Abstract
Objective
The objective of this study was to determine the effect of presenting symptom types on 30-day periprocedural outcomes of carotid endarterectomy (CEA) and carotid artery stenting (CAS) in contemporary vascular practice.
Methods
Retrospective review was undertaken of the Society for Vascular Surgery Vascular Registry database subjects who underwent CEA or CAS from 2004 to 2011. Patients were grouped by discrete 12-month preprocedural ipsilateral symptom type: stroke, transient ischemic attack (TIA), transient monocular blindness (TMB), or asymptomatic (ASX). Risk-adjusted odds ratios (ORs) were used to compare the likelihood of the 30-day outcomes of death, stroke, and myocardial infarction (MI) and the composite outcomes of death + stroke and death + stroke + MI.
Results
Symptom type significantly influences risk-adjusted 30-day outcomes for carotid intervention. Presentation with stroke predicted the poorest outcomes (death + stroke + MI composite: OR, 1.3; 95% confidence interval [CI], 0.83–2.03 vs TIA; OR, 2.56; 95% CI, 1.18–5.57 vs TMB; OR, 2.12; 95% CI, 1.46–3.08 vs ASX), followed by TIA (death + stroke + MI composite: OR, 1.97; 95% CI, 0.91–4.25 vs TMB; OR, 1.63; 95% CI, 1.14–2.33 vs ASX). For both CAS and CEA patients, presentation with stroke or TIA predicted a higher risk of periprocedural stroke than in ASX patients. Presentation with stroke predicted higher 30-day risk of death with CAS but not with CEA. MI rates were not affected by presenting symptom type. The 30-day outcomes for the TMB and ASX patient groups were equivalent in both treatment arms.
Conclusions
Presenting symptom type significantly affects the 30-day outcomes of both CAS and CEA in contemporary vascular surgical practice. Presentation with stroke and TIA predicts higher rates of periprocedural complications, whereas TMB presentation predicts a periprocedural risk profile similar to that of ASX disease.
In addition to prior completed ipsilateral stroke, hemispheric transient ischemic attack (TIA) ipsilateral to significant carotid bifurcation stenosis has long been known to predict subsequent ipsilateral stroke and excess cardiovascular mortality.1,2 In similar fashion, transient monocular blindness (TMB, also known as amaurosis fugax) associated with carotid bifurcation stenosis foretells an elevated risk of subsequent stroke, although less than that described for TIA.3
The North American Symptomatic Carotid Endarterectomy Trial (NASCET) firmly established the benefit of carotid endarterectomy (CEA) for symptomatic moderate to severe carotid stenoses.4,5 NASCET also added to our knowledge of the natural history of symptomatic carotid disease; analysis of the medical treatment arm of NASCET demonstrated a higher 2-year risk of stroke for patients presenting with hemispheric TIA (43.5% ± 6.7%) in comparison to TMB (16.6% ± 5.6%).6 Separate examination of surgical results from NASCET showed that procedural stroke outcomes were poorer for patients presenting with hemispheric TIA rather than TMB,7 confirming the findings of earlier investigators.8
Yet even as NASCET and the Asymptomatic Carotid Atherosclerosis Study9 established the primacy of CEA for stroke reduction in symptomatic and asymptomatic lesions, early experiences with angioplasty and stent placement for carotid disease were being reported.10,11 During the next decade, carotid artery stenting (CAS) was compared with CEA in randomized trials ranging from the Stenting and Angioplasty with Protection in Patients with High Risk for Endarterectomy (SAPPHIRE) study to the more recent and better powered Carotid Revascularization Endarterectomy vs Stenting Trial (CREST).12,13 Both trials enrolled asymptomatic and symptomatic patients, and their publication has provided further insight into the comparative benefits of CAS and CEA. However, neither study examined the relationship of presenting symptom type (stroke, TIA, or TMB) to procedural outcomes.
The Society for Vascular Surgery Vascular Registry (SVS-VR) carotid module collected demographic, procedural, and outcomes data from contributing centers for CEA and CAS from 2004 through 2011. By the nature of registry design, patients entered into the SVS-VR are unmatched, yet risk-adjusted data from this “real-world” experience provide valuable insight into current vascular surgical outcomes.14 Using the SVS-VR, we sought to determine the effect of presenting symptom type on early outcomes of CEA and CAS in contemporary vascular practice.
METHODS
The derivation of 30-day periprocedural outcomes data from the SVS-VR, inclusive of procedural and predischarge data, has previously been reported.14 All registry patients who underwent CEA or CAS with available 30-day outcomes reporting were identified. For clarity of comparison, carotid procedures undertaken for atherosclerotic, radiation-induced, or restenotic lesions of the carotid bifurcation and internal carotid artery were included, but procedures undertaken for trauma, dissection, or unspecified causes were excluded. Procedures undertaken only on the common carotid or external carotid arteries were excluded. CEA and CAS patients were grouped by discrete preprocedural ipsilateral symptom type occurring within the 12 months before intervention: stroke, TIA, TMB, or asymptomatic (ASX). Patients reporting more than one symptom (eg, TIA and stroke) were excluded from analysis. Risk-adjusted odds ratios (ORs) were used to compare the likelihood of the 30-day outcomes of death, stroke, and myocardial infarction (MI) and the composite outcomes of death + stroke and death + stroke + MI.
Statistical methods
Descriptive statistical comparisons were conducted with χ2 tests for categorical variables and analysis of variance for continuous variables. Descriptive statistics are listed as mean ± standard deviation for continuous variables and percentage (frequency) for categorical variables. Outcomes analyses comparing across symptom groups were conducted in subsets of the cohort with the Fisher exact test for discrete/categorical data. Adjusted ORs found through multivariable logistic regression were used to compare the selected outcomes measures between the symptom-defined groups. Adjusted ORs for the multiple symptom group comparisons were adjusted for significant baseline factors that were retained after applying backwards elimination methods. Differences in multiple symptom group comparisons were considered significant if P < .0083 (using a Bonferroni correction factor of 6). All other differences were considered significant if P < .05. All statistical analyses were performed by New England Research Institute (NERI, Watertown, Mass) with SAS Statistical Software (Cary, NC).
All data entered into the SVS-VR are fully compliant with the Health Insurance Portability and Accountability Act regulations and are auditable. All data reports and analyses performed include only de-identified and aggregated data. NERI maintains the online database, and funding for the administration and database management of the Vascular Registry has been provided by the Society for Vascular Surgery.
RESULTS
A total of 5758 CEA procedures and 2882 CAS procedures from the SVS-VR met the specified inclusion criteria. Demographics and medical history for the exclusive presenting symptom groups (Table I) reflect the heterogeneity of these unmatched registry populations. The CAS treatment group contains higher percentages of interventions for restenosis and postradiation changes than the CEA group does (Table II), consistent with this technique’s ability to avoid the surgical challenges associated with the hostile or previously operated on neck. Unadjusted event rates for CAS and CEA procedures, delineated by presenting symptom type, are displayed in Table III. Although certain unadjusted event rates in Table III are compelling (such as the 11.6% periprocedural incidence of the composite death + stroke + MI in patients presenting with stroke who were treated with CAS), the unmatched nature of the enrolled CAS and CEA patient populations invalidates comparisons between the endovascular and surgical treatment groups.
Table I.
Demographics and medical history
| Mutually exclusive symptom in last 12 months
|
P valuea | ||||
|---|---|---|---|---|---|
| Stroke (n = 947) | TIA (n = 1449) | TMB (n = 508) | ASX (n = 5736) | ||
| Age, years (range) | 69.7 (29–94) | 71.6 (18–98) | 69.7 (47–94) | 71.2 (39–95) | <.0001 |
| Gender, male (%) | 61.1% (579/947) | 59.8% (866/1449) | 68.3% (347/508) | 58.2% (3336/5736) | <.0001 |
| Race, white | 87.9% (832/947) | 92.8% (1344/1449) | 92.7% (471/508) | 93.7% (5374/5736) | <.0001 |
| Ethnicity, Hispanic | 4.6% (44/947) | 3.1% (45/1449) | 2.8% (14/508) | 3.2% (185/5736) | .1100 |
| Coronary artery disease | 45.3% (429/947) | 49.6% (718/1449) | 47.8% (243/508) | 54.4% (3121/5736) | <.0001 |
| MI | 16.7% (158/947) | 16.7% (242/1449) | 16.9% (86/508) | 19.2% (1104/5736) | .0437 |
| Valvular heart disease | 6.1% (58/947) | 7.5% (109/1449) | 6.5% (33/508) | 7.6% (438/5736) | .3401 |
| Cardiac arrhythmia | 12.2% (116/947) | 13.0% (188/1449) | 12.8% (65/508) | 14.0% (805/5736) | .3684 |
| Congestive heart failure | 10.7% (101/947) | 10.0% (145/1449) | 9.6% (49/508) | 10.0% (574/5736) | .9184 |
| Hypertension | 84.6% (801/947) | 83.8% (1214/1449) | 76.4% (388/508) | 84.7% (4861/5736) | <.0001 |
| Diabetes | 36.3% (344/947) | 32.2% (467/1449) | 28.0% (142/508) | 32.7% (1877/5736) | .0118 |
| Chronic obstructive pulmonary disease | 18.0% (170/947) | 19.3% (280/1449) | 22.6% (115/508) | 17.7% (1013/5736) | .0285 |
| Chronic renal failure | 4.0% (38/947) | 3.5% (50/1449) | 2.0% (10/508) | 3.3% (192/5736) | .2294 |
| Peripheral vascular disease | 33.5% (317/947) | 37.6% (545/1449) | 36.0% (183/508) | 43.9% (2520/5736) | <.0001 |
| Current or past smoker | 59.6% (564/947) | 58.5% (847/1449) | 66.5% (338/508) | 60.9% (3494/5736) | .0120 |
| Cancer | 15.2% (144/947) | 16.9% (245/1449) | 19.1% (97/508) | 14.2% (816/5736) | .0037 |
| Coagulopathy | 1.5% (14/947) | 1.4% (20/1449) | 0.6% (3/508) | 1.3% (72/5736) | .5013 |
| ASA grade | |||||
| ≤3 | 88.1% (834/947) | 91.5% (1326/1449) | 92.9% (472/508) | 92.1% (5282/5736) | .0004 |
| >3 | 11.9% (113/947) | 8.5% (123/1449) | 7.1% (36/508) | 7.9% (454/5736) | |
| New York Heart Association class | |||||
| ≤2 | 93.7% (887/947) | 92.1% (1335/1449) | 93.3% (474/508) | 93.7% (5372/5736) | .2147 |
| >2 | 6.3% (60/947) | 7.9% (114/1449) | 6.7% (34/508) | 6.3% (364/5736) | |
| Aspirin | 79.3% (751/947) | 84.9% (1230/1449) | 82.9% (421/508) | 84.9% (4872/5736) | .0001 |
| Clopidogrel | 46.5% (440/947) | 50.0% (725/1449) | 51.0% (259/508) | 39.0% (2235/5736) | <.0001 |
| Aspirin or clopidogrel | 88.4% (837/947) | 92.0% (1333/1449) | 92.1% (468/508) | 90.3% (5179/5736) | .0146 |
| Age ≥80 years | 20.3% (192/947) | 23.2% (336/1449) | 18.3% (93/508) | 19.5% (1117/5736) | .0115 |
| Embolic protection (CAS only) | 95.8% (322/336) | 97.0% (554/571) | 97.9% (184/188) | 98.0% (1751/1787) | .0992 |
ASA, American Society of Anesthesiologists; ASX, asymptomatic; CAS, carotid artery stenting; MI, myocardial infarction; TIA, transient ischemic attack; TMB, transient monocular blindness.
P value for age was found by analysis of variance. All others found by χ2 tests.
Table II.
Distribution by exclusive presenting symptom type and etiology
| Etiology | Stroke, No. (%) | TIA, No. (%) | TMB, No. (%) | ASX, No. (%) | Total, No. (%) |
|---|---|---|---|---|---|
| CEA patients | |||||
| Atherosclerosis | 604 | 862 | 313 | 3904 | 5683 (98.7) |
| Radiation | 0 | 1 | 2 | 3 | 6 (0.1) |
| Restenosis | 7 | 15 | 5 | 42 | 69 (1.2) |
| Total | 611 (10.6) | 878 (15.3) | 320 (5.5) | 3949 (68.6) | 5758 (100) |
| CAS patients | |||||
| Atherosclerosis | 293 | 413 | 140 | 1202 | 2048 (71.1) |
| Radiation | 10 | 24 | 16 | 91 | 141 (5.9) |
| Restenosis | 33 | 134 | 32 | 494 | 693 (24.0) |
| Total | 336 (11.7) | 571 (19.8) | 188 (6.5) | 1787 (62.0) | 2882 (100) |
ASX, Asymptomatic; CAS, carotid artery stenting; CEA, carotid endarterectomy; TIA, transient ischemic attack; TMB, transient monocular blindness.
Table III.
Unadjusted event rates by procedure and presenting symptom type
| Thirty-day adverse event | CAS patients
|
P valuea | |||
|---|---|---|---|---|---|
| Symptom in last 12 months (mutually exclusive)
| |||||
| Stroke ipsilateral (n = 336) | TIA ipsilateral (n = 571) | TMB ipsilateral (n = 188) | ASX (n = 1787) | ||
| Death | 6.3% (21/336) | 1.1% (6/571) | 0.5% (1/188) | 1.3% (24/1787) | <.0001 |
| Stroke | 8.6% (29/336) | 7.9% (45/571) | 3.2% (6/188) | 3.2% (57/1787) | <.0001 |
| MI | 1.2% (4/336) | 1.2% (7/571) | 1.6% (3/188) | 1.1% (20/1787) | .8652 |
| Death + stroke + MI | 11.6% (39/336) | 9.6% (55/571) | 4.8% (9/188) | 4.9% (88/1787) | <.0001 |
| Death + stroke | 11.0% (37/336) | 8.4% (48/571) | 3.2% (6/188) | 4.3% (77/1787) | <.0001 |
| Thirty-day adverse event | CEA patients
|
P valuea | |||
|---|---|---|---|---|---|
| Symptom in last 12 months (mutually exclusive)
| |||||
| Stroke ipsilateral (n = 611) | TIA ipsilateral (n = 878) | TMB ipsilateral (n = 320) | ASX (n = 3949) | ||
| Death | 1.8% (11/611) | 0.8% (7/878) | 0.0% (0/320) | 0.8% (30/3949) | .0277 |
| Stroke | 4.6% (28/611) | 3.3% (29/878) | 1.6% (5/320) | 1.6% (64/3949) | <.0001 |
| MI | 0.8% (5/611) | 1.6% (14/878) | 0.9% (3/320) | 1.2% (49/3949) | .6213 |
| Death + stroke + MI | 6.7% (41/611) | 5.0% (44/878) | 2.5% (8/320) | 3.1% (124/3949) | <.0001 |
| Death + stroke | 6.2% (38/611) | 4.0% (35/878) | 1.6% (5/320) | 2.2% (87/3949) | <.0001 |
ASX, Asymptomatic; CAS, carotid artery stenting; CEA, carotid endarterectomy; MI, myocardial infarction; TIA, transient ischemic attack; TMB, transient monocular blindness.
Events were defined as any event occurring intraoperatively, before discharge, or between discharge and 30 days. The event rates in the table are per patient.
P values were based on Fisher exact test.
After application of the stringent risk-adjusting methods described before, multiple symptom group comparisons were conducted in bivariate fashion, illuminating the effect of presenting symptom type on each of the major periprocedural outcomes measures (Tables IV and V). Of note is that 30-day outcomes of TMB and ASX presentation are indistinguishable for both CAS and CEA patients. The significant findings from Tables IV and V regarding symptom type presentation on periprocedural adverse event outcomes measure are summarized here.
Table IV.
Risk-adjusted bivariate comparison of outcomes based on presenting symptom types within the carotid artery stenting (CAS) treatment group
| Thirty-day outcome | CAS: Multivariable (adjusted) logistic models
|
Adjusted covariates | ||||
|---|---|---|---|---|---|---|
| Ipsilateral symptom
|
OR | 95% CI | P value | |||
| Comparison group | Reference group | |||||
| Death/stroke/MI outcome | Stroke | vs TIA | 1.08 | 0.69–1.71 | .7274 | Controls for age, white race, diabetes, ASA grade, clopidogrel, and CPD use |
| vs TMB | 2.03 | 0.95–4.37 | .0689 | |||
| vs ASX | 2.30 | 1.52–3.47 | <.0001a | |||
| TIA | vs TMB | 1.88 | 0.90–3.92 | .0941 | ||
| vs ASX | 2.12 | 1.48–3.04 | <.0001a | |||
| TMB | vs ASX | 1.13 | 0.55–2.30 | .7386 | ||
| Death/stroke outcome | Stroke | vs TIA | 1.16 | 0.72–1.86 | .5481 | Controls for age, white race, diabetes, ASA grade, aspirin, clopidogrel, and CPD use |
| vs TMB | 2.88 | 1.17–7.07 | .0209 | |||
| vs ASX | 2.43 | 1.58–3.73 | <.0001a | |||
| TIA | vs TMB | 2.49 | 1.04–5.99 | .0412 | ||
| vs ASX | 2.10 | 1.43–3.08 | .0002a | |||
| TMB | vs ASX | 0.84 | 0.36–1.98 | .6933 | ||
| Death outcome | Stroke | vs TIA | 5.21 | 2.04–13.33 | .0006a | Controls for ASA grade, aspirin, and CPD use |
| vs TMB | 10.03 | 1.32–76.25 | .0259 | |||
| vs ASX | 3.87 | 2.08–7.23 | <.0001a | |||
| TIA | vs TMB | 1.92 | 0.23–16.27 | .5476 | ||
| vs ASX | 0.74 | 0.30–1.85 | .5238 | |||
| TMB | ASX | 0.39 | 0.05–2.90 | .3549 | ||
| Stroke outcome | Stroke | vs TIA | 0.97 | 0.58–1.61 | .9050 | Controls for age, white race, diabetes, ASA grade, clopidogrel, and CPD use |
| vs TMB | 2.20 | 0.88–5.47 | .0905 | |||
| vs ASX | 2.59 | 1.60–4.18 | <.0001a | |||
| TIA | vs TMB | 2.27 | 0.94–5.45 | .0674 | ||
| vs ASX | 2.67 | 1.77–4.03 | <.0001a | |||
| TMB | vs ASX | 1.18 | 0.50–2.79 | .7102 | ||
| MI outcome | Stroke | vs TIA | 1.04 | 0.30–3.58 | .9514 | Controls for CAD |
| vs TMB | 0.75 | 0.17–3.42 | .7150 | |||
| vs ASX | 1.12 | 0.38–3.32 | .8324 | |||
| TIA | vs TMB | 0.73 | 0.19–2.84 | .6461 | ||
| vs ASX | 1.08 | 0.45–2.57 | .8595 | |||
| TMB | vs ASX | 1.49 | 0.44–5.07 | .5241 | ||
ASA, American Society of Anesthesiologists; ASX, asymptomatic; CAD, coronary artery disease; CI, confidence interval; CPD, cerebral protection device; MI, myocardial infarction; OR, odds ratio; TIA, transient ischemic attack; TMB, transient monocular blindness.
P < .0083 considered significant after Bonferroni correction for multiple comparisons.
Table V.
Risk-adjusted bivariate comparison of outcomes based on presenting symptom types within the carotid endarterectomy (CEA) treatment group
| Thirty-day outcome | CEA: Multivariable (adjusted) logistic models
|
Adjusted Covariates | ||||
|---|---|---|---|---|---|---|
| Ipsilateral symptom
|
OR | 95% CI | P value | |||
| Comparison group | Reference group | |||||
| Death/stroke/MI outcome | Stroke | vs TIA | 1.30 | 0.83–2.03 | .2466 | Controls for age, white race, CAD, diabetes, ASA grade, and aspirin use |
| vs TMB | 2.56 | 1.18–5.57 | .0174 | |||
| vs ASX | 2.12 | 1.46–3.08 | <.0001a | |||
| TIA | vs TMB | 1.97 | 0.91–4.25 | .0844 | ||
| vs ASX | 1.63 | 1.14–2.33 | .0073a | |||
| TMB | vs ASX | 0.83 | 0.40–1.72 | .6120 | ||
| Death/stroke outcome | Stroke | vs TIA | 1.50 | 0.93–2.41 | .0935 | Controls for ASA grade and aspirin use |
| vs TMB | 3.85 | 1.50–9.91 | .0052a | |||
| vs ASX | 2.73 | 1.84–4.05 | <.0001a | |||
| TIA | vs TMB | 2.56 | 0.99–6.62 | .0517 | ||
| vs ASX | 1.82 | 1.22–2.72 | .0035a | |||
| TMB | vs ASX | 0.71 | 0.29–1.76 | .4598 | ||
| Death outcome | Stroke | vs TIA | 2.07 | 0.79–5.41 | .1361 | Controls for ASA grade and aspirin use |
| vs ASX | 2.10 | 1.04–4.23 | .0388 | |||
| TIA | vs ASX | 1.01 | 0.44–2.32 | .9796 | ||
| Stroke outcome | Stroke | vs TIA | 1.36 | 0.80–2.32 | .2546 | Controls for diabetes, TMB, and ASA grade |
| vs TMB | 4.58 | 1.36–15.42 | .0140 | |||
| vs ASX | 2.79 | 1.77–4.40 | <.0001a | |||
| TIA | vs TMB | 3.36 | 1.01–11.17 | .0478 | ||
| vs ASX | 2.05 | 1.31–3.20 | .0016a | |||
| TMB | vs ASX | 0.61 | 0.19–1.98 | .4113 | ||
| MI outcome | Stroke | vs TIA | 0.51 | 0.18–1.42 | .1972 | Controls for age, white race, CAD, and diabetes |
| vs TMB | 0.84 | 0.20–3.55 | .8098 | |||
| vs ASX | 0.67 | 0.27–1.71 | .4041 | |||
| TIA | vs TMB | 1.65 | 0.47–5.83 | .4346 | ||
| vs ASX | 1.33 | 0.73–2.43 | .3584 | |||
| TMB | vs ASX | 0.80 | 0.25–2.61 | .7156 | ||
ASA, American Society of Anesthesiologists; ASX, asymptomatic; CAD, coronary artery disease; CI, confidence interval; MI, myocardial infarction; OR, odds ratio; TIA, transient ischemic attack; TMB, transient monocular blindness.
P < .0083 considered significant after Bonferroni correction for multiple comparisons.
Composite outcome: Death + stroke + MI
For both CAS and CEA treatment groups, stroke or TIA presentation predicted a higher risk of this composite outcome than ASX presentation.
Composite outcome: Death + stroke
For both CAS and CEA treatment groups, stroke or TIA presentation predicted a higher risk than ASX presentation.
Individual components of the composite outcomes
Death
In the CAS treatment group only, stroke presentation predicted a higher risk than TIA or ASX presentation.
Stroke
For both CAS and CEA treatment groups, stroke or TIA presentation predicted a higher risk than ASX presentation, as was reflected in the composite outcomes.
MI
No effect of presenting symptom type was noted on this outcome.
DISCUSSION
These data challenge practitioners to reconsider whether the binary classification of carotid lesions as either “asymptomatic” or “symptomatic” oversimplifies a more complex spectrum of disease. Our analyses demonstrate that specific presenting symptom types are powerful predictors of 30-day outcomes for CEA and CAS. In particular, the broad umbrella of symptomatic carotid disease encompasses several presenting symptoms with widely divergent perioperative risk profiles. Familiarity with these additional prognostic factors may allow clinicians to better counsel patients about treatment options and expected outcomes.
Are these expected outcomes broadly applicable to the multiple specialties that engage in performance of CEA and CAS? In their recent analysis of CREST, Timaran et al15 demonstrated that vascular surgeon outcomes were statistically similar to those generated by other participating specialties. Thus, although vascular surgeons represent the majority of clinicians entering patient data into the SVS-VR, we expect that the influence of presenting symptom type on periprocedural outcomes would be a durable finding, regardless of the physician operator.
There are weaknesses of this study that deserve mention. The cohort of patients presenting with TMB (n = 508) is less well powered than in the other study groups, relatively limiting the strength of bivariate comparisons in those instances. The SVS-VR data are self-reported, with inherent potential for bias. In addition, the CEA and CAS patient groups are unmatched, and thus direct comparisons between such cohorts must be conducted with caution, even after concerted efforts at risk adjustment. For that reason, we have primarily sought to identify the prognostic significance of symptom type within the separate CAS and CEA treatment groups and limited comparisons between the therapeutic modalities. Longer term outcomes data would be desirable, but beyond the 30-day perioperative period, the SVS-VR data collection for carotid subjects becomes attenuated.
With regard to risk adjustment techniques, we strove to maximally risk adjust these populations before engaging in bivariate comparisons (Tables IV and V). The rationale for not using a cerebral protection device (CPD) is often not found within the SVS-VR. Thus, the reviewer cannot reliably determine whether failure to use a CPD was secondary to the elective choice of the interventionalist or due to anatomic constraints that rendered CPD use impossible. Of the 2882 CAS procedures examined, 71 (2.5%) did not use a CPD. The refusal or inability to use a CPD was associated with higher risk of stroke (11.3% vs 4.6 % with CPD use; P = .009). As we could not be certain whether CPD nonuse was elective or mandated by anatomy, we chose to risk adjust for CPD use in these analyses; that decision may have introduced bias favoring the outcomes of CAS. In light of the higher adverse event rates seen when CPDs were not used for CAS, CEA should be preferentially employed when this constraint is anticipated.
Acknowledgments
The analysis of the Society for Vascular Surgery Vascular Registry carotid data set was supported exclusively by funds from the Society for Vascular Surgery.
The authors would like to express their gratitude to Sarah Murphy of the Society for Vascular Surgery for her assistance throughout this project and would also like to thank the other members of the Society for Vascular Surgery Outcomes Committee, including Rodney White, MD (Chair), Ellen Dillavou, MD, David Gillespie, MD, Jeffrey Jim, MD, Nicholas Osborne, MD, and Joseph Ricotta, MD.
Footnotes
Author conflict of interest: none.
Presented during the Plenary and Peripheral Vascular Surgery Society Program at the2013 Vascular Annual Meeting of the Society for Vascular Surgery, San Francisco, Calif, May 30–June 1, 2013.
The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest.
AUTHOR CONTRIBUTIONS
Conception and design: PJG, PPG
Analysis and interpretation: PJG, PPG, FS, CK
Data collection: FS, CK
Writing the article: PJG, TB, DG, GU, MS, FS, CK, PPG
Critical revision of the article: PJG, TB, DG, GU, MS, FS, CK, PPG
Final approval of the article: PJG, TB, DG, GU, MS, FS, CK, PPG
Statistical analysis: FS, CK
Obtained funding: PJG, PPG
Overall responsibility: PJG
References
- 1.Acheson J, Hutchinson EC. Observations on the natural history of transient cerebral ischaemia. Lancet. 1964;2:871–4. doi: 10.1016/s0140-6736(64)90734-2. [DOI] [PubMed] [Google Scholar]
- 2.Goldner J, Wisnant JP, Taylor WF. Long-term prognosis of transient cerebral ischemic attacks. Stroke. 1971;2:160–7. doi: 10.1161/01.str.2.2.160. [DOI] [PubMed] [Google Scholar]
- 3.Poole CJ, Ross Russell RW. Mortality and stroke after amaurosis fugax. J Neurol Neurosurg Psychiatry. 1985;48:902–5. doi: 10.1136/jnnp.48.9.902. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med. 1991;325:445–53. doi: 10.1056/NEJM199108153250701. [DOI] [PubMed] [Google Scholar]
- 5.Barnett HJ, Taylor DW, Eliasziw M, Fox AJ, Ferguson GG, Haynes RB, et al. Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. North American Symptomatic Carotid Endarterectomy Trial Collaborators. N Engl J Med. 1998;339:415–25. doi: 10.1056/NEJM199811123392002. [DOI] [PubMed] [Google Scholar]
- 6.Streifler JY, Eliasziw M, Benavente OR, Harbison JW, Hachinski VC, Barnett HJ, et al. The risk of stroke in patients with first-ever retinal vs hemispheric transient ischemic attacks and high-grade carotid stenosis. North American Symptomatic Carotid Endarterectomy Trial. Arch Neurol. 1995;52:246–9. doi: 10.1001/archneur.1995.00540270034016. [DOI] [PubMed] [Google Scholar]
- 7.Ferguson GG, Eliasziw M, Barr HW, Clagett GP, Barnes RW, Wallace MC, et al. The North American Symptomatic Carotid Endarterectomy Trial: surgical results in 1415 patients. Stroke. 1999;30:751–8. doi: 10.1161/01.str.30.9.1751. [DOI] [PubMed] [Google Scholar]
- 8.Bernstein EF, Dilley RB. Late results after carotid endarterectomy for amaurosis fugax. J Vasc Surg. 1987;6:333–40. [PubMed] [Google Scholar]
- 9.Executive Committee for the Asymptomatic Carotid Atherosclerosis Study. Endarterectomy for asymptomatic carotid artery stenosis. JAMA. 1995;273:1421–8. [PubMed] [Google Scholar]
- 10.Theron JG, Payelle GG, Coskun O, Huet HF, Guimaraens L. Carotid artery stenosis: treatment with protected balloon angioplasty and stent placement. Radiology. 1996;201:627–36. doi: 10.1148/radiology.201.3.8939208. [DOI] [PubMed] [Google Scholar]
- 11.Wholey MH, Wholey M, Bergeron P, Diethrich EB, Henry M, Laborde JC, et al. Current global status of carotid artery stent placement. Cathet Cardiovasc Diagn. 1998;44:1–6. doi: 10.1002/(sici)1097-0304(199805)44:1<1::aid-ccd1>3.0.co;2-b. [DOI] [PubMed] [Google Scholar]
- 12.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. doi: 10.1056/NEJMoa040127. [DOI] [PubMed] [Google Scholar]
- 13.Brott TG, Hobson RW, 2nd, Howard G, Roubin GS, Clark WM, Brooks W, et al. Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med. 2010;363:11–23. doi: 10.1056/NEJMoa0912321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sidawy AN, Zwolak RM, White RA, Siami FS, Schermerhorn ML, Sicard GA, et al. Risk-adjusted 30-day outcomes of carotid stenting and endarterectomy: results from the SVS Vascular Registry. J Vasc Surg. 2009;49:71–9. doi: 10.1016/j.jvs.2008.08.039. [DOI] [PubMed] [Google Scholar]
- 15.Timaran CH, Mantese VA, Malas M, Brown OW, Lal BK, Moore WS, et al. Differential outcomes of carotid stenting and endarterectomy performed exclusively by vascular surgeons in the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST) J Vasc Surg. 2013;57:303–8. doi: 10.1016/j.jvs.2012.09.014. [DOI] [PMC free article] [PubMed] [Google Scholar]