Endovascular Thrombectomy Versus Best Medical Therapy for Late Presentation Acute Ischemic Stroke With Proximal Large‐Vessel Occlusion Selected on the Basis of Noncontrast Computed Tomography: A Retrospective Analysis of 2 Prospectively Defined Cohorts

Abstract

Background The efficacy and safety of endovascular thrombectomy (EVT) >6 hours from acute ischemic stroke (AIS) onset for patients selected without computed tomography (CT) perfusion or magnetic resonance imaging is undetermined in routine clinical practice. Methods In this single‐center study, we identified consecutive late‐presenting patients with AIS who were eligible for EVT on the basis of noncontrast CT/CT angiography (without CT perfusion or magnetic resonance imaging) using an Alberta Stroke Program Early CT Score of ≥6, >6 hours from stroke onset, between January 2018 and March 2022. During the study period, EVT capacity limitations meant EVT‐eligible patients presenting out of regular working hours, consistently received best medical management (BMM). Functional outcomes (modified Rankin Scale at 90 days), symptomatic intracranial hemorrhage, and mortality at 90 days were compared between patients receiving EVT or BMM following multivariable adjustment for age, sex, baseline stroke severity, Alberta Stroke Program Early CT Score, onset‐to‐neuroimaging time, intravenous thrombolysis, and clot location. Results Among 4802 patients with AIS, 150 patients (3.1%) presenting beyond 6 hours of onset were eligible for EVT: 74 (49%) treated with EVT and 76 (51%) with BMM. Compared with the BMM group, patients treated with EVT had significantly improved functional outcome (modified Rankin Scale) (adjusted common odds ratio, 2.23 [95% CI, 1.18–4.22]; P =0.013), and higher rates of functional independence (modified Rankin Scale ≤2; 39.2.% versus 9.2%; adjusted odds ratio, =4.73 [95% CI, 1.64–13.63]; P =0.004). No significant difference was observed between the EVT and BMM groups in the symptomatic intracranial hemorrhage (5.4% versus 2.6%; P =0.94) or mortality (20.2% versus 47.3%; P =0.16) rates, respectively. Conclusion In routine clinical practice, of the 3.1% of patients in our AIS population presenting after 6 hours from stroke onset who were deemed eligible for EVT by noncontrast CT/CT angiography alone, those treated with EVT achieved significantly improved functional outcome, compared with patients treated with BMM only. No significant differences were noted between the 2 groups with respect to symptomatic intracranial hemorrhage and mortality. While confirmatory randomized trials are awaited, these findings suggest that EVT is effective and safe when performed in patients with AIS selected without CT perfusion or magnetic resonance imaging >6 hours from stroke onset.