Grading of internal carotid artery stenosis: validation of Doppler/duplex ultrasound criteria and angiography against endarterectomy specimen.
OBJECTIVES: duplex ultrasound has replaced angiography prior to carotid endarterectomy (CEA) in many institutions. However, the indications for CEA are based on angiographically controlled studies and widely accepted ultrasound criteria do not exist. Consequently, the reliability of Doppler and/or duplex ultrasound to predict a high-grade ICA stenosis has to be proven. DESIGN: prospective validation study. MATERIALS: one hundred and fifty carotid bifurcations assessed by ultrasound and selective angiography and 68 acrylat outcasts of carotid specimen after eversion CEA. METHODS: ICA stenosis was measured angiographically according to the ECST criteria. Combined Doppler acoustic standard criteria (CDASC), peak systolic frequency (PSF), peak systolic velocity (PSV) and end-diastolic velocity (EDV) served as criteria for the ultrasound assessment. These criteria and the results of angiography were compared to the degree of ICA stenosis determined by specimen measurements. RESULTS: the median degree of ICA stenosis as assessed by angiography (82%, range 56-97%) and CDASC (83%, range 50-99%) corresponded well to the specimen measurements (80%, range 50-95%). The sensitivity of angiography and CDASC to predict a 70-90% ICA stenosis (ECST criteria) compared to the specimen measurements was 88% and 95%, respectively. The positive predictive value (PPV) reached 92% and 96%, respectively. CDASC were equivalent to angiography and were superior to the best single frequency or velocity parameters. If CDASC do not indicate a >/=70% ICA stenosis in spite of a PSV >/=180 cm/s and/or an EDV >/=50 cm/s, angiography may detect patients with a >70% ICA stenosis. CONCLUSIONS: CDASC are valid in the quantification of high-grade ICA stenosis. They are more reliable than single velocity and/or frequency measurements. However, if velocity criteria and CDASC do not agree, angiography should be performed.