(Stroke. 1997;28:1981-1987.)
© 1997 American Heart Association, Inc.
Articles |
From the Department of Neurology, Klinikum Mannheim, University of Heidelberg (Germany). Dr Artemis was a visiting fellow from the Department of Neurology, AHEPA Hospital, Thessaloniki, Greece.
Correspondence to PD Dr W. Steinke, Department of Neurology, Marien-Hospital Düsseldorf, Rochusstr 2, 40479 Düsseldorf, Germany.
| Abstract |
|---|
|
|
|---|
Methods One hundred patients with a total of 128 ICA stenoses (50% to 69%, n=37; 70% to 79%, n=27; 80% to 99%, n=64) and 12 ICA occlusions were consecutively investigated by means of PDI, CDFI, and intra-arterial angiography (n=48). Reduction of the intrastenotic lumen was measured on longitudinal and transverse views of PDI and CDFI for the calculation of the degree of diameter and area stenosis, respectively. Angiographic stenosis was determined with the use of the North American Symptomatic Carotid Endarterectomy Trial (NASCET), European Carotid Surgery Trial (ECST), and common carotid (CC) methods.
Results PDI provided significantly more excellent or good (92% versus 79%; P<.01) displays of the intrastenotic lumen than CDFI, particularly in complicated high-grade stenosis. While linear regression analysis demonstrated a high overall correlation between PDI and CDFI for diameter (r=.88; P<.001) and area stenosis (r=.79; P<.001), categorization of ICA stenosis revealed best agreement for 80% to 99% area stenoses. Since angiography frequently either underclassified (NASCET method) or overclassified (ECST, CC methods) the degree of ICA stenosis in comparison to both PDI and CDFI, the sonographic-angiographic correlation was only moderate (regression coefficients ranged from .62 to .70; P<.001).
Conclusions PDI further improves the assessment of ICA stenosis by providing better visualization of the stenotic vascular lumen than CDFI. Sonographic imaging of the stenotic plaque on both PDI and CDFI provided a direct measurement of the local degree of stenosis, while the angiographic grade of stenosis essentially depended on the method used for evaluation.
Key Words: angiography carotid artery diseases ultrasonics
| Introduction |
|---|
|
|
|---|
For more than a decade, Doppler sonography and conventional duplex scanning using distinct parameters of the Doppler frequency spectrum have been well accepted as reliable diagnostic tools for the detection and classification of ICA stenosis.13 14 15 16 17 18 19 20 21 More recently, CDFI, which provides a color-coded visualization of the velocity and flow direction of moving blood cells superimposed on a gray-scale image of vascular structures, has become the leading routine method in most neurovascular departments because CDFI improves the diagnostic accuracy and reproducibility of sonographic assessment of carotid stenosis.6 22 23 24 25 26 27 28 29 However, plaque calcification remained a significant limitation for the visualization of the intrastenotic diameter and area reduction.25 29 In contrast to CDFI, PDI, the most recently introduced sonographic technique, generates intravascular color signals based on the reflected echo amplitude, which essentially depends on the amount of red blood cells within the sample volume. Thus, the color display on PDI is independent from the angle of insonation and, more importantly, from the velocity and direction of moving blood cells. In addition, because of the different technique of image generation, an increased sensitivity to display the continuity of the stenotic lumen and improved definition of the intravascular surface are claimed for PDI. Since experience in the assessment of carotid stenosis by PDI is limited to date,30 31 this study was performed to systematically evaluate the display quality of PDI in comparison to CDFI in patients with symptomatic and asymptomatic ICA stenoses. The second objective was to correlate measurements of ICA stenosis on PDI and CDFI with the three different angiographic techniques to investigate the clinically important intermethod relationship.
| Subjects and Methods |
|---|
|
|
|---|
PDI and CDFI examinations were performed by three experienced examiners (W.S., S.R., N.A.) using an Acuson 128 XP ultrasound device with a 7.0-MHz linear array transducer. PDI and CDFI were performed without knowledge of the degree and side of ICA stenosis and of the presence of ICA occlusion on CW Doppler sonography. The technical principles of CDFI have been reported elsewhere.32 33 The crucial difference between PDI and CDFI is that PDI uses the amplitude of the echo signal reflected from red blood cells and does not process information of frequency or phase shift for the generation of intravascular color signals. Thus, the visualization of blood flow essentially does not depend on the velocity and direction of moving blood cells and is free from color aliasing phenomenon. In addition, special filter systems further improve the blood-tissue discrimination, resulting in a relatively homogeneous angiography-like visualization of the vascular lumen surface superimposed on a conventional B-mode display of tissue and vessel structures.34
Sequential parallel longitudinal views and perpendicular cross sections
were displayed on PDI and CDFI. On-line measurements of the vessel wall
and the minimal residual lumen of the ICA stenosis contrasted
by color signals of either method on longitudinal and transverse
displays were used to calculate the degree of stenosis as
percentage of the diameter and area reduction, respectively (Fig 1
). Measurement of the diameter reduction
on longitudinal views was consistently performed at the
narrowest site of the stenosis perpendicularly to the long axis
of the vessel regardless of the angle to the intrastenotic flow
direction. Measurements were performed at exactly the same site of
stenosis by directly switching from CDFI to PDI display.
Pulsed-wave Doppler spectra were not recorded with the
integrated Doppler system to avoid bias for the measurement of ICA
stenosis. The display quality of both sonographic techniques
was classified as excellent, good, poor, and not visualized depending
on the continuity and intensity of intravascular color signals in the
prestenotic, intrastenotic, and poststenotic
vessel segments and on the delineation of the plaque surface and the
residual stenotic lumen by color contrast.
|
Intra-arterial selective biplane digital subtraction
angiography was conducted according to clinical indications in 43
symptomatic patients (48 ICA stenoses). In all
cases angiography followed the sonographic examinations (mean interval,
4.8 days; range, 1 to 17 days). The neuroradiologist who evaluated the
angiograms was unaware of the measurement of carotid stenosis
on PDI and CDFI. Calculation of the degree of diameter stenosis
was performed with the use of the NASCET and ECST criteria, as well as
the CC method.1 2 10 For statistical evaluation of the
correlation between measurements of stenosis by different
sonographic and angiographic methods, simple regression
analysis was used. Frequencies were analyzed with the
use of the
2 test.
| Results |
|---|
|
|
|---|
2 [df=1]=9.13, P<.01)
(Fig 3
2
[df=1]=2.33).
|
|
Measurement of ICA Stenosis
PDI Versus CDFI
Linear regression analysis demonstrated a high
correlation between PDI and CDFI for the measurement of both diameter
stenosis (r=.88, P<.001) and area
stenosis (r=.79, P<.001). After
classification of the ICA stenoses into distinct categories
according to the degree of stenosis, overall agreement between
both methods was better for area stenosis (75.0%) than for
diameter stenosis (67.5%) (Table 1
). While the rate of stenoses
classified into a higher category by PDI compared with CDFI did not
differ between diameter (9.4%) and area stenoses (9.8%),
underclassification by PDI occurred more frequently in diameter
(23.1%) (Fig 1
) than in area stenoses (15.2%). This was
particularly evident in diameter measurements of high-grade
stenoses: 44.1% of the 80% to 99% ICA stenoses
according to CDFI were classified into the 70% to 79% category by PDI
(Table 1
).
|
PDI and CDFI Versus CW Doppler
Correlation of CW Doppler with the measurement of diameter and
area reduction on PDI and CDFI was significant for both imaging methods
(P<.001). On CDFI, measurement of ICA stenosis on
longitudinal views correlated slightly better (r=.74) than
measurement on transverse views (r=.70), while correlation
with CW Doppler sonography was not different for diameter and area
stenosis on PDI (r=.75) (Fig 4
). Further evaluation demonstrated that
both PDI and CDFI had a relevant tendency to underestimate the degree
of stenosis in comparison to CW Doppler.
|
PDI and CDFI Versus Angiography
Measurement of carotid stenosis between both sonographic
techniques and the three angiographic methods correlated significantly
(P<.001). Coefficients of correlation demonstrated only
minor differences between PDI and CDFI (Fig 5
). They were highest for PDI versus ECST
(r=.70) and CC (r=.69) results, while correlation
of PDI versus NASCET (r=.63) and of CDFI versus NASCET
(r=.64), ECST (r=.64), and CC (r=.62)
results was slightly worse.
|
Classification of ICA diameter stenosis by PDI, CDFI, and
angiography revealed only a moderate overall agreement between
sonographic and angiographic methods (Tables 2 through 4![]()
![]()
).
Mismatch
was essentially due to a significant rate of overclassification of ICA
stenosis measured according to the ECST and CC methods in
comparison to both PDI (30.9% and 33.3%) and CDFI (28.9% and 42.1%)
(Tables 3
and 4
). In contrast, evaluation of angiography with the
NASCET method resulted in a significant underclassification of
stenosis compared with PDI (38.1%) and CDFI (39.6%),
respectively (Table 2
). However, if all the 70% to 79%
stenoses and 80% to 99% stenoses were grouped
together into one category, the overall agreement increased
significantly and was highest for PDI versus CC (95.2%). Angiography
performed in 7 of 12 patients with the sonographic diagnosis of ICA
occlusion confirmed the results of PDI and CDFI in all cases.
Furthermore, none of the ICA stenoses assessed by sonography
were occluded on the angiograms.
|
|
|
The comparative analysis of measurement of the individual ICA
stenosis by PDI and angiography revealed that application of
the NASCET method resulted in a lower degree of stenosis than
the ECST method in 95% of the angiograms (Fig 6
). In those cases in which NASCET
significantly underestimated the ICA stenosis compared with
PDI, ECST was regularly closer to the sonographic results. Conversely,
when the angiographic degree of stenosis was higher than the
PDI measurement, NASCET results correlated slightly better with PDI
than ECST (Fig 6
). However, angiographic overestimation of
stenosis was commonly less significant.
|
| Discussion |
|---|
|
|
|---|
Different from CDFI, PDI provides relatively homogeneous
color signals even in high-grade stenosis and generates an
angiography-like visualization of the vascular lumen surface (Figs 1
and 3
). Since PDI thus facilitates the visual perception of the
configuration of the ICA stenosis, angiograms will be less
frequently required before carotid endarterectomy.
In this respect, the greater acceptance of modern sonographic
techniques instead of angiography for the preoperative assessment and
visualization of ICA stenosis is particularly important because
of the well-known risks of intra-arterial
angiography.11 12 However, although we included a
considerable number of very high-grade ICA stenoses and
contralateral occlusions, the diagnostic significance of
PDI for the assessment of near occlusion in comparison to CDFI and
angiography could not be determined sufficiently from the present
data.
Independence from the angle of insonation and absence of aliasing phenomena are other advantages of PDI compared with CDFI.30 31 However, since the color signals of PDI do not provide any information about the velocity and direction of moving blood cells, the site of maximal stenotic flow velocity is not indicated, and areas of abnormal hemodynamics cannot be identified directly. Therefore, PDI and CDFI are complimentary sonographic methods and should be used together for optimal diagnostic results.
In the present study measurement of the diameter stenosis correlated significantly between PDI and CDFI when we used linear regression analysis. However, classification of ICA stenosis into distinct categories of severity revealed highest agreement for area reduction in high-grade (80% to 99%) ICA stenosis. This high concordance of both sonographic imaging techniques further supports the view of a recent study of PDI and CDFI in ICA stenosis that measurements of lumen reduction on transverse views comes probably closest to the "true anatomic degree of stenosis." Accordingly, Alexandrov et al6 reported that color-assisted duplex sonography correlated significantly with planimetric measurements of cross sections of surgical specimens from endarterectomy of carotid stenosis. A corresponding conclusion was consequently drawn in an international consensus meeting on the quantification of ICA stenosis that suggested measurement of the area ratio on transverse views of CDFI as a valid criterion for the assessment of the degree of stenosis.4
Measurement of diameter stenosis performed on PDI and CDFI correlated only moderately with angiography in this study regardless of the method used for evaluation of the angiograms. However, among all sonographic-angiographic correlations, agreement was best between PDI and ECST and CC methods, respectively. Griewing et al31 also found a higher correlation of PDI and angiography evaluated according to the ECST method in comparison to the correlation between CDFI and angiography. However, the significantly better performance of PDI in the study by Griewing et al was at least partially due to unusual discrepancies between measurements on CDFI and angiography.31
The cause of only moderate accordance between sonographic imaging and angiographic methods in our study became more evident after classification of the ICA stenoses into distinct categories of diameter reduction. Both ECST and CC methods frequently overestimated the stenoses, whereas the NASCET method led to significant underestimation compared with both PDI and CDFI. Corresponding findings were reported by Rothwell et al,10 who demonstrated that ECST and CC methods classified twice as many stenoses as severe (70% to 99%) than did the NASCET method in a comparison of the three angiographic methods for the measurement of carotid stenosis on 1001 angiograms. Conversely, more than 60% of the 30% to 69% stenoses according to the ECST and CC methods were assigned to the 0% to 29% category by the NASCET method.10 With regard to these findings, which confirmed previous analyses,5 6 9 and the results of the present study, it may be postulated that modern sonographic techniques such as PDI and CDFI provide the best approximation of the degree of carotid stenosis because of the capability to visualize directly the intrastenotic residual lumen diameter and the distance between vessel walls at the same site. In contrast, calculation of the angiographic degree of ICA stenosis uses denominators that frequently are either too small (NASCET), too wide (CC), or roughly estimated (ECST). Unfortunately, CDFI and PDI were not available when the large endarterectomy trials started. Retrospective analysis of duplex scanning with Doppler spectrum analysis in selected NASCET and ECST patients did not provide adequate information to judge the state of the art of sonographic imaging.35 36
However, there are still some limitations of sonographic measurement of carotid stenosis. Although the display of the intrastenotic continuity of blood flow is improved by PDI, severe plaque calcification remains a major obstacle for adequate sonographic visualization of carotid stenosis in a few cases. It may also be difficult to exactly outline the vessel wall on transverse views at the narrowest site of the stenosis, potentially resulting in some error of measurement of area stenosis. In addition, difficulties in determining the long axis of the vessel at the site of stenosis, which is important for perpendicular measurement of diameter reduction, may represent a source of underestimation of the degree of stenosis depending on the geometry of the carotid bifurcation and the configuration of the stenotic plaque.
In conclusion, although the present study revealed superiority of PDI over CDFI in providing color contrast of the stenotic vascular lumen, in particular in high-grade calcified ICA stenosis, agreement with angiography was only slightly better for PDI when correlated with the ECST and CC methods. In comparison to both PDI and CDFI, which directly visualize the local diameter and area lumen reduction, angiographic classification either underestimated (NASCET method) or overestimated (ECST and CC methods) the degree of ICA stenosis, thus indicating that sonographic imaging probably provides the best approximation of the actual grade of stenosis. However, since the present study did not investigate the benefit of carotid endarterectomy with regard to the sonographic degree of stenosis, the potential role of PDI and CDFI in this respect needs further confirmation.
| Selected Abbreviations and Acronyms |
|---|
|
| Footnotes |
|---|
Received February 21, 1997; revision received May 12, 1997; accepted June 5, 1997.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
M. Kamouchi, K. Kishikawa, Y. Okada, T. Inoue, S. Ibayashi, and M. Iida Reappraisal of Flow Velocity Ratio in Common Carotid Artery to Predict Hemodynamic Change in Carotid Stenosis AJNR Am. J. Neuroradiol., April 1, 2005; 26(4): 957 - 962. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. H. Wierks and N. Labropoulos Noninvasive Carotid Imaging Perspectives in Vascular Surgery and Endovascular Therapy, June 1, 2004; 16(2): 89 - 99. [Abstract] [PDF] |
||||
![]() |
R. A. Bucek, M. Reiter, A. Dirisamer, M. Haumer, A. Fritz, E. Minar, and J. Lammer Three-Dimensional Color Doppler Sonography in Carotid Artery Stenosis AJNR Am. J. Neuroradiol., August 1, 2003; 24(7): 1294 - 1299. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. J. Nederkoorn, Y. van der Graaf, and M.G. M. Hunink Duplex Ultrasound and Magnetic Resonance Angiography Compared With Digital Subtraction Angiography in Carotid Artery Stenosis: A Systematic Review Stroke, May 1, 2003; 34(5): 1324 - 1331. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. M. Romero, M. H. Lev, S.-T. Chan, M. M. Connelly, R. C. Curiel, A. E. Jackson, R. G. Gonzalez, and R. H. Ackerman US of Neurovascular Occlusive Disease: Interpretive Pearls and Pitfalls RadioGraphics, September 1, 2002; 22(5): 1165 - 1176. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Umemura and K. Yamada B-Mode Flow Imaging of the Carotid Artery Stroke, September 1, 2001; 32(9): 2055 - 2057. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Szabo, R. Kern, A. Gass, J. Hirsch, and M. Hennerici Acute Stroke Patterns in Patients With Internal Carotid Artery Disease : A Diffusion-Weighted Magnetic Resonance Imaging Study Stroke, June 1, 2001; 32(6): 1323 - 1329. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Koga, K. Kimura, K. Minematsu, and T. Yamaguchi Diagnosis of Internal Carotid Artery Stenosis Greater than 70% with Power Doppler Duplex Sonography AJNR Am. J. Neuroradiol., February 1, 2001; 22(2): 413 - 417. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. B. Anderson, R. Ashforth, D. E. Steinke, R. Ferdinandy, and J. M. Findlay CT Angiography for the Detection and Characterization of Carotid Artery Bifurcation Disease Stroke, September 1, 2000; 31(9): 2168 - 2174. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Devuyst, N. Cals, V. de Borchgrave, P. Bara, and M. Vandooren Advantages of Transcranial Power Duplex Imaging After Contrast Injection to Detect Low Flow in a Moyamoya Syndrome Stroke, April 1, 1999; 30(4): 897 - 899. [Full Text] [PDF] |
||||
![]() |
C. Arning, W. Steinke, M. Hennerici, S. Ries, A. Schwartz, and N. Artemis Evaluation of Carotid Artery Stenosis by Power Doppler Imaging • Response Stroke, October 1, 1998; 29(10): 2211 - 2213. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1997 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |