Background and Purpose Results from large multicenter studies have shown that carotid endarterectomy, performed with low perioperative morbidity and mortality, is beneficial for patients with symptomatic carotid stenosis ≥70% as calculated according to strict angiographic criteria. To apply these results in clinical practice, individual institutions should determine whether locally implemented duplex ultrasonography adequately identifies patients with ≥70% stenosis and whether the degree of stenosis reported by local angiographers correlates with strict angiographic measurements.
Methods We compared estimates of carotid stenosis obtained by duplex ultrasonography and the radiologists’ reports from conventional cerebral angiography with each other and with results obtained using North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria.
Results One hundred seventy-one vessels were available for review. In 155 (91%) of the cases, the reports from the ultrasound and angiogram were in agreement with regard to whether the stenosis was ≥70% or <70%. In 11 of the 16 cases where there was a disparity between the studies, the ultrasound was in closer agreement with measurements obtained using NASCET criteria. Nine of the angiography reports overestimated the degree of stenosis compared with NASCET measurements; twice angiography underestimated the stenosis. Twice the ultrasound underestimated the stenosis, and three times it overestimated the stenosis.
Conclusions Duplex ultrasonography was highly sensitive for detecting significant carotid stenosis at our institution; however, angiography reports often graded the degree of stenosis to be more severe than measurements obtained using NASCET criteria. Institutions that evaluate patients for carotid endarterectomy should investigate the correlation between their ultrasound and angiographic studies so that the results of carotid endarterectomy trials can be accurately applied.
Three recent randomized clinical trials have clearly established the benefit of carotid endarterectomy for patients with symptomatic high-grade stenosis.1 2 3 The most influential of these studies for North American physicians was NASCET. In NASCET, symptomatic patients with ipsilateral carotid stenosis of 70% to 99% who were randomized to carotid endarterectomy and medical management had significantly better outcomes than those randomized to medical management alone. The NASCET protocol requires strict and specific criteria for measuring carotid stenosis. NASCET stenosis is calculated from the ratio of the linear luminal diameter of the narrowest segment of the diseased portion of the artery to the diameter of the artery beyond any poststenotic dilatation.1 Since publication of this study, it has become standard practice to recommend carotid endarterectomy for patients who are good surgical candidates with symptomatic carotid stenosis of 70% or greater (as measured by NASCET criteria).4 It is therefore necessary to identify the subset of patients with ipsilateral carotid stenosis of ≥70% from the large population of patients who present with hemispheric transient ischemic attacks or minor strokes. While this appears to be a relatively straightforward task, it raises some practical questions: (1) If carotid ultrasound is used for screening patients, how often will a 70% or greater stenosis be overlooked? and (2) How does the degree of stenosis reported in a typical angiographic study compare with the degree of stenosis determined using NASCET criteria?
Previous studies suggest that an excellent correlation between carotid ultrasound and cerebral angiography can be obtained in some institutions5 6 7 8 9 10 11 ; however, other centers have found poor agreement between these two techniques.2 12 Carotid stenosis values determined with ultrasound may differ from the values obtained using conventional angiography for several reasons. These include dependence on the skill of the individuals performing the studies and the fact that ultrasonographic measurements estimate the reduction in cross-sectional area, whereas angiographic measurements are based on differences in vessel diameter. Furthermore, the degree of stenosis reported by conventional angiography may also differ substantially from values obtained with NASCET criteria because NASCET criteria require strict reproducible measurements rather than the “eyeballing” method used by many radiologists.13
Therefore, if individual centers are to apply the results of the NASCET study in clinical practice, it is important that they investigate the correlation between their ultrasonographic and angiographic measurements of carotid stenosis. In addition, the degree of stenosis reported by both methods should be compared with measurements obtained with NASCET criteria. To investigate these issues at our institution, we performed a study comparing ultrasound and angiographic results.>
Subjects and Methods
All patients who had undergone both carotid duplex ultrasound and cerebral angiography within 2 months between December 1992 and June 1993 at Stanford University Hospital were evaluated for entry into the study. Also included as a separate sample were patients who had undergone both procedures within a 6-month period at the SUH-affiliated Palo Alto Veterans Affairs Medical Center. Patients were excluded if they had undergone carotid endarterectomy surgery or carotid angioplasty in the time period between the two studies. Patients for whom radiology reports were missing or incomplete and patients whose radiological films could not be located for review were also excluded. In patients who had had a surgical procedure on only one vessel, the opposite vessel was eligible for inclusion in the study.
For each carotid artery, the ultrasound and angiography reports were reviewed, and the degree of stenosis documented in each report was recorded. If a range was given for the degree of stenosis, the average of the two values was used. If a descriptive term rather than a numerical value was given, the following criteria were used: severe, >90%; significant, 70% to 89%; moderate, 50% to 69%; and mild, <50%. Data were analyzed to investigate the statistical correlation between the two techniques when results were categorized as total occlusions, 70% to 99% stenosis, or <70% stenosis, and also when the results were categorized as severe (≥90%), significant (70% to 89%), moderate (50% to 69%), or mild (<50%) stenosis. For all cases in which there was a disparity between the angiographic and ultrasound measurements, the patients’ medical records and ultrasound and angiographic films were reviewed. The degree of stenosis was determined from the angiographic films with NASCET criteria by an evaluator who was blinded to the ultrasound and angiography reports (G.W.A.).
During the review of each angiographic report, any significant distal lesions (carotid siphon or intracranial) that were detected ipsilateral to a carotid stenosis were recorded. In addition, any additional information regarding vessel anatomy that was mentioned in the angiographic report (dissection, intracranial stenosis, fibromuscular dysplasia, webs or bands) was also recorded. The same ultrasound criteria were used for determining the degree of stenosis at both SUH and the PAVAMC; however, the vascular technicians and physicians interpreting the studies differed. In general, the degree of stenosis was determined to be <50% if the PSV was <150 cm/s, EDV was <50 cm/s, and the ratio of the peak systolic velocities in the ICA compared with the CCA (ICA/CCA) was <1.8. A 50% to 70% stenosis was assigned if the PSV was 150 to 250 cm/s, EDV was 50 to 90 cm/s, and ICA/CCA was 1.8 to 2.8. A 70% to 90% stenosis was assigned if the PSV was 250 to 400 cm/s, EDV was 90 to 150 cm/s, and ICA/CCA was 2.8 to 5.0. A 90% to 99% stenosis was assigned if the PSV was >400 cm/s, EDV was >150 cm/s, and ICA/CCA was >5.0.
Fifty-two patients from SUH and 40 patients from the PAVAMC met the inclusion criteria, yielding 97 and 74 vessels, respectively, for review (Table 1⇓). The mean time between ultrasound and angiography was 8 days for SUH patients and 35 days for PAVAMC patients. The time between procedures ranged from the same day to 56 days at SUH and from 1 day to 165 days for patients at the PAVAMC. Eighty-three percent of patients at SUH and 46% of patients at the PAVAMC had both procedures performed within a 2-week period.
For the SUH patients, in 90 of the 97 vessels examined (93%), the ultrasound and angiographic report were in agreement as to whether the patient’s degree of carotid stenosis was ≥70% (Table 2⇓). In five of the seven vessels where discrepancies were noted, the stenosis was rated <70% by ultrasound, while the angiographic report indicated a ≥70% stenosis (Table 3⇓). When NASCET criteria was used to grade these vessels, in four of these five cases the ultrasound report correlated more closely with the NASCET measurement (Table 3⇓).
In the remaining two cases where there was a disparity between the ultrasound and angiography reports, the ultrasound estimates were ≥70%, while the angiography estimate was <70% (Table 2⇑). NASCET criteria measurement agreed more closely with the ultrasound in one case and the angiography report in the other, although differences were relatively small (Table 3⇑).
For the PAVAMC patients, in 65 of the 74 vessels studied (88%), the ultrasound and angiography reports were in agreement as to whether the carotid stenosis was ≥70% (Table 1⇑). Ultrasound showed the vessel to have a nonsignificant (<70%) stenosis in six vessels in which the angiography estimate was ≥70% (Table 4⇓). In five of these six cases, NASCET measurements correlated more closely with the ultrasound estimate (Table 5⇓). There were also three vessels with ≥70% stenosis on ultrasound and angiography reports indicating a <70% stenosis. For these vessels, NASCET criteria was in closer agreement with the angiography estimate in two and the ultrasound in one.
In one SUH patient, the angiogram revealed a complete occlusion of the left internal carotid artery. However, an ultrasound performed 8 days earlier had indicated a 70% stenosis of the vessel. A repeated ultrasound obtained approximately 2 weeks later reconfirmed that the ICA was occluded, and the ultrasonographer suggested that the original ultrasound may have mistaken a large occipital branch of the external carotid artery for the ICA.
Ultrasound results indicated a carotid occlusion in 18 vessels (8 SUH, 10 PAVAMC). However, in three of these cases (all from PAVAMC), angiography revealed that the vessels were actually patent with very high-grade stenoses. In two other cases, angiography noted severe intracranial stenoses ipsilateral to significant carotid stenoses. In addition, angiography was able to document a carotid dissection in one patient that was reported only as a stenosis by ultrasound.
A high level of statistical correlation between the ultrasound and angiography reports was found when results were graded as severe, significant, moderate, and mild stenoses (Spearman’s correlation coefficient R=.92, P=.0001 [SUH], and R=.80, P=.0001 [PAVAMC]).
The results of this study indicate that at our institution duplex ultrasonography is an excellent screening tool for detecting clinically significant carotid stenosis. In fact, in only two of the vessels that we evaluated was a carotid stenosis that was graded >70% by NASCET criteria not detected by carotid ultrasound. In these two cases, the ultrasound did note moderate stenosis (50% and 65%) that would likely lead to either a follow-up ultrasound or angiogram. Therefore, it appears that potential “surgical lesions” are rarely missed by ultrasound at our institution. In addition, in only three of 171 vessels did carotid ultrasound report a significant stenosis (>70%) in vessels found to have <70% stenosis by NASCET criteria.
Another important finding of the study is that the degree of carotid stenosis reported on our angiography reports does not necessarily correlate closely with the values obtained using NASCET criteria. As suggested by others,14 the strict criteria used by NASCET may produce lower stenosis values than are typically determined by other methods, particularly with lesions in the moderate-to-severe range. Therefore, if the results of the NASCET study are to be translated into clinical practice, angiographic films should be assessed using NASCET criteria rather than routinely accepting the degree of stenosis as reported by the angiographer.
This study was not designed to address whether ultrasound alone is an adequate screening test before carotid endarterectomy, although we noted that in the 92 patients studied, ultrasound missed one dissection and two severe ipsilateral intracranial stenoses, one occlusion was read as a 70% stenosis, and three vessels with very high-grade stenosis were considered to be occluded. Important data that would likely influence surgical management could have been missed if angiography had not been performed in these seven cases. It is of note, however, that only three patients with a NASCET stenosis of <70% would have been subjected to surgery if angiographic confirmation had not been performed to verify the degree of stenosis.
We encourage other centers to perform similar comparisons between their ultrasound and angiographic results so the conclusions of clinical trials can be accurately applied to clinical practice.
Selected Abbreviations and Acronyms
|CCA||=||common carotid artery|
|ICA||=||internal carotid artery|
|NASCET||=||North American Symptomatic Carotid Endarterectomy Trial|
|PAVAMC||=||Palo Alto Veterans Affairs Medical Center|
|PSV||=||peak systolic velocity|
|SUH||=||Stanford University Hospital|
The authors thank Phyllis Grant for preparation of the manuscript and Paige Bracci for statistical advice.
- Received March 8, 1995.
- Revision received April 24, 1995.
- Accepted May 9, 1995.
- Copyright © 1995 by American Heart Association
Moore W, Barnett H, Beebe H, Bernstein E, Brener B, Brott T, Caplan L, Day A, Goldstone J, Hobson R, Kempczinski R, Matchar D, Mayberg M, Nicolaides A, Norris J, Ricotta J, Robertson J, Rutherford R, Thomas D, Toole J, Trout H, Wiebers D. Guidelines for carotid endarterectomy: a multidisciplinary consensus statement from the Ad Hoc Committee, American Heart Association. Stroke. 1995;26:188-201.
Moneta GL, Edwards JM, Chitwood RW, Taylor LM, Lee RW, Cummings CA, Porter JM. Correlation of North American Symptomatic Carotid Endarterectomy Trial (NASCET) angiographic definition of 70% to 99% internal carotid artery stenosis with duplex scanning. J Vasc Surg. 1993;17:152-159.
Tsuruda JS, Saloner D, Anderson C. Noninvasive evaluation of cerebral ischemia: trends for the 1990s. Circulation. 1991;83:176-189.
Mittle RL, Broderick M, Carpenter JP, Goldberg HI, Listerud J, Mishkin MM, Berkowitz HD, Atlas SW. Blinded-reader comparison of magnetic resonance angiography and duplex ultrasonography for carotid artery bifurcation stenosis. Stroke. 1994;25:4-10.
Alexandrov AV, Bladin CF, Maggisano R, Norris JW. Measuring carotid stenosis: time for a reappraisal. Stroke. 1993;24:1292-1296.