Stroke Severity and Functional Benefit of Thrombectomy in Acute M2 Middle Cerebral Artery Occlusion: A Multicenter Cohort Study.

Abstract

Recent randomized trials reported no overall functional benefit of endovascular treatment (EVT) for distal medium-vessel occlusion (DMVO) and did not identify consistent effect modifiers to guide patient selection. Consequently, the role of EVT-particularly for M2 occlusions-remains controversial. We investigated whether baseline clinical severity modifies the association between successful recanalization and 90-day outcome in patients with M2 occlusion. Multicenter retrospective cohort study at 2 tertiary stroke centers including consecutive adults with acute ischemic stroke due to M2 occlusion (January 2015-January 2023) triaged by multimodal CT and treated with EVT. The primary end point was functional independence (modified Rankin Scale [mRS] ≤ 2) at 90 days. Secondary end points included symptomatic intracerebral hemorrhage (sICH), mRS 0-1, and penumbra salvage volume (PSV). The primary analysis used multivariable logistic regression with baseline National Institutes of Health Stroke Scale (NIHSS) modeled linearly and an NIHSS×recanalization (modified Thrombolysis in Cerebral Infarction [mTICI] ≥2b) interaction term. Johnson-Neyman probing and inverse probability weighting (IPW) were applied; a supplementary restricted cubic spline analysis was performed to explore potential nonlinearity. Among 147 patients, 85 (58%) achieved successful recanalization. The mean age was 74 years (SD 13), and 47% were female. Higher baseline NIHSS (adjusted odds ratio [aOR] 0.83 per point, 95% CI 0.73-0.94) and older age (aOR 0.96 per year, 95% CI 0.94-0.99) were associated with lower odds of functional independence. NIHSS significantly modified the association between recanalization and outcome (interaction <i>p</i> = 0.03). The magnitude of the association between successful recanalization and functional independence was larger at higher NIHSS. Model-based estimates suggested a descriptive crossover around NIHSS 10, whereas statistical evidence of benefit emerged only at higher NIHSS values. Successful recanalization was associated with greater PSV (+33 mL, <i>p</i> = 0.01). In the PSV interaction model, onset-to-imaging time differed by recanalization status (<i>p</i> < 0.01): time was positively associated with PSV in the mTICI ≤2a group, but near zero in mTICI ≥2b. IPW analyses were concordant. sICH occurred in 8.2% vs 4.8% (<i>p</i> = 0.42). In M2 occlusions, the magnitude of the association between successful recanalization and functional independence was larger at higher NIHSS and not reliably demonstrable at lower NIHSS. These findings support a severity-informed, individualized EVT approach while remaining hypothesis-generating rather than prescriptive for specific NIHSS thresholds. Major limitations include the retrospective observational design and potential residual confounding.