Duplex Scanning and Spectral Analysis of Carotid Bifurcation Atherosclerotic Disease

Important Concepts

Guidelines recommend expressing the degree of ICA stenosis as a percentage of stenosis in relation to the distal normal ICA (NASCET [North American Symptomatic Carotid Endarterectomy Trial] stenosis) in routine clinical practice (see Chapter 6). Expressing the stenosis as a percentage of the bulb diameter (ECST [European Carotid Surgery Trial] stenosis) is needed in certain studies because it reflects the plaque volume and it is directly related to stroke risk. To avoid confusion, it should be remembered that the same velocities can be expressed in both NASCET and ECST grades of stenosis ( Table 1 ).

The Doppler velocity recorded at the site of maximum stenosis depends on the pressure gradient across the stenosis. This pressure gradient is a function of both the severity of the stenosis and the collateral circulation. Let us assume that three patients have a 70% diameter angiographic stenosis of the left ICA. The right carotid bifurcation is normal in the first patient, but in the second it is occluded. In the third patient the right carotid bifurcation is normal, but the circle of Willis is incomplete. In these three patients the velocities at the stenosis in the left ICA are different because the pressure gradients are not the same. The degree of stenosis can be overestimated in the second and third patients if the physician adheres to absolute velocities. For these reasons, the use of defined stenotic strata is recommended, and duplex scanning is found to be more accurate when lesions are classified as being above or below a single level, such as 60% stenosis or 70% stenosis.

Velocity ratios were introduced following the realization that absolute velocity criteria could be inaccurate in the presence of cardiac arrhythmia, aortic valve disease, tandem plaques, recent hemispheric stroke, carotid dilatation, or aneurysm, which could result in underestimation of the degree of stenosis. Conversely, carotid coiling or kinking, arteriovenous malformations, carotid body tumors, and contralateral severe stenosis or occlusion could produce an overestimation of the stenosis.

The criteria for different velocity ratios were initially tailored to screening and selecting patients for subsequent angiography. Thus, they had a high sensitivity at the expense of specificity. The aim was to not miss a patient with a stenosis greater than 60% (NASCET). However, subsequently, with the need to make decisions to proceed to carotid endarterectomy without angiography, the criteria were tailored to have equally high sensitivity and specificity. The criteria listed in Table 1 are based on the latter option, with emphasis on obtaining the highest combined sensitivity and specificity for surgically important grades greater than 70% (ECST) as indicated by the ECST trial, 60% (NASCET) as indicated by the Asymptomatic Carotid Atherosclerosis Study (ACAS) trial, and 70% (NASCET) as indicated by the NASCET trial.

Considerable progress has been made in the quality of ultrasound imaging since the 1990s. Ultrasound has undergone great advances, and the medical community has gained further expertise in performing carotid scans and interpreting the results through the widespread use of better technology, research, and continuing medical education. In addition, various accrediting bodies have been established, such as the Intersocietal Commission for Accreditation of Vascular Laboratories, the American Institute of Ultrasound in Medicine, and the American College of Radiology, in an attempt to improve and standardize the quality of vascular examinations. In an attempt to minimize the existing variability in the performed examination, expert panels from various faculties have suggested a series of guidelines on grading carotid stenosis with ultrasound (see later).

Grading of Stenosis

Grading of different degrees of stenosis that are less than 50% in relation to the normal distal ICA with velocity measurements is inaccurate because such lesions are of no hemodynamic significance. These stenoses should be reported as less than 50%, and they should be subcategorized with B-mode estimation in transverse section with the use of color flow when the plaque borders are not easily seen (see later). Color flow is important when following up patients in order to monitor plaque progression or regression. For practical surgical-decision purposes, the degree of stenosis is stratified on the basis of B-mode and Doppler velocity measurements, which results in the following strata: normal, less than 50% stenosis, 50% to 69%, 70% to 79%, 80% to 89%, more than 90% stenosis, and occlusion. Stratification can be achieved using the criteria listed in Table 1 . Figure 1 is an example of a stenosis greater than 90%.

The residual lumen (area or diameter reduction) can also be determined in transverse section using color flow for the majority of lesions producing higher grades of stenosis (see Figure 1 C), provided there is no calcification that produces acoustic shadowing. Several studies have produced sensitivities in the range of 85% to 87% and specificities in the range of 89% to 97% when compared with angiography.

The diagnosis of trickle flow (95%–99% stenosis) is made by detecting low velocities (usually peak systolic velocity of the ICA [PSVic] <30 cm/sec in the presence of a large plaque). This is because as the blood goes through a very narrow channel it loses most of its energy and fails to accelerate. Rarely, the velocities can be very low beyond the threshold of detection by the ultrasound machine and the vessel could appear occluded. These stenoses with a lumen reduction of greater than 95% are referred to as subtotal stenoses.

Color-flow Doppler is superior to velocity measurements in differentiating complete occlusions from pseudoocclusions because it offers the ability to detect low flow in difficult imaging situations, because flow can be detected without the accurate placement of a sample volume. By using color-flow Doppler, the accuracy of carotid duplex ultrasound in diagnosing total carotid occlusion is 97%, with a positive predictive value (PPV) of 96%, negative predictive value (NPV) of 98%, sensitivity of 91%, and specificity of 99%.

When occlusion is suspected based on the absence of velocity (complete absence of flow just distal to the stenosis), color-flow Doppler permits further evaluation of the proximal and distal ICA. However, in some instances evaluation of such a condition, even with the use of color flow, is hampered owing to plaque composition, exceptionally narrowed residual lumen, and possibly the patient’s clinical condition such as low blood pressure or reduced cardiac output.

A method to overcome these difficulties aiming for a 100% accuracy is to inject reflectors that are brighter than blood into the vascular system. Gas-filled microbubbles are one such reflector. This technique is currently used in some laboratories and avoids the need for angiography in patients with suspected pseudoocclusion. Alternatively, computed tomography (CT) angiography or magnetic resonance angiography (MRA) may be used to demonstrate flow in the ICA distal to the stenosis and therefore confirm ICA patency.