Peak systolic velocity and color aliasing are important in the development of duplex ultrasound criteria for external carotid artery stenosis

Abstract

ObjectiveThe external carotid artery (ECA) serves as a major collateral pathway for ophthalmic and cerebral artery blood supply. It is routinely examined as part of carotid duplex ultrasound, but criteria for determining ECA stenosis are poorly characterized and typically extrapolated from internal carotid artery data. This is despite the fact that the ECA is smaller in diameter, with a higher resistance and lower volume flow pattern. We hypothesized that using the cutoff of a peak systolic velocity (PSV) ≥125 cm/s, extrapolated from internal carotid artery data, will overestimate the prevalence of ≥50% ECA stenosis and aimed to determine a more appropriate criterion. MethodsFrom December 2016 to July 2017, consecutive carotid duplex ultrasound studies performed in our university hospital Intersocietal Accreditation Commission-accredited vascular laboratory were prospectively identified and categorized with respect to prevalence and distribution of ECA PSVs and color aliasing, an indication of turbulent flow or flow acceleration. Presence of color aliasing was determined by two individual reviewers and agreement assessed by Cohen κ coefficient. ECA stenosis was calculated by the North American Symptomatic Carotid Endarterectomy Trial (NASCET) method in patients with computed tomography angiography (CTA) performed within 3 months of carotid duplex ultrasound without an intervening intervention. Receiver operating characteristic analysis was performed to identify best criteria for determining ≥50% ECA stenosis. ResultsThere were 1324 ECAs from 662 patients analyzed; 174 patients had a total of 252 ECAs with PSV ≥125 cm/s (19% of the total sample). Of those ECAs with PSVs ≥125 cm/s, 30.5% were between 125 and 149 cm/s, 22.2% were between 150 and 174 cm/s, 13.1% were between 175 and 199 cm/s, and 34.1% were ≥200 cm/s. There were 341 ECAs that were analyzed for the presence of color aliasing. In 86 ECAs with PSV ≥200 cm/s, 58.1% had color aliasing, whereas in 255 ECAs with PSV <200 cm/s, only 19.2% had color aliasing (P = .0001). There were 325 CTA studies reviewed and assessed for the presence of a ≥50% ECA stenosis as determined by CTA. Overall, the combination of an ECA PSV ≥200 cm/s with the presence of color aliasing provided the highest combination of sensitivity (90%), specificity (96%), positive predictive value (83%), and negative predictive value (98%) and the greatest area under the curve of 0.971 for determining the presence of a ≥50% ECA stenosis based on CTA. ConclusionsA PSV ≥125 cm/s alone probably overestimates the prevalence of ≥50% ECA stenosis. A PSV ≥200 cm/s combined with color aliasing is highly predictive of >50% ECA stenosis based on correlation with CTA.