Journal of Vascular Surgery
Volume 49, Issue 2 , Pages 331-339.e1, February 2009

Outcomes after carotid endarterectomy: Is there a high-risk population? A National Surgical Quality Improvement Program report

Presented at the Annual Meeting of the Society for Vascular Surgery, San Diego, Calif, Jun 5-8, 2008.

  • Jeanwan L. Kang, MD

      Affiliations

    • Division of Vascular and Endovascular Surgery, Department of Surgery, Massachusetts General Hospital and Harvard Medical School, Boston, Mass
    • Corresponding Author InformationReprint requests: Jeanwan Kang, MD, Department of Surgery, Massachusetts General Hospital, 55 Fruit St, GRB-425, Boston, MA 02114
    • ,
  • Thomas K. Chung, MA

      Affiliations

    • Division of Vascular and Endovascular Surgery, Department of Surgery, Massachusetts General Hospital and Harvard Medical School, Boston, Mass
    • ,
  • Robert T. Lancaster, MD

      Affiliations

    • Ernest Amory Codman Center for Clinical Effectiveness in Surgery, Department of Surgery, Massachusetts General Hospital and Harvard Medical School, Boston, Mass
    • ,
  • Glenn M. LaMuraglia, MD

      Affiliations

    • Division of Vascular and Endovascular Surgery, Department of Surgery, Massachusetts General Hospital and Harvard Medical School, Boston, Mass
    • ,
  • Mark F. Conrad, MD

      Affiliations

    • Division of Vascular and Endovascular Surgery, Department of Surgery, Massachusetts General Hospital and Harvard Medical School, Boston, Mass
    • ,
  • Richard P. Cambria, MD

      Affiliations

    • Division of Vascular and Endovascular Surgery, Department of Surgery, Massachusetts General Hospital and Harvard Medical School, Boston, Mass

Received 29 June 2008; accepted 11 September 2008.

Article Outline

Objective

Carotid endarterectomy (CEA) is the standard treatment of carotid stenosis for symptomatic and asymptomatic patients. Carotid angioplasty and stenting (CAS), however, has been proposed as alternative therapy for patients deemed at high-risk for CEA. This study examined 30-day adjudicated outcomes in a contemporary series of CEAs and assessed the validity of criteria used to define a potential high-risk patient population for CEA.

Methods

Patients undergoing isolated CEA in private sector hospitals between Jan 1, 2005, and Dec 31, 2006, were identified using the prospectively gathered National Surgical Quality Improvement Program database. The primary study end points were 30-day stroke and death rates. Demographic, preoperative, and intraoperative variables were examined using multivariate models to identify variables associated with the study end points. Variables used to define systemic “high-risk” patients in the Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) study (active cardiac disease, severe chronic obstructive pulmonary disease, and octogenarian status) were examined individually and in composite fashion for association with study endpoints.

Results

Of the 3949 CEAs performed, 59% were in men, 30% were “high-risk” (19% age >80), and 43% had a previous neurologic event. The 30-day stroke rate was 1.6%, the death rate was 0.7%, and combined stroke/death rate was 2.2%. Multivariate analysis showed that intraoperative transfusion (odds ratio [OR], 5.95; 95% confidence interval [CI], 1.71-20.66; P = .005), prior major stroke (OR, 5.34; 95% CI, 2.96-9.64; P < .0001), shorter height (surrogate for small artery size; OR, 1.09; 95% CI, 1.02-1.16; P = .010), and increased anesthesia time (OR, 1.02; 95% CI, 1.00-1.03; P = .008) were predictive of stroke. Critical limb ischemia (OR, 12.72; 95% CI, 3.49-46.40; P < .0001) and poor functional status (OR, 7.05; 95% CI, 2.95-16.82; P < .0001) were independent correlates of death. Systemic high-risk variables, either combined or individually, did not increase risk of stroke or death on multivariate analysis.

Conclusion

CEA is associated with favorable 30-day outcomes across a spectrum of patient comorbidity features including octogenarian status. Anatomic and technical features are the important predictors of perioperative stroke, whereas critical limb ischemia and poor functional status are important predictors of death for patients undergoing CEA. These data refute the concept that CAS is preferred for patients deemed high-risk by virtue of systemic comorbidities.

 

Randomized controlled trials have shown that carotid endarterectomy (CEA) significantly reduces the long-term risk of subsequent stroke from carotid artery stenosis for both symptomatic1, 2, 3 and asymptomatic4, 5, 6 patients. Recent advances in endovascular therapies, however, have led some to propose carotid angioplasty and stenting (CAS) as appropriate alternative treatment for carotid artery stenosis.

In 2004 the United States Food and Drug Administration (FDA) approved the use of CAS for treatment of carotid stenosis for so-called high-risk surgical patients. The FDA's decision was based primarily on the Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) trial,7, 8 a multicenter, randomized controlled trial designed to prove the noninferiority of CAS vs CEA in high-risk surgical patients. The SAPPHIRE trial has been criticized for a number of methodologic and statistical problems.9, 10, 11, 12 One such criticism is the high 30-day stroke/death rate among patients undergoing CEA (5.4%) even though >70% of the patients included in the study had asymptomatic disease. Previous consensus documents indicate that for patients with asymptomatic carotid stenosis to realize benefit after CEA, the perioperative stroke/death rate should be <3%.13

Recent data from statewide and national registries as well as large single-center series have shown that perioperative complication rates after surgical treatment of carotid stenosis14, 15, 16, 17, 18, 19, 20, 21 are in fact, quite low. However, many of these studies have been criticized for lack of 30-day adjudication by an independent examiner. In addition, many such studies used discharge data from administrative databases and therefore may have underestimated the true perioperative complication rate after CEA.

In 2004 the American College of Surgeons (ACS) started its own arm of the National Surgical Quality Improvement Program (NSQIP) and offered it to any interested private sector hospitals, such that >120 academic and community medical centers were participating in the program by 2006. This presented us with a unique opportunity to examine 30-day adjudicated outcomes in a contemporary series of CEAs using NSQIP methodology, which has previously been validated.22, 23 We also sought to assess the validity of criteria used to define high-risk surgical patients in studies such as the SAPPHIRE trial.

Back to Article Outline

Methods 

Database 

Approval to use the ACS NSQIP database was obtained from both the ACS NSQIP and the Massachusetts General Hospital Institutional Review Board. A detailed description of the NSQIP study methods has been previously published and validated.22 Since its inception in 1994, the program's success within the Veterans Affairs (VA) hospitals has led to its subsequent extension into and validation within the private sector hospitals.23 By October 2004, the ACS had started its own arm of the NSQIP and offered the program to all interested private hospitals, including both academic and community centers. In 2005, 37 institutions participated in the program, and the number had increased to 121 by 2006.

For each participating hospital, patients undergoing major surgical procedures are identified. A sample of patients is then selected using the 8-day cycle to prevent bias in choosing cases.22 Preoperative risk factors, intraoperative variables, and 30-day postoperative mortality and morbidity outcomes are collected, validated, and submitted by a trained and audited surgical clinical nurse–reviewer designated by the ACS. The database is maintained by third-party organizations contracted by the ACS.

Patient selection 

The ACS NSQIP database was queried to identify patients undergoing CEA between 2005 and 2006 using the Current Procedural Terminology (CPT) codes 35301 and 35390. Cases were selected in which CEA was the primary procedure. Patients undergoing other major concurrent procedures, including cardiac surgeries and carotid artery angiograms, were excluded. Indication for undergoing CEA (ie symptomatic vs asymptomatic and percent stenosis) is not available in the database. In particular, although history of previous neurologic event is available, neither the laterality nor the timing of such event can be obtained from the database and thus we were not able to determine which patients were symptomatic at the time of CEA.

Definitions and end points 

Definitions of preoperative variables included in the NSQIP database are listed in the Appendix. We identified a subgroup of systemic “high-risk” patients using criteria similar to that defined by the SAPPHIRE study.7 The “high-risk” group included those aged >80 years, those with significant cardiac disease (active congestive heart failure, myocardial infarction [MI] ≤6 months before CEA, or angina ≤1 month before CEA), and those with severe chronic obstructive pulmonary disease. Various anatomic high-risk factors used in the SAPPHIRE study (eg, contralateral carotid occlusion, contralateral laryngeal-nerve palsy, previous radical neck surgery/radiation therapy to the neck) are not available in the NSQIP database and therefore were not addressed in this study.

The principal study end points were 30-day stroke and death. Thirty-day stroke is defined as “patient develops an embolic, thrombotic, or hemorrhagic vascular accident or stroke with motor, sensory, or cognitive dysfunction (eg hemiplegia, hemiparesis, aphasia, sensory deficit, impaired memory) that persists for 24 or more hours.” The assessment of these outcomes were made by the trained clinical nurse–reviewer using a variety of methods, including medical record reviews, and if necessary, letters or phone calls to patients.

Statistical analysis 

Statistical analysis was performed using SPSS 15.0 software (SPSS Inc, Chicago, Ill). The Fisher exact test and χ2 analysis were used to compare categoric variables, and the t test was used for continuous variables. Univariate analysis was performed between the 30-day outcome vs preoperative and intraoperative variables. Variables included as “high-risk” were assessed combined or individually in univariate analysis. The significant variables (P < .05) were then considered in a multivariate logistic regression model to identify independent predictors of stroke or death. In cases where preoperative and intraoperative variables might duplicate each other in assessing risks, only the one with the highest significance was entered into the multivariate model. Odds ratios were reported with 95% confidence intervals. The additive effect of “high-risk” variables was assessed using linear by linear association test.

Back to Article Outline

Results 

During the study period, 3949 isolated CEAs were performed, consisting of 872 CEAs in 2005 and 3077 CEAs in 2006, with no significant difference in stroke/death rate by year of treatment. Demographic and clinical features as well as intraoperative variables of the study group are compiled in Table I. The criteria for “high risk” were met by 30% of the patients, with 19% being those aged >80 years. A previous neurologic event had occurred in 43%, and 6.7% were hemiplegic. The 30-day stroke rate was 1.6% (62 of 3949), the death rate was 0.7% (26 of 3049), and the combined stroke/death rate was 2.2% (85 of 3949). Only three patients had a fatal stroke. The rates for stroke, death, or combined stroke/death were not significantly different between octogenarians and non-octogenarians, with stroke at 1.5% vs 1.6% (P = .85), death at 0.9% vs 0.6% (P = .28), and stroke/death at 2.3% vs 2.1% (P = .75). Among patients without any previous neurologic events, the 30-day stroke rate was 1.3%, the death rate was 0.5%, and the combined stroke/death rate was 1.7%. The average total length of stay was 2.7 ± 4.2 days, and 39 of 62 strokes (62.9%) and 12 of 26 deaths (46.2%) were observed before discharge from the initial hospitalization.

Table I. Democratic, clinical, and intraoperative features
Characteristic% (No)Mean ± SD [range]
Age, y70.4±9.5(21-89)⁎
Male59.0(2330/3949)
“High-risk”29.7(1174/3949)
Age >80 y18.7(737/3949)
Significant cardiac disease5.0(198/3949)
Active congestive heart failure1.3(53/3949)
Recent myocardial infarction1.5(60/3949)
Recent angina2.7(107/3949)
Severe COPD8.9(351/3949)
Previous PTCA17.5(691/3949)
Previous cardiac surgery24.7(977/3949)
Hypertension85.6(3380/3949)
Chronic renal insufficiency⁎⁎16.2(600/3708)
Dialysis dependent0.9(35/3949)
IDDM8.8(347/3949)
Peripheral vascular disease10.0(394/3949)
Critical limb ischemia1.1(42/3949)
Previous neurologic event43.2(1704/3949)
Hemiplegia6.7(263/3949)
Cerebrovascular accident
With deficit15.1(595/3949)
Without deficit7.6(302/3949)
Transient ischemic attack38.5(1127/3949)
Current smoker37.3(1078/3949)
Dependent functional status⁎⁎⁎6.3(247/3949)
Previous CEA0.9(34/3949)
Total anesthesia time(min)187.3±61.5[44-1592]
Total length of stay(days)2.7±4.2[0-92]

CEA, Carotid endarterectomy; COPD, chronic obstructive pulmonary disease; IDDM, insulin dependent diabetes mellitus; PTCA, percutaneous transluminal coronary angioplasty; PVD, peripheral vascular disease.

⁎Patients aged >90 years were coded as 90+ and were excluded from average age calculation.

⁎⁎Defined as serum creatinine ≥ 1.5 mg/dL.

⁎⁎⁎Dependent function status is defined as needing assistance to perform activities of daily living and does not include an individual who functions independently but uses prosthesis, equipment, or devices.

Univariate analysis for stroke is summarized in Table II. Although history of major neurologic event, defined as hemiplegia and cerebral vascular accident (CVA) with residual deficit, was associated with an increased risk of postoperative stroke, history of a minor neurologic event, such as transient ischemic attack (TIA) and CVA with no residual deficit, was not. Variables included in “high-risk,” either combined or individually, did not affect the rate of stroke after CEA.

Table II. Univariate analysis of 30-day stroke in various subgroups undergoing carotid endarterectomy
Variable30-Day stroke, % (No) or mean ± SD (range)aPb value
With variableWithout variable
“High-Risk”1.4(16/1174)1.7(46/2775).50
Age ≥80 y1.5(11/737)1.6(51/3212).85
Significant cardiac disease
Active CHF0(0/53)1.6(62/3896)>.99
Recent MI1.7(1/60)1.6(61/3889).62
Recent angina1.9(2/107)1.6(60/3842).69
Severe COPD1.1(4/351)1.6(58/3598).65
Female gender2.0(32/1619)1.3(30/2330).087
Height, (inches)a65.1±4.4[56-74]66.4±4.1[48-92].013
Previous PTCA2.2(15/691)1.4(47/3258).16
Previous cardiac surgery1.8(18/977)1.5(44/2972).43
Hypertension1.7(57/3380)0.9(5/569).20
CRIc0.8(5/600)1.7(52/3109).15
Dialysis dependent2.2(1/45)1.6(61/3904).51
IDDM2.3(8/347)1.5(54/3602).25
PVD1.0(4/394)1.6(58/3555).52
CLI0(0/42)1.6(62/3907)>.99
Previous neurologic event
Hemiplegia6.1(16/263)1.2(46/3686)<.001
CVA with deficit2.7(16/595)1.4(46/3354).017
CVA without deficit1.7(5/302)1.6(57/3647).81
TIA2.0(23/1127)1.4(39/2822).13
Current smoker1.9(21/1078)1.4(41/2871).24
Dependent functional status2.4(6/247)1.5(56/3702).26
Previous CEA0(0/34)1.6(62/3915)>.99
Intraoperative transfusion8.6(3/35)1.5(59/3914).017
Total anesthesia time, mina214.5±188.8[71-1558]186.9±57.2[44-1592]<.001

CEA, Carotid endarterectomy; CHF, congestive heart failure; CLI, critical limb ischemia; CVA, cerebral vascular accident; COPD, chronic obstructive pulmonary disease; IDDM, insulin-dependent diabetes mellitus; MI, myocardial infarction; PTCA, percutaneous coronary angioplasty; PVD, peripheral vascular disease; TIA, transient ischemic attack.

aFor continuous variables the “With variable” column denotes average value among those who had postoperative stroke and the “Without variable” column among those who did not have postoperative stroke.

bP < .05 indicates significance.

cChronic renal insufficiency defined as serum creatinine >1.5 mg/dL.

Candidate variables from the univariate analysis of stroke were selected and used to construct a multivariate model (Table III). History of CVA with neurologic deficit was not included in the model because of overlap with hemiplegia. Multivariate analysis showed need for intraoperative transfusion, previous history of hemiplegia, shorter height (surrogate for small artery size), and increased anesthesia time to be predictive of stroke.

Table III. Independent risk factors associated with 30-day stroke and death using logistic regression
VariableOR95% CIP
Stroke
Intraoperative transfusion5.951.71-20.66.005
Hemiplegia5.342.96-9.64<.0001
Shorter height1.091.02-1.16.010
Increased anesthesia time1.021.00-1.03.008
Death
Critical limb ischemia12.723.49-46.40<.0001
Poor functional status7.052.95-16.82<.0001

CI, Confidence interval; OR, odds ratio.

Univariate analysis for variables associated with perioperative mortality is reported in Table IV. A larger number of systemic comorbidities were significantly associated with death compared with stroke. The “high-risk” group, for example, had higher rate of perioperative mortality after CEA. However, when individual variables included in the “high-risk” group were examined, only a history of major cardiac disease was associated with higher rate of postoperative death.

Table IV. Univariate analysis of 30-day mortality in various subgroups undergoing carotid endarterectomy
Variable30-Day mortality, % (No.) or mean ± SD (range)aPb value
With variableWithout variable
“High-risk”1.3(15/1174)0.4(11/2775).002
Age >80 y0.9(7/737)0.6(19/3212).28
Significant cardiac disease
Active CHF7.5(4/53)0.6(22/3896)<.001
Recent MI5.0(3/60)0.6(23/3889).007
Recent angina0.9(1/107)0.7(25/3842).51
Severe COPD1.4(5/351)0.6(21/3598).06
Female gender0.8(13/1619)0.6(13/2330).35
Height, (inches)a64.9±4.5[56-72]66.4±4.1[48-82].07
Previous PTCA0.9(6/691)0.6(20/3258).045
Previous cardiac surgery0.5(5/977)0.7(21/2972).51
Hypertension0.7(24/3380)0.4(2/569).33
CRIc1.5(9/600)0.5(16/3108).007
Dialysis dependent4.4(2/45)0.6(24/3904).035
IDDM1.4(5/347)0.6(21/3602).059
PVD1.5(6/394)0.6(20/3555).025
CLI7.1(3/42)0.6(23/3907).002
Previous neurologic event
Hemiplegia2.7(7/263)0.5(19/3686)<.001
CVA with deficit1.7(10/595)0.5(16/3354).001
CVA without deficit0.7(2/302)0.7(24/3647)>.99
TIA0.3(3/1127)0.8(23/2822).08
Current smoker0.8(9/1078)0.6(17/2871).40
Dependent functional status3.6(9/247)0.5(17/3702)<.001
Previous CEA2.9(1/34)0.6(25/3915).20
Intraoperative transfusion2.9(1/35)0.6(25/3914).21
Total anesthesia time, mina211.1±71.6[100-376]187.2±61.4[44-1592].048

CEA, Carotid endarterectomy; CHF, congestive heart failure; CLI, critical limb ischemia; CVA, cerebral vascular accident; COPD, chronic obstructive pulmonary disease; IDDM, insulin-dependent diabetes mellitus; MI, myocardial infarction; PTCA, percutaneous coronary angioplasty; PVD, peripheral vascular disease; TIA, transient ischemic attack.

aFor continuous variables “With variable” column denotes average value among those who had postoperative stroke and “Without variable” column among those who did not have postoperative stroke.

bP < .05 indicates significance.

cChronic renal insufficiency(CRI) defined as serum creatinine ≥1.5 mg/dL.

Candidate variables from the univariate analysis of perioperative mortality were selected and used to construct a multivariate model (Table III). Again, in cases where preoperative/intraoperative variables might duplicate each other in assessing risks, only the one with the highest significance was entered into the multivariate model. Because only 26 deaths occurred, only the first two variables with the highest significance—critical limb ischemia (CLI) and poor functional status—were considered to be reliable independent predictors of death. Variables included as “high-risk,” either combined or individually, did not increase the risk of death on multivariate analysis.

The additive effect of “high-risk” variables was assessed using linear by linear association test. No significant relationship was found between the number of “high-risk” variables and the stroke rate. However, a significant linear relationship existed between the number of “high-risk” variables and mortality rate, where mortality was 14.3% among those with all three of the “high-risk” variables; whereas patients with only one or two of the “high-risk” variables had mortality rates of 1.0% and 2.8%, respectively.

Back to Article Outline

Discussion 

Although the evidence base supporting CEA as the gold standard in the treatment of both symptomatic and asymptomatic carotid stenosis is sound,1, 2, 3, 4, 5, 6 CAS has been proposed as appropriate alternative treatment for carotid stenosis. A number of randomized controlled trials comparing CAS with CEA followed, with variable results, perhaps reflecting the heterogeneous patient populations studied.7, 8, 24, 25, 26, 27, 28 The FDA, however, used the findings of these trials as the basis for its approval of the use of CAS for treating both symptomatic and asymptomatic patients who are considered high-risk surgical candidates.

The FDA's decision was largely based on the SAPPHIRE trial,7, 8 a multicenter, prospective, randomized controlled trial designed to test the noninferiority of CAS vs CEA in high-risk surgical patients. A number of methodologic and statistical problems with the trial have made drawing any definitive conclusions problematic and controversial.9, 10, 11, 12 One such criticism has been the apparently excessive 30-day stroke/death rate of 5.4% among patients undergoing CEA, even though >70% of the patients were those with asymptomatic disease. This rate is much higher than the 3% rate recommended in the American Heart Association (AHA) guideline statement for those with asymptomatic disease.13 The authors of the SAPPHIRE study nonetheless concluded that CAS was not inferior to CEA in treating high-risk patients with carotid artery stenosis.

Two other studies sought to examine outcomes after CAS in treating the so-called high-risk patients. The ACCULINK for Revascularization of Carotids in High-Risk patients (ARCHeR) study defined its high-risk patients using a combination of physiologic and anatomic factors.29 It is interesting to note that more than one-third of the patients in this high-risk group were those with recurrent stenosis after CEA. In such patients the inherent risk of embolic stroke from stenting should be quite low because these lesions are not the typical ulcerated plaques seen in patients with primary carotid stenosis. Despite this, the 30-day stroke/death rate was 6.9% for the 581 study patients. The rate was 11.6% for the symptomatic patients and 5.4% for the asymptomatic patients, perioperative complication rates again far higher than those recommended by the AHA guidelines.13 Furthermore, the authors used an extremely high 1-year adverse event rate of 14.4% as the “historical control” for carotid surgery in concluding that the outcome after CAS was more favorable. This methodology is scientifically unsound, however, because many of the studies used in deriving the “historical control” event rate were those with combined CEA and coronary artery bypass grafting surgeries.

Another high-risk CAS registry study is the Boston Scientific EPI: A Carotid Stenting Trial for High-risk Surgical Patients (BEACH) trial.30 High-risk was again defined using a combination of medical comorbidities and unfavorable anatomy. Among 480 such high-risk patients, where 34% were those with restenosis after CEA, combined 30-day stroke/death/MI rate was 5.4%, with only one patient (0.2%) sustaining an MI. Stroke/death/MI rates among symptomatic and asymptomatic patients were 7.7% (0% MI) and 4.7% (0.3% or 1 patient with MI), respectively. Those with medical comorbidities had a significantly higher 30-day event rate than those with anatomic high-risk factors (14.3% vs 5.3%; P = .002), suggesting that patients with medical comorbidities do not benefit from the supposedly less invasive endovascular treatment. In addition, advanced age (≥75 years) was associated with significantly worse outcomes. Nonetheless, the authors concluded noninferiority of CAS to CEA, again based on a flawed literature control as in the ARCHeR trial.

Contemporary series of CEAs consistently demonstrate excellent perioperative outcomes. This was the case in our study, with 30-day stroke/death rate of 2.2% among 3949 CEAs performed across a spectrum of both academic and community hospitals. Results are likewise favorable from an earlier report of the NSQIP database by Stoner et al,14 where the 30-day stroke/death rate was 3.4% among 13622 CEAs performed between 2000 and 2003 (95% men, 91% VA hospitals). Similar results are reported from a variety of large population-based studies. After review of 3259 CEAs (42% symptomatic) from the Society for Vascular Surgery carotid vascular registry, Sidawy et al15 reported 30-day stroke/death/MI rate of 3.75% in symptomatic patients and 1.97% in asymptomatic patients. Timaran et al16 reported in-hospital mortality rate of 0.6% and stroke rate of 1.1% among 113,000 CEAs (7.6% symptomatic) performed in the United States in 2005 in the Nationwide Inpatient Sample (NIS) database; whereas, McPhee et al17 reported an in-hospital mortality rate of 0.9% and a stroke rate of 0.4% among 245,000 CEAs (8% symptomatic) performed in 2003 and 2004 (NIS database).17

We previously published a series of 2236 isolated CEAs (36% symptomatic) performed between 1989 and 1999 with a 30-day stroke/death rate of 1.4%.30 Furthermore, a 36% reduction in perioperative morbidity/mortality occurred in the last 5 years of the study compared with the previous 5 years (7.5% vs 4.8%; P < .006). Other studies have also demonstrated improving outcomes after CEA. Matsen et al18 reported an in-hospital stroke/death rate of 1.3% among nearly 24,000 CEAs (15.2% symptomatic) performed in Maryland between 1994 and 2003 and observed improvement in stroke rate during the study period of 2.12% in 1994, 1.47% in 1995, and 0.29% to 0.65% from 1996 to 2003. Similarly, Sheikh et al19 reported decreasing 30-day mortality after CEA among Medicare beneficiaries from 1991 to 2000: 1.95% in 1991, 1.44% in 1995, and 0.89% in 2000 (P < .001). Finally, Kragsterman et al20 reviewed CEA outcomes on all asymptomatic patients undergoing CEA in Sweden between 1994 and 2003 and reported that the stroke/death rate of 2.1% in asymptomatic patients during the entire period improved to 0.9% from 1999 to 2003 (P = .026).

Taken together, these studies prove that excellent results are achieved in contemporary practice across a broad spectrum of surgical practices, not just in tertiary care centers.

A number of recent studies also document favorable results even among the so-called high-risk patients. In our study, 30-day stroke/death rate among “high-risk” patients was 2.5% compared with 2.0% in the non-“high-risk” group (P = .371). Similarly, Flanigan et al31 compared results of 207 SAPPHIRE-eligible high-risk vs 235 normal-risk patients (39% symptomatic) and saw no difference in 30-day outcomes. In a retrospective review of 776 CEAs performed at the Mayo Clinic between 1998 and 2002 (27% symptomatic; 42% SAPPHIRE-eligible), Mozes et al32 reported no significant difference in perioperative stroke/death rates between the high- and low-risk patients (2.5% vs 1.1%). Boules et al33 defined high-risk using similar criteria consisting of medical or anatomic factors, or both.33 After reviewing 499 CEAs from 1996 to 2001 (39% symptomatic; 17% high-risk), the authors found no difference in 30-day poor outcome (stroke/TIA/death) between high- and low-risk patients (4.8% vs 4.1%; P = .77). Others also reported no significant difference in perioperative outcomes when high-risk was defined as North American Symptomatic Carotid Endarterectomy Trial (NASCET)/Asymptomatic Carotid Atherosclerosis Study (ACAS) ineligible or ARCHeR eligible.34, 35

Advanced age is often included in high-risk criteria. However, a review of contemporary results for CEAs performed among 2564 octogenarians revealed a 30-day stroke/death rate of 3.5%.36 This was recapitulated in our study, where the 30-day stroke/death rate among patients aged >80 years was 2.3% compared with 2.1% among those aged <80 years (P = .749). In contrast, evidence suggests that the risk among octogenarians after CAS is much higher. The Carotid Revascularization Endarterectomy vs Stenting Trial (CREST) is an ongoing, multicenter, randomized controlled trial comparing CAS vs CEA. In their lead-in phase report, Hobson et al28 examined outcomes after 749 CAS procedures (31% symptomatic) and reported that the 30-day stroke/death rate was significantly higher among octogenarians (12.1% vs 3.2%; P < .0001), mainly secondary to a significantly higher 30-day stroke rate (12.1% vs 2.8%; P < .0001).28 Others have also reported increased adverse periprocedural outcomes among older patients undergoing CAS and suggested that increasing age may be a surrogate for unfavorable anatomy.37, 38, 39 The bulk of evidence establishes CEA as the safer carotid intervention in octogenarians.

In our study, independent risk factors associated with stroke were the need for intraoperative transfusion, prior major stroke, shorter height, and increased anesthesia time. Need for intraoperative transfusion and increased anesthesia time indicate technical difficulties during the procedure. To assess whether increased anesthesia time was due to patient comorbidities, analysis was performed after adjusting for American Society of Anesthesiologists (ASA) class; longer anesthesia time was still associated with increased risk for stroke. Shorter height, a surrogate for smaller artery size, also increased the risk of stroke. Previous studies have shown that height affects carotid artery size, and that shorter height is associated with increased risk after CEA.40, 41 The perioperative risk after CEA has been reported to be higher in women than in men, and the benefit of CEA for women has been questioned.42, 43, 44 This difference in perioperative risk may have been at least partly secondary to the difference in body size between men and women, where both the artery diameter and surgical exposure can be related to the patient's height. In our study, men were significantly taller than women and there was a trend toward increased risk for stroke in women (P = .087). After adjusting for height, this trend was no longer observed.

Risk factors associated with perioperative mortality were poor functional status and CLI, a surrogate for diffuse advanced atherosclerotic disease. Functional status is a unique variable available in the NSQIP database, where a patient's preoperative functional status is clearly defined and categorized (Appendix). These two variables point to a patient population with advanced systemic illness that is at increased risk of death after CEA; a logical conclusion would be that carotid intervention for asymptomatic disease is inappropriate in such patients. This is supported by the significant linear relationship between the number of high-risk variables and perioperative death, such that among patients with all three high-risk systemic variables, 30-day mortality was 14.3%, although only seven patients met these criteria of having all three variables. This relationship was not present for stroke.

Only three patients had a fatal stroke after CEA, suggesting that the patient groups at risk of stroke vs death after CEA were quite different. Although lacking definitive anatomic data such as history of previous neck irradiation or radical neck dissection, our data suggest that variables associated with increased risk of stroke may be related to anatomic and technical features, whereas variables associated with increased risk of death are related to serious systemic comorbidities.

Systemic comorbidities did not increase risk of stroke after CEA. Previous studies examining outcomes after CAS among the so-called high-risk surgical patients (SAPPHIRE, ARCHeR, and BEACH trials), defined such patients using both anatomic and physiologic criteria. Although there may be technical advantages of CAS over CEA among patients with “anatomic” risk factors, it is unknown if CAS will result in better or even equal outcomes for those with advanced systemic illness. Such patients, if asymptomatic, are likely better served by conservative medical management rather than undergoing any intervention at all; thus far, no study has adequately addressed this issue. Because much of the literature on CAS defines high-risk patients using both anatomic and systemic criteria, it is unclear if the noninferiority of CAS to CEA was secondary to anatomic risk factors alone.7, 8, 29, 30 It will be important to distinguish between the two groups and assess outcomes separately in future studies.

Limitations of this study are those of the NSQIP and include inherent problems of analysis using a large database, such as imprecise coding. Information on the degree of carotid artery stenosis, timing, and laterality of previous neurologic event, and various other anatomic information including history of previous neck irradiation and contralateral carotid occlusion are not available and thus limit the interpretation of the study findings.

Although 30% of the patients in the study met the criteria for “high-risk,” only a small proportion of patients met the criteria for active cardiac disease, which may represent surgical selection bias. However, it is also important to note that the inclusion criteria for the SAPPHIRE trial was not a rigorous one in that patients only needed to have one of the “high-risk” variables to be considered eligible for the trial.

One of the main advantages of the NSQIP database is that it offers independent adjudication of 30-day outcomes by a nonpartial party, namely by a trained clinical nurse–reviewer. This methodology has been previously validated.22, 23 We acknowledge that this is different from examination by an independent neurologist, but neurologic outcomes affecting activities of daily living will not be missed. In addition, our study reports true 30-day outcomes compared with studies based on state and nation-wide registries that rely on discharge data and may therefore underestimate true perioperative complication rates after CEA. In fact, in our study only 39 of 62 strokes (62.9%) and 12 of 26 deaths (46.2%) were observed before discharge from the initial hospitalization, indicating the valid capture of postdischarge events.

Back to Article Outline

Conclusions 

CEA is associated with favorable 30-day outcomes across a spectrum of patient comorbidity features, including octogenarian status. These data indicate that anatomic and intraoperative technical features are important predictors of perioperative stroke. Advanced systemic illness, in the form of CLI and poor functional status, is an important predictor of death for patients undergoing CEA, representing circumstances wherein a noninterventional posture should be undertaken for those with asymptomatic disease. While CAS may have a potential role in treating patients with certain anatomic features, our data refute the concept that CAS is preferred for patients deemed high-risk by virtue of systemic comorbidities.

Back to Article Outline

Author contributions 


Conception and design: JK, GL, MC, RC

Analysis and interpretation: JK, TC, GL, MC, RC

Data collection: JK, TC, RL

Writing the article: JK, RC

Critical revision of the article: JK, TC, RL, GL, MC, RC

Final approval of the article: JK, TC, RL, GL, MC, RC

Statistical analysis: JK, TC

Obtained funding: RC

Overall responsibility: JK

Back to Article Outline

Appendix 

Appendix. Definition of preoperative variables
VariableDefinition
CHFNewly diagnosed CHF within the previous 30 days or a diagnosis of chronic CHF with new signs or symptoms in the 30 days before surgery.
MIHistory of a non-Q wave or a Q wave infarct ≤6 months before surgery.
AnginaHistory of angina ≤30 days before surgery.
COPDCOPD resulting in any one or more of the following:
•Functional disability from COPD (eg dyspnea, inability to perform ADLs)

•Hospitalization in the past for treatment of COPD

•Requires chronic bronchodilator therapy with oral or inhaled agents

•An FEV1 of <75% of predicted on pulmonary function testing

Previous PTCAPrevious coronary intervention, including any attempted interventions. Includes balloon dilatation or stent placement, but not valvuloplasties.
Previous cardiac surgeryAny major cardiac surgical procedure performed either as an “off-pump” repair or using CPB. Includes CABG, valve replacement or repair, repair of ASD or VSD defects, great thoracic vessel repair, cardiac transplant, left ventricular aneurysmectomy, insertion of LVAD, etc. Does not include pacemaker insertions or automatic implantable cardioverter defibrillator insertions.
HypertensionPersistent elevation of SBP >140 mm Hg or a DBP >90 mm Hg, or requires an antihypertensive treatment at the time the patient is being considered as a candidate for surgery (which should be ≤30 days before surgery).
CRISerum creatinine <1.5 mg/dL.
Dialysis dependentAcute or chronic renal failure requiring treatment with peritoneal dialysis, hemodialysis, hemofiltration, hemodiafiltration, or ultrafiltration ≤2 weeks before surgery.
IDDMDiabetes requiring daily insulin therapy.
PVDAny type of angioplasty or revascularization procedure for atherosclerotic peripheral vascular disease or a patient who has had any type of amputation procedure for peripheral vascular disease.
CLIRest pain or gangrene at the time of surgery.
Previous neurologic event
HemiplegiaTotal or partial paralysis or paresis of one side of the body (not a single limb).
CVA with deficitEmbolic, thrombotic, or hemorrhagic CVA with persistent residual motor, sensory, or cognitive dysfunction (eg, hemiplegia, hemiparesis, aphasia, sensory deficit, impaired memory).
CVA without deficitEmbolic, thrombotic, or hemorrhagic CVA with neurologic deficit(s) lasting at ≥30 minutes, but no current residual neurologic dysfunction or deficit.
TIAFocal neurologic deficits of sudden onset and brief duration. These attacks may be recurrent.
Dependent functional statusRequires some or total assistance from another person for ADLs at the time the patient is being considered for surgery (no longer than 30 days before surgery). This does not include a person who is able to function independently with prosthetics, equipment, or devices. ADLs include: bathing, feeding, dressing, toileting, and mobility.

ADL, Activities of daily living; ASD, atrial septal defect; CABG, coronary artery bypass grafting; CHF, congestive heart failure; CLI, critical limb ischemia; COPD, chronic obstructive pulmonary disease; CPB, cardiopulmonary bypass; CRI, chronic renal insufficiency; CRI, chronic renal insufficiency; CVA, cerebral vascular accident; CVA, cerebrovascular accident; DBP, diastolic blood pressure; FEV1, forced expiratory volume in 1 second; IDDM, insulin-dependent diabetes mellitus; LVAD, left ventricular assist device; MI, myocardial infarction; PTCA, percutaneous coronary angioplasty; PVD, peripheral vascular disease; PVD, peripheral vascular disease; SBP, systolic blood pressure; TIA, transient ischemic attack; VSD, ventricular septal defect.

Back to Article Outline

References 

  1. North American Symptomatic Carotid Endarterectomy Trial Collatorators. Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med. 1991;325:445–453
  2. Barnett HJ, Taylor W, Eliasziw M, Fox AJ, Ferguson GG, Haynes RB, et al. North American Symptomatic Carotid Endarterectomy Trial Collatorators Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. N Engl J Med. 1998;339:1415–1425
  3. European Carotid Surgery Trialists' Collaborative Group. Randomised trial of endarterectomy for recently symptomatic carotid stenosis: final results of the MRC European Carotid Surgery Trial (ECST). Lancet. 1998;351:1379–1387
  4. Executive Committee for the Asymptomatic Carotid Atherosclerosis Study. Endarterectomy for asymptomatic carotid artery stenosis. JAMA. 1995;273:1421–1428
  5. Halliday A, Mansfield A, Marro J, Peto C, Peto R, Potter J, et al. MRC Asymptomatic Carotid Surgery Trial Collaborative Group Prevention of disabling and fatal strokes by successful carotid endarterectomy in patients without recent neurological symptoms: randomized controlled trial. Lancet. 2004;363:1491–1502
  6. Hobson RW, Weiss DG, Fields WS, Goldstone J, Moore WS, Towne JB, et al. Veterans Affairs Cooperative Study Group Efficacy of carotid endarterectomy for asymptomatic carotid stenosis. N Engl J Med. 1993;328:221–227
  7. Yadav JS, Wholey MH, Kuntz RE, Fayad P, Katzen BT, Mishkel GJ, et al. Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy investigators (Protected carotid-artery stenting versus endarterectomy in high-risk patients). N Engl J Med. 2004;351:1493–1501
  8. Gurm HS, Yadav JS, Fayad P, Katzen BT, Mishkel GJ, Bajwa TK, et al. SAPPHIRE investigators Long-term results of carotid stenting versus endarterectomy in high-risk patients. N Engl J Med. 2008;358:1572–1579
  9. Cambria RP. Stenting for carotid-artery stenosis. N Engl J Med. 2004;351:1565–1567
  10. Thomas DJ. Protected carotid artery stenting versus endarterectomy in high-risk patients: reflections from SAPPHIRE. Stroke. 2005;36:912–913
  11. LoGerfo FW. Carotid stents: unleashed, unproven. Circulation. 2007;116:1596–1601
  12. Samuelson RM, Yamamoto J, Levy EI, Siddiqui AH, Hopkins LN. The argument to support broader application of extracranial carotid artery stent technology. Circulation. 2007;116:1602–1610
  13. 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. Circulation. 1998;97:501–509
  14. Stoner MC, Abbott WM, Wong DR, Hua HT, LaMuraglia GM, Kwolek CJ, et al. Defining the high-risk patient for carotid endarterectomy: an analysis of the prospective National Surgical Quality Improvement Program database. J Vasc Surg. 2006;43:285–296
  15. Sidawy AN, Zwolak RM, White RA, Siami FS, Schermerhorn ML, Sicard GA. SVS Carotid Vascular Registry: CAS vs CEA outcomes comparison. 2008;Presented at the Annual Meeting of the SVS, San Diego, Calif, June 4-8,
  16. Timaran CH, Rosero EB, Veith FJ, Clagett JG, Valentine RJ, Smith ST, et al. Intracranial hemorrhage after carotid endarterectomy and carotid artery stenting in the United States in 2005. 2008;Presented at the Annual Meeting of the SVS, San Diego, Calif, June 4-8,
  17. McPhee JT, Hill JS, Ciocca RG, Messina LM, Eslami MH. Carotid endarterectomy was performed with lower stroke and death rates than carotid artery stenting in the United States in 2003 and 2004. J Vasc Surg. 2007;46:1112–1118
  18. Matsen SL, Chang DC, Perler BA, Roseborough GS, Williams GM. Trends in the in-hospital stroke rate following carotid endarterectomy in California and Maryland. J Vasc Surg. 2006;44:488–495
  19. Sheikh K, Bullock C. Variation and changes in state-specific carotid endarterectomy and 30-day mortality rates, United States, 1991-2000. J Vasc Surg. 2003;38:779–784
  20. Kragsterman B, Parsson H, Lindback J, Berggvist D, Bjorck M. Outcomes of carotid endarterectomy for asymptomatic stenosis in Sweden are improving: Results from a population-based registry. J Vasc Surg. 2006;44:79–85
  21. 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;240:535–546
  22. Khuri SF, Daley J, Henderson W, Barbour GJ, Lowry P, Irvin G, et al. National Veterans Administration Surgical Risk Study (The National Veterans Administration Surgical Risk Study: risk adjustment for the comparative assessment of the quality of surgical care). J Am Coll Surg. 1995;180:519–531
  23. Fink AS, Campbell DA, Mentzer RM, Henderson WG, Daley J, Bannister J, et al. The National Surgical Quality Improvement Program in non-veterans administration hospitals: initial demonstration of feasibility. Ann Surg. 2002;236:344–353
  24. CAVATAS investigators. Endovascular versus surgical treatment in patients with carotid stenosis in the carotid and vertebral artery transluminal angioplasty study (CAVATAS): a radomised trial. Lancet. 2001;357:1729–1737
  25. Alberts MJ. Results of a multicenter prospective randomized trial of carotid artery stenting vs carotid endarterectomy [abstract]. Stroke. 2001;32:325;-d
  26. Ringleb PA, Allenberg J, Bruckmann H, Eckstein HH, Fraedrich G, Hartmann M, et al. SPACE collaborative group 30 day results from the SPACE trial of stent-protected angioplasty versus carotid endarterectomy in symptomatic patients: a randomised non-inferiority trial. Lancet. 2006;368:1239–1247
  27. Mas JL, Chatellier G, Beyssen B, Branchereau A, Moulin T, Becquemin JP, et al. EVA-3S investigators Endarterectomy versus stenting in patients with symptomatic severe carotid stenosis. N Engl J Med. 2006;355:1660–1671
  28. Hobson RW, Howard VJ, Roubin GS, Brott TG, Ferguson RD, Popma JJ, et al. CREST Investigators Carotid artery stenting is associated with increased complications in octogenarians: 30-day stroke and death rates in the CREST lead-in phase. J Vasc Surg. 2004;40:1106–1111
  29. Gray WA, Hopkins LN, Yadav S, Davis T, Wholey M, Atkinson R, et al. ARCHeR trial collaborators Protected carotid stenting in high-surgical-risk patients: the ARCHeR results. J Vasc Surg. 2006;44:258–269
  30. Iyer SS, White CJ, Hopkins LN, Katzen BT, Safian R, Wholey MH, et al. BEACH investigators Carotid artery revascularization in high-surgical-risk patients using the Carotid WALLSTENT and FilterWire EX/EZ: 1-year outcomes in the BEACH Pivotal Group. J Am Coll Cardiol. 2008;51:427–434
  31. Flanigan DP, Flanigan ME, Dorne AL, Harward TRS, Razavi MK, Ballard JL. Long-term results of 442 consecutive, standardized carotid endarterectomy procedures in standard-risk and high-risk patients. J Vasc Surg. 2007;46:876–882
  32. Mozes G, Sullivan TM, Torres-Russotto DR, Bower TC, Hoskin TL, Sampaio SM, et al. Carotid endarterectomy in SAPPHIRE-eligible high-risk patients: Implications for selecting patients for carotid angioplasty and stenting. J Vasc Surg. 2004;39:958–966
  33. Boules TN, Proctor MC, Aref A, Upchurch GR, Stanley JC, Henke PK. Carotid endarterectomy remains the standard of care, even in high-risk surgical patients. Ann Surg. 2005;241:356–363
  34. Illig KA, Zhang R, Tanski W, Benesch C, Sternbach Y, Green RM. Is the rationale for carotid angioplasty and stenting in patients excluded from NASCET/ACAS or eligible for ARCHeR justified?. J Vasc Surg. 2003;37:575–581
  35. Lepore MR, Sternbergh WC, Salartash K, Tonnessen B, Money SR. Influence of NASCET/ACAS trial eligibility on outcome after carotid endarterectomy. J Vasc Surg. 2001;34:581–586
  36. Miller MT, Comerota AJ, Tzilinis A, Daoud Y, Hammerling J. Carotid endarterectomy in octogenarians: Does increased age indicate “high risk?”. J Vasc Surg. 2005;41:231–237
  37. Sayeed S, Stanziale SF, Wholey MH, Makaroun MS. Angiographic lesion characteristics can predict adverse outcomes after carotid artery stenting. J Vasc Surg. 2008;47:81–87
  38. Lam RC, Lin SC, DeRubertis B, Hynecek R, Kent KC, Faries PL. The impact of increasing age on anatomic factors affectin carotid angioplasty and stenting. J Vasc Surg. 2007;45:875–880
  39. Stanziale SF, Marone LK, Boules TN, Brimmeier JA, Hill K, Makaroun MS, et al. Carotid artery stenting in octogenarians is associated with increased adverse outcomes. J Vasc Surg. 2006;43:297–304
  40. Krejza J, Arkuszewski M, Kasner SE, Weigele J, Ustymowicz A, Hurst RW, et al. Carotid artery diameter in men and women and the relation to body and neck size. Stroke. 2006;37:1103–1105
  41. Messe SR, Kasner SE, Mehta Z, Warlow CP, Rothwell PM European Carotid Surgery Trialists. Effect of body size on operative risk of carotid endarterectomy. J Neurol Neurosurg Psychiatry. 2004;75:1759–1761
  42. Rothwell PM, Eliasziw M, Gutnikov SA, Warlow CP, Barnett HJ Carotid Endarterectomy Trialists Collaboration. Endarterectomy for symptomatic carotid stenosis in relation to clinical subgroups and timing of surgery. Lancet. 2004;363:915–924
  43. Alamowitch S, Eliasziw M, Barnett HJ. The risk and benefit of endarterectomy in women with symptomatic internal carotid artery disease. Stroke. 2005;36:27–31
  44. Rothwell PM, Goldstein LB. Carotid endarterectomy for asymptomatic carotid stenosis: asymptomatic carotid surgery trial. Stroke. 2004;35:2425–2427

 Competition of interest: none.

PII: S0741-5214(08)01602-9

doi:10.1016/j.jvs.2008.09.018

Journal of Vascular Surgery
Volume 49, Issue 2 , Pages 331-339.e1, February 2009

Article Tools