General vs Nongeneral Anesthesia for Endovascular Thrombectomy in Patients With Large Core Strokes: A Prespecified Secondary Analysis of SELECT2 Trial

Abstract

Background and objectives: The association of anesthesia approach during endovascular thrombectomy (EVT) with clinical outcomes in large strokes is unexplored. We aimed to evaluate whether general anesthesia (GA), compared with non-GA, was associated with better functional outcomes in the SELECT2 trial.

Methods: In a prespecified secondary analysis of the SELECT2 trial that enrolled patients with large strokes on noncontrast CT (Alberta Stroke Program Early CT Score [ASPECTS] 3-5), CT perfusion/MRI (core volume ≥50 mL), or both, functional outcomes were compared in EVT-treated patients who received GA or non-GA and whether this association was modified by stroke severity (NIH Stroke Scale score), ischemic injury estimates, and collateral status was evaluated. The primary outcome was 90-day functional status (ordinal modified Rankin Scale [mRS]). Secondary outcomes were functional independence (mRS scores 0-2), independent ambulation (mRS scores 0-3), complete dependence or death (mRS scores 5-6), and mortality.

Results: Of 178 EVT patients (median [interquartile range] age 66 [58-75] years, stroke severity 19 [15-23], CT-ASPECTS 4 [3-5], and core volume 101.5 [70-138] mL, 71 women [39.9%]), 104 (58%) received GA. Time from randomization to arterial puncture was longer with GA (40 [23-59] minutes) vs non-GA (27 [18-47] minutes), but procedural duration (GA: 57 [31.5-77] minutes vs non-GA: 49.5 [30-71] minutes) was similar. Successful reperfusion (modified treatment in cerebral infarction [mTICI] score 2b-3) rates were similar (GA 81 (78%) vs non-GA 62 (84%), adjusted relative risk [aRR] 0.91, 95% CI 0.79-1.06). In addition, mRS distribution did not differ between GA and non-GA groups (adjusted generalized odds ratio 1.21, 95% CI 0.86-1.70), as well as independent ambulation (GA: 41% vs non-GA: 34%, aRR 1.22, 95% CI 0.86-1.74) and functional independence (GA: 22% vs non-GA: 18%, aRR 1.32, 95% CI 0.75-2.35). Stroke severity, ASPECTS, ischemic core volume, or collaterals did not modify the association between anesthesia and functional outcome (all p-interaction >0.05). Patients experienced systolic blood pressure (SBP) variability ≥40 mm Hg and minimum intraprocedural SBP (<100 mm Hg) more frequently with GA, but this did not modify GA association with functional outcomes (p-interaction = 0.77 and 0.89, respectively).

Discussion: In patients with large core strokes randomized in SELECT2, EVT outcomes did not differ significantly based on anesthesia approach (GA or non-GA) without heterogeneity across stroke severity and size. While GA was associated with higher SBP variability and lower minimum SBP, this did not modify GA association with functional outcomes. While allocation to anesthesia approach was nonrandomized, our findings suggest that optimizing institutional protocols for preferred anesthesia technique, whether GA or non-GA, may enhance EVT procedural outcomes.

Trial registration information: ClinicalTrials.gov ID: NCT03876457.

Classification of evidence: This study provides Class II evidence that in patients presenting within 24 hours with large vessel occlusion strokes undergoing EVT, the 90-day mRS score is comparable in those with or without GA.