Abstract TP255: Comparative Analysis of Carotid Artery Stenosis Diagnosis: Concordance Between Carotid Duplex Ultrasound and Computed Tomography Angiography and Its Impact on Intervention&Patient Outcomes

Background: Carotid artery stenosis (CAS) remains a crucial treatable contributor to ischemic stroke. Accurate imaging is the cornerstone of determining candidacy for revascularization. While computed tomography angiography (CTA) and carotid ultrasound (CUS) are routinely used, discrepancies between these modalities can affect treatment decisions. Despite frequent tandem use, concordance between CTA and CUS is not well understood. Thus, this study explored rates of concordance between CTA and CUS and associations between imaging concordance and plaque characteristics/location. Methods: A total of 95 adult stroke patients hospitalized between January 2017 and June 2024 who underwent both CTA and CUS within a 1-month interval were included. CAS severity was stratified as normal, mild, moderate, severe, and occlusion using NASCET criteria. Concordance was defined as agreement in the stenosis severity category between CTA and CUS. Patient and plaque characteristics&plaque location were examined by concordance of each imaging side: left (L-side), right (R-side), and intervention side (INT-side; n=34). Results: Concordance rates were about half: 51.6% in L-side, 55.8% in R-side, and 44.1% in INT-side. Among discordant cases, CTA more frequently overestimated CAS severity compared to CUS—60.9% in L-side, 64.3% in R-side, and 78.9% in INT-side. In the L-side, a higher moderate stenosis of CUS in discordance vs. concordance (41.3% vs. 14.6%; p=0.031) was found. Interestingly, a higher mild stenosis (49.1% vs. 9.5%; p<0.001) of CTA was observed in concordance vs. discordance of the R-side. No differences in patient/plaque characteristics or plaque location were found between concordant and discordant imaging findings for the three sites, except for a higher proportion of anticoagulant users in the discordant R-side group (23.8% vs. 7.5%; p=0.026). Among discordant intervention cases (n=19), current smokers (100.0% vs. 20.0%; p=0.013), CKD (100.0% vs. 13.3%; p=0.004), and anticoagulant use (75.0% vs. 13.3%; p=0.037) were associated with overestimated CAS on CUS. Conclusions: Our study shows that discordance between CTA and CUS is prevalent (~50%) and suggests that patients who smoke, have CKD, or use anticoagulants tend to have overestimated stenosis in the INT-side on CUS. This indicates CUS and CTA findings be used together to determine intervention candidacy in certain groups and future studies with larger sample sizes should explore the low concordance in the INT-side.