Search

To determine whether choice of material used for patch closure following carotid artery endarterectomy (CAE) influences rates of early or late restenosis, stroke, and death, 274 consecutive CAEs were retrospectively reviewed. Saphenous vein (SV) was used in 159 (58.0%) procedures; everted, double-thickness jugular vein (JV) was used in 25 (9.1%); and knitted Dacron (KD) was used in 90 (32.9%). Primary closure was not used in this series. There were four perioperative strokes: two (1.3%) in SV, one (4%) in JV, and one (1.1%) in KD (NS). Follow-up was obtained on 263 (96%) operated arteries (mean 41.5 months). Duplex scan results were available for 236 (89.7%) of these arteries (mean follow-up time 33.7 months). There were three (2%) late strokes in SV and two (2.2%) in KD (NS). In long-term follow-up, one patient (0.7%) in SV and two (2.4%) in KD developed >80% stenosis (NS). One patient (0.7%) in SV, one (5.3%) in JV, and one (1.2%) in KD had total occlusion of the operated vessel (NS). Three procedures (2.2%) in SV, 1 (5.3%) in JV, and 7 (8.5%) in KD demonstrated moderate stenosis (50–79%) (NS). Three-year follow-up shows that choice of patch material does not affect early or late stroke rate, stroke-related death rate, rate of high-grade (>80%) restenosis, or rate of total occlusion. There is a higher incidence of moderate stenosis in KD. Although our results and a review of the literature do not indicate that these patients are at increased risk for symptoms or progression of stenosis, they should be followed by duplex scanning to ensure that this is the case.

Purpose: When managing a new neurologic deficit after carotid endarterectomy (CEA), the surgeon is often preoccupied with determining the cause of the problem, requesting diagnostics tests, and deciding whether the patient should be surgically reexplored. The goal of this study was to analyze a series of perioperative neurologic events and to determine if careful analysis of their timing and mechanisms can predict which cases are likely to improve with reoperation. Methods: A review of 2024 CEAs performed from 1985 to 1997 revealed 38 patients who manifested a neurologic deficit in the perioperative period (1.9%). These cases form the focus of this analysis. Results: The causes of the events included intraoperative clamping ischemia in 5 patients (13.2%); thromboembolic events in 24 (63.2%); intracerebral hemorrhage in 5 (13.2%); and deficits unrelated to the operated artery in 4 (10.5%). Neurologic events manifesting in the first 24 hours after surgery were significantly more likely to be caused by thromboembolic events than by other causes of stroke (88.0% vs 12.0%, P

Objective: The ideal method of arterial reconstruction in operations for recurrent carotid disease after prior endarterectomy is unknown. The goal of this study was to review a series of carotid reoperations and to determine whether the surgical technique influenced the rate of perioperative stroke, late stroke, or secondary restenosis. Methods: A retrospective review was conducted of 82 carotid reoperations performed on 74 patients at our institution. Results: The patient population included 39 men (52.7%) and 35 women (47.3%), with a mean age of 67.5 years. The indications for redo surgery included transient ischemic attack or amaurosis fugax in 35.3% of the patients, stroke in 6.1%, and asymptomatic restenosis (>80%) in 58.5%. Patch angioplasty with or without redo endarterectomy was used in 47 cases (57.3%), with saphenous vein in 26 (31.7%), Dacron in 15 (18.3%), and polytetrafluoroethylene in 6 (7.3%). Interposition grafting was used in 35 cases (42.7%), with saphenous vein in 9 (11.0%), Dacron in 10 (12.2%), and polytetrafluoroethylene in 16 (19.5%). The perioperative complications included three strokes (3.7%). There was a trend toward increased perioperative neurologic complications with interposition grafting when compared with patch angioplasty (8.6% vs 2.1%), although this did not reach statistical significance. Long-term clinical follow-up was obtained in all cases with a mean duration of 35 months, with follow-up duplex scanning performed in 89.2%. The late failures of redo surgery included four significant secondary restenoses and five total occlusions. There was a trend towards improved long-term results with interposition grafting as opposed to patch angioplasty. However, the cases in which reconstruction was performed with a vein had a significantly higher rate of late failures (stroke, secondary recurrent stenosis, or occlusion) than those in which reconstruction was performed with any prosthetic material (26.7% vs 2.3%; P = .002 by Fisher exact test). Conclusion: The use of autologous material for redo carotid surgery in any configuration appears to significantly increase the rate of subsequent recurrent stenosis or total occlusion of the operated artery. The reason for this finding is unclear but may be related to both host and technical factors. Prosthetic material may be more durable in the long-term for redo carotid surgery. Interposition grafting for redo carotid surgery may increase the perioperative neurologic complication rate to some degree; however, this was not statistically significant in this series. Interposition grafting may be a more durable solution in long-term follow-up than redo endarterectomy and patch angioplasty. A longer follow-up period will be needed to confirm this conclusion. (J Vasc Surg 1999;29:72-81.)

It has been suggested that general anesthesia is the preferred method for reoperative carotid surgery for several reasons, including: the difficulty of the reoperative dissection; the disease may extend unusually high into the internal carotid artery; and the reconstruction required may be more complex than a typical endarterectomy. The purpose of this study is to show that reoperative carotid surgery can be performed safely under regional anesthesia. The records of 109 reoperative carotid operations performed on 96 patients over the past 25 years were reviewed. Procedures performed under regional anesthesia were compared to those performed under general anesthesia with respect to patient characteristics, intraoperative courses, and perioperative results. Regional anesthesia was utilized in 79 operations (72.5%); 30 operations were performed with general anesthesia (27.5%). The two patient groups were essentially equivalent with regard to atherosclerotic risk factors, preoperative neurologic symptoms, and the prevalence of contralateral total occlusion. The etiologies for recurrent disease included recurrent atherosclerosis (50.4%), intimal hyperplasia (30.3%), and vein patch aneurysm (9.2%). The methods of reconstruction employed included saphenous vein patch (47.7%), vein interposition graft (11.9%), prosthetic patch (20.2%), and prosthetic graft (20.2%). Perioperative strokes occurred in one case performed under regional anesthesia (1.3%), and in two cases under general anesthesia (6.6%); this difference was not statistically significant. Reoperative carotid artery surgery can be performed under regional anesthesia safely in the majority of instances. The aforementioned theoretical factors in favor of general anesthesia could also lead to technical difficulties with intraarterial shunt insertion. Having the patient awake, even if just long enough to prove that the patient will tolerate carotid artery clamping, might simplify many of these operations by avoiding shunt insertion. Regional anesthesia should therefore be considered an acceptable option in cases of reoperative carotid surgery.

Purpose: Although it has been widely accepted as the evidence supporting prophylactic carotid endarterectomy, aspects of the Asymptomatic Carotid Atherosclerosis Study have left unease among clinicians who must decide which individuals without symptoms should undergo surgery. Additional confusion has been created by the fact that the several large randomized trials investigating the efficacy of carotid endarterectomy have classified and analyzed different categories of carotid stenosis. In an effort to provide more information on the natural history of asymptomatic, moderate carotid artery stenosis (50% to 79%), we have reviewed data on approximately 500 arteries.Methods: Records of our vascular laboratory from 1990 to 1992 were reviewed. We identified 425 patients with asymptomatic, moderate carotid artery stenosis; 71 patients had bilateral stenoses in this category, resulting in 496 arteries for study.Results: The mean length of follow-up was 38 ± 18 months. New ipsilateral strokes occurred in 16 (3.8%) patients. New ipsilateral transient ischemic attacks occurred in 25 (5.9%) patients. Documented progression of stenosis occurred in 48 (17%) of the 282 arteries for which a repeat duplex examination was available. Arteries that progressed to >80% stenosis were significantly more likely to have caused strokes than those that remained in the 50% to 79% range (10.4% vs 2.1%, p < 0.02). Conversely, arteries that remained stable in the degree of stenosis were significantly more likely to have remained asymptomatic than those that progressed (92.7% vs 62.5%, p < 0.001). With life-table analysis the estimated cumulative ipsilateral stroke rate was 0.85% at 1 year, 3.6% at 3 years, and 5.4% at 5 years. The respective estimated cumulative transient ischemic attack rates were 1.9%, 5.5%, and 6.3%. The respective estimated cumulative rates for progression of stenosis were 4.9%, 16.7%, and 26.5%. Life-table comparison of ipsilateral stroke revealed a significantly higher cumulative rate among arteries that progressed in the degree of stenosis than among those that remained stable (p < 0.001).Conclusions: Based on the low rate of permanent neurologic events in these cases, prophylactic carotid endarterectomy for the asymptomatic, moderately stenotic internal carotid artery cannot currently be recommended. The only factor that appears to predict increased risk for future stroke is progression of stenosis. Careful follow-up with serial repeat duplex examinations must be performed in these patients. Until there are widely accepted duplex parameters that can provide all clinicians with accurate identification of arteries with narrowing corresponding to 60% stenosis as defined by the Asymptomatic Carotid Atherosclerosis Study, all surgeons will need to be aware of specifically how their noninvasive laboratories are deriving their results. For the many laboratories that continue to use the University of Washington criteria, 80% should remain the level above which prophylactic carotid endarterectomy is warranted. (J Vasc Surg 1997;25:423-31.)

To determine whether the choice of material used for patch closure following carotid artery endarterectomy (CAE) affected the immediate operative results, the early follow-up results, or the incidence of early restenosis, a retrospective study of 275 consecutive carotid endarterectomies by two vascular surgeons was performed. Among 275 primary CAEs performed between July 1991 and August 1993, 159 (57.8%) were closed with saphenous vein (SV), 25 (9.1%) with double-thickness internal jugular vein (JV), and 91 (33.1%) with knitted Dacron (KD). Primary closure was not used in any of the arteries in this series. The overall perioperative mortality rate was 1.1% and the rate of major and minor morbidity was 4.4%. There were four (1.5%) perioperative strokes: two (1.3%) in the SV group, one (4.0%) in the JV group, and one (1.1%) in the KD group. Two-hundred fifty-eight (93.8%) of the 275 endarterectomies were followed postoperatively for 2 to 35 months (mean 14.4). Two-hundred nineteen (79.6%) were evaluated using duplex scans during follow-up with a mean interval of 13.7 months. Of the arteries studied, four (3.6%) in the SV group, none in the JV group, and six (8.4%) in the KD group demonstrated restenosis of > 50% at the time of follow-up (NS). In addition, one (0.9%) artery in the SV group, one (5.6%) in the JV group, and none in the KD group demonstrated complete occlusion. Retrospective analysts of the data showed no statistically significant differences in perioperative morbidity, mortality, or early postoperative restenosis whether the artery was closed with saphenous vein, jugular vein, or knitted Dacron patches. Longer follow-up is needed to determine whether rates of late restenosis and aneurysmal dilatation will differ between synthetic and autologous patches.

From a registry of 2406 carotid endarterectomies performed on 1818 patients over a 19-year period, 29 patients (1.6%) underwent reoperations for recurrent stenosis. Reoperations were performed for symptomatic stenosis for 23 and asymptomatic greater than 80% stenosis for six patients. Compared to the entire series, there was no difference in the incidence of restenosis for men and women. The pathologic findings were myointimal hyperplasia in 27%, atherosclerosis in 53%, thrombus with vessel dilatation in 17% and extrinsic scar in 3%. Redo endarterectomy with patch angioplasty was used for reconstruction in 27 patients and patch angioplasty alone in two. There were no operative deaths or strokes. Late follow-up (mean 50 months) revealed only one stroke and six other deaths. Although 21 (75%) were alive and stroke-free, follow-up studies suggest a high incidence (21%) of tertiary lesions among patients who have undergone redo endarterectomy for recurrent stenosis.

Reluctance to use saphenous vein (SV) for patch closure of the carotid arteriotomy is due in part to the concern of vascular surgeons that the SV should be preserved for possible future coronary or lower extremity reconstruction. Even among those who favour vein patch closure of the arteriotomy for improved immediate and late results, an effort has been made to use the ankle portion of the SV, preserving the upper, larger segment for later surgery. Recent reports of rupture of the patches formed from ankle SV and a study showing a decrease in strength of the ankle segment of vein, raised the question of the importance of trying to preserve the proximal SV and the impact of use of this segment in those patients requiring secondary vascular procedures. We reviewed 134 consecutive carotid patients from 1981 who had proximal SV harvested for patch angioplasty. Of those 122 were available (mean 44.9 months). Thirteen had secondary vascular procedures. Adequate saphenous vein was available in twelve. We conclude from this study that (1) less than 15% of patients undergoing carotid surgery will require a secondary vascular surgery within 5 years and (2) harvesting SV from the thigh rarely compromises future revascularisation.

ObjectivesStroke puts a major financial burden on our healthcare system. However, carotid duplex scanning performed as a screening test for occult carotid artery stenosis (CAS) currently is not reimbursed by Medicare. The goals of this study were to develop a cost-effective stroke screening program, to determine the prevalence of potential causes of stroke in this population, and to define a population at high risk in which screening would be most effective.MethodsIn a community-based stroke screening program, patients were eligible if they were older than 60 years and had a history of either hypertension, heart disease, or cigarette smoking, or a family history of stroke. Screening included blood pressure determination, an electrocardiographic rhythm strip, and a previously validated modified carotid duplex ultrasound examination to detect CAS 50% or greater. The relationships between standard demographic risk factors and screening outcomes were analyzed.ResultsScreening was performed in 610 patients. Unilateral or bilateral CAS was detected in 66 patients (10.8%). The finding of occult CAS was more prevalent than that of new hypertension (2.6%) or new atrial fibrillation (0.5%). Patients with known hypertension were significantly more likely to have CAS than were those without hypertension (12.7% vs 7.8%; P = .05). Patients with heart disease were significantly more likely to have CAS than were those without heart disease (18.2% vs 8%; P < .0001). Patients with both risk factors were significantly more likely to have occult carotid artery disease than were patients without either risk factor (22.1% vs 8.5%; P < .0001). Multivariate analysis with logistic regression revealed a history of heart disease as an independent predictor of occult carotid artery disease (odds ratio 95% confidence interval, 1.4-4.4). Type of heart disease was not a significant factor in predicting occult CAS. Direct cost of the screening, including community outreach, nurses, technicians, support staff, and miscellaneous expenses, was less than $75 per patient.ConclusionsIn a screening program for treatable causes of potential stroke, CAS was the most commonly diagnosed disease. More than one of every five patients with known hypertension and heart disease had occult CAS. Known heart disease of any type was a significant independent predictor of occult CAS. Screening for treatable causes of potential stroke can be cost-effective. This information could help to further target populations to screen for occult CAS and to justify reimbursement for screening carotid duplex scanning examinations.

ObjectivesThe diagnosis and treatment of carotid artery disease is an integral part of stroke prevention. However, a population of patients who would benefit from screening for carotid artery stenosis has not been well defined. As part of an institutional stroke-screening program, a modified, rapid duplex scan was developed to evaluate patients for occult carotid stenosis. The goal of this study was to evaluate risk factors predictive of carotid stenosis in a selected population, and to identify patients who would benefit from carotid screening.MethodsPatients were eligible for the study if they were >60 years of age and had a history of hypertension, heart disease, current smoking, or family history of stroke. A modified carotid duplex scan that had been previously validated against formal duplex scanning was utilized; this involved visualization of the carotid bulb and proximal internal carotid artery where Doppler flow velocities were obtained and recorded.ResultsScreening was performed on 394 patients. Thirty-eight patients (9.6%) had either unilateral or bilateral carotid stenosis of ≥50%. Risk factors evaluated included smoking, hypertension, cardiac disease, or hypercholesterolemia. If none of these risk factors was present, the incidence of carotid stenosis was 1.8%. This increased to 5.8% with one risk factor, 13.5% with two risk factors, and 16.7% with three risk factors. Two of three patients with all four risk factors had carotid stenosis (66.7%). Logistic regression and prespecified contrast statements for multiple comparisons were used to assess the relationship between the presence of risk factors and occult carotid artery stenosis. The presence of any one of these risk factors was associated with a statistically significant increase in the presence of occult carotid stenosis (P < .01). This was also statistically significant for the presence of any two risk factors (P < .01) or three risk factors (P < .05).ConclusionThe prevalence of carotid stenosis significantly increases with the presence of one or more identifiable demographic risk factors in a selected population. Assuming the diagnosis and treatment of carotid stenosis are fundamental to stroke prevention, screening for carotid artery disease is justified in this group of patients.

Proponents of carotid angioplasty and stenting (CAS) believe that this technique would be preferred over carotid endarterectomy (CEA) for the high-risk patient. Presumably this would include patients over 80 years of age. However, a recent large series of patients undergoing CAS revealed a 16% incidence of nonfatal strokes and deaths for patients over the age of 80; these results were significantly worse than those for younger patients undergoing CAS. The objective of this study was to reassess results of CEA in patients over 80, and to compare surgical results with the published results of CAS in this patient group. A review was conducted of a prospectively maintained database of all carotid surgery performed at our institution. Primary CEA that took place from 1997 through 1999 were included for analysis (n = 698). Our institutional results were compared with representative results from a recently published large series of CAS. Our analysis showed that CEA can be performed safely in the octogenarian, and results are equivalent to those of younger patients. CEA appears to have significantly better results in the octogenarian than CAS. The reasons for the poor outcomes of CAS in the octogenarian are unclear. The results of CAS in the older patient population are worrisome, and this "less invasive" technique may prove to be an inferior alternative in this patient group.

The natural history and clinical significance of type II or branch vessel endoleaks following endovascular aortic aneurysm (AAA) repair remain unclear. Some investigators have suggested that these endoleaks have a benign course and outcome and that they can be safely observed. The purpose of this study was to document the natural history and outcome of all type II endoleaks that have occurred following endovascular AAA repair at our institution. A review of a prospectively compiled database of all endovascular AAA repairs performed at our institution was performed. From this review, we determined that type II endoleaks appear to have a relatively benign course, with a reasonable chance of spontaneously sealing within a 2-year period. No cases of rupture or aneurysm enlargement were documented in patients with open type II leaks. However, almost one-third of the patients did not manifest a type II leak until after their initial CT scan. The implications of such a "delayed" leak are unclear. Careful follow-up remains mandatory in patients with type II endoleaks to better define outcome.

Prior studies have suggested that young patients may be more prone to recurrent disease after carotid endarterectomy (CEA). The goal of this study was to review a series of CEAs performed on younger patients (or = 55 years) and to determine if these patients are more likely to develop recurrent stenosis. A review was conducted of CEAs performed from 1985 through 1994. Analysis was performed on a study group of 94 young patients who underwent 109 CEAs during this time. A control group of 222 patients older than 55 years who underwent 256 CEAs during the years 1991 through 1993 was selected for comparison. During a mean of nearly 4 years of follow-up, younger patients were significantly more likely to experience a late failure of CEA, including total occlusion of the operated artery, or recurrent stenosis requiring redo surgery. Careful patient evaluation is important in choosing younger patients who require CEA. Implications of these data include mandating careful noninvasive follow-up examinations for younger patients undergoing CEA.

In order to maximize the efficacy of carotid endarterectomy (CEA), the rate of perioperative stroke must be kept to a minimum. A recent analysis of carotid surgery at our institution found that most perioperative strokes were due to technical errors resulting in thrombosis or embolization. From 1992 through 1997 we have performed nearly 1200 additional CEAs; the purpose of this study was to examine recent trends in the causes of perioperative stroke, with specific attention to differences in symptomatic and asymptomatic patients. The records of 1041 patients undergoing 1165 CEAs were reviewed from a prospectively compiled database. Analysis of these data showed that a history of preoperative stroke appears to increase the risk of perioperative stroke after CEA. Surgical factors associated with perioperative stroke include an inability to tolerate clamping, use of an intraarterial shunt, and having surgery performed under general anesthesia; these factors are clearly interrelated and only the use of intraarterial shunting remains a risk factor by multivariate analysis. Over half of all perioperative strokes (54%) appear to be caused by intraoperative or postoperative thrombosis and embolization. The patient requiring use of intraarterial shunting and/or with a preoperative stroke most likely has a significant watershed area of brain at increased risk of infarction. However, technical errors are still the most common cause of perioperative stroke in these high-risk patients. Such high-risk patients may manifest clinical stroke from small emboli that may be tolerated by asymptomatic clamp-tolerant patients. Technical precision and appropriate cerebral protection are particularly critical for successful outcomes in high-risk patients.

Objectives: Although the results of the Asymptomatic Carotid Atherosclerosis Study clearly demonstrated the benefit of surgical over medical management of severe carotid artery stenosis, the results for women in particular were less certain. This was to some extent because of the higher perioperative complication rate observed in the 281 women (3.6% vs 1.7% in men). The objective of this study was to review a large experience with carotid endarterectomy in female patients and to determine whether the perioperative results differed from those of male patients. Methods: A review was conducted of a prospectively compiled database on all carotid endarterectomies performed between 1982 and 1997. Operations performed in 991 female patients were compared with those performed in 1485 male patients. Results: Female patients had a significantly lower incidence of diabetes, coronary artery disease, and contralateral carotid artery occlusion than did male patients. Female patients had a significantly higher incidence of hypertension. There were no significant differences in the age, smoking history, anesthetic route, shunt use, or clamp tolerance between the two groups. Of 991 female patients, 659 (66.5%) had preoperative symptoms, whereas 332 (33.5%) cases were performed for asymptomatic stenosis. Among 1485 male patients, 1041 (70.1%) had symptoms, and 444 (29.9%) were symptom free before surgery. There were no significant differences noted in the perioperative stroke rates between men and women overall (2.3% vs 2.4%, P =.92), or when divided into symptomatic (2.5% vs 3.0%, P =.52) and asymptomatic (2.0% vs 1.2%, P =.55) cases. Conclusions: Carotid endarterectomy can be performed with equally low perioperative stroke rates in men and women in both symptomatic and asymptomatic cases. In this series, symptom-free female patients had the lowest overall stroke rate. The concerns of the Asymptomatic Carotid Atherosclerosis Study regarding the benefit of carotid endarterectomy in female patients should therefore not prevent clinicians from recommending and performing carotid endarterectomy in appropriately selected symptom-free female patients. (J Vasc Surg 2001;33:236-41.)

The purpose of this study was to review a single-institution experience with the Endovascular Technologies [(EVT) Menlo Park, CA] transfemoral, endovascular system of abdominal aortic aneurysm repair. This study was performed at a medical center participating in the phase 1 and phase 2 evaluations of the EVT device. We reviewed the 25 cases performed at our institution. The patient population consisted of 21 males (84%) and 4 females (16%), with a mean age of 73.4 years. A total of eight tube grafts (32%) and 17 bifurcated grafts (68%) were attempted. Twenty-two of the twenty-five grafts were successfully implanted endovascularly (88%). Implantation failures were due to tortuosity or inadequate caliber of the iliac arteries, or incorrect positioning of the graft. The results show that endovascular repair of abdominal aortic aneurysms is an appropriate treatment for selected patients. Conversions to open repair have decreased as experience has grown; careful patient selection can minimize the number of unsuccessful implantations. Patient selection and accurate technique can also minimize the number of endoleaks.

Patients who have sustained a preoperative stroke are at increased risk for perioperative stroke after carotid endarterectomy. At our institution this risk was recently shown to be increased two-to threefold. The purpose of this study was to investigate the reasons for the increased surgical risk in these patients. Records of 606 patients undergoing 704 consecutive carotid endarterectomies from 1988 through 1993 were reviewed. Patients who suffered preoperative strokes (n = 183) were compared to those who were either asymptomatic or experienced only transient ischemic attacks (TIAs) preoperatively (n = 423). Of the 183 patients who had suffered preoperative strokes, eight patients who experienced perioperative strokes after endarterectomy were compared with 175 who successfully underwent surgery. Patients with a prior stroke had an increased perioperative stroke rate (4.4% versus 1.2%, p = 0.01). They had a significantly higher incidence of hypertension (62.6% versus 47.9%, p0.001), cardiac disease (54.7% versus 40.7%, p = 0.001), and positive smoking history (52% versus 40.6%, p = 0.01) than did the asymptomatic/TIA patients. The presence of contralateral total occlusion was also significantly increased (22% versus 10.3%, p0.001). Although not statistically significant due to the overall small number of patients who sustained perioperative strokes, the preoperative stroke patients who sustained perioperative strokes had a higher incidence of hypertension (87.5% versus 61.5%) and contralateral total occlusion (37.5% versus 21.3%) than did those who successfully underwent surgery. Patients with both a prior stroke and contralateral total occlusion had a 7.5% perioperative stroke rate. Patients with both a prior stroke and hypertension had a 6.1% perioperative stroke rate. The perioperative strokes in patients with prior strokes were not related to the severity of the prior stroke, the interval between the stroke and surgery, the use of a shunt, or the type of anesthesia employed. Patients who have sustained preoperative strokes have a higher incidence of significant medical illnesses and overall cerebrovascular disease. Hypertension and total occlusion of the contralateral carotid artery appear to be particularly poor prognostic indicators of outcome after endarterectomy in these patients. Patients who have sustained preoperative strokes may be more likely to display clinical neurologic symptoms in response to any form of cerebral ischemia. In this higher risk subgroup, intraoperative and surgeon-dependent factors appear to play less of a role.

Purpose: The purpose of this study was to examine the cause of perioperative stroke after carotid endarterectomy. Methods: The records of 2365 patients undergoing 3062 carotid endarterectomies from 1965 through 1991 were reviewed. Sixty-six (2.2%) operations were associated with a perioperative stroke. The mechanism of stroke was determined in 63 of 66 cases. Patient risk factors and surgeon-dependent factors were analyzed. Results: More than 20 different mechanisms of perioperative stroke were identified, but most could be grouped into broad categories of ischemia during carotid artery clamping ( n = 10), postoperative thrombosis and embolism ( n = 25), intracerebral hemorrhage ( n = 12), strokes from other mechanisms associated with the surgery ( n = 8), and stroke unrelated to the reconstructed artery( n = 8). Dividing the operative experience approximately into thirds, during the years 1965 to 1979, 1980 to 1985, and 1986 to 1991 the perioperative stroke rates were 2.7%, 2.2%, and 1.5%, respectively. This, in part, is associated with a better selection of patients (more symptom free, fewer with neurologic deficits). There has been a notable decrease in perioperative stroke caused by ischemia during clamping and intracerebral hemorrhage, but postoperative thrombosis and embolism remain the major cause of neurologic complications. Conclusions: Although patient selection seems to play a role, most perioperative strokes were due to technical errors made during carotid endarterectomy or reconstruction and were preventable. (J VASC SURG 1994;19:206-16.)