| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Neuroradiology |
1 From the Departments of Radiology (S.M.E., E.G.G., G.M.H., P.T.Z.), Neurology (S.N.C.), and Surgery (J.D.B.), West Los Angeles Veterans Administration Medical Center, 11301 Wilshire Blvd, Los Angeles, CA 90073. From the 1999 RSNA scientific assembly. Received October 2, 2000; revision requested November 15; revision received February 27, 2001; accepted March 30. Address correspondence to S.M.E. (e-mail: sels@mednet.ucla.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Consecutive patients with occlusion or near occlusion of ICA at catheter angiography and who underwent MR angiography and US were included. MR angiography and US were compared with catheter angiography, the standard, for the ability to help distinguish occlusion from near occlusion. Noninvasive examinations were evaluated for the ability to classify near occlusions as having severe focal stenosis with distal luminal collapse versus diffuse nonfocal disease. The 95% CIs were calculated.
RESULTS: In 55 of 274 patients with 548 ICAs, catheter angiography depicted 37 total occlusions and 21 near occlusions. US depicted all total occlusions; MR angiography depicted 34 (92%) (95% CI: 0.78, 0.98). US depicted 18 (86%) of 21 (95% CI: 0.64, 0.97) near occlusions; MR angiography depicted all (100%). Of 18 vessels that were determined to be patent at US, 17 (94%) (95% CI: 0.73, 0.99) were classified as having focal stenosis or diffuse disease. Because flow gaps were identified in vessels with focal and diffuse disease, MR angiography was not effective in helping to differentiate these lesions.
CONCLUSION: Assuming US is the initial imaging examination, when occlusion is diagnosed, MR angiography can depict it. If occlusion is confirmed, no further imaging is necessary. US performed well in helping to differentiate vessels with focal severe stenosis from those with diffuse disease. MR angiography added little in this group. Catheter angiography remains beneficial for vessels with diffuse nonfocal narrowing.
Index terms: Angiography, comparative studies, 172.12142, 172.1245, 172.12981, 172.12983 • Carotid arteries, angiography, 172.12142, 172.12143, 172.1245, 172.12981, 172.12983 • Carotid arteries, stenosis or obstruction, 172.721
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The optimum algorithm for performing imaging in patients with high-grade stenosis remains controversial, with increasing numbers of patients being referred directly for CEA on the basis of ultrasonographic (US) results alone. Magnetic resonance (MR) angiography and/or conventional angiography are often added to the work-up (5). However, patients with occlusion versus those with near occlusion (also termed pseudo-occlusion, preocclusive stenosis, or string sign) (3,6–9) represent a diagnostic dilemma in that the ability of US and/or MR angiography to help differentiate these two entities has not been definitively established. Catheter angiography is still considered by many to be the standard for helping to distinguish occlusion from near occlusion, but it carries the usual risks associated with such invasive procedures (10).
The purpose of this study was to evaluate the individual performance of US and MR angiography in the examination of patients with complete versus those with near occlusion of the ICA by using conventional angiography as the standard.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Catheter Angiography
Digital subtraction angiography was performed in all patients by using selective injections in the common carotid artery. A minimum of two views of the common carotid artery bifurcation were obtained, but often, more than two were necessary to image the proximal ICA optimally. Delayed imaging and prolonged injections were performed in all patients. Our technique consisted of an exposure rate of one image per second for up to 20 seconds and a manual injection volume of up to 20 mL of contrast material (Isovue 300 [iopamidol]; Bracco Diagnostics, Princeton, NJ). The images were reviewed by two neuroradiologists (S.M.E., G.M.H.), and the results were determined in consensus.
For this study, an occlusion was diagnosed when the ICA was found to terminate anywhere along its course. A near occlusion was diagnosed when the ICA lumen was smaller at its widest diameter than the ipsilateral internal maxillary artery (11) and/or demonstrated delayed filling relative to the ipsilateral external carotid artery. Patent vessels were further classified as having luminal atrophy or being collapsed beyond a proximal extremely high-grade stenosis or having diffuse disease or classic string signs, which had a collapsed lumen throughout their course and no obvious area of focal narrowing at or near the bifurcation.
Ultrasonography
All patients underwent gray-scale color and spectral Doppler US in a single facility accredited by the American College of Radiology. After 1994, power Doppler US was included in all examinations in which occlusion was suspected at the time of the examination. Commercially available US units (Advanced Technology Laboratories, Bothell, Wash, and Acuson, Mountain View, Calif) were used for all examinations. US examinations were performed in accordance with an established laboratory protocol.
The presence of flow was assessed by using all available techniques. Inability to identify flow in the ICA was taken to imply occlusion. Only the inframandibular portions of the ICA could be directly evaluated at US. Absent diastolic flow at spectral Doppler US in an otherwise patent cervical ICA lumen was taken to imply a possible distal occlusion (12). The color and power Doppler US appearances of the arterial lumen were observed for markedly decreased size in relation to the original size of the lumen. Additionally, evidence of focal narrowing with respect to the more distal lumen or narrowing throughout the visualized portions of the vessel was documented. Angle-adjusted spectral Doppler US samples were obtained from proscribed sites in the common carotid artery and ICA and from any areas of suspected vessel narrowing. Doppler US parameters that were evaluated included peak systolic velocity (PSV), end-diastolic velocity, and ratio of the PSV in the ICA to that in the ipsilateral distal common carotid artery. Lesions were classified as being in need of CEA (>70% stenotic) on the basis of a PSV greater than 250 cm/sec and a systolic ratio greater than 4.0 (13).
MR Angiography
All MR angiographies were performed with a 1.5-T magnet. MR angiography usually began with two-dimensional (2D) (repetition time msec/echo time msec of 25/9, 35° flip angle) and three-dimensional (3D) (30/6, 20° flip angle) time-of-flight (TOF) techniques and included gadolinium-enhanced MR angiography as of January 1998. 3D TOF MR angiography (30/6.5, 20° flip angle) was performed through the circle of Willis in all patients. Gadolinium-enhanced MR angiography was performed as of December 1997 in 21 patients (16 with near total occlusions and five with total occlusions) by using a 3D subtracted gradient-recalled echo sequence and turbo fast low-angle shot (FLASH) sequence (4/1.6, 25° flip angle, 120 x 256 matrix). Before June 1999, the total dose of gadolinium-based contrast material (ProHance; Bracco Diagnostics) was 20 mL, injected by hand and subsequently with a power injector (Spectris; Medrad, Indianola, Pa) at a rate of 3 mL/sec, following a timing bolus of 6 mL at 3 mL/sec, which was flushed with 15 mL of saline. Source images of all MR angiographic studies were evaluated in all patients.
Total occlusion was defined as a flow signal termination on all sequences at any point along the intra- or extracranial ICA without any identifiable flow signal intensity distally. In the case of near occlusion, the presence of a focal flow gap was noted and taken to imply the presence of a focal stenosis. Flow gap was defined as a finite segment of artery with no perceptible flow signal but displaying visible flow signal intensity proximal and distal to this area. Lumen was considered to be patent if flow was identified within the ICA at any of the MR angiographic sequences performed. The definition of a string sign was similar to that for angiography, as defined earlier.
Two of the authors (S.M.E., E.G.G.) compared the results of US and MR angiography with those of conventional angiography, which was used as the standard for this study. For each patient, US and MR angiographic results were evaluated together to determine if the addition of MR angiography, which was always the second imaging study performed in the work-up, improved diagnosis. Ninety-five percent CIs were calculated (StatXact-4; Cytel Software, Cambridge, Mass).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Total Occlusions
On the basis of the combination of criteria defined earlier, US correctly depicted all (100%) total occlusions. In two of the three cases of occlusion distal to the ICA origin, the diagnosis was based on the spectral pattern of low-velocity flow, with an absent diastolic component in a patient with a patent cervical ICA (Fig 1). In the third distal occlusion, flow in the ICA could not be identified in the neck, and the vessel was diagnosed as occluded despite the presence of a minute patent proximal lumen at catheter angiography.
|
|
|
Near Occlusions at US
US correctly depicted 18 (86%) of 21 near occlusions (95% CI: 0.64, 0.97); three were overestimated as completely occluded. These three ICAs were a mixed group consisting of one high-grade stenosis with distal atrophy, one string lesion, and one siphon stenosis with a collapsed proximal lumen. Of the 18 near occlusions that were correctly identified, 17 were diagnosed at color Doppler US. Power Doppler US allowed diagnosis in one additional vessel in which color Doppler US failed to depict flow. Two of the three vessels in which flow was not identified were examined with power Doppler US, which also failed to depict the residual lumen.
Catheter angiography depicted 15 of 21 near occlusions as having focal high-grade stenoses with distal collapse. US helped to correctly determine that 14 of these were patent, and one was overestimated as a total occlusion. Twelve of these 14 lesions had PSVs that were sufficiently high to allow them to be classified as surgical lesions (
70% stenosis) on the basis of the spectral Doppler US findings alone. PSVs in this group ranged between 250 and 706 cm/sec (mean, 411 cm/sec; median, 331 cm/sec). The remaining two lesions had velocities that would have been classified as middle range stenoses (50%–60%) with PSV (202 and 229 cm/sec). Other spectral Doppler US parameters were also not in the surgical range. Among the two lesions, one had color and power Doppler US findings typical of a high-grade stenosis, with areas of focal narrowing proximally with aliasing, and a very small distal lumen (Fig 2). The other lesion was incorrectly classified as a string, because no identifiable area of focal narrowing was found.
|
|
|
|
|
|
|
|
|
|
|
Two-dimensional TOF imaging depicted 18 patent arteries among the 19 vessels in which it was used. Fourteen of these vessels had focal disease at catheter angiography. In this group, 12 vessels had flow gaps and two did not. Among the three vessels with diffuse disease thought to be patent at 2D TOF imaging, two had flow gaps and one did not. The single patient with a collapsed lumen proximal to a siphon stenosis had no flow gap at the ICA origin. 3D TOF imaging depicted 14 patent arteries among the 19 vessels in which it was used. Thirteen of these vessels had focal disease at catheter angiography. In this group, 12 had flow gaps and one did not. Only one of the vessels with diffuse disease was seen at 3D TOF imaging; this vessel had a flow gap.
Gadolinium-enhanced MR angiography depicted 15 patent arteries among the 16 vessels in which it was used. Twelve of these had focal disease at catheter angiography. Seven gaps were found in this group, and the remaining five had no gaps. At review of the vessels without flow gaps, however, four clearly demonstrated a focal narrowed area at the ICA origin and would have been correctly classified as focal disease. Among the two vessels with diffuse disease, which were thought to be patent at gadolinium-enhanced MR angiography, both had flow gaps.
In general, since flow gaps were identified in vessels having both focal and diffuse disease, this finding did not appear effective in differentiating these lesions.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The ability of US to successfully help differentiate total from near total occlusions has been investigated in several recent reports (15–19). Most of these authors advocated surgery on the basis of US as the sole imaging technique. Some added the caveat that the US study must be "technically adequate" if other imaging studies are to be bypassed (15,18). This approach, however, is not universally accepted, and, on basis of our observations, cannot be recommended. In our study, US successfully depicted all angiographically identified total occlusions. On the other hand, by relying on US alone, three of 21 (14%) subtotal lesions (ie, near occlusions) would have been called occluded, at least one of which was a clear candidate for CEA.
To our knowledge, little information about the flow patterns in near occlusions is available in the US literature, but it has often been implied that they have a slow flow based on the principle that beyond a certain degree of stenosis the velocity begins to decrease (20). In our series, this was not the case in most patients with high-grade proximal lesions and distal collapse. Eleven of 13 such vessels exhibited velocity patterns that would have placed them clearly in the operative (70% stenosis) group. Even among classic string lesions, two demonstrated very high velocity. Of concern, with regard to patients with focal lesions is the small subgroup of patients with middle range or low velocities who could be excluded from surgery on the basis of spectral velocity alone. In this group, US depiction of an extremely narrowed lumen at color or power Doppler US was an important part of the diagnosis.
It is often stated that power Doppler US may be superior to color Doppler US in depicting the extremely narrowed lumen of near occlusions (21). In our experience, while power Doppler US did depict one patent lumen that was not seen with conventional color Doppler US, two lesions were still missed despite its use.
The ability of US to depict occlusions beyond the available sonographic window in the neck is also not well documented, and the existence of such pathologic findings has been essentially ignored in some of the series mentioned earlier (15). Low-velocity flow with a high resistance pattern when measured in a patent cervical ICA has been described as suggesting a distal lesion or dissection (12). Indeed, occlusions beyond the visible area of the cervical ICA were implied in two cases by virtue of this pattern. While both patients in our series with this pattern had occlusion, it is certainly possible that similar waveforms could be found in association with a severe high-grade distal stenosis, although this did not occur in our single case of distal stenosis. This case was somewhat unusual in that the proximal vessel was severely narrowed, and no flow was seen at US.
MR angiography performed well in the identification of total occlusions that occur at the ICA origin; all were correctly classified. Despite the ability of MR angiography to enable imaging from the arch to the terminal ICA bifurcation, it fell short in delineating supraclinoid disease, whether occlusive or tightly stenotic. Reasons for this are likely related to resolution and the inability of MR angiography to depict segmental occlusions within such a short segment as the supraclinoid ICA. This MR angiographic diagnosis is further complicated by the presence of collateral flow distal to the occlusion, without the ability to define the flow direction or sequence of vessel filling. The ability to define the minute lumen of such small vessels may be difficult to demonstrate with any noninvasive modality, and given current technology, still requires the fine detail and temporal information provided by means of conventional angiography.
MR angiography performed better than did US in depicting near occlusions, as it successfully depicted all 21. Furthermore, 2D TOF imaging was better at depicting near occlusion than 3D TOF imaging. As has been previously described, there is improved sensitivity to slow flow with 2D TOF imaging, when compared with 3D TOF imaging, due to saturation effects (16). Gadolinium-enhanced MR angiography was better than either of the TOF techniques for depicting near occlusion, correctly identifying all 16 cases in which it was used. The superiority of gadolinium-enhanced MR angiography is due to increased signal-to-noise ratio and decreased intravoxal dephasing (22,23). It should be noted that in one case, gadolinium-enhanced MR angiography appeared to lead to an incorrect interpretation of an angiographically diagnosed total occlusion that occurred in the midcervical ICA.
However, on the basis of a thorough review of both MR and catheter angiographic findings, the question of whether the angiogram might have eventually demonstrated flow is raised. Even optimally performed catheter angiography may have difficulty in helping to differentiate complete and complicated near occlusion (9). In our experience, in keeping with previous literature, the optimum performance of all MR angiographic techniques is heavily dependent on the review of source images and the inclusion of MR angiography of the circle of Willis (24,25). In several cases, only on the basis of source images could the lumen of the ICA be followed from the cervical region to the skull base.
We classified vessels with near occlusions into those with proximal high-grade stenosis and distal luminal collapse and those with diffuse luminal narrowing and no visible proximal stenosis. Diffuse carotid disease has been categorized in the older angiography literature as resulting from such entities as dissection, postradiation change, subacute partial thrombosis, and chronic subtotal thrombosis (9). In our series, five such lesions were identified. Additionally, two stringlike lesions were found in patients with distal occlusions, and one was seen in a patient with a tight stenosis in the area of the siphon. Regardless of whether these lesions were occlusive or nonocclusive, none of the patients was a candidate for CEA, as the disease was not located in or confined to a surgically accessible region. Such information is obviously vital to treatment planning. In these patients, treatment options may include reconstruction of the artery by means of stent placement, angioplasty, or surgical ligation of the vessel, or medical management may be the only option. In fact, attempts at CEA in these patients may be detrimental (14). While surgical exploration of all cases would have provided a more accurate standard for this and similar studies, in practice, diagnosis must be made prior to patient treatment decisions, including surgery. This is particularly true in patients with lesions located beyond a viable surgical field, in whom angiography may still be the only standard.
On the basis of the results of our study, several conclusions can be drawn about the use of noninvasive examinations versus angiography in patients with total and near total occlusions. Assuming US to be the initial imaging technique, if a complete occlusion is suspected at or near the bifurcation, MR angiography should be used to confirm its presence. The MR angiographic study should include imaging of the circle of Willis. If occlusion is confirmed, no further imaging is necessary. Vessels found to have focal high-grade stenosis and suspected distal collapse at US can be dealt with in a fashion similar to that for any other high-grade lesion. This may vary somewhat, depending on the philosophy of the referring physician and the institution. However, in the present study, MR angiography added little in this group due to its limited ability to help differentiate between vessels with focal very high stenoses and those with diffuse disease.
Finally, a small group of vessels that have true string lesions remain. These may be found at MR angiography in patients thought to have occlusion at US or culled from the general population of near occlusions when US fails to demonstrate focal disease in an otherwise very narrow ICA. Patients with low-velocity high-resistance flow in the ICA would also fall into this category, regardless of the appearance of the bifurcation, due to the implication that severe distal disease is present. This group of patients with diffusely narrowed ICA lumina, which comprised eight (13.7%) of 58 vessels, often have unusual lesions, and informed therapeutic decisions in this unique population may still be best made by using conventional angiography.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, S.M.E., E.G.G.; study concepts, S.M.E., E.G.G., S.N.C., J.D.B.; study design, S.M.E., E.G.G.; literature research, E.G.G., S.M.E., G.M.H.; clinical studies, S.M.E., E.G.G., G.M.H., P.T.Z.; data acquisition, S.M.E., E.G.G., G.M.H.; data analysis/interpretation, S.M.E., E.G.G., G.M.H., P.T.Z.; statistical analysis, S.M.E., E.G.G.; manuscript preparation, S.M.E., E.G.G.; manuscript definition of intellectual content, all authors; manuscript editing, S.M.E., E.G.G.; manuscript revision/review and final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. J. Alexander, J. Moawad, and D. Super Outcome Analysis of Carotid Artery Occlusion Vascular and Endovascular Surgery, November 1, 2007; 41(5): 409 - 416. [Abstract] [PDF]
|
|
|
|
 |
|
 L. Saba, G. Caddeo, R. Sanfilippo, R. Montisci, and G. Mallarini CT and Ultrasound in the Study of Ulcerated Carotid Plaque Compared with Surgical Results: Potentialities and Advantages of Multidetector Row CT Angiography AJNR Am. J. Neuroradiol., June 1, 2007; 28(6): 1061 - 1066. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Thanvi and T. Robinson Complete occlusion of extracranial internal carotid artery: clinical features, pathophysiology, diagnosis and management Postgrad. Med. J., February 1, 2007; 83(976): 95 - 99. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A.J. Thomas, R. Gupta, A.H. Tayal, M.B. Horowitz, and T.G. Jovin Stenting and Angioplasty of the Symptomatic Chronically Occluded Carotid Artery AJNR Am. J. Neuroradiol., January 1, 2007; 28(1): 168 - 171. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Ohm, P. J. Bendick, J. Monash, P. G. Bove, O. W. Brown, G. W. Long, G. B. Zelenock, and C. J. Shanley Diagnosis of Total Internal Carotid Occlusions with Duplex Ultrasound and Ultrasound Contrast Vascular and Endovascular Surgery, May 1, 2005; 39(3): 237 - 243. [Abstract] [PDF]
|
|
|
|
 |
|
 E. G. Grant, C. B. Benson, G. L. Moneta, A. V. Alexandrov, J. D. Baker, E. I. Bluth, B. A. Carroll, M. Eliasziw, J. Gocke, B. S. Hertzberg, et al. Carotid Artery Stenosis: Gray-Scale and Doppler US Diagnosis--Society of Radiologists in Ultrasound Consensus Conference Radiology, November 1, 2003; 229(2): 340 - 346. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M Hathout, M. J. Duh, and S. M. El-Saden Accuracy of Contrast-Enhanced MR Angiography in Predicting Angiographic Stenosis of the Internal Carotid Artery: Linear Regression Analysis AJNR Am. J. Neuroradiol., October 1, 2003; 24(9): 1747 - 1756. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Kriegshauser, M. D. Patel, and K. D. Nelson Carotid Pseudostring Sign From Vasa Vasorum Collaterals J. Ultrasound Med., September 1, 2003; 22(9): 959 - 963. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Lev, J. M. Romero, D. N.F. Goodman, R. Bagga, H. Y. K. Kim, N. A. Clerk, R. H. Ackerman, and R. G. Gonzalez Total Occlusion versus Hairline Residual Lumen of the Internal Carotid Arteries: Accuracy of Single Section Helical CT Angiography AJNR Am. J. Neuroradiol., June 1, 2003; 24(6): 1123 - 1129. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
