endarterectomy for asymptomatic carotid artery stenosis · table 1.—baseline characteristicsof...

Endarterectomy for AsymptomaticCarotid Artery StenosisExecutive Committee for the Asymptomatic Carotid Atherosclerosis Study

Objective.\p=m-\To determine whether the addition of carotid endarterectomy toaggressive medical management can reduce the incidence of cerebral infarction inpatients with asymptomatic carotid artery stenosis.

Design.\p=m-\Prospective, randomized, multicenter trial.Setting.\p=m-\Thirty-nineclinical sites across the United States and Canada.Patients.\p=m-\Between December 1987 and December 1993, a total of 1662 pa-

tients with asymptomatic carotid artery stenosis of 60% or greater reduction in di-ameter were randomized; follow-up data are available on 1659. At baseline, rec-ognized risk factors for stroke were similar between the two treatment groups.

Intervention.\p=m-\Dailyaspirin administration and medical risk factor managementfor all patients; carotid endarterectomy for patients randomized to receive surgery.

Main Outcome Measures.\p=m-\Initially,transient ischemic attack or cerebral infarc-tion occurring in the distribution of thestudy arteryand any transient ischemic attack,stroke, or death occurring in the perioperative period. In March 1993, the primaryoutcome measures were changed to cerebral infarction occurring in the distributionof the study artery or any stroke or death occurring in the perioperative period.

Results.\p=m-\Aftera median follow-up of 2.7 years, with 4657 patient-years of ob-servation, the aggregate risk over 5 years for ipsilateral stroke and any periopera-tive stroke or death was estimated to be 5.1% for surgical patients and 11.0% forpatients treated medically (aggregate risk reduction of 53% [95% confidence inter-val, 22% to 72%]).

Conclusion.\p=m-\Patientswith asymptomatic carotid artery stenosis of 60% orgreater reduction in diameter and whose general health makes them good candi-dates for elective surgery will have a reduced 5-year risk of ipsilateral stroke if ca-rotid endarterectomy performed with less than 3% perioperative morbidity andmortality is added to aggressive management of modifiable risk factors.

(JAMA. 1995;273:1421-1428)

MORE THAN 500 000 new strokes occurannually in the United States, and it hasbeen estimated that carotid artery dis¬ease may be responsible for 20% to 30%of them.1 The annual stroke event ratefor asymptomatic patients with hemody-namically significant carotid artery ste¬nosis ranges from 2% to 5%.2"6 Carotidartery stenosis usually is identified aftertransient ischemie attack (TIA), but formany patients, cerebral infarction causedby artery-to-artery embolism or carotidocclusion is the initial event. Progressionof asymptomatic carotid artery stenosisto occlusion is unpredictable and can bedisastrous; at the time of occlusion, dis-

A complete list of the collaborators in the Asymp-tomatic Carotid Atherosclerosis Study appears at theend of this article.

Reprint requests to Stroke Center and Department ofNeurology, Bowman Gray School of Medicine of WakeForest University, Medical Center Blvd, Winston\x=req-\Salem, NC 27157-1068 (James F. Toole, MD).

abling stroke may occur in 20% of pa¬tients, and thereafter in 1.5% to 5% an¬

nually.6"8 On the other hand, the 30-daymajor morbidity and mortality for pa¬tients who undergo surgery for asymp¬tomatic stenosis ranges from 0.0% to 3.8%,and that for patients with symptomaticstenosis may be 6%.9 Because the role ofcarotid endarterectomy (CEA) for asymp¬tomatic carotid artery stenosis had notbeen proved,10-12 the Asymptomatic Ca¬rotid Atherosclerosis Study (ACAS) wasinitiated in 1987.13 The ACAS is an in¬vestigator-initiated randomized trial de¬signed to test whether CEA should be a

component of management for selectedpatients with asymptomatic stenosis ofthe common carotid bulb, the internal ca¬rotid sinus, or both. The question ad¬dressed was: Will CEA added to aggres¬sive reduction of modifiable risk factorsand administration of aspirin reduce the5-year risk of ipsilateral cerebral infarc-

tion in individuals with asymptomatic he-modynamically significant carotid arterystenosis?

For editorial comment see 1459.

Secondary objectives were to deter¬mine the surgical success in lesion re¬moval and the incidence of recurrent ca¬rotid stenosis, the rate of progression or

regression of carotid atherosclerosis inthe medically treated comparison group,and the incidence of all other vascularevents, such as TIA, myocardial infarc¬tion, and death related to vascular dis¬ease during follow-up.METHODS

The design and organization of theACAS are detailed elsewhere.13 Thirty-nine clinical centers were chosen from an

applicant pool of 55. All obtained institu¬tional review board approval ofthe studyprotocol.Recruitment

Study participants were recruited fromultrasound vascular laboratories, practi¬tioners who auscultated carotid bruits,and physicians who found carotid steno¬sis during evaluation for peripheral vas¬cular surgery or contralateral CEA.

Inclusion criteria were age between40 and 79 years; compatible history andfindings on physical and neurologicalexaminations; performance of requiredlaboratory and electrocardiograph^ ex¬aminations no earlier than 3 months be¬fore randomization; patient accessibil¬ity and willingness to be followed for 5years; and valid informed consent.

Exclusion criteria were cerebrovascu¬lar events in the distribution of the studycarotid artery or in that of the vertebro-basilar arterial system; symptoms refer¬able to the contralateral cerebral hemi¬sphere within the previous 45 days;contraindication to aspirin therapy; a dis¬order that could seriously complicate sur¬gery; or a condition that could preventcontinuing participation or was likely toproduce disability or death within 5 years.(Detailed information regarding eligibil-

Downloaded From: http://jama.jamanetwork.com/ by a University of Rochester User on 01/21/2014

ity and exclusion is available on requestfrom the corresponding author.)

The ACAS definition of hemodynami-cally significant carotid stenosis requiredthat at least one of three criteria was met:arteriography within the previous 60 daysindicating stenosis of at least 60% reduc¬tion in diameter (if the arteriogram was

performed 61 to 364 days before random¬ization, Doppler ultrasonography was re¬

quired to verify that the artery had notoccluded); Doppler examination within thepreceding 60 days showing a frequency or

velocity greater than the instrument-spe¬cific cut point with 95% positive predic¬tive value (PPV); or Doppler examinationshowing a frequency or velocity greaterthan the instrument-specific 90% PPV cutpoint confirmed by ocular pneumople-thysmographic (OPG-Gee) examinationperformed within the previous 60 days.

A patient could enter the study withunilaterally or bilaterally asymptomatic,hemodynamically significant stenosis, butonly one artery was the study artery. Iftwo arteries were eligible, the one withthe greater stenosis was selected. If thestenoses were identical, the left carotidartery was chosen. Patients randomizedto surgery on the basis of Doppler or

Doppler with OPG-Gee were required tohave an arteriogram prior to CEA. If a

postrandomization arteriogram revealedthe contralateral artery to have thegreater stenosis, it then became the studyartery. The nonstudy artery was man¬

aged medically unless a cerebrovascularevent occurred, at which time CEA couldbe considered.

Artériographie MeasurementsThe minimal residual lumen (MRL)

and the distal lumen (DL) were mea¬sured on the same radiograph. The MRLwas the smallest lumen diameter at thesite of the stenotic lesion. The DL wasthe diameter at the first point distal tothe MRL at which the arterial wallsbecame parallel. Percentage of stenosiswas calculated as 100 x (1-[MRL/DL]).Ultrasound Measurements

Because of the heterogeneity amongultrasound devices and techniques, we es¬tablished a cut point for each by compar¬ing Doppler ultrasonography with arte-riograms performed within 42 days ofeachother. Doppler cut points were computedfor peak systolic frequency or, if indeter-minant, end diastolic frequency, based ondata from 50 consecutive patients.14Randomization

An ACAS neurologist and an ACASsurgeon gave joint approval for enteringpatients. Once the eligibility criteria hadbeen confirmed and after informed con¬sent was obtained, the patient was

randomized using the permuted blockmethod with at least three different blocksizes determined randomly, stratified bycenter, gender, number of eligible arter¬ies, and previous contralateral CEA. Theassignment category was communicatedto each clinical center by the statisticalcoordinating center through an individu¬alized computer program arranged sothat the clinical center could neither pre¬dict nor reject an assignment.Medical Treatment

All patients received 325 mg of regularor enteric-coated aspirin daily (providedby Sterling Health USA, New York, NY).Stroke risk factors and their modificationwere reviewed with all patients at thetime of randomization and again duringsubsequent interviews and telephone fol¬low-up. This included discussion of dia-stolic and systolic hypertension, diabetesmellitus, abnormal lipid levels, excessiveconsumption of ethanol, and tobacco use.Whenever possible, the recommendationsofthe ACAS Risk Factor Reduction Com¬mittee were followed (available on re¬

quest from the corresponding author).Surgical Treatment

In addition to 325 mg of aspirin dailyand risk factor modification counseling,patients randomized to the surgical armreceived the normal evaluation and careof a surgical patient. They were sched¬uled to undergo CEA within 2 weeks ofrandomization. If an arteriogram or cra¬nial computed tomogram (CCT) had notbeen performed, the patient underwentthe procedure(s) before CEA. The arte¬riogram must have demonstrated a ste¬nosis of 60% or greater. Patients with a

postrandomization, presurgery arterio¬gram demonstrating less than 60% ste¬nosis or a distal abnormality such as an-

eurysm, arteriovenous malformation, or

siphon stenosis exceeding the proximalstenosis did not undergo surgery but wereretained in the surgical arm for compari¬son analyses. Asymptomatic cerebral in¬farction demonstrated by CCT was notan exclusion for surgery.15 No attemptwas made to standardize or control an¬esthesia or surgical techniques used bythe 117 ACAS-credentialed surgeons.16

The surgeon, the ACAS neurologist,and the ACAS patient coordinator ex¬amined each patient 24 hours after CEA.All deficits occurring through the 30-day perioperative period required theadministration of the end point reviewprocess (described below).Follow-up

Follow-up evaluations were conductedat 1 month and thereafter every 3 months,alternating between clinic visits and tele¬phone contacts. During the clinic visit,

patients completed a medical history ques¬tionnaire and TIA/stroke questionnaireand underwent physical and neurologicalexaminations and a Mini-Mental State Ex¬amination.17 Risk reduction managementwas reviewed and aspirin adherence wasdetermined by pill count.

Doppler ultrasound studies were re¬

peated at the 3-month follow-up, every 6months thereafter during the first 24months, then yearly, and at potential orverified end point or at exit from thestudy after 5 years; CCT was repeated atpotential end point or exit. Electrocar¬diogram was repeated when clinically in¬dicated and at exit.

Patients were instructed to notify thecoordinator ifsymptoms suggesting pos¬sible TIA or stroke occurred. The co¬ordinator scheduled urgent evaluationsby the ACAS neurologist and surgeon,and activated the end point verificationsystem.

In addition to identification of eventsfrom clinic visits and telephone contacts,hospital discharge diagnoses and deathcertificates were reviewed for coronaryevents and strokes.

End Point Definitionand Verification

A TIA was defined as a focal ischemieneurological deficit of abrupt onset last¬ing at least 30 seconds and resolving com¬

pletely within 24 hours. Deficits persist¬ing longer than 24 hours were classifiedas stroke.18 All strokes or deaths occur¬

ring within 30 days after randomizationin the surgical and 42 days in the medicalgroups were included as end points toreflect operative morbidity and mortal¬ity. The difference in times reflected an

average 12-day interval between random¬ization and surgery.

Initial review was conducted under a

stringent timetable involving one exter¬nal expert masked for local diagnosis,treatment assignment, clinical center, andtemporal relationship to surgery (ifdone).In addition, every potential event was

abstracted, masked, and reviewed to¬gether by all six experts on the End PointReview Committee. Our analyses arebased on their diagnoses.

Secondary analyses considered anystroke and perioperative death; any strokeand any death; and any ipsilateral TIAand stroke and any perioperative TIA,stroke, or death. The ACAS used cat¬egories 2 through 5 of the Glasgow scaleto determine a major stroke, defined as astroke resulting in moderate or severe

disability, persistent vegetative state, ordeath.19Statistical Analyses

Initially, the primary end points forevaluation of the two treatments were


Table 1.—Baseline Characteristics of RandomizedPatients by Treatment Assignment, in Percentages*

TreatmentAssignment

Baseline CharacteristicSurgical(n=825)

Medical(n=834)

Age, y40-4950-5960-6970-79

RaceWhiteBlackOther

SexMF

HistoryCoronary artery dlseasetHypertension:):CancerDiabetes mellitus

Lung disease at entryCurrent cigarette smokerBilaterally eligible arteriesPrevious contralateral

EndarterectomyTIA or stroke§EC-IC bypassSubclavlan bypass

Bruits In neck]|IpsilateralContralateral

Infarct on CT scannAny locationIpsilateral silent

Contralateral occlusionby Doppler

2135036

9433

6634

69641225

62810

2022<1<1

7644

22

10

2154638

9532

6634

69641021

524

1927

0<1

7442

249

*TIA indicates transient ischemie attack; EC-IC, exter¬nal carotid-internal carotid; and CT, computed tomographic.

tDefined as positive history of angina, coronary arterybypass, previous myocardial Infarction, or abnormal elec¬trocardiogram.

ÍPositive response to "Has your doctor ever told youthat you had high blood pressure or were hypertensive?"

§Significant difference between groups at .05 level(unadjusted for multiple comparisons).

¡Bruit question not evaluated for 51 surgical and 52medical patients. Ipsilateral/contralateral refers to ran¬domized artery.

HCT scan was not available on 71 surgical and 60medical patients.

ipsilateral TIA, stroke, or any periopera¬tive TIA, stroke, or death. The primaryanalysis is the comparison of the 5-yearrisk of cerebrovascular events in the twogroups. Because both treatments were inwide use, a two-sided test of the null hy¬pothesis (a=.05) was chosen. Power cal¬culations included all randomized patientsaccording to original treatment assign¬ments (intention-to-treat analysis). As¬suming an annual event rate in the worse-outcome group of 3% for TIA and 1% forcerebral infarction, these calculations in¬dicated that 750 patients were needed ineach treatment arm for 90% power todetect a 35% difference in 5-year eventrates, allowing for as much as 20% loss tofollow-up.

The results of the Veterans Affairstrial20 demonstrated that CEA is pref¬erable to medical management for pre¬venting TIA in asymptomatic carotidstenosis, and the North American Symp¬tomatic Carotid Endarterectomy Trial21demonstrated that infarction followingTIA is better managed surgically. Nei-

Table 2.—Artériographie Stenosis of the Ipsilateral Carotid ArteryPrerandomization Arteriogram

I Postrandomization:-o Stenosis_Medical_Surgical_Total (%)_Presurgery Arteriogram

0-59_Notapplicable_NA_NA_32(8)_60-69_131_137_268 (42)_139 (34)_70-79_94_93_187(29)_110(26)_80-89_75_79_154(24)_107(26)_90-99_13_20_33(5)_24(6)_Total 313 329 642* 412f

*Two patients were missing, one in each group.tTwo patients were missing.

ther resolved the issue ofwhether CEAprevents unheralded cerebral infarction.Therefore, in March 1993, the ACASExecutive Committee and the Data andSafety Monitoring Committee voted torestrict the primary end point to strokeand perioperative complications ordeath.

Forbaseline comparisons, we used two-tailed t tests for comparing the means ofcontinuous variables and 2 for compar¬ing distributions of categorical variables,with no adjustment for multiple compari¬sons. Kaplan-Meier estimates of 5-yearaggregate risk were compared betweentreatment groups using either Green¬wood's formula for variances, for a large-sample test ignoringrandomization strati¬fication,22 or randomization tests, respect¬ing randomization strata.23 In the initialyears of treatment comparison, 1991through 1993, the tests were for 2-, 3-, or

4-year aggregate risk. The randomiza¬tion test was the primary method for in¬terim treatment comparisons (see below).By the time of study closure valuesfrom the two methods agreed within .002,so that all test results and confidence in¬tervals (CIs) reported are based on large-sample tests unless otherwise noted.

Semiannual treatment comparison anal¬yses were used to advise the Data Safetyand Monitoring Committee whether a sig¬nificance boundary had been crossed. Thestopping rule was a modified O'Brien-Fleming24 rule for mamtaining the desiredoverall significance level despite repeatedtesting. The modification was for testingat selected intervals rather than prede¬termined numbers of events. The fiveanalyses originally planned were changedto 10, because recruitment lagged andTIAwas deleted as a primary end point. Thecritical value for the test statistic was 6.00at the first test and 2.07 for the 10th, as

compared with 1.96 for a single test at thesignificance level .05. The study was

stopped after the eighth test, when thecritical value was 2.38, corresponding to anominal significance level of .017.

Treatment comparisons are reportedherein in terms of relative risk reduc¬tion, the 5-year risk reduction due tosurgery as a proportion of risk in themedical group.25 Absolute treatment

group-specific risk levels are also pro¬vided for calculation of absolute risk re¬duction. For the primary event the num¬ber of patients treated to prevent oneevent over 5 years was calculated as theinverse of the absolute risk reduction.26

Intention-to-treat analyses were usedfor all comparisons unless otherwise in¬dicated, regardless of postrandomizationineligibility or crossover. End points formedical patients who received CEA af¬ter verified ipsilateral end point were cat¬egorized as perioperative ifthey occurredwithin 30 days of CEA. AU tests weretwo tailed.

COHORT CHARACTERISTICSDuring the 6 years ofthe study, more

than 42 000 patients were screened and1662 patients were randomized. Twenty-four centers contributed more than 30patients each, and 13 contributed morethan 50 patients. The average numberof patients recruited per center was 43.

From March 1988 through October1993, 12080 CEAs were performed atthe sites. Six percent (683) were per¬formed on "likely eligible nonrandom-ized" patients of ACAS physicians, 6%(758) were performed on already ran¬domized ACAS patients, and the restwere performed on symptomatic pa¬tients, ineligible patients, or patients ofsurgeons not collaborating in the ACAS.

Patient characteristics are presentedin Table 1. Of the 1662 randomized pa¬tients, three in the surgery group werelost to follow-up after randomization andare excluded from analysis, leaving 1659.The 825 surgical and 834 medical patientswere compared for 189 baseline charac¬teristics, withonly six tests yielding nomi¬nal statistically significant differences atthe .05 level. Two thirds of the patientswere men, 95% were white, and 48% were

aged 60 through 69 years. Mean age was67 years; mean weight, 81 kg for men and67 kg for women; mean systolic bloodpressure, 146 mm Hg; mean diastolic bloodpressure, 78 mm Hg; and mean total cho¬lesterol concentration, 5.90 mmol/L (228mg/dL). Approximately 75% of patientshad a bruit associated with the study ar¬

tery, and in 43%, a contralateral carotidbruit was heard; 21% had a previous myo-


Table 3.—Number of Observed Events in Median 2.7-Year Follow-up, Estimated Number and Percentage of Events in 5 Years, Reduction Due to Surgery in 5-YearRisk as a Proportion of Risk in the Medical Group (95% CI), and Large-Sample Value for Treatment Group Difference, by Event Type*

Event Type

ObservedNo. of Events

in Median 2.7-yFollow-up

Medical (n=834) Surgical (n=825)Kaplan-Meier

Estimate of 5-yEvent Risk,

No. (%)

Observed

No. of Eventsin Median 2.7-y

Follow-up

IKaplan-Meier

Estimate of 5-yEvent Risk,

No. (%)

Reduction Dueto Surgery in 5-y Risk

as a Proportion of Riskin Medical Group

(95% CI)Ipsilateral stroke or any perloperative stroke or death 52 92(11.0) 33 42(5.1) 0.53 (0.22 to 0.72) .004

Major Ipsilateral stroke or anyperloperative major stroke or death 24 50 (6.0) 28 (3.4) 0.43 (-0.17 to 0.72) .12

Ipsilateral TIA or stroke or any perioperative TIA orstroke or death 102 160(19.2) 55 67 (8.2) 0.57 (0.39 to 0.70) <.001

Any stroke or any perioperative death 86 146(17.5) 60 102(12.4) 0.29 (-0.05 to 0.52) .09

Any major stroke or perioperative death 40 76(9.1) 28 53 (6.4) 0.30 (-0.30 to 0.62) .26

Any stroke or death 155 266(31.9) 127 211 (25.6) 0.20 (-0.02 to 0.37) .08

Any major stroke or death 213(25.5) 100 171 (20.7) 0.19 (-0.08 to 0.39) .16

*CI Indicates confidence interval; and TIA, transient ischemie attack.

cardial infarction, and 21% a previouscoronary artery bypass. Sixty-four per¬cent had hypertension, 26% were ciga¬rette smokers, and 23% had diabetes mel-litus.

Four hundred seven patients (25%)had had a previous hemispheric eventcontralateral to the study artery, and1155 (70%) were asymptomatic in thedistribution of both arteries.

Thirty-nine percent of patients wererandomized on the basis of an arterio¬gram showing at least 60% stenosis ofthecarotid artery. Fifty-five percent wererandomized with a Doppler PPV cut pointofat least 95%, and 6% with a Doppler cutpoint of at least 90% confirmed by OPG-Gee. The positive predictive value ofDop¬pler, estimated from the postrandomiza-tion presurgery angiogram, was 93%.

Table 2 shows the distribution of per¬centage of stenosis for prerandomiza-tion and postrandomization arteriogramsbefore CEA. Because the health statusof patients who received prerandomiza-tion arteriograms may differ from thatof those who did not, a weighted esti¬mate based on both categories is in¬cluded. Five percent ofpatients had ste¬nosis of the randomized artery less than60%; 39%, 60% to 69% stenosis; 28%,70% to 79% stenosis; 25%, 80% to 89%stenosis; and 5%, 90% to 99% stenosis,for a mean percentage ofstenosis of 73%.

The Central Reading Center classified536 baseline CCTs15 as showing cerebralinfarction. If an infarcì was present, itwas further classified by age, size, dis¬tribution, and volume. Local and centralreaders agreed on 89% of the cases. Us¬ing the Central Reading Center as thestandard, the sensitivity of local readingwas 71% and the specificity was 94%.

RESULTSOfthe 825 surgical patients, 101 did not

have ipsilateral arteriography or CEA,45 because ofpatient refusal despite prioragreement to accept either treatment.Twelve patients were rejected for sin·-

gery because of severe cardiac disease.Three had a stroke or died before arte-riography or surgery was performed. Ar-teriograms found 33 patients to be ineli¬gible, six because of intracranial abnor¬malities and 27 because of less than 60%carotid artery stenosis. Eight patients didnot have surgery for various other rea¬sons. Of the 834 patients randomized tomedical treatment, 45 received CEAwith¬out a verified ipsilateral TIA or stroke.Thus, 146 (9%) patients did not receivethe assigned treatment. Eleven patientsdropped out from follow-up in the medi¬cal and nine in the surgical group.

During the perioperative period, 19 sur¬

gical patients (2.3%) had a stroke or died.Two patients had a stroke, one died priorto hospitalizaron, and five had a cerebralinfarction as a direct result of arteriog-raphy, one of whom died. There were 10nonfatal strokes and one fatal myocardialinfarction during the 30-day postsurgeryperiod. In the comparable perioperativeperiod for the medical group, three pa¬tients (0.4%) had a cerebral infarction (twopatients) or died (one patient). For thesurgical group, the risk in the periopera¬tive period was 2.3% (95% CI, 1.28% to3.32%), whereas for the medical group itwas 0.4% (95% CI, 0.0% to 0.8%).

All patients randomized to the surgicalgroup were required to have arteriogra-phy. Of the 414 patients who underwentarteriography prior to CEA, five expe¬rienced a cerebral infarction, for an ar¬

tériographie complication rate of 1.2%. Itis estimated that if all 724 patients re¬

ceiving CEA had undergone arteriogra¬phy as a part of the ACAS, 8.7 artério¬graphie cerebral infarctions would haveoccurred in addition to the 11 primaryevents in the 30 days following surgery,for an overall rate of 2.7% for cerebralinfarction or death from the procedure.

Sixteen fatalities, potentially due tostrokes, were reviewed by the Cere-brovascular End Point Review Commit¬tee. In no case was there a differencebetween the End Point Review Com-

mittee diagnosis and the local physiciandiagnosis. These events included sixhemorrhagic strokes, two cerebral in¬farctions in the distribution of the ran¬domized artery, three cerebral infarc¬tions in the nonrandomized distribution,and five deaths not due to stroke.

Treatment ComparisonsThe study achieved its significance

boundary after a median of 2.7 years offollow-up, with 9% ofpatients having com¬

pleted 5 years; 26%, 4; 44%, 3; 68%, 2; and87%, 1 year of follow-up. Because surgi¬cal patients were at greatest risk duringthe first month after endarterectomy,whereas the risk for medical patients wasdistributed throughout 5 years, compari¬sons near term greatly understated thedifferences expected after 5 years. Table3 presents the observed number ofeventsand also the Kaplan-Meier estimates pre¬dicted ifall patients had been followed for5 years. The estimated 5-year risk of ip¬silateral stroke and any perioperativestroke or death was 11.0% for the medicalgroup and 5.1% for the surgical group.The reduction in 5-year ipsilateral strokerisk in the surgical group was 53% of theestimated 5-year risk in the medical group(95% CI, 22% to 72%). The value for thetest of the difference between the treat¬ment groups in 5-year risk of primaryevent was .004 by the large-sample testand the randomization test. For the pri¬mary end point of ipsilateral stroke andany perioperative stroke or death, thesurvival curves in the Figure cross near10 months and become significantly re¬duced in the surgical group by 3 years(P<.05).

The results for secondary end pointsare in the same direction although notalways statistically significant (Table 3and Figure). Ipsilateral TIA or stroke or

any perioperative TIA, stroke, or death—the original primary end point for theACAS—showed a 57% reduction in 5-yearrisk for the surgery group (95% CI, 39%to 70%). In terms of any stroke or death,


0.8

0.7

Ipsilateral Stroke or PerioperativeDeath or Stroke

Any Stroke orPerioperative Death

eo ¬

Major Ipsilateral Stroke or Any PerioperativeMajor Stroke or Perioperative Death

0.9

I

0.8

0.7

0.8

0.7

Any Major Stroke orPerioperative Death

Ipsilateral TIA or Stroke or PerioperativeTIA or Stroke or Death

1.0-K 1.0Any Stroke

or Death

Years of Follow-up

Proportion of patients without end point at a given time during follow-up, by treatment group, using Kaplan-Meier estimation method. Solid line indicates medical patients; broken line, surgical patients; and TIA, tran¬sient ischemie attack.

the surgery group had a 20% reduction inevents (95% CI, -2% to 37%). The reduc¬tion due to CEAin major ipsilateral strokeor perioperative death was 43% (95% CI,-17% to 72%). Table 4 shows the causesof death were similar for the two groups,except for death not proved to result frommyocardial infarction.

Table 5 summarizes results for ipsilat¬eral stroke or any perioperative stroke ordeath bypatient subgroups. In men, CEAreduced the 5-year event rate by 66%(95% CI, 36% to 82%); in women, theevent rate was reduced by 17% (95% CI,-96% to 65%). However, the differencebetween genders was not statistically sig¬nificant (P=.10). The proportion of wom¬en with perioperative complications was

3.6%, compared with 1.7% formen CP=.12).However, among patients who had no

perioperative event, 5-year risk was re¬duced by 56% for women (95% CI, -50%to 87%), compared with a reduction of79% for men (95% CI, 52% to 91%) (data

not shown). Table 5 shows a larger riskreduction due to CEA for younger pa¬tients, but the difference is not statisti¬cally significant (P=.50).

Reanalysis excluding the 146 cross¬

overs, ie, restricted to those patients whoreceived the assigned treatment, showsthat surgery reduced 5-year stroke riskby 55% (95% CI, 23% to 74%). Alterna¬tively, for the 1155 patients who had nocontralateral TIA, stroke, or endarter-ectomy prior to randomization, surgeryreduced the 5-year stroke risk by 46%(95% CI, 0% to 71%) (Table 5).

The percentage of stenosis for the 642patients who received an arteriogramwithin 6 months preceding randomiza¬tion is shown in Table 5. Throughout thethree groups, ie, for patients with 60% to69%, 70% to 79%, and 80% to 99% ste¬nosis, there was no statistically signifi¬cant gradation in reduction of5-year riskof primary event, but sample sizes weresmall. The patients for this comparison

Table 4.—Number of Deaths, Overall and byTreatment Group, in the ACAS, December 1987Through December 31, 1994*

Cause of Death

Treatment GroupI ISurgical Medical

Perioperative onlyIpsilateral strokeStroke on contralateral side

or posterior circulationCerebral hemorrhageAcute myocardial infarctionOther cardiac diseaseOther vascular disorderRespiratory failureCancerTuberculosisRenal failureGastrointestinal system

diseaseNervous system diseaseTraumaUnknownTotal DeathsDeaths per 100 person-years

of follow-up

32

13

2115

11015

1

4120

83

3.5

13

24

2424

29

1302

3002t

3.8

*ACAS indicates Asymptomatic Carotid Atheroscle¬rosis Study.

tOne patient died at home, the other at a non-ACASfacility. The patient coordinators were unable to obtaininformation as to the causes of deaths.

all had an arteriogram performed beforerandomization, so the risk of stroke fromundergoing an arteriogram was not in¬cluded in the calculations. If a 1.2% ar¬

teriogram risk were added to the sur¬gery groups at the three stenosis levels,the risk reductions become 0.35,0.49, and0.13, respectively, which are consistentwith the overall results of the ACAS.COMMENT

The ACAS was designed to test theefficacy of CEA for preventing ipsilat¬eral stroke during a 5-year period. Eventhough this report includes patients fol¬lowed up for a median of only 2.7 years,with 9% having completed 5 years of fol¬low-up, the data demonstrate a statisti¬cally significant (P=.004) difference be¬tween the estimated 5-year ipsilateralstroke rates of 11.0% for the medical and5.1% for the surgical group. Moreover,the results are in the same direction forall subgroups considered, including decilesof stenosis (although not statistically sig¬nificant because ofsmall sample size), andfor various secondary cerebrovascular endpoints. Furthermore, the results are vir¬tually the same when restricted to allpatients receiving the assigned treatment,and are almost identical for patients with¬out previous contralateral symptoms or

endarterectomy.Approximately 70% ofour medical and

surgical patients had artériographie ste¬noses less than 80%. Even so, the esti¬mated 5-year ipsilateral stroke rate inthe ACAS medical group was 11.0%(about 2.3% annually). The stroke ratein the medically managed group de¬creased to the lower end of the previ¬ously reported range, perhaps as a re-


Table 5.—Number and Percentage of Perioperative Strokes or Deaths, Number of Observed Events in Median 2.7-Year Follow-up, Estimated Number andPercentage of 5-Year Ipsilateral Strokes or Perioperative Strokes or Deaths, and Reduction Due to Surgery in 5-Year Risk as a Proportion of Risk in the MedicalGroup, With 95% CI, by Subgroup*

Total Events

Patient GroupTreatment

GroupNo. atRisk

PerioperativeEvents,No. (%)

No. of EventsObserved in Median

2.7-y Follow-up

Estimated Eventstfor 5-y Follow-up,

No. (%)

Reduction Dueto Surgery in 5-y Riskas Proportion of Riskin the Medical Group

Estimate 95% CIAll Surgical 825 19(2.3) 33

Medical 834 3 (0.4) 5242(5.1)92(11.0)

0.53 0.22 to 0.72

Men Surgical 544 9(1.7)Medical 547 3 (0.5) 38

22(4.1)66(12.1)

0.66 0.36 to 0.82

Women Surgical 281 10(3.6) 15

Medical 287 0 (0.0)20 (7.3)25 (8.7)

0.17 -0.96 to 0.65

Age <68 y Surgical 408 6(1.5) 13Medical 394 2 (0.5) 23

19(4.7)47(11.8)

0.60 0.11 to 0.82

Age *68 y Surgical 13(3.1) 20Medical 440 1 (0.2) 29

23 (5.5)43 (9.7)

0.43 -0.07 to 0.70

Bilaterally asymptomatic Surgical 585 12(2.1) 24

Medical 3 (0.5) 3432 (5.5)58(10.2)

0.46 0.00 to 0.71

Previous contralateral endarterectomyor previous TIA or stroke

Surgical 240 7 (2.9)Medical 264 0 (0.0) 18

11 (4.5)33(12.6)

0.65 0.13 to 0.86

Patients receiving assigned treatment Surgical 724 16(2.2)Medical 789 3 (0.4) 50

37(5.1)91 (11.5)

0.55 0.23 to 0.74

% Stenosis* 60.0-69.9 Surgical 137 4 (2.9)Medical 131 0 (0.0)

9 (6.3)15(11.4)

0.45 -0.70 to 0.82

% Stenosis* 70.0-79.9 Surgical 93 1(1.1)Medical 94 0 (0.0)

2 (2.2)6 (6.7)

0.67 -0.65 to 0.94

% Stenosis* 80.0-99.9 Surgical 99 1 (1.0)Medical 0 (0.0)

2 (2.0)3 (3.7)

0.45 -2.19 to 0.91

*CI indicates confidence interval; and TIA, transient ischemie attack.tUsing Kaplan-Meier estimation.^Percentage of stenosis of randomized artery at baseline, for patients with prerandomization angiogram within 6 months of randomization.

suit ofvigorous risk factor managementand exclusion of high-risk patients.

There were no significant differencesin primary event rates between patientgroups with and without symptoms or

previous CEA of the contralateral ca¬rotid artery. With an annual mortalityrate of 3.7%, approximately 89% of pa¬tients survived long enough to benefitfrom the protective effect of the opera¬tion, because the crossover in favor ofsurgery occurred within the first year.

Four other randomized prospectivestudies of CEA for asymptomatic ca¬rotid artery stenosis have been reported.One did not include stenosis exceeding90%,26 another was terminated early be¬cause of excess cardiac events,27 and a

third, the European Asymptomatic Ca¬rotid Surgery Trial, is ongoing.28

The fourth, the Veterans Affairs Co¬operative Trial, randomized 444 patientsand published results based on a mean

follow-up of 47.9 months. The VeteransAffairs study differed from the ACASin that only men were studied and allpatients had an arteriogram.20

Like the Veterans Affairs trial, theACAS showed an advantage for CEA inpreventingTIAs, cerebral infarctions, anddeath in men. In addition, the ACASshowed an advantage in reducing the riskofipsilateral stroke alone. Ourresults are

consistent with others that establishedthat symptomatic patients with carotidstenosis greater than 70% were besttreated by CEA.21^

Because a 10% difference in lumendiameter on arteriogram is approxi¬mately 0.5 mm, and the lumen area ste¬nosis difference is only 6%, this cannotbe measured accurately, and when min¬iaturized images are used, these differ¬ences cannot be discerned. Therefore,we believe that stenoses with 60% and70% reductions in diameter are bothhemodynamically significant, and thatputative difference by decile is withinthe range of observer variability.3032

It has been suggested that arteriog-raphy should have been required for allACAS patients to ensure that only thosewith greater than 60% stenosis were en¬tered. However, it was the judgment ofthe ACAS group that the hazards andcosts ofarteriographywere not warrantedfor patients in our medical group. Thiswas borne out by the 1.2% stroke ratefrom arteriography. Our Doppler criteriawere established to maintain a PPV of95%.14 Retrospective analysis of all post-randomization, presurgery arteriogramsdemonstrated that our actual PPV was93%. This indicates that our medical pa¬tients did indeed have significant carotidartery stenosis, with fewer than 5% hav-

ing less than the required 60%.If all patients who underwent surgery

had received arteriography as part of thesurgical treatment, the absolute risk re¬duction would have been from 11.0% to5.6%. Using ACAS eligibility require¬ments, 19 CEAs would be necessary toprevent one stroke over 5 years.25 Thisratio would be less if patient subsets athigher risk for stroke could be identified.

A CEA can be performed with a lowcomplication rate even in elderly pa¬tients. In selected instances, some ACASsurgeons now operate without arteri¬ography on the basis ofnoninvasive stud¬ies33·34 and sometimes discharge patients24 hours following surgery.35 These andother measures may reduce costs ifproved generally feasible.CONCLUSIONS

The ACAS has demonstrated that theincidence of cerebral infarction can bereduced by CEA and that stringent qual¬ity control measures can reduce surgicalmorbidity and mortality. A major reasonwas the 30-day morbidity and mortalityofACAS patients, estimated to have been2.7% if all surgery patients had under¬gone arteriography as part of the study.This includes artériographie complicationsof 1.2%. These results may be improvedfurther by reducing risk associated with


contrast arteriography. The ACAS hasestablished that men with a good life ex¬

pectancy who have asymptomatic carotidartery stenosis with at least 60% reduc¬tion in diameter are protected from strokeby CEA, whereas the results for womenare less certain. Following CEA, the rela¬tive stroke risk reduction for men andwomen combined is 53%, with an abso¬lute 5-year risk reduction from 11.0% to5.1%. The 5-year reduction in stroke riskamong men was 66% and among women17%, perhaps because of the higher peri-operative complication rate inwomen. Ex¬cluding artériographie and perioperativecomplications, the risk reduction was 79%for men and 56% for women.

The Asymptomatic Carotid Atherosclerosis Studyclinical trial was supported by an investigator\x=req-\initiated research grant (RO1 NS22611) from the USPublic Health Service National Institute of Neuro-logical Disorders and Stroke.

The authors acknowledge the volunteerism ofallpatients in the Asymptomatic Carotid Atheroscle-rosis Study.

Funding Agency.\p=m-\National Institute of Neu-rological Disorders and Stroke, Bethesda, Md: Mi-chael D. Walker, MD; John R. Marler, MD; MurrayGoldstein, DO; Patricia A. Grady, PhD.

Executive Committee.\p=m-\James F. Toole, MD,chair; William H. Baker, MD; John E. Castaldo,MD; Lloyd E. Chambless, PhD; Wesley S. Moore,MD; James T. Robertson, MD; Byron Young, MD;Virginia J. Howard, MSPH (ex officio).

Executive Committee Liaison.\p=m-\John R. Marler,MD.

Operations Center.\p=m-\Bowman Gray School ofMedicine of Wake Forest University, Winston\x=req-\Salem, NC: James F. Toole, MD; Virginia J.Howard, MSPH; Suzanne Purvis; Dee Dee Vernon;Kelley Needham; Pam Beck; Marty Dozier; DavidS. Lefkowitz, MD; George Howard, DrPh; John R.Crouse III, MD; David M. Herrington, MD; Curt D.Furberg, MD, PhD; Karla Essick; Ralph M. Hicks,Jr, MA, MEd.

Statistical Coordinating Center.\p=m-\University ofNorth Carolina at Chapel Hill: Lloyd E. Chamb-less, PhD; J. J. Nelson, MSPH; Will Ball; ErnestineBland; Sean Condon, MS; Thomas Elliott, MS (de-ceased); James E. Grizzle, PhD; Dick Hayes, MCJ;Scott Henley; Jeffrey Johnson, MSPH; JamesLocklear, MS; Margaret S. Misch, MS; Catherine C.Paton, MSPH; Skai Schwartz, MA; ClimmonWalker; O. Dale Williams, PhD.

Data and Safety Monitoring Committee.\p=m-\J. Donald Easton, MD, chair; Jerry Goldstone, MD;John M. Hallenbeck, MD; Julian T. Hoff, MD; Her-bert R. Karp, MD; Richard A. Kronmal, PhD.

Cranial Computerized Tomography ReadingCenter.\p=m-\Universityof Cincinnati (Ohio): ThomasG. Brott, MD; Thomas A. Tomsick, MD; JosephBroderick, MD; Laura Sauerbeck, RN; ChristineBlum, RN.

Cerebrovascular End Point Review Commit-tee.\p=m-\MarkDyken, MD, chair; John C. M. Brust, MD;Arthur R. Dick, MD; Robert A. Gotshall, MD; AlbertHeyman, MD; Phillip D. Swanson, MD.

Publications Committee.\p=m-\JamesF. Toole, MD,chair; Harold P. Adams, Jr, MD; Lloyd E. Chamb-less, PhD; Robert Dempsey, MD; Calvin B. Ernst,MD; John Rothrock, MD (with special appreciationto Stanley N. Cohen, MD, for desk editing).Participating Centers (in order ofnumber ofpa-tients randomized, number ofpatients in brackets,and personnel listed in the following order: Princi-pal Investigator, Co-Investigator, Coordinator, andOther Key Personnel, Current and Past)

Lehigh Valley Hospital, Allentown, Pa [142]:John Castaldo, MD; Gary Nicholas, MD; JoanLongenecker; Peter Barbour, MD; Alan Berger,

MD; Victor J. Celani; Nancy Eckert, RN; JamesGoodreau, MD; Judith Hutchinson, RN; DonnaJenny, RN; Zwu-Shin Lin, MD; Alice Madden, RN;James L. McCullough, MD; Kenneth McDonald,MD; William Pistone, MD; Alexander D. Rae\x=req-\Grant, MD; James Redenbaugh, MD; James Rex,MD; Christopher J. Wohlberg, MD.

Marshfield Clinic, Wisconsin [121]: PercyKaranjia, MD; Mark Swanson, MD; Sandra Lobner,LPN; R. Lee Kolts, MD; Marvin E. Kuehner, MD;Bradley C. Hiner, MD; Kenneth P. Madden, MD;Robert D. Carlson, MD; J. Steven Davis, MD; Tho-mas Gallant, MD; John J. Warner, MD; Ann Faust,RVT; Nancy Fryza, RVT; John Hasenauer, RVT;Mary Regner, RVT; Linette Ronkin, RVT; SharonSchaefer, RVT; Dawn Strack, RVT; Lynn C.Turner, RVT; Ann Walgenbach, RVT; JulieGraves; Sara Michalski; Linda Schuette.

Columbia University, New York, NY [91]: J. P.Mohr, MD; Donald Quest, MD; Annette Cruz;Ralph Sacco, MD; T. K. Tatemichi, MD; RandolphMarshall, MD; Henning Mast, MD; Oscar Ramos,MD; James Correll, MD; Richard Libman, MD;George Petty, MD; Anilda Cabrere, MD; LorraineOropeza, RN, RVT; Tina Gonzalez.

University ofKentucky Chandler Medical Cen-ter, Lexington [88]: Byron Young, MD; Creed Pet-tigrew, MD; Ruthie Sadler, RN; Robert Dempsey,MD; Eric Endean, MD; Jerry Sherrow, RVT; Mar-cie Hauer, RN; Charles Lee, MD; Jane Norton; Mi-chael McQuillen, MD; Sally Mattingly, MD; StevenDekosky, MD; Andrew Massey, MD.

H\l=o^\pitalde l'Enfant Jesus, Qu\l=e'\becCity, Quebec[86]: Denis Simard, MD; Jean-Fran\l=c;\oisTurcotte,MD; Charlotte Benguigui, RN; Jacques C\l=o^\te,MD(deceased); Jean-Marie Bouchard, MD; ClaudeRoberge, MD; Denis Brunet, MD; Fernand B\l=e'\dard;R\l=e'\ginaldLangelier, MD; Marcel Lajeunesse; JeanMarc Bigaouette; Jasmin Parent.

University of California at San Diego [78]: Pa-trick Lyden, MD; Robert Hye, MD; Stacy Lewis,RN; Gerry Cali, RN; Traci Babcock, RN; BarbaraTaft-Alvarez, RN; John Rothrock, MD; MarkBrody, MD; Richard Zweifler, MD; Mark Sedwitz,MD; Bruce Stabile, MD; Julie A. Freischlag, MD;Yehuda G Wolf, MD; James Sivo; John Forsythe;Melody Adame.

Loyola University Medical Center, Maywood,Ill [77]: William H. Baker, MD; Sudha Gupta, MD;Katy Burke; Howard P. Greisler, MD; Fred N.Littooy, MD; Michael A. Kelly, MD.

University ofTennessee, Memphis [74]: James T.Robertson, MD; William Pulsinelli, MD; Judith A.Campbell, RN, BSN; John Crockarell, MD; ClarenceWatridge, MD; J. Acker, MD; Samard Erkulwater,MD; Michael Jacewicz, MD; Gail Walker; PatrickO'Sullivan, MD; Curtis Sauer, MD; Ken Vasu, MD;Kenneth Gaines, MD; Bijan Bakhitian, MD; Tulio E.Bertorini, MD; Susan Bennett, PA-C; Terrye Tho-mas, BSN; Nan Stahl, RN; Connie Taylor, RN; MaryAnn Giampapa; Joan Connell; Judy Riley; AngeleahBradley; Jeanette Davis, RN; Kathy Newman; Re-becca Manning; Marcella McCrea.

University Hospital, London, Ontario [63]:Vladimir Hachinski, MD; Gary Ferguson, MD;Cheryl Mayer, RN; Henry J. M. Barnett, MD; S. J.Peerless, MD; A. M. Buchan, MD; Howard Reich-man, MD; Andrew Kertesz, MD; Stephen Lownie,MD; Carole White, RN; Allan Fox, MD; RichardRankin, MD.

Victoria Hospital, London, Ontario [61]: J.David Spence, MD; H. W. K. Barr, MD; LesliePaddock-Eliasziw, RN; L. J. P. Assis, MD; J. H. W.Pexman, MD; Maria DiCicco, RVT; Barb Tate, RN;Caroline James, RVT.

Virginia Mason Clinic, Seattle, Wash [57]: Ed-mond Raker, MD; James Coatsworth, MD; SandyHarris, RN; Hugh Beebe, MD; Richard Birchfield,MD; Kathy Butler-Levy; Robert Crane, MD; DavidFryer, MD; James MacLean, MD; Laird Patterson,MD; Terence Quigley, MD; John Ravits, MD;Lynne Taylor, MD; Stacy Pullen, RDMS; ShannonBoswell, RDMS; Karen Kenny, RDMS, RVT.

University ofCincinnati Ohio [55]: Thomas Brott,MD; Joseph Broderick, MD; Thomas Tomsick, MD;

Laura Sauerbeck, RN; L. R. Roedersheimer, MD;Richard Fowl, MD; John Tew, MD; Richard Kemp-czinski, MD; Robert Reed, MD; Richard Welling,MD; Christine Blum, RN; Bill Schomaker, RN.

Bowman Gray School ofMedicine ofWake For-est University, Winston-Salem, NC [52]: DavidLefkowitz, MD; J. Michael McWhorter, MD; CharlesL. Branch, MD; Jean Satterfield; Robert Cordell,MD; Richard Dean, MD; George Plonk, MD; GaryHarpold, MD; James F. Toole, MD; Francis Walker,MD; Cathy Nunn, RN RVT; Larry Myers, MBA,RDMS, RVT; Charles Tegeler, MD; Sharon Hardin,RVT, RCMS; Dana Meads, RT-R, RVT.

University of Iowa Hospital and Clinics, IowaCity [49]: Harold P. Adams,Jr, MD; Christopher M.Loftus, MD; Lynn A. Vining, RN; Birgette H.Bendixen, MD, PhD; Jos\l=e'\Biller, MD; John D. Cor-son, MD; Patricia H. Davis, MD;John C. Godersky,MD; David L. Gordon, MD; Michael R. Jacoby, MD;Laurens J. Kappelle, MD; Timothy F. Kresowik,MD; E. Eugene Marsh III, MD; Betsy B. Love, MD;Asad R. Shamma, MD; Karla J. Grimsman, RN;Dawn M. Karboski, RN; Ed V. Miller, RVT.

Johns Hopkins Bayview Medical Center, Balti-more, Md [46]: Constance Johnson, MD; CalvinJones, MD; Brenda Stone, CRNP; Mairi PatMaguire, RN; Christopher Earley, MD; PeterKaplan, MD; John Cavaluzzi, MD; Gerald Waters,PA-C; Betty Chachich, RN.

St John's Mercy Medical Center, St Louis, Mo[44]: Arthur Auer, MD; William Logan, MD; MaryWilcox, RN, BSN; Barbara Green, MD; JosephHurley, MD; Richard Pennell, MD; John Woods,MD; Richard Levine, MD; James Nepute, MD; Jef-frey Thomasson, MD; Claire Blackburn, RN, RVT;Marcia Foldes, BSN, RVT; Kathy Klemp, RN,RVT; Becky Nappier, RT, RVT; Karen Ruther-ford, BSN, RVT; Sally Schroer, RN, BSN; JenaHogan, RN, RVT; Lynne Thorpe, RN, RVT.

University of Arizona Health Sciences Center,Tucson [42]: William M. Feinberg, MD; Glenn C.Hunter, MD; Denise C. Bruck, AAS; Victor M.Bernhard, MD; Kenneth E. McIntyre, MD; L. PhilipCarter, MD; Scott S. Berman, MD; Joseph L. R.Mills, MD; Enrique L. Labadie, MD; Darry S.Johnson, MD; Constantine G. Moschonas, MD; Rob-ert H. Hamilton, MD; Scott C. Forrer, MD; JoachimF. Seeger, MD; Raymond F. Carmody, MD; BrendaK. Void, RN, BSN; Richard L. Carlson; Joan E. La-guna; John P. Krikawa; Jenifer J. Devine, RN, BSN,RVT; Amalia M. Castrillo; Sheryl L. Kistler, RN,RVT; Betty Ledbetter; Kathleen (Sue) Dorr.

University ofMississippi Medical Center, Jack-son [39]: Robert R. Smith, MD; Armin F. Haerer,MD; Robin L. Brown, RN, BSN; William Russell,MD; Edward Rigdon, MD; Robert Rhodes, MD;Evelyn Smith, RVT; Michael Graeber, MD; DavidDoorenbos, MD; S. H. Subramony, MD; David L.Gordon, MD.

Milton S. Hershey Medical Center, Hershey, Pa[35]: Robert G. Atnip, MD; Robert W. Brennan,MD; Diane Friedman, RN, MSN; Marsha M. Neu-myer, RVT; Brian L. Thiele, MD; Florence Smith,RN; John D. Barr, MD; R. Bradford Duckrow, MD;Cindy Janesky, MD; Jon W. Meilstrup, MD; KevinP. McNamara, MD; Lawrence D. Rodichok, MD;Leslie Stewart, MD; Maureen Sullivan, MD; MarkWengrovitz, MD.

University ofTexas Southwestern Medical Cen-ter, Dallas [35]: G. Patrick Clagett, MD; Hal Unwin,MD; Wilson Bryan, MD; Chris Matkins; CarolynPatterson; Candy Alway, RDMS; Patty Boyd;Mary Inman, RN; Christie Albiston; Eva Scoggins;John Swilling.

University of California at Los Angeles [35]:Wesley S. Moore, MD; Stanley N. Cohen, MD;Kathleen G. Walden, RN; Samuel S. Ahn, MD; Ed-win C. Amos, MD; J. Dennis Baker, MD; Bruce H.Dobkin, MD; Carlos E. Donayre, MD; Julie A.Freischlag, MD; Hugh A. Gelabert, MD; Sheldon E.Jordan, MD; Herbert I. Machleder, MD; William J.Quinones-Baldrich, MD; Jeffrey L. Saver, MD;Suzie M. El-Saden, MD; Richard C. Holgate, MD;Bradley A. Jabour, MD; J. Bruce Jacobs, MD;Theresa M. Abraham, RN; Candace L. Vescera,


RN; Jeanine A. von Rajcs, RN; Vicki L. Carter,RN, RVT; Dale T. Carter, RVT; De Ette Dix-Goss,RVT; Eugene C. Hernandez, RVT.

Oregon Health Sciences University, Portland[34]: Lloyd Taylor, MD; Bruce Coull, MD; LethaLoboa, RN, RVT; Gregory Moneta, MD; John Por-ter, MD; Richard Yeager, MD; Lucy Whittaker, RN.

Yale University, New Haven, Conn [34]:Lawrence M. Brass, MD; Richard J. Gusberg, MD;Anne M. Lovejoy, PA-C; Pierre B. Fayad, MD;Bauer Sumpio, MD, PhD; George H. Meier, MD;Vicky M. Chang, ANP, MSN; Karen Marzitelli,ANP, MSN; Douglas Chyatte, MD; Lynwood Ham-mers, DO; Fran Lepore, RDMS, RT; Frank J.Pavalkis, PA-C; Jill Mele; Donna Kisiel, PA-C.

University of Arkansas for Medical Sciences,Little Rock [32]: Robert W. Barnes, MD; Michael Z.Chesser, MD; R. Lee Archer, MD; Bernard W.Thompson, MD; Colette MacDonald; Gary W. Bar-one, MD; John F. Eidt, MD; David Harshfield, MD;David McFarland, MD; Jess R. Nickols, MD; E.Carol Howard, BSN, RN, RVT; M. Lee Nix, BSN,RN, RVT; Judy Kaye Overstreet, RN, RVT;Rhonda Troillett, RDMS, RVT.

Medical College ofVirginia (Virginia Common-wealth University), Richmond [27]: John Taylor,MD; H. M. Lee, MD; Patricia Akins; John W. Har-bison, MD; Rhonda M. Pridgeon, MD; Warren L.Felton, MD; Marc Posner, MD; M. Sobel, MD; GuyClifton, MD; Chris Conway; Anne Cockrell; War-ren Stringer, MD; Jean Wingo; Brenda Nichols;Wendy Smoker, MD; Ruth Fisher.

Barrow Neurological Institute at St Joseph'sHospital and Medical Center, Phoenix, Ariz [26]:Robert F. Spetzler, MD; James L. Frey, MD;Joseph M. Zabramski, MD; Sonna Lea Hunsley,RN; Heidi Jahnke, BSN; Kathryn L. Plenge, MD;Rosemary Holland, RVT; Roberta Turner, RVT;Duane Strava, RVT; Susan Stumpff, RVT; JohnHodak, MD; Richard A. Flom, MD; Bruce L. Dean,MD; Richard A. Thompson, MD.

Oschner Clinic, New Orleans, La [21]: RichardHughes, MD; Bruce Lepler, MD; John Bowen, MD;Cheryl Benoit, CCRC; Larry Hollier, MD; JohnOchsner, MD; Richard Strub, MD; Vickie Lang;Vicki Cahanin.

University of Medicine and Dentistry of NewJersey, Newark [20]: Robert W. Hobson II, MD;Fred Weisbrot, MD; Stephen Kamin, MD; TomBack; Zafar Jamil, MD; Carolyn Rogers; Bea Lain-son; Larry Hart.

New England Medical Center, Boston, Mass [19]:Louis Caplan, MD; Thomas O'Donnell, MD; LorettaBarron, RN; Michael Pessin, MD; Dana DeWitt,MD; William Mackey, MD; Michael Belkin, MD;Robert McGlaughlin, RVT; Paula Heggerick, RVT.

Henry Ford Hospital, Detroit, Mich [17]: CalvinB. Ernst, MD; K. M. A. Welch, MD; Judith Wil-czewski, RN; Wendy Robertson, PA-C; Sheila Da-ley, RN; Joseph P. Elliot, Jr, MD; Daniel J. Reddy,MD; Alexander D. Shephard, MD; Roger F. Smith,MD; Steven Levine, MD; Nabih Ramadan, MD;Gretchen Tietjen, MD; Panaytios Mitsias, MD;Mark Gorman, MD; Michalene McPharlin, RN;Suresh Patel, MD; Rajeev Deveshwar, MD; NoraLee, MD; James Kokkinos, MD.

University of New Mexico, Albuquerque [13]:Askiel Bruno, MD; Eric Weinstein, MD; JamesKunkel, MD (deceased); Anna Kratochvil, RN;Edie Johnson, RN, MSN, FNP; Susan Steel, RN.

Sunnybrook Health Science Centre, Universityof Toronto, North York, Ontario [11]: John Norris,MD; David Rowed, MD; Beverley Bowyer, RN;Marek Gawel, MD; Perry W. Cooper, MD; DianneBrodie, RVT.

Harbin Clinic, Rome, Ga [10]: John S. Kirkland,MD, PhD; Jay A. Schecter, MD; Nell W. Farrar,PA-C; Raymond Capps, MD; E. Leeon Rhodes, MD;D. Michael Rogers, MD; Jeffrey T. Glass, MD; Rob-ert Naguszewski, MD; William Naguszewski, MD;Bobbie Maddox; Bobbie Dollison; Lynda Moulton;Polly Cole; Paul Kinsella; Alisa Ansley; Noell Britz.

Roanoke Neurological Associates, Virginia[10]: Don H. Bivins, MD; Edwin L. Williams II, MD;Jesse T. Davidson III, MD; William Elias, MD;

Donna Atkins, RN; Phyllis B. Turner, RN, RVT; J.Gordon Burch, MD; Donald B. Nolan, MD; ReneeSpeese; Candace D. Foley, RN.

Singing RiverHospital, Pascagoula, Miss [6]: Ter-rence J. Millette, MD; Dewey H. Lane, MD; CynthiaAlmond, RN, RVT; Roland J. Mestayer III, MD.

California Pacific Medical Center, San Fran-cisco [4]: Phillip Calanchini, MD; Robert Szarnicki,MD; Pat Radosevich; Linda Elias; Peggy McCor-mick; Charles Gould, MD; Forbes Norris, MD; EricDenys, MD; Robert Bernstein, MD; Donna Du-bono; Keith Atkinson; Mark Peters.

Northwestern University Medical School, Chi-cago, Ill [4]: Bruce Cohen, MD; James Yao, MD;Susan Roston, RN; Donna Blackburn, RN; Jos\l=e'\Biller, MD; Jeffrey Saver, MD; Linda Chadwick,RN; Douglas Chyatte, MD; Walter McCarthy, MD;William Pearce, MD; Jeff Frank, MD; ErnestoFernandez-Beer, MD; James Patrick, MD.

University ofRochester, New York [3]: RichardGreen, MD; Richard Satran, MD; John Ricotta,MD; James DeWeese, MD; Joshua Hollander, MD;Mollie O'Brien, RN; JoAnne McNamara, RN; San-dra Rose, RT, RDMS; Dahne Cohen, MD.

Cleveland Clinic, Ohio [1]: Anthony Furlan,MD; John Little, MD; Bernadine Bryerton, RN;Cathy Sila, MD; Isam Awad, MD; Marc Chimowitz,MD; Sylvia Robertson, ARRT, RDMS; CindyBecker, ARRT, RDMS; David Paushter, MD.

Arteriogram Quality Control.\p=m-\BruceL. Dean,MD; Daniel H. O'Leary, MD.

Doppler and OPG-Gee Quality Control.\p=m-\AnneM. Jones, BSN, RVT, RDMS; John J. Ricotta, MD;William Gee, MD; Nancy Shebel, BSN, RN, RVT.

Pathology.\p=m-\MarkJ. Fisher, MD; Eric A.Schenk, MD.

TIA/Stroke Validation.\p=m-\AnthonyJ. Furlan,MD; Nancy N. Futrell, MD; Michael Kelly, MD;Clark H. Millikan, MD.

General Consultants.\p=m-\H.C. Diener, MD;William S. Fields, MD; Marshall F. Folstein, MD;Jean-Claude Gautier, MD; Michael J. G. Harrison,MD; William K. Hass, MD; Michael G. Hennerici,MD; Richard Satran, MD; Merrill P. Spencer, MD;Gerhard M. von-Reutern, MD.

References1. Timsit SG, Sacco RL, Mohr JP, et al. Early clini-cal differentiation of cerebral infarction from se-vere atherosclerotic stenosis and cardioembolism.Stroke. 1992;23:486-491.2. Hennerici M, H\l=u'\lsb\l=o'\merH-B, Hefter H, Lam-merts D, Rautenberg W. Natural history of asymp-tomatic extracranial arterial disease. Brain. 1987;110:777-791.3. Autret A, Saudeau D, Bertrand PH, PourcelotL, Marchal C, DeBoisvilliers S. Stroke risk in pa-tients with carotid stenosis. Lancet. 1987;1:888-890.4. O'Hollerhan LW, Kennelly MM, McClurken M,Johnson JM. Natural history of asymptomatic ca-rotid plaque. Am J Surg. 1987;154:659-662.5. Norris JW, Zhu CZ, Bornstein NM, ChambersBR. Vascular risks of asymptomatic carotid steno-sis. Stroke. 1991;22:1485-1490.6. Jorgensen L, Torvik A. Ischaemic cerebrovas-cular diseases in an autopsy series, 2. J Neurol Sci.1969;9:285-320.7. Cote R, Barnett HJM, Taylor DW. Internal ca-

rotid occlusion. Stroke. 1983;14:898-902.8. Nicholls SC, Kohler TR, Bergelin RO, PrimozichJF, Lawrence RL, Strandness DE Jr. Carotid ar-

tery occlusion. J Vase Surg. 1986;4:479-485.9. Moore WS, Barnet HJM, Beebe HG, et al. Guide-lines for carotid endarterectomy. Stroke. 1995;26:188\x=req-\201.10. Riles TS, Adelman MA, Gibstein LA. Incidenceof premonitory transient ischemic attacks beforeischemic strokes. In: Veith FJ, ed. Current CriticalProblems in Vascular Surgery. St Louis, Mo: Qual-ity Medical Publishing Inc; 1994;6:264-268.11. Till JS, Toole JF, Howard VJ, Ford CS, Willi-ams D. Declining morbidity and mortality of ca-

rotid endarterectomy. Stroke. 1987;18:823-829.12. Wilson SE, Mayberg MR, Yatsu FM. Defining

the indications for carotid endarterectomy. Sur-gery. 1988;104:932-933.13. The Asymptomatic Carotid AtherosclerosisStudy Group. Study design for randomized pro-spective trial of carotid endarterectomy for asymp-tomatic atherosclerosis. Stroke. 1989;20:844-849.14. Howard G, Chambless LE, Baker WH, et al. Amulticenter validation study of Doppler ultrasoundversus angiogram. J Stroke Cerebrovasc Dis. 1991;1:166-173.15. Brott T, Tomsick T, Feinberg W, et al, for theAsymptomatic Carotid Atherosclerosis Study In-vestigators. Baseline silent cerebral infarction inthe Asymptomatic Carotid Atherosclerosis Study.Stroke. 1994;25:1122-1129.16. Moore WS, Vescera CL, Robertson JT, BakerWH, Howard VJ, Toole JF. Selection process forsurgeons in the Asymptomatic Carotid Atheroscle-rosis Study. Stroke. 1991;22:1353-1357.17. Folstein MF, Folstein SE, McHugh PR. 'Mini-Mental State': a practical method for grading thecognitive state of patients for the clinician. J Psy-chiatr Res. 1975;12:189-198.18. Lefkowitz DS, Brust JC, Goldman L, et al. Apilot study of the end point verification system inthe Asymptomatic Carotid Atherosclerosis Study.J Stroke Cerebrovasc Dis. 1992;2:92-99.19. Jennett B, Bond M. Assessment of outcomeafter severe brain damage: a practical scale. Lan-cet. 1975;1:480-484.20. Hobson RWII, Weiss DG, Fields WS, et al, forthe Veterans Affairs Cooperative Study Group. Ef-ficacy of carotid endarterectomy for asymptomaticcarotid stenosis. N Engl J Med. 1993;328:221-227.21. North American Symptomatic Carotid Endar-terectomy Trial Collaborators. Beneficial effect of ca-rotid endarterectomy in symptomatic patients withhigh-grade stenosis. NEngl JMed. 1991;325:445-453.22. Kalbfleisch JD, Prentice RL. The StatisticalAnalysis of Failure Time Data. New York, NY:John Wiley & Sons Inc; 1980:14.23. Lachin JM. Statistical properties of random-ization in clinical trials. Control Clin Trials. 1988;9:289-311.24. O'Brien PC, Fleming TR. A multiple testing pro-cedure for clinical trials. Biometrics. 1979;35:549-556.25. Laupacis A, Naylor CD, Sackett DL. How shouldthe results of clinical trials be presented to clini-cians? ACP J Club. May/June 1991:A12-A14.26. CASANOVA Study Group. Carotid surgeryversus medical therapy in asymptomatic carotidstenosis. Stroke. 1991;22:1229-1235.27. Mayo Asymptomatic Carotid EndarterectomyStudy Group. Results of a randomized controlledtrial of carotid endarterectomy for asymptomaticcarotid stenosis. Mayo Clin Proc. 1992;67:513-518.28. Halliday AW, Thomas D, Mansfield A. TheAsymptomatic Carotid Surgery Trial (ACST): ra-tionale and design. Eur J Vasc Surg. 1994;8:703-710.29. European Carotid Surgery Trialists' Collabo-rative Group. European Carotid Surgery Trial: in-terim results for symptomatic patients with severe

(70-99%) or with mild (0-29%) carotid stenosis. Lan-cet. 1991;337:1235-1243.30. Toole JF, Castaldo JE. Accurate measurementof carotid stenosis: chaos in methodology. J Neu-roimaging. 1994;4:222-230.31. Rothwell PM, Gibson RJ, Slattery J, Sellar RJ,Warlow CP, for the European Carotid Surgery Tri-alists' Collaborative Group. Equivalence ofmeasure-ments ofcarotid stenosis. Stroke. 1994;25:2435-2439.32. Alexandrov AV, Bladin CF, Magissano R, Nor-ris JW. Measuring carotid stenosis: time for a re-

appraisal. Stroke. 1993;24:1292-1296.33. Horn M, Michelini M, Greisler HP, Littooy FN,Baker WH. Carotid endarterectomy without arte-riography. Ann Vasc Surg. 1994;8:221-224.34. Chervu A, Moore WS. Carotid endarterectomywithout arteriography. Ann Vasc Surg. 1994;8:296\x=req-\302.35. Morasch MD, Hodgett D, Burke K, Baker WH.Selective use of the intensive care unit followingcarotid endarterectomy. Ann Vasc Surg. In press.


endarterectomy for asymptomatic carotid artery stenosis · table 1.—baseline characteristicsof...

Documents