The International Cooperative Study of
Anastomosis (EC/IC Bypass Study):
Methodology and Entry Characteristics
THE EC/IC BYPASS STUDY GROUP
SUMMARY This report summarizes the protocol of a randomized trial of superficial temporal artery-middle cerebral artery (STA-MCA) bypass procedure and presents the entry characteristics of its patients. The trial has been designed to determine whether this EC/IC bypass reduces the rate of subsequent stroke among patients with recent hemispheric or retinal strokes and/or transient ischemic attacks who have angiographically proven atherosclerotic narrowing or occlusion of the ipsilateral internal carotid or middle cerebral artery.
Of the 1377 eligible patients entered from the 71 participating centers, 714 (52%) have been assigned to medical treatment alone (daily aspirin and aggressive hypertension control) while 663 (48%) have been assigned to receive STA-MCA bypass in addition to medical therapy. The two treatment groups have been well balanced for important prognostic factors. Bypass patency rates of 95% have been documented. At the end of the study in mid-1985, an average follow-up of five years and a minimum follow-up of 33 months will have been achieved.
On both clinical and methodologic grounds, this study will have provided a rigorous test of the STA-MCA bypass procedure.
Stroke Vol 16, No 3, 1985 THE FIRST EXTRACRANIAL TO
INTRACRANI-AL ARTERIINTRACRANI-AL ANASTOMOSIS was performed in 1967 and this procedure came into widespread use during the next decade. In 1977, the International Co-operative Study of Extracranial/Intracranial Arterial Anastomosis (EC/IC Bypass Study) was initiated in order to test the value of this procedure among patients with symptomatic atherosclerotic lesions of the inter-nal and/or middle cerebral arteries. Entry of patients began on August 19, 1977 and the last patient was entered September 30, 1982; final follow-up will be completed in March 1985.
This paper describes the methods employed in the study and the entry characteristics of its patients.
Objective and Organization
The EC/IC Bypass Study is a randomized trial de-signed to determine whether anastomosis of the super-ficial temporal artery to the middle cerebral artery can reduce, despite perioperative stroke and death, subse-quent events of stroke and stroke-related death among patients with symptomatic, surgically-inaccessible (to endarterectomy) atherosclerotic stenosis or occlusion of the internal carotid or middle cerebral arteries. "Perioperative" refers to stroke and death (regardless of cause) occurring intraoperatively and during the 30 days immediately following surgery. Stroke resulting from the initial angiography is not included as an out-come event because angiography was required to es-tablish eligibility for the study and thus preceded the
A list of the participating centres, investigators and study committee members appears in the Appendix.
This study is funded by the U.S. National Institutes of Health, Grant #ROIN S14164.
Address correspondence to: EC/IC Bypass Study, University Hospi-tal, P.O. Box 5339, Postal Station A, London, Ontario, N6A-5A5.
actual entry of both medical and surgical patients. The trial, funded in 1977 by the U.S. National Insti-tute of Neurological Communicative Disease and Stroke (NINCDS), has been coordinated clinically by a Central Office in the Department of Clinical Neuro-logical Sciences at the University of Western Ontario in London, Ontario and has been coordinated method-ologically from a Methods Center in the Department of Clinical Epidemiology and Biostatistics at McMaster University in Hamilton, Ontario. Investigators from these two centers have comprised a Steering Commit-tee which has been responsible for the overall adminis-tration and day-to-day execution of the study. Major decisions concerning study policy have been vetted by an Executive Committee consisting of members of the Steering Committee, investigators from participating centers in North America, Europe and the Far East, and representatives from the NINCDS.
A clinical center has been accredited for participa-tion in the trial if it could demonstrate close collabora-tion between specialists in neurosurgery, neurology, and neuroradiology. Moreover, prior to participation every surgeon was required to demonstrate the achievement of at least 80% graft patency (proven by angiography) on at least 10 consecutive EC/IC proce-dures on patients with appropriate cerebrovascular lesions.
Annual meetings and workshops have been held in North America, Europe and Asia to bring together clinical investigators, to inform them of the progress of the trial, to resolve any questions concerning its execu-tion, and to maintain a high degree of enthusiasm and collaboration.
The conduct and results of the study have been re-viewed twice yearly by a Monitoring Committee ap-pointed by the Stroke and Trauma Program of the NINCDS.
398 STROKE VOL 16, No 3, MAY-JUNE 1985
Methods 1. Patient Eligibility
To be eligible for the trial, patients have had to satisfy clinical, radiological, and exclusion criteria: a. Clinical Inclusion Criteria:
Patients must have experienced, within three months prior to entry, one or both of the following: (i) transient ischemic attack(s) (TIA) in the carotid distri-bution (one or more episodes of distinct focal neuro-logical dysfunction or monocular blindness (amaurosis fugax), the symptoms and signs of which cleared com-pletely in less than 24 hours); (ii) minor completed stroke(s) in the carotid distribution (one or more events of distinct focal neurologic dysfunction or amaurosis fugax, the signs of which persisted for more than 24 hours). Patients without useful residual function in the affected territory have not been entered.
b. Radiological Inclusion Criteria:
An angiogram demonstrating an atherosclerotic le-sion in the appropriate territory has had to be submitted for central confirmation and adjudication by the princi-pal neuroradiologist; bilateral common carotid arterio-grams have been requested on all patients, and angiog-raphy of the vertebro-basilar circulation has been optional.
One or more of the following atherosclerotic lesions must have been demonstrated in vessels appropriate to the patient's symptoms: (i) stenosis or occlusion of the middle cerebral artery trunk; (ii) stenosis of the internal carotid artery at or above the C2 vertebral body (i.e. inaccessible to carotid endarterectomy); or (iii) inter-nal carotid artery occlusion.
For radiological eligibility, stenosis has been de-fined as "any recognizable atherosclerotic lesion of the surgically inaccessible portion of the internal carotid artery or middle cerebral artery which, in the opinion of the attending neurosurgeon, might reasonably be expected to benefit from bypass surgery."
c. Exclusion Criteria:
Patients have been excluded from the study if they were unable to meet at least the following functional standards: (i) capability of self-care for most activities of daily living (may require some assistance); (ii) re-tention of some useful residual function in the affected arm or leg; (iii) comprehension intact with no evidence of Wernicke's receptive aphasia; (iv) no, or only mild, motor (expressive, Broca's) aphasia; (v) ability to han-dle their own oropharyngeal secretions.
Patients also have been excluded from the trial if any of the following pertained: inability to provide in-formed consent; evidence that the original stroke was due to cerebral hemorrhage; within eight weeks of an acute cerebral ischemic event; exhibition of non-atherosclerotic conditions causing or likely to cause cerebral dysfunction (fibromuscular dysplasia, arteri-tis, blood dyscrasia, a cardiac source of cerebral em-boli, chronic atrial fibrillation, complete heart block, significant valvular heart disease, cardiomyopathy, or nonatherosclerotic dissection); the presence of any morbid condition(s) likely to lead to death within five years (cancer, renal failure (BUN > 5 0 mg%),
cardio-megaly (cardiothoracic ratio of >0.50 (>0.55 in Japa-nese patients),2"* or any hepatic or pulmonary disease constituting an unacceptable anesthetic risk); the oc-currence of ischemic symptoms isolated to the verte-brobasilar circulation; prior participation in the study (regardless of the occurrence of new ischemic events or success or failure of previous therapy); myocardial infarction within the preceding six months; a fasting blood sugar of 300 mg% or more on the most recent assessment despite appropriate therapy; diastolic blood pressure > 110 mm Hg (using disappearance of sounds for diastolic pressure) despite appropriate medical therapy. Once uncontrolled diabetes or hypertension were corrected, otherwise eligible patients could be entered.
Every entry form has been reviewed by both the Central Office and the Methods Center. All angio-grams have been reviewed (without knowledge of the treatment group to which the patient had been random-ized) by the principal neuroradiologist at the Central Office. When this review has suggested that an ineligi-ble patient had been entered, an external group of adjudicators (who were not trial participants and were "blind" to the patient's allocation) has reviewed the entry data and angiograms and decided whether the patient should be excluded from the trial,
d. Specific Clinical Situations:
Because cerebrovascular disease often involves multiple sites, guidelines for managing, and arbitrary rules for analyzing specific clinical situations have been established. When "tandem lesions" (two lesions in the same vessel or sequence of vessels, one proxi-mal to the other) have existed, and both lesions have fulfilled the entry criteria, the patient has been entered for the more distal lesion. If the proximal lesion has been amenable to endarterectomy, the decisions of whether to perform endarterectomy, and the selection of the initial site for surgery (i.e. endarterectomy or EC/IC bypass), have been left to the judgment of the participating surgeon. In patients judged to require external carotid endarterectomy as a preparation for bypass, it has been recommended that this be per-formed only after the patient has been randomized to the surgical group. If external carotid endarterectomy had been performed prior to randomization, then 30 days have had to elapse following endarterectomy be-fore the patient could have been entered.
When contralateral carotid disease has existed and has been accessible to endarterectomy, the decision of whether and when to perform contralateral endarterec-tomy has again been left to the participating neurosur-geon. Once again, if contralateral endarterectomy has been carried out first, 30 days have had to elapse fol-lowing contralateral endarterectomy before the patient could be entered.
When eligible patients have had appropriate symp-toms and angiographic lesions in both carotid distribu-tions, the patient has been entered for the most recent clinically eligible event. If more than one TIA or stroke has occurred prior to entry, the most recent has served as the basis for enrollment.
In all of the above circumstances, if a 30-day wait-ing period has applied, the patient was still eligible for randomization, despite the passage of more than three months since the last episode of cerebral ischemia. 2. Baseline Investigations
At entry, detailed neurological and medical histories and examinations have been carried out and recorded on standardized forms. The examination has included a 12-item functional status assessment in which the patient's ability to perform activities of daily living such as eating, toileting, and ambulation have been rated on a three- point scale: 1. able to perform task without difficulty; 2. able to perform task with difficul-ty; or 3. unable to perform task without mechanical or personal assistance. Additional historical data have been gathered about employment status, history of dia-betes mellitus, hypertension, angina pectoris, myocar-dial infarction, intermittent claudication, cardiac sur-gery or other serious illness, current medications, and smoking habits. The patient's blood pressure, heart rate and rhythm, cardiac murmurs, and neck bruits have been recorded. Finally, the following baseline investigations have been carried out: hemoglobin; platelet count; prothrombin time; random blood glu-cose; blood urea nitrogen; cholesterol; electrocardio-gram (for left ventricular hypertrophy, new or old myocardial infarction and rhythm); chest x-ray to esti-mate the cardiothoracic ratio; and, if available, a com-puterized tomographic (CT) scan of the head. 3. Random Allocation to Therapy
If the patient has been judged eligible on clinical and radiographic grounds, informed consent has been re-quested. If obtained, a tentative date for surgery has been booked (to minimize the time interval between randomization and surgery) and the Methods Center has been contacted by telephone for registration and random allocation (for logistic reasons, a separate ran-domization center has been set up in Kyoto, Japan, for patients from Japan and Taiwan). To ensure balance between the medical and surgical limbs of this trial, a stratified randomization has been carried out. The stra-ta have been defined by the underlying vascular lesion (stenosis or occlusion of the middle cerebral or internal carotid artery), the presence or absence of a related neurologic deficit and, in the case of patients with internal carotid occlusion, whether related symptoms have occured since its angiographic demonstration (some centers have joined the trial with a commitment to exclude patients with no symptoms since their inter-nal carotid occlusions were demonstrated). Then, based on a computer-generated randomization scheme established at the Methods Center for each participat-ing center and stratum, the patient has been assigned to either the medical or surgical limb of the trial. 4. Treatment
Patients randomized to the surgical limb have under-gone microsurgical, end to side, anastomosis of the superficial temporal or occipital artery to a cortical
branch of the middle cerebral artery. All participating surgeons have agreed to follow an identical procedure in the performance of the extra- to intracranial bypass, but minor variations in the surgical technique have been permitted and left to the individual surgeon's discretion.
Because previous randomized trials have estab-lished the efficacy of aspirin in patients with TIA and minor stroke,6 all patients (both medical and surgical) have been prescribed acetylsalicylic acid 325 mg q.i.d. throughout the trial, unless contraindicated or not tol-erated. Furthermore, the control of hypertension has been stressed and monitored in both medical and surgi-cal patients; the treatment of other risk factors has been left to individual clinical judgment.
All patients (both medical and surgical) have been examined by a participating neurologist six weeks after randomization and at three-month intervals thereafter (surgical patients have received an additional review approximately 30 days following surgery). At each visit, an interim history has been obtained and a de-tailed neurologic examination has been performed, in-cluding the functional status assessment. In addition, repeat assessments have been made of the patient's employment status, medical and smoking history, medications, blood pressure and cardiac status.
Surgical patients have undergone postoperative an-giography; the recommended timing has been three to six months following bypass. All the post-operative angiograms have been reviewed by the principal neu-roradiologist at the Central Office. The success of the anastomosis has been assessed by the following fea-tures: the comparative size (smaller; same; or larger) of the superficial temporal artery proximal to the anasto-mosis on preoperative and postoperative angiograms; the degree of retrograde flow in the middle cerebral circulation (none; slight; to the middle cerebral artery bifurcation; or to the internal carotid) and the number of branches of the middle cerebral artery filled. Overall flow through the bypass has been estimated as: none, slight, moderate, good, or excellent. Further follow-up angiograms have been recommended when the first post-operative angiogram has suggested poor bypass function.
A random sample of 20 preoperative and 20 postop-erative angiograms have been reviewed a second time by the principal neuroradiologist (who had been blind-ed to his original report) to assess intraobserver agree-ment.
Other medical problems, medications and perioper-ative complications have been monitored to detect un-expected adverse effects of any medical and/or surgi-cal treatment that patients have undergone during the trial.
6. Withdrawals and Crossovers
Only patients who refused to continue under obser-vation were to have been withdrawn from the trial; all others were to be retained. EC/IC surgery on the side
400 STROKE VOL 16, No 3, MAY-JUNE 1985
opposite to the qualifying lesion and internal carotid endarterectomy on either side have been discouraged; however, patients who have undergone these addition-al procedures have not been withdrawn from the triaddition-al. A second EC/IC bypass on the same side has been permitted in a small number of surgical patients in whom it has appeared likely to improve bypass func-tion; however, these surgical procedures have not, by themselves, constituted events, and such patients have not been withdrawn.
Patients who initially accepted their randomization to medical therapy but later underwent EC/IC bypass on the randomized side, or patients who accepted ran-domization to surgical therapy and then declined the operation were labelled "crossovers". Strokes occur-ring to such patients before their crossover are to be charged to the medical limb and the surgical limb re-spectively in the primary analysis. Those who had no events prior to the crossover were judged to have been randomized improperly and to have failed to give prop-er informed consent. They wprop-ere considprop-ered as one of the exclusion categories.
The main study events are the occurrence of fatal or nonfatal stroke following randomization. However, in order to include any excess perioperative risk, deaths from all causes have been included if they occurred between randomization and 30 days following EC/IC bypass among surgical patients and for a comparable duration among patients in the medical group.
Events have been identified and classified in one or TABLE 1 Stroke Severity Scale
Sever-ity Grade 1 2 3 4 5 6 7 8 9 10 11 Impairment* none none none Minor, in one or more domains Major, in only one domain Major, in any two domains Major, in any three domains Major, in any four domains Major, in all five domains Reduced level of consciousness Death Neuro-logic Symptoms Present Absent Present Present nat na na na na na na Neuro-logic Signs Absent Present Present Present na na na na na na na Impairment in the domains of swallowing, self-care, ambula-tion, communicaambula-tion, and comprehension. If independence is main-tained despite the impairment, it is classified as minor; if independ-ence is lost, it is classified as major.
tNeurologic signs and symptoms are integrated into the higher grades of impairment.
more of three ways: First, participating neurologists have identified and reported them during routine pa-tient follow-up. Second, the principal investigators at the Central Office have identified events during their Weekly Clinical Reviews of the follow-up data, sup-plemented when necessary by further communication with the participating neurologist. Third, a computer program at the Methods Center has identified possible events by comparing the functional status, signs and symptoms on each follow-up with those recorded on previous follow-up assessments.
The best current means for diagnosis (including CT scan where available) have been used to differentiate strokes due to infarction from those due to hemor-rhage. The severity of the stroke, in terms of the im-pairment of functional status, has been rated on a 10-point scale, using information derived from the neurologic follow-up forms. This Stroke Severity Scale is summarized in table 1.
Transient ischemic attacks, other major cardiovas-cular morbidity and deaths from non-cerebrovascardiovas-cular causes have also been recorded but have not constitut-ed main events. Thorough documentation has been sought for every death from any cause, including: au-topsy findings; emergency room, operative and hospi-tal discharge reports; death certificates; and the obser-vations of any witnesses.
8. Adjudication of Events
Adjudication of the cause of death and the occur-rence and severity of strokes has been performed inde-pendently by a neurologist and a neurosurgeon who have not cared for patients in the study and have been blind both to the patients' treatments and to each oth-er's judgments. Disagreements between the two adju-dicators have been resolved by consensus. Disagree-ments between the adjudicators' consensus and the results of the Central Office's Weekly Clinical Re-views have been resolved through discussions with the principal clinical investigators.
Sample Size and Interim Analyses
The required sample size and duration of follow-up initially had been determined on the basis of data avail-able when the protocol had been designed (February 1977). The following estimates and specifications have been employed:
(1) For medically-treated patients: an expected com-bined stroke and stroke-death rate of 23.6% over five years.7"10
(2) For surgically-treated patients: an estimated 30-day perioperative stroke and death rate of 4%, followed by a 50% reduction in the subsequent rate of fatal and non-fatal stroke compared with medi-cally-treated patients, for a net surgical benefit of 33% reduction in the five-year risk of fatal and non-fatal stroke. It has been agreed that this degree of risk reduction constituted a clinically important surgical benefit which, if achieved in the trial, would justify advocating the EC-IC bypass proce-dure.
(3) A risk of 5% of drawing the false-positive
sion that surgery is beneficial when, in truth, it is not (alpha = 0.05, one-tailed).
(4) A 10% risk of drawing the false-negative conclu-sion that surgery is no better than medical therapy when, in truth, surgery results in the clinically important benefit stated above (beta = 0.10; pow-er = 90%).
These considerations had required a sample size of 442 patients per treatment group, followed for an aver-age of five years. Allowing for drop-outs, an initial total sample size target of 1,000 patients had been established.
By February 1981, 35% of trial patients had been entered with internal carotid artery occlusion without further ischemic symptoms since angiography. Be-cause this group had been thought to be at a lower risk of subsequent stroke, it had been feared that large numbers of such cases might have obscured a surgical benefit to other patients. This situation was discussed by Executive Committee members who were blind with respect to events in both treatment groups, and they proposed to increase the total sample size target to 1,400 patients. This proposal was accepted by the Monitoring Committee and was implemented.
In order to discharge the ethical responsibility of stopping the trial as soon as a clear cut conclusion is evident, the accumulating data have been analyzed at six-month intervals. An "alerting" procedure has been developed and adopted that calls for the early stopping of the study to have been considered if two consecutive interim analyses have shown either that the surgically treated group are faring significantly better (at the p = 0.02 level) or that it has become very unlikely that any surgical benefit could have been demonstrated.12 The results of these interim analyses, known only to the principal epidemiologic investigator and chief biostat-istician at the Methods Center, have been summarized and reported to the Monitoring Committee.
Appropriate allowance has been made for the statis-tical effect of these interim challenges of the data; if the trial has not been terminated early, ap-value of 0.04 or less would have constituted statistical significance in the final analysis.
Results 1. Entry Characteristics
Seventy-one centers, on three continents, have en-tered a total of 1495 patients between August 1977 and September 1982. The review of entry documentation by the Central Office and Methods Center, supported by an independent review by external adjudicators, has led to the exclusion of 118 (7.9%) of these patients who, usually on the basis of pre-randomization data, have failed to meet all entry criteria. The reasons for these exclusions appear in table 2. An additional ten patients in the surgical group and nine in the medical group have died or experienced a stroke or a major medical disorder (such as pulmonary embolism) that has prevented them from receiving their assigned treat-ment; however, these patients have remained in the
TABLE 2 Trial Exclusions
Informed consent not properly obtained
Refused surgery* Demanded surgery*
No informed consent obtained from patient
Surgeon not informed prior to randomization Angiographically ineligible
Distal branch occlusion No lesion
Nonatherosclerotic disease Lesion accessible to
conventional endarterectomy Inadequate angiography Lack of appropriate symptoms
Event more than 3 months prior to randomization No evidence of Cerebral vascular
disease (1 migraine, 1 seizure) No evidence of appropriate
Events in vertebral basilar, not carotid, distribution Co-morbid conditions
Tumor Cardiomegaly Atrial fibrillation Renal failure
Open TB (not suitable for surgery)
Inappropriate surgery Vein graft performed, not
Innominate to ICA bypass done in lieu of EC/IC procedure
Bypass not done following carotid endarterectomy Poor functional status at entry Suspension/withdrawal of centre Total Percent Excluded Surgical 22 1 3 9 7 7 . 3 1 3 1 1 1 1 1 3 1 2 67 = 118/1495 Medical 18 2 5 5 2 1 1 3 1 1 1 1 6 2 1 1 51 = 7.9% *No preceding stroke or major medical "event."
study, and their events have been charged to the group to which they were randomized.
Of the 1,377 eligible patients, 714 (52%) have been randomized to medical and 663 (48%) to surgical ther-apy. Randomization has created balanced treatment groups with respect to both important prognostic char-acteristics (table 3) and underlying vascular lesions (table 4). There have been no striking differences be-tween the two treatment groups for any of these fea-tures.
402 STROKE VOL 16, No 3, MAY-JUNE 1985
Over 70% of trial patients have had abnormal neuro-logical findings at entry, but the majority have had little or no functional impairment (table 5). [Minor impairment means difficulty in performing (but not in need of external assistance for) any activity of daily living in: swallowing, self-care (feeding, dressing, personal hygiene), ambulation, communication or comprehension]. Major impairment means a loss of independence in one or more of the above activities of daily living, but not of sufficient severity to preclude entry into the trial). Again, the balance between medi-cal and surgimedi-cal groups has been excellent.
As expected, there have been regional differences in
TABLE 3 Entry Characteristics of Eligible Study Participants
TABLE 3 (continued)
Medical (n = 714)
Surgical (n = 663) Age (mean years)
Sex male female Race white black asian
north american indian Randomizable event
TIA (retinal as % of TIA) stroke (retinal as % of
stroke) Previous stroke(s) Previous TIA(s) Other medical problems
hypertension diabetes angina pectoris prior myocardial
infarc-tion intermittent claudication Medications at entry antiplatelet aggregating agents acetylsalicylic acid other anticoagulants antihypertensive agents Smoking history never smoked former smoker cigarette pipe current smoker cigarettes pipe/cigar 56 82% 18% 77% 7% 16% 0% 34%(16%) 66% (2%) 78% 62% 48% 18% 8% 9% 11% 51% 6% 4% 31% 17% 25% < 1 % 55% 3% 56 81% 19% 78% 6% 16% < 1 % 33%(12%) 67% (1%) 78% 66% 52% 17% 10% 11% 13% 47% 7% 3% 33% 19% 22% 0% 57% 2% Medical (n = 714) Surgical (n = 663) Employment status employed full time part time unemployed
due to cerebral vas-cular disease other reason both
reason unknown Vascular examination
heart rate (mean beats/ min)
blood pressure at entry systolic (mean mm Hg) diastolic (mean mm Hg) carotid bruit(s) ipsilateral contralateral bilateral 56% 51% 5% 44% 15% 28% < 1 % 1% 74 144 85 4% 7% 7% 53% 50% 3% 47% 30% 75 145 85
the site of vascular lesions, as shown in table 6. Inter-nal carotid occlusion has been more common in North American and European patients while disease of the middle cerebral arteries has been present in more than 50% of Asian patients, who also have exhibited more tandem lesions.
Some manifestations of atherosclerosis, such as coronary and peripheral vascular disease, have been more common among North American patients (table 7). There have been no differences, however, in the rates of prior cerebral vascular disease among patients from the three regions.
2. Surgical Performance
The only data currently available for release on sur-gical performance concern patency rates following the TABLE 4 Patterns of Atherosclerotic Carotid Arterial Disease in Medical and Surgical Patients*
Medical (n = 714)
Surgical (n = 663) Isolated MCA stenosis 4%
Isolated MCA occlusion 4% Isolated ICA stenosis 17%
Isolated ICA occlusion 59% Tandem lesions 17%
15% 58% •Definitions: MCA = Middle cerebral artery; ICA = internal carotid artery; Stenosis = recognizable atherosclerotic lesion (see Methods section, "Radiological Inclusion Criteria"); Tandem le-sions = more than one atherosclerotic lesion in the same vessel or sequence of vessels, one proximal to the other.
TABLE 5 Neurological and Functional Status* at Entry
Neurological exam normal Neurological exam abnormal
no functional impairment minor functional impairment major functional impairment
Medical (n = 714) 27% 3 1 % 36% 6% Surgical (n = 663) 25% 30% 36% 8% *Functional status: see Methods section, "Baseline Investiga-tions" and "Entry Characteristics" definitions.
bypass procedure; a patency rate of 95% has been achieved. The agreement between duplicate interpreta-tions of pre- and post-operative angiograms has been high, and will be reported elsewhere.
3. Patient Follow-Up
No patients have been withdrawn from the trial, and only 2 patients have been lost to follow-up. Reproduc-ibility studies, to be published elsewhere, have demon-strated high agreement in the application of the func-tional status assessment and stroke severity scales. 4. Center Performance
The number of eligible patients entered per center has varied considerably, with 2 1 % of the 71 centers having contributed 50% of the study patients. Because this variation in patient entry from center to center has raised the possibility that technical expertise may also vary from center to center, the bypass graft patency rates for small centers (entering less than 25 patients each) have been compared with those for large centers (entering 25 or more patients). These patency rates have been identical (95% and 96%, respectively).
Graft patency rates have also been examined for regional differences. They have been equally high in all three regions: 92% in North America, 94% in Eu-rope and 97% in Asia.
The International Cooperative Study of Extracrani-al/Intracranial Arterial Bypass is the first randomized trial of the EC/IC bypass procedure. Patient entry has been completed. Highly competent neurosurgical, neurological and neuroradiological teams from around the world have recruited and fully documented the clinical state of a sufficient number of appropriate pa-tients to test the value of the bypass procedure in a rigorous fashion. The surgical procedure has been ex-ecuted in a randomized half of these patients with technical success, as demonstrated by very high graft patency rates. No patients have withdrawn from the trial, and only 1 has been lost to follow-up. Thus, whatever the result of the final analysis, it should be both bias-free and of direct applicability to patients similar to those who participated.
In addition to information concerning the clinical value of the EC/IC bypass procedure, the study will provide new information on several other important topics. For example, the cohort of patients assigned to
medical therapy will provide valuable information about the clinical course and prognosis of both middle cerebral artery stenosis and occlusion and internal ca-rotid artery stenosis and occlusion, as well as deter-mining whether there are any important geographic differences in the course of cerebrovascular disease. Furthermore, the detailed review of pre- and post-op-erative angiograms should identify features that pre-dict success in the prevention of future stroke and stroke-related death. Finally, the three-monthly as-sessment of the functional status of patients should reveal the quality of life experienced by patients with these varieties of cerebrovascular disease as well as whether this quality is affected by surgery.
Comparisons between this trial and previous investi-gations of the EC/IC bypass procedure are scientifical-ly tenuous; these earlier studies were not controlled trials, involved fewer patients, and usually included non-consecutive patients with a far wider variety of cerebrovascular conditions (including moyamoya dis-ease, fibromuscular hyperplasia, and venous stasis ret-inopathy). Nonetheless, in the only previous series in which there is sufficient published data to permit com-parison,13 the patients are remarkably similar to those in the present trial. For example, 80% of Sundt and colleagues' patients are males, with an average age of 55 years, and internal carotid artery occlusion is pres-ent in 62%; in patipres-ents randomized to surgery in the present trial, 81 % are males, with an average age of 57 years, and 59% have ICA occlusion. Thus, the patients entered into this trial may be closely comparable to analogous patients seen in clinical centers not partici-pating in the study.
The superficial temporal artery to middle cerebral artery anastomosis procedure is an elegant result of modern technology and a testimonial to the sophistica-tion of modern surgical expertise. Although measure-ments of cerebral blood flow and metabolism suggest that the EC/IC bypass works as a conduit, they cannot determine whether the benefits of the procedure outweigh its risks or whether it is superior to medical management.14 Given the extraordinary variability in the clinical course of cerebrovascular disease, these
Patterns of Vascular Disease in Patients by Geographic
Isolated MCA stenosis Isolated MCA occlusion Isolated ICA stenosis
Isolated ICA occlusion Tandem lesions North America (n = 679) 3% 2% 16% 62% 17% Europe (n = 477) 3% 4% 15% 67% 12% Asia (n = 221) 9% 12% 19% 28% 33% *Definitions: MCA = Middle cerebral artery; ICA = internal carotid artery; Stenosis = recognizable atherosclerotic lesion (see Methods section, "Radiological Inclusion Criteria"); Tandem le-sions = more than one atherosclerotic lesion in the same vessel or sequence of vessels, one proximal to the other.
404 STROKE VOL 16, No 3, MAY-JUNE 1985
TABLE 7 Clinical Description of the Patients at Entry by Region
Age (mean years) Sex
Qualifying condition TIA (retinal as % of TIA) stroke (retinal as % of stroke) Previous stroke(s)
Previous TIA(s) Other medical problems
hypertension diabetes angina pectoris
prior myocardial infarction intermittent claudication Medications at entry
antiplatelet aggregating agents acetylsalicylic acid other anticoagulants antihypertensive agents Smoking history never smoked former smoker cigarette pipe smoker current smoker cigarettes pipe/cigar Employment status employed full time part time unemployed
due to cerebral vascular disease other reason both unknown reason Vascular examination blood pressure systolic (mean mm Hg) diastolic (mean mm Hg) heart rate (mean beats/min) carotid bruit(s) ipsilateral contralateral bilateral North America (n = 679) 59 78% 22% 42%(15%) 58% (2%) 72% 72% 58% 22% 15% 16% 17% 53% 5% 5% 42% 19% 21% 0% 57% 4% 45% 41% 5% 55% 16% 37% 1% < 1 % 144 84 76 6% 10% 10% Europe (n = 477) 53 87% 13% 25%(16%) 75% (1%) 84% 61% 39% 13% 3% 5% 10% 45% 9% 1% 22% 13% 29% < 1 % 56% 2% 57% 54% 3% 43% 16% 25% <I% 1% 146 87 73 4% 6% 3% Asia (n = 221) 56 81% 19% 24%(6%) 76%(1%) 82% 45% 47% 16% 4% 4% 1% 43% 4% 1% 19% 28% 20% 0% 52% < 1 % 77% 74% 3% 23% 11% 12% 0% 0% 141 85 72 1% 1% 0%
broader questions can only be answered through rigor-ous comparisons with randomized controls. The Inter-national Cooperative Study of Extracranial/Intracrani-al ArteriExtracranial/Intracrani-al Anastomosis will provide the best test to date of this procedure when used for atherosclerotic lesions of the carotid and middle cerebral arteries.
1. Yasargil MG: Anastomosis between the superficial temporal artery and a branch of the middle cerebral artery. In Microsurgery Ap-plied to Neuro- surgery. M.G. Yasargil (ed.), George Thieme Verlag, Stuttgart, pp. 105-115, 1969
2. Oberman A, Myers AR, Karunas TM, Epstein FH: Heart size of adults in a natural population-Tecumseh, Michigan: Variation by sex, age, height and weight. Circulation 35: 724, 1967 3. Shibata H, Matsuzaki T, Shiohida K, Saito N: Study on usefulness
of cardiothoracic ratio in the aged. Jpn J Geriatrics 13: 406, 1976 4. Yano K, Ueda S: Coronary heart disease in Hiroshima, Japan: Analysis of the data at the initial examination, 1958-1960. Yale J Biol Med 35: 504, 1962-63
5. Peerless SJ: Techniques of cerebral revascularization. Clin Neuro-surgery 23: 258-269, 1976
6. Canadian Cooperative Study Group: A randomized trial of aspirin and sulfinpyrazone in threatened stroke. N Engl J Med 299: 53-59, 1978
7. Toole JF, Janeway R, Choi K, etal: Transient ischemic attacks due to atherosclerosis: a prospective study of 160 patients. Arch Neural 32: 5-12, 1975
8. Heyman A, Leviton A, Millikan CH, et al: Transient focal cerebral ischemia: epidemiologic and clinical aspects. Stroke 5: 277-287, 1974
9. Matsumoto N, Whisnant JP, Kurland LT, Okazaki H: Natural history of stroke in Rochester, Minnesota, 1955 through 1969. Stroke 4: 20-29, 1973
10. Fields WS, Maslenikov V, Meyer JS, Hass WK, Remington RD, MacDonald M: Joint study of extracranial arterial occlusion. V. Progress report of prognosis following surgery or nonsurgical treat-ment of transient cerebral ischemic attacks and cervical carotid artery lesions. JAMA 211: 1993-2003, 1970
11. Delong WB: Microsurgical revascularization for cerebrovascular insufficiency. Stroke 9: 15-20, 1976
12. Taylor DW, Haynes RB, Sackett DL: Stopping rules for long-term clinical trials, (abstract) Controlled Clinical Trials 1: 170, 1980 13. Sundt TM Jr, Siekert RG, Piepgras DS, Sharbrough FW, Houser
OW: Bypass surgery for vascular disease of the carotid system. Mayo Clin Proc 51: 677-692, 1976
14. Hachinski VC: Extracranial/intracranial bypass surgery: technique or therapy? In Cerebral Vascular Disease 4, Proceedings of the 1 lth International Salzburg Conference, edited by JS Meyer, H. Lechner, M. Reivich, EO Ott. Amsterdam: Exerpta Medica, 1983
Investigators: Central Office (University of Western Ontario (UWO), London, Ontario, Canada): Principal Investigator, HJM Baraett; Princi-pal Neurosurgical Investigator, SJ Peerless; PrinciPrinci-pal Neuro-Radiologi-cal Investigator, A.J. Fox; Senior Staff: Barbara Valberg, Jean Pea-cock.
Methods Center (McMaster University (MU), Hamilton, Ontario, Can-ada): Principal Epidemiologic Investigator, DL Sackett; Chief Epidemi-ologist, RB Haynes; Chief Statistician, DW Taylor; Senior Staff: Chris-tine Collis, Jayanti Mukherjee.
Steering Committee: UWO: HJM Barnett (chairman), SJ Peerless, AJ Fox, B. Valberg, VC Hachinski. MU: RB Haynes, DW Taylor, DL Sackett.
Monitoring Committee: M Goldstein (Chairman 1977-82; NINCDS), MD Walker (Chairman 1982- , NINCDS), JB Benedict (NINCDS), W Weiss (NINCDS), JR Marler (NINCDS), JP Whisnant (Mayo Clinic),
HG Schwartz (Washington U), A Heyman (Duke U Medical Center), WH Feindel (Montreal Neurological Institute).
Participating Centers (in order of number of eligible patients entered): U. of Western Ontario, London, Ontario; H.J.M. Barnett, C.W.
McCormick, V.C. Hachinski, S.J. Peerless, G.G. Ferguson Neuroradiologists: J. Allcock, A.J. Fox
Additional Contributors: K. Meguro, R. Cote, D. Moulin, P.C. Gates, S. Lauzier
University of Toronto, Toronto, Ontario; R. Wilson, G. Sawa, H. Schutz
University of Tennessee, Memphis, Tennessee; A. Heck, J. Robertson National Institute of Neurosurgery, Budapest, Hungary; L. Ronai, E.
Pasztor*, J. Vajda, M. Horvath, I. Nyary
University of Essen, Essen, Federal Republic of Germany; A. Buch, H.M. Mehdorn
Universta di Firenze, Florence, Italy; L. Amaducci, D. Inzitari, S. Brianit, R. Gagliardi
Neurochir. und Neurolog., University Klinik, Giessen, Federal Repub-lic of Germany; O. Busse, C. Hornig, E. Grote, R. Schonmayr Kyoto U. Medical School, Kyoto, Japan; M. Kameyama, I. Akiguchi,
H. Shio, H. Handa*, Y. Yonekawa
Hopital Pellegrin, Bordeaux, France; J.M. Orgogozo, J.J. Pere, J.P. Castel
New York University, New York, New York; W.K. Hass, E.S. Flamm University of Pecs, Pecs, Hungary; M. Bodosi, G. Gacs, F T . Merei WesteindeZiekenhuis, Den Haag, Holland; J.Th.T. Tans, C.A.F.
U. of Mississippi, Jackson, Mississippi; A.F. Haerer, R.R. Smith Institute of Brain & Blood Vessels, Gunma, Japan; G. Araki, K.
Na-gata, C. Yunoki, M. Mizukami
Tokyo Medical & Dental University, Tokyo, Japan; H. Tsukagoshi, U. Ito, Y. Inaba*, T. Fujimoto, K. Komatsu
Neurochir. Univ., Klinik, Universitatsspital, Zurich, Switzerland; H. Zumstein, H. Keller, H.G. Imhof, B. Zumstein
Upstate Medical Center, Syracuse, New York; A. Culebras, C.J. Hodge Case Western Reserve U., Cleveland, Ohio; D.L. Jackson, K. Chandar,
Univ. of Tokushima, Tokushima, Japan; S. Yasuoka, K. Matsumoto*, S. Ueda
Neurological Institute of Savannah, Savannah, Georgia; O.E. Ham, E.F. Downing, F.P. Wirth
Universita di Milano, Milan, Italy; P. Perrone, G. Cabrini
State University of New York at Buffalo, Buffalo, N.Y.; E.J. Manning, D. Ehrenreich, L.N. Hopkins
U. of Minnesota, Minneapolis, Minnesota; M.C. Lee, D. Erickson Royal Victoria Infirm., Newcastle-Upon-Tyne, England; D.A. Shaw,
D. Bates, G. Venables, R. Sengupta
Veterans General Hosp., Taipei, Taiwan; Fu-Li Chu, Han-Hua Hu, W. Wen-Jang Wong, A.L. Shen
Johns Hopkins University, Baltimore, Maryland; T.J. Preziosi, M.H. Epstein
University of Oregon, Portland, Oregon; B. Coull, F. Yatsu, F. Waller, C. Tanabe
Sophia Hospital, Zwolle, The Netherlands; W.G.M. Teunissen, P.W. Gelderman
Universita degli Studi, Naples, Italy; R. Cotrufo, P. Conforti, F. Toma-sello, V. Albanese
Neurosurgical U. Hosp.. Belgrade, Yugoslavia; M. Panic, B. Milosavl-jevic
Fujita-Gakuen University, School of Medicine, Nagoya, Japan; M. Nomura, T. Kanno, H. Sano
Harvard Medical School, Boston, Massachusetts; P. Kistler, R. Crowell
University of Pavia, Pavia, Italy; G. Brambilla, D. Locatelli, R.R. yBaena, P. Paoletti
U. of North Carolina, Chapel Hill, N. Carolina; J.N. Hayward, S.C. Boone, J.D. Mann
Emory U. School of Med., Atlanta, Georgia; H. Karp, R. Schnapper, A. Fleischer
University of Iowa, Iowa City, Iowa; H. Adams, C. Gross University of Texas, Dallas, Texas; E. Ross, D. Samson
*Sponsoring Surgeon t Deceased
406 STROKE VOL 16, No 3, MAY-JUNE 1985
Legnano General Hosp., Legnano, Italy; G. Tonnarelli, I. Piazza Tufts New Eng. Med. Cent., Boston, Massachusetts; M. Pessin, R.M.
University of Missouri, Columbia, Missouri; J. Byer, C. Watts, M. Ditmore
University of Mainz, Mainz, West Germany; G. Kramer, G. Meinig U. of South Alabama, Mobile, Alabama; J.P. Mohr, C.S. Kase, H.C.
National Cardiovasc. Cent., Osaka, Japan; T. Yamaguchi, T. Sawada, H. Kikuchi
Mayfield Neurolog. Inst., Cincinnati, Ohio; R. Reed, J. Tew Dalhousie University, Halifax, Nova Scotia; T.J. Murray, C.W.
McCormick, W.J. Howes
Cleveland Clinic, Cleveland, Ohio; A. Furlan, J. Little, D. Dohn Inst. of Brain Diseases, Tohoku University, Sendai, Japan; H. Saito, J.
Suzuki*, T. Yoshimoto, N. Kodama
U. of California, San Francisco, California; M.S. Edwards Hokkaido U. Hospital, Sapporo, Japan; K. Tashiro, M. Tsuru*, Y.
North Manchester General Hospital, Manchester, England; D. Shep-herd, G.M. Yuill, C. Bannister
Henry Ford Hospital, Detroit, Michigan; R. Teasdall, J. Ausman Research Institute of Brain & Blood Vessels, Akita, Japan; T.
Kut-suzawa, N. Nakajima, T. Kobayashi, N. Yasui, Z. Ito
U. of Cincinnati, Cincinnati, Ohio; C. Olinger, R. Singh, G. Khodadad Duke U. Medical Center, Durham, North Carolina; W.C. Olanow,
Ciudad Sanitara V. del Rocio, Seville, Spain; R. Alberca-Serrano, F. Morales-Ramos
Barrow's Neurolog. Inst., Phoenix, Arizona; A. Yudell, R. Thompson, P. Carter
Mississippi Baptist Medical Center, Jackson, Mississippi; W. Bowlus, L. Mahalak, D. Stringer
Neurol. Inst., Tokyo Women's Med. Coll., Tokyo, Japan; S. Mar-uyama, K. Kitamura, M. Kagawa
U. of Pittsburgh, Pittsburgh, Pennsylvania; O. Reinmuth, R. Heros U. of Nagasaki Med. Sen., Nagasaki, Japan; H. Matsumura, M.
Taka-mori, K. Mori*, H. Ono
Osaka U. Med. School, Osaka, Japan; M. Imaizumi, S. Yoneda, H. Mogami, T. Hayakawa
Nassau County Med. Cent., East Meadow, New York; R. Carruthers, R. Decker
Naval Regional Medical Center, Oakland, California; A. Chalmers, T.H. Rockel, R. Hodosh
U. of Arizona, Tucson, Arizona; J. Laguna, P. Weinstein
Tokyo U. Med. School, Tokyo, Japan; T. Takasu, T. Eguchi, H. Sugiyama, N. Basugi, T. Asano
Queens U., Kingston, Ontario; H.B. Dinsdale, P. Murray Albert Einstein, College of Medicine, Bronx, New York; L.J. Thai Hopital Neurologique, Lyon, France; D. Deruty
Daniel Freeman, Memorial Hospital, Inglewood, California; B. Dobkin Scarborough Gen. Hosp., Scarborough, Ontario; M R . Goldman Long Island-Jewish Hillside Med. Cent., New Hyde Park, N.Y.; M.
Writing Committee (for this article): R.B. Haynes, D.W. Taylor, J. Mukherjee and D.L. Sackett (MU); R. Cote, K. Meguro, P.C. Gates, C.W. McCormick, A.J. Fox, S.J. Peerless, and H.J.M. Bamett. (UWO).
The Principal Investigators gratefully acknowledge the invaluable contribution of all the neurologists, neurosurgeons, neuroradiologists, research fellows, residents, family physicians, nurse practitioners, sec-retaries and data management staff who have served the Bypass Study with loyalty and dedication. Their constant help and support enabled us to carry this large, demanding and important project to a successful conclusion.
Persantine Aspirin Trial In Cerebral Ischemia
Part II: Endpoint Results
T H E A M E R I C A N - C A N A D I A N CO-OPERATIVE S T U D Y G R O U P
SUMMARY The Persantine9 Aspirin Trial focused on the question of whether the administration of the combination of aspirin and dipyridamole (Persantine) would result in a lower incidence of cerebral or retinal infarction or death than the administration of aspirin alone for persons with a history of recent carotid territory transient ischemic attacks (TIAs).
Fifteen centers in the United States and Canada participated and 890 individuals were admitted and randomly allocated to either aspirin (325 mg) plus placebo or aspirin (325 mg) plus Persantine (75 mg) four times daily. Ninety eight percent of the subjects were followed for at least one year; many were followed for four to five years.
The results of life table analysis indicate that the overall endpoint rates for the "aspirin only" and "aspirin plus Persantine" groups are identical. Thus, for TIA patients taking aspirin, the addition of Persantine contributes nothing. There was a clustering of stroke endpoints during the first month after randomization. Deaths from all causes were essentially equally divided between the two treatment groups.
Stroke Vol 16, No 3, 1985 THIS COOPERATIVE CLINICAL TRIAL was
de-signed to determine whether persons with a history of transient ischemic attacks (TIAs) have a lower risk of stroke, retinal infarction, or death if they are taking a Address correspondence to: William S. Fields, M.D., University of Texas, M.D. Anderson Hospital and Tumor Institute, 6723 Bertner Avenue, Houston, Texas 77030.
Received August 21, 1984; revision # 1 accepted December 19, 1984.
combination of aspirin and dipyridamole (Persantine) rather than aspirin alone. In the initial publication,1 the rationale for the trial and details of the design and methods were described. This report presents the re-sults of the five-year endeavor.
More than 1400 individuals were screened as poten-tial candidates for the trial. Some declined to partici-pate; others were not considered eligible for various reasons: