Cause-Specific Endovascular Thrombectomy Time Windows in Basilar Artery Occlusion: ATTENTION Trial Post Hoc Analysis

BACKGROUND: Endovascular thrombectomy (EVT) outperforms best medical treatment (BMT) for basilar artery occlusion (BAO), but the influence of stroke cause on the time sensitivity of EVT remains unclear. We hypothesized nonatherosclerotic BAO would require earlier EVT, whereas atherosclerotic BAO—with chronic stenosis and more robust collaterals—would retain EVT benefit in extended windows. We conducted a post hoc 3-way interaction analysis of the ATTENTION trial (Trial of Endovascular Treatment of Acute Basilar Artery Occlusion) to explore cause-specific EVT time patterns. METHODS: Using randomized ATTENTION trial data (36 Chinese centers, 2021–2022), patients were stratified by onset-to-puncture time (<6 versus ≥6 hours) and TOAST (Trial of ORG 10172 in Acute Stroke Treatment) cause (atherosclerotic: large-artery atherosclerosis; nonatherosclerotic: cardioembolism/undetermined/other causes). The primary outcome was 90-day favorable functional outcome (modified Rankin Scale score 0–3), with relative risk as the primary estimand; the 3-way interaction was tested via joint Wald test. A weighted 1-hour interval odds ratio trend analysis served as an exploratory tool, with multivariable regression adjusting for confounders. RESULTS: Among 340 patients (226 EVT+BMT, 114 BMT), for onset-to-puncture time <6 hours, patients without atherosclerosis with EVT achieved a significantly higher favorable outcome rate (46.7% versus 20.4%; relative risk, 2.29 [95% CI, 1.25–4.18]; P =0.007); patients with atherosclerosis showed numerical outcome benefit (40.4% versus 26.1%, P =0.259). For onset-to-puncture time ≥6 hours, numerical benefit of patients with atherosclerosis increased from 14.3% to 30.6% (56.9% versus 26.3%; P =0.199), accompanied by ordinal modified Rankin Scale score improvement (common odds ratio, 3.30 [95% CI, 1.13–9.62]; P =0.029) and National Institutes of Health Stroke Scale score reduction; patients without atherosclerosis’ benefit decreased from 26.3% to 17.8% with loss of significance (39.5% versus 21.7%, P =0.341). The treatment time×cause interaction was not significant ( P =0.662). The weighted 1-hour interval odds ratio trend analysis identified a provisional 5.3-hour reference point in patients without atherosclerosis. CONCLUSIONS: Compared with nonatherosclerotic BAO, atherosclerotic BAO appears less time-sensitive, potentially retaining EVT responsiveness over longer windows. These findings are hypothesis-generating and require prospective validation. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT04751708.