|Home | About | Journals | Submit | Contact Us | Français|
The safety of carotid artery stenting (CAS) and carotid endarterectomy (CEA) has varied by symptomatic status in previous trials. The Carotid Revascularization Endarterectomy versus Stenting Trial (CREST) data were analyzed to determine safety in symptomatic and asymptomatic patients.
CREST is a randomized trial comparing safety and efficacy of CAS versus CEA in patients with high-grade carotid stenoses. Patients were defined as symptomatic if they had relevant symptoms within 180 days of randomization. The primary endpoint was stroke, myocardial infarction, or death within the periprocedural period or ipsilateral stroke up to 4 years.
For 1321 symptomatic and 1181 asymptomatic patients, the periprocedural aggregate of stroke, myocardial infarction, and death did not differ between CAS and CEA (5.2% versus 4.5%; hazard ratio, 1.18; 95% CI, 0.82 to 1.68; P=0.38). The stroke and death rate was higher for CAS versus CEA (4.4% versus 2.3%; hazard ratio, 1.90; 95% CI, 1.21 to 2.98; P=0.005). For symptomatic patients, the periprocedural stroke and death rates were 6.0%±0.9% for CAS and 3.2%±0.7% for CEA (hazard ratio, 1.89; 95% CI, 1.11 to 3.21; P=0.02). For asymptomatic patients, the stroke and death rates were 2.5%±0.6% for CAS and 1.4%±0.5% for CEA (hazard ratio, 1.88; 95% CI, 0.79 to 4.42; P=0.15). Rates were lower for those aged <80 years.
There were no significant differences between CAS versus CEA by symptomatic status for the primary CREST endpoint. Periprocedural stroke and death rates were significantly lower for CEA in symptomatic patients. However, for both CAS and CEA stroke and death rates were below or comparable to those of previous randomized trials and were within the complication thresholds suggested in current guidelines for both symptomatic and asymptomatic patients.
Extracranial atherosclerosis is responsible for up to 10% of all ischemic strokes.(1) Clinical trials demonstrate that carotid endarterectomy (CEA) significantly reduces the risk of stroke in patients with recent stroke or transient ischemic attack and in patients without prior symptoms who have moderate to severe carotid stenoses.(2) The Carotid Revascularization Endarterectomy versus Stenting Trial (CREST) compared CEA and carotid artery stenting (CAS) in symptomatic and asymptomatic patients.(3) In this secondary analysis of data from CREST, we compare the periprocedural complications of CEA and CAS in symptomatic and asymptomatic patients. These results are discussed in relationship to prior clinical trials and current clinical practice guidelines.
CREST is a multicenter randomized clinical trial with blinded endpoint adjudications that compared the safety and efficacy of CAS versus CEA. Symptomatic patients were required to have ≥50% ipsilateral carotid stenosis by angiography, ≥70% by duplex ultrasound, or ≥70% by CT angiography or MR angiography if the stenosis on ultrasonography was 50% to 69%. Asymptomatic patients had to have ≥60% stenosis by angiography, ≥70% by ultrasound, or ≥80% by CT angiography or MR angiography if the stenosis on ultrasonography was 50% to 69%. Symptomatic patients had transient ischemic attack or ischemic stroke ipsilateral to a stenosed carotid artery within 180 days of randomization. The full eligibility criteria have been previously published.(4)
Patients were enrolled at 117 sites in the United States and Canada. Operators were carefully selected. Surgeons had to perform >12 CEAs annually. Interventionalists had to demonstrate experience in CAS, receive hands-on experience with the RX Acculink stent and the RX Accunet embolic-protection device, and/or successfully complete a lead-in phase.(5)
The primary endpoint was any stroke, myocardial infarction (MI), or death in the periprocedural period or ipsilateral stroke thereafter up to 4 years. Neurological evaluations were conducted preprocedurally, at 18 to 54 hours postprocedure, at 1 month, and every 6 months thereafter; cardiac enzymes were obtained preprocedurally, at 6 to 8 hours postprocedure, and for chest pain lasting >15 minutes; electrocardiograms were completed preprocedurally, 6 to 48 hours postprocedurally, and for chest pain lasting >15 minutes.
Intention-to-treat endpoint analyses adjusting for major baseline covariates were conducted using standard time-to-event statistical modeling. The periprocedural period was defined as the 30-day period after the procedure for those participants receiving their assigned procedure within 30 days, or 36 days after randomization for those participants not receiving their assigned treatment within 30 days. Because the periprocedural period was short, minimizing the need for censoring, event proportion and the absolute differences in event proportions were calculated as the percentage of patients with events. Standard survival techniques were used to estimate hazard ratios as an index of the relative risk of the two procedures.
A total of 1,321 symptomatic patients and 1,181 asymptomatic patients were enrolled in the study. There were no significant differences in age (mean, 69 years) and sex (64% male) between symptomatic and asymptomatic subjects (Table 1). Atherosclerotic risk factors were common in both subject groups, but more prevalent in the asymptomatic subjects. Asymptomatic subjects were more likely than symptomatic to have severe (≥ 70%) carotid stenosis (92% versus 80%).
The periprocedural primary endpoint rates (stroke, MI, or death) for the entire cohort did not differ between CAS and CEA (5.2% versus 4.5%; hazard ratio [HR], 1.18; 95% CI, 0.82 to 1.68; P=0.38). Similarly, there was no difference in the periprocedural primary endpoint rates for CAS versus CEA for either symptomatic patients (6.7%±1.0% versus 5.4%±0.9%; HR, 1.26; 95% CI, 0.81 to 1.96; P=0.30) or asymptomatic patients (3.5%±0.8% versus 3.6%±0.8%; HR, 1.02; 95% CI, 0.55 to 1.86; P=0.96; Table 2). The periprocedural rate of stroke and death was significantly higher in CAS versus CEA for symptomatic patients (6.0%±0.9% versus 3.2%±0.7%; HR, 1.89; 95% CI, 1.11 to 3.21; P=0.02) but not for asymptomatic patients (2.5%±0.6% versus 1.4%±0.5%; HR, 1.88; 95% CI, 0.79 to 4.42; P=0.15). The rate of MI was lower after CAS versus CEA for symptomatic patients (1.0%±0.4% versus 2.3%±0.6%; HR, 0.45; 95% CI, 0.18 to 1.11; P=0.08) and for asymptomatic patients (1.2%±0.3% versus 2.2%±0.6%; HR, 0.55; 95% CI, 0.22 to 1.38; P=0.20); however, the differences were not significant.
Table 3 reports event rates after removing the oldest patients (≥80 years) and allows the CREST data to be compared with previous clinical trials in which octogenarians were not enrolled or in which the number enrolled was not specified.(6-10) For the symptomatic cohort aged <80 years, the 30-day stroke and death rate was 2.6%±0.7% for CEA and 5.6%±1.0% for CAS. For the asymptomatic cohort aged <80 years, the 30-day stroke and death rate was 1.5%±0.5% for CEA and 2.4%±0.7% for CAS.
Other adverse events excluding stroke, MI, and death are outlined in Table 4. Neck hematomas (1.5%), surgical wound complications (1.6%), and cranial nerve palsies (4.7%) were expected complications of CEA. Femoral bleeding events (3.6% versus 1.2%) and non-hemorrhagic femoral complications (0.8% versus 0.2%) were higher for CAS. Bradycardia requiring a permanent pacemaker occurred in 0.5% of patients who underwent CAS. There were no differences in complication rates between symptomatic and asymptomatic patients.
These CREST safety results demonstrate the comparative safety of CAS and CEA in patients with carotid artery disease. The periprocedural stroke and death rates for CAS and CEA are the lowest reported from population-based studies or from large randomized trials of either symptomatic or asymptomatic patients undergoing carotid revascularization (Figure).(6-17) Furthermore, these rates are within the target of <6% for symptomatic patients suggested in the recent American Heart Association / American Stroke Association guidelines(18) and below the 3% target for asymptomatic patients as recommended by the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology.(19) Periprocedural stroke and death are even lower when the patients aged ≥80 years are excluded (Table 3), the population comparable to those of the North American Symptomatic Carotid Endarterectomy Trial and the Asymptomatic Carotid Atherosclerosis Study, 2 of the landmark studies from which guidelines have been derived.
Unlike the other published carotid revascularization trials,(13-16) CREST interventionalists were only permitted to use 1 stenting system and the corresponding distal protection device was used in 96% of the stented patients. The high use of embolic protection and the standardization of the stenting system may have contributed to the low periprocedural complication rate in CREST; however, this also limits the generalizability of the study to other stenting systems.
The CREST results imply that both CAS and CEA can be done with acceptable periprocedural risks by experienced surgeons and interventionalists. However, in the symptomatic patients, CREST surgeons performed CEA with a significantly lower periprocedural risk of stroke and death as compared with the interventionalists performing CAS. The HRs for CAS versus CEA were almost identical in the symptomatic and asymptomatic patients (1.89 versus 1.88), but the periprocedural stroke and death rate was lower for asymptomatic patients. These HRs, in favor of surgery, reached significance only in the symptomatic group. This advantage of surgery was counterweighed by a higher MI rate and the added risk of postoperative cranial nerve palsies. Cranial nerve palsies almost always resolve with time, and MI did not have the same impact as stroke on physical and mental health-related quality of life as assessed by the SF-36 scale.(3) Although the rate of periprocedural stroke was higher in stented patients for the symptomatic patients, the Figure shows that, like the periprocedural risk of stroke related to CEA, the rate has fallen over time. CEA has had a 40-year head start over CAS; patient selection, technique, and technology continue to improve.
CREST is currently the only published randomized clinical trial comparing CAS and CEA that has included conventional risk patients with both symptomatic and asymptomatic carotid disease. The study investigators, although remaining blinded to event rates, worked carefully with the Data and Monitoring Board to ensure that the anticipated lower event rates in the asymptomatic patients included in CREST after 2005 did not endanger the statistical power of the study. The inclusion of asymptomatic patients was considered paramount to ensure the generalizability of findings because it is estimated that ≥70% of patients treated with CEA in the United States are asymptomatic.(20)
Future treatment guidelines will likely call for safety results better than what we report here for CREST. The risk of stroke in the medical arms of the past CEA clinical trials has diminished over time, likely as a consequence of improvements in medical treatments for secondary stroke prevention such as better blood pressure control and the widespread use of statins. It is also reasonable to infer that improvements in medical treatments have led to improvements in the safety of CAS and CEA. In addition, improvements in technique and technology will impact both CAS and CEA over the coming decade. By 2020, it may be reasonable to require stroke and death rates <3% to 4% for symptomatic patients and 1% to 2% for asymptomatic patients. Even with these lower complication rates, as the effectiveness of medical therapies improve, additional clinical trials will be needed to demonstrate a benefit of carotid revascularization in asymptomatic patients in whom the risk of stroke is already low.
CREST has demonstrated that, with experienced surgeons and interventionalists, both CEA and CAS are viable options for carotid revascularization because the overall complication rates for both procedures are within current treatment guidelines. Although the primary complication rates (MI, stroke and death) were similar for CEA and CAS in both symptomatic and asymptomatic patients, the rate of stroke in the periprocedural period was higher for CAS for symptomatic patients. This difference may currently favor CEA; however, the disparity could potentially be reduced as new stent systems are introduced and as endovascular techniques improve.
Sources of Funding
Supported by the National Institute of Neurological Disorders and Stroke (NINDS) and the National Institutes of Health (R01 NS 038384); supplemental funding was received from Abbott Vascular Solutions, Inc. (formerly Guidant).
Clinical Trial Registration Information: ClinicalTrials.gov website Identifier: NCT00004732
Conflict of Interest/Disclosure(s)
F.L. Silver: Honoraria, Boehringer Ingelheim Canada, Merck Frosst Canada, Pfizer Canada, Servier Canada. Consultant/Advisory Board, Boehringer Ingelheim Canada. A. Mackey: None. W.M. Clark: None. W. Brooks: None. C.H. Timaran: Honoraria, Abbott Vascular. D. Chiu: None. L.B. Goldstein: Consultant/Advisory Board, Abbott Vascular, ACT-1 Trial Clinical Oversight Committee. J.F. Meschia: None. R.D. Ferguson: None. W.S. Moore: None. G. Howard: Consultant/Advisory Board, Bayer Healthcare, Member of the ARRIVE Executive Committee. T.G. Brott, for the CREST Investigators: Honoraria, Sahs Memorial Lecture University of Iowa, American Academy of Neurology 62nd Annual Meeting, Heritage Valley Health System, American Society of Neuroradiology 48th Annual Meeting, Massachusetts General Hospital, Pennsylvania Advances in Stroke.
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.
Frank L. Silver, University Health Network, University of Toronto, Toronto, ON, Canada.
Ariane Mackey, CHA, Hôpital de l'Enfant-Jésus, Québec QC, Canada.
Wayne M. Clark, Oregon Health and Science University, Portland, OR.
William Brooks, Central Baptist Hospital, Lexington, KY.
Carlos H. Timaran, University of Texas Southwestern Medical Center and VA North Texas Care System, Dallas, TX.
David Chiu, The Methodist Hospital, Houston, TX.
Larry B. Goldstein, Duke University Medical Center, Durham, NC.
James F. Meschia, Mayo Clinic, Jacksonville, FL.
Robert D. Ferguson, MetroHealth Medical Center, Cleveland, OH.
Wesley S. Moore, UCLA Medical Center, Los Angeles, CA.
George Howard, University of Alabama at Birmingham, Birmingham, AL.
Thomas G. Brott, Mayo Clinic, Jacksonville, FL.