Thrombectomy With Conscious Sedation Increases Functional Independence Compared to General Anesthesia: A DEFUSE-3 Post-hoc Analysis

Introduction: The optimal method of patient sedation for mechanical thrombectomy in large vessel occlusion, particularly in the delayed time window, remains unknown. We evaluated the endovascular arm patients of the DEFUSE-3 trial for outcome differences between general anesthesia (GA) and conscious sedation (CS). While CS was the suggested method in the trial, the choice was ultimately left to the discretion of the treating interventionalist. Methods: A post-hoc analysis of patients enrolled in the DEFUSE-3 study was performed to assess differences in functional independence (modified Rankin scale of 0-2) at 90 days comparing patients undergoing GA to CS. The groups were also compared for baseline clinical and imaging characteristics and procedural process measures. Statistical analyses were performed using the Mann-Whitney U test, χ 2 -test, and logistic regression analysis. Results: Among patients undergoing thrombectomy, 26 (28%) utilized GA, while 66 (72%) underwent CS. Baseline demographics, clinical and radiologic characteristics were similar between the groups. Functional independence at 90 days (Figure 1) was observed in 53% of patients undergoing CS, compared to 23% of those receiving GA (adjusted-odds ratio of 0.27 (0.08-0.93), p=0.037). Patients undergoing CS had better NIHSS scores at 24 hours compared to GA (8 versus 15.5, p=0.032), but not at discharge (4 v 8, p=0.133). Importantly, CS yielded a faster time from arrival in the angiography suite to femoral puncture (median 14 minutes, interquartile range (IQR) 9-18) compared to GA (18 min, IQR 12-22, p=0.0497). Time from femoral puncture to reperfusion was also shorter with CS (median 36 min, IQR 25-51) than with GA (48 min, IQR 32-83, p=0.004). Conclusions: Patients undergoing thrombectomy between 6 to 16 hours following stroke onset exhibit improved functional independence at 90 days when treated with CS as compared to GA. This difference may be mediated by faster time to reperfusion in this cohort.