Stroke is one of the most devastating complications of

Transcription

1 Risk Factors for Early or Delayed Stroke After Cardiac Surgery Charles W. Hogue, Jr, MD; Suzan F. Murphy, RN, BSN; Kenneth B. Schechtman, PhD; Victor G. Dávila-Román, MD Background Stroke after cardiac surgery is a devastating complication that leads to excess mortality and health resource utilization. The purpose of this study was to identify risk factors for perioperative stroke, including strokes detected early after cardiac surgery or postoperatively. Methods and Results Data were obtained from 2972 patients undergoing coronary artery bypass graft and/or valve surgery. Patients 65 years old and those with a history of symptomatic neurological disease underwent preoperative carotid artery ultrasound scanning. Intraoperative epiaortic ultrasound to assess for ascending aorta atherosclerosis was performed in all patients. New strokes were considered as a single end point and were categorized with respect to whether they were detected immediately after surgery (early stroke) or after an initial, uneventful neurological recovery from surgery (delayed stroke). Strokes occurred in 48 patients (1.6%); 31 (65%) were delayed strokes. By multivariate analysis, prior neurological event, aortic atherosclerosis, and duration of cardiopulmonary bypass were independently associated with early stroke, whereas predictors of delayed stroke were prior neurological event, diabetes, aortic atherosclerosis, and the combined end points of low cardiac output and atrial fibrillation. Female sex was associated with a 6.9-fold increased risk of early stroke and a 1.7-fold increased risk of delayed stroke. In-hospital mortality of patients with early (41%) and delayed (13%) strokes was higher than that of other patients (3%, P ). Conclusions Most strokes after cardiac surgery occurred after initial uneventful recovery from surgery. Women were at higher risk to suffer early and delayed perioperative strokes. Atrial fibrillation had no impact on postoperative stroke rate unless it was accompanied by low cardiac output syndrome. (Circulation. 1999;100: ) Key Words: stroke surgery atherosclerosis Stroke is one of the most devastating complications of cardiac surgery, and it can lead to a decreased quality of life and excess mortality The frequency of this complication is reported to be as high as 5% in patients undergoing coronary artery bypass graft (CABG) surgery, almost 9% in CABG patients 75 years of age, and nearly 16% in patients undergoing valve surgery or those with preexisting cerebrovascular disease Adverse neurological events also have important economic consequences, with estimated costs that exceed $2 to $4 billion annually worldwide for patients with stroke after CABG surgery. 7 Moreover, the impact of stroke on patient outcome is likely to remain substantial in light of the predicted increase in elderly patients, who often suffer from comorbidity predisposing to stroke and who will require cardiac surgery in the next century. 11,12 Therefore, identification of individuals at risk for perioperative stroke is increasingly important not only to accurately assess patient risk for surgery but also to foster the development of new strategies to reduce the frequency of this complication. Investigations have identified multiple risk factors for stroke after cardiac surgery, but the clinical applicability of these findings has been restricted by methodological limitations, including the frequent failure to include in the analysis an accurate assessment for important stroke predictors, such as atherosclerosis of the ascending aorta and carotid arteries. 1 10,13 15 Previous studies have also in most cases considered strokes occurring during and after surgery as a single end point, despite reports suggesting that many strokes occur after an initial uneventful neurological recovery from surgery The causes of these delayed postoperative strokes may differ from the causes of those that occur during surgery. Furthermore, the consideration of all strokes as a single end point, regardless of timing of the event, could lead to underestimation of the importance of variables specific to a particular perioperative period. The purpose of this study was to identify risk factors for early and delayed stroke in a cohort of cardiac surgical patients to whom an aggressive strategy was applied to identify atherosclerosis of the carotid arteries and ascending aorta. Methods Patient Population The population consisted of 2972 of 3321 consecutive patients (1900 men and 1072 women) 50 years old who underwent cardiac Received December 22, 1998; revision received May 5, 1999; accepted May 19, From the Department of Anesthesiology (C.W.H., V.G.D.-R.); the Cardiovascular Division, Department of Internal Medicine (V.G.D.-R.); and the Department of Biostatistics (K.B.S.), Washington University School of Medicine and Barnes-Jewish Hospital, St Louis, Mo. Correspondence to Charles W. Hogue, Jr, MD, Department of Anesthesiology, Washington University School of Medicine, 660 S Euclid Ave, Box 8054, St Louis, MO hoguec@notes.wustl.edu 1999 American Heart Association, Inc. Circulation is available at 642

2 Hogue et al Risk Factors for Stroke After Cardiac Surgery 643 TABLE 1. Characteristics of the Patients All Patients (n 2972) Age, y (mean SD) 68 9 Male/female, n 1900/1072 History of neurological event, n (%) 203 (6.8) Diabetes mellitus, n (%) 826 (27.8) Hypertension, n (%) 1941 (65.3) Left ventricular score 9 4 Pulmonary disease, n (%) 282 (9.5) Carotid artery stenosis, n (%) Moderate right stenosis 371 (12.5) Severe right stenosis 92 (3.1) Moderate left stenosis 383 (12.9) Severe left stenosis 98 (3.3) Coronary stenosis, n (%) None 128 (4.3) One vessel 220 (7.4) Two vessels 633 (21.3) Three vessels 1991 (67.0) Left main 612 (20.6) Ascending aorta atherosclerosis, n (%) Normal 1982 (66.7) Mild 535 (18.0) Moderate/severe 455 (15.3) Types of surgery, n (%) CABG 2627 (88.4) CABG/valvular 446 (15.0) Valvular 790 (26.6) CABG/carotid endarterectomy 90 (3) Cardiopulmonary bypass time, min Cross-clamp time, min Postoperative events, n (%) Low cardiac output 383 (12.9) Myocardial infarction 65 (2.2) Atrial fibrillation 1022 (34.4) Moderate and severe carotid artery stenosis refer to stenosis of 50% but 80% and 80% to 99%, respectively. surgery at Barnes-Jewish Hospital from January 1990 through August Patients were excluded (n 349) if aortic root replacement was planned before surgery, if emergency procedures were necessary, or if epiaortic ultrasound equipment was unavailable (see below). This study was approved by the Human Studies Committee at Washington University. Number of strokes detected immediately after surgery (early strokes) and after initial uneventful neurological recovery from surgery (delayed strokes) by day neurological event was detected. Note: Postoperative day 0 refers to day of surgery, which begins after arrival in intensive care unit. Preoperative Data Patient data were analyzed as previously reported, including the recording of information on previous neurological events, such as stroke. 8,16 19 Documentation of a prior stroke required verification by each patient s primary care physician, review of medical records, and review of results of CT and/or MRI when available. Left ventricular function was assessed angiographically and graded by use of criteria of the Coronary Artery Surgery Study. 20 Carotid artery duplex scanning was performed as previously described in patients 65 years old and in younger patients with carotid artery bruits and/or symptoms or history of neurological events, including transient ischemic episodes. 15 Carotid artery stenosis was graded as follows: insignificant or no disease (luminal narrowing 50%); moderate disease (narrowing 50% but 80%); severe disease (narrowing 80 but 99%); and complete occlusion. For the analysis, severe disease and complete occlusion were combined. Intraoperative Data Epiaortic ultrasound of the ascending aorta was performed to evaluate for atheromatous disease, and the information was used at the time of surgery to avoid atheroma during aortic manipulations. 8,16 19 Changes in surgical technique based on epiaortic ultrasound results were classified as minor and major alterations. Minor alterations included a change in any of the following: site of aortic cannulations, aortic cross-clamping, or proximal bypass graft anastomosis. Major alterations included replacement of a portion of the severely atherosclerotic ascending aorta with a Dacron graft under hypothermic circulatory arrest, as previously described. 8 The severity of atherosclerosis was graded independently by 2 blinded investigators as follows: insignificant or no atherosclerosis; mild atherosclerosis (intimal thickening 3.0 mm without intimal irregularities); or moderate to severe atherosclerosis ( 3.0 mm thickening with diffuse irregularities, large mobile or protruding atheromata, ulcerated plaques, and/or thrombi). 8,16 19 Postoperative Data Complications documented included myocardial infarction (new Q waves on the 12-lead ECG or ratio of fractionated lactic dehydrogenase [LDH 1 /LDH 2 ] 1 during the first 72 hours), low cardiac output syndrome (cardiac index of 2.0 L min 1 m 2 for 24 hours after surgery regardless of treatment), renal failure (requiring dialysis), and death. Continuous telemetry ECG monitoring was performed until the time of hospital discharge to document atrial fibrillation. Neurological Complications Stroke was defined as any new permanent global or focal neurological deficit that could not be attributed to other neurological (eg, dementia) and/or medical (ie, metabolic abnormalities, hypoxia, or drugs) processes. Reversible cerebral ischemic events were not included in the analysis because evidence of these events cannot be identified under general anesthesia and their detection is hindered postoperatively owing to residual anesthetics, analgesics, and sedative drugs. Strokes were diagnosed by a neurologist, and in the majority of patients they were confirmed by CT head scan. All stroke data were reviewed by 3 investigators, and the temporal onset of the deficits was classified by consensus as either an early stroke, if the

3 644 Circulation August 10, 1999 TABLE 2. Number and Temporal Pattern of Strokes Based on Types of Surgical Procedure All Strokes (n 48) neurological deficit was present after emergence from anesthesia, or a delayed stroke, if the patient developed the neurological deficit after first awaking from surgery without a neurological deficit. Statistical Analysis Data were analyzed by version 6.12 of SAS. Univariate comparisons between subjects with and without stroke were performed with 2 tests for dichotomous variables and ANOVA for ordered categorical and continuous variables. The latter analyses were performed nonparametrically when regression residuals suggested that the model fit was poor. Stepwise logistic regression was used to select a best set of independent predictors of both early and delayed stroke. Variables entered into the initial logistic models were those with a univariate probability value of P 0.2. The final model included all variables with an independent significance level of P 0.1. The quality of the fit of the logistic model was tested with the Hosmer and Lemeshow goodness-of-fit test. Data for continuous variables are presented as mean SD. A significant difference was considered to exist when P Results Early Strokes (n 17) Delayed Strokes (n 31) CABG surgery, n (%) 29 (1.3) 9 (0.4) 20 (0.9) Valve surgery, n (%) 4 (1.2) 1 (0.3) 3 (0.9) CABG/valve surgery, n (%) 15 (3.4) 7 (1.6) 8 (1.8) P between surgical procedures for all strokes; P 0.01 between procedures for early stroke; P 0.2 between all groups for delayed strokes. Strokes Demographic and other characteristics of all patients are listed in Table 1. Stroke occurred in 48 patients (1.6%). Seventeen strokes (35% of strokes, 0.6% of patients) were detected early after surgery; 31 (65% of strokes, 1.0% of patients) were delayed (Figure). Fourteen early strokes (82%) and 22 delayed strokes (71%) occurred in patients 65 years old. The frequency of stroke after the different surgical procedures that the study patients underwent (Table 2) was significantly different for all strokes (P 0.006) and early strokes (P 0.02) but not for delayed strokes (P 0.2). The number of strokes that occurred in patients in whom alterations in the surgical technique was made for ascending aorta atherosclerosis is shown in Table 3. Brain CT examinations were performed in 93% of patients suffering a stroke. On the TABLE 3. Number of Strokes According to Alterations in Surgical Technique Due to Ascending Aorta Atherosclerosis All Strokes, n (%) Early Strokes, n (%) Delayed Strokes, n (%) No alterations 31 (1.2) 9 (0.3) 22 (0.9) (n 2544) Minor alterations 12 (3.4) 5 (1.4) 7 (2) (n 349) Major alterations (n 79) 5 (6.3) 3 (3.8) 2 (2.5) Minor alterations consisted of change of the site of aortic cross-clamp placement, aortic cannulations, and/or proximal bypass graft anastomoses to avoid atherosclerotic areas. Major alterations consisted of replacement of a portion of the severely atherosclerotic ascending aorta by a Dacron graft. basis of clinical interpretations of the results, the cause of early strokes was believed to be embolic in 11 patients and due to cerebral hypoperfusion in 5 patients. The cause of delayed strokes was believed to be embolic in 18 patients and due to hypoperfusion in 11. Predictors of Stroke The characteristics of patients who suffered strokes and univariate predictors of these events are listed in Table 4. Further analysis demonstrated that female patients were significantly older than male patients (70 8 years versus 67 9 years, P 0.05) and were more likely to have diabetes, hypertension, and low cardiac output syndrome (P 0.05). Atrial fibrillation after surgery was prevalent in patients with and those without stroke. Because of the relationship between atrial fibrillation and stroke in the general population and in cardiac surgical patients, the data were examined to evaluate for covariates that, when present with atrial fibrillation, increased the risk of stroke. 21,22 Because this analysis suggested that postoperative atrial fibrillation is a risk factor for delayed stroke only in the presence of low cardiac output syndrome (Table 5), all multivariate analyses of data on delayed stroke included a variable that combined low cardiac output and postoperative atrial fibrillation. Results from the multivariate logistic regression analysis are listed in Table 6. History of stroke was the strongest independent predictor of perioperative stroke, regardless of whether strokes were considered as a single end point or whether early and delayed strokes were considered separately. Female sex was also independently associated with stroke, regardless of the timing of the event, as was ascending aorta atherosclerosis. Other independent risk factors for stroke, however, were dependent on the timing of the neurological event: duration of cardiopulmonary bypass was an independent predictor of early stroke, whereas diabetes and the combined variable of low cardiac output and atrial fibrillation were additional independent predictors of delayed stroke. The presence of significant carotid artery stenosis was an independent predictor of early stroke only when prior stroke was excluded from the multivariate analysis. Because women were found to have a higher risk of stroke, we evaluated the covariate-adjusted role of sex in greater detail. Multivariate logistic analysis was repeated by including variables that differed between the sexes (P value of 0.2) as well as stroke risk factors such as prior stroke, atrial fibrillation, ascending aortic atherosclerosis, carotid artery stenosis, and hypertension. After correction for these potentially confounding factors, female sex was still independently associated with a 3-fold increased risk of perioperative stroke. Mortality Seven and 4 deaths occurred in patients with early and delayed strokes, respectively. This in-hospital mortality (early strokes, 41%; delayed strokes, 13%) was higher than that observed in the control group (3%, P ). The mortality rate for women (5.4%) was higher (P ) than that for men (2.9%).

4 Hogue et al Risk Factors for Stroke After Cardiac Surgery 645 TABLE 4. Characteristics of Patients With Early and Delayed Strokes and of Patients With No Stroke No Stroke (n 2924) All Strokes (n 48) P (vs No Stroke) Discussion In this study, we sought to evaluate risk factors for perioperative stroke in a cohort of cardiac surgical patients in whom an aggressive strategy was implemented to identify atherosclerotic disease of the carotid arteries and the ascending Early stroke (n 17) P (Early vs No Stroke) Delayed stroke (n 31) P (Delayed vs No Stroke) Age, y (mean SD) Female, % History of neurological event, % Diabetes mellitus, % Hypertension, % Left ventricular score Pulmonary disease, % Carotid artery stenosis, % Moderate right stenosis Severe right stenosis Moderate left stenosis Severe left stenosis Coronary stenosis, % None One vessel Two vessels Three vessels Left main Ascending aorta atherosclerosis, % Normal Mild Moderate/severe Types of surgery, % CABG CABG/valve Valve CABG/carotid endarterectomy Cardiopulmonary bypass time, min Cross-clamp time, min Postoperative events, % Low cardiac output syndrome Myocardial infarction Atrial fibrillation TABLE 5. Incidence of Delayed Stroke in Relation to Low Cardiac Output Syndrome and Postoperative Atrial Fibrillation Low Cardiac Output Syndrome Atrial Fibrillation Delayed Stroke, % Absent Absent 0.88 Absent Present 0.71 Present Absent 1.85 Present Present 3.68 P aorta. We found that the majority of strokes occurred after an initial, uneventful neurological recovery from cardiac surgery. A new finding of this study was that women were at a higher risk for both early and delayed stroke. Also, postoperative atrial fibrillation was associated with increased risk of delayed stroke only when associated with low cardiac output syndrome, and diabetes was associated with increased risk of delayed but not early stroke. A history of previous stroke and ascending aorta atherosclerosis were independent predictors of either early or delayed stroke. Results of the present study confirm that perioperative stroke is associated with a significantly higher in-hospital mortality, regardless of when the event occurs. The stroke rate observed (1.6%) is lower than that reported from other studies (3% to 5.6%) Despite this fact, the percentage of delayed strokes that occurred in the present study (nearly 66% of strokes) is similar to that previously

5 646 Circulation August 10, 1999 TABLE 6. Independent Predictors of Stroke After Cardiac Surgery Variable Odds Ratio (95% CI) P All strokes History of stroke 14.0 ( ) ( ) Female sex 1.6 ( )* Ascending aorta atherosclerosis 1.1 ( ) Cardiopulmonary bypass time Early strokes History of stroke 11.6 ( ) Female sex 6.9 ( ) Ascending aorta atherosclerosis 2.0 ( )* Cardiopulmonary bypass time 1.1 ( ) Delayed strokes History of stroke 27.6 ( ) Diabetes 2.8 ( ) Female sex 2.4 ( ) Low cardiac output syndrome 1.7 ( ) and atrial fibrillation Ascending aorta atherosclerosis 1.4 ( )* Strokes are classified depending on whether the neurological deficit was identified either immediately after surgery (early events) or after initial uneventful neurological recovery (delayed events). *Odds ratio reflects risk of stroke with increase in a single level in the aortic scan results. The odds of an increase of 2 levels (eg, normal to moderate/ severe) is the square of the reported odds ratio. reported, suggesting that the number of strokes that occur after initial recovery from surgery has not changed in more than a decade. 2,3,5 An understanding of the mechanisms for early and delayed stroke and whether they differ has important implications for potential preventive strategies and thus requires further investigation. Moreover, these data suggest that future clinical trials should consider temporal onset of stroke to accurately judge the efficacy of strategies aimed at specifically preventing early and/or delayed strokes. The finding that ascending aorta atherosclerosis was an independent predictor of delayed strokes suggests that risk of stroke associated with this condition may result from mechanisms other than direct atheroembolism. In addition to being a potential cause of cerebral embolism, ascending aorta atherosclerosis may be a marker of widespread atherosclerosis of the aortic arch and cerebral vessels. 8,16 19,23 26 Prior neurological event, carotid artery stenosis, diabetes mellitus, and advanced age have been found in many studies to increase susceptibility to perioperative stroke, possibly by identifying individuals with widespread cerebrovascular disease, impaired cerebral blood flow, and/or increased susceptibility to atheroembolism or thromboembolism. 1 10,27 29 The relative importance of these risk factors for stroke in the present study in comparison with previous reports might result from the aggressive detection of atherosclerosis of the carotid arteries and ascending aorta. These findings support the notion that many of these previously identified risk factors for stroke may represent surrogate markers for risk factors not previously evaluated, such as atherosclerosis of the ascending aorta. The lack of an independent association between age and perioperative stroke in this study suggests that the relationship between these variables may be associated with age-related risk factors and not age per se. The identification of female sex as an independent risk factor for stroke is new In this study, women were more likely to have comorbid conditions and were more likely to have undergone valve surgery. However, after adjustment for potential confounding variables by use of multivariate analysis, female sex remained an independent predictor of stroke. It is possible that female sex has not been identified as a risk factor in previous studies because of the small number of women included in these studies or because of the failure to account for atherosclerosis of the ascending aorta and/or carotid arteries. It is also possible that the sex-related stroke risk that we observed represents an emerging phenomenon arising from the changing characteristics and general aging of cardiac surgical patients. Atrial fibrillation is a frequent complication of cardiac surgery that has been reported to increase the risk of perioperative stroke in some, but not all, studies. 1 10,22 The contribution of postoperative atrial fibrillation to stroke risk may have been underestimated in previous investigations because the timing of the neurological event was not taken into consideration. Indeed, because early strokes precede the onset of postoperative atrial fibrillation, this arrhythmia cannot be a predictor of these events. An equally important explanation may be the strong interaction we observed between postoperative atrial fibrillation combined with low cardiac output syndrome and delayed stroke, an interaction that has not been reported previously (Table 5). Because both complications are associated with cardiac thrombus formation and cerebral hypoperfusion, aggressive therapy may be beneficial for patients with both conditions. Limitations of the Study We were unable to evaluate the efficacy of epiaortic ultrasound in reducing perioperative stroke because patients were not randomly assigned to undergo this procedure and thus there was no control group. These same considerations also limit the utility of comparisons of stroke rates in patients for whom no alterations, minor alterations, or major alterations in aortic manipulations were made (Table 3). Although the diagnosis of stroke was made by a neurologist, a detailed preoperative neurological assessment was not performed in our patients, and thus, paired neurological evaluations were not available. More detailed neurological and neurocognitive examinations might have revealed subtle neurological events, but it is unlikely that strokes went undetected. Carotid artery ultrasound was performed in 75% of all patients, most of whom represented a high-risk group of those with the disease. It is possible that the prevalence of carotid artery disease could have been underestimated, but the 25% of patients who underwent carotid artery ultrasound can be considered to be a low-risk group for significant carotid artery disease on the basis of clinical criteria. 15 Because of the inherent exploratory nature of stepwise regression analysis, we acknowledge that some variables identified to be independently associated with stroke risk (based on probability

6 Hogue et al Risk Factors for Stroke After Cardiac Surgery 647 value close to 0.05) might not be predictors in other studies. This limitation, however, would not apply to predictive variables with a strong level of significance. Conclusions Most strokes after cardiac surgery occur after initial uneventful neurological recovery from surgery. Previous stroke and ascending aorta atherosclerosis were associated with increased risk for perioperative stroke regardless of the timing of onset of the event, but other risk factors appeared to be associated with the time of occurrence. Women were found to be at higher risk for early and delayed perioperative stroke and in-hospital mortality. Atrial fibrillation was found to have no impact on postoperative stroke rate unless accompanied by low cardiac output syndrome. Acknowledgments We acknowledge Benico Barzilai, MD, and Nicholas T. Kouchoukos, MD, for their thoughtful review and advice regarding the manuscript. References 1. Gardner TJ, Horneffer PJ, Manolio TA, Pearson TA, Gott VL, Baumgartner WA, Borkon AM, Watkins L, Reitz BA. Stroke following coronary artery bypass grafting: a ten-year study. Ann Thorac Surg. 1985; 40: Reed GL III, Singer DE, Picard EH, DeSanctis RW. Stroke following coronary artery bypass surgery: a case-control estimate of the risk from carotid bruit. N Engl J Med. 1988;319: Frye RL, Kronmal R, Schaff HV, Myers WO, Gersh BJ. Stroke in coronary artery bypass graft surgery: an analysis of the CASS experience. Int J Cardiol. 1992;36: Tuman KJ, McCarthy RJ, Najafi H, Ivankovich AD. Differential effects of advanced age on neurologic and cardiac risks of coronary artery operations. J Thorac Cardiovasc Surg. 1992;104: Ricotta JJ, Faggioli GL, Castilone A, Hassett JM. Risk factors for stroke after cardiac surgery. J Vasc Surg. 1995;21: Newman MF, Wolman R, Kanchuger M, Marschall K, Mora-Mangano C, Roach G, Smith LR, Aggarwal A, Nussmeier N, Herskowitz A, Mangano DT. Multicenter preoperative stroke risk index for patient undergoing coronary artery bypass graft surgery. Circulation. 1996;94(suppl II):II- 74 II Roach GW, Kanchuger M, Mora-Mangano C, Newman M, Nussmeier N, Wolman R, Aggarwal A, Marschall K, Graham SH, Ley C, Ozanne G, Mangano DT. Adverse cerebral outcomes after coronary bypass surgery. N Engl J Med. 1996;335: Wareing TH, Dávila-Román VG, Daily BB, Murphy SF, Schechtman KB, Barzilai B, Kouchoukos NT. Strategy for the reduction of stroke incidence in cardiac surgical patients. Ann Thorac Surg. 1993;55: Beall AC, Jones JW, Guinn GA, Svensson LG, Nahas C. Cardiopulmonary bypass in patients with previously completed stroke. Ann Thorac Surg. 1993;55: Craver JM, Weintraub WS, Jones EL, Guyton RA, Hatcher CR. Predictors of mortality, complications and length of stay in aortic valve replacement for aortic surgery. Circulation. 1988;78(suppl I):I-85 II Jones EL, Weintraub WS, Craver JM, Guyton RA, Cohen CL. Coronary bypass surgery: is the operation different today? J Thorac Cardiovasc Surg. 1991;101: Peterson ED, Cowper PA, Jollis JG, Bebchuk JD, DeLong ER, Muhlbaier LH, Mark DB, Pryor DB. Outcomes of coronary artery bypass graft surgery in 24,461 patients aged 80 years or older. Circulation. 1995; 92(suppl II):II-85 II Faggioli GL, Curl GR, Ricotta JJ. The role of carotid screening before coronary artery bypass. J Vasc Surg. 1990;12: Salasidis GC, Latter DA, Steinmetz OK, Blair JF, Graham AM. Carotid artery duplex scanning in preoperative assessment for coronary artery revascularization: the association between peripheral vascular disease, carotid artery stenosis, and stroke. J Vasc Surg. 1995;21: Berens ES, Kouchoukos NT, Murphy SF, Wareing TH. Preoperative carotid artery screening in elderly patients undergoing cardiac surgery. J Vasc Surg. 1992;15: Dávila-Román VG, Barzilai B, Wareing TH, Murphy SF, Kouchoukos NT. Intraoperative ultrasonographic evaluation of the ascending aorta in 100 consecutive patients undergoing cardiac surgery. Circulation. 1991; 84(suppl III):III-47 III Kouchoukos NT, Wareing TH, Daily BB, Murphy SF. Management of the severely atherosclerotic aorta during cardiac operations. J Card Surg. 1994;9: Dávila-Román VG, Phillips KJ, Daily BB, Dávila RM, Kouchoukos NT, Barzilai B. Intraoperative transesophageal echocardiography and epiaortic ultrasound for assessment of atherosclerosis of the thoracic aorta. J Am Coll Cardiol. 1996;28: Dávila-Román VG, Barzilai B, Wareing TH, Murphy SF, Schechtman KB, Kouchoukos NT. Atherosclerosis of the ascending aorta: prevalence and role as an independent predictor of cerebrovascular events in cardiac patients. Stroke. 1994;25: The Principal Investigators of CASS and their Associates. The National Heart, Lung, and Blood Institute Coronary Artery Surgery Study (CASS). Circulation. 1981;63(suppl I):I-1 I Wolf PA, Abbott RD, Kannel W. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke. 1991;22: Creswell LL, Schuessler RB, Rosenbloom M, Cox JL. Hazards of postoperative atrial arrhythmias. Ann Thorac Surg. 1993;56: The French Study of Aortic Plaques in Stroke group. Atherosclerotic disease of the aortic arch as a risk factor for recurrent ischemic stroke. N Engl J Med. 1996;334: Katz ES, Tunick PA, Rusinek H, Ribakove G, Spencer FC, Kronzon I. Protruding aortic atheromas predict stroke in elderly patients undergoing cardiopulmonary bypass: experience with intraoperative transesophageal echocardiography. J Am Coll Cardiol. 1992;20: Blauth CI, Cosgrove DM, Webb BW, Ratliff NB, Boylan M, Piedmonte MR, Lytle BW, Loop FD. Atheroembolism from the ascending aorta: an emerging problem in cardiac surgery. J Thorac Cardiovasc Surg. 1992; 103: Amarenco P, Duyckaerts C, Tzourio C, Henin D, Bousser MG, Hauw JJ. The prevalence of ulcerated plaques in the aortic arch in patients with stroke. N Engl J Med. 1992;362: Davis SM, Ackerman RH, Correia JA, Alpert NM, Chang J, Buonanno F, Kelley RE, Rosner B, Taveras JM. Cerebral blood flow and cerebrovascular CO 2 reactivity in stroke-age normal controls. Neurology. 1983; 33: Shaw TG, Mortel KF, Meyer JS, Rogers RL, Hardenberg J, Cutaia MM. Cerebral blood flow changes in benign aging and cerebrovascular disease. Neurology. 1984;34: Bentsen N, Larsen B, Lassen NA. Chronically impaired autoregulation of cerebral blood flow in long-term diabetics. Stroke. 1975;6: