| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
Department of Radiology, OMESA, 166 Luis Thayer Ojeda, Oficina 706, Santiago 9, Chile; e-mail: mgarcia@omesa.cl
Editor:
It is a fact accepted in the literature and clinical practice that a correlation exists between the degree of angiographic stenosis and the Doppler ultrasonographic (US) velocity in the internal carotid artery (ICA). It is fiction, amply discussed in the article by Dr Grant and colleagues in the January 2000 issue of Radiology (1) and accepted in clinical practice, that this correlation is simple and reliable. It is also fiction, documented in the February 2000 issue of Radiology by Lee et al (2), that better instrumental measurements and/or larger samples of patients will improve the accuracy of Doppler US velocity estimation of ICA stenosis in individual patients. The physics of the Doppler US experiment tells us that the relationship between both parameters is simple but not direct, its quantification is highly unreliable, and the restrictions on its clinical implementation are theoretic, not experimental.
The US and angiographic methods for the estimation of ICA stenosis are both based on the continuity of the rate of transport of blood along the carotid vessels: What goes in must go out. The angiographic quantification of stenosis is done by using two measurements on a single vessel. For the Doppler US quantification of stenosis, two methods are used: the maximum peak systolic velocity (PSV) at a single point in the ICA and the ratio of the PSV at a point in the ICA to the PSV at a point on the common carotid artery.
If the ICA measurements are identified with the i subindex and those in the common carotid artery are identified with the c subindex, for the two angiographic locations where the subindex i1 designates the stenotic location and i2 the normal location, it is valid that dmi1/dt = dmi2/dt, where m is mass and t is time.
For a constant density, cross-sectional areas Ai1 and Ai2, and speeds vi1 and vi2, we can approximately write Ai1 · vi1 = Ai2 · vi2.
We can express the areas as equivalent circles of radii ri as follows: ri12 · vi1 = ri22 · vi2, or vi1 = vi2 · (ri2/ri1)2.
The stenosis index (SI) is defined angiographically as follows: SI = 1 - (ri1/ri2). Then, vi1 = vi2 · 1/(1 - SI)2.
In the Doppler US protocols, blood speeds are estimated with PSV measurements. For the single maximum PSV stenotic vi1 protocol, it is clear that its relationship to the stenosis index is ambiguous if the second PSV measurement vi2 in the ICA is missing. Plots of vi1 versus the stenosis index will show the typical scattergrams obtained in the North American Symptomatic Carotid Endarterectomy Trial (NASCET) study (3) because of the variability of the missing vi2 parameter.
For the second Doppler US protocol, the mass continuity equation is dmc/dt = dmi1/dt + dme/dt, where the subindex e identifies the external carotid artery. That is, our problem can have only a one- or three-vessel geometry. By introducing speeds and radii as before, we can write rc2 · vc = ri12 · vi1 + re2 · ve. For the US ratio of PSVs, we can write vi1/vc = (rc/ri1)2 - (ve/vc) · (re/ri1)2.
Finally, by introducing the stenosis index parameter, we can write vi1/vc = [rc2 - (ve/vc) · re2]/[ri22 · (1 - SI)2].
That is, to estimate the stenosis index parameter with the ICA and the common carotid artery speed measurements ratio is a hazardous job: Four additional data are missing, more than enough to introduce important ambiguities in its clinical correlations.
In summary, in Doppler US, the single PSV measurement protocol is a more robust estimator of the angiographic stenosis index parameter than the ratio of the PSV protocol. To improve its robustness, attempts should be made to obtain a second PSV measurement vi2 at a second normal location in the same ICA vessel. The second approach to the problem is to obviate the classic angiographic stenosis index standard and search for a new (angiographically defined or not) stenosis index that has a more reliable relationship with US parameters.
REFERENCES
Department of Radiology, West Los Angeles Veterans Affairs Medical Center, 11301 Wilshire Boulevard, Los Angeles, CA 90073; e-mail: egrant@ucla.edu
On the whole, we agree with the insightful analysis provided by Dr Vergara, and we certainly think that it may help to explain the lack of a simple and reliable correlation between US criteria and measured angiographic stenosis, the latter of which is considered the standard.
With the mass-balance approach suggested by Dr Vergara and by carrying the algebra a step further, we indeed see that SI = 1 - (vi2/vi1)
, by using the same notation.
With the PSV method, we agree that, for the accurate prediction of the stenosis index by using US velocities, there is a missing parameter, which is the value of vi2, the velocity in the more distal normal ICA. Soulez et al (1), in fact, recently evaluated the effect of adding a second data point taken from the distal ICA to the single PSV measurement method. They found a slight improvement in results compared with those obtained by using the routine PSV and ratio when predicting certain types of stenoses. As Dr Vergara pointed out, the problem of missing parameters and, hence, the lack of accuracy in the prediction of carotid stenosis, is even more compelling for the systolic ratio method (ratio of the PSV in the ICA to that in the ipsilateral common carotid artery). Certainly, most experienced sonologists are familiar with the dramatic effect of outflow blockage on systolic ratio measurements when both the ICA and external carotid artery are severely stenosed. On the other hand, most clinical studies in which the PSV and ratio are compared by using the receiver operating characteristic techniques reveal that the two perform equally in the classification of ICA lesions with more or less than a given degree of angiographically determined stenosis (2).
There are several other considerations not accounted for in the equations presented that may prevent major increases in accuracy. First, attempting to capture the complex biologic compensations that occur in the face of significant stenosis with a single PSV measurement is difficult at best. Clearly, as a vessel becomes narrower, velocity must increase to maintain flux. While we agree with a mass-balance approach to analysis, the true mass-balance equations stipulate that bulk flow into the carotid equals bulk flow out; this, then, must take into account the entire velocity profile, not just PSV.
If there are changes in the US waveform between the stenotic proximal ICA and the distal normal ICA (eg, a greater distal diastolic flow component in the setting of tighter stenoses), then we are dealing with a more complex phenomenon than can be captured by measuring a second PSV to provide vi2. A mathematic mass-balance analysis dictates the integration of the entire area under the waveform. To further complicate matters, the cross-sectional area A is actually also a function of time, A(t), varying with cardiac pulsations.
The correlation of Doppler velocity measurements with angiographically determined estimates of stenoses is an extremely complex problem, and the missing data points described by Dr Vergara represent several pieces of the overall puzzle that have been largely overlooked. However, we believe that the root of the problem remains in the essentially independent development of the NASCET-style measurement protocols and the US criteria for significant stenosis, with an almost retrospective attempt to fit together two disjoint measurement methods. The former is obviously based on an anatomic measurement, while the latter is physiologic, and it is unlikely that a perfect match can ever be achieved.
Finally, it is worthwhile to note that the accepted standard of angiography, which we are trying to mirror with US, may in fact not be a standard in terms of accuracy, as the recent article by Elgersma et al (3) showed. Rotational angiography shows that NASCET measurements may cause underestimation of the degree of stenosis. It may well be time to realign the methods, particularly in light of the availability of newer techniques such as CT angiography. We encourage further research to clarify whether the addition of the missing parameters will improve the accuracy of US.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  G. M. Hathout, J. R. Fink, S. M. El-Saden, and E. G. Grant Sonographic NASCET Index: A New Doppler Parameter for Assessment of Internal Carotid Artery Stenosis AJNR Am. J. Neuroradiol., January 1, 2005; 26(1): 68 - 75. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
