Abstract WP33: DEFUSE 3 Trial Outcomes not Affected by the Enrollment Rates of the Participating Centers

Abstract

Introduction: Heterogeneity of outcomes between enrolling sites can affect the results of multicenter trials. We investigated if key DEFUSE 3 trial outcomes differed between low and high enrolling sites. Methods: DEFUSE 3, a randomized trial assessing the efficacy of mechanical thrombectomy in patients who could be treated 6 to 16 hours from last known to be well, was conducted at 38 US centers from the NIH StrokeNet network. We grouped centers by the number of the enrolled patients and compared rates of independent functional outcome (mRS 0-2 at 90-days) and successful recanalization (TICI score of 2b-3) between groups using Fisher’s exact test. Results: The 38 DEFUSE 3 centers enrolled a median of 4 (IQR 3-5) patients (range 1 to 15; Table). Rates of functional independence did not differ between centers when they were dichotomized by total enrollments. The rates of functional independence in the EVT arm were 38% in sites with 1-5 enrollments vs 52% in sites with 6-15 enrollments (p=0.21). Rates of functional independence in the medical arm were 19% in sites with 1-5 enrollments and 15% in sites with 6-15 enrollments (p=0.78). Similarly, there was no difference in rates of functional independence between the EVT and medical groups when centers were grouped by tertiles of total enrollments (1-4, 5-8, 9-15), p=0.40 for EVT and 0.78 for medical arm; or quartiles (1-3, 4-5, 6-12, 13-15) p=0.17 for EVT and 0.89 for medical arm. Similarly, there was no difference in the rates of recanalization between groups when dichotomized: 75% (1-5 patients) vs 77% (6-15 patients), p=0.50; trichotomized (71%, 79%, 78%, p=0.75); or assessed by quartiles (74%, 76%, 73%, 82%, p=0.91). Conclusions: Enrollment rate heterogeneity among the StrokeNet centers participating in the DEFUSE 3 trial did not affect the rates of functional independence or successful recanalization.