Carotid endarterectomy (CE) is a well-established treatment for the prevention of stroke. However, CE has inherent limitations that make it applicable only to treatment of the surgically accessible internal carotid artery. Recently, percutaneous transluminal carotid artery stenting (CAS) deployment has emerged as an alternative modality for the treatment of carotid artery disease. Current data for CAS are comparable with CE. Trials comparing CE with CAS are under way. The results of these studies will not reflect the state-of-the-art technology as new stents and new embolic protective devices will become available during these studies. Therefore, the decision to use CAS will depend on patient risks and lesion characteristics.
Husam A. Noor, MD
Ramdas G. Pai, MD
is a Professor of Medicine and the Cardiology Fellowship Program
Director, Loma Linda University Medical Center, Loma Linda, CA.
Carotid endarterectomy (CE) is a well-established
treatment for the prevention of stroke. However, CE has inherent
limitations that make it applicable only to treatment of the
surgically accessible internal carotid artery. Recently,
percutaneous transluminal carotid artery stenting (CAS)
deployment has emerged as an alternative modality for the
treatment of carotid artery disease. Current data for CAS are
comparable with CE. Trials comparing CE with CAS are under way.
The results of these studies will not reflect the
state-of-the-art technology as new stents and new embolic
protective devices will become available during these studies.
Therefore, the decision to use CAS will depend on patient risks
and lesion characteristics.
Stroke remains the leading cause of disability in the
industrialized world, and carotid artery disease (CAD) is a major
cause of stroke. Given the tragedy and morbidity of stroke, any
additional treatment modalities that will enable us to improve the
care we provide are welcome. In the United States alone, the
incidence of strokes is approximately 600,000 per year, of which
60% are due to CAD. This leads to approximately 160,000 deaths and
leaves many more patients with major disabilities.
Carotid endarterectomy (CE) is a well-established treatment of
extracranial carotid atherosclerotic disease in symptomatic and
asymptomatic lesions. Approximately 200,000 CEs are performed
annually. However, CE is limited to treatment of surgically
The technology used for endovascular interventions has undergone
a rapid evolution over the past decade in that coronary and
peripheral arterial interventional techniques have been used for
carotid, vertebral, and other cerebral interventions.
The progress in the treatment of cerebrovascular disease has
been much slower than that of CAD. This is probably related to the
more intractable nature of cerebrovascular disease and the
diversity of specialists involved in the care of these
Carotid artery stenosis
The treatment of stroke resulting from atherosclerotic stenotic
lesions can be categorized as medical, surgical, or endovascular.
The first two types of treatments have been available
traditionally. The medical therapy centers on antiplatelet and
antithrombotic agents and aggressive risk-factor modification.
Currently, the antiplatelet agents available for stroke prevention
include aspirin, ticlopidine, clopidogrel, and an
aspirin/dipyridamole combination. The oral anticoagulant, warfarin,
has been used mainly in the context of cardiogenic brain
For extracranial carotid atherosclerosis, the studies that will
be discussed showed the superiority of surgical therapy compared
with the best medical therapy. In addition, the lessons learned
from these landmark studies have allowed the construction of an
evidence-based measure against which future medical and ongoing
interventional trials should be compared. The degree of the
stenosis and whether the lesions are associated with symptoms
significantly challenge the approach to treatment.
Symptomatic carotid artery stenosis
Two large trials and other small ones that were conducted in the
1980s and 1990s involved a total of approximately 6000 patients
with cerebral or retinal symptoms and corresponding atherosclerotic
The European Carotid Surgery Trial (ECST)
and the North American Symptomatic Carotid Endarterectomy Trial
stopped recruiting patients with severe stenosis (70% to 99%) after
fewer than 700 patients had been randomized in each trial. Positive
results favoring surgery were evident for patients with 70% to 90%
stenosis measured angiographically. However, at that time, the
benefit had not been detected for symptomatic patients with lesser
degree of stenosis (moderate stenosis 50% to 69%, mild stenosis
<50%). In these large trials, patients with nonathe-rosclerotic
disease, organ failure, unde-rlying potentially fatal cancers,
embolic cardiac lesions, and remote (>6 months) ischemic
symptoms were excluded.
It is important to realize that ECST and NASCET used different
denominators to assess the degree of stenosis (Figure 1). Barnett
remeasured the degree of the stenosis of the ECST patients using
the NASCET method. Similar treatment results and long-term benefits
NASCET found CE to be beneficial in patients with 50% to 99%
stenosis with the greatest benefits for lesions with 70% to 99%
ECST found that CE was beneficial if the degree of stenosis was
above 80% (NASCET 60%).
Extrapolation of the results of NASCET and ECST into clinical
practice requires the acknowledgment of two facts. First, the
reported surgical complication rate of stroke and death was the
result of the care taken by the investigators to select surgeons
with proven expertise. In NASCET, surgeons were required to provide
evidence that they had achieved a perioperative complication rate
of <6%. When the results of the 328 patients with severe carotid
artery stenosis who underwent CE were analyzed, the perioperative
rate of death or stroke was 5.8%. Substituting a hypothetical
perioperative complication rate above that observed in NASCET shows
that the higher surgical complication rates soon negate benefits.
The second fact is that the results were acquired from angiographic
data. Extrapolation of the results to patients investigated using
noninvasive methods remains unproven.
What about the patients with less than 70% stenosis? In NASCET,
226 patients were randomized to medical therapy or surgical therapy
and followed up for an average of 5 years. The patients were
stratified into two groups: those with 50% to 69% stenosis, and
those with <50% stenosis. Benefits of CE were less dramatic in
the 50% to 69% group than in the severe group with >70%
stenosis. No benefits could be seen in patients with <50%
stenosis. Similarly negative results came from the ECST for
patients with what ECST calls 70% stenosis and NASCET measures at
Patients with a moderate degree of stenosis and high operative
risks should be managed medically with close follow-up. This
includes patients with an occluded carotid artery on the opposite
side of the symptoms, diabetes, thrombus containing stenotic
artery, ulcerated lesions,
diastolic blood pressure >90 mm Hg, and objective evidence of
brain infarction in the territory affected by the stenosis.
No strong evidence is available currently to support the use of
CE to prevent stroke in patients with nonspecific, nonhemispheric,
or vertebrobasilar symptoms. Similarly, no data are available to
support the use of CE in patients with symptoms related to
fibromuscular dysplasia, postradiation fibrosis, or carotid artery
Asymptomatic carotid artery disease
The frequency of asymptomatic athe-rosclerotic carotid artery
stenosis increases with age. It has been estimated that 2 million
individuals in the United States older than 50 years have >50%
stenosis of the carotid artery on one or both sides.
The Asymptomatic Carotid Atherosclerosis Study (ACAS) randomized
half the patients to the medical therapy arm and the other half to
the CE arm.
Individuals with >60% stenosis had a benefit at 5 years favoring
CE. Ipsilateral stroke occurred in 11% of the patients in the
medical arm compared with 5.1% in the surgical group (Table 1).
Limitations of CE
The benefit of CE holds true only if the risk of death and
stroke from the procedure is <6% for symptomatic patients and
<3% for asymptomatic patients.
Furthermore, in routine clinical practice, these numbers are widely
quoted as they set the standard for whether or not patients are
referred for surgery based on the institution's experience with CE.
However, postoperative complications of CE are not only limited to
stroke and death. For example, in NASCET, the incidence of cranial
nerve injury was 7%, resulting in a group of patients with new
neurologic deficits related to the surgery. In ECST, the rate of
major stroke and death was approximately 7%. Although this trial
used somewhat different methodology than NASCET, this 7% rate of
major stroke and death reported by ECST is three times the 2.1%
rate reported by the NASCET group. The rate of CE complications
reported is dependent on the specialty of the study investigators.
Having a neurologist among the investigators of CE studies may
result in a higher rate of neurologic complications because other
specialists may have difficulty in acknowledging subtle neurologic
Another important issue regarding CE is whether the patient
population in these large trials represents the patients seen in
daily practice. In NASCET, the majority of patients were below the
age of 80 and their clinical comorbidities lacked significant
coronary artery disease--only 19% had a previous myocardial
infarction, 2.6% had congestive heart failure, 2% had valvular
heart disease, and 5% had cardiac arrhythmias; patients with renal
failure, liver failure, or cancer were excluded from entry into the
study. Also, <20% were diabetics and hyperlipidemics and
approximately 30% were smokers.
The ACAS participants were extremely low-risk, relatively healthy
Therefore, it is obvious that the results of these trials cannot be
applied to patients at higher risk or those with complex lesions
(Table 2). In NASCET, the existence of contralateral carotid artery
occlusion and tandem siphon lesions was associated with a higher
risk of postoperative complications, 14% and 9%, respectively.
Carotid angioplasty and stenting: Evolving
The advent of percutaneous treatment of significant carotid
artery stenosis dates back to the early 1980s with several groups
reporting their experience in a handful of cases with reasonable
success for both atheromatous and nonatheromatous lesions.
In 1987, Theron and colleagues
reported the first large series of carotid angioplasty including 6
patients with atherosclerotic lesions and 5 postsurgical restenotic
lesions. In 1990, the same investigators treated 13 patients with
significant carotid artery stenosis with distal protection devices.
The Carotid and Vertebral Artery Transluminal Angioplasty Study
was a prospective randomized study that compared carotid
angioplasty with CE. Stents were not used routinely. Only 26% of
patients received a stent. The 30-day risk of stroke and death was
10% in both groups. The surgical group had significantly more
cranial nerve palsies, 8.7% versus 0%, and had more hematomas
requiring treatment, 6.7% versus 1.2%.
In 1995, several groups published their experiences with carotid
Roubin and colleagues published 5-year prospective data on carotid
artery stenting (CAS) in 528 patients with either symptomatic
(>50% stenosis) or asymptomatic (>60% stenosis).
Three-year follow-up data were available for 518 patients. Results
are summarized in Table 3.
An international registry was formed that compiled the
experiences of the centers involved in carotid stenting. Data for a
total of 5200 carotid stenting procedures were presented. Most of
these series included patients that would have been excluded from
randomized surgery versus medical treatment trials because of
additional risk factors. The technical success rate was >98%.
The 30-day mortality was 0.84% and all stroke rates (major and
minor) were 3.94%. In addition, the follow-up data from the
international registry showed a stroke rate of 3% during a period
of 24 to 36 months and a restenosis rate of approximately 4%.
These numbers seem to be very competitive with those of CE. In
NASCET, the stroke rate in the surgical arm over 3 years was 14%
and that of ACAS was 10% over 5 years.
The enthusiasm and interest in CAS has grown tremendously in
recent years. This has paralleled the development of stent and
catheter technology dedicated to the carotid arteries. These
technologic advances pose challenges not limited to stents suitable
for carotid use but also to carotid access systems and embolic
protection devices (EPDs). Eventually, this will lead to successful
performance of safe and less traumatic carotid procedures.
Fundamentals of stent angioplasty in carotid artery
The procedural goal of stent angioplasty of the carotid artery
is the elimination of the stenosis and the prevention of embolism
and stroke. In principle, the technique is similar to that used in
coronary and peripheral interventions.
Carotid artery ultrasound or magnetic resonance imaging (MRI)
can provide information about the severity and the composition of
the atherosclerotic plaque. Head computed tomography or MRI should
be obtained before the procedure to ascertain the extent of the
previous infarctions and rule out any other structural
abnormalities, particularly any evidence of intracranial
hemorrhage. Four-vessel cerebral angiography is essential to plan
for CAS. The information that should be obtained is the best
angiographic view for the intervention, the significance of any
arch disease if present, and the extent of the collateral cerebral
Standard transcranial Doppler (TCD) monitoring during the
procedure provides real-time information about ipsilateral middle
cerebral artery perfusion and embolization.
It may also be useful in showing the amount of blood flow during
dilatation and in obtaining information about the efficiency of the
collateral blood flow.
Antiplatelet and anticoagulant therapy
Patients who are to undergo CAS should receive aspirin (325 mg)
daily starting at least 1 week before the procedure. Patients
should also receive clopidogrel (75 mg) 5 days before CAS.
Carotid access and general procedural techniques
The common femoral vein is accessed in order to facilitate
prompt infusion of medications and fluid as needed in anticipation
of hypotension and bradycardia as a result of the carotid sinus
baroreceptor reflex. The vein access also is important because a
temporary pacer wire needs to be inserted occasionally to address
To access the arterial system, usually the transfemoral approach
is used. After visualization of the target internal carotid artery
(ICA) and the intracranial arteries, an extra long (125 cm) 5F
diagnostic catheter is inserted into the lumen of a regular length
(110 cm) 7F (eg, multipurpose type) guiding catheter, protruding
from it distally and used as a "search" catheter. This search
catheter is engaged into the ostium of either the left or right
common carotid artery (CCA). Heparin (100 U/kg) is administered
intra-arterially to maintain the activated clotting time between
200 and 250 seconds. A regular 0.035-inch guidewire is advanced
close to or into the external carotid artery (ECA), and the
diagnostic search catheter is advanced over it to a position just
proximal to the CCA bifurcation or the stenosis. Using the search
catheter and the guidewire as a support, the guide catheter is
advanced to position in the distal CCA and the search catheter and
the guidewire are withdrawn. It may be necessary to predilate the
lesion with a coronary balloon over an angioplasty wire or the wire
of an EPD. One or more stents are then deployed across the lesion
as necessary. The stents are further dilated at higher pressures to
firmly embed them into the vessel wall (Figure 2).
Severe hypotension and bradycardia may occur during the
poststent dilatation. The frequency of the hemodynamic instability
is less if the balloon does not stretch the carotid sinus. Atropine
(0.5 mg) and norepinephrine should be available for immediate
administration. In most cases, the stented area of the ICA involves
the takeoff of the ECA and it ends caudally in the CCA. It is
important not to have the distal end of the stent come to lie in a
prominent vessel curvature because this may induce arterial spasm
or dissection. Treatment (dilatation or stenting) of the ECA is not
usually necessary, unless the ECA is providing significant
collaterals or unless the patient develops symptoms due to the
occlusion of the ECA, such as jaw pain. In this situation, the flow
can be restored using common balloon techniques through the stent
Clopidogrel is continued for 4 to 6 weeks and aspirin is
continued indefinitely. The patient is fully awake during the
procedure with little or no sedation, as constant neurologic
monitoring is essential.
Procedure-related cerebral embolization
In CE, embolism is still the most common cause (55%) of all the
cerebrovascular complications, followed by hyperperfusion syndrome
(29%), and hypoperfusion (17%) in a series of 500 patients with CE
who had a neurologic complication rate of 4.8%.
Neurologic deficits during carotid stenting are infrequent and
are usually transient and resolve in minutes to hours. Echolucent
carotid plaques and lesions with >90% stenosis produce a higher
number of embolic particles.
Jordan and colleagues
reviewed 105 patients who underwent TCD monitoring during CE (n =
75) or CAS (n = 40). In the CAS group, there was a mean of 7.4
emboli per stenosis and 4 neurologic events. In the CE group, there
was a mean of 8.8 emboli per stenosis with 1 neurologic event. The
mean number of microemboli was 56.8 in patients with neurologic
complications and 31.2 in those without (
A study by Mathur et al
showed that the independent predictors of procedural stroke are
advanced age, lesion severity, and long or multiple stenotic
lesions. No correlation was found for plaque ulceration, sex,
presence of coronary artery disease, diabetes,
hypercholesterolemia, smoking, prior CE, contralateral carotid
occlusion, or type of stent used.
However, when a clinically symptomatic embolism occurs, a
loading dose of 10 mg of recombinant tissue plasmin activator
(rt-PA) is administered in the ICA followed by an immediate
cerebral angiogram. If an embolus is evident at a major branch of
the middle cerebral artery, a microcatheter is advanced over a
hydrophilic 0.014-inch wire to the occluded branch to restore a
minimal blood flow. Then rt-PA is continued to a maximum of 50 mg.
If the angiogram did not reveal an embolus, rt-PA is administered
into the ICA. The risk of intracerebral bleeding in these patients
It cannot be emphasized enough that, unlike other organs, the
brain is very unforgiving when microembolization occurs. Embolic
particles can be created at any time during carotid angioplasty.
Cerebral protection should reduce embolization risks, and
therefore, should widen the indications for carotid angioplasty to
include high-risk lesions.
Theron and colleagues
have introduced a concept that involves the use of a distally
placed protection balloon designed to catch and remove embolic
material. Wholey et al
were issued the first patent on the design of a filter device that
was positioned between the dilating balloon and the distal end of
the catheter. Currently, not one of the devices is approved for
routine use in CAS in the United States.
There are three different classes of EPDs: the balloon
occluders, the filters, and the circulatory control devices (Table
4, Figures 3 and 4).
Data from a subset of 2038 cases from the Carotid Artery
Stenting Global Registry showed that EPDs decreased the risk of
procedure-related death from 4.1% (unprotected) to 2.4%
The preliminary clinical data on the application of the
GuardWire Plus (Medtronic AVE; Santa Rosa, CA) system in the
Carotid Angioplasty Free of Emboli (CAFE) study are promising.
Seventy-five patients were treated; 56% had symptomatic disease and
31% had severe contralateral carotid disease. In addition, 36% of
patients had angina and 28% had suffered cerebrovascular accidents.
Procedural success was achieved in all cases. No strokes were
recorded and only 1 patient suffered from a wire-related
dissection. Patients tolerated a mean balloon occlusion time of
15.3 ± 5.5 min.
Data from the PercuSurge Global Registry are even more impressive.
Of 463 procedures performed in 19 institutions, only 7 (1.5%)
resulted in death or stroke.
Filter devices have high capture efficiency with unparalleled
particle retention and do not interrupt cerebral blood flow. Armor
presented the experience of 3 institutions on 138 carotid
interventions. The 30-day death and stroke rate with filter devices
Published clinical data with the Parodi Anti-Embolization System
(PAES; ArteriA Medical Science, Inc.; San Francisco, CA) consist of
an international series of 60 patients. Angiographic evidence of
carotid artery stenosis with thrombus or ulceration was present in
45% of patients. The deployment of the PAES was successful in all
the patients. There was no periprocedural stroke or death.
Potential advantages of the flow reversal technique include no
distal problems, such as spasm and dissections (Figure 5). In
addition, embolic protection starts before crossing the lesion.
PAES allows for the treatment of tight tortuous lesions and
captures particles of all sizes. However, some patients may not
tolerate retrograde flow, and it requires the insertion of an 11F
sheath, which may lead to more vascular complications (Table
Roubin et al
presented data on 329 CAS procedures using PercuSurge GuideWire
(Medtronic AVE; Santa Rosa, CA) (n = 232), PAES (n = 10),
NeuroShield (MedNova, Inc.; Galway, Ireland) (n = 74), Accunet
(Guidant; Indianapolis, IN) (n = 9), and AngioGuard (Cordis Corp.;
Warren, NJ) (n = 4). They reported 30-day event rates of 0% for
major stroke, 2% for minor stroke, and 0.9% for retinal emboli; for
a total embolic event rate of 3%.
Current limitations of EPD
Seizures, altered sensorium, transient ischemic attacks, and
stroke resulting from transient cerebral ischemia have been
documented for all three types of EPD.
The series reported by Henry et al,
the largest single-center study of PercuSurge EPD, showed
intolerance to balloon occlusion of the cerebral flow in 4.9% of
the patients. This was in the context of a mean occlusion time of
422 seconds. In the PAES series, 5% of the patients were intolerant
to flow reversal.
Filters occasionally accumulate enough debris to plug up the pores
and impede the cerebral flow. Patients with contralateral carotid
artery occlusion are likely to be intolerant to EPD. The EPDs add
additional bulk to the guidewire and require manipulations and the
traversal of the target lesions. Bulkiness occasionally results in
failure to pass these devices across tight lesions, negating the
benefit of devices in the patient subsets that would potentially
benefit most from cerebral protection. Microscopic furrows in the
carotid artery intima and vasospasm may occur with some filters.
Filters with much smaller pore sizes may come at the cost of more
pore plugging and cessation of flow. Spillage or overflow of debris
can also occur with filters. Clot formation on the distal surface
of the device is possible. This is especially true in cases in
which the procedural time is increased or the distal flow is
The current generation of EPDs does not offer protection from
embolic phenomena that precede the deployment. In addition, they
may add to blood loss, procedural time, and the complexity of the
technique. At this time, no devices are approved by the Food and
Drug Administration for cerebral protection. There are no studies
that compare the strategies of CAS with EPD versus CAS without
cerebral protection. This assumption is based on EPDs being
all-protective, which may not be the case.
Who should be offered CAS?
It is quite obvious, as with other endovascular therapies, that
CAS could potentially benefit patients by causing less risk,
trauma, and patient discomfort. At this time, CAS may be considered
in poor surgical candidates or patients who refuse surgery and
understand their treatment options and associated risks.
What is not yet clear is the comparison of CE and CAS in low-risk
The future of CAS
Carotid angioplasty and stenting is rapidly becoming an
alternative to traditional CE. The potential complications and
their management will require a multispecialty team approach. In
the global carotid stent survey, surveys were sent to 36 major
carotid stenting centers in the world. Information was obtained
regarding the number of procedures, the stents used, the
complications rate, and follow-up information. The results showed
that cardiologists performed 63% of the cases, radiologists 25%,
and surgeons 12%.
Cerebral protection devices will play a major role in the future.
The rapid technologic improvements should further refine these
devices and should improve outcomes. Elimination of embolic debris,
inflammation, or oxidative cellular damage represents an additional
therapeutic goal. The role of platelet glycoprotein receptor
antagonists as an adjunctive therapy is not yet clear.
The key issue will be the results of the study of Carotid
Endarterectomy Versus Stent Trial (the CREST trial), funded by the
National Institutes of Health.
It is conceivable that if this study confirms that CAS is not
inferior to CE, then the interest in and the number of CAS
procedures will be further enhanced. Some leaders in this field
have argued that this trial is premature because CAS is still
Carotid angioplasty and stenting is evolving from its initial
controversial status to that of an acceptable alternative for the
treatment of CAD. The procedure is feasible and safe even in
high-risk patients. As stents, guide catheters, and EPDs improve
technical success and patency, complication rates should diminish.
Current results of CAS in nonrandomized control trials are
comparable to CE.
The authors are indebted to Dr. Felipe Cecena for providing the
advice and images for this report.