Influence of time to endovascular stroke treatment on outcomes in the early versus extended window paradigms

Background The effect of time from stroke onset to thrombectomy in the extended time window remains poorly characterized. Aim We aimed to analyze the relationship between time to treatment and clinical outcomes in the early versus extended time windows. Methods Proximal anterior circulation occlusion patients from a multicentric prospective registry were categorized into early (<= 6 h) or extended (>6-24 h) treatment window. Patients with baseline National Institutes of Health Stroke Scale (NIHSS) >= 10 and intracranial internal carotid artery or middle cerebral artery-M1-segment occlusion and pre-morbid modified Rankin scale (mRS) 0-1 ("DAWN-like" cohort) served as the population for the primary analysis. The relationship between time to treatment and 90-day mRS, analyzed in ordinal (mRS shift) and dichotomized (good outcome, mRS 0-2) fashion, was compared within and across the extended and early windows. Results A total of 1603 out of 2008 patients qualified. Despite longer time to treatment (9[7-13.9] vs. 3.4[2.5-4.3] h, p < 0.001), extended-window patients (n = 257) had similar rates of symptomatic intracranial hemorrhage (sICH; 0.8% vs. 1.7%, p = 0.293) and 90-day-mortality (10.5% vs. 9.6%, p = 0.714) with only slightly lower rates of 90-day good outcomes (50.4% vs. 57.6%, p = 0.047) versus early-window patients (n = 709). Time to treatment was associated with 90-day disability in both ordinal (adjusted odd ratio (aOR), >= 1-point mRS shift: 0.75; 95%CI [0.66-0.86], p < 0.001) and dichotomized (aOR, mRS 0-2: 0.73; 95%CI [0.62-0.86], p < 0.001) analyses in the early- but not in the extended-window (aOR, mRS shift: 0.96; 95%CI [0.90-1.02], p = 0.15; aOR, mRS0-2: 0.97; 95%CI [0.90-1.04], p = 0.41). Early-window patients had significantly lower 90-day functional disability (aOR, mRS shift: 1.533; 95%CI [1.138-2.065], p = 0.005) and a trend towards higher rates of good outcomes (aOR, mRS 0-2: 1.391; 95%CI [0.972-1.990], p = 0.071). Conclusions The impact of time to thrombectomy on outcomes appears to be time dependent with a steep influence in the early followed by a less significant plateau in the extended window. However, every effort should be made to shorten treatment times regardless of ischemia duration.