Ugo Avati
Servizio di Neuroradiologia A.O. Cardarelli - Napoli
Cerebral ischaemia and infarcts due to atherosclerosis are major causes
of mortality and morbidity in Europe and North America. Two thirds of strokes
result from atherosclerotic disease which has a predilection to develop
in the extracranial carotid arteries, particularly at the bifurcation of
the common carotid artery. In patients with transient ischemic attacks
(TIA), amaurosis fugax and previous cerebrovascular accident (CVA) there
is a significant risk of a future CVA (Table 1).
| Transient ischemic attack (TIA) | 45 % |
| Amaurosis fugax | 19 % |
| TIA and amaurosis fugax | 4 % |
| Stroke with good functional recovery | 32 % |
Two recent major trials, the European Carotid Surgery Trial (ECST) and
the North American Symptomatic Carotid Endarterectomy Trial (NASCET) have
shown benefit from carotid endarterectomy in some patients with symptomatic
carotid stenosis of 70% or greater.
The NASCET trial set the standard for preoperative assessment of the
carotid bifurcation: an angiographically demonstrated and specifically
measured stenosis of the internal carotid artery in the absence of significant
tandem lesions (intracranial stenosis in the vessel ipsilateral to an extracranial
carotid stenosis).
Other studies (Asymptomatic Carotid Artery Stenosis Study, 1995) have
shown benefit from carotid endarterectomy also in patients with asymptomatic
stenosis greater than 60%.
As a result there has been increased interest in identifying the most
appropriate method of investigating patients with carotid atherosclerosis
prior to possible surgery.
In most cases cerebral ischaemia results from embolic mechanism. Nevertheless,
haemodynamic changes due to impaired intra- and extra-cranial anastomotic
systems are important causes in determining both the onset and the severity
of symptoms.
Anastomotic systems include a complex vascular network connecting each
other the extra-cranial vessels and completed at the cranial base by the
circle of Willis (anterior and posterior communicating arteries).
Some cases of complex steno-occlusive disease may rarely benefit from
surgery (intra-extracranial or extra-extracranial anastomoses). In these
cases a complete assessment of the extra- and intra-cranial anastomotic
systems is required, to both confirm the indication for surgery and plan
the most appropriate operative technique.
Intra-arterial digital subtraction angiography (IADSA) procedures have
been considered the gold standard in the evaluation of carotid stenosis
and are used in preoperative assessment in most centers in USA and Europe,
but their attendant risks have led to the search for noninvasive imaging
tecniques which can be used in their place.
For a noninvasive imaging modality to replace catheter angiography
in the preoperative assessment of carotid atherosclerosis, it must be able
to identify stenoses of 70-99% severity, distinguish severe stenosis from
complete occlusion and exclude tandem stenoses as accurately as angiography.
Alternatively, noninvasive imaging may be used as a screening method
to select patients with significant disease who may potentially benefit
from carotid endarterectomy, to whom further invasive investigation would
be limited, accurately distinguishing them from patients with minimal or
no disease, in whom angiography would be unnecessary.
Current treatment options mean that the degree of accuracy for less
severe degrees of stenosis (< 70%) is not as critical, although this
may change in the future.
The debate as to the role of the different imaging modalities in the
assessment of symptomatic carotid atherosclerosis is still continuing.
In many centers colour-doppler ultrasound examination is the screening
method of choice for carotid stenosis. Colour-doppler provides grey scale
demonstration of the vessel, display of flow by colour and spectral analysis
of sample volumes within the vessel to produce a doppler velocity waveform.
Standard duplex ultrasound is said to be able to quantify vessel stenosis
with a high sensitivity and specificity; colour-flow doppler alone is equally
accurate in assessing stenotic lesions.
A problem identified with duplex ultrasound is that high grade stenoses
may be misinterpreted as occlusion. This is a significant error, as patient
in the first category could benefit from endarterectomy, whereas those
with occlusion would not. The rate of concordance between ultrasound and
angiography is high (70%) in patients with normal vessels or mild stenoses
(0-29%). In patients with angiographically demonstrated severe stenosis
(70-99%) ultrasound result correlated in only 52% of the cases. (Fig.
1) In these patients it is essential that they are correctly identified
and referred for surgery; if ultrasound were used as the sole preoperative
evaluation, some arteries would have been considered suitable for endarterectomy
which were in fact totally occluded and would not have been benefited from
surgery.
Other arteries with severe stenosis which would appropriately be treated
by endarterectomy would not have been identified as a result of subtotal
occlusion being reported as total occlusion. Furthermore, other arteries
would not have been considered further due to an apparently less severe
stenosis.
Another limitation of ultrasound for carotid stenosis is its inability
to demonstrate the full extent of the cerebral circulation. Ultrasound
examination is limited to the extracranial carotid and vertebral arteries
and cannot therefore demonstrate intracranial tandem lesions. The haemodynamic
changes following endarterectomy may lead to an increased risk of rupture
of intracranial aneurysms in the same arterial territory, and is therefore
essential to correctly diagnose such important intracranial pathology.
Other problem with ultrasound examination is the difficulty in obese
patients and the attenuation of the ultrasound beam by tissue calcium.
Also, ultrasound examination is highly dependent on operator skill.
Therefore, ultrasound examination is to be considered a noninvasive
screening tool that provides an accurate indicator of minimal or absent
disease, thereby avoiding the cost and risks of other imaging modalities
in patients in whom no further investigation of the carotid arteries is
indicated.
Intra-arterial angiography with selective catheterisation of the carotid
vessels is currently the accepted standard for imaging the extracranial
carotid arteries, being the best method able to detect stenoses sensitively,
determine the anatomical site involved accurately, quantify the stenoses
accurately and also to assess the intracranial circulation and the extra-
and intra-cranial anastomotic systems.
It carries a 4% risk of TIA and a 1% risk of permanent neurological
deficit, the risk being greatest in patients with severe carotid stenosis.
In view of this, selective intra-arterial angiography is currently reserved
for the investigation of patients being considered for endarterectomy.
A suitable diagnostic technique to screen for the presence of carotid
stenosis is therefore required. Intravenous digital subtraction angiography
(IVDSA) has been suggested as a screening method but despite a lower rate
of neurological complications, the rate of systemic complications such
as myocardial infarction or pulmonary oedema is unacceptably high. Problems
with poor image quality also mean that IVDSA cannot be used reliably to
quantify stenoses. These drawbacks make IVDSA an inappropriate screening
investigation.
The measurement of stenosis severity derived from angiography images
is assumed to represent the true value. However, it should be noted that
although intra-arterial angiography is the accepted standard, inter- and
intraobserver variability in reporting >50% stenoses is not insignificant
and may be up to 15%.
In investigating the extracranial carotid arteries, angiography is
useful to clarify diagnostic problems not resolved by noninvasive imaging
modalities. Despite a lesser spatial resolution, digital subtraction angiography
uses lower doses of contrast medium than conventional angiography; using
thinner catheters it may be proposed also in ambulatory patients. The incidence
of drug intolerance is mild to moderate, and may be reduced using non-ionic
and low-osmolarity contrast agents.
IADSA provides demonstration of the presence and distribution of the
atherosclerotic plaque, with a detailed display of the vessel wall and
the ulcerative changes, dissecation and calcifications. (Fig.
2) It provides an accurate measurement of stenosis severity;
the degree of stenosis is measured on magnified hard copy of either the
oblique or the lateral image, whichever showed the most severe stenosis.
The diameter of the residual lumen is compared with the original diameter
of the normal carotid bulb, extrapolated on the angiogram, as in the ECST.
(Fig. 3) The percentage stenosis is calculated
by (normal lumen - residual lumen)/ normal lumen x 100. The technique is
especially useful in displaying the "string sign" in subocclusive lesions,
distinguishing them from total occlusion. (Fig.
4)
Angiography demonstrates the full extent of the cerebral circulation,
to detect tandem lesions or other non-atherosclerotic lesions such as dissections
(Fig. 5) , fibrodysplasia, arteritis or aneurysms
(Fig. 6) ; it also provides functional evaluation
of the intracranial anastomoses, of great value in planning the surgical
treatment and in prognosis.
Angiography provides an accurate picture of the internal profile of
the vessel lumen; the wall and the surrounding tissue cannot be investigated.
For this reason its role has been debated. Nevertheless, angiography plays
an important role due to three intrinsic characteristics:
1) high sensitivity in evaluating the wall profile, also in cases of
tortuous vascular course; (Fig. 7)
2) best evaluation of the aortic arch and the origin of epiaortic vessels;
3) high accuracy in evaluating the anatomy and the abnormalities of
the circle of Willis. Although the circle of Willis is complete in only
21% of patients, and may be accurately evaluated by magnetic resonance
angiography, selective angiography provides best evaluation of intracranial
stenoses and aneurysms.
Anyway, the current role of selective angiography is more precise and
limited than before. Some subtle lesions of the vessel wall and the intracranial
tandem stenoses may be correctly evaluated only by selective injection,
that is also able to distinguish accurately complete occlusion from severe
stenosis. In all other cases catheter angiography may be replaced by noninvasive
imaging modalities.
Due to ongoing technological advances, computed tomography now offers
detailed information on the carotid bifurcation with short acquisition
times and high spatial resolution, after low dose bolus administration
of contrast agents. Spiral (helical) acquisition devices further improve
image quality and subsequent multiplanar reformatting.
CT accurately displays the site of the stenotic lesion and its type,
being able to distinguish high density "hard" calcified plaque from "soft"
fibrolipid build up, which is hypodense with respect to the contrast enhanced
lumen. (Fig. 9 and 10)
The endoluminal surface and the internal arrangement of the plaque along
the vessel wall are easily explored, sometimes displaying small surface
ulcerations resulting in emboli. (Fig .11)
CT also provides an accurate measurement of the percentage of stenosis
applying both NASCET and ECST criteria, by calculating a ratio between
lumen diameters and between lumen areas.
Multiplanar and three-dimensional reconstructions offer a longitudinal
view of the vessel and further information on the arrangement and size
of the plaque. (Fig. 8)
CT can identify a carotid dissection and its evolution; in the acute
phase the subintimal haematoma is slightly hyperdense on non-enhanced scans;
then it becomes isodense with respect to the lumen, making reduced vessel
diameter the most significant finding.
CT readily displays extracranial carotid aneurysms and pseudo-aneurysms
as well as offering detailed information on surrounding structures.
(Fig. 12)
In the next future magnetic resonance angiography (MRA) will probably
represent the best preoperative imaging modality in evaluating carotid
stenoses. It allows an accurate investigation of the brain, the extracranial
vessel and the cerebral circulation in the same session.
Since its early experimental applications, nuclear magnetic resonance
phenomenon has shown sensitivity to fluid movement; recent MRA techniques
are based on flow effects, either with respect to the sequence timing (time
of flight effects, TOF) or to the magnetic field gradient being applicated
(phase effects or phase contrast, PC).
Assessing the degree of internal carotid stenosis at its bifurcation,
MRA provides reliable information on normal or mildly stenotic vessel without
false negatives. However, MRA tends to overestimate the degree of stenosis,
especially the 2D-TOF technique. Nevertheless, recent studies found MRA
more reliable than ultrasound in distinguishing complete occlusion from
severe stenosis, especially when both 2D and 3D TOF techniques are used.
A major drawback in MRA investigation is poor resolution in characterizing
atherosclerotic plaque and in detecting ulcerations, but technological
advances (e.g. phased array surface coils) should soon overcome problems
linked with spatial resolution, signal to noise ratio and flow-turbulence
artifacts, which currently prevent MR and MRA in displaying ulcerated plaque.
MRA is a panoramic technique, being able to show tandem lesions and
also the intracranial circulation at the circle of Willis. Assessment of
the origin of epiaortic vessel is difficult, although dedicated coils (head-neck
or quadrature coils) are already available to demonstrate the origin of
the vessels from the aortic arch, carotid bifurcations and carotid siphons
in one acquisition. (Fig. 13 and
14)
The high negative predictive value and the sensitivity in detecting
stenotic findings, coupled with the ability in providing detailed information
on brain structure, make MR and MRA particularly suited to the noninvasive
preoperative work-up in symptomatic patients with carotid stenosis. As
a screening tool, MR is more expensive than ultrasound, but is less operator-dependent
and gives a better overall display both at cervical and cerebral levels.
There is no doubt that a combination of ultrasound and catheter angiography
can provide an efficient and clinically acceptable basis on which to make
decisions in the management of carotid artery disease. Some recent studies
suggest to refer for catheter angiography patients with stenosis greater
than 30% on ultrasound; other studies suggest that only patients with a
stenosis of 70% or greater should undergo angiography. In fact, one should
choose a threshold for the detection of abnormality; the problem is that
any option has implications for outcome. If a low threshold is chosen,
many patient who do not require it will undergo catheter angiography, but
if only patients with a stenosis >70% on ultrasound underwent angiography,
some patients could be denied appropriate treatment.
In fact, it does not seem a rational course to limit diagnostic options
to ultrasound and/or catheter angiography, especially if we do consider
legal implications, recently pointing on precise surgical indication and
accurate anatomic and functional preoperative evaluation. These modality
are to be considered at the beginning and at the end of a series of imaging
tools, including MR and TC in intermediate position.
Colour-doppler examinations will probably continue to be the screening
method of choice, due to low cost, wide diffusion and good characterization
of the plaque structure. MR is more expensive, but il allows an investigation
of extra- and intra-cranial vessels and the brain structures in one session.
At present, despite a time-spending post-processing session performed by
the operator, MRA images are already richest in reliable information. The
ongoing technological advances (dedicated coils and acquisition sequences)
will further improve its diagnostic value, making MR probably to replace
every other modality in a next future.
Volumetric acquisition CT seems to be very useful; however, comparing
it with MR a wide evaluation is required regarding cost and repercussion
of the use of contrast agents. Angio-CT gives an excellent demostration
of the plaque structure and the degree of stenosis, due to the superior
spatial resolution.
Catheter angiography will probably continue to be considered the reference
standard in vascular imaging; the representation of the vessel lumen it
provides is accurate, panoramic enough and the one we are most familiar
with.
In fact, the number of angiographic examinations in the steno-occlusive
disease is progressively reducing in most centres; the diagnostic role
of catheter angiography will probably be "targeted" and more and more limited
to selected cases, whereas its therapeutics and interventional applications
will increase.
1) Anderson CM, Saloner D, et al. (1992) Assessment of carotid artery
stenosis by MR angiography: comparison with x-ray angiography and color-coded
Doppler ultrasound. Am J Neuroradiol 13:989-1003.
2) Avati U., Esposito G, et al. Trattamento chirurgico dell'insufficienza
cerebro-vascolare: criteri di valutazione neuroradiologica. Neuroradiologia
tra clinica e terapia, Ischia, 1994.
3) Bornstein NM, Beloev ZG, et al. (1988) The limitations of diagnosis
of carotid occlusion by Doppler ultrasound. Ann Surg 207:315-317.
4) Cellerini M., Chiti E. et al. (1996) RM ed AngioRM nello studio
della patologia steno-occlusiva (aterosclerotica) delle arterie carotidi
nel tratto cervicale. Riv. Neuroradiol 9 (Suppl 2):45-54.
5) Chikos P, Fisher L. et al. (1983) Observer variability in evaluating
extracranial carotid artery stenosis.Stroke 14:885-892.
6) Culebras A, Magana R , et al. (1988) Computed tomography of the
cervical carotid artery: significance of the lucent defect. Stroke 19;723-727.
7) Eikelboom BC, Riles TR et al (1983) Inaccuracy of angiography in
the diagnosis of carotid ulcerations. Stroke 14:882-887.
8) Fox AJ (1993) How to measure carotid stenosis. Radiology 186:316-318.
9) Heiserman JE, Drayer BP, et al. (1992) Carotid artery stenosis:
clinical efficacy of two dimensional time of flight MR angiography. Radiology
182;761-768.
10) Huston J III, Lewis BD, et al. (1993) Carotid artery: prospective
blinded comparison of two-dimensional time of flight MR angiography with
conventional angiography and duplex US. Radiology 186;339-344.
11) Leeson MD, Cacayorin E (1985) Atheromatous extra-cranial carotid
arteries: CT evaluation correlated with arteriography and pathologic examination.
Radiology 156;397-402.
12) North American Symptomatic Carotid Endarterectomy Trial Collaborators
(1991) Beneficial effect of carotid endarterectomy in symptomatic patients
with high grade stenosis. N Engl J Med 325:445-452.
13) Pellicanò G, Cellerini M, et al. (1996) Il ruolo della TC
e dell'angioTC. Riv. Neuroradiol 9 (Suppl 2): 35-43.
14) Rega L, Santoni M (1996) Ruolo dell'angiografia digitale. Riv.
Neuroradiol 9 (Suppl 2):27-34.
15) Turner WH, Murie JA (1989) IVDSA for extracranial carotid artery
disease. Br J Surg 76:1247-1250.
16) Waugh JR, Sacharias N (1992) Arteriographic complications in the
DSA era. Radiology 182:243-246.
17) Worthy S.A, Henderson J. et al (1997). The role of duplex sonography
and angiography in the investigation of carotid artery disease. Neuroradiology
39:122-6.
18) Yuan C, Tsuruda JS, et al. (1994) Techniques for high resolution
MR imaging of atherosclerotic plaque. J Magn Reson Imaging 4:43-49.
19) Zwiebel WJ (1992) Duplex sonography of the cerebral arteries: efficacy,
limitations and indications. AJR 158;29-36.