Refining endovascular thrombectomy for large vessel occlusion in active cancer: predictors of death despite successful recanalization.

BACKGROUND: Endovascular thrombectomy (EVT) in large vessel occlusion (LVO) with active cancer remains uncertain due to poor outcomes. This retrospective cohort study aimed to identify predictors of (1) early post-stroke death and death despite successful recanalization (SR) and (2) favorable outcomes among survivors. METHODS: We analyzed 153 patients with active cancer and EVT-eligible LVO from four tertiary stroke centers (2017-2024). We performed an exploratory 1:1 propensity score-matched comparison of EVT versus non-EVT management, as well as separate analyses of the EVT-treated cohort. The primary outcome was death within 3 months despite SR (modified Thrombolysis in Cerebral Infarction (mTICI) 2b/3), while secondary outcomes included early post-stroke death and favorable functional outcome among 3-month survivors. RESULTS: Propensity score matching (PSM) yielded a 1:1 matched cohort (non-EVT, n=30; EVT, n=30) with improved, though not perfect, covariate balance. In this exploratory matched cohort, EVT showed only a non-significant trend toward favorable outcome and was not associated with lower mortality. In the EVT-treated cohort, systemic metastasis and elevated D-dimer predicted early mortality. Among patients with SR, death despite SR occurred in 43% (37/86) and was associated with systemic metastasis and elevated D-dimer rather than procedural factors. Among 3-month EVT survivors, first-pass effect (FPE) was the only procedural factor independently associated with favorable functional outcome. CONCLUSIONS: In cancer-related LVO, EVT decisions should consider the risk of death despite SR based on metastasis and hypercoagulability. For patients with survival potential, achieving FPE is critical for favorable outcomes.