Endovascular revascularization for progressive stroke beyond conventional time windows: safety, efficacy, and predictors of technical failure and restenosis.

Progressive stroke (PS) due to intracranial large vessel lesions (LVLs) has limited treatment options. While endovascular revascularization (EVR) is proposed as a salvage therapy, evidence beyond the conventional therapeutic window is scarce. This multicenter retrospective study evaluated the safety, efficacy, and durability of EVR in PS patients attributable to intracranial LVLs beyond the conventional therapeutic window. We retrospectively analyzed PS patients (NIHSS increase ≥ 4) with intracranial LVLs (severe stenosis > 70% or occlusion), infarct core ≤ 70 mL, and confirmed hypoperfusion who underwent EVR at three centers (2020-2024). Outcomes included technical success, perioperative complications, restenosis, and stroke recurrence. A total of 76 patients (median age 62 years; 71.4% male) were included: with 46 presenting with stenosis and 30 with occlusion. The mean interval from symptom onset to progression was 3.92 ± 1.59 days. Technical success was achieved in 96.1% (73/76) of patients; the three failures were all Mori C lesions with blunt stump morphology on baseline angiography, consistent with chronic occlusions of relatively long duration. Perioperative complications occurred in 7.9% (6/76), with only 2.6% (2/76) resulting in disabling symptoms at discharge. The mean NIHSS score decreased significantly from 10.11 ± 3.43 preoperatively to 5.14 ± 3.07 at discharge (p < 0.001). Over a mean follow-up of 11.9 months, the overall restenosis rate was 22.9% (16/70), with 6.8% (5/70) being symptomatic. Restenosis rates varied significantly by treatment modality: 31.3% for balloon angioplasty alone, 22.6% for self-expanding stents, and 11.5% for balloon-expandable stents (p for trend = 0.041). Critically, within the balloon-expandable cohort, no in-stent restenosis was observed in the drug-eluting NOVA stent subgroup (0/12, 0%), compared to 21.4% (3/14) in the bare-metal Apollo stent subgroup. Smoking was identified as an independent risk factor for restenosis, with continued smoking after EVR associated with significantly higher ISR rates (69.2% vs. 29.4%; OR = 5.40, p = 0.038). The rate of recurrent ischemic events in the target territory was 8.2% (6/73). EVR appears to be feasible, safe and effective in carefully selected PS patients with LVLs beyond conventional time windows. However, long-term durability seems dependent on the revascularization modality, with intracranial DES showing encouraging patency. Additionally, smoking cessation may significantly reduce restenosis risk. These findings support the integration of EVR into the therapeutic options for PS and highlight the necessity of preoperative imaging evaluation. Nevertheless, these results require confirmation in larger prospective studies.