Abstract Number: Esoc2026a1983 Morning Reperfusion in Large-Volume Ischaemic Stroke Yields Better Neurological and Functional Outcomes: A Tension Subanalysis

Abstract Background and aims In acute stroke, circadian biological processes modulate both cerebral susceptibility and the effectiveness of reperfusion. Observational studies report more favorable outcomes after morning thrombectomy. However, evidence from randomized trials examining diurnal effects remains sparse, especially among patients with large-volume infarcts. Methods Of 253 patients in the TENSION trial, 125 patients underwent thrombectomy. Reperfusion time was stratified by consensus into morning (05:00-10:59), daytime (11:00-16:59), evening (17:00-22:59), and nighttime (23:00-04:59). Associations with 90-day mRS and NIHSS were evaluated using multivariable regression models adjusted for prespecified key temporal and clinical covariates. Results Only morning reperfusion was associated with more favorable neurological and functional outcomes at 90 days with lower mRS (aOR 0.42; 95%-CI 0.22-0.79, p=0.007) and NIHSS (aOR 0.32; 95%-CI 0.14-0.72, p=0.006) (Figure 1A-B). This association remained significant among patients achieving complete reperfusion (mTICI3; aOR 0.34; 95%-CI 0.12-0.99; p=0.048, Figure 1C). Unlike daytime and evening thrombectomy, morning thrombectomy did not exhibit an increase in one-point mRS worsening per hour of treatment delay (morning: aOR 0.999; 95%-CI 0.962-1.037, p=0.095; evening: aOR 1.019; 95%-CI 1.004-1.035, p=0.012, Figure 1D). No significant time-of-day effect were detected in patients with poorest functional outcomes (p=0.29, Figure 1E). Conclusions In a randomized trial, time of endovascular reperfusion emerged as an independent determinant of outcome in patients with large-volume ischaemic stroke. Morning thrombectomy was consistently associated with better neurological and functional recovery at 90 days, indicating a clinically relevant diurnal pattern consistent with circadian modulation of post-reperfusion recovery. Conflict of interest All authors have nothing to disclose regarding this post-hoc analysis Figure 1 - belongs to Conclusions