Abstract Body: Introduction: Multiple trials recently established efficacy and safety of endovascular thrombectomy (EVT) in patients with large strokes, with different eligibility requirements for time to randomization. However, the effect of time from last known well (LKW) to randomization on functional outcomes and EVT treatment effect is not fully characterized in these patients.

Methods: In a pre-specified analysis of SELECT2 , the association of time from LKW to randomization with imaging characteristics, EVT treatment effect and clinical outcomes was assessed. All analyses were adjusted for age, stroke severity, CT ASPECTS and CTP/MRI core volume.

Results: Median (IQR) time from LKW to randomization was 9.31 (5.66-15.33) hours [EVT: 9.07 (5.27-15.33) hours, MM: 9.79 (5.82-15.32) hours], with 100 patients (28.4%) randomized within early (0-6 hours) and 252 (71.6%) randomized within late window (6-24 hours), with similar baseline clinical and imaging characteristics. As time from LKW to randomization increased, CTP core volume (rCBF<30%) appeared to decrease (p-for-trend<0.0001). However, CT hypodensity volume tended to increase (p-for-trend: 0.059) representing underestimation of ischemic injury by CT early and by CTP late due to partial reperfusion and collateral recruitment (fig1). EVT treatment effect was consistent among those presenting in early (aGenOR: 1.84, 95% CI: 1.23 - 2.75, p: 0.003) and late (aGenOR: 1.68, 95% CI: 1.28 - 2.22, p<0.001) time windows, without significant heterogeneity (p_int: 0.607) (fig2). Treatment effect estimates also favored EVT across various imaging strata based on CT ASPECTS and CTP (rCBF<30%) or MRI (Apparent Diffusion Coefficient <620x10^-6 mm²/s) core volume in both early and late time windows (forest plots - fig3), along with age and stroke severity. Furthermore, among patients receiving EVT, time to randomization was not significantly associated with worse functional outcomes (aGenOR: 0.98, 95% CI: 0.96 - 1.01 per hour, p: 0.276), independent ambulation (aRR: 0.99, 95% CI: 0.96 - 1.03, p: 0.727 per hour) or mRS 5-6 (aRR: 1.01, 95% CI: 0.98 - 1.04 per hour, p: 0.588).

Conclusion: Thrombectomy benefit was maintained in both early and late time window in patients with large core randomized in SELECT2, without significant heterogeneity of treatment effect based on imaging modality, key clinical and imaging subgroups. Furthermore, longer time to randomization was not associated with worsened clinical outcomes among EVT patients.