Author + information
- Received August 9, 2013
- Revision received September 25, 2013
- Accepted October 4, 2013
- Published online February 1, 2014.
- Jay Giri, MD, MPH∗∗ (, )
- Kevin F. Kennedy, MS†,
- Ido Weinberg, MD‡,
- Beau M. Hawkins, MD‡,
- Marcella Calfon Press, MD, PhD§,
- Douglas Drachman, MD‡,
- Daniel J. McCormick, DO∗,
- Herbert D. Aronow, MD, MPH‖,
- Christopher J. White, MD¶,
- Kenneth Rosenfield, MD‡ and
- Robert W. Yeh, MD, MSc‡
- ∗Cardiovascular Medicine Division, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania
- †St. Luke's Mid-America Heart Institute, University of Missouri—Kansas City, Kansas City, Missouri
- ‡Division of Cardiology, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts
- §Ronald Reagan UCLA Medical Center, University of California–Los Angeles, Los Angeles, California
- ‖St. Joseph Mercy Hospital, Ann Arbor, Michigan
- ¶John Ochsner Heart and Vascular Institute, Ochsner Medical Center, New Orleans, Louisiana
- ↵∗Reprint requests and correspondence:
Dr. Jay Giri, Hospital of the University of Pennsylvania, 3400 Spruce Street, Gates Building 9053, Philadelphia, Pennsylvania 19104.
Objectives This study sought to characterize usage and outcomes of carotid stenting platforms.
Background A variety of stents and embolic protection devices (EPDs) are used for carotid artery stenting. Little is known about current usage patterns and differences in outcomes with these devices.
Methods We analyzed 12,135 consecutive carotid stent procedures in the NCDR (National Cardiovascular Data Registry) CARE (Carotid Artery Revascularization and Endarterectomy) registry performed between January 1, 2007 and March 31, 2012. We compared baseline characteristics and crude and multivariable-adjusted rates of in-hospital combined death/stroke among patients treated with Acculink/Accunet (Abbott Laboratories, Abbott Park, Illinois), Xact/Emboshield (Abbott), and Precise/Angioguard (Cordis Corporation, Bridgewater, New Jersey) stent/EPD combinations.
Results In 78.2% of cases, stents were used in conjunction with their specific, corresponding U.S. Food and Drug Administration–approved EPD. The Acculink/Accunet (n = 2,617, 21.6%), Xact/Emboshield (n = 3,507, 28.9%), and Precise/Angioguard (n = 2,696, 22.2%) stent/EPD combinations were used in 72.7% of all cases. The Protégé/SpiderFx (ev3 Endovascular Inc., Plymouth, Minnesota) (n = 453, 3.7%) and Wallstent/Filterwire (Boston Scientific, Natick, Massachusetts) (n = 213, 1.8%) devices were used in a minority of cases. In unadjusted analyses, the Precise/Angioguard system was associated with higher rates of the primary outcome than were the Acculink/Accunet (2.5% vs. 1.8%; p = 0.058) and Xact/Emboshield (2.5% vs. 1.9%; p = 0.14) systems that were not statistically different. In adjusted analyses, differences between Precise/Angioguard and Accunet/Acculink (odds ratio [OR]: 1.48, 95% confidence interval [CI]: 0.89 to 2.47; p = 0.065), Precise/Angioguard and Xact/Emboshield (OR: 1.16, 95% CI: 0.77 to 1.76; p = 0.38), and Xact/Emboshield and Accunet/Acculink (OR: 1.28, 95% CI: 0.82 to 1.97; p = 0.18) remained nonsignificant.
Conclusions In modern U.S. practice, the Acculink/Accunet, Xact/Emboshield, and Precise/Angioguard carotid stenting systems are used in most cases and are associated with similarly low rates of adverse events.
Carotid artery stenting (CAS) is a commonly used revascularization procedure to reduce the risk of stroke among patients with asymptomatic or symptomatic carotid stenosis. In the United States, nearly all procedures are performed with the use of embolic protection devices (EPDs), and EPD use is required for Center for Medicare and Medicaid Services reimbursement in cases of elective stenting (1). Operators may choose from multiple U.S. Food and Drug Administration (FDA)–approved and investigational stents and EPDs. A variety of factors may influence the choice of stent and EPD, including device availability, clinical trial or post-marketing registry participation, stent geometry considerations (i.e., open- vs. closed-cell structure or tapered vs. nontapered design), and specific EPD characteristics. Little is known about current U.S. usage patterns of the various carotid stents and EPDs. The CARE (Carotid Artery Revascularization and Endarterectomy) registry provides a unique opportunity to analyze contemporary usage patterns as well as the comparative effectiveness of the most commonly used carotid stenting/EPD systems.
The CARE registry is a national registry enrolling patients with carotid stenosis who have undergone revascularization with either carotid endarterectomy or CAS. It was created to monitor clinical practice, assess patient outcomes, and provide a framework for quality improvement initiatives (2). As of November 2012, the registry included 14,343 CAS procedures performed at 174 hospitals. The current research was supported by the American College of Cardiology Foundation's NCDR (National Cardiovascular Data Registry).
All patients undergoing CAS from January 2007 through March 2012 were initially evaluated for inclusion in this analysis. Patients with acute evolving stroke (n = 367, 2.7%), with spontaneous carotid artery dissection (n = 124, 0.9%), or undergoing general anesthesia (n = 543, 4.1%) were excluded from analysis, as these patients represented a distinct subgroup of patients with substantially higher procedural risk that were likely nonelective cases. Patients who received proximal embolic protection (n = 305, 2.42%) or distal occlusion balloon (n = 24, 0.19%) were excluded from this analysis due to the very low usage rates of these devices during the period analyzed. The primary outcome of interest was the occurrence of in-hospital major adverse events, defined as the composite of stroke and all-cause death. Stroke was defined as a new neurologic deficit persisting for more than 24 h.
For each patient, information on demographics, comorbid conditions, cardiac history, neurologic history, neurologic risk factors, anatomical and procedural information, and lesion characteristics were collected. Demographics and other general descriptive variables included age, sex, body mass index, and ethnicity. Comorbid conditions assessed included the following: hypertension, diabetes, renal insufficiency (glomerular filtration rate <60 ml/min), dyslipidemia, peripheral artery disease, chronic lung disease, smoking status, major surgery planned within 8 weeks, previous neck irradiation, and previous neck surgery. Cardiac history variables included ischemic heart disease, history of heart failure, and history of atrial fibrillation/flutter. Neurologic variables assessed included the following: dementia, seizure disorders, previous ipsilateral carotid endarterectomy, previous ipsilateral CAS, previous transient ischemic attack, previous ischemic stroke, and pre-procedure National Institutes of Health stroke scale. Site-reported anatomical and procedural variables included the following: target vessel, need for urgent cardiac surgery within 30 days, symptomatic for target lesion within 6 months, contralateral carotid artery occlusion, and aortic arch type. Lesion characteristics assessed included the following: presence of visible thrombus, dense calcification, target lesion location, lesion length, minimal luminal diameter, and pre-procedure stenosis severity. Carotid artery restenosis was defined as >50% diameter stenosis at or adjacent to the site previously treated with balloon angioplasty or stent.
An initial cross match was performed of stent type and distal filter EPD type. Baseline comparisons and unadjusted analyses of patients undergoing CAS using different stent/EPD combinations were performed using chi-square tests for categorical variables and t tests for continuous variables. We then conducted pairwise comparisons of adjusted rates of stroke or death using a logistic regression. For these comparisons, we adjusted for the predicted risk of in-hospital stroke or death using a validated CAS risk score that incorporated the following variables: impending major surgery, previous stroke, age, symptomatic lesion, atrial fibrillation, and absence of previous ipsilateral carotid endarterectomy (3). Additional analyses included an adjusted analysis of 30-day event rates for the portion of the cohort in whom this was available and a test for interaction of symptomatic/asymptomatic status with our comparison of device types. All analyses were conducted using SAS (version 9.2, SAS Institute, Cary, North Carolina).
The Acculink/Accunet (Abbott Laboratories, Abbott Park, Illinois) (n = 2,617, 21.6%), Xact/Emboshield (Abbott) (n = 3,507, 28.9%), and Precise/Angioguard (Cordis Corporation, Bridgewater, New Jersey) (n = 2,696, 22.2%) stent/EPD combinations were used in 72.7% of all cases; the Protégé/SpiderFx (ev3 Endovascular Inc., Plymouth, Minnesota) (n = 453, 3.73%) and Wallstent/Filterwire (Boston Scientific, Natick, Massachusetts) (n = 213, 1.76%) devices were used in fewer cases. In 78.2% of cases, stents were used in conjunction with their specific, corresponding FDA-approved EPD (Fig. 1). Baseline characteristics of the patients receiving the 3 most commonly used carotid stent systems are listed in Table 1.
In unadjusted analyses, the Precise/Angioguard system was associated with higher rates of the primary outcome than the Acculink/Accunet (2.5% vs. 1.8%; p = 0.058) and Xact/Emboshield (2.5% vs. 1.9%; p = 0.14) systems, although neither comparison met statistical significance. There was no difference in the primary outcome between the Acculink/Accunet and Xact/Emboshield systems (1.8% vs. 1.9%; p = 0.63). The Precise/Angioguard system was associated with a significantly higher rate of stroke than was the Xact/Emboshield system (2.3% vs. 1.7%; p = 0.02) and with a nonsignificant trend toward a higher stroke rate than the Accunet/Acculink system (2.3% vs. 1.6%; p = 0.06). There was no difference in the stroke rate between the Accunet/Acculink and Xact/Emboshield systems (1.6% vs. 1.7%; p = 0.84). Unadjusted outcomes are shown in Table 2.
In adjusted analyses, differences in the primary outcome between Precise/Angioguard and Accunet/Acculink (odds ratio [OR]: 1.48, 95% confidence interval [CI]: 0.89 to 2.47; p = 0.065), Precise/Angioguard and Xact/Emboshield (OR: 1.16, 95% CI: 0.77 to 1.76; p = 0.38), as well as those between Xact/Emboshield and Accunet/Acculink (OR: 1.28, 95% CI: 0.82 to 1.97; p = 0.18) were not significant (Fig. 2). A test for interaction of symptomatic/asymptomatic status with our comparison of device types was negative (p value for interaction = 0.76).
An additional secondary analysis was performed on patients in whom 30-day outcome data was available, which was 78% of the total cohort (Table 3). This demonstrated similar results to the primary in-hospital event analysis with a trend toward increased stroke/death with use of Angioguard/Precise that did not meet statistical significance after adjustment.
Our analysis revealed that, in the great majority of cases, carotid stents are used with their matched FDA-approved EPD. Several factors may influence this pattern of use. First, the FDA approves carotid stenting systems as a unit of stent and EPD. Operators may be more comfortable using an FDA-approved unit rather than “mixing and matching” stents with other EPDs. Additionally, reimbursement restrictions from the Center for Medicare and Medicaid Services likely have a large impact on this usage pattern. Currently, despite FDA approvals, Medicare reimburses patients with asymptomatic severe carotid stenosis for CAS only if they are enrolled in a pre-approval or post-marketing surveillance study. Clinical trials and post-marketing surveillance studies are sponsored by medical device companies that require use of a specific paired stent/EPD combination. The availability of such studies may greatly influence observed patterns of U.S. carotid stent/EPD use in the future (e.g., use rates of the Acculink and Xact stents may decrease following closure of the CHOICE [Carotid Stenting For High Surgical-Risk Patients; Evaluating Outcomes Through The Collection Of Clinical Evidence] post-marketing study, whereas those for Precise stents may increase given continuation of the SAPPHIRE [Stenting and Angioplasty With Protection in Patients at High Risk for Endarterectomy] WW [Worldwide] post-marketing study). It is possible that a strategy of selecting carotid stents and EPD individually according to procedural characteristics might benefit operators and patients. The current usage pattern is analogous to a situation in which coronary interventional operators could only use a single wire with a matching stent as opposed to choosing stents and wires independently of one another, each optimally suited for the particular anatomical situation encountered.
To our knowledge, this is the first large-scale assessment of carotid stenting device usage patterns. We sought to compare devices against one another in the fashion in which they are used in real-world clinical practice. Previous analyses of carotid stents have been smaller in scope and have grouped patients according to other designations such as open-/closed-cell stent type. Open-cell stents provide more conformability to vessel contours but cover less total vessel area. In contrast, closed-cell stents are more rigid, which may prove unfavorable for delivery and deployment in vessels with significant tortuosity, but may more completely cover carotid plaque.
Data regarding open- versus closed-cell stents has been mixed. Two previous analyses have suggested a benefit of closed-cell stents that may be more pronounced in patients with symptomatic lesions (4,5). In contrast, Schillinger et al. (6) analyzed a registry including 1,684 carotid stenting cases from 10 European centers and found no significant differences in adverse event rates between patients treated with open- and closed-cell stents. More recently, a small randomized trial showed no difference in transcranial Doppler-detected microembolization rates between the 2 stent types (7).
Traditionally, stents with free cell areas <5 mm2 have been classified as closed-cell, which is admittedly a somewhat arbitrary cutoff. The Xact stent has a free cell area of 2.74 mm2, and the Precise stent has a free cell area of 5.89 mm2. The Acculink stent has the largest free cell area of any currently available carotid stent at 11.48 mm2 (8). In the >9,000 patients analyzed in our study, outcomes were not statistically different between patients treated with each of these 3 stent types, representing the spectrum of available free cell areas, when the stent platform was used in concert with its respective EPD. This finding should provide support for operators to use their judgment to select the stent most favorable for a particular anatomic or clinical situation.
The various EPDs also possess unique technical characteristics that have been postulated to affect the outcomes of carotid stenting. There are several distal filter EPDs currently available, and they exhibit variability in pore size, crossing profile, and filter diameter. In addition, some EPDs feature a concentric design, in which the guidewire passes directly through the middle of the filter. Others possess an eccentric design, with a guidewire attached to an edge of the filter off the central axis.
Several previous analyses of EPDs have been performed. An analysis by Zahn et al. (9) revealed distal filter EPDs to be equally effective at preventing adverse events as a distal occlusion balloon. We observed in our current analysis of the CARE registry that distal occlusion balloons are almost never used in contemporary U.S. practice. Indeed, the PercuSurge GuardWire (Medtronic, Minneapolis, Minnesota) distal occlusion balloon is no longer clinically available in the United States. Previous studies of eccentric versus concentric filter designs have been limited in size or confounded by selection bias (4,10). The 3 EPDs that we evaluated all have concentric design with mild variance in pore size and crossing profiles. No significant differences were seen in outcomes with use of these devices in concert with their matching stents.
Our study has several strengths. In contrast to existing data, our analysis represents a modern U.S. experience. We have clarified usage patterns across a wide range of patient, institutional, and operator characteristics. Our study also represents the largest comparative analysis performed of patients undergoing CAS. Previous analyses have included less data and grouped patients according to designations such as open-/closed-cell stent type and distal balloon versus filter embolic protection. These comparative designations do not have as much relevance to contemporary practice given the previously mentioned usage patterns.
First, we were not able to assess separately outcome effects that could be individually associated with either a stent or an EPD. In nearly 80% of cases, stents were used with their matched FDA-approved EPDs. This precluded separate analysis of these 2 variables. It should be noted that this is an important confounder of previous comparative analyses of stents and EPDs. As the CARE registry currently has complete data available only for in-hospital outcomes, we cannot exclude the possibility of significant differences in either intermediate- or long-term outcomes, including restenosis rates between the stents/EPDs we compared. However, nearly 80% of our cohort had 30-day adverse event data available, and these results mirrored our in-hospital outcomes. We also cannot exclude the possibility of differences in outcome with eccentric versus concentric EPDs as all 3 of our analyzed EPDs have a concentric design; given the low event rates across our 3 groups, though, it is unlikely that eccentric distal filter EPDs would have a significantly positive impact on event rates. Nearly three-quarters of our patients had post-procedure National Institutes of Health stroke scales performed and recorded by certified personnel. The fact that independent neurologic assessment did not exist for every patient in this registry is in keeping with the experience in published and ongoing randomized trials of carotid artery stenting. Specific noninvasive imaging lesion characteristics that may be associated with a higher risk of stroke such as “hypoechoic plaque” and “intraplaque hemorrhage” were not available for analysis. Variability in carotid stenting technique, including the use of pre- and post-dilation, represents an additional potential confounder. Also, it is possible that a statistically significant difference in the primary outcome between the stent/EPD systems would emerge with analysis of several thousand additional cases. However, given the small absolute differences in adverse event rates seen in this study (<1%), such a difference is unlikely to have strong clinical relevance. Despite our assessment of 35 clinical variables, we also cannot exclude the possibility of unmeasured confounders influencing our analysis.
Overall, our analysis demonstrates that carotid stenting for symptomatic and asymptomatic disease is being performed with low adverse event rates across a wide range of institutions domestically using 3 particular stent/EPD systems. Periprocedural stroke prevention remains the most important goal in new technology evaluation for carotid stenosis. Our study suggests that a continued focus on stent and distal filter EPD characteristics is not likely to lead to dramatically improved event rates, and efforts at periprocedural stroke reduction can focus elsewhere. Given the known learning curve associated with carotid stenting, operator comfort with individual stenting systems and with case selection is likely of more import in reducing event rates than are technical differences between devices (11).
This research was supported by funding for statistical support from the National Cardiovascular Data Registry. The views expressed in this paper represent those of the authors and do not necessarily represent the official views of the NCDR or its associated professional societies identified on its website. Dr. Drachman has received research grants from iDev Technologies, Inc. and Lutonix/Bard; serves on the Clinical Events Committee for PLC Medical Systems, Inc.: and serves on the Data Safety and Monitoring Board for Prairie Education and Research Cooperative. Dr. McCormick has received research grants from Abbott Vascular Corp, W. L. Gore, and Boston Scientific Corp. Dr. Aronow is an unpaid consultant for Silk Road Medical, Inc. Dr. White is the Steering Committee Chair for the NCDR CARE registry; he accepts no compensation for this position. Dr. Rosenfield has received research grants from Abbott Vascular, Lutonix/Bard Peripheral Vascular, IDEV, Cordis, and Atrium; has received consulting/advisory board fees from Abbott Vascular, VORTEX/AngioDynamics, Complete Conference Management, and Becker Ventures; has equity in Medical Stimulation Corp., Angioguard (Cordis), and Micell; and serves on the board of directors for VIVA Physicians (501c3). Dr. Yeh has received institutional research support for and is an investigator at the Harvard Clinical Research Institute. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Sotirios Tsimikas, MD, served as Guest Editor for this paper.
- Abbreviations and Acronyms
- carotid artery stenting
- confidence interval
- embolic protection device
- U.S. Food and Drug Administration
- odds ratio
- Received August 9, 2013.
- Revision received September 25, 2013.
- Accepted October 4, 2013.
- American College of Cardiology Foundation
- Giri J.,
- Yeh R.W.,
- Kennedy K.F.,
- et al.
- Hawkins B.M.,
- Kennedy K.F.,
- Giri J.,
- et al.
- Schillinger M.,
- Gschwendtner M.,
- Reimers B.,
- Trenkler J.,
- Stockx L.,
- Mairet J.
- Bosiers M.,
- Deloose K.,
- Verbist J.,
- Peeters P.
- Zahn R.,
- Ischinger T.,
- Mark B.,
- et al.,
- for the ALKK Investigators