Grading of internal carotid artery stenosis: comparative analysis of different flow velocity criteria and multidetector computed tomographic angiography

Abstract

Purpose: To evaluate multidetector computed tomographic angiography (CTA) versus published color Doppler sonography (CDS) velocity criteria in the grading of internal carotid artery (ICA) stenosis.

Methods: Sixty-eight consecutive patients (50 men; mean age 70.2 +/- 8.1 years, range 51- 85) with known ICA stenosis and complete CTA and CDS data for 127 carotid arteries were enrolled in this retrospective analysis. The degree of stenosis was determined using CDS velocities according to 5 published sets of criteria, as well as the criteria used at the authors' institution. These outcomes were then correlated using kappa-statistics with the results of multidetector CTA according to NASCET.

Results: The best overall agreement was achieved applying the criteria sets of Hwang (kappa = 0.70) and AbuRahma (kappa = 0.68). All 5 occlusions were correctly identified with both modalities. CTA detected 73 ICA stenoses > 70%; the best correlation was with the application of Hwang criteria, which correctly identified 69 (94.5%) > 70% stenoses. In order of increasing tendency to underscore the grade of stenosis, the corresponding results for the other criteria sets were 62 (84.9%) for Mittl, 59 (80.8%) for AbuRahma, 55 (75.3%) each for Nicolaides and Filis, and 50 (68.5%) for Nederkoorn. CTA detected 85 stenoses >50%; the sensitivity of all applied CDS criteria sets exceeded 90%.

Conclusion: Grading of ICA stenosis > 70% with CDS and CTA results in clinically relevant discrepancies, with higher grades of stenoses assessed by CTA. The choice of CDS grading criteria is of significant clinical importance, especially in the identification of high-grade ICA stenosis.