Cerebral tomographic findings in patients undergoing carotid endarterectomy for asymptomatic carotid stenosis: short-term and long-term implications

Cerebral tomographic findings in patients undergoing carotid endarterectomy for asymptomatic carotid stenosis: short-term and long-term implications Piergiorgio Cao, MD, Simona Zannetti, MD, Giuseppe Giordano, MD, Paola De Rango, MD, Gianbattista Parlani, MD, and Nevia Caputo, MD, Perugia, Italy Purpose: Preoperative cerebral imaging has been considered not to be cost-effective in carotid endarterectomy (CEA) for asymptomatic carotid stenosis. Yet, silent brain infarction (SBI) has been associated with the embolization potential of a severe carotid stenosis. Thus the presence of SBI may represent an additional indication for CEA in asymptomatic patients. We examined the predictive value of preoperatively detected silent cerebral lesions on early and late outcomes in patients undergoing CEA for asymptomatic carotid stenosis. Methods: Preoperative cerebral tomographic (CT) scans performed on 301 asymptomatic patients undergoing 346 CEAs from 1986 to 1995 were reviewed by a single neuroradiologist blinded to patients records. Mean follow-up was 67.3 months (range, 24-130 months). The degree of internal carotid lumen reduction was measured bilaterally in all patients (602 carotid arteries); carotid stenosis of 60% or more was found in 399 carotid arteries. Results: Of the 103 (34%) CT scans positive for cerebral lesions, 58% were lacunar. No significant association was observed between the side of the cerebral lesion on CT scan and the severity of the corresponding carotid stenosis; 38 silent lesions were detected in the 203 hemispheres ipsilateral to carotid stenoses that were less than 60% versus 95 SBIs in the 399 hemispheres ipsilateral to carotid stenoses that were 60% or more (19% vs 24%; P =.2). There were no significant differences in the perioperative stroke/death rate in patients with or without cerebral CT lesions (2% vs 1%; odds ratio, 1.94; P =.6). Mortality rate during follow-up was 22% in patients with preoperative SBI and 15% in patients without SBI (P =.1). However, actuarial survival at 10 years was shorter (P =.02) in patients with SBI. Late stroke occurred in 11% of patients with preoperative SBI and in 3% of patients without preoperative SBI (P =.006). Cox regression analysis showed that both preoperative lacunar and nonlacunar infarctions were independent predictors of late stroke (hazard ratio, 3.6; P =.04; and hazard ratio, 7.1; P =.001; respectively). Conclusion: In our experience, preoperative SBI did not occur more frequently in the hemisphere ipsilateral to asymptomatic severe carotid stenosis. Although our study lacks a medically treated control group, our data show that SBI is predictive of poor neurologic outcome in asymptomatic patients undergoing CEA. We conclude that CT before CEA, selectively applied, provides information on long-term neurologic prognosis and that a less aggressive attitude towards CEA in asymptomatic patients with SBI may be justified. (J Vasc Surg 1999;29:995-1005.) From the Unit of Vascular Surgery (Drs Cao, Zannetti, Giordano, De Rango, and Parlani) and the Unit of Neuroradiology (Dr Caputo), Policlinico Monteluce, Perugia, Italy. Presented at the Twenty-fifth Annual Meeting of the New England Society for Vascular Surgery, Toronto, Ontario, Canada, Sep 24 25, 1998. Reprint requests: Piergiorgio Cao, MD, Unità Operativa di Chirurgia Vascolare, Policlinico Monteluce, Via Brunamonti, Perugia, 06122 Italy. Copyright 1999 by the Society for Vascular Surgery and International Society for Cardiovascular Surgery, North American Chapter. 0741-5214/99/$8.00 + 0 24/6/96985 The clinical appropriateness of preoperative cerebral tomographic (CT) scan for patients undergoing carotid endarterectomy (CEA) is still under debate. Preoperative cerebral imaging has been considered not to be cost-effective in CEA for asymptomatic stenosis because both CT scan and magnetic resonance imaging (MRI) are expensive and seldom, if ever, condition surgical treatment. 1-3 Silent brain infarction (SBI) (ie, a low-density area on cerebral CT scan not related to previous neurological symptoms) has been defined as a marker of vascular dis- 995

996 Cao et al June 1999 ease; thus CT scan could be helpful in selecting asymptomatic patients who require CEA. 4-8 The reported incidence of SBI in different studies varies widely from 15% to 84% of patients. 5,6,8-10 This discrepancy may be explained by the different techniques of cerebral imaging used (ie, CT or MRI) by the different age groups, or by different definitions of SBI. The significance of lesions found with CT scans in asymptomatic patients with carotid stenosis who are undergoing CEA remains an open issue, particularly when considering surgical treatment. 7,11-15 In a previous paper we reviewed CT scanning records of 844 patients who underwent CEA for symptomatic and asymptomatic disease. 7 In the present study preoperative CT scans of asymptomatic patients of our previous study cohort, together with CT scans of an additional 48 asymptomatic patients who were recruited subsequently, were reviewed by a neuroradiologist to further define the role of preoperative cerebral imaging before CEA. The aim of this study is to evaluate whether cerebral infarction detected on preoperative CT scan influences early and late outcomes in patients undergoing CEA for asymptomatic carotid stenosis. PATIENTS AND METHODS From January 1986 to December 1995, data were collected in a database from 1199 patients undergoing CEA at our unit. Four hundred-eight CEAs were performed on 364 asymptomatic patients who were referred to our unit for carotid stenosis that was detected during screening for generalized atherosclerosis (peripheral and coronary artery disease) and related risk factors. Preoperative CT scans were reviewed by a single neuroradiologist (N.C.) who was blinded to clinical findings, surgical outcome, and additional imaging (angiography, MRI, or duplex scan). In 63 patients, preoperative CT scans were not available, or the review was not reliable because of poor quality of the film; thus the study cohort comprised 301 patients. Mean age was 66.8 years (range, 41-81 years); 27% were women. Patients were considered asymptomatic in the absence of a previous history of hemispheric, ocular, or vertebrobasilar symptoms. Preoperative work-up at our unit includes neurologic evaluation by an independent audit in all cases when the clinical status of the patient (symptomatic or asymptomatic) is uncertain in the operating surgeon s judgment. The diagnosis of postoperative stroke was based on clinical evaluation with the criteria of the World Health Organization: rapidly developing clinical signs or focal disturbance of cerebral function lasting more than 24 hours or leading to death with no apparent cause other than vascular origin. 16 The residual disability was evaluated according to a modified Rankin scale, 17 and strokes were subdivided into disabling and nondisabling, based on the handicap score of patients 6 months after the event (disabling, Rankin 3). Preoperative CT scans were defined as positive or negative depending on the presence of one or more hypodense lesions based on classic CT templates. 18-20 Two subgroups of cerebral lesions were considered: lacunar and nonlacunar infarctions. Lacunar infarction was defined as a deep, sharply delineated hypodense lesion less than 10 mm in size on CT scan, compatible with the occlusion of a small perforating artery at the base of the brain, as previously defined. 8,21 On the basis of the relationship between nonlacunar infarctions and atherosclerotic disease of the extracranial vessels, when both lacunar and nonlacunar infarctions were present, lesions were classified as nonlacunar. 22 Carotid artery stenosis was diagnosed by angiography in 262 patients and by duplex scan in the remaining patients. The degree of lumen reduction was measured bilaterally in all patients, in a total of 602 carotid arteries. Our carotid database includes data prospectively entered since 1986. The European Carotid Surgery Trial method for measurement of carotid stenosis by angiogram is used. 23 Rothwell et al 24 analyzed the equivalence between the European Carotid Surgery Trial and North American Symptomatic Carotid Endarterectomy Trial methods for measurements of carotid stenosis. Because the present study focuses on asymptomatic patients, in reporting endpoints, stenoses were classified as less than 60% versus 60% or more, according to the Asymptomatic Carotid Atherosclerosis Study guidelines. 25 The criteria for defining stenosis by duplex scan and the velocity cut-points were based on previous validation studies. 26-28 The association between the degree of stenosis and the presence of cerebral ipsilateral hemispheric lesions on CT scan was considered in all 602 carotid arteries and for the type of cerebral infarction. Intracerebral arterial stenosis was defined as lumen reduction of 50% or more of the intracerebral vessel, as detected on cerebral angiography. All postoperative disabling strokes and deaths occurring within 30 days of CEA were considered as major perioperative complications. Minor neurologic and nonneurologic complications were likewise recorded. Mean follow-up was 67.3 months (range, 24-130 months), and no patients were lost to the study. Each patient evaluation consisted of a clinical

Volume 29, Number 6 Cao et al 997 Table I. Demographics and risk factors of 301 patients undergoing CEA for asymptomatic carotid stenosis SBI+ * (n=103) SBI (n=198) n % n % P value Mean age.01 Male 79 77 140 71.3 Alcohol consumption 13 13 27 14.9 Smoking 22 21 50 25.5 Hypertension 55 53 104 53.9 Diabetes 23 22 30 15.1 Hyperlipidemia 40 39 73 37.8 Peripheral occlusive disease 42 41 66 33.2 Coronary artery disease 19 18 42 21.6 Atrial fibrillation 5 5 4 2.2 Contralateral carotid occlusion 15 15 10 5.08 Intracranial arterial stenosis 17 18 22 13.3 * Patients with preoperative SBI on CT scan. Mean age, 68.4 ± 6.5 years. Patients without preoperative SBI on CT scan. Mean age, 66.1 ± 7.8 years. Detected only in patients undergoing preoperative arteriography (SBI+, 92 patients; SBI, 170 patients). examination or telephone interview at 1, 6, and 12 months after surgery and yearly thereafter. When possible, duplex scanning was also part of the follow-up examination. A neurologist evaluated each postoperative neurologic event. CT scan was used to define the nature of postoperative stroke. Data were analyzed with the statistical package SPSS (SPSS Inc, Chicago, Ill) and EpiInfo (Center for Disease Control, Atlanta, Ga) software. Demographics, risk factors, and comorbidities of patients with CT lesions were compared with those of patients without cerebral infarctions, using χ 2 Yates corrected, Fisher s exact test, the Student s t test, and odds ratio (OR) with 95% confidence intervals (CIs). Variables were considered statistically significant at a level of P <.05. The prognostic impact of ischemic CT lesions on subsequent risk of stroke and death was investigated with the use of life tables and Cox regression models. If patients experienced more than one neurologic event during follow-up, only the first event was considered for purposes of life table analysis. Corresponding log rank test, hazard ratio, and 95% CI were used in data analysis. Multivariate analysis was performed taking outcome measures (early and late stroke, late death), as dependent variables, and entering the independent variables (age, sex, diabetes, alcohol, hypertension, hyperlipidemia, smoking, atrial fibrillation, coronary artery disease, peripheral vascular disease, contralateral carotid occlusion, intracranial vessel stenosis, any cerebral infarction, and lacunar and nonlacunar) into the model to identify possible independent predictors. RESULTS A total of 301 CT scans were re-evaluated; no infarctions were found in 198 scans (66%). Of 103 CT scans (34%) positive for cerebral lesions, 60 (58%) were lacunar and 43 (42%) were of nonlacunar type. A total of 133 lesions were detected. Eight CT scans revealed both lacunar and nonlacunar lesions and were classified as nonlacunar infarctions. Fortyfive patients underwent staged bilateral CEA (13 patients in the CT positive group and 32 patients in the CT negative group). Ninety-seven percent of the operated stenoses were 60% or more. Baseline data and risk factors of patients with and without SBIs are shown in Table I; there were no statistically significant differences among groups, with the exception of age. Older patients (P =.01) were more prone to experience SBI before surgery. All patients were receiving antiplatelet therapy after surgery (aspirin or, in the case of gastric intolerance, ticlopidine). In 602 carotid arteries examined, the frequency of carotid stenoses of 60% or more was 399 of 602. On analysis of the side of the cerebral lesions on CT scan relative to the degree of the corresponding carotid stenoses, no significant associations were observed when both the 60% or more and the less than 60% groups of stenoses were evaluated; 38 silent lesions were detected in the 203 hemispheres ipsilateral to carotid stenoses less than 60% versus 95 SBIs in the 399 hemispheres ipsilateral to carotid stenoses of 60% or more (19% vs 24%; P =.2; Table II). The perioperative rate of major complications was not significantly higher for patients with preoperative

998 Cao et al June 1999 Table II. Congruity between carotid stenosis and SBI in 602 carotid arteries (301 patients) Carotid artery stenosis 0%-59% (n=203) 60%-100% (n=399) n % n % P value SBI (n=133) 38 19 95 24.2 Lacunar 19 9.5 58 15.09 Nonlacunar 19 9.5 37 9.9 Table III. Incidence of early (30 days) and late complications in 301 patients SBI+ (n=103) SBI (n=198) n % n % P value OR 95% CI Early events Major stroke/death 2 2 2 1.6 1.94 0.1-27.1 Minor stroke 3 3 2 1.3 2.9 0.3-35.6 Late events Stroke 11 11 5 3.01 4.6 1.4-17.4 Ischemic 6 6 4 2.09 3.0 0.7-14.7 Hemorrhagic 5 5 1 1.01 10.1 1.1-477.7 Ipsilateral to CEA 7 7 1 1 0 14.4 1.8-650.6 Ipsilateral to SBI 9 9 Death 23 22 30 15.1 1.6 0.8-3.1 Vascular death 13 13 19 10.5 1.4 0.6-3.1 SBI, relative to patients without preoperative SBI (2% vs 1%; OR, 1.94; CI, 0.14-27.1; P =.6). Two patients died perioperatively; one of ipsilateral stroke, and the other of pulmonary embolism. Late stroke occurred in 16 patients (5%). Late death occurred in 53 patients (22%). Eight of the late strokes (50%) were ipsilateral to the operated carotid artery. None of these patients had a recurrent stenosis of the operated carotid artery at the time of the neurologic event. In 9 of 103 patients with SBI, late stroke was ipsilateral to the preoperative CT lesion (9%). Causes of late death included myocardial infarction in 22 patients (7%), stroke in 7 patients (2%), cancer in 14 patients (5%), ruptured aortic aneurysm in 3 patients (1%), and other causes in 7 patients (2%). The occurrence of early and late events in patients with and without SBI is summarized in Table III. With the use of life table analysis, risk of death at 10 years was 39% in the group with SBI as compared with 31% in the group without cerebral lesions (log rank test; P =.02; Fig 1; Table IV). Similarly, over the same period the risk of stroke (ipsilateral or contralateral, disabling or not) was 21% and 11% (log rank test; P =.01; Fig 2; Table V). Fig 3 and the corresponding life tables (Table VI) show the cumulative risk of ipsilateral stroke. Multivariate analysis, with the use of Cox regression models, showed that preoperative SBI was an independent predictor of late stroke (Table VII). This risk was evident for both lacunar and nonlacunar lesions and was increased when CT scan was positive for silent nonlacunar infarctions, as compared with lacunar infarctions. Multivariate analysis also showed that lacunar infarction, age more than 70 years, and diabetes were independent predictors of late death. DISCUSSION The concept of SBI and its theoretic and physiopathologic implications has been extensively explored mainly by neurologists. 4-6,8-10,14,15,21,22,29-32 An important study on this issue by Brott et al 6 reported the prevalence of and the radiologic and clinical characteristics of patients with asymptomatic brain infarctions, within the setting of the Asymptomatic Carotid Atherosclerosis Study trial. Likewise, in a detailed review Caplan 5 analyzed the frequency of unexpected infarctions, their detection on MRI as compared with CT, their location and type, and their correlation with clinical features. Conversely, few data are available on the clinical significance of SBIs in patients with asymptomatic carotid disease and the influence of unexpected infarcts on surgical outcome. 2,13 Our study focuses on this aspect of the issue and to our knowledge has the longest follow-up.

Volume 29, Number 6 Cao et al 999 Fig 1. Long-term survival of patients with SBI compared with patients with negative preoperative CT scan. Table IV. Life table analysis of survival Interval Entering interval Withdrawn Exposed to Events Interval survival Cumulative survival Standard error (mo) (n) (n) risk (n) (n) (%) (%) (%) SBI+ 0-12 103 0 103.0 4 96.12 96.12 1.9 12-24 99 0 99.0 4 95.96 92.23 2.64 24-36 95 15 87.5 7 92.0 84.85 3.61 36-48 73 9 68.5 1 98.54 83.62 3.77 48-60 63 14 56.0 1 98.21 82.12 3.98 60-72 48 11 42.5 2 95.29 78.26 4.64 72-84 35 9 30.5 4 86.89 67.99 6.26 84-96 22 7 18.5 2 89.19 60.64 7.43 96-108 13 8 9.0 0 100.0 60.64 7.43 108-120 5 3 3.5 0 100.0 60.64 7.43 120+ 1 1 0.5 0 100.0 60.64 7.43 SBI 0-12 198 0 198.0 5 97.47 97.47 1.11 12-24 193 0 193.0 1 99.48 96.97 1.22 24-36 192 28 178.0 3 98.31 95.34 1.52 36-48 161 12 155.0 6 96.13 91.64 2.08 48-60 143 27 129.5 8 93.82 85.98 2.75 60-72 108 30 93.0 0 100.0 85.98 2.75 72-84 78 17 69.5 1 98.56 84.75 2.98 84-96 60 19 50.5 3 94.06 79.71 3.97 96-108 38 16 30.0 2 93.33 74.4 5.19 108-120 20 11 14.5 1 93.1 69.27 6.92 120+ 8 8 4.0 0 100.0 69.27 6.92

1000 Cao et al June 1999 Fig 2. Stroke-free interval in patients with and without preoperative SBI. Table V. Life table analysis of stroke Interval Entering interval Withdrawn Exposed Events Stroke-free Cumulative Standard (mo) (n) (n) to risk (n) interval (%) stroke free (%) error (%) SBI+ 0-12 103 3 101.5 3 97.04 97.04 1.68 12-24 97 2 96.0 5 94.79 91.99 2.72 24-36 90 21 79.5 1 98.74 90.83 2.92 36-48 68 10 63.0 1 98.41 89.39 3.21 48-60 57 15 49.5 1 97.98 87.59 3.62 60-72 41 11 35.5 0 100.0 87.59 3.62 72-84 30 10 25.0 1 96.0 84.08 4.88 84-96 19 8 15.0 1 93.3 78.48 7.08 96-108 10 7 6.5 0 100.0 78.48 7.08 108-120 3 2 2.0 0 100.0 78.48 7.08 120+ 1 1 0.5 0 100.0 78.48 7.08 SBI 0-12 198 5 195.5 5 97.44 97.44 1.13 12-24 188 1 187.5 1 99.47 96.92 1.24 24-36 186 30 171.0 0 100.0 96.92 1.24 36-48 156 14 149.0 0 100.0 96.92 1.24 48-60 142 35 124.5 0 100.0 96.92 1.24 60-72 107 30 92.0 1 98.91 95.87 1.61 72-84 76 18 67.0 0 100.0 95.87 1.61 84-96 58 22 47.0 0 100.0 95.87 1.61 96-108 36 16 28.0 2 92.86 89.02 4.9 108-120 18 11 12.5 0 100.0 89.02 4.9 120+ 7 7 3.5 0 100.0 89.02 4.9

Volume 29, Number 6 Cao et al 1001 Fig 3. Ipsilateral stroke-free interval in patients with and without preoperative SBI. Table VI. Life table analysis of ipsilateral stroke Ipsilateral Cumulative Interval Entering Withdrawn Exposed Events stroke-free ipsilateral Standard (mo) interval (n) (n) to risk (n) (n) interval stroke-free % error (%) SBI+ 0-12 103 3 101.5 3 97.04 97.04 1.68 12-24 97 4 95.0 2 97.89 95.0 2.18 24-36 91 22 80.0 0 100.0 95.0 2.18 36-48 69 10 64.0 1 98.44 93.52 2.6 48-60 58 15 50.5 1 98.02 91.67 3.14 60-72 42 11 36.5 0 100.0 91.67 3.14 72-84 31 11 25.5 1 96.08 88.07 4.64 84-96 19 8 15.0 1 93.33 82.2 7.14 96-108 10 7 6.5 0 100.0 82.2 7.14 108-120 3 2 2.0 0 100.0 82.2 7.14 120+ 1 1 0.5 0 100.0 82.2 7.14 SBI 0-12 198 5 195.5 3 98.47 98.47 0.88 12-24 190 1 189.5 0 100.0 98.47 0.88 24-36 189 31 173.5 0 100.0 98.47 0.88 36-48 158 16 150.0 0 100.0 98.47 0.88 48-60 142 35 124.5 0 100.0 98.47 0.88 60-72 107 30 92.0 0 100.0 98.47 0.88 72-84 77 18 68.0 0 100.0 98.47 0.88 84-96 59 22 48.0 0 100.0 98.47 0.88 96-108 37 17 28.5 1 96.49 95.01 3.5 108-120 19 12 13.0 0 100.0 95.01 3.5 120+ 7 7 3.5 0 100.0 95.01 3.5

1002 Cao et al June 1999 Table VII. Multivariate analysis: Cox regression (backward stepwise logistic regression) Late death Late stroke P value HR 95% CI P value HR 95% CI SBI.08 1.6 0.9 2.8.003 4.9 1.7 14.5 Lacunar infarct.03 1.9 1.1-3.7.04 3.6 1.1 12.8 Nonlacunar infarct.7 1.1 0.5 2.7.001 7.1 2.1 23.3 Age > 70 years.005 2.2 1.3 3.8.1 2.6 0.8 8.9 Male patients.02 2.6 1.1 6.1.2 3.5 0.4 29.2 Diabetes.04 1.9 1.1 3.5.1 2.5 0.8 8.1 Smoking.07 1.9 0.9 3.7.5 0.6 0.1 2.9 Hypertension.8 0.9 0.6 2.1.7 1.3 0.4 3.7 Hyperlipidemia.6 1.1 0.6 2.1.8 1.0 0.3 3.2 Peripheral artery disease.4 1.2 0.7 2.2.07 2.8 0.9 8.7 Coronary artery disease.08 1.8 0.9 3.5.9 1.1 0.3 3.9 Atrial fibrillation.6 0.5 0.07 4.2.8 0.9 0.04 5.2 Contralateral carotid occlusion.1 1.9 0.8 4.2.4 0.4 0.05 3.8 Intracranial stenosis.9 1.0 0.4 2.1.4 1.6 0.5 5.4 Although there is a tendency in the literature to consider SBIs as occurring more often distally to a severe carotid stenosis, 5,21,23,29,33,34 the present study does not support this conclusion, in accordance with others. 6,30 On analysis of all 602 carotid arteries and comparison of the degree of stenosis, many cerebral infarctions, including nonlacunar infarctions, were found to occur distally to a mild or moderate carotid disease. The data also indicate that the neurologic long-term outcome is remarkably affected in patients with SBI relative to patients without SBI. In a previous paper, we reviewed CT scanning records of 844 patients undergoing CEA for symptomatic and asymptomatic disease and found that the risk of perioperative stroke and death and of late stroke and death in patients with preoperative SBI was increased as compared with patients with negative CT findings. 7 Interestingly, asymptomatic patients with unexpected infarcts showed a higher risk of the development of early and late neurologic events. In the mentioned study, patients who did not experience hemispheric symptoms were considered asymptomatic (ie, patients with vertebrobasilar symptoms). 7 In the present study, we analyzed totally asymptomatic patients of the previous series. Despite the restrictions imposed by the new enrollment criteria and the participation in this study of a neuroradiologist who analyzed all CT scans, our previous findings were confirmed with respect to late stroke and survival. As for early major morbidity and deaths, although different from our previous review, there were no statistically significant differences between patients with and without SBI, it should be appreciated that the number of events was rather small and that a nonstatistically significant figure may have masked a clinical significance (2% major stroke/death rate in patients with SBI vs 1%; OR, 1.94; Table III). Clearly, asymptomatic patients with SBI belong in a category with an unfavorable long-term prognosis. In this regard, we would like to emphasize that, on considering the type of brain lesions, multivariate analysis revealed that the presence of lacunar infarction was an independent predictor of late stroke and death, whereas nonlacunar infarction was an independent predictor of late stroke only. These considerations are in line with the known relationship between lacunar infarction and advanced age. 35 At a time when indications for CEA in asymptomatic patients are still undergoing definition, such findings may be useful because they question the value of surgical repair in specific subgroups. Our data showed that most late events in patients with SBI were ipsilateral to the preoperative silent lesion and to the operated carotid artery and that about one half of the late strokes were hemorrhagic. Although a stroke-free interval of 80% at 10 years in our patients can be considered a satisfactory result, it should be noted that this interval is significantly shorter than in patients without preoperative SBI. In our series the presence of infarction on CT compromised the effect of surgical repair; patients with SBI were almost five times more likely to experience a stroke during follow-up when compared with patients with a negative preoperative CT scan (Table VII). Yet, we cannot exclude that such patients might have experienced even worse outcome if treated medically. Although we are convinced that unexpected infarcts should be taken into account when the choice of treatment is being considered, to derive therapeutic guidelines from our data would not be

Volume 29, Number 6 Cao et al 1003 an easy task. It remains a controversial issue whether or not it would be appropriate to operate on these patients. Only a randomized trial with specific analysis of subgroups of patients can clarify this issue. From a surgical standpoint, it would be useful to possibly provide suggestions on the management of asymptomatic patients with SBI and severe carotid stenosis. It is reasonable to question whether it might be appropriate to perform a CT scan, searching for silent infarcts. Nevertheless, because of the many facets of this partially explored issue and the lack of a control group in ours and previous studies, it is difficult to draw definite conclusions. There is no evidence that asymptomatic patients with SBI would fare better if not undergoing an operation. A poised interpretation of our data and an objective therapeutic attitude, until additional data will be available from randomized trials, may be that of restricting the use of a preoperative CT scan to asymptomatic patients with borderline surgical indication (ie, patients with stenosis between 60% and 70%, patients with severe diabetes, older patients). In such cases, a positive CT may be taken as a decisive factor against surgery. In conclusion the presence of SBI in asymptomatic patients undergoing carotid surgery is a meaningful negative prognostic factor in long-term outcome. In the case of borderline surgical indication, a preoperative CT scan may help refine decision making in the treatment of asymptomatic patients undergoing CEA. We thank Dr. Richard P. Cambria for continuous support of our studies. REFERENCES 1. Ammar AD. Cost-efficient carotid surgery: a comprehensive evaluation. J Vasc Surg 1996;24:1050-6 2. Martin JD, Valentine RJ, Myers SI, Rossi MB, Patterson CB, Clagett GP. Is routine CT scanning necessary in the preoperative evaluation of patients undergoing carotid endarterectomy? J Vasc Surg 1991;14:267-70 3. Dawson DL, Roseberry CA, Fujitani RM. Preoperative testing before carotid endarterectomy: a survey of vascular surgeons attitudes. Ann Vasc Surg 1997;11:264-72 4. Norris JW, Zhu CZ. Silent stroke and carotid stenosis. Stroke 1992;23:483-5. 5. Caplan LR. Significance of unexpected (silent) brain infarcts. In: Caplan LR, Shifrin EG, Nicolaides AN, Moore WS, editors. Cerebrovascular ischemia: investigation and management. London: Med Orion; 1996. p. 423-33. 6. Brott T, Tomsick T, Feinberg W, et al for the Asymptomatic Carotid Atherosclerosis Study Investigators. Baseline silent cerebral infarction in the asymptomatic carotid atherosclerosis study. Stroke 1994;25:1122-9. 7. Cao P, Giordano G, De Rango P, et al. Computerised tomography findings as a risk factor in carotid endarterectomy: early and late results. Eur J Vasc Endovasc Surg 1996;12:37-45. 8. Ricci S, Celani MG, La Rosa F, Righetti E, Duca E, Caputo N. Silent brain infarctions in patients with first-ever stroke: a community-based study in Umbria, Italy. Stroke 1993;24: 647-51 9. Hougaku H, Matsumoto M, Handa N, et al. Asymptomatic carotid lesions and silent cerebral infarction. Stroke 1994;25:566-70. 10. Salgado E, Weinstein M, Furlan AJ, et al. Proton magnetic resonance imaging in ischemic vascular disease. Ann Neurol 1986;20:502-7. 11. Nicolaides A, Kalodiki E, Ramaswami G, Geroulakos G, Stevens J. The significance of cerebral infarcts on CT scans in patients with transient ischaemic attacks. In: Bernstein EF, Callow AD, Nicolaides AN, Shifran EQ, editors. Cerebral revascularisation. London: Med-Orion; 1993. p. 159-78. 12. Ricotta JJ, Ouriel K, Green RM, DeWeese JA. Use of computerized cerebral tomography in selection of patients for elective and urgent carotid endarterectomy. Ann Surg 1985;202:783-7. 13. Graber JN, Vollman RW, Johnson WC, et al. Stroke after carotid endarterectomy: risk as predicted by preoperative computerized tomography. Am J Surg 1984;147:492-7. 14. Sise MJ, Sedwitz MM, Rowley WR, Shackford SR. Prospective analysis of carotid endarterectomy and silent cerebral infarction in 97 patients. Stroke 1989;20:329-32. 15. Habozit B. The silent brain infarct before and after carotid surgery. Ann Vasc Surg 1990;4:485-9. 16. Report of the WHO task force on stroke and other cerebrovascular disorders: recommendations on stroke prevention, diagnosis and therapy. Stroke 1989;20:1407-31. 17. Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. A prospective study of acute cerebrovascular disease in the community: the Oxfordshire Community Stroke Project 1981-86. Part 2. Incidence, case fatality rates and overall outcome at one year of cerebral infarction, primary intracerebral and subarachnoid haemorrhage. J Neurol Neurosurg Psychiatry 1990;53:16-22. 18. Culebras A, Kase CS, Masdeu JC, et al. Practice guidelines for the use of imaging in transient ischemic attacks and acute stroke: a report of the stroke council, American Heart Association. Stroke 1997;28:1480-97. 19. Herderscheë D, Hijdra A, Algra A, Koudstaal PJ, Kappelle LJ, Van Gijn J, for the Dutch TIA Trial Study Group. Silent stroke in patients with transient ischemic attack or minor ischemic stroke. Stroke 1992;23:1220-4. 20. Schneider R, Rademacher M, Wolf S. Lacunar infarcts and white matter attenuation: ophthalmologic and microcirculatory aspects of the pathophysiology. Stroke 1993;24:1874-9. 21. Eliasziw M, Paddock-Eliasziw LM, Barnett HJM. Application of the CHAT classification to patients in the NASCET. In: Caplan LR, Shifrin EG, Nicolaides AN, Moore WS, editors. Cerebrovascular ischemia: investigation and management. London: Med Orion; 1996. p. 697-702. 22. Bogousslavsky J. The plurality of subcortical infarction. Stroke 1992;23:629-31. 23. European Carotid Surgery Trialists Collaborative Group. MRC European Carotid Surgery Trial: interim results for symptomatic patients with severe (70-99%) or with mild (0-29%) carotid stenosis. Lancet 1991;337:1235-43. 24. Rothwell PM, Gibson EJ, Slattery J, Sellar RJ, Warlow CP,

1004 Cao et al June 1999 for the European Carotid Surgery Trialists Collaborative Group. Equivalence of measurements of carotid stenosis: a comparison of three methods on 1001 angiograms. Stroke 1994;25;2435-9. 25. Executive Committee for the Asymptomatic Carotid Atherosclerosis Study. Endarterectomy for asymptomatic carotid artery stenosis. JAMA 1995;273:1421-8. 26. Nicolaides AN, Shifrin E, Dhanjil S, Griffin M. Duplex grading of internal carotid stenosis. In: Caplan LR, Shifrin EG, Nicolaides AN, Moore WS, editors. Cerebrovascular ischemia: investigation and management. London: Med Orion; 1996. p. 101-10. 27. Appleberg M. Reappraisal of duplex criteria to assess significant carotid stenosis with special reference to reports from the North American Symptomatic Carotid Endarterectomy Trial and the European Carotid Surgery Trial. J Vasc Surg 1994;20:642-9. 28. de Bray JM, Glatt B. Quantification of atheromatous stenosis in the extracranial internal carotid artery. Cerebrovasc Dis 1995;5:414-26. 29. Evans GW, Howard G, Murros KE, Rose LA, Toole JF. Cerebral infarction verified by cranial computed tomography and prognosis for survival following transient ischemic attack. Stroke 1991;22:431-6. 30. Boon A, Lodder J, Heuts van Raak L, Kessels F. Silent brain infarct in 755 consecutive patients with a first-ever supratentorial ischemic stroke: relationship with index-stroke subtype, vascular risk factors, and mortality. Stroke 1994;25:2384-90. 31. Waxman SG, Toole JF. Temporal profile resembling TIA in the setting of cerebral infarction. Stroke 1983;14:433-7. 32. Nicolaides AN. Asymptomatic carotid stenosis and risk of stroke: identification of a high risk group (ACSRS): a natural history study. Int Angiol 1995;14:21-3. 33. Eliasziw M, Streifler JY, Spence JD, Fox AJ, Hachinski VC, Barnett HJM, for the North American Symptomatic Carotid Endarterectomy Trial (NASCET) Group. Prognosis for patients following a transient ischemic attack with and without a cerebral infarction on brain CT. Neurology 1995; 45:428-31. 34. Hupperts RMM, Warlow CP, Slattery J, Rothwell PM. Severe stenosis of the internal carotid artery is not associated with borderzone infarcts in patients randomised in the European Carotid Surgery Trial. J Neurol 1997;244:45-50. 35. Chamorro A, Saiz A, Vila N, et al. Contribution of arterial blood pressure to the clinical expression of lacunar infarction. Stroke 1996;27:388-92. Submitted Sep 24, 1998; accepted Dec 8, 1998. DISCUSSION Dr William Flynn. You have clearly shown that people who have had stroke disease, documented by CT scan, have more stroke disease. You have jumped to the conclusion that, in your patients, your data support a generally conservative approach to CEA. I have not seen any specific breakdown in your cases as to whether the CT lesions were on the side of the carotid lesions or not, other than a statement that they were evenly distributed, and there was no association. Certainly some CT lesions were on the side of the carotid lesion; I would guess probably one half. You also have no controls. You do not have any nonoperated patients. I am not certain your data support an argument against surgery. One might suspect that, in a subgroup of patients who had a lesion on the side of a tight carotid lesion, the patients probably would have been shown to have done better with surgery than with conservative treatment. I am not sure your data support your conclusions. Dr Cao. With respect to congruity between carotid stenosis and silent brain lesions, 23% of the brain lesions were ipsilateral to severe carotid stenosis, and 20% of the brain lesions were ipsilateral to stenosis less than 70%. This difference was not statistically significant. We do not have medically treated patients as a control. We did not draw any definite conclusions, and we acknowledge that medically treated patients may have had a worse outcome. As I said, only a randomized trial with specific analysis of subgroups can establish the benefit of surgery in asymptomatic patients with SBI. We found that CT preoperative lesions are associated with a poor long-term prognosis. Late strokes were strongly associated with perioperative CT lesions. What we can say is that there is some different biologic behavior on the hemisphere ipsilateral to the silent brain lesion. Dr Yvon Baribeau (Manchester, NH). This was a very interesting study. Do you have any documentation of arch disease on those patients, for example by transesophageal electrocardiogram? I am a cardiac and vascular surgeon, and I will tell you that a lot of those people have significant aortic arch disease, and carotid disease is concomitant to this diffuse atherosclerotic process. It is the same patient who has a very high risk of subsequent stroke and in prospective studies, the risk of stroke was higher than for any other risk factor, including carotid disease, on a yearly basis. I think that the study would be much more informative if you had a transesophageal echocardiogram concomitant with your carotid evaluation to evaluate their proximal aorta. Maybe the conclusion should be that we can help those patients by intervening on both carotid and arch level for their severe atherosclerosis and reduce the risk of embolization. Dr Cao. You made a very good point. However, only 80% of our patients underwent angiograms and the majority underwent bilateral selective carotid angiogram. I have no data with respect to arch disease. Dr Baribeau. Just a short comment. The angiogram is a bad test, in our experience, to qualify severe atherosclerotic proximal aorta. We have had experience with over 2000 epi-aortic echocardiograms during heart surgery and since a lot of our patients have angiograms of arch and carotid preoperatively for significant carotid disease or

Volume 29, Number 6 Cao et al 1005 brachial gradient, I will tell you that the angiogram might show a very smooth aorta, which is very bad by epi-aortic echocardiogram, particularly for small mobile debris that is not seen through dye contrast. Dr Frank Logerfo (Boston, Mass). I found your data interesting and point out that, years ago, Dr Pat Clagget looked at the value of preoperative and postoperative CT scans of patients undergoing carotid surgery. He found the same incidence of SBIs, I think it was 18% in his series, which is very similar to yours. In that case it was prospective, so all patients undergoing CEA had a CT scan and 18% had SBIs, very similar to yours, even though yours was not done prospectively. I think that adds some support to your data. I am curious about the late strokes. Were these late strokes or late infarcts? Was the diagnosis made on a follow-up CT scan, or did they actually have clinical stroke? Dr Cao. The diagnosis of stroke was made clinically in all cases, both perioperatively and later during follow up. Announcing a Home Page on the WWW for the Vascular Surgical Societies To enter the exciting new world of cyberspace, simply point your computer to : http://www.vascsurg.org and hang on to your hat! You can scan back issues of the Journal of Vascular Surgery, look up a colleague who is a member of most regional vascular societies, review abstracts for upcoming vascular meetings, analyze a challenging Case of the Month, and enjoy many other interesting features. Don t forget to visit the Welcome area for the latest information on navigating the site, and please register for your user name and password if you have not already received these as a member of either The Society for Vascular Surgery or the North American Chapter of the International Society for Cardiovascular Surgery. SEE YOU ON THE WEB! Richard F. Kempczinski, MD