MR angiographic and sonographic indications for endarterectomy.

PURPOSE Our objective was to determine whether appropriate criteria could be developed for performing an endarterectomy on the basis of sonographic and MR angiographic findings. METHODS Fifty patients were examined prospectively with sonography, MR angiography, and conventional angiography. All three imaging studies were performed within 2 weeks of one another, and conventional angiography served as the reference standard. RESULTS All 10 carotid occlusions were detected with sonography and MR angiography. Sonography accurately showed flow in two arteries, and MR angiography showed flow in one of three nearly occluded arteries with extremely slow flow. Multislab three-dimensional time-of-flight MR angiographic sequences underestimated the degree of stenosis in 12 arteries, and in two cases this resulted from high T1 signal within the atherosclerotic plaque. With conventional angiography as the reference standard for 70% to 99% stenosis, sonography had a sensitivity of 96%, a specificity of 91%, and a positive predictive value of 90%, while concordant sonographic findings and the presence of a signal void on multislab 3-D time-of-flight sequences had a sensitivity of 72%, a specificity of 98%, and a positive predictive value of 97%. CONCLUSION Endarterectomy performed on the basis of sonographic findings of 70% to 99% stenosis and of a signal void on multislab 3-D time-of-flight MR angiographic sequences is appropriate.