Time to Treatment in Stroke Thrombectomy and Outcomes in the Extended Time Window: A Meta-Analysis.

Abstract (English)

BACKGROUND AND OBJECTIVES: Time to endovascular thrombectomy (EVT) is a critical determinant of outcomes for large vessel occlusion (LVO) strokes presenting within 6 hours of onset. Its impact in the extended (6-24-hour) window remains uncertain. We aimed to evaluate the association between treatment times and outcomes in this window. METHODS: Individual patient-level data from 6 randomized trials enrolling patients in the extended window were pooled. The primary outcome was degree of disability at 90 days (modified Rankin Scale [mRS] 0-6). Secondary outcomes included functional independence (mRS 0-2), mortality, and symptomatic intracranial hemorrhage. RESULTS: Among 505 participants (median age 70 years; baseline NIH Stroke Scale 16; 51.3% female; advanced imaging selection 86.5%), 266 (52.7%) received EVT and 239 (47.3%) medical therapy alone. In EVT-treated patients, longer onset-to-randomization times were not associated with differences in disability (adjusted odds ratio [aOR] per 60 minutes 1.02; 95% CI 0.96-1.08; <i>p</i> = 0.53) or functional independence (aOR 1.06; 95% CI 0.95-1.17; <i>p</i> = 0.28). Conversely, control patients exhibited worse outcomes with increasing onset-to-randomization times (aOR for disability 0.93; 95% CI 0.87-1.00; <i>p</i> = 0.041; functional independence 0.84; 95% CI 0.73-0.97; <i>p</i> = 0.019), resulting in greater treatment benefit at later times (<i>p</i>-interaction = 0.033 and 0.003, respectively). No association was observed between onset-to-puncture or onset-to-reperfusion and outcomes in EVT patients. However, longer randomization-to-reperfusion times correlated with worse disability (aOR 0.58; 95% CI 0.37-0.91; <i>p</i> = 0.018) and lower functional independence (aOR 0.51; 95% CI 0.28-0.96; <i>p</i> = 0.038). DISCUSSION: This study demonstrates that the benefit of EVT is consistently preserved across the extended (6-24-hour) time window when patients are selected using advanced imaging criteria. In this highly selected population, enriched with slow progressors, the apparent increase in treatment effect with longer onset-to-presentation times reflects the validity of physiologic selection rather than true time insensitivity. However, this observation should not diminish the critical importance of time: imaging-based selection effectively resets the treatment clock, masking the harmful impact of prehospital delays. Notably, longer in-hospital delays were strongly associated with worse outcomes, reinforcing that "time is brain," even in extended-window cohorts.