Circulation. 2006;113: ; originally published online January 23, 2006; doi: /CIRCULATIONAHA Data Supplement (unedited) at:

Transcription

1 Predictors of Ischemic Stroke in the Territory of a Symptomatic Intracranial Arterial Stenosis Scott E. Kasner, Marc I. Chimowitz, Michael J. Lynn, Harriet Howlett-Smith, Barney J. Stern, Vicki S. Hertzberg, Michael R. Frankel, Steven R. Levine, Seemant Chaturvedi, Curtis G. Benesch, Cathy A. Sila, Tudor G. Jovin, Jose G. Romano and Harry J. Cloft Circulation. 2006;113: ; originally published online January 23, 2006; doi: /CIRCULATIONAHA Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX Copyright 2006 American Heart Association, Inc. All rights reserved. Print ISSN: Online ISSN: The online version of this article, along with updated information and services, is located on the World Wide Web at: Data Supplement (unedited) at: Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Circulation can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: Subscriptions: Information about subscribing to Circulation is online at: Downloaded from by guest on February 27, 2014

2 Stroke Predictors of Ischemic Stroke in the Territory of a Symptomatic Intracranial Arterial Stenosis Scott E. Kasner, MD; Marc I. Chimowitz, MBChB; Michael J. Lynn, MS; Harriet Howlett-Smith, RN; Barney J. Stern, MD; Vicki S. Hertzberg, PhD; Michael R. Frankel, MD; Steven R. Levine, MD; Seemant Chaturvedi, MD; Curtis G. Benesch, MD; Cathy A. Sila, MD; Tudor G. Jovin, MD; Jose G. Romano, MD; Harry J. Cloft, MD, PhD; for the Warfarin Aspirin Symptomatic Intracranial Disease (WASID) Trial Investigators* Background Antithrombotic therapy for intracranial arterial stenosis was recently evaluated in the Warfarin versus Aspirin for Symptomatic Intracranial Disease (WASID) trial. A prespecified aim of WASID was to identify patients at highest risk for stroke in the territory of the stenotic artery who would be the target group for a subsequent trial comparing intracranial stenting with medical therapy. Methods and Results WASID was a randomized, double-blinded, multicenter trial involving 569 patients with transient ischemic attack or ischemic stroke due to 50% to 99% stenosis of a major intracranial artery. Median time from qualifying event to randomization was 17 days, and mean follow-up was 1.8 years. Multivariable Cox proportional hazards models were used to identify factors associated with subsequent ischemic stroke in the territory of the stenotic artery. Subsequent ischemic stroke occurred in 106 patients (19.0%); 77 (73%) of these strokes were in the territory of the stenotic artery. Risk of stroke in the territory of the stenotic artery was highest with severe stenosis 70% (hazard ratio 2.03; 95% confidence interval 1.29 to 3.22; P ) and in patients enrolled early ( 17 days) after the qualifying event (hazard ratio 1.69; 95% confidence interval 1.06 to 2.72; P 0.028). Women were also at increased risk, although this was of borderline significance (hazard ratio 1.59; 95% confidence interval 1.00 to 2.55; P 0.051). Location of stenosis, type of qualifying event, and prior use of antithrombotic medications were not associated with increased risk. Conclusions Among patients with symptomatic intracranial stenosis, the risk of subsequent stroke in the territory of the stenotic artery is greatest with stenosis 70%, after recent symptoms, and in women. (Circulation. 2006;113: ) Key Words: stroke atherosclerosis cerebrovascular circulation Atherosclerotic stenosis of the major intracranial arteries is an important cause of ischemic stroke, especially in blacks, Asians, and Hispanics. 1,2 In the United States, intracranial stenosis causes 8% to 10% of ischemic strokes, and the risk of recurrent stroke in these patients may be as high as 15% per year. 3 8 The Warfarin versus Aspirin for Symptomatic Intracranial Disease (WASID) study was a randomized clinical trial that compared warfarin and aspirin for preventing stroke and vascular death in patients with stenosis of a major intracranial artery. 9 The study determined that warfarin Clinical Perspective p 563 was associated with significantly higher rates of adverse events and provided no benefit over aspirin for preventing stroke and vascular death. Another therapeutic strategy is intracranial angioplasty or stenting, but there are no studies on the comparative safety and efficacy of angioplasty or stenting versus medical therapy. A few retrospective studies have suggested that there are subgroups of patients with intracranial arterial stenosis at Received July 25, 2005; revision received November 21, 2005; accepted November 23, From the Department of Neurology, University of Pennsylvania Medical Center (S.E.K.), Philadelphia; Department of Neurology (M.I.C., H.H.-S., B.J.S., M.R.F.) and Department of Biostatistics (M.J.L., V.S.H.), Rollins School of Public Health, Emory University School of Medicine, Atlanta, Ga; Department of Neurology, University of Maryland (B.J.S.), Baltimore; Department of Neurology, Mount Sinai School of Medicine (S.R.L.), New York, NY; Department of Neurology, Wayne State University (S.C.), Detroit, Mich; Department of Neurology, University of Rochester School of Medicine (C.G.B.), Rochester, NY; Department of Neurology, Cleveland Clinic Foundation (C.A.S.), Cleveland, Ohio; Department of Neurology, University of Pittsburgh School of Medicine (T.G.J.), Pittsburgh, Pa; Department of Neurology, University of Miami Medical School (J.G.R.), Miami, Fla; and Department of Radiology, Mayo Clinic (H.J.C.), Rochester, Minn. *For a list of the WASID Investigators, see the Appendix in the online-only Data Supplement, available at Correspondence to Scott E. Kasner, MD, Department of Neurology, University of Pennsylvania, 3 W Gates Bldg, 3400 Spruce St, Philadelphia, PA kasner@mail.med.upenn.edu 2006 American Heart Association, Inc. Circulation is available at DOI: /CIRCULATIONAHA Downloaded from by guest on February 27, 2014

3 556 Circulation January 31, 2006 TABLE 1. Baseline Characteristics of Patients With and Without Subsequent Stroke in the Territory of the Symptomatic Stenotic Artery Characteristic No. of Patients With Data Stroke in the Territory (n 77) No Stroke in the Territory (n 492) P Age, y Gender 0.06 Male (11) 310 (89) Female (17) 182 (83) Race 0.49 White (13) 289 (87) Nonwhite (15) 203 (85) Height, in Weight, lb Body mass index, kg/m Drinks alcohol 0.78 No (14) 292 (86) Yes (13) 200 (87) Ever smoked 0.67 No (14) 173 (86) Yes (13) 319 (87) Activity level 0.80 Sedentary (13) 128 (87) Not sedentary (14) 364 (86) History of ischemic stroke 0.46 No (14) 360 (86) Yes (12) 122 (88) History of TIA 0.44 No (14) 354 (86) Yes (12) 121 (88) History of coronary artery disease 0.91 No (14) 351 (86) Yes (14) 130 (86) History of hypertension 0.94 No (13) 78 (87) Yes (14) 412 (86) History of diabetes No (11) 312 (89) Yes (17) 179 (83) History of lipid disease 0.70 No (14) 138 (86) Yes (13) 340 (87) NIH Stroke Scale score (10) 328 (90) (20) 164 (80) Qualifying event 0.21 Stroke (15) 295 (85) TIA (11) 197 (89) Symptomatic vessel* 0.85 Anterior (14) 265 (86) Posterior (13) 212 (87)

4 Kasner et al Stroke in the Territory of Intracranial Stenosis 557 TABLE 1. Continued Characteristic No. of Patients With Data Stroke in the Territory (n 77) No Stroke in the Territory (n 492) Percent stenosis % (10) 320 (90) 70% (19) 166 (81) Length of stenosis (13) 237 (87) (14) 236 (86) On antithrombotic medication at 0.71 qualifying event No (14) 231 (86) Yes (13) 260 (87) Time from qualifying event to enrollment 17 days (17) 240 (83) 17 days (10) 252 (90) Treatment assignment 0.31 Aspirin (15) 238 (85) Warfarin (12) 254 (88) Values are n (%) or mean SD. Numerical thresholds were defined as the median of the study population. *Symptomatic vessels were defined as either anterior (middle cerebral artery and intracranial internal carotid artery) or posterior (intracranial vertebral artery and basilar artery). Length of stenosis is defined as the ratio of the length of the stenotic segment divided by the diameter of the normal referent artery. P particularly high risk of stroke in the territory of the stenotic artery who may be the target population for intracranial angioplasty or stenting. These include patients with severe stenosis (70% to 99%), 8 those with vertebrobasilar disease, 8,10,11 and those who had an ischemic event while undergoing antithrombotic therapy. 12 Furthermore, studies of extracranial carotid stenosis have suggested that stroke rather than transient ischemic attack (TIA) 13 and recent symptoms 14 may portend greater risk. These 5 potential high-risk features require validation in a prospective study to have clinical utility for patients with intracranial stenosis. The WASID study provided a unique opportunity to identify predictors of stroke in the territory of a stenotic intracranial artery. As such, a prespecified aim of WASID was to identify patients at sufficiently high risk of stroke in the territory of the stenotic artery who would be the target group for a subsequent trial comparing intracranial stenting with medical therapy. Methods Study Design and Patient Eligibility Details of the WASID study design and results of the comparison of warfarin and aspirin have been published previously. 7,9 Briefly, WASID was an investigator-initiated, randomized, double-blind, multicenter clinical trial conducted at 59 sites in North America that was funded by the National Institute of Neurological Disorders and Stroke (NINDS). The study protocol was approved by the institutional review board at each site, and all patients gave written informed consent to participate. Patients were enrolled between February 1999 and July Inclusion criteria included TIA or nondisabling stroke within 90 days of randomization that was attributable to angiographically verified 50% to 99% stenosis of a major intracranial artery (internal carotid, middle cerebral, vertebral, or basilar artery), modified Rankin score 3, and age 40 years. Exclusion criteria included extracranial internal carotid stenosis (50% to 99%) tandem to an intracranial carotid or middle cerebral artery stenosis; nonatherosclerotic stenosis; a cardiac source of embolism (eg, atrial fibrillation); contraindication to aspirin or warfarin; and a comorbid condition that limited the outlook for survival to 5 years. Patients were randomly assigned treatment with warfarin (target international normalized ratio [INR] between 2.0 and 3.0) or aspirin (1300 mg daily). Study subjects and investigators were blinded to treatment allocation. Angiographic Evaluation Patients were enrolled in WASID on the basis of local interpretation of angiograms that demonstrated 50% to 99% stenosis of a major intracranial artery. One copy of the angiogram frame that best depicted the symptomatic stenosis was sent for central reading. The central neuroradiologist (HJC), blinded to the site reading, measured percent diameter stenosis according to the WASID measurement technique. 7 On the same angiographic image, the length of stenosis was measured and then normalized by dividing by the diameter of the normal referent arterial segment. The measurements used in this analysis were those of the central neuroradiologist. On the basis of the intent-to-treat principle, patients whose central readings indicated 50% stenosis or complete 100% occlusion were included in this analysis. Follow-Up and Assessment of End Points Patients were contacted monthly to determine whether any events had occurred. Every 4 months, patients were examined by a neurologist blinded to treatment allocation. If a stroke was suspected, patients underwent brain computerized tomography (CT) or magnetic resonance imaging (MRI). Patients were to be followed up to a primary end point, death, or a common termination date. The primary end point for the comparison of warfarin and aspirin in WASID was ischemic stroke (in any vascular territory), brain hemorrhage, or nonstroke vascular death.

5 558 Circulation January 31, 2006 TABLE 2. Univariate Associations of Baseline Characteristics With Stroke in the Territory of the Symptomatic Stenotic Artery Characteristic HR (95% CI) P Age ( 64 vs 64 y) 0.72 ( ) 0.16 Gender (female vs male) 1.57 ( ) Race (other vs white) 1.23 ( ) 0.37 Height ( 67 vs 67 in) 0.87 ( ) 0.54 Weight ( 180 vs 180 lbs) 1.05 ( ) 0.84 Body mass index ( 25 vs 25 kg/m 2 ) 1.82 ( ) Drinks alcohol (yes vs no) 0.91 ( ) 0.67 Ever smoked (yes vs no) 0.88 ( ) 0.59 Activity level (sedentary vs other) 0.93 ( ) 0.79 History of ischemic stroke (yes vs no) 0.87 ( ) 0.62 History of TIA (yes vs no) 0.80 ( ) 0.43 History of coronary artery disease (yes vs no) 1.05 ( ) 0.85 History of hypertension (yes vs no) 1.08 ( ) 0.81 History of diabetes (yes vs no) 1.58 ( ) History of lipid disease (yes vs no) 0.90 ( ) 0.66 NIH Stroke Scale score ( 1 vs 1) 2.13 ( ) Qualifying event (stroke vs TIA) 1.42 ( ) 0.15 Symptomatic vessel (posterior vs anterior)* 0.97 ( ) 0.89 Percent stenosis ( 70% vs 70%) 2.11 ( ) On antithrombotic medication at qualifying 0.89 ( ) 0.61 event (yes vs no) Time from qualifying event to enrollment 1.80 ( ) ( 17 vs 17 days) Treatment assignment (aspirin vs warfarin) 1.26 ( ) 0.31 Numerical thresholds were defined as the median of the study population. *Symptomatic vessels were defined as either anterior (middle cerebral artery and intracranial internal carotid artery) or posterior (intracranial vertebral artery and basilar artery). For the present analysis, the end point of interest was ischemic stroke in the territory of the symptomatic intracranial stenosis. Ischemic stroke was defined as a new focal neurological deficit of sudden onset that lasted 24 hours and was not caused by hemorrhage on brain imaging. Ischemic stroke was considered to be definitely in the same territory of the symptomatic artery when the neurological signs correlated with a new infarct on CT or MRI in an area of the brain supplied by that artery. Ischemic stroke was considered probably in the territory of the symptomatic artery when the neurological signs were localized to an area of the brain supplied by that artery but there was no new infarct on brain imaging. Ischemic stroke was considered in an indeterminate territory when the neurological signs could localize to 2 or more distinct vascular territories and there was no new infarct on brain imaging. Ischemic stroke could likewise be classified as probably not or definitely not in the territory of the symptomatic artery. For the purpose of the present analysis, ischemic strokes that were definitely or probably in the territory of the stenotic artery were considered in the territory. The local investigator classified the location of the end-point ischemic stroke. In addition, the locations of all ischemic strokes were independently determined by a central investigator (MIC) at the end of the study. In cases in which there was disagreement, a second central investigator (BJS) independently determined the location, and the classification made by 2 of the 3 investigators was used. Statistical Methods The original power analysis to address the main hypothesis pertinent to this study (in the best medical therapy group, patients with 70% to 99% stenosis would have a significantly higher rate of ischemic Product-limit estimate of the cumulative probability of an ischemic stroke in the territory of the stenotic artery vs years after randomization, according to percent stenosis ( 70% stenosis shown as solid line, 70% stenosis as dashed line); log-rank test P stroke in the territory of the stenotic artery than patients with 50% to 69% stenosis) assumed a sample size of 403 patients followed up for an average of 2.5 years. Assuming a 60:40 ratio in patients with moderate versus severe stenosis, the power was 94% to detect a difference in rates similar to that observed in the retrospective WASID pilot study, 8 18% with severe stenosis versus 6% with moderate stenosis over 3 years. In addition, we calculated that there was 80% power to detect hazard ratios (HRs) between 2.0 and 2.3 for other risk factors (eg, anterior versus posterior circulation stenosis). WASID was halted after 569 patients had been enrolled and followed up for an average of 1.8 years. Because no difference was discerned between the 2 treatment arms, all patients were included in the present analysis. We determined the power of the study with 569 patients to detect HRs of 2 using the method of Lachin and Foulkes 15 for the log-rank test. Among all 569 patients, the probability of ischemic stroke in the territory of the stenotic artery at 2 years was To estimate the detectable differences between groups of patients, we determined the proportions with an event by 2 years in each of 2 groups that would yield an overall proportion of 0.14 when 1 proportion was twice the other proportion and when the sample size varied with 33%, 50%, or 67% of the 569 patients in the group with the smaller proportion. The smaller proportion at 2 years was translated to a hazard rate, and the power to detect a hazard rate twice as large was determined for the 3 sets of sample sizes. With a type I error of 0.05, the power was 0.76, 0.86, and 0.88 when the sample size in the group with the smaller hazard rate was 33%, 50%, and 67% of the 569 patients, respectively. Thus, the study had adequate power to detect HRs of 2 when the sample size of the smaller of 2 groups was at least 190. Univariate analyses to assess the effect of each baseline factor were done with Cox proportional hazards regression. A series of nested multivariable Cox proportional hazards models were considered. The primary model was based on the 5 hypothesized contributors to the risk of stroke in the territory of the symptomatic artery: type of qualifying event (stroke versus TIA), location of vessel (anterior versus posterior circulation), percent stenosis, treatment with antithrombotic medications at the time of the qualifying event, and time from qualifying event to enrollment. A second model incorporated these 5 variables and was also adjusted for age, gender, and race. Additional variables were considered for inclusion in a third exploratory multivariable model if they were associated with stroke in the territory in univariate analysis at the P 0.05 level.

6 Kasner et al Stroke in the Territory of Intracranial Stenosis 559 TABLE 3. Risk of Stroke in the Territory of the Stenotic Intracranial Artery According to Percent Stenosis Percent Stenosis No. of Patients No. of Patients With Stroke (%) Probability of Stroke at 1 Year (95% CI) Probability of Stroke at 2 Years (95% CI) (4.2) 0.05 ( ) 0.05 ( ) (4.4) 0.02 ( ) 0.06 ( ) (11.2) 0.07 ( ) 0.14 ( ) (11.3) 0.08 ( ) 0.10 ( ) (18.2) 0.17 ( ) 0.18 ( ) (30.3) 0.31 ( ) 0.31 ( ) (9.1) 0.05 ( ) 0.12 ( ) (33.3) 0.33 ( ) 0.33 ( ) Total (13.4) 0.11 ( ) 0.14 ( ) Cochran-Armitage trend test P Eight patients were excluded from this analysis because their films were not available for review by the central neuroradiologist. The cumulative probability of a stroke in the territory of the stenotic intracranial artery versus time was estimated by the product-limit method. The relationship between degree of stenosis and the risk of stroke in the territory of the stenotic artery was tested with the Cochran-Armitage test. 16 Patients lost to follow-up were censored at the last contact date. Baseline characteristics were compared between patients with and without stroke in the territory with a 2 test (for proportions) or independent groups t test (for means). All reported probability values are 2-sided, without adjustment for multiple testing; probability values 0.05 were considered statistically significant. Results In WASID, 569 patients were enrolled a median of 17 days after the qualifying TIA or ischemic stroke. The mean duration of follow-up was 1.8 years. Thirteen patients (2.3%) were lost to follow-up (6) or withdrew consent (7) at an average of 6 months after enrollment. The primary WASID end point of ischemic stroke, brain hemorrhage, or nonstroke vascular death occurred in 125 patients (22%). Ischemic stroke accounted for the majority of events in WASID and occurred in 106 patients (19.0%); 77 (73%) of these strokes were in the territory of the stenotic artery (51 definite, 26 probable). Sixty (78%) of these 77 strokes occurred within the first year. Table 1 summarizes the baseline characteristics for patients with and without a stroke in the territory of the symptomatic stenotic artery. Table 2 shows event rates and univariate relationships between baseline characteristics and the rates of stroke in the territory of the stenotic artery. Univariate analysis (Table 2) showed that severity of stenosis ( 70% versus 70%; Figure), time from qualifying event to enrollment ( 17 days versus 17 days), female gender, National Institutes of Health (NIH) stroke scale ( 1 versus 1), and history of diabetes mellitus were significantly associated (P 0.05) with stroke in the territory of the stenotic artery, whereas body mass index was of borderline significance (P 0.068). Age, race, location of stenosis, length of stenosis, other vascular risk factors, comorbidities, treatment with antithrombotic agents at the time of the qualifying event, and treatment assignment were not significantly associated with an increased risk of stroke in the territory of the stenotic artery. In the primary multivariable model, severe stenosis 70% resulted in a significantly higher subsequent risk of stroke in the territory of a symptomatic intracranial stenotic artery than stenosis 70% (HR 2.03; 95% confidence interval [CI] 1.29 to 3.22; P ). Risk increased linearly (P for trend ) with greater percent stenosis but may have declined when stenosis was 90% to 99% (Table 3). Furthermore, risk was significantly greater for patients enrolled 17 days after the qualifying event (HR 1.69; 95% CI 1.06 to 2.72; P 0.028). Type of qualifying event (stroke or TIA) was not statistically associated with risk of stroke in the territory of the symptomatic stenosis (HR 1.50; 95% CI 0.90 to 2.49; P 0.12). However, when we stratified the qualifying event according to severity of stenosis, we observed a low rate of stroke in patients with initial TIA and stenosis 70% and a very high rate of stroke in patients with initial stroke and stenosis 70% (Table 4), but there was no statistical evidence of interaction (P 0.11). There was also no interaction between type of qualifying event and time from qualifying event to enrollment (P 0.46). Among the 342 patients whose qualifying event was stroke, lacunar versus nonlacunar type had no impact (P 0.20). Location of stenosis (posterior versus anterior circulation) was not associated with risk (HR 1.05; 95% CI 0.66 to 1.68; P 0.97) nor was treatment with antithrombotic medications at the time of the qualifying event (HR 1.06; 95% CI 0.66 to 1.69; P 0.83). In the second multivariable model adjusted for age, sex, and race (Table 5), the hazard estimates of the 5 putative risk TABLE 4. Probability of Stroke in the Territory of the Symptomatic Stenotic Artery Related to Severity of Stenosis and Type of Qualifying Event Qualifying Event Stenosis 50% to 69% Stenosis 70% to 99% TIA 1 year 0.03 ( ) 0.14 ( ) 2 years 0.08 ( ) 0.14 ( ) Stroke 1 year 0.08 ( ) 0.23 ( ) 2 years 0.11 ( ) 0.25 ( ) Data are presented as mean probability (95% CI).

7 560 Circulation January 31, 2006 TABLE 5. Multivariable Model of Characteristics Hypothesized to Be Associated With Subsequent Risk of Stroke in the Territory of the Symptomatic Stenotic Artery, Adjusted for Age, Gender, and Race Characteristic HR (95% CI) P Qualifying event (stroke vs TIA) 1.42 ( ) 0.18 Symptomatic vessel (posterior vs anterior)* 1.12 ( ) 0.65 Percent stenosis ( 70% vs 70%) 2.08 ( ) On antithrombotic medication at qualifying event (yes vs no) 1.05 ( ) 0.86 Time from qualifying event to enrollment ( 17 vs 17 days) 1.67 ( ) Age ( 64 vs 64 y) 0.76 ( ) 0.24 Gender (female vs male) 1.59 ( ) Race (other vs white) 1.17 ( ) 0.53 Numerical thresholds were defined as the median of the study population. *Symptomatic vessels were defined as either anterior (middle cerebral artery and intracranial internal carotid artery) or posterior (intracranial vertebral artery and basilar artery). factors were not significantly altered, which suggests no major confounding by demographic characteristics. However, female gender was marginally associated with greater risk of stroke in the territory of the symptomatic artery (HR 1.59; 95% CI 1.00 to 2.55; P 0.051). Additional variables identified as possibly associated with stroke in the territory of an intracranial stenosis in univariate analysis were diabetes mellitus and the NIH Stroke Scale. When incorporated into the exploratory third multivariable model (Table 6), diabetes was no longer associated with risk (HR 1.18; 95% CI 0.74 to 1.90; P 0.49), but NIH Stroke Scale scores greater than the median score of 1 were associated with subsequent risk (HR 2.16; 95% CI 1.26 to 3.70; P ). The NIH Stroke Scale strongly correlated with the type of qualifying event (P 0.001), and its inclusion in the model completely negated the marginal trend associated with the type of qualifying event. The HRs and probability values for the other parameters of the second model were essentially unchanged by the inclusion of these 2 additional variables. To further evaluate the relationship between the type of qualifying event and NIH Stroke Scale, we compared subsequent risk between patients with single versus multiple ischemic events before randomization. Multiple prior events were not associated with a greater risk of stroke in the territory of the stenotic artery than was a single event (HR 1.31; 95% CI to 2.06; P 0.25). All the above analyses were repeated with the exclusion of patients with centrally measured 50% stenosis (n 70) or complete occlusion (n 3), and no substantive changes in the primary results were observed. Discussion Patients with symptomatic intracranial stenosis are at high risk of subsequent stroke, predominantly in the territory of the stenotic artery. Despite the use of either warfarin or aspirin and vascular risk factor modification, the overall rates of stroke in the territory of the stenotic artery were 11% at 1 year and 14% at 2 years. Given the high risk of stroke in the territory of a symptomatic intracranial stenosis, alternative TABLE 6. Multivariable Model of Characteristics Hypothesized to Be Associated With Subsequent Risk of Stroke in the Territory of the Symptomatic Stenotic Artery, Adjusted for Age, Gender, and Race, Including Variables Statistically Significant in Univariate Analysis Characteristic HR (95% CI) P Qualifying event (stroke vs TIA) 0.93 ( ) 0.82 Symptomatic vessel (posterior vs anterior)* 1.17 ( ) 0.53 Percent stenosis ( 70% vs 70%) 2.00 ( ) On antithrombotic medication at qualifying event (yes vs no) 0.94 ( ) 0.78 Time from qualifying event to enrollment ( 17 vs 17 days) 1.72 ( ) Age ( 64 vs 64 y) 0.77 ( ) 0.27 Gender (female vs male) 1.59 ( ) Race (other vs white) 1.06 ( ) 0.81 History of diabetes (yes vs no) 1.18 ( ) 0.49 NIH Stroke Scale score ( 1 vs 1) 2.16 ( ) Numerical thresholds were defined as the median of the study population. *Symptomatic vessels were defined as either anterior (middle cerebral artery and intracranial internal carotid artery) or posterior (intracranial vertebral artery and basilar artery).

8 Kasner et al Stroke in the Territory of Intracranial Stenosis 561 therapies are needed. In this regard, intracranial angioplasty and stenting have emerged as promising treatments for intracranial stenosis that are increasingly being used in clinical practice in the United States and other countries However, these therapies have not been evaluated in a controlled clinical trial. One recent uncontrolled multicenter phase I study, Stenting of Symptomatic atherosclerotic Lesions in the Vertebral or Intracranial Arteries (SSYLVIA), with similar eligibility criteria to WASID, showed that the frequency of stroke in the territory of a stented intracranial artery was 7.2% at 30 days and 10.9% at 1 year. 6 Two (4.7%) of 43 patients had strokes during the stenting procedure. Notably, the point estimates of the 1-year rates of stroke in WASID and SSYLVIA were virtually identical. Given the inherent risks of intracranial stenting, it is likely that the role, if any, for stenting will emerge from randomized controlled trials of patients at particularly high risk of stroke in the territory despite medical treatment. In the present study, we identified risk factors associated with stroke in the territory of a stenotic intracranial artery. The major high-risk features were stenosis 70% to 99% and recent symptoms. The increased risk with severe stenosis validates the findings of the retrospective WASID pilot study and parallels the findings of symptomatic extracranial carotid artery stenosis trials. 13,14 Furthermore, risk increased linearly with percent stenosis, except perhaps among patients with stenosis 90% to 99%. This again parallels findings with severe extracranial carotid stenosis and may be related to low flow in the severely stenotic vessel with little potential for thromboembolism or the development of collateral blood supply. 26 Similarly, the association between recent symptoms and the risk of subsequent stroke in the territory of the stenotic artery has also been observed with extracranial carotid disease. 14 It is possible that a time-selection bias existed in WASID, because some patients deemed to be at high risk of stroke may not have been enrolled early after the initial event to allow for a period of observation, and then were enrolled later after they were thought to be stable enough to participate in a randomized trial. This could have resulted in preferential inclusion of those subjects who remained stroke-free into the latter part of the enrollment period, and they may have represented a lower-risk population. However, this seems relatively unlikely, because the impact of timing was not confounded by the other characteristic most robustly tied to subsequent risk, namely, stenosis 70%. Although timing cannot be regarded as a factor that can discriminate among high- and low-risk patients during evaluation of the initial event, our observations suggest that potential intervention should be considered very soon after clinical presentation, unless early intervention also increases the short-term risk. 27 In contrast, if a patient presents for evaluation relatively late after the initial symptom, without intercurrent events, then perhaps a more conservative approach is warranted. Although the qualifying event type was not a statistically independent predictor of recurrent stroke in the territory of the stenotic artery, patients with stroke and 70% to 99% stenosis seemed to be at particularly high risk, whereas patients with TIA and 50% to 69% stenosis were at low risk. Other studies of patients with symptomatic extracranial carotid stenosis have suggested that initial stroke carries a higher risk of subsequent stroke than TIA, 13,14 although other studies of patients with medically treated carotid stenosis 28 and those with undefined vascular pathology 29,30 suggested that TIA could pose greater risk than stroke. Our empiric finding of an independent association of the baseline NIH Stroke Scale score with increased risk of subsequent stroke suggests that patients with stroke-related deficits are more prone to subsequent stroke than patients presenting with TIA or stroke without residual deficits, although the reason for this is unclear. We suggest caution in the interpretation of the NIH Stroke Scale impact because it was not a prespecified variable, it can be influenced by preexisting and nonstroke factors (such as cognitive deficits or arthritis), and it may be insensitive to brain stem findings that are common in patients with vertebrobasilar stenosis, the circulation involved in nearly half of the patients in the present study. The trend toward greater risk in women is an interesting finding that is not readily explained. 31 In some studies of extracranial carotid stenosis, women had a lower risk of stroke than men and reaped a smaller benefit from intervention. 14 This remains controversial, and potential explanations for the gender gap with extracranial disease include differential arterial anatomy, 32 body size and habitus, 33 and hormonal factors, although considerable uncertainty remains. Women may also have smaller intracranial arteries, which could pose greater risk, as has been suggested in studies of patients with coronary or femoral artery disease in which smaller body size and smaller arteries were associated with greater risk of vascular events. 34,35 Further research is needed to determine why women with intracranial disease appear to be at greater risk than men. Numerous other factors including age, race, location of stenosis, length of stenosis, antithrombotic agent at time of qualifying event, vascular risk factors, prior cerebrovascular events, comorbidities, and treatment assignment were considered, but none of these had a large impact on the risk of stroke in the territory of the stenotic artery. Of note, patients with intracranial stenosis of the vertebral and basilar arteries were not at higher risk of stroke in the territory than those with stenosis of the internal carotid or middle cerebral artery, which contradicts findings of previous retrospective studies. 8,10,11 Additionally, patients who had their qualifying event while they were undergoing antithrombotic therapy had virtually the same rate of stroke in the territory as patients who were not taking antithrombotic therapy at the time of their qualifying event. This finding is also in contrast with the results of a previous retrospective study that suggested that patients taking antithrombotic medications at the time of presentation were at substantially higher risk of recurrent stroke. 12 Potential limitations of the present analysis are related primarily to statistical power and generalizability. We had enough power to detect relatively large effects of baseline characteristics on subsequent risk of stroke in the territory of the stenotic intracranial artery, but smaller effects such as

9 562 Circulation January 31, 2006 those potentially related to the type of qualifying event and interactions among variables may not have been statistically evident. Furthermore, randomized clinical trials such as WASID may have limited generalizability, because patients were carefully selected for enrollment according to strict eligibility criteria. It is therefore possible that some patients may have been systematically excluded from the trial who had different clinical courses. This seems unlikely to be a major drawback, because WASID was the largest study of symptomatic intracranial stenosis to date and included patients with a wide variety of baseline characteristics. Ultimately, randomized trials still provide the best evidence for establishing clinical practice recommendations. The results of the present study have important implications for clinical practice and for future research. Given the high risk of stroke in patients who were randomized soon after their qualifying events, early identification of intracranial stenosis is important for prognosis and possible intervention. Patients with 70% to 99% stenosis and recent symptoms are at particularly high risk of stroke and should constitute the target group for a future randomized trial comparing stenting with medical therapy. Conversely, some patients face a relatively low risk of stroke in the territory of the stenotic artery, and the exclusion of such patients from trials of intracranial stenting may be justified because the upfront risk of this procedure may exceed any potential subsequent benefit. Women may also represent a high-risk group that should be aggressively recruited to participate in intracranial stenting trials. Acknowledgments This study was funded by a research grant (1R01 NS36643, Principal Investigator: Dr Chimowitz) from the US Public Health Service, NINDS. In addition, the following General Clinical Research centers, funded by the NIH, provided local support for the evaluation of patients in the trial: Emory University (M01 RR00039), Case Western University, MetroHealth Medical Center (5M01 RR00080), San Francisco General Hospital (M01 RR ), Johns Hopkins University School of Medicine (M01 RR000052), Indiana University School of Medicine (5M01 RR ), Cedars-Sinai Hospital (M01 RR00425), and the University of Maryland (M01 RR165001). Bristol-Myers-Squibb (after incorporating DuPont Pharma) supplied the warfarin (Coumadin) and placebo warfarin tablets, and Bayer supplied the aspirin and placebo aspirin tablets for the trial. Neither of these companies supplied direct funding for the trial. The Food and Drug Administration assigned an IND number of 57,138 for Coumadin (warfarin sodium) for this trial. Disclosures Dr Kasner reports having received grant support from NINDS and Boehringer Ingelheim, consulting fees from Boehringer Ingelheim and the Sankyo/Lilly Partnership, and lecture fees from Boehringer Ingelheim and Bristol-Myers Squibb; all corporate affiliations are with companies that make antithrombotic agents not evaluated in this study. Dr Chimowitz is the recipient of research grants (R01 NS36643 and R01 NS051688) from the US Public Health Service NINDS to fund this trial. He is also supported by grant 1 K24 NS from the NIH/NINDS and reports being paid fees by the Bristol-Myers Squibb/Sanofi Pharmaceuticals Partnership, Astra- Zeneca, and the Sankyo/Lilly Partnership for consulting on antithrombotic agents that were not evaluated in this trial, and from Guidant Corporation and Cine-Med for consulting on medical devices not evaluated in this trial. M.J. Lynn reports receiving grant support from NINDS and the National Eye Institute (grant U10EY013287). Dr Stern reports receiving grant support from NINDS and has been paid fees by the Bristol-Myers Squibb/Sanofi Pharmaceuticals Partnership. Dr Hertzberg reports receiving grant support from NINDS. Dr Frankel reports receiving grant support from NINDS and the Centers for Disease Control and Prevention and has been paid lecture fees by Boehringer Ingelheim and Sanofi Pharmaceuticals, makers of antithrombotic agents not evaluated in this trial. Dr Levine reports receiving grant support from NINDS (grant K24NS43392), Ono Pharmaceuticals, and the Gaisman Frontiers of Biomedical Research; he also has received fees from AstraZeneca and lecture fees from Boehringer Ingelheim and Inspire for consulting on issues unrelated to this study. Dr Chaturvedi reports receiving grant support from Boehringer Ingelheim and has been paid lecture fees by Bristol-Myers Squibb, Sanofi Pharmaceuticals and Boehringer Ingelheim. Dr Sila reports receiving grant support from NINDS for participating in various multicenter stroke trials and has been paid lecture fees by Bristol-Myers Squibb. Dr Romano reports receiving grant support from the American Heart Association and has been paid lecture fees by Bristol-Myers Squibb and Boehringer Ingelheim. The other authors report no conflicts. References 1. Wityk RJ, Lehman D, Klag M, Coresh J, Ahn H, Litt B. Race and sex differences in the distribution of cerebral atherosclerosis. Stroke. 1996; 27: Feldmann E, Daneault N, Kwan E, Ho KJ, Pessin MS, Langenberg P, Caplan LR. Chinese-white differences in the distribution of occlusive cerebrovascular disease. Neurology. 1990;40: Sacco RL, Kargman DE, Gu Q, Zamanillo MC. Race-ethnicity and determinants of intracranial atherosclerotic cerebral infarction: the Northern Manhattan Stroke Study. Stroke. 1995;26: Chimowitz MI. Angioplasty or stenting is not appropriate as first-line treatment of intracranial stenosis. Arch Neurol. 2001;58: Gomez CR, Orr SC. Angioplasty and stenting for primary treatment of intracranial arterial stenoses. Arch Neurol. 2001;58: The SSYLVIA Study Investigators. Stenting of symptomatic atherosclerotic lesions in the vertebral or intracranial arteries (SSYLVIA): study results. Stroke. 2004;35: The Warfarin-Aspirin Symptomatic Intracranial Disease (WASID) Trial Investigators. Design, progress and challenges of a double-blind trial of warfarin versus aspirin for symptomatic intracranial arterial stenosis. Neuroepidemiology. 2003;22: Chimowitz MI, Kokkinos J, Strong J, Brown MB, Levine SR, Silliman S, Pessin MS, Weichel E, Sila CA, Furlan AJ. The Warfarin-Aspirin Symptomatic Intracranial Disease Study. Neurology. 1995;45: Chimowitz MI, Lynn MJ, Howlett-Smith H, Stern BJ, Hertzberg VS, Frankel MR, Levine SR, Chaturvedi S, Kasner SE, Benesch CG, Sila CA, Jovin TG, Romano JG. Comparison of warfarin and aspirin for symptomatic intracranial arterial stenosis. N Engl J Med. 2005;352: Browne TR III, Poskanzer DC. Treatment of strokes. I. N Engl J Med. 1969;281: Qureshi AI, Ziai WC, Yahia AM, Mohammad Y, Sen S, Agarwal P, Zaidat OO, Suarez JI, Wityk RJ. Stroke-free survival and its determinants in patients with symptomatic vertebrobasilar stenosis: a multicenter study. Neurosurgery. 2003;52: Thijs VN, Albers GW. Symptomatic intracranial atherosclerosis: outcome of patients who fail antithrombotic therapy. Neurology. 2000;55: North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med. 1991;325: Rothwell PM, Eliasziw M, Gutnikov SA, Warlow CP, Barnett H. Endarterectomy for symptomatic carotid stenosis in relation to clinical subgroups and timing of surgery. Lancet. 2004;363: Lachin JM, Foulkes MA. Evaluation of sample size and power for analyses of survival with allowance for nonuniform patient entry, losses to follow-up, noncompliance, and stratification. Biometrics. 1986;42: Armitage P. Tests for linear trends in proportions and frequencies. Biometrics. 1955;11:

10 Kasner et al Stroke in the Territory of Intracranial Stenosis Gomez CR, Misra VK, Liu MW, Wadlington VR, Terry JB, Tulyapronchote R, Campbell MS. Elective stenting of symptomatic basilar artery stenosis. Stroke. 2000;31: Gomez CR, Misra VK, Campbell MS, Soto RD. Elective stenting of symptomatic middle cerebral artery stenosis. AJNR Am J Neuroradiol. 2000;21: Rasmussen PA, Perl J II, Barr JD, Markarian GZ, Katzan I, Sila C, Krieger D, Furlan AJ, Masaryk TJ. Stent-assisted angioplasty of intracranial vertebrobasilar atherosclerosis: an initial experience. J Neurosurg. 2000;92: Levy EI, Hanel RA, Boulos AS, Bendok BR, Kim SH, Gibbons KJ, Qureshi AI, Guterman LR, Hopkins LN. Comparison of periprocedure complications resulting from direct stent placement compared with those due to conventional and staged stent placement in the basilar artery. J Neurosurg. 2003;99: Kim JK, Ahn JY, Lee BH, Chung YS, Chung SS, Kim OJ, Kim WC, Joo JY. Elective stenting for symptomatic middle cerebral artery stenosis presenting as transient ischaemic deficits or stroke attacks: short term arteriographical and clinical outcome. J Neurol Neurosurg Psychiatry. 2004;75: de Rochemont RM, Turowski B, Buchkremer M, Sitzer M, Zanella FE, Berkefeld J. Recurrent symptomatic high-grade intracranial stenoses: safety and efficacy of undersized stents: initial experience. Radiology. 2004;231: Liu JM, Hong B, Huang QH, Xu Y, Zhao WY, Zhang L, Zhao R, Zhou XP. Safety and short-term results of stent-assisted angioplasty for the treatment of intracranial arterial stenosis [in Chinese]. Zhonghua Wai Ke Za Zhi. 2004;42: Zhang QZ, Miao ZR, Li SM, Zhu FS, Hua Y, Wang ML, Song QB, Ling F. Complications of stent-assistant angioplasty of symptomatic intracranial artery stenosis [in Chinese]. Zhonghua Yi Xue Za Zhi. 2003;83: Jiang WJ, Du B, Wang YJ, Wang SX, Wang GH, Jin M. Symptomatic intracranial artery stenosis: angiographic classifications and stent-assisted angioplasty [in Chinese]. Zhonghua Nei Ke Za Zhi. 2003;42: Morgenstern L, Fox A, Sharpe B, Eliasziw M, Barnett H, Grotta J. The risks and benefits of carotid endarterectomy in patients with near occlusion of the carotid artery: North American Symptomatic Carotid Endarterectomy Trial (NASCET) group. Neurology. 1997;48: Gupta R, Schumacher HC, Mangla S, Meyers PM, Duong H, Khandji AG, Marshall RS, Mohr JP, Pile-Spellman J. Urgent endovascular revascularization for symptomatic intracranial atherosclerotic stenosis. Neurology. 2003;61: Eliasziw M, Kennedy J, Hill MD, Buchan AM, Barnett HJ. Early risk of stroke after a transient ischemic attack in patients with internal carotid artery disease. CMAJ. 2004;170: Johnston SC. Transient ischemic attack. N Engl J Med. 2002;347: Johnston SC, Gress DR, Browner WS, Sidney S. Short-term prognosis after emergency department diagnosis of TIA. JAMA. 2000;284: Caplan LR, Gorelick PB, Hier DB. Race, sex and occlusive cerebrovascular disease: a review. Stroke. 1986;17: Schulz UG, Rothwell PM. Sex differences in carotid bifurcation anatomy and the distribution of atherosclerotic plaque. Stroke. 2001;32: Messe SR, Kasner SE, Mehta Z, Warlow CP, Rothwell PM. Effect of body size on operative risk of carotid endarterectomy. J Neurol Neurosurg Psychiatry. 2004;75: Watson HR, Schroeder TV, Simms MH, Horrocks M. Association of sex with patency of femorodistal bypass grafts. Eur J Vasc Endovasc Surg. 2000;20: Ferraris VA, Harrah JD, Moritz DM, Striz M, Striz D, Ferraris SP. Long-term angiographic results of coronary endarterectomy. Ann Thorac Surg. 2000;69: CLINICAL PERSPECTIVE Antithrombotic therapy for intracranial arterial stenosis was evaluated recently in the Warfarin versus Aspirin for Symptomatic Intracranial Disease (WASID) trial, which showed that warfarin was associated with significantly higher rates of adverse events and provided no benefit over aspirin for preventing stroke and vascular death. Importantly, the risk of future stroke in the territory of the stenotic intracranial artery was 14% at 2 years with either antithrombotic therapy. This prompts the question whether new treatment modalities such as intracranial angioplasty and stenting should become first-line therapies for this disease. In this regard, a prespecified aim of WASID, described in the current report, was to identify patients at highest risk for stroke in the territory of the stenotic artery who would be the target group for a subsequent trial comparing intracranial stenting with medical therapy. The risk of subsequent stroke in the territory of the stenotic artery was highest with severe stenosis 70% and in patients enrolled early ( 17 days) after an initial transient ischemic attack or ischemic stroke. Women also appeared to be at increased risk. Variables previously believed to be associated with greater risk, including vertebrobasilar stenosis, type of qualifying event, and prior use of antithrombotic medications, had no significant impact in this study. Future trials of intracranial angioplasty and stenting should target recently symptomatic patients with severe stenosis 70%, and women should be actively recruited to participate.