Diagnostic accuracy of DSA in carotid artery stenosis: a comparison between stenosis measured on carotid endarterectomy specimens and DSA in 644 cases

Abstract

Background and purpose DSA (digital subtraction angiography) is the gold standard for measuring carotid artery stenosis (CS). Yet, the correlation between DSA and stenosis is not well documented. Material and methods We compared CS as measured by DSA to carotid artery specimens obtained from carotid endarterectomy surgery. Patients were divided into three groups according to NASCET criteria (North American Symptomatic Carotid Endarterectomy Trial): stenosis of 30–49% (mild), stenosis of 50–69% (moderate), and stenosis of 70–99% (severe). Results This prospective cohort study involved 644 patients. The mean stenosis in the mild stenosis group (n = 128 patients) was 54% ECST (European Carotid Surgery Trial), 40% NASCET, and 72% ESs (endarterectomy specimens). The mean absolute difference between ECST and NASCET was 14%. The mean stenosis in the moderate stenosis group (n = 347 patients) was 66% ECST, 60% NASCET, and 77% ES. The mean absolute difference between ECST and NASCET was 6%. The mean stenosis in the severe group (n = 169 patients) was 80% ECST, 76% NASCET, and 79% ES. No significant correlation coefficients were found between DSA and ES methods. In the mild group, the CC was 0.16 (ESCT) and 0.13 (NASCET); in the moderate group, the CC was 0.05 (ESCT) and 0.01 (NASCET); and in the severe group, the CC was 0.23 (ESCT) and 0.10 (NASCET). For all groups combined, CC was 0.22 for the ECST and 0.20 for the NASCET method. Conclusion The relationship between DSA and ES methods to measure CS is almost random. This lack of a relationship between the DSA and ES techniques questions the validity of current DSA-based guidelines.