Correlation of Computed Tomography Angiography to Digital Subtraction Angiography in Carotid Stenosis with Real-World Assessment of Overestimation of Carotid Stenosis on Computed Tomography Angiography

Abstract

PurposeComputerized tomography angiography (CTA) is a well-established diagnostic modality for carotid stenosis (CS). However, false-positive CTA results may expose patients to unnecessary procedural complications in cases where surgical intervention is not warranted. We aim to assess the correlation of CTA to DSA in CS and characterize patients who were referred for intervention based on CTA and did not require it based on DSA. MethodsWe retrospectively reviewed 186 patients who underwent carotid angioplasty and stenting following pre-procedural CTA at our institution from April 2017 to December 2022. ResultsTwenty-one of 186 patients (11.2%) were found to have <50% carotid stenosis on DSA (discordant group). Severe plaque calcification on CTA was associated with a discordant degree of stenosis on DSA (LR+ = 7.4). Among 186 patients, agreement between the percentage of stenosis from CTA and DSA was weak-moderate (r2 = 0.27, p <0.01). Among concordant pairs, we found moderate-strong agreement between CTA and DSA (adj r2=0.37)(p<0.0001). Of 186 patients, 127 patients had CTA stenosis of ≥70%, and 59 had CTA of 50-69%. Correlation between CTA and DSA in severe CTA stenosis was weak (r2=0.11, p <0.01). ConclusionIn patients with stenosis found on CTA, over 88% also had stenosis on DSA, with this PPV in line with previous studies. The percent-stenosis value from CTA and DSA was weakly correlated but does not affect clinical judgement of stenosis overall. Severe calcification found on CTA may potentially indicate non-stenosis on DSA.