIS). Patients with suspected large vessel occlusion (LVO) and infarct core size measured by computed tomography perfusion (CTP) or non-contrast computed tomography (NCCT) were included. Large core was defined as an Alberta Stroke Program Early CT Score (ASPECTS) ≤5 on NCCT, or a core volume ≥70 mL with cerebral blood flow <30% on CTP. Multivariable logistic regression compared in-hospital mortality and discharge ambulation (independent or assisted vs unable or not documented) between large-core patients treated with EVT and those untreated, adjusting for demographics, NIHSS score, pre-stroke mRS, presentation timing, onset-to-arrival time, center type, admission blood pressure, and intravenous thrombolysis. Results: Of 2,212 AIS patients, 174 (7.9%) had large-core infarcts (91% identified by CTP); 102 (59%) received EVT (Figure 1). The median age was 70.5 years (IQR 61-79), 56.3% were male, the median onset-to-arrival time was 130.5 minutes (IQR 47-518), and the median NIHSS score was 20 (IQR 14-25). Ambulation at discharge occurred in 70 patients (40.2%), and in-hospital mortality occurred in 28 patients (16.1%). In multivariable regression, EVT was associated with lower odds of death (adjusted odds ratio [aOR] 0.29; 95% CI 0.10-0.88) (Table 1a) and higher odds of ambulation at discharge (aOR 2.45; 95% CI 1.10-5.46) (Table 1b). The model demonstrated good discrimination, with an area under the receiver operating characteristic curve of 0.78 (95% CI 0.72-0.85). Conclusions: In this multicenter cohort of AIS patients with large-core infarcts, EVT was independently associated with higher odds of ambulation at discharge and lower in-hospital mortality after adjustment for stroke severity and presentation factors. These findings support emerging trial data that EVT improves survival and functional outcomes in large-core strokes in real-world practice.