North American SOLITAIRE Stent-Retriever Acute Stroke Registry: choice of anesthesia and outcomes.

BACKGROUND AND PURPOSE: Previous work that predated the availability of the safer stent-retriever devices has suggested that general anesthesia (GA) may have a negative impact on outcomes in patients with acute ischemic stroke undergoing endovascular therapy. METHODS: We reviewed demographic, clinical, procedural (GA versus local anesthesia [LA], etc), and site-adjudicated angiographic and clinical outcomes data from consecutive patients treated with the Solitaire FR device in the investigator-initiated North American SOLITAIRE Stent-Retriever Acute Stroke (NASA) Registry. The primary outcomes were 90-day modified Rankin Scale, mortality, and symptomatic intracranial hemorrhage. RESULTS: A total of 281 patients from 18 centers were enrolled. GA was used in 69.8% (196/281) of patients. Baseline demographic and procedural factors were comparable between the LA and GA groups, except the former demonstrated longer time-to-groin puncture (395.4±254 versus 337.4±208 min; P=0.04), lower National Institutes of Health Stroke Scale (NIHSS; 16.2±5.8 versus 18.8±6.9; P=0.002), lower balloon-guide catheter usage (22.4% versus 49.2%; P=0.0001), and longer fluoroscopy times (39.5±33 versus 28±22.8 min; P=0.008). Recanalization (thrombolysis in cerebral infarction ≥2b; 72.94% versus 73.6%; P=0.9) and rate of symptomatic intracranial hemorrhage (7.1% versus 11.2%; P=0.4) were similar but modified Rankin Scale ≤2 was achieved in more LA patients, 52.6% versus 35.6% (odds ratio, 1.4 [1.1-1.8]; P=0.01). In multivariate analysis, hypertension, NIHSS, unsuccessful revascularization, and GA use (odds ratio, 3.3 [1.6-7.1]; P=0.001) were associated with death. When only anterior circulation and elective GA patients were included, there was a persistent difference in good outcomes in favor of LA patients (50.7% versus 35.5%; odds ratio, 1.3 [1.01-1.6]; P=0.04). CONCLUSIONS: The NASA Registry has demonstrated that clinical outcomes and survival are significantly better in patients treated with LA, without increased symptomatic intracranial hemorrhage risk. Future trials should prospectively evaluate the effect of GA on outcomes.