Safety and efficacy of reoperative carotid endarterectomy: A 14-year experience

From the New England Society for Vascular Surgery

Safety and efficacy of reoperative carotid endarterectomy: A 14-year experience Michael C. Stoner, MD, Richard P. Cambria, MD, David C. Brewster, MD, Kendra L. Juhola, BS, Michael T. Watkins, MD, Christopher J. Kwolek, MD, Hong T. Hua, MD, and Glenn M. LaMuraglia, MD, Boston, Mass Background: Reoperative carotid endarterectomy (CEA) is an accepted treatment for recurrent carotid stenosis. With reports of a higher operative morbidity than primary CEA and the advent of carotid stenting, catheter-based therapy has been advocated as the primary treatment for this reportedly “high-risk” subgroup. This study reviews a contemporary experience with reoperative CEA to validate the high-risk categorization of these patients. Methods: From 1989 to 2002, 153 consecutive, isolated (excluding CEA/coronary artery bypass graft and carotid bypass operations) reoperative CEA procedures were reviewed. Clinical and demographic variables potentially associated with the end points of perioperative morbidity, long-term durability, and late survival were assessed with multivariate analysis. Results: There were 153 reoperative CEA procedures in 145 patients (56% men, 36% symptomatic) with an average age of 69 ⴞ 1.3 years. The average time from primary CEA (68% primary closure, 23% prosthetic, 9% vein patch) to reoperative CEA was 6.1 ⴞ 0.4 years (range, 0.3 to 20.4 years). At reoperation, patch reconstruction was undertaken in 93% of cases. The perioperative stroke rate was 1.9%, with no deaths or cardiac complications. Other complications included cranial nerve injury (1.3%) and hematoma (3.2%). Average follow-up after reoperative CEA was 4.4 ⴞ 0.3 years (range, 0.1 to 12.7 years), with an overall total stroke-free rate of 96% and a restenosis rate (>50%) by carotid duplex of 9.2%. Among variables assessed for association with restenosis after reoperative CEA, only younger age was found to be significant (66 ⴞ 2.5 years vs 70 ⴞ 0.7 years, P < .05). The all-cause long-term mortality rate was 29%. Multivariate analysis of long-term survival identified diabetes mellitus as having a negative impact (hazard ratio, 3.4 ⴞ 0.3, P < .05) and lipid-lowering agents as having a protective effect (hazard ratio, 0.42 ⴞ 0.4, P < .05) on survival. Conclusion: Reoperative CEA is a safe and durable procedure, comparable to reported standards for primary CEA, for long-term protection from stroke. These data do not support the contention that patients who require reoperative CEA constitute a “high-risk” subgroup in whom reoperative therapy should be avoided. ( J Vasc Surg 2005;41:942-9.)

Supported by level 1 evidence, carotid endarterectomy (CEA) remains the gold standard compared to medical therapy for stroke prevention in patients with both symptomatic and asymptomatic flow-limiting stenosis of the internal carotid artery.1-3 The surgical treatment of carotid artery stenosis is predicated on both a low operative risk and a durable reconstruction. The perioperative risk of CEA has been well elucidated, and many contemporary series continue to affirm the safety of this operation.4,5 With the advent of new treatment modalities such as carotid artery stenting (CAS), efforts to define the specific subgroup who might benefit from such alternative therapy have received much attention.6-8 With respect to durability of CEA, several variables have emerged such as patch reconstruction, renal function, serum lipid profile, gender, and pharmacotherapy.9-11 ReFrom the Division of Vascular and Endovascular Surgery, Massachusetts General Hospital. Supported in part by the Harold and June Geneen Vascular Surgery Research Fund and the Rita and Monte Goldman Vascular Surgery Research Fund. Competition of interest: none. Presented at the Thirty-first Annual Meeting of the New England Vascular Society, Whitefield, NH, September 17-19, 2004. Reprint requests: Glenn M. LaMuraglia, MD, Massachusetts General Hospital, Division of Vascular and Endovascular Surgery, 15 Parkman St, WAC 464, Boston, MA 02114. 0741-5214/$30.00 Copyright © 2005 by The Society for Vascular Surgery. doi:10.1016/j.jvs.2005.02.047

942

stenosis can occur after CEA, and the patient progresses to either symptomatic disease or severe recurrent stenosis. The actual incidence of recurrent carotid stenosis is ill-defined because of variable follow-up and diagnostic standards. In addition, the natural history of recurrent stenosis may be more benign than the primary atherosclerotic plaque, owing to the histologic differences.12 Although reoperative CEA is reported to be appropriate for patients with significant restenosis,13-15 the need for re-dissection of the prior operative field and other technical hazards has led some to label these patients as “high-risk”. Indeed, a literature exists that suggests local anatomic complications such as cranial nerve injury occur more frequently in these patients.16 Consequently, virtually every CAS trial includes reoperative CEA among the high-risk inclusion criteria. Towards this end, the recently published Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) study has documented the noninferiority of CAS versus CEA in a high-risk subgroup of patients.6 This high-risk classification, though supported in the literature, has been a subject of continued debate. The purpose of this study is to delineate the operative risk, long-term durability, and stroke-free survival benefit of reoperative CEA in a contemporary surgical series. This is particularly timely with the advent of endovascular therapies for carotid stenosis and the inclusion of patients with recurrent carotid disease in the initial indication for CAS.

JOURNAL OF VASCULAR SURGERY Volume 41, Number 6

METHODS Patient selection and data collection. The computerized databases and medical records of the Massachusetts General Hospital (retrospective), Vascular Surgery Registry, and operative logs (prospective) were cross-referenced to identify patients who underwent reoperative CEAs between January 1989 through December 2002 performed by surgeons in the Division of Vascular and Endovascular Surgery. Only isolated, reoperative CEAs were included in the study; those performed with a bypass or in conjunction with cardiac surgery were excluded from analysis. The primary end points of this study included perioperative complications and death, anatomic durability as defined by duplex exam, and stroke-free survival. Patient longevity was also examined as a secondary end point. The temporal relationship from primary CEA and type of reconstruction (patch or primary closure) was recorded. Patient clinical and laboratory data, operative details, postoperative carotid noninvasive studies, and clinical course for those undergoing reoperative CEA were recorded. A clinical criteria determination was made at the time of reoperative CEA, and their definitions are: ●

hypertension: taking antihypertensive medication, consistent blood pressure ⬎150 mm Hg systolic or ⬎90 mm Hg diastolic;

●

diabetes mellitus: receiving insulin, oral hypoglycemic medication, or having two serum blood glucose values ⬎150 mg/dL;

●

coronary artery disease: history of myocardial infarction, congestive heart failure, coronary artery bypass graft or intervention, symptoms of angina, or identification of a positive stress test, and

●

chronic obstructive pulmonary disease: routine use of inhalers or symptoms of lifestyle limiting dyspnea.

In addition, the use of lipid-reduction pharmacotherapy was recorded. The laboratory values used were those at the time of reoperative surgery. Operative procedure. Operative procedures were performed under general anesthesia with electroencephalographic monitoring and selective shunting. During the study period, all CEAs were performed by longitudinal arteriotomy through the common and internal carotid arteries. Venous or prosthetic patches for the closure of the carotid artery were used selectively. Intraoperative assessment of the reconstruction was done by continuous wave Doppler ultrasound scanning. Perioperative outcome. End points in this study included perioperative (30 day) complications and patient longevity. Hospital and patient office records were screened for any adverse event during the hospitalization that was associated with the surgical procedure. Particular importance was afforded to postoperative cardiovascular and neurologic status as documented in the record. For the purpose of this study, a perioperative stroke was defined as a central neurologic deficit ipsilateral to the operative side.

Stoner et al 943

Recurrent (tertiary) stenosis. Postoperative outpatient follow-up was generally within 6 weeks with carotid duplex ultrasound of the operative site. Follow-up continued at least to 2 years, unless evidence of carotid restenosis developed. Any abnormal finding consistent with a stenosis of ⱖ50% (internal-to-common carotid artery ratio ⬎2.0 or an internal carotid artery end-diastolic velocity ⬎100 cm/s) was considered recurrent disease. In addition, progression to ⬎75% stenosis (internal to common carotid artery ratio ⬎4.0 or an internal carotid artery end diastolic velocity ⬎140 cm/s), total occlusion, or the need for subsequent reoperation was considered anatomic failure. A final measure of durability was assessed via the occurrence of a cerebrovascular event associated with the operative side. Long-term outcome. Patient medical records were searched longitudinally for evidence of stroke, recurrent symptoms, or death. When patient long-term follow-up was lacking, it was supplemented by direct telephone contact with the patient. For patients who lacked recent clinical data and could not be reached, the date and causes of death were requested and procured from the National Death Index in Bethesda, Md. These data were used to determine overall longevity and stroke-free survival. Data analysis. The effect of variables on the development of recurrent stenosis and stroke were examined with univariate methods (␹2 or Student’s t test as appropriate). With the exception of age, time from primary CEA, serum cholesterol, and clinical variables were dichotomized. To identify variables associated with the development of recurrent carotid disease requiring reoperation, a previously described database was used.17 A multivariate descending logistic analysis was done to examine clinical differences at the time of the primary CEA that were associated with the need for eventual reoperation. Life-table techniques were used to calculate stroke-free survival and total survival for the reoperative population. Clinical variables that impact total survival were determined by multivariate analysis with the Cox hazard ratio (HR). The analysis was generated using Statistical Analysis System (SAS) software (SAS Institute Inc., Cary, NC). The protocols of this study were independently reviewed and approved by the institutional human study committee. In addition, the review board of the National Death Index in Bethesda, Md approved the protocol for release of the patient data. RESULTS Patient and operative profile. During the 14-year study period, 145 patients underwent 153 reoperative carotid endarterectomies (Tables I and II). Only two operations were performed for tertiary restenosis. Details of the primary CEA indicate that 32% had patch closure, and the average time from primary CEA to reoperation was 6.1 ⫾ 0.4 years. The temporal relationship from primary CEA and indication is shown in Fig 1. With respect to operative indication (symptoms), no difference was noted between early (⬍2 years from primary CEA, 41% of patients) and

JOURNAL OF VASCULAR SURGERY June 2005

944 Stoner et al

Table I. Patient characteristics n Demographics Patients Age Female gender Symptoms TMB TIA Stroke Clinical Diabetes mellitus CAD Hypertension Hyperlipidemia COPD Tobacco (any history) Lipid-lowering therapy Cholesterol (mg/dL)

145 69.9 ⫾ 0.7 61 53 25 18 10 30 78 134 105 19 102 85 199.3 ⫾ 4.1

% 100 41.8 36.3 17.1 12.3 6.9 20.9 53.8 91.8 71.9 13.0 70.3 58.2

TMB, Transient monocular blindness; TIA, transient ischemic attack; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease.

Table II. Surgical characteristics of 153 reoperative carotid endarterectomies n General Total cases Contralateral CEA Time from CEA (yrs) Patch at first CEA Imaging Angiography Computed tomography Magnetic resonance Operative General anesthesia EEG monitoring Shunt Patch reconstruction Dacron ePTFE Vein

%

153 61 6.1 ⫾ 0.4 49

100 39.9

73 10 34

47.7 6.6 22.5

153 145 39 143 79 8 56

100 94.7 25.4 93.5 51.6 5.2 36.6

32

CEA, Carotid endarterectomy; EEG, electroencephalography; ePTFE, expanded polytetrafluoroethylene.

late (⬎2 years, 59% of patients). At reoperation, patch reconstruction was used in 93% of cases. Analysis of factors associated with reoperative CEA. A database that has previously been described13 was used to construct a multivariate model to examine variables associated with the requirement for reoperative CEA (Table III). A total of 2,065 endarterectomies were compared to the 153 in the current study, and hypertension was identified as a significant deleterious variable (odds ratio [OR], 2.8; 95% CI, 1.5 to 5.2; P ⬍ .01) and use of a patch at primary CEA was protective (OR, 0.7; 95% CI, 0.5 to 0.9; P ⬍ .05). Age, gender, diabetes mellitus, coronary disease, smoking history, and lipid profile did not have a significant influence on the requisite for reoperative CEA.

Fig 1. Histogram of time from index carotid endarterectomy (CEA) to reoperative CEA. Indication for reoperation: symptomatic (black) and asymptomatic restenosis (grey).

Morbidity and mortality. True 30-day data was available for 140 (96.5%) of the 145 patients; in-hospital perioperative data were used for the remaining five patients. Three strokes occurred in the study patients: two postoperative and one after diagnostic angiography. The two postoperative strokes resulted in one hemiparesis and one functionally minor hand weakness, whereas the postangiography stroke left the patient with a decreased dorsiflexion of the affected foot. Cranial nerve injury was seen in only two patients (1.3%); one was a permanent hypoglossal nerve injury, the other a transient facial droop and minor tongue deviation. Significant wound hematoma occurred in five (3.2%), with four patients requiring reoperation. The total stroke and death rate was 1.9%, with no perioperative mortality or cardiac events (Table IV). The all-cause mortality (early and late) was 28.7%, and the total stroke rate (early and late) was 3.9% over an average follow-up of 4.4 ⫾ 0.2 years. Secondary restenosis and stroke. Fourteen cases of restenosis (⬎50%) were seen, with two cases of anatomic failure (⬎75%) over an average 4.4-year follow-up period (Table IV). Duplex follow-up of ⬎2 years was available for 122 patients (84.5%). Univariate analysis for restenosis after reoperative CEA or anatomic failure is summarized in Table V. Of the patient and surgical variables examined, only patient age was inversely associated with restenosis (66.1 ⫾ 0.3 years vs 70.0 ⫾ 0.7 years, P ⬍ .05). None of the other comorbidities examined approached statistical significance. Statistical analysis was also undertaken to examine the role of patient and operative factors on the development of either early or late ipsilateral stroke, and no significant variables were identified.

JOURNAL OF VASCULAR SURGERY Volume 41, Number 6

Stoner et al 945

Table III. Multivariate analysis of variables associated with progression to reoperative carotid endarterectomy

Patient variables Age (yrs) Female gender Diabetes mellitus Coronary artery disease Hypertension Hyperlipidemia COPD Tobacco Lipid-lowering therapy Operative variables Patch reconstruction

OR

95% CI

P

0.996 1.309 0.769 0.846 2.878 0.704 1.037 1.035 1.383

0.983-1.009 0.922-1.858 0.505-1.170 0.599-1.193 1.574-5.262 0.432-1.150 0.620-1.734 0.709-1.510 0.948-2.017

.52 .13 .22 .34 .0006* .16 .89 .86 .092

0.664

0.471–0.937

.019†

OR, Odds ratio; CI, confidence interval; COPD, chronic obstructive pulmonary disease. *P ⬍ .01. † P ⬍ .05.

Table IV. Perioperative and long-term outcomes after reoperative carotid endarterectomy

Perioperative Stroke Mortality Myocardial infarction Hematoma/bleeding Nerve injury Other Durability Restenosis Anatomic failure Total (short and long-term) Stroke Mortality

n

%

3 0 0 5 2 3

1.9 0 0 3.2 1.3 1.9

14 2

9.1 1.3

6 44

3.9 28.7

Longevity and stroke-free survival. Life-table analysis techniques were used to examine total survival and stroke-free survival for the study population (Figs 2 and 3). Five- and 10-year survival was 74% ⫾ 5% and 42% ⫾ 8%, with a stroke-free survival at the same time points of 96% ⫾ 3% and 95% ⫾ 4%. Multivariate analysis was undertaken to examine the role of both patient and surgical variables on total survival (Table VI). This demonstrated a significant deleterious effect with advancing age (HR, 1.1 ⫾ 0.02 per year; P ⬍ .05), the presence of diabetes mellitus (HR, 3.4 ⫾ 0.4; P ⬍ .01) and a protective effect of lipid-reduction pharmacotherapy (HR, 0.4 ⫾ 0.4, P ⬍ .01). DISCUSSION Reoperative CEA has been affirmed by many as a safe and durable procedure.13-15 Several studies have documented a low perioperative procedural risk with good durability. However, with the emerging field of extracranial catheter-based therapies, these results have been questioned. Damage to nearby anatomic structures is com-

monly cited as a reason to avoid reoperative surgery in the setting of secondary stenosis. Also, the recurrent lesion may represent a lower risk compared with primary carotid disease, owing to its histologic features.12 In addition, it can be hypothesized that patients with recurrent carotid disease may represent a high-risk cohort because they have demonstrated a propensity for the development of atherosclerotic disease that may be translated to other vascular beds.18,19 This study was undertaken to evaluate both the perioperative risk inherent to reoperative CEA and the durability of this secondary procedure in the context of minimally invasive approaches to recurrent cerebrovascular disease. The average time from the first CEA was 6.1 ⫾ 0.4 years, with 36% demonstrating symptoms with their recurrent disease. As demonstrated by others, this study showed an asymmetric temporal distribution, with 41% of the cases occurring within 2 years of the primary CEA. The timing of restenosis is of biologic importance, with early (⬍2 years) cases being consistent with myointimal hyperplasia and late (⬎2 years) recurrence generally ascribed to atherosclerotic disease.12,14 In this study, time from primary CEA did not have an impact on perioperative complications, development of stroke, or patient longevity. In addition, we were unable to identify any cases of residual disease in the current study. At the primary CEA (many done at other institutions), most patients underwent primary closure of the carotid arteriotomy. During the same period, patch closure was used in about half of our primary CEA cases (52% vs 32%).17 Many of the reoperative cases in the current series were referred to our institution, thus it is impossible to know how many patients with primary closure done at other hospitals did not develop recurrent disease or how many of the carotid reconstructions had residual disease after operative intervention. However, the multivariate analysis comparing primary CEA cases progressing to reoperation with those not requiring further intervention confirmed the benefit of patch reconstruction. Indeed, several prospective trials have demonstrated the protective effects of patch angioplasty, and most consider patch closure the preferred technique.11 In this current study, a subset analysis failed to delineate differences based on type of patch closure at primary CEA. In addition, we identified hypertension as a variable associated with reoperative CEA (80% in those not proceeding to reoperation, 92% in those requiring reoperation), demonstrated also by others.20-22 The relatively high percentage of patients with hypertension and cerebrovascular disease makes this association of limited value, but does suggest the importance of blood pressure control after CEA. Although implicated by other authors, this study did not demonstrate a significant relationship between gender, tobacco, diabetes mellitus, or lipid profile and progression to reoperative CEA. Many of these variables may indeed impact disease progression, but for this cohort of patients, there was no impact on the eventual requirement for reop-

JOURNAL OF VASCULAR SURGERY June 2005

946 Stoner et al

Table V. Univariate analysis of patient and surgical variables associated with subsequent restenosis or stroke Any Restenosis

Patient variables Number Age (yrs) Female gender Diabetes mellitus CAD Hypertension Hyperlipidemia COPD Tobacco Lipid-lowering therapy Cholesterol (mg/dL) Time from primary CEA (yrs) Operative variables Symptomatic Shunt Patch reconstruction

Anatomic Failure

Stroke

Yes (%)

No (%)

Yes (%)

No (%)

Yes (%)

No (%)

14 66.1 ⫾ 0.3* 9 (64) 2 (14.2) 8 (57.1) 12 (85.7) 13 (92.8) 0 (0) 8 (61.7) 12 (85.7) 199.2 ⫾ 11.4 5.9 ⫾ 1.4

139 70.0 ⫾ 0.7 56 (40.3) 28(20.6) 74 (53.6) 129 (91.3) 99 (71.2) 20 (14.4) 99 (71.2) 78 (56.1) 200.2 ⫾ 4.4 6.3 ⫾ 0.4

2 69.9 ⫾ 0.7 2 (100) 0 (0) 1 (50) 2 (100) 2 (100) 0 (0) 2 (100) 2 (100) 226.5 ⫾ 9.5 12.5 ⫾ 4.5

151 67.9 ⫾ 2.2 63 (41) 30 (20) 81 (54) 139 (92) 110 (72) 20 (13) 105 (70) 88 (58) 197.7 ⫾ 4.1 6.1 ⫾ 0.4

6 69.9 ⫾ 3.6 4 (66.7) 3 (50) 2 (33.3) 6 (100) 3 (50) 1 (16.6) 5 (83.3) 2 (33.3) 197.8 ⫾ 19.8 6.2 ⫾ 2.4

147 69.1 ⫾ 1.0 61 (41.5) 27 (18.8) 80 (54.8) 135 (91.8) 109 (74.2) 19 (12.9) 102 (69.4) 88 (59.9) 205.5 ⫾ 3.8 4.6 ⫾ 0.3

5 (35.7) 4 (28.6) 12 (85.7)

50 (35.9) 35 (25.2) 131 (94.2)

1 (50) 1 (50) 0 (0)†

54 (35) 38 (25) 143 (94)

4 (66.7) 1 (16.6) 6 (100)

51 (34.7) 38 (25.8) 137 (93.2)

CAD, Coronary artery disease; COPD, chronic obstructive pulmonary disease; CEA, carotid endarterectomy. *P ⬍ .05. † P ⬍ .01.

Fig 2. Kaplan-Meier curve demonstrates stroke-free survival in the study population as measured from the date of reoperative carotid endarterectomy (CEA) (broken line shows standard error).

erative CEA. In a larger study, our group has demonstrated the negative impact of female gender, renal insufficiency, primary closure, and a protective effect of lipid-lowering agents on the development of restenosis.9 These discrepancies with the present study may be ascribed simply to sample size, temporal collection of the patient cohort, or may be related to the more ambiguous nature of recurrent carotid disease and the clinical indications for reoperation.

Fig 3. Kaplan-Meier curve demonstrates total survival in the study population as measured from the date of reoperative carotid endarterectomy (CEA) (broken line shows standard error).

Periprocedural morbidity in this series is comparable to our larger report of primary CEA.17 Guidelines suggested by the American Heart Association define an acceptable stroke/death rate of ⬍6% for symptomatic patients, and ⬍3% for asymptomatic patients.23 Although these guidelines are intended for primary CEA, another consensus

JOURNAL OF VASCULAR SURGERY Volume 41, Number 6

Stoner et al 947

Table VI. Multivariate analysis of variables associated with all-cause mortality

Patient variables Age (years) Female gender Diabetes mellitus Coronary artery disease Hypertension Hyperlipidemia COPD Tobacco Lipid-lowering therapy ⬍2 years from primary CEA Operative variables Symptomatic Shunt Patch reconstruction

Hazard ratio

P

1.053 ⫾ 0.02 0.826 ⫾ 0.38 3.439 ⫾ 0.38 1.311 ⫾ 0.35 0.453 ⫾ 0.57 0.754 ⫾ 0.45 0.919 ⫾ 0.46 1.049 ⫾ 0.30 0.423 ⫾ 0.38 0.492 ⫾ 0.40

.017* .61 .0018† .44 .16 .53 .85 .88 .026* .078

0.881 ⫾ 0.35 1.598 ⫾ 0.37 1.111 ⫾ 0.64

.71 .21 .87

COPD, Chronic obstructive pulmonary disease; CEA, carotid endarterectomy. *P ⬍.05. † P ⬍.01.

statement has defined the upper limit of acceptable stroke/ death rate for reoperative CEA to be 10%.24 In addition, there was no evidence of postoperative myocardial infarction in these patients, albeit routine screening with electrocardiography or cardiac enzyme assay was not done unless clinically indicated. The low rate of major adverse outcome in this study is echoed by other reoperative series, with stroke/death rates noted between 0% and 7%.14,25-30 The neck hematoma rate was 5%, again consistent with previously published and acceptable rates as defined by the randomized trials.1,23 The two cases of nerve injury seen in this study (3.2%) are also well within cited guidelines for primary CEA and demonstrate that meticulous technique can yield excellent results, despite the presence of a reoperative field.31 However, others have demonstrated that with formal neurologic evaluation, including direct laryngoscopy, a much higher cranial nerve injury rate is observed.16 Although most of these are transient and without clinical significance, it is quite possible that with independent otolaryngology evaluation, the rate of cranial nerve injury would have been higher in this study. This study defined any restenosis after reoperative CEA as a duplex examination documenting restenosis of ⬎50%. There were 14 cases of any restenosis (9.1%) incidence throughout an average follow-up of 4.4 ⫾ 0.2 years. None of these patients went on to require tertiary reoperation or intervention nor did they have a stroke. Univariate analysis demonstrated that secondary restenosis tended to occur in younger patients, but no other covariates were identified. Younger patients were more likely to develop subsequent restenosis, which suggests that this cohort of patients may have innate differences in their vascular biology and mandates close postprocedure follow-up.

Using a more stringent criterion for anatomic failure (⬎75% stenosis or occlusion), we identified only two cases. Univariate analysis demonstrated use of patch angioplasty as the only significant variable influencing the development of failure after reoperative CEA. Although the small number of cases makes this analysis difficult, it does again suggest the protective effects of patch reconstruction for anatomic durability. Patch closure of the carotid arteriotomy was used in almost all of the reoperative cases in this study. Stroke-free survival must be considered the ultimate measure of operative durability. The study population had six strokes (three perioperative and three long-term). The stroke-free survival by life-table analysis at 5 and 10 years was 96% and 95% respectively, thus demonstrating excellent clinical effectiveness from repeat CEA. Other authors, also demonstrating a strong ipsilateral stroke-risk reduction in patients undergoing reoperative CEA, have reported similar results.32 The present study did not do formal neurologic testing and, therefore, most likely underestimates the true incidence of stroke. A second consideration for the cohort of patients developing recurrent carotid stenosis is their expected longevity. Clearly, many patients with recurrent disease are displaying a predilection towards aggressive arteriosclerosis, and carotid disease can be seen as a marker for this systemic process. By life-table analysis, the total survival at 5 and 10 years from reoperation was 74% and 42%. Multivariate analysis demonstrated several important variables that influence this survival. First, age was an obvious factor, with chronologic age at the time of reoperation having a negative impact on survival. This finding raises the question of reoperative surgery in patients with advanced age. In fact, some authors have used operative risk and expected longevity to questioned the role of CEA in the octogeneraian.33,34 An assessment of expected perioperative risk coupled with expected longevity is a requisite. Diabetes mellitus is another significant variable that has been well documented as an independent risk factor for the development of cardiovascular-related adverse events, and subsequent survivial.35,36 Finally, the use of a lipid-lowering agent during the postoperative period had a significant positive impact on survival. This finding was independent of the serum lipid profile, and in fact, hyperlipidemia was not found to be a significant predictor in the multivariate analysis. Population based studies have demonstrated a marked reduction in all-cause mortality and cardiovascular risk in patients treated with lipid-lowering agents.37,38 In addition, HMG-CoA reductase inhibitors have been found to have a pleiotropic effect both in vitro and in vivo, and through this presumptive action are able to influence the remodeling of atherosclerotic disease in a variety of vascular beds.39,40 In the current study, the presence of lipid-lowering pharmacotherapy was found to be protective, and illustrates a point of intervention that must be

JOURNAL OF VASCULAR SURGERY June 2005

948 Stoner et al

considered in these patients, as has been previously reported of all CEAs.9 Some authors have suggested catheter-based interventions for patients with recurrent disease. The proposed benefit is a lower procedural cardiovascular risk coupled with the complete avoidance of cranial nerve injury. Furthermore, it can be argued that stenting in the setting of myointimal hyperplasia has far less potential for atheroembolic complications. Despite these concerns, no myocardial infarctions or deaths occurred in this series after redo CEA. In addition, only one permanent cranial nerve injury was identified, demonstrating that reoperative CEA can be undertaken safely. In fact, recent data suggests that although there is essentially no risk for cranial nerve injury with carotid artery stenting, the risk for restenosis compared with surgically treated patients to be increased.41 Many variables will influence the choice of therapy for recurrent carotid stenosis; important among these is surgical experience and expertise. A valid criticism of the present study can be the objection as to whether these results can be achieved across a broad spectrum of surgeons and practice settings. Such questions are difficult to answer, but these results are similar to other reoperative CEA series.13-15,21,25,27-29 In summary, these data do not support the contention that patients with recurrent carotid disease constitute a high-risk population of patients for whom stenting should be considered initial therapy. Clearly, this proposition will only be definitively answered in the future by results of comparative clinical trials. REFERENCES 1. Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. North American Symptomatic Carotid Endarterectomy Trial Collaborators. N Engl J Med 1991;325:445-53. 2. Endarterectomy for asymptomatic carotid artery stenosis. Executive Committee for the Asymptomatic Carotid Atherosclerosis Study. JAMA 1995;273:1421-8. 3. Halliday A, Mansfield A, Marro J, Peto C, Peto R, Potter J, et al. Prevention of disabling and fatal strokes by successful carotid endarterectomy in patients without recent neurological symptoms: randomised controlled trial. Lancet 2004;363:1491-502. 4. Reed AB, Gaccione P, Belkin M, Donaldson MC, Mannick JA, Whittemore AD, et al. Preoperative risk factors for carotid endarterectomy: defining the patient at high risk. J Vasc Surg 2003;37:1191-99. 5. Halm EA, Chassin MR, Tuhrim S, et al. Revisiting the appropriateness of carotid endarterectomy. Stroke 2003;34:1464-71. 6. Yadav JS, Wholey MH, Kuntz RE, Fayad P, Katzen BT, Mishkel GJ, et al. Protected carotid artery-stenting versus endarterectomy in high risk patients. N Engl J Med 2004;35:1493-501. 7. Carotid revascularization using endarterectomy or stenting systems (CARESS): phase I clinical trial. J Endovasc Ther 2003;10:1021-30. 8. Ouriel K, Yadav JS. The role of stents in patients with carotid disease. Rev Cardiovasc Med. Spring 2003;4(2):61-7. 9. LaMuraglia GM, Stoner MC, Brewster DC, Watkins MT, Juhola KL, Kwolek CJ, et al. Determinants of carotid endarterectomy anatomic durability: effects of serum lipids and lipid lowering drugs. J Vasc Surg 2004:in press. 10. Sarac TP, Hertzer NR, Mascha EJ, O’Hara PJ, Krajewski LP, Clair DG, et al. Gender as a primary predictor of outcome after carotid endarterectomy. J Vasc Surg 2002;35:748-53. 11. AbuRahma AF, Robinson PA, Saiedy S, Kahn JH, Boland JP. Prospective randomized trial of carotid endarterectomy with primary closure

12.

13. 14.

15.

16. 17.

18.

19.

20.

21. 22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

and patch angioplasty with saphenous vein, jugular vein, and polytetrafluoroethylene: long-term follow-up. J Vasc Surg 1998;27: 222-32; discussion 233-4. Hunter GC. Edgar J. Poth Memorial/W.L. Gore and Associates, Inc. Lectureship. The clinical and pathological spectrum of recurrent carotid stenosis. Am J Surg 1997;174:583-8. Harris RA, Stow N, Fisher CM, Neale ML, Appleberg M. Carotid redo surgery: both safe and durable. ANZ J Surg 2003;73:1000-3. Hill BB, Olcott Ct, Dalman RL, Harris EJ Jr, Zarins CK. Reoperation for carotid stenosis is as safe as primary carotid endarterectomy. J Vasc Surg 1999;30:26-35. O’Hara PJ, Hertzer NR, Karafa MT, Mascha EJ, Krajewski LP, Beven EG. Reoperation for recurrent carotid stenosis: early results and late outcome in 199 patients. J Vasc Surg 2001;34:5-12. AbuRahma AF, Choueiri MA. Cranial and cervical nerve injuries after repeat carotid endarterectomy. J Vasc Surg 2000;32:649-54. LaMuraglia GM, Brewster DC, Moncure AC, Dorer DJ, Stoner MC, Trehan SK, et al. Carotid endarterectomy at the millennium: what interventional therapy must match. Ann Surg 2004;240:535-44; discussion 544-6. Hallerstam S, Larsson PT, Zuber E, Rosfors S. Carotid atherosclerosis is correlated with extent and severity of coronary artery disease evaluated by myocardial perfusion scintigraphy. Angiology 2004;55:281-8. Simons PC, Algra A, Bots ML, Banga JD, Grobbee DE, van der Graaf Y. Common carotid intima-media thickness in patients with peripheral arterial disease or abdominal aortic aneurysm: the SMART study. Second Manifestations of ARTerial disease. Atherosclerosis 1999;146: 243-8. Reilly LM, Okuhn SP, Rapp JH, Bennett JB, Ehrenfeld WK, Goldstone J, et al. Recurrent carotid stenosis: a consequence of local or systemic factors? The influence of unrepaired technical defects. J Vasc Surg 1990;11:448-9; discussion 459-60. Mansour MA. Recurrent carotid stenosis: prevention, surveillance, and management. Semin Vasc Surg 1998;11:30-35. Rapp JH, Qvarfordt P, Krupski WC, Ehrenfeld WK, Stoney RJ. Hypercholesterolemia and early restenosis after carotid endarterectomy. Surgery 1987;101:277-82. Biller J, Feinberg WM, Castaldo JE, Whittemore AD, Harbaugh RE, Dempsey RJ, et al. Guidelines for carotid endarterectomy: a statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke 1998;29: 554-62. Beebe HG, Clagett GP, DeWeese JA, Moore WS, Robertson JT, Sandok B, et al. Assessing risk associated with carotid endarterectomy. A statement for health professionals by an Ad Hoc Committee on Carotid Surgery Standards of the Stroke Council, American Heart Association. Circulation 1989;79:472-3. Coyle KA, Smith RB 3rd, Gray BC, Salam AA, Dodson TF, Chaikof EL, et al. Treatment of recurrent cerebrovascular disease. Review of a 10-year experience. Ann Surg 1995;221:517-21; discussion 521-4. Bartlett FF, Rapp JH, Goldstone J, Ehrenfeld WK, Stoney RJ. Recurrent carotid stenosis: operative strategy and late results. J Vasc Surg 1987;5:452-6. AbuRahma AF, Snodgrass KR, Robinson PA, Wood DJ, Meek RB, Patton DJ. Safety and durability of redo carotid endarterectomy for recurrent carotid artery stenosis. Am J Surg 1994;168:175-8. AbuRahma AF, Jennings TG, Wulu JT, Tarakji L, Robinson PA. Redo carotid endarterectomy versus primary carotid endarterectomy. Stroke 2001;32:2787-92. Mansour MA, Kang SS, Baker WH, Watson WC, Littooy FN, Labropoulos N, et al. Carotid endarterectomy for recurrent stenosis. J Vasc Surg 1997;25:877-83. Rockman CB, Riles TS, Landis R, Lamparello PJ, Giangola G, Adelman MA, et al. Redo carotid surgery: an analysis of materials and configurations used in carotid reoperations and their influence on perioperative stroke and subsequent recurrent stenosis. J Vasc Surg 1999;29:72-80; discussion 80-1. Maroulis J, Karkanevatos A, Papakostas K, Gilling-Smith GL, McCormick MS, Harris PL. Cranial nerve dysfunction following carotid endarterectomy. Int Angiol 2000;19:237-41.

JOURNAL OF VASCULAR SURGERY Volume 41, Number 6

32. Cho JS, Pandurangi K, Conrad MF, Shepard AS, Carr JA, Nypaver TJ, et al. Safety and durability of redo carotid operation: an 11-year experience. J Vasc Surg 2004;39:155-61. 33. Alozairi O, MacKenzie RK, Morgan R, Cooper G, Engeset J, Brittenden J. Carotid endarterectomy in patients aged 75 and over: early results and late outcome. Eur J Vasc Endovasc Surg 2003;26:245-9. 34. Norman PE, Semmens JB, Laurvick CL, Lawrence-Brown M. Longterm relative survival in elderly patients after carotid endarterectomy: a population-based study. Stroke 2003;34:e95-8. 35. Brand FN, Kannel WB, Evans J, Larson MG, Wolf PA. Glucose intolerance, physical signs of peripheral artery disease, and risk of cardiovascular events: the Framingham Study. Am Heart J 1998;136:919-27. 36. Brand FN, Abbott RD, Kannel WB. Diabetes, intermittent claudication, and risk of cardiovascular events. The Framingham Study. Diabetes 1989;38:504-9. 37. Athyros VG, Papageorgiou AA, Mercouris BR, Athyrou VV, Symeonidis AN, Basayannis EO, et al. Treatment with atorvastatin to the

Stoner et al 949

38.

39.

40. 41.

National Cholesterol Educational Program goal versus ‘usual’ care in secondary coronary heart disease prevention. The GREek Atorvastatin and Coronary-heart-disease Evaluation (GREACE) study. Curr Med Res Opin 2002;18:220-8. Pedersen TR, Tobert JA. Benefits and risks of HMG-CoA reductase inhibitors in the prevention of coronary heart disease: a reappraisal. Drug Saf 1996;14:11-24. Napoli C, Sica V. Statin treatment and the natural history of atherosclerotic-related diseases: pathogenic mechanisms and the risk-benefit profile. Curr Pharm Des 2004;10:425-32. Wolfrum S, Jensen KS, Liao JK. Endothelium-dependent effects of statins. Arterioscler Thromb Vasc Biol 2003;23:729-36. AbuRahma AF, Bates MC, Wulu JT, Stone PA. Early postsurgical carotid restenosis: redo surgery versus angioplasty/stenting. J Endovasc Ther 2002;9:566-72.

Submitted Oct 22, 2004; accepted Feb 5, 2005.

INVITED COMMENTARY

A. Ross Naylor, MD, FRCS, Leicester, United Kingdom Despite being the most evidence-based procedure in surgical history, carotid endarterectomy (CEA) maintains an unparalleled reputation for controversy, none more so than the management of recurrent stenosis. In the United Kingdom and Scandinavia, few advocate intervention for recurrent stenosis unless the patient is symptomatic. However, in mainland Europe and the United States, there is a totally opposing viewpoint. Here the only debate is whether treatment should be surgical or interventional. How is it possible that the same scientific literature can be interpreted in such a polarized way? The paper by Stoner et al is typical of supporters of redo surgery in its fundamental assumption that an asymptomatic, recurrent stenosis (ie, neointimal hyperplasia) confers an equivalent stroke risk to atherosclerotic lesions. (Astute readers will have observed, however, that Stoner et al did not have the same overwhelming need to intervene on the 14 patients who developed recurrent stenoses after redo surgery and who remained completely asymptomatic!) Thereafter, practice is usually justified by the results of the Asymptomatic Carotid Atherosclerosis Study (ACAS) and Asymptomatic Carotid Surgery Trial (ACST).1,2 This is reflected by the authors’ comments that “reoperative CEA is an accepted treatment for recurrent stenosis” and “reoperative CEA has been affirmed as a safe and durable procedure.” Do these claims stand up to close scrutiny? Are the authors’ results generalizable to routine clinical practice? Notwithstanding the inevitable problems associated with retrospective studies (apparently no one lost to follow-up despite some living far away, no clinical cardiac events, no data on patients with recurrent stenoses not subjected to surgery, only two cranial nerve injuries identified, and a nonstandard definition of perioperative stroke), the principle results are extremely good and the authors are to be commended. In this particular surgical unit, it would be wrong to say that a policy of redo surgery was inappropriate in the face of a 1.9% complication rate. However, I suspect that this very low level of risk, which is less than that observed after primary CEA in ACAS and ACST, is unlikely to be typical of overall clinical practice, and this observation is supported by the latest multistate audit by Kresowik et al.3 In the Kresowik et al series of 401 contemporary, redo CEAs in 10 states in the United States, the risk of perioperative death and stroke was 5.7%. If this is a true reflection of practice—and it probably is—and one assumes that the natural history (stroke risk) of patients with recurrent disease is similar to that of asymptomatic atherosclerotic patients, a simple reworking of the ACAS data shows that only 25 ipsilateral strokes would be prevented at 5 years

by operating on 1,000 patients with restenosis where the perioperative risk is 5.7%. Similarly, Stoner et al acknowledged the emergence of angioplasty as an alternative to redo CEA and cited the Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) trial as showing noninferiority between the two treatment strategies.4 In fact, 78% of SAPPHIRE patients were asymptomatic (many were recurrent stenoses) in whom the procedural risk averaged almost 6% for both treatment limbs. As with the Kresowik et al study, this is a level of risk where there is no natural history evidence that any intervention is warranted at all! In summary, vascular units with results comparable to that of Stoner et al can quite easily justify treating patients with recurrent stenosis after CEA. However, those with complication rates ⬎4% cannot simply extrapolate the ACAS or ACST data to justify intervention in patients with asymptomatic recurrent stenoses after CEA. This statement clearly questions whether existing American Heart Association (AHA) Guidelines should still apply. As Stoner et al noted, the AHA advises surgeons that the stroke risk after surgery for recurrent stenosis should be ⱕ10%.5 If we assume that medically treated patients in ACAS and ACST do represent the true natural history risk for patients with asymptomatic, recurrent disease—which remains a generous assumption—surgeons performing redo CEA with a 10% stroke risk will cause approximately 20 more strokes per 1,000 redo CEAs at 5 years than would have occurred if patients had been left on medical treatment alone. Stoner et al are to be commended for their results, but national guidelines need to be revised. REFERENCES 1. Asymptomatic Carotid Surgery Trial Collaborators. The MRC Asymptomatic Carotid Surgery Trial (ACST): Carotid endarterectomy prevents disabling and fatal carotid territory strokes. Lancet 2004;363:1491-502. 2. Executive Committee for the Asymptomatic Carotid Atherosclerosis Study. Endarterectomy for asymptomatic carotid artery stenosis. JAMA 1995;273:1421-8. 3. Kresowik TF, Bratzler DW, Kresowik RA, Hendel ME, Grund SL, Brown KR, et al. Multistate improvement in process and outcomes of carotid endarterectomy. J Vasc Surg 2004;39:372-80. 4. Yadav JS, Wholey MH, Kuntz RE Fayad P, Katzen BT, Mishkel GJ, et al Protected carotid artery stenting versus endarterectomy in high-risk patients. N Engl J Med 2004;351:1493-501. 5. Biller J, Feinberg WM, Castaldo JE, Whittemore AD, Harbaugh RE, Dempsey RJ, et al. Guidelines for carotid endarterectomy: A statement for healthcare professionals from a special writing group of the Stroke Council, American Heart Association. Stroke 1998;29:554-62.