IMPORTANCE
Stroke is a major complication of surgical aortic valve replacement.
OBJECTIVE
To determine the effectiveness and safety of cerebral embolic protection devices in reducing ischemic central nervous system injury during surgical aortic valve replacement
DESIGN, SETTING and PARTICIPANTS
A parallel-group trial conducted in 18 North American centers, randomizing patients with calcific aortic stenosis undergoing surgical aortic valve replacement between March 2015 and July 2016.
INTERVENTIONS
Use of one of two cerebral embolic protection devices (suction-based extraction;n=118 or intra-aortic filtration device;n=133) versus a standard aortic cannula (n=132) at the time of surgical aortic valve replacement.
MAIN OUTCOMES AND MEASURES
The primary endpoint was freedom from clinical or radiographic central nervous system (CNS) infarction at 7±3 days post-procedure. Secondary endpoints included a composite of clinical ischemic stroke, acute kidney injury, and death ≤30 days after surgery; delirium; mortality; serious adverse events and neurocognition.
RESULTS
Among 383 randomized patients (mean age 73.9, 38.4% women), freedom from CNS infarction at 7 days did not differ between suction-based extraction and controls (32.0 vs. 33.3%,difference −1.3,95%CI−13.8, 11.2) nor between intra-aortic filtration and controls (25.6% vs 32.4%, −6.9,95%CI−17.9, 4.2). Clinical stroke occurred in 5.1% suction-based extraction vs 5.8% controls (−0.7, 95%CI −6.5, 5.1) and 8.3% intra-aortic filtration vs. 6.1% controls (2.2, 95%CI −4.1, 8.4). Delirium trajectories (baseline to day 7) differed for suction-based extraction vs. control (p= 0.03) and intra-aortic filtration vs. control (p=0.02); by day 7, 6.3% of suction-based extraction vs 15.3% of control patients (−9.1, 95%CI −17.1, −1.0), and 8.1% of intra-aortic filtration vs 15.6% of control patients (−7.4, 95%CI−15.5, 0.6) experienced delirium. The 30-day composite endpoint did not differ between suction-based extraction and controls (21.4% vs 24.2%; −2.8, 95%CI −13.5,7.9) nor between intra-aortic filtration and controls (33.3% vs 23.7%; 9.7, 95%CI −1.2,20.5). Neurocognitive outcomes were not different between groups except for less decline in executive function in intra-aortic filtration vs controls (p=0.05).
CONCLUSIONS AND RELEVANCE
The incidence of CNS infarction after surgical aortic valve replacement was not altered by either embolic protection device. Potential benefits for delirium reduction, cognition and symptomatic stroke merit larger trials with longer follow-up.