Regarding "Evaluation of carotid artery stenosis: is duplex ultrasonography sufficient?".

Abstract

The recent article by Dr. Paula M. Muto et al. (1996;24:17-24) requires some comment. There are deficiencies in the magnetic resonance angiographic (MRA) technique used, which no longer represents what most would regard as appropriate MRA for carotid artery disease. Dr. Muto et al. used two-dimensional time-of-flight images for the extracranial vessels and three-dimensional time-of-flight for the Circle of Willis and intracranial vessels. This technique limits spatial resolution and also flow sensitivity in the extracranial vessels, which reduces the accuracy of stenosis quantification. It is also not reasonable to assume that a signal void indicates a stenosis ≥70%. Signal voids may be a result of other causes, such as abnormal vessel orientation or flow turbulence from ulceration. In addition, it is unclear whether the imaging was extended down to the level of the aortic arch to detect proximal aortic branch vessel stenosis. The number of cases is also small, and the failure to detect any tandem lesions, proximally or distally, may reflect a sampling error as a result of the small sample volume. To state that duplex ultrasound alone provides all the information that is provided by MRA would require a larger patient population. We have found in our patient population that a significant number of tandem lesions, some of which are surgically relevant, are detected by MRA but are not detected by duplex ultrasound. There is also a problem when comparing the degree of stenosis by duplex ultrasound with that found by MRA. The categories described by Muto et al. are different. Results of duplex ultrasound are reported as a percentage of stenosis, whereas results of MRA are reported as mild, moderate, or high-grade stenosis. Surely, if these two methods are going to be compared, the same grading protocol and nomenclature should be used. The interpretation of these examinations should be performed by the same readers throughout the population rather than by one of a pool of four neuroradiologists (who apparently were not directly involved in this study, as they are not acknowledged in the list of authors). Some of the discrepancy between MRA and duplex ultrasound could represent interobserver variability rather than any limitations on the MRA technique. We have demonstrated that using modern MRA techniques, including three-dimensional MRA in the neck and electrocardiographic MRA for the aortic arch and extracranial carotid arteries, relevant extra lesions remote from the carotid bifurcation can be detected by MRA that are not detected by duplex ultrasound.1 The conclusions of the study by Dr. Muto et al. may be the result of their small patient population and limited MRA technique rather than being a result of the absence of any value of performing MRA in these patients. 24/41/80564