ASPECTS-based selection for late endovascular treatment: a retrospective two-site cohort study

Introduction The DAWN trial demonstrated the effectiveness of late endovascular treatment in acute ischemic stroke patients selected on the basis of a clinical-core mismatch. We explored in a real-world sample of endovascular treatment patients if a clinical-ASPECTS (Alberta Stroke Program Early CT Score) mismatch was associated with an outcome benefit after late endovascular treatment. Methods We retrospectively analyzed all consecutive acute ischemic stroke patients admitted 6-24 h after last proof of good health in two stroke centers, with initial National Institutes of Health Stroke Scale (NIHSS) >= 10 and an internal carotid artery or M1 occlusion. We defined clinical-ASPECTS mismatch as NIHSS >= 10 and ASPECTS >= 7, or NIHSS >= 20 and ASPECTS >= 5. We assessed the interaction between the presence of the clinical-ASPECTS mismatch and late endovascular treatment using ordinal shift analysis of the three-month modified Rankin Scale and adjusting for multiple confounders. Results The included 337 patients had a median age of 73 years (IQR = 61-82), admission NIHSS of 18 (15-22), and baseline ASPECTS of 7 (5-9). Out of 196 (58.2%) patients showing clinical-ASPECTS mismatch, 146 (74.5%) underwent late endovascular treatment. Among 141 (41.8%) mismatch negative patients, late endovascular treatment was performed in 72 (51.1%) patients. In the adjusted analysis, late endovascular treatment was significantly associated with a better outcome in the presence of clinical-ASPECTS mismatch (adjusted odd ratio, aOR = 2.83; 95% confidence interval, CI: 1.48-5.58) but not in its absence (aOR = 1.32; 95%CI: 0.61-2.84). The p-value for the interaction term between clinical-ASPECTS mismatch and late endovascular treatment was 0.073. Conclusions In our retrospective two-site analysis, late endovascular treatment seemed effective in the presence of a clinical-ASPECTS mismatch, but not in its absence. If confirmed in randomized trials, this finding could support the use of an ASPECTS-based selection for late endovascular treatment decisions, obviating the need for advanced imaging.