Abstract
Background
The Centers for Medicare and Medicaid Services (CMS) have established guidelines that outline patients who are considered “high risk” for complications following CEA for which CAS may provide benefit. The validity of these high risk criteria are yet unproven. In this study, we stratified patients who underwent either CAS or CEA by CMS high risk criteria and symptom status, and examined their 30-day outcomes.
Methods
A non-randomized, retrospective cohort study was performed by chart review of all patients undergoing CEA or CAS from January 1, 2005 to December 31, 2010 at our institution. Demographic data, as well as data pertaining to the presence or absence of high risk factors were collected. Patients were stratified using symptom status and high risk status as variables and 30-day adverse events (stroke, death, and/or myocardial infarction (MI)) were compared.
Results
271 patients underwent CAS with 30-day complication rates of stroke (3.0%), death (1.1%), MI (1.5%), stroke/death (3.7%), and stroke/death/MI (5.2%). 830 patients underwent CEA with 30-day complication rates of stroke (2.0%), death (0.1%), MI (0.6%), stroke/death (1.9%), and stroke/death/MI (2.7%). Among symptomatic patients, physiologic high risk status was associated with increased stroke/death (6/42, 14.3% vs. 2/74, 2.7%, P<0.01), and anatomic high risk status was associated with a trend towards increased stroke/death (5/31, 16.1% vs. 0/20, 0.0%, P=0.14) in patients who underwent CAS compared to CEA. Analysis of asymptomatic patients showed no differences among the two groups overall, except for a trend towards higher rate of MI following CAS compared to CEA (3/71, 4.2% vs. 0/108, 0.0%, P=.06) in those who were physiologically high risk. Among symptomatic patients who underwent CAS, patients with physiologic and anatomic high risk factors had a higher rate of stroke/death compared to non-high risk patients (6/42, 14.3% vs. 0/24, 0.0% and 5/31, 16.1% vs. 0/24, 0.0%, respectively, both P≤.05)
Conclusions
Physiologic high risk status was associated with increased stroke/death, while anatomic high risk status showed a trend towards increased stroke/death, in symptomatic patients undergoing CAS compared to non-high risk patients undergoing CAS or physiologically high risk patients undergoing CEA. Our results suggest that the current national criteria for carotid artery stenting overestimates its efficacy in patients who are symptomatic and high risk.
INTRODUCTION
Carotid endarterectomy (CEA) has been established as the gold standard treatment for reducing the risk of stroke in patients with severe carotid artery stenosis1–4. More recently, carotid artery stenting (CAS) has become an accepted treatment alternative in those considered high risk for complications following CEA. Using data from previous registries, the Centers for Medicare and Medicaid Services (CMS) have established and reaffirmed guidelines regarding the use of CAS as approved reimbursement coverage criteria5. These criteria outline patients who are considered “high risk” for complications following CEA for which CAS may provide benefit. However, data from subsequent studies has led to questions regarding the validity of these high risk criteria6–9. We have found that studies known to date have stratified data according to high risk status or symptom status but not both. Therefore, we hypothesized that a two-tiered stratification approach that include high risk and symptom status will further delineate a subset of patients in which CEA or CAS will confer reduced risk. In an attempt to improve patient selection for the treatment of carotid artery disease, we performed a chart review of all patients who underwent either CAS or CEA at our tertiary medical center from 2005 to 2010, stratified them according to CMS high risk criteria and symptom status, and examined their 30-day outcomes.
METHODS
Patients
A non-randomized, retrospective cohort study was performed by chart review of all patients undergoing CEA or CAS from January 1, 2005 to December 31, 2010 at our institution. The Vascular Registry, a national carotid procedures registry maintained by the Society for Vascular Surgery (SVS) was used to identify all patients undergoing CEA or CAS. Additionally, the hospital database was searched using ICD-9-CM (International Classification of Diseases, 9th Revision, Clinical Modification) diagnosis and procedure codes (CAS: 00.63, CEA: 38.12), to identify cases performed prior to joining the SVS registry and to confirm that all cases were entered into the SVS Vascular Registry.
Data Acquisition
Demographic data and other data outlined by the CMS guidelines were obtained. These data included symptom status, degree of stenosis, specific physiologic or anatomic risk factors deemed high risk for CEA per CMS guidelines5, and adverse outcomes within 30 days of the procedure. Under CMS guidelines, “patients at high risk for CEA who have symptomatic carotid artery stenosis >70% are covered for procedures performed using Food and Drug Administration-approved CAS systems with embolic protection devices in facilities approved by CMS to perform CAS procedures. In addition, patients at high risk for CEA with symptomatic carotid artery stenosis between 50% and 70% and patients at high risk for CEA with asymptomatic carotid artery stenosis >80% are covered in accordance with the Category B Investigational Device Exemption clinical trials regulation (42 CFR 405.201), as a routine cost under the clinical trials policy”5. Physiologic high risk variables included age > 80, class III congestive heart failure (CHF), left ventricular ejection fraction (LVEF) < 30%, unstable angina, MI within 30 days, forced expiratory volume < 30% of predicted or home oxygen use, contralateral internal carotid artery occlusion, recent coronary artery bypass grafting (CABG) or valve repair, and hemodialysis. Anatomic high risk factors included contralateral laryngeal nerve palsy, restenosis, history of neck radiation, high or low lesion, and prior neck surgery. Adverse events, including stroke, death, or myocardial infarction (MI) within 30-days of the procedure served as primary outcomes. Strokes excluded TIAs. Patients with nonspecific symptoms or undocumented symptom status were considered asymptomatic for the purposes of this study. These patients were known to have carotid stenoses of greater than 80% but without focal symptoms that would constitute amaurosis fugax, TIA, RIND, or stroke. The chart review included all available notes (including the H&P and ROS). Less than 3% of patients in the asymptomatic cohort fit this category.
Statistical Analysis
We analyzed 30-day adverse events in patients undergoing CAS versus CEA stratified by symptom status (symptomatic or asymptomatic) and high risk status (physiologic or anatomic). We then analyzed the 30-day adverse events in symptomatic patients underging CAS or CEA based on high risk status (physiologic or anatomic as compared to non-high risk status). Patient variables were compared using univariate analysis. Categorical variables were analyzed using the chi-square test and Fisher's exact test where appropriate. Statistical significance was determined as P<.05. All statistical analyses were performed using STATA 12 software (StataCorp, College Station, TX). The institutional review board at Beth Israel Deaconess Medical Center approved this study.
RESULTS
A total of 271 CAS patients and 830 CEA patients were reviewed for this study. Table 1 shows patient demographics, symptom status, CMS high risk status (including co-morbidities that qualify patients into physiologic high risk status) and gender in CAS and CEA patients. A significantly higher proportion of patients who underwent CAS were high risk, both in overall numbers or categorized into physiologic, anatomic, or both (63.5%, 41.7%, 30.3%, and 8.5% respectively), when compared to the CEA patient population (25.3%, 21.9%, 5.1%, and 1.7% respectively; P<.01 all). There was no difference in the proportion of symptomatic patients who underwent CEA (32.9%) compared to CAS (31.4%).
The majority of physiologic high risk factors represented in the CEA patient population were octogenarian status (19.8%), end-stage renal disease (1.3%) and contralateral internal carotid artery occlusion (5.9%). Each of the other co-morbidities was present in ≤1% of the total CEA population. This trend was also seen in the CAS patient population (octogenarian status, 17.3%; contralateral occlusion, 8.5%), but a wider distribution of co-morbidities was found in the CAS patient population. A significantly higher percentage of patients received carotid stents in the settings of significant CHF, poor LVEF, unstable angina, recent MI, severe lung disease, and in conjunction with CABG. There were no significant differences in the percentage of octogenarians, patients with end-stage renal disease, or patients with contralateral occlusion between the CAS and CEA patient populations.
Of the 271 patients who underwent CAS, 8 strokes (3.0%), 3 deaths (1.1%) and 4 MIs (1.5%) were recorded at 30-days. Among the 830 CEA patients, there were 17 strokes (2.0%), 1 death (0.1%), and 5 MIs (0.6%) recorded at 30-days. Ten stroke or deaths (3.7%) and 13 stroke, death, or MIs (4.8%) were counted following CAS and 17 stroke or deaths (2.0%) and 22 stroke, death, or MIs (2.7%) following CEA. Globally, statistical significance was only seen with death (CAS 3/271 vs. CEA 1/830, p=0.05) and stroke/death/MI (CAS 14/271 vs. CEA 22/830, p=0.04.
Stratified Analysis based on Symptom Status and High Risk Status
Differences in outcome between CAS and CEA were measured after stratifying the data by symptom status and high risk status (Table 2). Among the symptomatic patients, physiologic high risk status was associated with increased stroke/death (6/42, 14.3% vs. 2/74, 2.7%, P<.01) in patients who underwent CAS compared to CEA, with a trend for higher stroke (4/42, 9.5% vs. 2/74, 2.7%, P=.19) and stroke/death/MI rates (6/42, 14.3% vs. 3/74, 4.2%, P=0.07) following CAS. Of the 42 symptomatic, physiologic high risk patients who underwent CAS, 23 were octogenarians and accounted for three of the six stroke/deaths. There were no differences in adverse events among the symptomatic, non-high risk patients undergoing CAS versus CEA. Symptomatic, anatomic high risk patients showed a trend toward higher rates of stroke (4/31, 12.9% vs. 0/20, 0.0%, P=.15) and stroke/death (5/31, 16.1% vs. 0/20, 0.0%, P=.14) following CAS.
Table 2.
30-day adverse events after CAS vs. CEA stratified by symptom status and high risk status
| Symptomatic | Stroke | Death | MI | Stroke/Death | Stroke/Death/MI | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||||||||
| CAS | CEA | p-value | CAS | CEA | p-value | CAS | CEA | p-value | CAS | CEA | p-value | CAS | CEA | P-value | |
| Physiologic High Risk (n=116) | (4/42) 9.5% | (2/74) 2.7% | 0.19 | (2/42) 4.8% | (0/74) 0% | 0.13 | (0/42) 0% | (1/74) 1.4% | 1.00 | (6/42) 14.3% | (2/74) 2.7% | 0.01 | (6/42) 14.3% | (3/74) 4.1% | 0.07 |
| Non High Risk (208) | (0/24) 0% | (7/184) 3.8% | 1.00 | (0/24) 0% | (0/184) 0% | 1.00 | (0/24) 0% | (1/184) 0.5% | 1.00 | (0/24) 0% | (7/184) 3.8% | 1.00 | (0/24) 0% | (8/184) 4.4% | 0.60 |
| Anatomic High Risk (n=51) | (4/31) 12.9% | (0/20) 0% | 0.15 | (2/31) 6.5% | (0/20) 0% | 0.51 | (0/31) 0% | (0/20) 0% | 1.00 | (5/31) 16.1% | (0/20) 0% | 0.14 | (5/31) 16.1% | (0/20) 0% | 0.14 |
| Both Anatomic & Physiolgic HR (n=17) | (1/12) 8.3% | (0/5) 0% | 0.51 | (1/12) 8.3% | (0/5) 0% | 0.51 | (0/12) 0% | (0/5) 0% | 1.00 | (2/12) 16.7% | (0/5) 0% | 0.33 | (2/12) 16.7% | (0/5) 0% | 0.33 |
| Asymptomatic | Stroke | Death | MI | Stroke/Death | Stroke/Death/MI | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
| |||||||||||||||
| CAS | CEA | p-value | CAS | CEA | p-value | CAS | CEA | p-value | CAS | CEA | p-value | CAS | CEA | P-value | |
| Physiologic High Risk (n=179) | (1/71) 1.4% | (2/108) 1.9% | 1.00 | (0/71) 0% | (1/108) 0.9% | 1.00 | (3/71) 4.2% | (0/108) 0% | 0.06 | (1/71) 1.4% | (2/108) 1.9% | 1.00 | (3/71) 4.2% | (2/108) 1.9% | 0.39 |
| Non High Risk (511) | (0/75) 0% | (6/436) 1.4% | 0.59 | (0/75) 0% | (0/436) 0% | 1.00 | (1/75) 1.3% | (3/436) 0.7% | 0.47 | (0/75) 0% | (6/436) 1.4% | 0.59 | (1/75) 1.3% | (9/436) 2.1% | 1.00 |
| Anatomic High Risk (n=73) | (0/51) 0% | (1/22) 4.6% | 0.30 | (0/51) 0% | (0/22) 0% | 1.00 | (2/51) 3.9% | (0/22) 0% | 1.00 | (0/51) 0% | (1/22) 4.6% | 0.30 | (1/51) 2% | (1/22) 4.6% | 0.52 |
| Both Anatomic & Physiolgic HR (n=20) | (0/11) 0% | (1/9) 11.1% | 0.26 | (0/11) 0% | (0/9) 0% | 0.26 | (2/11) 18.2% | (0/9) 0% | 0.18 | (0/11) 0% | (1/9) 11.1% | 0.26 | (2/11) 18.2% | (1/9) 11.1% | 0.66 |
Asymptomatic patients had a trend towards higher rate of MI following CAS compared to CEA (3/71, 4.2% vs. 0/108, 0.0%, P=.06) if they were physiologically high risk. Among asymptomatic patients who were non high risk or anatomically high risk, there were no differences in adverse events at 30 days. Overall, stroke, death, and stroke/death rates were low in asymptomatic patients following CAS while rates of MI were low in asymptomatic patients following CEA.
Among symptomatic patients who underwent CAS (Table 3), physiologic high risk and anatomic high risk patients had a higher rate of stroke/death than their non-high risk counterparts (6/42, 14.3% vs. 0/24, 0.0%, P=.05 and 5/31, 16.1% vs. 0/24, 0.0%, P=.04 respectively). There were no MIs in symptomatic patients following CAS. Furthermore, symptomatic non high risk patients were complication-free during the 30-day investigative window. In comparison, symptomatic patients who underwent CEA had no differences in 30-day adverse events when comparing risk status.
Table 3.
30-day adverse events in symptomatic patients undergoing CAS or CEA comparing physiologic high risk and anatomic high risk to non-high risk
| CAS | Symptomatic Physiologic High Risk (n=42) | Symptomatic Non High Risk (n=24) | P value | Symptomatic Anatomic High Risk (n=31) | Symptomatic Non High Risk (n=24) | P value |
|---|---|---|---|---|---|---|
| Stroke | (4/42) 9.5% | (0/24) 0% | 0.12 | (4/31) 12.9% | (0/24) 0% | 0.07 |
| Death | (2/42) 4.8% | (0/24) 0% | 0.28 | (2/31) 6.5% | (0/24) 0% | 0.20 |
| MI | (0/42) 0.0% | (0/24) 0% | 1.00 | (0/31) 0.0% | (0/24) 0% | 1.00 |
| Stroke/Death | (6/42) 14.3% | (0/24) 0% | 0.05 | (5/31) 16.1% | (0/24) 0% | 0.04 |
| Stroke/Death/MI | (6/42) 14.3% | (0/24) 0% | 0.05 | (5/31) 16.1% | (0/24) 0% | 0.04 |
| CEA | Symptomatic Physiologic High Risk (n=74) | Symptomatic Non High Risk (n=184) | P value | Symptomatic Anatomic High Risk (n=20) | Symptomatic Non High Risk (n=184) | P value |
|---|---|---|---|---|---|---|
| Stroke | (2/74) 2.7% | (7/184) 3.8% | 0.66 | (0/20) 0% | (7/184) 3.8% | 0.37 |
| Death | (0/74) 0.0% | (0/184) 0% | 1.00 | (0/20) 0% | (0/184) 0% | 1.00 |
| MI | (1/74) 1.4% | (1/184) 0.5% | 0.50 | (0/20) 0% | (1/184) 0.5% | 0.74 |
| Stroke/Death | (2/74) 2.7% | (7/184) 3.8% | 0.66 | (0/20) 0% | (7/184) 3.8% | 0.37 |
| Stroke/Death/MI | (3/74) 4.1% | (8/184) 4.3% | 0.92 | (0/20) 0% | (8/184) 4.3% | 0.34 |
There were no significant differences among asymptomatic patients who underwent CAS or CEA, when comparing physiologic high risk or anatomic high risk vs. non-high risk (Table 4). The lone exception was the observation of one death among asymptomatic, physiologically high risk patients that led to a statistically significant result (1/108, 0.9% vs. 0/436, 0.0%, P=0.04).
Table 4.
30-day adverse events in asymptomatic patients undergoing CAS or CEA comparing physiologic high risk and anatomic high risk to non-high risk
| CAS | Asymptomatic Physiologic High Risk (n=71) | Asymptomatic Non High Risk (n=75) | P value | Asymptomatic Anatomic High Risk (n=51) | Asymptomatic Non High Risk (n=75) | P value |
|---|---|---|---|---|---|---|
| Stroke | (1/71) 1.4% | (0/75) 0.0% | 0.30 | (0/51) 0.0% | (0/75) 0.0% | 1.00 |
| Death | (0/71) 0.0% | (0/75) 0.0% | 1.00 | (0/51) 0.0% | (0/75) 0.0% | 1.00 |
| MI | (3/71) 4.2% | (1/75) 1.3% | 0.28 | (2/51) 3.9% | (1/75) 1.3% | 0.35 |
| Stroke/Death | (1/71) 1.4% | (0/75) 0.0% | 0.30 | (0/51) 0.0% | (0/75) 0.0% | 1.00 |
| Stroke/Death/MI | (4/71) 5.6% | (1/75) 1.3% | 0.15 | (2/51) 3.9% | (1/75) 1.3% | 0.35 |
| CEA | Asymptomatic Physiologic High Risk (n=108) | Asymptomatic Non High Risk (n=436) | P value | Asymptomatic Anatomic High Risk (n=22) | Asymptomatic Non High Risk (n=436) | P value |
|---|---|---|---|---|---|---|
| Stroke | (2/108) 1.9% | (6/436) 1.4% | 0.71 | (1/22) 4.6% | (6/436) 1.4% | 0.24 |
| Death | (1/108) 0.9% | (0/436) 0.0% | 0.04 | (0/22) 0.0% | (0/436) 0.0% | 1.00 |
| MI | (0/108) 0.0% | (3/436) 0.7% | 0.39 | (0/22) 0.0% | (3/436) 0.7% | 0.70 |
| Stroke/Death | (2/108) 1.9% | (6/436) 1.4% | 0.71 | (1/22) 4.6% | (6/436) 1.4% | 0.24 |
| Stroke/Death/MI | (2/108) 1.9% | (9/436) 2.1% | 0.88 | (1/22) 4.6% | (9/436) 2.1% | 0.44 |
The Influence of Octogenarian Status on Outcome
Of the 271 patients who underwent CAS, 47 were greater than 80 years old and 224 were less than 80 years old. Rates of stroke/death was similar 3/47 (6.4%) vs. 7/224 (3.1%), p=0.28. Of the 830 patients who underwent CEA 164 were greater than 80 years old and 668 were less than 80 years old. Again, rates of stroke/death was similar 5/164 (3.1%) vs. 12/668 (1.8%), p=0.31.
Sixty six physiologic high risk patients under the age of 80 underwent CAS and 48 of such patients underwent CEA. Among this population, there was no difference in stroke [4.6% (3/66) vs. 0% (0/48), p=0.26], MI [3.0% (2/66) vs. 0% (0/48), p=0.51], death [1.5% (1/66) vs. 0% (0.48), p=1.00], or stroke/death [6.1% (4/66) vs. 0% (0/48), p=0.14]. However, a statistical difference was detected in stroke/death/MI [9.1% (6/66) vs. 0% (0/48), p=0.04]. Further stratification to include symptom status yielded 19 symptomatic physiologic high risk patients less than 80 years old who underwent CAS and 16 that underwent CEA. There were no differences in 30-day outcomes (CAS vs. CEA): stroke [(2/19) 10.5% vs. (0/16) 0.0%, p=0.49], MIs (none), death [(1/19) 5.3% vs. (0/16) 0.0%, p=1.00], stroke/death [(3/19) 15.8% vs. (0/16) 0.0%, p=0.23], and stroke/death/MI [(3/19) 15.8% vs. (0/16) 0.0%, p=0.23].
DISCUSSION
Our study has two important findings. First, significantly higher rates of stroke/death were observed in physiologically high risk, symptomatic patients following CAS compared to CEA. Second, an increased risk of stroke/death was observed in physiologically high risk, symptomatic patients compared to non-high risk, symptomatic patients undergoing CAS, but this was not observed in patients undergoing CEA. Symptomatic patients are thought to incur higher rates of morbidity and mortality following CAS compared to CEA10–12. Our results suggest that among symptomatic patients, those who are physiologically high risk may have the highest risk of 30-day adverse events following CAS. The CMS physiologic high risk criteria, based on the criteria first published in the SAPPHIRE trial13, are intended to identify patients that are at increased risk for complications with CEA. These patients are considered better suited for an endovascular procedure to avoid potential peri-operative risks associated with open surgery. Our analysis suggests that the current CMS physiologic high risk criteria may not identify the population best suited for CAS.
We observed a trend toward higher incidences of stroke/death in the symptomatic, anatomic high risk patients who underwent CAS compared to CEA, but statistical significance was not reached. Further study with greater power is indicated to determine whether symptomatic, anatomic high risk patients are at greater risk for 30-day adverse events after CAS as surgeons would be more likely to undertake redo surgery if this were clearly demonstrated.
According to our data, differences in outcome may be tied to both high risk status and symptom status, and prospective data are needed to provide evidence-based recommendations for patients. Previous randomized control trials have stratified patients according to symptom status or high risk status but not both. The Stent-Protected Angioplasty versus Carotid Endarterectomy (SPACE) trial randomized 1183 symptomatic patients11. The study excluded patients with restenosis and previous radiation. No other risk status restrictions were instituted, however, some high risk patients were likely excluded under the criterion of severe concomitant disease with poor prognosis as determined by individual physicians. The results showed no differences in outcomes of stroke/death between CAS (6.84%) and CEA (6.34%) at 30 days, yet failed to prove non-inferiority of carotid stents. The Endarterectomy Versus Angioplasty in Patients with Symptomatic Severe Carotid Stenosis (EVA-3S) study also studied symptomatic patients and is known for prematurely stopping the trial early for futility and safety concerns10. Thirty day stroke/death rate for the CAS arm was 9.6% compared to 3.9% in the CEA arm at the conclusion of the study. High risk status was not accounted for in the EVA-3 trial, although this study excluded patients with unstable angina, restenosis, and need for additional surgical procedures within 30 days. Only 12 CAS deployments were required to be included as operators in the study compared to a median number of 64 in SAPPHIRE, and many have criticized this as bias favoring CEA. The International Carotid Stenting Study (ICSS) randomized 1649 symptomatic, low risk patients14. According to the study protocol, patients with re-stenosis, planned coronary artery bypass grafting, those medically not fit or anatomically not suitable for surgery, were excluded from the study. The trial concluded with 30-day stroke/death and stroke/death/MI rates that favored CEA (7.4% vs. 3.4% P=.0004 and 7.4% vs. 4.0% P=.003 respectively). The Carotid Revascularization Endarterectomy vs. Stent Trial (CREST) randomized 2502 patients, 52% of which were symptomatic15. Patients were mostly non-high risk as the study excluded patients with all major high risk criteria as set by CMS except age > 80 years and contralateral ICA occlusion16. There were no significant differences in rates of ipsilateral stroke, MI, or death during the four years following CAS or CEA (7.2% vs. 6.8%, P=.51). When the four year rate of ipsilateral stroke or death were compared, the results favored CEA (6.4% vs. 4.7%, P=.03). Peri-procedural strokes were more common following CAS (4.1% vs. 2.3%, P=.01), while peri-procedural MIs were more common after CEA (1.1% vs. 2.3%, P=.03)15. Later analysis showed stroke or death rates to be lower among symptomatic patients who underwent CEA (4.4% vs. 2.3%, P=.005)12. Overall incidences of 30-day adverse events were low compared to previous randomized control trials12, 15. Low event rates in CREST may be explained somewhat by the fact that most patients were “low risk”.
In the Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy (SAPPHIRE) Trial, 334 high risk patients were evaluated13. The high risk criteria were those incorporated by CMS and therefore, similar to the criteria used for this study. Results of 30-day adverse events favored CAS when comparing the incidences of stroke/death/MI (7/159, 4.4% vs. 15/151, 9.9%; P<.04). The rates of MI had a strong influence on outcome (3/159, 1.9% for CAS and 10/151, 6.6% for CEA) and no differences were seen in outcomes of stroke, death, or stroke/death. The SAPPHIRE trial is known as the first trial to show non-inferiority of CAS compared to CEA in the high risk population. However, the patients were not stratified by symptom status and over 70% were asymptomatic. The subsequent SAPPHIRE registry data showed that patients with physiologic high risk factors have increased risk of 30-day adverse events compared to patients with anatomic high risk factors (4.9% vs. 2.8% P=.0306)17. Our 30-day adverse event rates for CAS (Stroke 3.0%, Death 1.1%, MI 1.5%, Combined 5.2%) were similar to that reported in SAPPHIRE (3.1%, 0.6%, 1.9%, 4.4% respectively), however CEA results from their trial (Stroke 3.3%, Death 2.0%, MI 6.6%, Combined 9.9%) showed much higher rates of complications compared to our data (Stroke 2.0%, Death 0.1%, MI 0.6%, Combined 2.7%).
Interestingly, non-high risk patients and asymptomatic patients who underwent CAS were not at increased risk of 30-day complications compared to CEA. The CREST investigators also found no differences in peri-procedural adverse events among asymptomatic patients who underwent CAS versus CEA12. The current focus is to allow CAS in symptomatic or high risk patients, instead, carotid artery stenting may be a reasonable alternative in asymptomatic or non-high risk patients. The SVS vascular registry study8 compared between symptomatic and asymptomatic patients undergoing CAS and found increased stroke/death/MI rates in symptomatic patients. This study further compared data between CEA and CAS, but they did not stratify the data according to high risk status. Others have performed retrospective studies to question the validity of the high risk criteria in patients who underwent CAS9 or CEA18, 19 but peri-operative stroke/death rates were comparable among high risk and non-high risk cohorts. The key difference between this report and prior studies is the analysis of patient groups stratified by both symptom and high risk status. Furthermore, high risk status was stratified by physiologic versus anatomic high risk factors.
Advanced age is thought to incur worse outcomes in those who undergo CAS with octogenarians being the most vulnerable11, 14, 15, 20. Our results also recorded a trend towards worse outcomes in octogenarians who underwent CAS compared CEA (OR =2.06), however these results were not statistically significant (p=0.38). Octogenarians comprised a majority of those with physiologic high risk designations in the CEA group (164/182, 90.1%) and approximately half of those in the CAS group (47/113, 41.6%) in this study. We examined the data among physiologic high risk patients under the age of 80 and found no differences in stroke/death rates (4/66 vs. 0/48, p=0.51) but stroke/death/MI rates were increased in those who underwent CAS compared to CEA (6/66 vs. 0/48, p=0.04). Among non-octogenarians with physiologic high risk and symptoms, no statistically significant results were achieved between CAS and CEA in either stroke/death or stroke/death/MI (3/19 vs. 0/16, p=0.23). Trends toward better outcomes following CEA were evident, and further evaluation with larger numbers is warranted.
Among the non-octogenarian patients, the diversity of physiologic high risk factors varied between the two groups. The CEA group was mostly comprised of patients with contralateral occlusion (29/48) and nearly all (46/48, 95.8%) had just one high risk factor. The CAS cohort had more patients with multiple risk factors (20/66, 30.3%), and more patients with high risk factors that are likely to be more prohibitive of an operation such as CHF and unstable angina. The experimental and selective nature of CAS inherently leads to a patient population that is higher risk.
Retrospective studies that analyze non-randomized data are inherently biased against CAS because greater percentages of high risk and symptomatic patients are included in the CAS group secondary to CMS reimbursement criteria and patient selection by the provider. This imbalance in patient cohorts makes a fair comparison of outcomes difficult to perform. Our group previously published an analysis of Nationwide Inpatient Sample data stratifying patient data according to high risk status and found CAS to be associated with a higher incidence of stroke, death, and stroke/death in the medically high risk population7. However, database limitations led to the inability to 1) distinguish peri-operative from pre-operative stroke, and 2) discern the severity of comorbid medical conditions leading to potentially inaccurate assignment of high risk status. Furthermore, the analysis was limited to physiological high risk status and immediate in-hospital peri-operative complication rates. This study was aimed at overcoming these limitations by performing a chart review. We were able to show CAS to be associated with worse 30-day outcomes in physiologically high risk, symptomatic patients. Our results suggest the need for future trials to include this stratification. In our study, we did not detect differences in adverse event rates between CAS and CEA in the asymptomatic and non-high risk patient populations, however, stratification of data yielded low event rates that increased the chances of committing a type II error. Larger sample sizes are needed to overcome this limitation.
Finally, we acknowledge that the sample size among high risk patients was too small to compare the effects of each risk factor and its effect on 30-day morbidity or mortality rates. Previous research has demonstrated that a minimum of 5 events needs to be present for each independent variable included in a multivariable model21. This type of analysis could be performed albeit at the risk of overloading the model and thus making the results highly unreliable. Because of this limitation we felt it was best to limit our analysis to its current form. Continued collection of data will improve our ability to detect differences among individual risk factors in the future. Analysis using a large database to determine the effects of individual risk factors has been performed previously for patients who underwent CEA22. Whether the same risk factors hold true for those undergoing CAS is unknown. Refinement of data collection and analysis methods will improve our patient selection for CAS, CEA, or medical management.
CONCLUSION
Physiologic high risk status was associated with increased stroke/death in symptomatic patients undergoing CAS compared to non-high risk patients undergoing CAS or physiologically high risk patients undergoing CEA. Anatomic high risk status also showed increased rates of stroke/death in symptomatic patients although statistical significance was not reached. Current selection criteria for CAS versus CEA may be inappropriate. Physiologically high risk, symptomatic patients, as defined by the current CMS criteria for CAS reimbursement, may be better suited for CEA than CAS. CAS may be a reasonable alternative to CEA in symptomatic patients who are non-high risk. Future comparative analyses of CEA versus CAS should be stratified according to physiologic high risk status and symptom status to overcome bias in patient selection. Further work and greater number of patients are needed to identify appropriate patients for CAS, CEA, or medical management.
Table 1A.
Demographics and comorbidities of patients who underwent CAS or CEA from 01/01/2005 to 12/31/2010
| CAS (n=271) | CEA (n=830) | P-value | |||
|---|---|---|---|---|---|
| (n) | (%) | (n) | (%) | ||
| Male gender | 179 | 66.1% | 470 | 56.6% | <0.01 |
| Symptomatic | 85 | 31.4% | 273 | 32.9% | 0.64 |
| High Risk (any) | 172 | 63.5% | 210 | 25.3% | <0.001 |
| Physiologic High Risk | 113 | 41.7% | 182 | 21.9% | <0.001 |
| Age > 80 years | 47 | 17.3% | 164 | 19.8% | 0.38 |
| CHF (> class III) | 13 | 4.8% | 1 | 0.1% | <0.001 |
| LVEF < 30% | 21 | 7.7% | 7 | 0.8% | <0.001 |
| Unstable angina | 18 | 6.6% | 3 | 0.4% | <0.001 |
| Myocardial infarction w/in 30 days | 6 | 2.2% | 1 | 0.1% | <0.001 |
| Hemodialysis | 2 | 0.7% | 11 | 1.3% | 0.44 |
| Severe Lung Disease* | 7 | 2.6% | 6 | 0.7% | 0.01 |
| Contralateral Occlusion | 23 | 8.5% | 49 | 5.9% | 0.14 |
| Pre or Post CABG/VR | 23 | 8.5% | 6 | 0.7% | <0.001 |
| Anatomic High Risk | 82 | 30.3% | 42 | 5.1% | <0.001 |
| Contralateral laryngeal nerve palsy | 1 | 0.4% | 0 | 0.0% | 0.25 |
| Restenosis | 60 | 22.1% | 38 | 4.6% | <0.001 |
| History of neck radiation | 14 | 5.2% | 12 | 1.5% | <0.01 |
| High or Low lesion | 12 | 4.4% | 5 | 0.6% | <0.001 |
| Prior neck surgery | 18 | 6.6% | 3 | 0.4% | <0.001 |
| High Risk both | 23 | 8.5% | 14 | 1.7% | <0.001 |
CHF = congestive heart failure; LVEF = left ventricular ejection fraction; CABG/VR = coronary artery bypass graft/valve repair;
Forced Expiratory Volume in 1 second < 30% or home oxygen
Table 1B.
High-risk variables of symptomatic patients who underwent CAS or CEA from 01/01/2005 to 12/31/2010
| CAS (n=85) | CEA (n=273) | P-value | |||
|---|---|---|---|---|---|
| (n) | (%) | (n) | (%) | ||
| Physiologic High Risk | 42 | 49.4% | 74 | 27.1% | <0.001 |
| Age > 80 years | 23 | 27.1% | 60 | 22.0% | 0.38 |
| CHF (> class III) | 2 | 2.4% | 0 | 0.0% | 0.06 |
| LVEF < 30% | 8 | 9.4% | 2 | 0.7% | <0.001 |
| Unstable angina | 5 | 5.9% | 0 | 0.0% | <0.001 |
| Myocardial infarction w/in 30 days | 2 | 2.4% | 0 | 0.0% | 0.06 |
| Hemodialysis | 2 | 2.4% | 5 | 1.8% | 0.67 |
| Severe Lung Disease* | 1 | 1.2% | 2 | 0.7% | 0.56 |
| Contralateral Occlusion | 9 | 10.6% | 16 | 5.9% | 0.15 |
| Pre or Post CABG/VR | 3 | 3.5% | 2 | 0.7% | 0.09 |
| Anatomic High Risk | 31 | 36.5% | 20 | 7.3% | <0.001 |
| Contralateral laryngeal nerve palsy | 0 | 0.0% | 0 | 0.0% | 1.00 |
| Restenosis | 20 | 23.5% | 10 | 3.7% | <0.001 |
| History of neck radiation | 5 | 5.9% | 10 | 3.7% | 0.36 |
| High or Low lesion | 7 | 8.2% | 1 | 0.4% | <0.001 |
| Prior neck surgery | 8 | 9.4% | 2 | 0.7% | <0.001 |
| High Risk both | 12 | 14.1% | 5 | 1.8% | <0.001 |
CHF = congestive heart failure; LVEF = left ventricular ejection fraction; CABG/VR = coronary artery bypass graft/valve repair;
Forced Expiratory Volume in 1 second < 30% or home oxygen
Acknowledgments
This work was supported by the NIH T32 Harvard-Longwood Research Training in Vascular Surgery Grant HL007734
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Author Disclosures: S. Yoshida, None; R.P. Bensley, None; J.D. Glaser, None; C.S. Nabzdyk, None; A.D. Hamdan, None; M.C. Wyers, None; E.L. Chaikof, None; M.L. Schermerhorn, Endologix, Boston Scientific, and Medtronic Consultant.
Bibliography
- 1.Barnett HJM, Taylor DW, Haynes RB, Sackett DL, Peerless SJ, Ferguson GG, et al. The 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]
- 2.Toole JF, Baker WH, Castaldo JE, Chambless LE, Moore WS, Robertson JT, et al. The Executive Committee for the Asymptomatic Carotid Atherosclerosis Study Endarterectomy for asymptomatic carotid artery stenosis. JAMA. 1995;273:1421–8. [PubMed] [Google Scholar]
- 3.Farrell B, Fraser A, Sandercock P, Slattery J, Warlow CP, The 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;351:1379–87. [PubMed] [Google Scholar]
- 4.Halliday A, Mansfield A, Marro J, Peto C, Peto R, Potter J, et al. The MRC Asymptomatic Carotid Surgery Trial (ACST) Collaborative Group Prevention of disabling and fatal strokes by successful carotid endarterectomy in patients without recent neurological symptoms: randomised controlled trial. Lancet. 2004;363:1491–502. doi: 10.1016/S0140-6736(04)16146-1. [DOI] [PubMed] [Google Scholar]
- 5.Centers of Medicare & Medicaid Services [Last accessed 12/Aug/2012];Pub 100-03 Medicare National Coverage Determinations. http://www.cms.gov/Regulations-and-Guidance/Guidance/Transmittals/downloads/R115NCD.pdf. [PubMed]
- 6.Giacovelli JK, Egorova N, Dayal R, Gelijns A, McKinsey J, Kent KC. Outcomes of carotid stenting compared with endarterectomy are equivalent in asymptomatic patients and inferior in symptomatic patients. J Vasc Surg. 2010;52:906–13. 913.e1–4. doi: 10.1016/j.jvs.2010.05.010. [DOI] [PubMed] [Google Scholar]
- 7.Giles KA, Hamdan AD, Pomposelli FB, Wyers MC, Schermerhorn ML. Stroke and death after carotid endarterectomy and carotid artery stenting with and without high risk criteria. J Vasc Surg. 2010;52:1497–504. doi: 10.1016/j.jvs.2010.06.174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Sidawy AN, Zwolak RM, White RA, Siami FS, Schermerhorn ML, Sicard GA. 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]
- 9.Yuo TH, Goodney PP, Powell RJ, Cronenwett JL. “Medical high risk” designation is not associated with survival after carotid artery stenting. J Vasc Surg. 2008;47:356–62. doi: 10.1016/j.jvs.2007.10.046. [DOI] [PubMed] [Google Scholar]
- 10.Mas JL, Chatellier G, Beyssen B, Branchereau A, Moulin T, Becquemin JP, et al. The EVA-3S Investigators Endarterectomy versus stenting in patients with symptomatic severe carotid stenosis. N Engl J Med. 2006;355:1660–71. doi: 10.1056/NEJMoa061752. [DOI] [PubMed] [Google Scholar]
- 11.Ringleb PA, Allenberg J, Berger J, Bruckmann H, Eckstein HH, Fraedrich G, et al. The The 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;368:1239–47. doi: 10.1016/S0140-6736(06)69122-8. [DOI] [PubMed] [Google Scholar]
- 12.Silver FL, Mackey A, Clark WM, Brooks W, Timaran CH, Chiu D, et al. The CREST Investigators Safety of stenting and endarterectomy by symptomatic status in the Carotid Revascularization Endarterectomy Versus Stenting Trial (CREST) Stroke. 2011;42:675–80. doi: 10.1161/STROKEAHA.110.610212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Yadav JS, Wholey MH, Kuntz RE, Fayad P, Katzen BT, Mishkel GJ, et al. The 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;351:1493–501. doi: 10.1056/NEJMoa040127. [DOI] [PubMed] [Google Scholar]
- 14.Ederle J, Dobson J, Featherstone RL, Bonati LH, van der Worp HB, de Borst GJ, et al. The International Carotid Stenting Study investigators Carotid artery stenting compared with endarterectomy in patients with symptomatic carotid stenosis (International Carotid Stenting Study): an interim analysis of a randomised controlled trial. Lancet. 2010;375:985–97. doi: 10.1016/S0140-6736(10)60239-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Brott TG, Hobson RW, II, Howard G, Roubin GS, Clark WM, Brooks W, et al. The CREST Investigators 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]
- 16.Sheffet AJ, Roubin G, Howard G, Howard V, Moore W, Meschia JF, et al. Design of the Carotid Revascularization Endarterectomy vs. Stenting Trial (CREST) Int J Stroke. 2010;5:40–6. doi: 10.1111/j.1747-4949.2009.00405.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Massop D, Dave R, Metzger C, Bachinsky W, Solis M, Shah R, et al. The SAPPHIRE Worldwide Investigators Stenting and angioplasty with protection in patients at high-risk for endarterectomy: SAPPHIRE Worldwide Registry first 2,001 patients. Catheter Cardiovasc Interv. 2009;73:129–36. doi: 10.1002/ccd.21844. [DOI] [PubMed] [Google Scholar]
- 18.Fitzgerald TN, Popp C, Federman DG, Dardik A. Success of carotid endarterectomy in veterans: high medical risk does not equate with high surgical risk. J Am Coll Surg. 2008;207:219–26. doi: 10.1016/j.jamcollsurg.2008.02.033. [DOI] [PubMed] [Google Scholar]
- 19.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;37:1191–9. doi: 10.1016/s0741-5214(03)00336-7. [DOI] [PubMed] [Google Scholar]
- 20.Bonati LH, Dobson J, Algra A, Branchereau A, Chatellier G, et al. The Carotid Stenting Trialist's Collaboration Short-term outcome after stenting versus endarterectomy for symptomatic carotid stenosis: a preplanned meta-analysis of individual patient data. Lancet. 2010;376:1062–73. doi: 10.1016/S0140-6736(10)61009-4. [DOI] [PubMed] [Google Scholar]
- 21.Vittinghoff E, McCulloch CE. Relaxing the rule of ten events per variable in logistic and Cox regression. Am J Epidemiol. 2007;165(6):710–8. doi: 10.1093/aje/kwk052. [DOI] [PubMed] [Google Scholar]
- 22.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;43:285–95. doi: 10.1016/j.jvs.2005.10.069. discussion 95–6. [DOI] [PubMed] [Google Scholar]