CTA quantification of internal carotid artery stenosis: Application of luminal area vs. luminal diameter measurements and assessment of inter-observer variability

Background and Purpose. In contrast to digital subtraction angiography (DSAdia), computed tomography angiography (CTA) provides exact delineation of the perfused lumen in the axial plane, thus allowing luminal (CTAdia) as well as cross-sectional area (CTAarea) internal carotid artery stenosis (ICAS) assessment. The purposes of the present study were to correlate CTAdia and CTAarea with DSAdia and to assess the inter-observer variabilities of both CTA techniques. Methods. In a retrospective analysis, CTA images were reviewed by two observers and ICAS was assessed according to North American Symptomatic Carotid Endarterectomy Trial applying CTAdia and CTAarea. DSAdia was assessed by a third observer. Results. Based on 54 consecutive patients (40 males [74.1%] and 14 females [25.9%]; median age 73.3 years), ICAS percentages of CTAdia and CTAarea revealed significant correlations with DSAdia (r = 0.79-0.87, all P <.001) with median differences in the range of +8% to -6%. Inter-observer agreement was moderate for CTAdia (kappa = 0.60) and excellent for CTAarea (kappa = 0.86). Sensitivity of CTAarea for the detection of ICAS > 70% was 100% for both observers, corresponding results for CTAdia were 97.1% and 71.4%, respectively, using DSAdia as the gold standard. Conclusion. CTAarea assessment of ICAS correlates well with the results of DSAdia and provides an excellent sensitivity for the detection of ICAS > 70% with superior inter-observer agreement compared to CTAdia.