Carotid Endarterectomy for Asymptomatic Stenosis Based on Duplex Ultrasound Alone Achieves Equivalent Perioperative and Long-Term Outcomes Relative to Advanced Imaging Based Endarterectomy

Abstract

The purpose of this study is to compare both perioperative as well as long-term outcomes of patients undergoing carotid endarterectomy (CEA) for asymptomatic carotid bifurcation stenosis based on duplex ultrasound in isolation relative to a combination of duplex and more advanced imaging. All CEA in the VQI between January 2003 and May 2022 were queried. We identified 171,816 CEAs in the database. Exclusions were symptomatic carotid lesion (57,742), lack of imaging documentation (908), lack of advanced imaging status (1,816), simultaneous additional arterial intervention in the carotid, coronary, or peripheral arterial system (n=4,118), and anatomic high-risk status for CEA (n=4,071). Included patients were then placed into 1 of 2 cohorts: patients undergoing CEA based on duplex imaging alone (n=33,437) and those undergoing CEA based on advanced imaging (CTA, MRA, or invasive angiography) with or without duplex (n=69,715). We performed multivariable analysis for the following outcomes utilizing CEA based on duplex in isolation as 1 of the variables: perioperative neurological ischemic event utilizing binary logistic regression; combined 90-day mortality and neurological ischemic event utilizing binary logistic regression; neurological event in long-term follow-up with date of surgery serving as time zero; time dependent Cox regression analysis; mortality in long-term follow-up utilizing time-dependent Cox regression. Carotid endarterectomy based on duplex alone and CEA based on advanced imaging had essentially equivalent rates of 90-day mortality (0.9% vs. 1.0%, P=0.108); combined perioperative neurological event and 90-day mortality (2.0% vs. 2.2%, P=0.042); and, return to the operating room (1.6% vs. 1.7%, P=0.154). On multivariable analysis CEA based on advanced imaging was noted to have a slightly higher absolute rate of perioperative neurological event without achieving multivariable significance (1.3% vs. 1.2%, adjusted odds ratio 1.11 (0.98-1.25), P=0.092. CEA based on advanced imaging had a higher rate of neurological event after index hospital admission relative to duplex in isolation (hazard ratio (HR) 1.44 (1.31-1.60), P<0.001). However, the absolute percentage difference was just 0.5% (1.6% vs. 2.1%). CEA based on duplex alone was associated with a slightly increased risk of mortality in LTFU (HR 1.16 (1.11-1.21), P<0.001). At 5years the absolute risk of mortality was less than 1% different between the cohorts. Performing CEA for asymptomatic bifurcation stenosis based on duplex ultrasound alone is a safe practice which achieves clinically equivalent perioperative and long-term freedom from cerebral ischemia and mortality relative to CEA based on advanced imaging. This has potential implications for health care cost saving as well as avoidance of radiation and iodinated contrast.