Challenges and Pitfalls in CT-Angiography Evaluation of Carotid Bulb Stenosis: Is It Time for a Reappraisal?

Abstract

We aimed to perform an anatomical evaluation of the carotid bulb using CT-angiography, implement a new reliable index for carotid stenosis quantification and to assess the accuracy of relationship between NASCET and ECST methods in a large adult population. The cross-sectional areas of the healthy carotid at five levels were measured by two experienced radiologists. A regression analysis was performed in order to quantify the relationship between the areas of the carotid bulb at different carotid bulbar level. A new index (Regression indeX, RegX) for carotid stenosis quantification was proposed. Five different stenoses with different grade in three bulbar locations were simulated for all patients for a total of 1365 stenoses and were used for a direct comparison of the RegX, NASCET, and ECST methods. The results of this study demonstrated that the RegX index provided a consistent and accurate measure of carotid stenosis through the application of the ECST method, avoiding the limitations of NASCET method. Furthermore, our results strongly depart from the consolidated relationships between NASCET and ECST values used in clinical practice and reported in extensive medical literature. In particular, we highlighted that a major misdiagnosis in patient selection for CEA could be introduced because of the large underestimation of real stenosis degree provided by the NASCET method. A reappraisal of carotid stenosis patients' work-up is evoked by the effectiveness of state-of-the-art noninvasive contemporary carotid imaging.

Keywords: CT-angiography; ECST; NASCET; carotid stenosis.

Figure 1

(a) Measurements of luminal carotid areas. The areas of the common carotid artery (a) and the internal carotid (c) were measured at 2 cm below the carotid bifurcation and 2 cm above the bulb, respectively. The areas of the carotid bulb were measured at the origin (b1), at the middle third (b2), and at the distal third (b3); (b) box-and whisker plots of measured carotid cross-sectional areas in the five locations shown in (a), reporting the median values and the interquartile range (where 50% of the data are found). Mean values are shown as crosses.

Figure 2

Linear regression between b1 and b2 variables with (a) the regression line and the confidence intervals on mean and on values, (b) the plot or the normalized residuals, and (c) the plot of the Cook’s distance.

Figure 3

Box-plots of the distributions of the NASCET, ECST, and ReGX stenosis degree indexes for all simulated stenosis in the b1 location obtained with a residual lumen ranging of (a) 10 mm², (b) 8 mm², (c) 6 mm², (d) 4 mm² and (e) 2 mm².

Figure 4

Correlation scatter plot of ECST and NASCET values in b1 (black), b2 (red), and b3 (blue) locations. An example of diagnostic test was reported considering the current equivalence relation between NASCET (70%) and ECST (82%) stenosis degree. TP (True Positives), FP (False Positives), FN (False Negatives), and TN (True Negatives).