Updated Society for Vascular Surgery guidelinesfor management of extracranial carotid diseaseJohn J. Ricotta, MD,a Ali AbuRahma, MD, FACS,b Enrico Ascher, MD,c Mark Eskandari, MD,d

Peter Faries, MD,e and Brajesh K. Lal MD,f Washington, DC; Charleston, WV; Brooklyn, NY; Chicago, Ill;New York, NY; and Baltimore, Md

Management of carotid bifurcation stenosis is a cornerstone of stroke prevention and has been the subject of extensiveclinical investigation, including multiple controlled randomized trials. The appropriate treatment of patients with carotidbifurcation disease is of major interest to the community of vascular surgeons. In 2008, the Society for Vascular Surgerypublished guidelines for treatment of carotid artery disease. At the time, only one randomized trial, comparing carotidendarterectomy (CEA) and carotid stenting (CAS), had been published. Since that publication, four major randomizedtrials comparing CEA and CAS have been published, and the role of medical management has been re-emphasized. Thecurrent publication updates and expands the 2008 guidelines with specific emphasis on six areas: imaging in identificationand characterization of carotid stenosis, medical therapy (as stand-alone management and also in conjunction withintervention in patients with carotid bifurcation stenosis), risk stratification to select patients for appropriate interven-tional management (CEA or CAS), technical standards for performing CEA and CAS, the relative roles of CEA and CAS,and management of unusual conditions associated with extracranial carotid pathology. Recommendations are made usingthe GRADE (Grades of Recommendation Assessment, Development and Evaluation) system, as has been done with otherSociety for Vascular Surgery guideline documents. The committee recommends CEA as the first-line treatment for mostsymptomatic patients with stenosis of 50% to 99% and asymptomatic patients with stenosis of 60% to 99%. Theperioperative risk of stroke and death in asymptomatic patients must be

rtorwamddc

eieoaigo1it

tmmavaosttCp

cngpnsla

ipatic

patmm

JOURNAL OF VASCULAR SURGERYSeptember 2011e2 Ricotta et alA. Assessing the risk associated with intervention1. Anatomic and lesion characteristics

a. Lesion locationb. Lesion characteristicsc. Other anatomic considerations

2. Patient characteristics

B. Neurologically asymptomatic patients with 60%carotid artery stenosis1. CEA for asymptomatic lesions2. CAS in asymptomatic lesions3. Medical management of asymptomatic carotid

stenosis

C. Neurologically symptomatic patients with 50%carotid artery disease1. CEA in symptomatic stenosis2. CAS in symptomatic stenosis

D. Meta-analysis: CEA vs CAS Recommendations for selecting therapy

VI. Unusual conditions associated with carotid steno-sisA. Acute neurologic syndromes

1. Management of acute strokea. Presentation within 0-6 hoursb. Presentation later than 6 hours

2. Stroke in evolution (fluctuating neurologicdeficits)

3. Crescendo TIA4. Acute postintervention stroke/occlusion

Recommendations for management of acute neu-rologic syndromes

B. ICA occlusion with persistent symptoms/externalcarotid stenosis

Recommendations for management of symptom-atic ICA occlusion

C. Carotid dissection Recommendations for management of carotid dis-

sectionD. Combined carotid and coronary disease Recommendations for management of combined

carotid and coronary disease

Management of extracranial carotid disease has beenthe focus of intense investigation and debate by multiplemedical specialists since the introduction of carotid endar-terectomy (CEA) as a therapeutic option for the treatmentand prevention of stroke more than half a century ago.Initial hopes that CEA could reverse the clinical course ofstroke were proven false, and the role of surgical manage-ment of extracranial carotid and vertebral obstructions wasdefined by one of the earliest efforts at a multicenteredrandomized clinical trial, The Joint Study on The Extracra-nial Circulation.1 The results of this decade-long study,involving 5000 patients, established the role of CEA in thetreatment of minor stroke, transient ischemic attack (TIA),

and amaurosis fugax, confirmed that surgery had a limited aole in the treatment of established stroke, and establishedhe limited role of vertebral reconstruction in the treatmentf cerebral insufficiency. Over the ensuing decades, surgicalesults of CEA improved, asymptomatic carotid stenosisas increasingly identified by noninvasive studies, and CEA

ssumed a primarily prophylactic role as prevention ofajor stroke in asymptomatic patients or those with evi-ence of transient cerebral or ocular ischemia. Large ran-omized trials2-6 have established the role and efficacy ofarotid endarterectomy (CEA) in stroke prevention.

In the last decade, carotid artery stenting (CAS) hasmerged as a catheter-based alternative to CEA, and med-cal therapy for stroke treatment and prevention hasvolved. Currently, approximately 135,000 interventionsn lesions in the carotid bifurcation are being performednnually in the United States, by a variety of specialists,ncluding vascular surgeons, general surgeons, thoracic sur-eons, neurosurgeons, cardiologists, interventional radiol-gists, and interventional neurologists.7,8 Approximately1% of these interventions are catheter-based, and 90% of

nterventions are in patients without neurologic symp-oms.7

As in any situation where there are multiple options forhe treatment of a single condition, defining optimal treat-ent can be difficult. This is further compounded whenultiple specialists, often with nonoverlapping expertise,

re involved in the treatment of the patient. As a result, aoluminous and often conflicting literature has developedround the current standards of diagnosis and managementf extracranial carotid stenosis. Recently two large, pro-pective, randomized trials have been published comparinghe efficacy of CEA and CAS in the management of ex-racranial carotid stenosis.9,10 A meta-analysis comparingAS and CEA, including these trials has recently beenublished in the Journal of Vascular Surgery.11

In 2008, the Society for Vascular Surgery publishedlinical practice guidelines for the management of extracra-ial carotid artery disease in the Journal of Vascular Sur-ery.12 More recently, a multispecialty document has beenublished on the Management of Patients with Extracra-ial Carotid and Vertebral Artery Disease.13 This exten-ive document represents an effort to evaluate the existingiterature on extracranial carotid and vertebral disease and isn important reference.

The data contained in the recently published random-zed trials has prompted the Society for Vascular Surgery toublish an update of its 2008 guidelines, confined to man-gement of extracranial carotid artery disease. This is par-icularly appropriate because vascular surgeons play a majorf not predominant role in the management of patients witharotid bifurcation disease.

In developing these recommendations, the committeelaced more weight on the reduction of stroke and deathnd less on the importance of nonfatal myocardial infarc-ion (MI). Because the latter end point often represents theain benefit of CAS, the recommendations in this docu-ent are more circumspect with regard to the role of CASnd more supportive of the role of CEA than the recom-

telwtnano(

bobclmotamdcdtp

I

cita(ittbidts

taopo

eTrtr

JOURNAL OF VASCULAR SURGERYVolume 54, Number 3 Ricotta et al e3mendations of the American Heart Association (AHA)guidelines committee. This document is divided into sixmajor sections:

I. Indications for imaging of the extracranial circulationII. Selection of imaging modality

III. The importance of medical therapy in the overallmanagement of patients with carotid stenosis, includ-ing medical management in the peri-intervention pe-riod.

IV. Technical considerations for performing CEA andCAS

V. The relative roles of medical management, CEA andCAS for stroke risk reduction in patients with carotidstenosis based on review of the literature, with partic-ular reference to risk factor stratification and the mostrecent completed trials

VI. The management of unusual conditions associatedwith extracranial carotid pathology, including acuteneurologic conditions, symptomatic carotid occlu-sion, carotid dissection, and patients with carotid ste-nosis in need of coronary artery revascularization

The committee reviewed the literature pertinent toeach of the six areas and provided recommendations fortreatment using the GRADE (Grades of RecommendationAssessment, Development and Evaluation) system.14 Thissystem, adopted by more than 40 other organizations,incorporates an evaluation of the strength of the evidenceand the risks/benefits of implementing the recommenda-tion. For the purposes of this review, we placed the highestpriorities on reducing overall stroke risk, periproceduralstroke risk, and periprocedural mortality. Lesser impor-tance was given to reducing nonfatal MI, cost, and theability to perform a percutaneous procedure. Recommen-dations are characterized as strong GRADE 1 or weakGRADE 2, based on the quality of evidence, the balancebetween desirable effects and undesirable ones, the valuesand preferences, and the resources and costs.

GRADE 1 recommendations are meant to identifypractices where benefit clearly outweighs risk. These rec-ommendations can be made by clinicians and accepted bypatients with a high degree of confidence. GRADE 2recommendations are made when the benefits and risks aremore closely matched and are more dependent on specificclinical scenarios. In general, physician and patient prefer-ence plays a more important role in the decision-makingprocess in these circumstances.

In addition to the GRADE of recommendation, thelevel of evidence to support the recommendation is noted.Evidence is divided into 3 categories: A (high quality), B(moderate quality), and C (low quality). Conclusions basedon high-quality evidence are unlikely to change with fur-ther study, those based on moderate-quality evidence aremore likely to be affected by further investigation, andthose based on low-quality evidence are the least supportedby current data and the most likely to be subject to change

in the future. rIt is important to note that a GRADE 1 recommenda-ion can be made based on low-quality (C) evidence by theffect on patient outcome. For example, although there areittle data on the efficacy of CEA in asymptomatic patientsith 60% stenosis, one can recommend with confidence

hat CEA not be performed in these patients. A full expla-ation of the GRADE system is presented in the recentrticle by Murad et al14 referenced earlier. It is important toote that this grading system differs somewhat from thene used in the recent American College of CardiologyACC)/AHA Task force report.13

Each member of the committee was assigned responsi-ility for compiling information pertinent to a specific areaf the document. These data were distributed to all mem-ers for review, and each area was subsequently discussed inonference calls. A consensus of the recommendation andevel of evidence to support it was reached. Each recom-

endation in this document represents the unanimouspinion of the task force. Although some recommenda-ions are GRADE 2 with Level 3 data, the task force felt itppropriate to present these as the unanimous opinion of itsembers regarding optimal current management. This wasone with the recognition that such recommendationsould change in the future but that it was unlikely that newata would emerge soon. These guidelines are likely to be aliving document that will change as techniques are fur-her refined, technology develops, medical therapy im-roves, and new data emerge.

. INDICATIONS FOR CAROTIDBIFURCATION IMAGING

Stroke is the third leading cause of death, behindoronary artery disease (CAD) and cancer, and is the lead-ng cause of disability in the United States and Wes-ern Europe. Approximately 80% of strokes are ischemicnd 20% are hemorrhagic.15 Significant carotid stenosis50%) is seen in 12% to 20% of all anterior circulationschemic strokes, which is two to three times higher thanhe risk for less severe asymptomatic stenosis.16,17 Unfor-unately, only 15% of stroke victims have a warning TIAefore stroke, and waiting until symptoms occur is not

deal.18 The purpose of carotid bifurcation imaging is toetect stroke-prone carotid bifurcation plaque and iden-ify a high-risk patient likely to benefit from therapy de-igned to reduce stroke risk.

Stroke risk is dependent on many factors, but for pa-ients with carotid bifurcation disease, the most importantre a history of neurologic symptoms, the degree of stenosisf the carotid bifurcation plaque, and to a lesser extent,laque characteristics such as ulcerations, intraplaque hem-rrhage, and lipid content.

The North American Symptomatic Carotid Endarter-ctomy Trial (NASCET) and European Carotid Surgeryrial (ECST) clearly demonstrated the efficacy of CEA in

educing stroke in patients with symptoms of carotid terri-ory cerebral ischemia and carotid bifurcation stenosis thateduced luminal diameter by 50%.2-4 In these studies, the

isk of stroke was higher in patients with a clear history of

mpsi

ddstptoprhcos

umSaitsrhpoaoSCorvg

aslwbibtsstfasdlm

JOURNAL OF VASCULAR SURGERYSeptember 2011e4 Ricotta et alcarotid territory ischemic events (as opposed to amaurosisfugax), and stroke risk increased with the severity of steno-sis. Asymptomatic Carotid Atherosclerosis Study (ACAS)and Asymptomatic Carotid Surgery Trial (ACST)5,6 foundthat CEA was also effective in reducing stroke risk inpatients with asymptomatic carotid stenosis 60%, al-though the stroke risk inherent in an asymptomatic stenosiswas much less than that in a symptomatic lesion. It followsthen that neurologically symptomatic patients and neuro-logically asymptomatic patients at high risk for harboring acarotid stenosis of 60% would be candidates for carotidbifurcation imaging.

A. Indications for imaging the neurologicallysymptomatic patient

Typical carotid territory ischemic symptoms includecontralateral weakness of the face, arm, or leg, or both;contralateral sensory deficit or paresthesia of the face, arm,or leg, or both; or transient ipsilateral blindness (amaurosisfugax). If the right cerebral hemisphere is involved, othermanifestations may be noted, including anosognosia, aso-matognosia, neglect, visual, or sensory extinction. If the lefthemisphere is involved, patients may show manifestation ofaphasia, alexia, anomia, and agraphesthesia. Symptoms nottypically associated with carotid territory events includevertigo, ataxia, diplopia, visual disturbances, dysarthria,nausea, vomiting, decreased consciousness, and weakness,which may include quadriparesis.

The physical examination may show signs of stroke:facial/eyelid drooping, motor or sensory deficits, andspeech disturbances. Ocular examinations can occasionallyidentify Hollenhorst plaques. Neck auscultation may elicitcarotid bruit; however, the absence of a neck bruit does notexclude the possibility of a significant carotid bifurcationlesion. Given the incidence of significant carotid stenosis inpatients who present with stroke15,19 and the effectivenessof CEA in reducing stroke in symptomatic patients with50% carotid stenosis,2-4 it is important to evaluate thecarotid bifurcation in every patient with symptoms of ca-rotid territory ischemia.

Amaurosis fugax or the finding of a Hollenhorst plaqueon funduscopic examination, or both, is also correlatedwith the presence of significant carotid bifurcation stenosis.However, neither amaurosis fugax nor identification of aHollenhorst plaque are associated with the same stroke riskas transient cerebral ischemia.20 Identification of carotidstenosis in that clinical scenario implies a stroke risk some-where between a neurologically symptomatic patient andone who is asymptomatic.

B. Indications for imaging the neurologicallyasymptomatic patient

Evaluation and treatment of patients who are neuro-logically asymptomatic is much more controversial. Thebenefit of carotid endarterectomy for stenosis 60%, al-though statistically significant in large trials, is much lessthan for neurologically symptomatic individuals and rests

on the premise that intervention can be performed with tinimal morbidity.5,6 Identification of these asymptomaticatients may occur by routine screening using duplex ultra-ound (DUS) imaging or selective application of DUSmaging to high-risk individuals.

1. Screening for asymptomatic carotid stenosis. Toate, there is no consensus on which patients should un-ergo carotid screening for the detection of carotid steno-is. The American Society of Neuroimaging21 concludedhat the efficacy of screening would be related to therevalence of the disease in the screened populations. Whenhe prevalence of stenosis is 20%, screening reduced riskf stroke in a cost-effective manner, with intermediaterevalence of between 5% and 20%: screening reduced theisk of stroke in a cost-effective manner in some studies;owever, the benefit was usually marginal and was lost ifomplications of the intervention 5%. With a prevalencef 5%, screening has not been shown to reduce the risk oftroke in a cost-effective manner and may be harmful.

Given these assumptions, screening of the general pop-lation is not indicated. This position is supported byultiple professional organizations, including the National

troke Association,22 the Canadian Stroke Consortium23

nd the U.S. Preventive Services Task Force.24 The Amer-can Stroke Association/AHA Stroke Council25 concludedhat highly selected patient populations may benefit, butcreening of the general population for asymptomatic ca-otid stenosis was unlikely to be cost-effective and mightave the potential adverse effect of false-negative or false-ositive results. Finally, the American College of Cardiol-gy Foundation, Society for Cardiovascular Angiographynd Interventions, Society for Vascular Medicine and Biol-gy, Society of Interventional Radiology, and Americanociety of Interventional & Therapeutic Neuroradiologylinical Expert Consensus Panel on Carotid Stenting rec-mmended screening for asymptomatic patients with ca-otid bruit who are potential candidates for carotid inter-ention and for those in whom coronary artery bypassrafting (CABG) is planned.26

a. Screening patients with asymptomatic bruit. Zhund Norris,27 in the largest reported study of carotidcreening in asymptomatic patients, reported the preva-ence of carotid stenosis 75% for those with a carotid bruitas 1.2%. Although the presence of a neck bruit has noteen found to predict carotid stenosis 60% in a neurolog-

cally asymptomatic population,27 focal ipsilateral carotidruits in neurologically symptomatic patients had a sensi-ivity of 63% and a specificity of 61% for high-grade carotidtenosis (range, 70%-99%).28 The absence of a bruit did notignificantly change the probability of significant stenosis inhis group of patients (pretest, 52%; post-test, 40%). Ratch-ord et al29 found in a selected high-risk subgroup ofsymptomatic patients that if a bruit was heard, 25% had a60% stenosis. The presence of carotid bruit has been

hown to increase the absolute risk of stroke,30-32 MI, andeath.33 In general population-based studies, the preva-

ence of severe bifurcation stenosis is not high enough toake bruit alone an indication for carotid screening. Withhese facts in mind, screening should be pursued only if a

va

rhansdtirCCtlm

hsrvnstifltH4tsi

t

1

2

3

JOURNAL OF VASCULAR SURGERYVolume 54, Number 3 Ricotta et al e5bruit is associated with other risk factors for stenosis andstroke in patients who have a low operative risk5,6,12,34 andare willing to undergo carotid intervention, whether CEAor CAS.

b. Potential high-risk groups who might benefit fromscreening for asymptomatic stenosis. Two studies haveidentified specific groups among the general populationwith a higher prevalence of significant carotid stenosis thatmay 30%. Jacobowitz et al35 developed a model identify-ing patients at high risk for 50% asymptomatic carotidstenosis. The screened patients were aged 60 years andhad one or more of the following risk factors: history ofhypertension, CAD, current smoking, and a first-degreefamily relative with a history of stroke. The prevalence ofcarotid artery stenosis was only 2% if no risk factor waspresent, 6% with one risk factor, which increased to 14% fortwo risk factors, to 16% for three risk factors, and to 67% forfour risk factors.

Qureshi et al36 identified the following variables asso-ciated with 60% asymptomatic carotid stenosis: age 65years (odds ratio, 4.1), current smoking (odds ratio, 2),CAD (odds ratio, 2.4), and hypercholesterolemia (oddsratio, 1.9). Patients undergoing coronary revascularizationare another group with an increased prevalence of carotidstenosis of 2% to 27%.37,38 Overall, the prevalence ofcarotid artery stenosis among patients undergoing CABG ishigher than the general population. In patients with symp-tomatic CAD and other risk factors, such as age 65 years,history of stroke or TIA, left main coronary stenosis, dia-betes mellitus, carotid bruit, peripheral arterial disease(PAD), and previous carotid operation, it is feasible that asubset of patients with a prevalence 20% can be identifiedwho might benefit from carotid screening.37,39-45 TheACC/AHA guidelines46 note that carotid screening beforeCABG is probably indicated in the following subset ofpatients: age 65 years, left main coronary stenosis, historyof smoking, history of TIA/stroke or carotid bruit, andPAD.

Several studies47-51 have suggested that the prevalenceof 60% carotid artery stenosis among patients with symp-tomatic PAD is 20%, regardless of the patients age.However, the prevalence of 60% carotid artery stenosisamong patients with abdominal aortic aneurysms (AAA) is20%.52-54 This suggests that screening patients with AAAwould have only a modest benefit and only if interventioncould be performed with low morbidity and mortality.23

Because there is no evidence that stroke risk after AAArepair is increased by the presence of carotid stenosis,routine carotid screening of AAA patients is not indicated.

In patients with prior head and neck radiotherapy, theprevalence of significant carotid artery stenosis may be highenough, depending on the time between radiotherapy ex-posure and screening, to justify routine carotid screen-ing.55-59 The highest incidence is generally observed 15years after radiotherapy exposure, with 21.3% and 5.3%rates of ipsilateral and contralateral stenosis, respectively.The data also suggest that the ipsilateral common carotid

(CCA) and internal carotid arteries (ICA) are both in-olved. The rate of contralateral carotid artery stenosis maylso be higher than that observed in the general population.

Unfortunately, limited data are available regarding ca-otid screening after radiotherapy among patients withead and neck cancer. However, the distribution of diseasend clinical course in patients after radiation for head andeck malignancy is different from that of the typical athero-clerotic population. There is a higher incidence of diffuseisease, often involving the CCA, and many of these pa-ients remain neurologically asymptomatic. Further, CEAn this group is considered relatively high risk, and prioradiotherapy is a relative indication for CAS rather thanEA. There are no robust data on the long-term results ofAS in asymptomatic stenosis associated with prior radio-

herapy. Therefore, issues other than the increased preva-ence of disease must be considered in formulating recom-

endations concerning screening in this group of patients.Brain imaging will occasionally identify patients who

ave evidence of focal cerebral infarction despite the ab-ence of any history of neurologic symptoms and a normalesult on the neurologic examination. These infarcts canary in size and are often found in the frontal lobes or theondominant temporal lobe. They may occur as smallymmetric lacunar infarcts, implying small-vessel disease, orhey may also be asymmetric, which tends to implicatepsilateral carotid stenosis. These can be secondary to bloodow changes distal to carotid occlusion, which may increasehe risk of lacunar infarcts in those with small vessel disease.owever, Kakkos et al60 reported a higher stroke rate of

.4% vs 1.3% in patients with 60% to 79% clinically asymp-omatic stenosis if a silent infarct was present. Carotidcreening is recommended in patients with asymptomaticnfarctions.

Recommendations for the use of carotid bifurca-ion imaging

. Imaging of the cervical carotid artery is recommendedin all patients with symptoms of carotid territory isch-emia. This recommendation is based on the significantincidence of clinically relevant carotid stenosis in thispatient group and the efficacy of CEA for clinicallysignificant lesions in reducing overall stroke (GRADE 1,Level of Evidence A).

. Imaging should be strongly considered for patients whopresent with amaurosis fugax, evidence of retinal arteryembolization on funduscopic examination, or asymp-tomatic cerebral infarction, and are candidates for CEA.This recommendation is based on the intermediatestroke risk in this group of patients and the efficacy ofCEA in reducing risk of subsequent stroke (GRADE 1,Level of Evidence A).

. Routine screening is not recommended to detect clini-cally asymptomatic carotid stenosis in the general pop-ulation. Screening is not recommended for presence of aneck bruit alone without other risk factors. This recom-mendation is based on the low prevalence of disease inthe population at large, including those with neck

bruits, as well as the potential harm of indiscriminate

ss7smlsimz

fhtimitsw

ao4loqhad

p(raw

A

vsutdvAsu

l(csCifI

JOURNAL OF VASCULAR SURGERYSeptember 2011e6 Ricotta et alapplication of carotid bifurcation intervention to a largenumber of asymptomatic individuals (GRADE 1, Levelof Evidence A).

4. Screening for asymptomatic clinically significant carotidbifurcation stenosis should be considered in certaingroups of patients with multiple risk factors that increasethe incidence of disease as long as the patients are fit forand willing to consider carotid intervention if a signifi-cant stenosis is discovered. The presence of a carotidbruit in these patients increases the likelihood of asignificant stenosis (GRADE 1, Level of Evidence B).Such groups of patients include:a. Patients with evidence of clinically significant periph-

eral vascular disease regardless of age.b. Patients aged 65 years with a history of one or

more of the following atherosclerotic risk factors:CAD, smoking, or hypercholesterolemia. In general,the more risk factors present, the higher the yield ofscreening should be expected.

5. Carotid screening may be considered in patients beforeCABG. This is most likely to be fruitful if the patientsare aged 65 years and have left main disease or ahistory of peripheral vascular disease. The strongestindication for screening these patients from the dataavailable is to identify patients at high risk for perioper-ative stroke (GRADE 2, Level of Evidence B).

6. Carotid screening is not recommended for patients withAAA who do not fit into one of the above categories(GRADE 2, Level of Evidence B).

7. Carotid screening is not recommended for asymptom-atic patients who have undergone prior head and neckradiotherapy. Although the incidence of disease is in-creased in this group of patients, the utility of interven-tion in the absence of neurologic symptoms has notbeen clearly established (GRADE 2, Level of Evidence B).

II. SELECTING IMAGING MODALITIES FORCAROTID EVALUATION

The two most important features of carotid bifurcationatheroma are the degree of diameter stenosis and the char-acter of the bifurcation plaque. In addition to informationabout the carotid bifurcation, there are clinical scenarioswhere the clinician requires information on the status of thevessels proximal or distal to the cervical carotid artery.These factors need to be considered when choosing be-tween imaging studies. It is commonbut not universalto use multiple modalities when evaluating a patient withsuspected cervical carotid stenosis.

In NASCET2,3 and ECST,4 a higher degree of stenosisin symptomatic patients was associated with a higher strokerisk. The ACAS5 found no correlation between the severityof carotid stenosis and the incidence of stroke; however,there were too few strokes in this study to permit a sub-group analysis of the effect of degree of stenosis on theability to benefit from CEA. Angiographic data from theECST study61 on contralateral asymptomatic carotid arter-ies from 2295 patients demonstrated a 2% annual stroke

risk in patients with 70% neurologically asymptomatic otenosis. Asymptomatic lesions with greater degrees oftenosis had a greater risk of stroke: 9.8% for patients with0% to 79% stenosis and 14.4% for those with 80% to 99%tenosis. These data suggest that the degree of stenosis is aarker of stroke risk in symptomatic and asymptomatic

esions. Pathologic studies have demonstrated that moretenotic carotid plaques are more likely to have ulceration,ntraplaque hemorrhage, and intraluminal thrombus for-

ation, all of which are clearly related to cerebral emboli-ation and stroke.62

Plaque morphology is an important feature in assessinguture risk of neurologic events. Heterogeneous plaquesave been shown to increase the risk of neurologic symp-oms (TIA/stroke)63,64 and were also associated with anncidence of TIA/stroke that was higher than that in ho-

ogenous plaques for all grades of stenosis.64 Using DUSmaging and computerized image analysis65 quantifyinghe gray scale median of the plaque, Biasi et al66 demon-trated that gray scale median values of 25 were associatedith an increased stroke risk of carotid stenting procedures.

Nicolaides et al67 recently concluded that morphologicssessment of plaque structure may allow the identificationf a subgroup of asymptomatic carotid stenoses with a.5-fold increase in the risk of developing ipsilateral neuro-

ogic symptoms compared with those with a similar degreef stenosis, which will reduce the number of patients re-uiring intervention to prevent one stroke. At present,owever, this type of plaque analysis66,67 is not widelyvailable and requires further prospective evaluation toetermine its ultimate clinical utility.

The imaging modalities most often used to evaluateatients for cervical carotid stenosis are carotid DUSCDUS), magnetic resonance imaging (MRI) and angiog-aphy (MRA), computed tomography angiography (CTA),nd digital subtraction angiography (DSA). Each of theseill be discussed in turn.

. Carotid duplex ultrasound imaging

DUS imaging provides an accurate and reliable nonin-asive tool to determine the degree of cervical carotidtenosis and plaque morphology in most patients. It issually the initial study in patients who present with symp-oms or a carotid bruit. Because the study is highly depen-ent on technique, testing should be done in an accreditedascular laboratory (eg, Intersocietal Commission for theccreditation of Vascular Laboratories), and the images

hould be reviewed by physicians experienced in vascularltrasound interpretation.

Determining the degree of carotid artery stenosis isargely based on an analysis of the peak systolic velocityPSV) or the end-diastolic velocity (EDV), or both, of thearotid artery. A panel of experts from several medicalpecialties convened in October 2002 in San Francisco,alifornia, under the auspices of the Society of Radiologists

n Ultrasound, to arrive at a consensus regarding the per-ormance of Doppler US imaging to aid in the diagnosis ofCA stenosis.68 This panel of experts recommended a cut-

ff PSV of the ICA of 125 cm/s for predicting angio-

cdawunr

ciihm6wrctoc

ciwbWsfi

C

sCtmmeMscqaSb

asiihilto(cb

JOURNAL OF VASCULAR SURGERYVolume 54, Number 3 Ricotta et al e7graphic 50% stenosis and 230 cm/s for predicting70% ICA stenosis. These recommended criteria are basedon an analysis of several published studies and the experi-ence of the panelists rather than values validated againstother imaging modalities.

AbuRahma et al69 recently analyzed the CDUS andangiography results of 376 carotid arteries in their institu-tion. Using the consensus criteria, they demonstrated asensitivity of 93%, specificity of 68%, and overall accuracy of85% for stenosis between 50% and 69%. A PSV of 230cm/s for 70% stenosis had a sensitivity of 99%, specificityof 86%, and overall accuracy of 95%. Receiver operatorcurves showed that the ICA PSV was significantly betterthan EDV or ICA/CCA ratio (P .036) in detecting70% stenosis and 50% stenosis. There was no improve-ment in accuracy by adding the EDV values or the ratios, orboth, to the PSV values.

Velocity-based estimation of carotid artery stenosis mayneed to be adjusted in certain circumstances, for example,higher velocities in women than in men and higher veloci-ties in the presence of contralateral carotid artery occlu-sion.70,71 High carotid bifurcation, severe arterial tortuos-ity, extensive vascular calcification, and obesity may alsoreduce the accuracy of DUS imaging. Carotid stents willdecrease compliance of the vessel wall and flow velocity.72

DUS imaging may also fail to differentiate between subto-tal and total carotid occlusion. Intravenous administrationof contrast agents may improve diagnostic accuracy,73,74

but the safety of these agents has been questioned.Power Doppler and contrast DUS imaging can be used

to differentiate between preocclusive stenosis and completeocclusion.75 Overall, each vascular laboratory should havein place an internal validation process of their own criteriafor their internal use.

DUS imaging of the carotid artery has two majorlimitations: quality dependence on the technicians exami-nation and limitations of visualization of the proximalcarotid artery and intracranial portions. Although the intra-cranial cerebral arteries can be assessed with transcranialDoppler imaging, this technique is not as widely available atmost institutions as other imaging modalities.

In addition to determining percent stenosis, DUS, asnoted above, has been used to determine plaque character-istics (echogenicity) using gray scale median values, whichpredict the stroke risk of a particular plaque. Plaque char-acterization is not routine in every vascular laboratory andrequires specific protocols to assure standardization of re-sults.

B. Magnetic resonance imaging and angiography

MRA has the advantage of being noninvasive, does notrequire iodinated contrast or ionizing radiation, and pro-vides an unlimited number of projections of the carotidlumen from a single acquisition. MRA can also assessintrathoracic and intracranial lesions that are not amenableto DUS interrogation. MRA does not visualize the sur-rounding soft tissue structures, unless additional MRI is

performed, and calcium within the plaque is not defined. It cannot be used in patients with implanted ferromagneticevices, such as implantable defibrillators and pacemakers,nd is of limited use in uncooperative patients and thoseith claustrophobia. The gadolinium-based compoundssed as a contrast agent for MRA have been associated withephrogenic systemic fibrosis in patients with pre-existingenal disorders.76

MRA has a tendency to overestimate the degree ofarotid stenosis. The sensitivity and specificity for diagnos-ng 70% to 99% stenosis with time-of-flight MRA aredentical to DUS imaging (88% and 84%, respectively);owever, MRA has a tendency to over-read stenosis,aking it difficult to differentiate more moderate (50% to9%) from severe stenosis. Similarly, high-grade stenosisill result in a loss of signal on MRA. This does not

epresent a carotid occlusion when the more distal cervicalarotid is visualized. However, when there is no reconsti-ution of the cervical carotid artery on MRA, the diagnosisf carotid occlusion can be made with a high degree ofertainty.76,77

MRI can be used to analyze plaque morphology, spe-ifically the structure of the atherosclerotic plaque. It candentify the lipid-rich necrotic core and the fibrous capsuleith high sensitivity and specificity78 and can distinguishetween an intact thick, thin, or ruptured fibrous cap.79

hen dedicated protocols are used, MR also can demon-trate specific plaque components, including calcium, lipid,brocellular element, or thrombus within the plaques.

. Computed tomography angiography

CTA is less susceptible than MRA to overestimating theeverity of carotid stenosis. The rapid acquisition of spiralT images allows excellent timing with contrast adminis-

ration and provides quality images that can be viewed inultiple planes. CTA is extremely fast and offers submilli-eter spatial resolution (0.3 vs 0.8 mm for contrast-

nhanced MRA), is less expensive than contrast-enhancedRA, provides a faster processing time, and can visualize

oft tissue, bone, and blood vessels at the same time. CTAan also demonstrate vascular anomalies, has the ability touantify the extent of calcification, and can interrogate therterial tree from the aortic arch to the circle of Willis.tenoses can be measured with electronic microcalipersased on NASCET or ECST methods.80

A meta-analysis of 28 studies analyzing the diagnosticccuracy of CTA compared with DSA showed a pooledensitivity of 85% and specificity of 93% for CTA in detect-ng 70% to 99% carotid stenosis and a sensitivity and spec-ficity for occlusions of 97% and 99%.81 CTA was alsoighly accurate in identifying calcification but less reliable

n describing carotid plaque morphology, specifically theipid component, or ulceration. CTA appears less reliablehan DUS imaging or MRA for assessing plaque morphol-gy.82 Other limitations of this technique include costcompared with DUS), contrast exposure, and the addedoncern of radiation exposure. In addition, a large calciumurden can limit the ability to distinguish contrast from

alcium during postprocessing imaging.

iCnsa1tclysa

tnrsteol

rhadtdbtcrfiC

i

1

2

3

4

JOURNAL OF VASCULAR SURGERYSeptember 2011e8 Ricotta et alD. Catheter-based DSA

Many authorities still consider carotid conventionaldigital angiography to be the gold standard against all otherimaging modalities in patients with extracranial cerebrovas-cular disease. Measurement of carotid stenosis using DSA isgenerally done using the NASCET method.2 Conventionalangiography is generally reserved for patients with conflict-ing imaging studies before CEA or in patients consideredfor CAS. DSA provides high-quality imaging that is accu-rate, objective, and easy to interpret. It can identify lesionsfrom the aortic arch to the intracranial vessels. Major limi-tations of angiography that make it inappropriate as ascreening modality include its cost and associated risks,specifically of stroke.83-85 Overall, DSA is most useful inpatients when less invasive imaging studies produce con-flicting results. When DUS imaging is equivocal, DSA ispreferred over CT and MR in evaluating patients with renaldysfunction (by minimizing contrast load), obesity, or in-dwelling ferromagnetic material, which render CTA orMRA technically inadequate or difficult.

E. Comparison of CDUS, MRA, CTA, and DSA

The U.K. Health Technology Assessment concludedthat although contrast-enhanced MRA was the most accu-rate imaging modality overall, it was limited by unavailabil-ity, inaccessibility, and delays. Therefore, they concludedthat color DUS imaging remained the preferred imagingmodality for identifying patients with 70% to 99% steno-sis.86 As such, CDUS is the preferred imaging modality forthe identification of asymptomatic stenosis.

This recommendation was based on several factors,including low cost, a much higher number of strokes likelyto be prevented in the long-term by the rapid availability ofCDUS imaging in contrast with other imaging, and thegood sensitivity of imaging in detecting significant stenosis.However, the Health Technology Assessment highlightedthe concern of the accuracy of DUS in diagnosing 50% to69% stenosis, which carries a sensitivity of only 36% and aspecificity of 91%.86

The utility of CDUS will depend on the clinical presen-tation of the patient. In neurologically symptomatic pa-tients, a diagnosis by CDUS of stenosis between 50% and69% is sufficient to proceed with surgery based on itsspecificity. However, a negative CDUS result would man-date another imaging study because of the low sensitivity ofCDUS in this setting. In neurologically asymptomatic pa-tients, a moderate stenosis (50% to 69%) diagnosed byCDUS should be confirmed by another imaging studybefore intervention is undertaken.

F. Imaging after carotid intervention

The prevalence of carotid artery restenosis after CEAvaries between 1% and 37%,87-90 although symptomaticrecurrent stenoses is infrequent (0% to 8%).87 Factors asso-ciated with restenosis include continued smoking, smallICA diameter, operative defect detected at intraoperative

assessment, and primary closure after CEA. The aggregatencidence of residual and recurrent carotid stenosis afterEA in ACAS was 13%.88 Of 136 patients who had reste-osis, 8 (5.9%) underwent reoperation, only 1 of whom wasymptomatic. There was no correlation between late strokend recurrent stenosis. Similarly, Cao et al89 randomized353 patients who underwent CEA. Of these, the eversionechnique was used in 678, and standard CEA with primarylosure was done in 419 and patch closure in 256. Theife-table estimate of the cumulative risk of restenosis at 4ears was 4% in the eversion CEA group and 9% in thetandard CEA group, and 98% of these patients weresymptomatic.

Several studies have reported the progression of con-ralateral stenosis after CEA.91-93 Contralateral carotid ste-osis progression was more frequent than ipsilateral recur-ent stenosis during the long-term follow-up in thesetudies. These studies also identified that the risk of con-ralateral carotid artery stenosis progression depends on thexisting disease at the time of the initial CEA.91-93 The riskf progression for moderate stenosis at the initial surveil-

ance to severe stenosis can be as high as five times.92

Several large prospective studies94-99 have analyzed theate of carotid in-stent restenosis after CAS. More patientsad 70% stenosis of the ipsilateral carotid artery 1 yearfter CAS than after CEA (19% vs 5%). Use of CDUS toiagnose post-CAS restenosis is confounded by changes inhe velocity criterion caused by the stent itself, and standardiagnostic criteria do not apply. Artifacts associated withoth CTA and MRA similarly limit the utility of theseechniques in the post-CAS patient. DSA is required toonfirm restenosis after CAS identified by CDUS wheneintervention is contemplated. In contrast, CDUS is suf-cient to diagnose and plan therapy for restenosis afterEA. Recommendations for selection of carotid imag-

ng modalities

. CDUS in an accredited vascular laboratory is the initialdiagnostic imaging of choice for evaluating the severityof stenosis in symptomatic and asymptomatic patients.Unequivocal identification of stenosis of 50% to 99% inneurologically symptomatic patients or 70% to 99% inasymptomatic patients is sufficient to make a decisionregarding intervention (GRADE 1, Level of Evidence A).

. CDUS in an accredited vascular laboratory is the imag-ing modality of choice to screen asymptomatic popula-tions at high risk (GRADE 1, Level of Evidence B).

. When CDUS is nondiagnostic or suggests stenosis ofintermediate severity (50% to 69%) in an asymptomaticpatient, additional imaging with MRA, CTA or DSAis required before embarking on any intervention(GRADE 1, Level of Evidence B).

. When evaluation of the vessels proximal or distal to thecervical carotid arteries is needed for diagnosis or to plantherapy, imaging with CTA, MRA, or catheter angiog-raphy in addition to CDUS is indicated. CTA is prefer-able to MRI or MRA for delineating calcium. When

there is discordance between two minimally invasive

Hfbr

brairJErasdpb

B

gisRaaUt(VEcrAilatt

C

cwcalpesc

ratg

JOURNAL OF VASCULAR SURGERYVolume 54, Number 3 Ricotta et al e9imaging studies (CDUS, MRA, CTA), DSA is indicatedto resolve conflicting results. DSA is generally reservedfor situations where there is inconclusive evidence ofstenosis on less invasive studies or when CAS is planned(GRADE 1, Level of Evidence B).

5. A postoperative DUS study 30 days is recommendedto assess the status of the endarterectomized vessel. Inpatients with 50% stenosis on this study, furtherfollow-up imaging to assess progression or resolution isindicated. In patients with a normal DUS study resultand primary closure of the endarterectomy site, ongoingimaging is recommended to identify recurrent stenosis.In patients with a normal DUS after patch or eversionendarterectomy, further imaging of the endarterecto-mized vessel may be indicated if the patient has multiplerisk factors for progression of atherosclerosis. There areinsufficient data to make recommendations on imagingafter CAS (GRADE 2, Level of Evidence C). Althoughthe data in this area are not robust concerning intervalsfor follow-up imaging, the committee was unanimous inthis recommendation, recognizing that follow-up DUScarries little risk.

6. Imaging after CAS or CEA is indicated to monitorcontralateral disease progression in patients with con-tralateral stenosis 50%. In patients with multiple riskfactors for vascular disease, follow-up DUS may beindicated with lesser degrees of stenosis. The likelihoodof disease progression is related to the initial severity ofstenosis (GRADE 2, Level of Evidence C). Althoughthe data in this area are not robust concerning intervalsfor follow-up imaging, the committee was unanimous inthis recommendation, recognizing that follow-up DUScarries little risk.

III. MEDICAL MANAGEMENT OF PATIENTSWITH CAROTID STENOSIS

Optimal medical management is an important part ofoverall treatment of all patients with carotid bifurcationdisease, regardless of the degree of stenosis or the plan forintervention. This therapy is directed both at the reductionof stroke and overall cardiovascular events, including car-diovascular-related mortality. The best medical manage-ment for stroke prevention was highlighted in clinical prac-tice guidelines issued jointly in 2006 by the AHA and theAmerican Stroke Association, and cosponsored by theCouncil on Cardiovascular Radiology and Intervention andthe American Academy of Neurology.100

A. Treatment of hypertension

Elevated blood pressure increases the risk for stroke,101

and reducing blood pressure decreases the risk forstroke.102 The relationship between blood pressure andstroke risk is continuous, consistent, and independent ofother risk factors.103 The Framingham Heart study,104

the Atherosclerosis Risk in Communities (ARIC) study,105

and the Cardiovascular Health Study106 each found thathypertension was independently associated with an ele-

vated risk for carotid artery atherosclerosis. Each 10-mm (g increase in blood pressure results in an increase in riskor stroke of 30% to 45%. Each 10-mm Hg reduction inlood pressure amongst hypertensive patients decreases theisk for stroke by 33%.102

Lowering blood pressure to a target 140/90 mm Hgy lifestyle interventions and antihypertensive treatment isecommended in individuals who have hypertension withsymptomatic carotid atherosclerosis.100 Aggressive lower-ng of blood pressure may harm patients who have had aecent stroke by reducing cerebral perfusion. In fact, theoint National Committee for the Prevention, Detection,valuation, and Treatment of High Blood Pressure-VII has

emained ambiguous regarding recommendations forntihypertensive management in patients with a recenttroke.103 However, antihypertensive therapy aimed at re-ucing blood pressures to 140/90 is recommended foratients who have had an ischemic stroke or TIA and areeyond the hyperacute period.100

. Treatment of diabetes mellitus

In the Cardiovascular Health Study, an elevated fastinglucose level was associated with an increased risk for stroken patients with carotid atherosclerosis.107 The Insulin Re-istance Atherosclerosis Study108 and the Atherosclerosisisk in the Community study109 showed that diabetes wasssociated with intima-media thickness of the carotid arterynd with progression in intima-media thickness. Thenited Kingdom Prospective Diabetes Study (UKPDS),110

he Action to Control Cardiovascular Risk in DiabetesACCORD)111 study, and the Action in Diabetes andascular Disease: Preterax and Diamicron MR Controlledvaluation (ADVANCE)112 trial all tested whether tightontrol of serum glucose levels in diabetic patients wouldeduce the risk for stroke. Despite achieving hemoglobin1C levels 6.5%, no reduction in stroke risk was identified

n these trials. Glucose control to nearly normoglycemicevels (target hemoglobin A1C 7%) is recommendedmong diabetic patients to reduce microvascular complica-ions and, with lesser certainty, macrovascular complica-ions other than stroke.

. Treatment of lipid abnormalities

The relationship between elevated cholesterol and in-ident MI in patients with coronary artery atherosclerosis isell established; however, the relationship between hyper-

holesterolemia and incident stroke is less clear. A meta-nalysis of 45 studies of strokes in patients with hypercho-esterolemia did not suggest an increased risk for stroke inatients with elevated serum cholesterol.113 However, sev-ral other prospective studies in men and women haveubsequently identified an increase in incident stroke asso-iated with elevated cholesterol levels.114-116

Patients with known atherosclerosis have demonstratededuced stroke rates when treated with lipid-lowering ther-py. The Multiple Risk Factor Intervention Trial of statinherapy in hypercholesterolemic patients observed thatreater reductions in levels of low-density lipoprotein

LDL) were associated with a reduction in stroke risk. A

Pc

E

icdTcmapcASeaw

pccat

rEornwdbwa

aCEtRaepdd

tsddfi

lw

JOURNAL OF VASCULAR SURGERYSeptember 2011e10 Ricotta et almeta-analysis of 26 trials observed that the risk of strokedecreased by 15% for every 10% reduction in serumLDL117 in patients with known coronary or other athero-sclerosis.

Further meta-analyses of statin trials that also reportedon stroke as an outcome have shown that statin therapyreduces the risk of stroke by 15% to 30%.118 The StrokePrevention by Aggressive Reduction in Cholesterol trialfound that atorvastatin treatment of patients with a recentstroke or TIA resulted in a reduction in stroke rate by 16%over 5 years.119

It is less clear if aggressive lipid-lowering therapy resultsin regression of carotid artery atherosclerosis. Aggressivestatin therapy in the Measuring Effects on Intima-MediaThickness: an Evaluation of Rosuvastatin study,120 theArterial Biology for the Investigation of the TreatmentEffects of Reducing Cholesterol121 study, and the Atorva-statin vs Simvastatin on Atherosclerosis Progression122

study all showed increased regression of the carotid arteryintima-media thickness compared with controls.

Although LDL is the primary determinant of cardio-vascular and stroke risk, low levels of high-density lipopro-tein (HDL) cholesterol also influence stroke risk, and ele-vation of HDL has been shown to reduce the risk ofstroke.123 In an analysis of a large (9200) series of pa-tients, treated by dual therapy aimed at decreasing LDL andraising HDL, elevation of HDL level was independentlyassociated with a reduction of stroke risk by a factor of 0.86.Conversely, an elevated total cholesterol/HDL ratio in-creased stroke risk by a factor of 1.22. Reducing cholesterolabsorption has also been shown to reduce stroke risk inpatients with familial hypercholesterolemia.124 Overall,however, the HDL level had much less effect on stroke riskthan did the level of LDL.

Elevated cholesterol, comorbid CAD, or evidence ofan atherosclerotic etiology of carotid stenosis should bemanaged according to National Cholesterol EducationProgram-Adult Treatment Panel III guidelines,125 whichinclude lifestyle modification or medications, or both. Sta-tin agents are recommended targeting LDL of 100 mg/dL, for those with coronary heart disease or symptomaticatherosclerotic disease, and LDL of 70 mg/dL for veryhigh-risk persons with multiple risk factors.

D. Smoking cessation

Smoking nearly doubles the risk of stroke.126,127

Smoking also acts synergistically on other risk factors thatare known to increase the risk of stroke, such as CAD andPAD. Conversely, smoking cessation results in a reductionin risk for CAD and for coronary mortality.128 Cessationalso reduces the risk of stroke in men and women.128-130

Counseling and smoking cessation medications are effec-tive in helping smokers to quit. Physician counseling is animportant and effective intervention that reduces smokingin patients by 10% to 20%131 but continues to be under-used.132 Nicotine replacement therapy, in the form of

patches and gums, is effective in reducing smoking.133 ratients with extracranial carotid stenosis who are smokingigarettes should be counseled to quit.

. Antithrombotic treatment

No adequately powered studies have been performedn asymptomatic patients with carotid atherosclerosis toonfirm a benefit with antithrombotic treatment in re-ucing incident stroke. The US Preventative Servicesask Force has recommended daily aspirin as cardiovas-ular prophylaxis in patients with anticipated cardiacorbidity of 3% for men aged 45 years and in women

ged 55 years.134 These recommendations are basedrimarily on an observed reduction in overall cardiovas-ular morbidity and death with aspirin therapy. TheHA Primary Prevention of Cardiovascular Disease andtroke agrees with this recommendation.135 There is novidence to suggest that antiplatelet agents other thanspirin have improved benefit in asymptomatic patientsith carotid atherosclerosis.

Evidence for antithrombotic treatment for secondaryrevention of recurrent stroke in symptomatic patients witharotid atherosclerosis is more robust.100,136-139 Thehoice of antiplatelet therapy among aspirin, clopidogrel,nd dipyridamole plus aspirin is not clearly defined becausehe data are uncertain.

The Clopidogrel in Unstable Angina to Prevent Recur-ent Events140 and Clopidogrel for the Reduction ofvents During Observation141 trials each showed a benefitf clopidogrel plus aspirin compared with aspirin alone ineducing vascular events in patients with prior acute coro-ary syndromes. The Management of Atherothrombosisith Clopidogrel in High-risk Patients142 subsequentlyemonstrated no significant difference in vascular eventsetween symptomatic patients with carotid stenosis treatedith clopidogrel plus aspirin compared with clopidogrel

lone.Clopidogrel alone was initially shown to have a small

dvantage over aspirin in this subset of patients in thelopidogrel versus Aspirin in Patients at Risk of Ischaemicvents143 trial. However, it costs more than aspirin, and

he subsequent Clopidogrel for High Atherothromboticisk and Ischemic Stabilization, Management, and Avoid-nce144 trial showed that clopidogrel plus aspirin wasquivalent to aspirin alone in preventing vascular events inatients with a prior stroke, TIA, or other cardiovascularisease or in patients with high risk for cardiovascularisease.

Antiplatelet agents are therefore recommended for pa-ients with non-cardioembolic ischemic stroke or TIA as-ociated with carotid atherosclerosis. Aspirin (50-325 mg/), the combination of aspirin and extended-releaseipyridamole, and clopidogrel, are all acceptable optionsor initial therapy; a combination of aspirin and clopidogrels not recommended.

Aspirin is currently the most commonly used antiplate-et agent and one of the most frequently prescribed drugs,ith as many as 30 million Americans on long-term aspirinegimens. A growing body of evidence suggests that some

pGci

arAbwfdCelah

demdpscit

ridsiourarTrttm

H

aibmrdpr

JOURNAL OF VASCULAR SURGERYVolume 54, Number 3 Ricotta et al e11patients compliant with aspirin therapy may still developatherothrombotic complications, such as stroke. Reducedantiaggregant effect is more common in patients takinglower-dose aspirin (81 mg) or 325 mg aspirin with anenteric coating than it is in patients taking 325 mg ofnoncoated aspirin daily.145 The lack of consensus for thedefinition of aspirin resistance and for the specific labora-tory test to identify it has led to large variability in itsreported prevalence.146 The routine laboratory evaluationof platelet reactivity is not justifiable. The pharmacologicresponse to clopidogrel also demonstrates significant inter-individual variability.

Patients with reduced platelet inhibition in response toclopidogrel appear to be at increased risk for cardiovascularevents. This resistance may result from reduced bioavail-ability, polymorphisms of cytochrome P450, additionalgenetic variants, or increased platelet turnover.147 Prasug-rel is an alternate platelet-inhibiting pharmacologic agentof the same class as clopidogrel that does not have theselimitations. It has been approved for use in acute coronarysyndromes,148 but no data are available on its use for strokeprevention.

As in the case of aspirin resistance, current evidencedoes not suggest the routine use of platelet function orgenetic testing for clopidogrel resistance. In the absence ofrandomized trial data, the general approach to patientsdeveloping clinical events while taking aspirin or clopi-dogrel has been to confirm compliance and increase thedosage, followed by the addition or substitution of anotherantiplatelet agent.149 However, data in this area are insuf-ficient to allow clear recommendations.

F. Anticoagulant therapy

Parenteral and oral anticoagulants are effective in theprevention of embolic stroke in patients with atrial fibrilla-tion or prosthetic heart valves. However, warfarin antico-agulation has been shown to be less effective than antiplate-let therapy for secondary prevention of neurologic events inpatients with carotid atherosclerosis who do not have ahistory of atrial fibrillation and is not indicated in patientswith symptoms of cerebral ischemia.150,151 The Warfarin-Aspirin Reduced Cardiac Ejection Fraction Study trial iscurrently investigating the potential advantage in prevent-ing cardioembolic stroke of anticoagulation vs aspirin inpatients with chronic congestive heart failure.152

G. Medical management for the perioperative periodof CEA

Hypertension is a common comorbidity in patientsundergoing CEA. Blood pressure fluctuations, bothabove and below normal, are a significant source ofmorbidity and may contribute to MI and postoperativereperfusion syndrome. Careful periprocedural bloodpressure management is critical to obtaining optimalresults from the operation. Although the most recentAHA guidelines do not classify CEA as a high-risk sur-gical procedure mandating -blockade, they do indicate

that all patients with known CAD should receive t-blockade therapy before CEA to achieve a stable bloodressure and heart rate of 60 to 80 beats/min.153,154

iven the near ubiquity of this condition in patients witharotid stenosis, -blockade is nearly universally requiredn this patient group.

Patients taking combined aspirin and clopidogrel ther-py in the perioperative period have a 0.4% to 1.0% higherisk of major bleeding compared with aspirin alone.155

spirin therapy alone does not have to be discontinuedefore CEA.156 The risks of periprocedural MI from aspirinithdrawal outweigh the risk of fatal or severe bleeding

rom aspirin use. The ACC Perioperative Guidelines en-orses the continued use of aspirin before and afterEA.153 A low dose (81 to 325 mg) appears at least as

ffective as higher doses, and higher doses may in fact beess effective.157 Patients should continue aspirin therapyfter CEA indefinitely, according to recommendations forigh-risk patients with atherosclerosis.100,136-139

There has been a consensus that preprocedural clopi-ogrel should be stopped approximately 5 days beforelective CABG.46 Recent data from a large, retrospective,ulticentered clinical experience158 suggest that clopi-ogrel may be safely continued through the perioperativeeriod without increased bleeding risk. It is therefore rea-onable to individualize the management of perioperativelopidogrel therapy. There is no clear information regard-ng the risks or benefits of continued clopidogrel mono-herapy in the periprocedural period for CEA.

One meta-analysis showed preoperative statin therapyesulted in a significant reduction in perioperative mortalityn patients undergoing vascular surgery.159 One small ran-omized trial found that perioperative death, MI, andtroke in patients undergoing vascular surgery was reducedn the group treated with atorvastatin.160 In one largebservational study of hospital records of 780,591 patientsndergoing noncardiac surgery, the risk-adjusted mortalityate was significantly lower in those who received perioper-tive statins than in those who were not taking statins (oddsatio, 0.62; 95% confidence interval [CI], 0.58-0.67).161

he evidence for continued use of statin therapy currentlyemains largely observational. Furthermore, the optimalime for starting therapy, the duration of therapy, dose, orarget LDL levels to be achieved still remain to be deter-ined.

. Medical management for the perioperative periodof CAS

Antihypertensive, -blocker, and lipid-lowering ther-py should be initiated in patients undergoing CAS accord-ng to the same recommendations for CEA. Patients shoulde started on dual antiplatelet therapy with aspirin (325g) and clopidogrel (75 mg) or ticlopidine (250 mg). No

andomized trial has yet compared CAS performed withual-antiplatelet therapy vs aspirin alone. However, theublished periprocedural stroke, MI, and death rates in allecent clinical trials have been achieved with this combina-

ion therapy.9,10,162,163 Dual-antiplatelet therapy should

ppemScarC

A

mripiridab

faFdscscsCe

raocisauto

pdilwhcoow

JOURNAL OF VASCULAR SURGERYSeptember 2011e12 Ricotta et albe continued for 1 month after the procedure, and aspirinshould be continued indefinitely.

Recommendations for medical management ofpatients with carotid atherosclerosis

1. In patients with carotid artery stenosis, treatment ofhypertension, hypercholesterolemia, and efforts atsmoking cessation are recommended to reduce overallcardiovascular risk and risk of stroke regardless ofwhether intervention is planned. Targets are those de-fined by the National Cholesterol Education Programguidelines (GRADE 1, Level of Evidence A).

2. Aggressive treatment of hypertension in the setting ofacute stroke is not recommended; however, treatment ofhypertension after this period has passed is associatedwith reduced risk of subsequent stroke. The target pa-rameters are not well defined (GRADE 1, Level ofEvidence C).

3. Treatment of diabetes with the goal of tight glucosecontrol has not been shown to reduce stroke risk ordecrease complication rates after CEA and is not recom-mended for these purposes (GRADE 2, Level of Evi-dence A).

4. Anticoagulation is not recommended for the treatmentof TIA or acute stroke, unless there is evidence of acardioembolic source (GRADE 1, Level of Evidence B).

5. Antiplatelet therapy in asymptomatic patients with ca-rotid atherosclerosis is recommended to reduce overallcardiovascular morbidity, although it has not beenshown to be effective in the primary prevention of stroke(GRADE 1, Level of Evidence A).

6. Antiplatelet therapy is recommended for secondarystroke prevention: aspirin, aspirin combined with dipy-ridamole, and clopidogrel are all effective. Clopidogrelcombined with aspirin is not more effective than eitherdrug alone (GRADE 1, Level of Evidence B).

7. Perioperative medical management of patients under-going carotid revascularization should include bloodpressure control (140/80 mm Hg), -blockade(heart rate, 60-80 beats/min), and statin therapy (LDL100 mg/dL) (GRADE 1, Level of Evidence B).

8. Perioperative antithrombotic therapy for CEA shouldinclude aspirin (81-325 mg) (GRADE 1, Level of Evi-dence A). The use of clopidogrel in the perioperativeperiod should be decided case-by-case (GRADE 2 Levelof Evidence B).

9. Perioperative antithrombotic management of CAS pa-tients should include dual-antiplatelet therapy with as-pirin and ticlopidine or clopidogrel. Dual-antiplatelettherapy should be initiated at least 3 days before CASand continued for 1 month, and aspirin therapyshould be continued indefinitely (GRADE 1, Level ofEvidence C).

IV. TECHNICAL RECOMMENDATIONS FORCAROTID INTERVENTIONS

The efficacy of carotid interventions depends on mini-

mizing perioperative complication rates. This involves ap- eropriate risk factor assessment and patient selection,erioperative therapy, and performance of a technicallyxcellent operation. Perioperative therapy and medicalanagement have been discussed in the previous sections.

pecific recommendations on techniques to reduce compli-ations of CEA and CAS are beyond the scope of this reportnd can be found elsewhere.164,165 However, some generalecommendations can be made regarding the conduct ofEA and CAS.

. Carotid endarterectomy

Among the variables that have been studied to deter-ine their effect on the outcome of CEA are local vs

egional anesthesia, routine vs selective use of shunts, mon-toring of brain function during the procedure, routineatch closure after endarterectomy, and completion imag-ng. Although several authors have suggested that use ofegional or local anesthesia is associated with a reducedncidence of perioperative hemodynamic changes and car-iac events, a prospective randomized trial166 and system-tic review of the literature167 failed to show any differenceetween the two anesthetic approaches.

An abundant literature exists on the indications identi-ying patients at risk of flow-related ischemia during CEAnd the role of shunting in reducing this complication.actors associated with increased risk of cerebral ischemiauring carotid cross-clamping, and therefore the increa-ed likelihood that a shunt will be needed, during CEA, in-lude recent stroke, contralateral carotid occlusion, andymptoms suggestive of hemodynamic cerebral insuffi-iency.168,169 Despite extensive study on the routine orelective use of shunts and cerebral monitoring duringEA, no clear benefit of one approach over the other hasmerged.170

The routine use of completion imaging after CEA alsoemains an area of controversy. Although a number ofuthors have reported detecting abnormalities in 5% to 10%f patients using completion DUS imaging,171,172 and aost-benefit analysis suggests completion DUS imagingncreases quality-adjusted life-years by 2%,173 the clinicalignificance of many of these abnormalities is uncertain,nd several series have reported excellent results withoutse of completion imaging.174-176 Like the choice of anes-hesia and shunting, completion imaging remains a matterf personal preference.

There are, however, data to recommend the use ofatch angioplasty or eversion endarterectomy over stan-ard endarterectomy with primary closure. Women and

ndividuals with small ICAs are at most risk of early neuro-ogic events and late restenosis if standard endarterectomyith primary closure is performed. Randomized studiesave shown the benefit of patch closure over primarylosure in patients undergoing standard CEA.177 The typef patch material does not appear to have a significant effectn outcome. Prospective comparisons of eversion CEAith primary closure178 have demonstrated a benefit of the

version technique for reduction of early and late stroke.

plt

poaaopdioi

dLscvrcttca

poctitslotoIt

rwHetngbe

sirrws

JOURNAL OF VASCULAR SURGERYVolume 54, Number 3 Ricotta et al e13This has been borne out by low rates of early stroke and laterestenosis in large single-center reports.179

A number of investigations have studied the relation-ship of operative volume and specialty training with out-come. Although data suggest that there is some relation-ship between operative volume and outcome, the effectappears less than with other procedures.180 In a large studyof Medicare populations in Maryland and California, sur-geons who perform 10 to 15 CEAs per year have betterresults than those who perform 5 procedures annually,but there was no added benefit to performing more thanthis relatively low threshold. There has been no consistentrelationship between surgical specialty and outcome, andany effect seen is likely related to volume rather thanspecialty designation.8,180,181

B. Carotid artery stenting

The periprocedural management of the CAS patienthas been discussed in a previous section. Periproceduralantiplatelet therapy is mandatory, and appropriate atten-tion to access vessels and the status of the aortic arch isrequired for optimal results. The technical conduct of theCAS involves access of the target vessel, crossing the targetlesion, and stent deployment. Technical issues related tocarotid stenting include achieving a stable platform for theprocedure, use of embolic protection devices, stent dilationbefore deployment, stent selection, and postdeploymentdilation.

Stable sheath access in the proximal CCA is required.This depends on appropriate patient selection, as describedabove. Once a stable platform is obtained, a decision mustbe made about use of a cerebral protection device. Ingeneral, cerebral protection device deployment has beensuggested to reduce the incidence of distal embolizationand, potentially, the risk of stroke.182,183 Although thisposition is not supported by robust data, it has been gen-erally accepted by the medical community, and use of anembolic protection device has been required by Centers forMedicare and Medicaid Services to qualify for reimburse-ment.

Several embolic protection devices are available, andselection depends on lesion characteristics and anatomicconsiderations. Options include proximal or distal occlu-sion devices designed to interrupt forward flow during theprocedure and filter devices placed distal to the lesiondesigned to trap debris released during the procedure.Distal occlusion devices have the advantage of a smallerdiameter than proximal occlusion devices, but the lesionmust be crossed before the device is placed in the ICA, amaneuver that can itself cause embolization. At the end ofthe procedure, suction is applied between the sheath in theCCA and the balloon occluding the distal internal carotidartery to remove debris.

Proximal occlusion devices require placement of twoocclusion balloons, one each in the common and externalcarotid artery, with flow reversal by suction or creation of aproximal arteriovenous connection.184 Placement of a

proximal occlusion device avoids crossing the lesion before brotection is in place and has been associated with theowest incidence of distal embolization as detected byranscranial Doppler imaging or postoperative MRI.184-186

The main procedural disadvantages of the currentroximal occlusion devices are the relatively large size (9F)f the access sheath, the need to occlude both the commonnd external carotid artery, and the need to establish venousccess for continuous flow reversal. Proximal and distalcclusion devices may both be problematic in patients withoor intracranial collateral circulation because they man-ate cessationand sometimes reversalof antegrade flow

n the ICA. This situation is generally encountered in 5%f patients and may be managed by short intermittent

nflation times.Distal filters are deployed in the distal ICA and trap

ebris released during angioplasty and stent deployment.ike distal occlusion devices, they have the advantage of a

maller diameter (6F), but the lesion must be initiallyrossed in an unprotected fashion. These filters come in aariety of configurations and pore sizes. Their efficacy iselated to the degree with which they can reliably achieveomplete apposition to the distal arterial wall. They havehe advantage that antegrade ICA flow can be maintainedhroughout the procedure; however, the filter may becomeompletely occluded if large amounts of debris are released,nd debris may escape distally during filter recapture.

The choice of filter device is often related to individualreference and familiarity. Some authors believe proximalcclusion with flow reversal is preferable in nearly all cir-umstances; however, this technique has its greatest advan-age in lesions with a high risk of embolization (markedlyrregular plaque, echolucent lesions, active symptoms) or inhose that may be difficult to cross due to tortuosity orevere narrowing because protection is in place before theesion is manipulated. Use of a filter over proximal or distalcclusion is preferred if there is a likelihood that interrup-ion of antegrade ICA flow will not be tolerated. Distalcclusion devices are preferred to filters when the distalCA anatomy suggests that compete apposition of filter tohe distal ICA wall may be difficult due to size or tortuosity.

Direct comparison of proximal occlusion with floweversal vs distal protection shows that proximal protectionith flow reversal results in the lowest embolic load165,185

owever, no device can completely eliminate the risk ofmbolization during CAS. The fact that no embolic pro-ection is completely effective and that some emboli origi-ate during cannulation of the aortic arch and the proximalreat vessels suggests that although improvements to em-olization after CAS can be made, the problem cannot beliminated.

Studies using postoperative diffusion-weighted MRIhown increased 17% incidence of MRI-identified infarctsn patients undergoing CAS compared with CEA (adjustedisk ratio, 5.21).10 Although these are generally subclinical,ecent reports suggest these lesions might be associatedith subtle long-term neurologic changes. Echolucent le-

ions are more likely to be associated with increased em-

olic risk, whereas recurrent stenoses or fibrous lesions are

eo

mncottilw

atdpoctasNtsiStprabc

A

spimstt

pimtanhmtaia

JOURNAL OF VASCULAR SURGERYSeptember 2011e14 Ricotta et alassociated with a decreased risk of procedural emboliza-tion.66,96,165,184

Stent predilation is not recommended unless a filtercannot be passed. Stent selection is often based on physi-cian preference. The only large study to test the effect of cellsize suggests that closed-cell stents more effectively con-strain the carotid plaque and reduce embolization,187 butthere is no consensus on that point. Open-cell stents aremore conformable than closed-cell stents and are preferredby some in tortuous anatomy. Once the stent has beendeployed, postdilation is used to ensure stent apposition tothe plaque, but vigorous postdilation to achieve anatomicperfection is avoided.

The learning curve associated with CAS has been theobject of considerable study. The Endarterectomy VersusAngioplasty in Patients With Symptomatic Severe CarotidStenosis162 (EVA-3S) study was criticized because of therequirement that interventionalists perform only 25 proce-dures to qualify for participation in the trial. The CarotidRevascularization Endarterectomy versus Stenting Trial(CREST) study required a significant lead-in phase forinterventionalists and documentation of low proceduralmorbidity.188,189 It is worth noting that CREST is associ-ated with the lowest periprocedural complication rate in theliterature. There are data in the literature that suggestachieving a low predictable complication rate after CASrequires a higher level of initial and ongoing experiencethan current guidelines suggest.97,190,191

No relationship has been demonstrated between spe-cialty and outcome. Data from CREST demonstrate thatvascular surgeons, cardiologists, interventional radiolo-gists, and interventional neurologists can all achieve com-parable results with CAS and that experience is more im-portant than specialty designation in assuring optimaloutcomes.188,189

Recommendations regarding CEA and CAS tech-nique

1. Patch angioplasty or eversion endarterectomy are rec-ommended rather than primary closure to reduce theearly and late complications of CEA (GRADE 1, Levelof Evidence A).

2. Use of an embolic protection device (proximal or distalocclusion, distal filter) is recommended during CAS toreduce the risk of cerebral embolization (GRADE 1,Level of Evidence B).

V. SELECTING THE APPROPRIATE THERAPY:MEDICAL MANAGEMENT, CAS, OR CEA

Once a patient with a clinically significant carotid ste-nosis is identified, appropriate treatment must be selected.Treatment is primarily directed at the reduction of strokerisk. The risks of an interventional treatment must beconsidered when treatment choices are made. In general,rates of stroke, MI, and death have been used when com-paring CAS with CEA. In most clinical trials comparing

CAS with CEA, stroke, MI, and death have been given iqual weight in determining a composite end point to testverall efficacy.

Data from CREST,9 however, indicate that stroke has aore significant effect on quality of life at 1 year thanonfatal MI. Because the primary goal of intervention inarotid stenosis is stroke prevention, in developing its rec-mmendations, the committee placed more emphasis onhe prevention of stroke and procedurally related deathhan the occurrence of periprocedural MI. This may resultn committee recommendations that differ from the pub-ished results of some trials where these three end pointsere given equal weight in analysis.

Treatment is chosen based on the assessment of riskssociated with intervention and the likelihood that a par-icular intervention will favorably alter the course of theisease. The major determinants of the clinical course ofatients with carotid bifurcation stenosis are the presencer absence of neurologic symptoms and the degree ofarotid bifurcation stenosis. The threat of stroke in asymp-omatic patients with 60% ICA stenosis and in symptom-tic patients with 50% stenosis is generally considered to bemall and does not warrant intervention. ECST andASCET2-4 demonstrated that CEA was unable to reduce

he subsequent neurologic event rates in patients withymptoms of cerebral ischemia and bifurcation stenosis of50% diameter reduction and was actually associated with

ncreased morbidity compared with medical management.tenoses of 60% diameter reduction were excluded fromhe asymptomatic studies,5,6 assuming that asymptomaticatients with stenosis 60% would not benefit from carotideconstruction. Given the findings of the symptomatic tri-ls, this proved to be an appropriate decision. There haveeen no studies supporting either CEA or CAS for thisohort of patients.16

. Assessing the risk associated with intervention

CEA and CAS are each associated with specific clinicalcenarios that increase their respective risks. This sectionrovides information to identify conditions that pose anncreased risk for CAS or CEA and thereby help select the

ost appropriate therapy. When the risk of intervention isufficiently increased due to the presence of one or more ofhese factors, medical therapy may be more appropriatehan CEA or CAS.

In the initial CAS trials, a series of anatomic andhysiologic criteria were developed by a consensus paneln an attempt to identify high-risk CEA patients who

ight be expected to benefit from CAS.192 Althoughhese criteria were used to enroll patients in CAS trialsnd registries, their ability to define high risk wasever validated in a prospective manner. In fact, someave suggested that CEA could be safely performed inost of these patients.193,194 As CAS experience ma-

ured, certain conditions have been shown to be associ-ted with increased complications after CAS. Risk strat-fication can generally be divided into two categories:natomic (including the lesion) characteristics and phys-

ologic characteristics.

ti

aImtpa9Spdswa

ridYvpamatii

iCcCcCtwciAfiChbe

ersHdl2tlr

JOURNAL OF VASCULAR SURGERYVolume 54, Number 3 Ricotta et al e151. Anatomic and lesion characteristics.

a. Lesion location. CEA provides excellent access tothe cervical carotid artery, but lesions that extend outsidethis zone can be difficult to treat surgically. Lesions at orabove the level of the C2 cervical vertebra or below theclavicle are generally more difficult to expose surgically forCEA without increasing the morbidity of the operation.Lesions of the distal cervical carotid artery can be exposedby division of the digastric muscle and subluxation ordivision of the mandible, as required.195,196 Althoughrarely required, these high carotid exposures may be asso-ciated with increased difficulty in directly visualizing theend point of the endarterectomy and with increased inci-dence of cranial nerve injury, particularly cranial nerveIX.195,196 Lesions of the very proximal CCAs are difficultor impossible to expose without extending the incision intothe chest. This must be considered when evaluating themorbidity of the procedure.

b. Lesion characteristics. Lesion-specific characteristicsare thought to increase the risk of cerebral vascular eventsafter CAS66,197 and include a soft lipid-rich plaque iden-tified on noninvasive imaging, extensive (15 mm or more)disease, a preocclusive lesion, and circumferential heavycalcification. A recent publication using multivariate logis-tic regression analysis of a large patient cohort demon-strated increased periprocedural stroke risk (odds ratios,2.5-5.6) among patients with lesions 15 mm, excessivecalcification, and ulceration.197 An earlier study has showna periprocedural stroke risk (odds ratio, 7.1) among pa-tients with a lipid-rich plaque treated with CAS.66

The CAS procedure requires manipulation of a wireand a self-expanding stent through the carotid lesion. Un-stable plaque increases the risk of embolization duringplacement of the wire or stent across the carotid lesion. Thiscan be reduced, but not eliminated, by using flow-reversalembolic protection rather than distal filter protection.185

Long segment lesions may require the placement of multi-ple stents, and this situation and preocclusive stenoses areboth associated with a higher risk of acute or late stentocclusion. Heavy circumferential calcification makes lesiondilation more difficult and also increases the risk of embo-lization with CAS. There are no lesion specific characteris-tics that increase the risk of CEA.

c. Other anatomic considerations. Several anatomicsituations may increase the difficulty of CEA. Theseinclude reoperation after prior CEA, existence of a cer-vical stoma, history of neck radiotherapy with resultantlocal fibrotic changes of the skin and soft tissues, andprevious ablative neck surgery, such as radical neck dis-section and laryngectomy.193-195,198

While CEA can be successfully performed in these situa-tions, particularly when the tissues of the ipsilateral neck arenot scarred and fibrotic, these situations can increase the riskof wound infection, difficulty of dissection, and potentially,the incidence of cranial nerve injury. The presence of a short,

thick neck in an obese patient may make dissection more sedious but has not in itself proven to be associated withncreased operative risk.

Anatomic factors to consider with CAS are related toccess issues. Successful CAS requires remote access of theCA artery using a stable platform to avoid the intravascularotion of sheaths, stents, and protection devices during

he procedure. Anatomic factors that may complicate thisrocess include aortoiliac tortuosity, a sharply angulatedortic arch (type III), or a carotid lesion with more than two0 bends within a short distance of the target lesion.197

ignificant distal ICA tortuosity may also complicate thelacement and stabilization of a distal embolic protectionevice. An aortic arch with heavy calcium or a high athero-clerotic burden is also associated with an increased riskith CAS. This is felt to be the main reason that CAS results

re worse in patients aged 80 years.199,200

2. Patient characteristics. It seems intuitive that theisk of periprocedural events after CEA or CAS might bencreased in patients presenting with severe comorbid con-itions, including dialysis-dependent renal failure, Nework Heart Association class III or IV heart disease, leftentricular ejection fraction 30%, class III or IV anginaectoris, left main or multivessel coronary disease, severeortic valvular disease, oxygen- or steroid-dependent pul-onary disease, or both, contralateral carotid occlusion,

nd advanced age. However, little data exist to support oneherapy over another in these patients.162,163 In fact, defin-ng a high-risk patient is much more subjective than defin-ng a high-risk lesion.193,194,201,202

As will be seen later, CAS is associated with a lowerncidence of cardiac events than is seen in CEA. Therefore,AS would be preferred over CEA when severe cardiac

omorbidities exist in neurologically symptomatic patients.hronic renal insufficiency has been associated with in-

reased risk of stroke and death after CAS203,204 andEA.205,206 Univariate and multivariate analysis both show

hat the risk of death, stroke, and MI after CAS at 6 monthsas associated with hazard ratios 2.5 among patients with

hronic kidney disease.204 Chronic renal insufficiency alsoncreases the risk of stroke after CEA (1.08% to 5.56%).mong asymptomatic patients with cardiac or renal insuf-ciency, best medical therapy may be preferable to CAS orEA. CEA or CAS may be considered among symptomaticigh-risk patients with moderate to severe carotid stenosis,ut the effectiveness over medical therapy is not wellstablished.

There are conflicting data on the influence of contralat-ral occlusion on the outcome of CEA or CAS. NASCETeported that a contralateral occlusion increased the risk oftroke after CEA from 5.8% to approximately 14%.207

owever, most reports regarding contralateral occlusiono not bear this observation out, and a meta-analysis of the

iterature suggests a much more modest increase, from.4% to 3.7%.208,209 This was statistically significant, buthe results remain within the AHA recommended guide-ines. Several single-center studies have shown excellentesults in patients with contralateral carotid occlu-

ion.210,211 A possible explanation for this discrepancy is an

rpfhwttprAtMtaog

hittCCtttoaattg

prEppb3sat

itwitssu

iCt

JOURNAL OF VASCULAR SURGERYSeptember 2011e16 Ricotta et alinadequate sample size in the single-center studies. Alter-natively, a more consistent technique of intraoperativemanagement in single-center reports, including, algorithmsfor maintaining intraoperative cerebral perfusion, are morelikely to occur in single-center experience than in multicen-tered studies.

CEA is associated with a lower stroke risk than CAS inpatients aged 80 years.152,161,162 A combined death,stroke, and MI rate of 10% has been seen in octogenarianstreated with CAS.152,161 In the CREST study, CAS wasassociated with an increased stroke risk in patients aged70 years9 and appeared to have a benefit compared withCEA in patients aged 70 years, although no other studieshave reported this association to date.

Because there is a demonstrable incidence of cranialnerve injury after CEA that is absent after CAS, patientswith a history of a contralateral vocal cord paralysis are atincreased risk with CEA vs CAS; thus, CAS would bepreferred in these patients.

B. Neurologically asymptomatic patients with carotidartery stenosis of 60% or more

1. CEA for asymptomatic lesions. Patients withasymptomatic lesions are currently responsible for nearly allof the carotid in