Author + information
- Received August 3, 2009
- Revision received February 17, 2010
- Accepted February 25, 2010
- Published online June 29, 2010.
- Theodore L. Schreiber, MD⁎,⁎ (, )
- Neil Strickman, MD†,
- Thomas Davis, MD‡,
- Vinay Kumar, MD§,
- Greg Mishkel, MD∥,
- Malcolm Foster, MD¶,
- Dennis Donohoe, MD⁎⁎,
- Suzanne Britto, RN⁎⁎,
- Gary Ansel, MD††,
- CASES-PMS Investigators
- ↵⁎Reprint requests and correspondence:
Dr. Theodore L. Schreiber, Detroit Medical Center, 8545 Common Road, Suite 150, Warren, Michigan 48093
Objectives The CASES-PMS (Carotid Artery Stenting With Emboli Protection Surveillance–Post-Marketing Study) multicenter, prospective, single-arm, surveillance study was designed to assess the safety and efficacy of carotid artery stenting (CAS) when performed by physicians with varied experience in CAS utilizing a formal training program. Whether the excellent results achieved at 30 days would be sustained to 1 year was the subject of the current investigation.
Background Previously, the pivotal SAPPHIRE (Stenting and Angioplasty with Protection of Patients with High Risk for Endarterectomy) trial demonstrated that CAS was not inferior to carotid endarterectomy (CEA) when performed by physicians experienced in carotid stenting.
Methods High surgical-risk patients with de novo atherosclerotic or post-endarterectomy restenotic lesions in native carotid arteries were enrolled at participating centers. Inclusion and exclusion criteria matched those of the SAPPHIRE trial. The primary end point was a composite of 30-day major adverse events (MAE) including death, any stroke, or myocardial infarction.
Results A total of 1,492 patients were enrolled at 73 sites. The primary end point of 30-day MAE was 5.0%, meeting criteria for noninferiority to the prespecified objective performance criteria (OPC) established by the SAPPHIRE trial. The 1-year cumulative percentage of MAE was 12.5% by Kaplan-Meier analysis. All strokes to 30 days plus ipsilateral stroke between 31 and 360 days with CASES-PMS (5.4%) was similar to the rate seen with the SAPPHIRE trial stent cohort (4.9%). There were no significant differences in outcomes at 1 year by symptom status and high-risk status.
Conclusions With the formalized training program utilized in this study, physicians with varied experience in carotid stenting can achieve similar short- and longer-term results to the highly experienced SAPPHIRE Investigators. (Carotid Artery Stenting With Emboli Protection Surveillance–Post-Marketing Study [CASES-PMS]; NCT00231231)
Multiple registries have reported favorable periprocedural results after carotid artery stenting (CAS) with distal embolic protection for those patients considered at high risk for complications following carotid endarterectomy (CEA) (1–12). One unique trial examining carotid stenting in a randomized fashion versus endarterectomy for high surgical risk patients, the SAPPHIRE trial (Stenting and Angioplasty With Protection of Patients With High Risk for Endarterectomy trial), did demonstrate that patients randomized to CAS with distal embolic protection had efficacy and outcomes at 3 years that were not inferior to patients randomized to CEA (13). Thirty-day outcomes have previously been reported with the CASES-PMS (Carotid Artery Stenting With Embolic Protection Surveillance–Post-Marketing Study) (10), which was initiated as a condition-of-approval study under investigational device exemption from the Food and Drug Administration. This study demonstrated that despite a large cohort of sites and physicians, who for the most part are not CAS trialists with a broad range of experience, when trained in carotid stenting, obtained safety and efficacy outcomes similar to those physicians expert in carotid stenting who participated in the pivotal SAPPHIRE trial (13,14). Whether the excellent results achieved at 30 days would be sustained to 1 year was the subject of the current investigation and report.
Details of the research methods have previously been described in the 30-day outcome report (10).
Centers that participated in this study performed either low (<30), medium (30 to 99), or high (≥100) volume annual carotid stenting procedures, including both academic and nonacademic sites, and were located in different geographic locations within the U.S. All sites received institutional review board approval to participate in this study.
Physicians were required to have prior expertise in endovascular procedures and to have completed the Carotid Artery Stenting Education System (CASES). The amount of training depended upon the physician's previous experience in both carotid stenting and in the use of the Cordis ANGIOGUARD XP emboli protection device (Bridgewater, New Jersey). Level 1 investigators were exempt from training based on having performed at least 25 CAS procedures with complication rates at or below those achieved in the SAPPHIRE trial, and completion of more than 10 procedures using the ANGIOGUARD device. Level 2 investigators had previously performed at least 25 CAS procedures with complication rates at or below those achieved in the SAPPHIRE trial, but completed fewer than 10 procedures using the ANGIOGUARD device; these physicians received training that included online didactic training and proctoring by a physician experienced in carotid stenting using the study devices. Level 3 investigators were those who had performed fewer than 25 CAS procedures; these physicians underwent full CASES training that included 2 days of didactic and simulator instruction at a regional education center, in addition to the training provided to Level 2 physicians.
Definitions of training levels were based on recommendations by a physician advisory board and the Society for Cardiovascular Angiography and Interventions/Society for Vascular Medicine and Biology/Society for Vascular Surgery requirements for performance of carotid stenting, which include a minimum of 25 supervised carotid interventions with at least 13 as the primary operator.
Inclusion criteria matched those of the SAPPHIRE trial and included patients greater than 18 years of age with de novo atherosclerotic or post-endarterectomy restenotic obstructive lesions in native carotid arteries with either ≥50% carotid stenosis if symptomatic or ≥80% carotid stenosis if asymptomatic, determined by ultrasound or angiogram. In addition, patients were required to have at least 1 anatomic or physiologic risk factor that placed them at high risk for surgery (Tables 1 and 2).⇓Exclusion criteria included an acute or recent stroke (<48 h), intraluminal thrombus, total carotid occlusion, ostial common carotid artery lesion, intracranial mass lesion, arteriovenous malformation, prior stent placement in the target artery, stent placement in the contralateral carotid artery within the past 30 days, reference segment diameter <4 or >9 mm, intracranial aneurysm >9 mm in diameter, life expectancy <1 year, allergies to components of the stent or emboli protection device, or arterial access issues precluding percutaneous intervention.
Aspirin (81 to 325 mg orally per day) was given at least 72 h prior to the procedure, and continued indefinitely. Either ticlopidine (250 mg orally twice a day) or clopidogrel (75 mg orally once a day, with a loading dose of 300 mg if begun within 48 h of the procedure) was given and continued for 2 weeks following the procedure. Unfractionated heparin was administered during the procedure to maintain an activated clotting time (ACT) ∼300 s; glycoprotein IIb/IIIa inhibitors were used at the discretion of the operator. Distal emboli protection was achieved by placement of the ANGIOGUARD XP Emboli Capture Guidewire, an umbrella-shaped structure consisting of a 0.014-inch guidewire and a nitinol self-expanding filter basket containing 100-μm pores, distal to the lesion prior to stenting. Carotid artery stenting was performed using the self-expanding PRECISE Nitinol Stent System (Cordis). A patient was considered enrolled into the CASES-PMS study once either study device had entered the body.
Patients received a carotid ultrasound and/or angiogram at baseline, at discharge, and at 30-day follow-up. Baseline, discharge, and 30-day (+14-day window) neurological examination by an independent neurologist were mandated by the protocol. Stroke scale examinations (National Institutes of Health, Barthel, and Rankin stroke scale quantitative assessments) were performed if the patient had evidence of a stroke on neurologic examination. Follow-up visits were conducted with the investigators for clinical event tracking at 1, 6, and 12 months. All interim hospitalization records were analyzed. A clinical events committee, independent of the sponsor, adjudicated all major adverse events (MAE). Data were monitored independent of the sponsor and collected and analyzed by an independent data management group (Harvard Clinical Research Institute, Boston, Massachusetts).
The primary end point of this study was the cumulative rate of MAE, defined as death, myocardial infarction (MI), or all stroke at 30 days following the procedure. The MAE rate at 1 year included death, MI, and all stroke to 30 days plus all-cause death and ipsilateral stroke from 31 to 360 days. Stroke was defined as a focal neurological deficit lasting more than 24 h and was classified as major or minor using the National Institutes of Health Stroke Scale (NIHSS) and the Modified Rankin and Barthel scales (major: Barthel ≤60, NIHSS ≥15, or Rankin > 2; minor: Barthel ≥90, NIHSS ≤4, and Rankin ≥2). MI was determined when signs or symptoms of cardiac ischemia were present and was defined as creatinine kinase (CK) elevation >2 times the upper limit of normal with a CK-MB fraction above normal. Electrocardiograms were evaluated by an independent core laboratory.
The 1-year MAE rate was determined and compared to the 30-day MAE rate. Both were compared to an objective performance criterion (OPC) derived from the combined 30-day MAE rates from the SAPPHIRE randomized and nonrandomized stent arms of 6.3%. The pre-specified noninferiority margin was 3.0%, based on the “50% rule” (50% of the treatment effect in the control arm) recommended by the Federal Drug Administration for statistical reasoning noninferiority margin determination. Noninferiority would be established if the upper bound of the 1-sided confidence interval (CI) of the MAE rate excluded the OPC + margin (9.3%).
Descriptive summary statistics were calculated: frequencies, percentages for categorical variables, and mean, standard deviations for continuous variables. Categorical variables were compared using the Fisher exact test, and continuous variables were compared using ttests. The cumulative incidence of MAE and other adverse events during the 1-year follow-up period were analyzed using the actuarial lifetable method. The difference between the survival curves were compared using the Wilcoxon as well as the log-rank tests. Odds ratios and their 95% CIs were summarized for the predictors selected through multivariate predictor analysis for MAE at 360 days. All statistical analyses were performed using SAS version 8.2 (SAS Institute, Cary, North Carolina).
The following subgroup comparisons were pre-specified for observation, but were not powered to detect statistically significant differences in MAE rates between subgroups due to small sample size: site academic versus nonacademic status and geographic location, and investigator level of device training and volume.
Seventy-three sites enrolled a total of 1,492 patients between August 2003 and October 2005. Baseline characteristics of these patients are compared with those of the SAPPHIRE trial stenting and endarterectomy arms (Table 3)and previously commented upon. Patients in CASES-PMS had a higher incidence of diabetes and were older compared with patients in the SAPPHIRE trial. Conversely, patients in SAPPHIRE had a higher incidence of contralateral carotid occlusion or coexistent severe cardiac and carotid disease requiring open heart surgery and carotid revascularization as compared with patients in CASES-PMS. The mean percent diameter stenosis was 85.6 ± 8.6%.
Baseline and hospital discharge information was available on all 1,492 patients enrolled; 30-day clinical follow-up was available for 99.2% of those patients; and 1-year follow-up for 95.2% patients. As previously reported, the primary end point of the cumulative incidence of MAE at 30 days was 5.0% (95% CI: 3.9% to 6.2%, p < 0.001), meeting the criteria for noninferiority to the pre-specified OPC + margin (9.3%) established from the stent cohort of the SAPPHIRE trial. At 1 year, the cumulative percentage of MAE defined as death, MI, and all stroke to 30 days plus all-cause death and ipsilateral stroke from 31 to 360 days was 12.5% by Kaplan-Meier analysis, similar to the SAPPHIRE stent cohort of 12.2% (Fig. 1A).Cumulative percentage of death to 1 year was 8.1% (Fig. 1C). Cardiac-related deaths accounted for 2.9%, noncardiac-related death 3.2%, and neurological death 1.6% of all deaths. All stroke (ipsilateral and contralateral) to 30 days plus ipsilateral strokes between 31 to 360 days with CASES-PMS (5.4%) were similar to rates seen with the SAPPHIRE trial stent cohort (4.9%) (Fig. 1B). Ipsilateral stroke between 31 to 360 days was relatively rare (1.0%). Target lesion revascularization at 1 year (Fig. 1D) was low (0.6%) and consistent with results previously reported in the SAPPHIRE randomized trial.
For the overall population, 1-year safety and efficacy outcomes did not significantly differ among physician training levels, carotid stent volume by site, or site characteristics (academic vs. nonacademic and geographic location) (Tables 4 to 7⇓⇓⇓). Although nonsignificant, there was a slight increase in MAE with less experienced physicians and at low volume sites; however, the stroke rate at low-volume sites consisted of only minor strokes, whereas both major and minor strokes occurred in the higher-volume sites. In a higher-risk subset such as octogenarians, there was a significant difference seen in the 1-year MAE rate between patients that were ≤80 years of age versus those >80 years of age treated by Level 1 and 2 physicians, but this was not seen with Level 3 physicians, who underwent the more extensive training program (Level 1: 9.6% vs. 15.6%, p = 0.0180), (Level 2: 9.2% vs. 18.8%, p = 0.0172), (Level 3: 13.5% vs. 13.4%, p = 1.000), respectively. Importantly, these data were purely observational, and subgroups were underpowered to draw conclusions.
Previous analysis of the 30-day risk of death and stroke showed a nonsignificant difference in patients with symptomatic (5.6%, 18 of 322) and asymptomatic (4.2%, 49 of 1,157; p = 0.2917) stenoses. More specifically, death and stroke in symptomatic patients with pure anatomic risk (2.6%, 2 of 76) and symptomatic patients with pure physiologic risk was (7.4%, 11 of 149; p = 0.2276); and asymptomatic patients with pure anatomic risk (2.7%, 7 of 258) and asymptomatic patients with pure physiologic high risk (5.8%, 25 of 428; p = 0.0635), did not significantly differ. The 1-year cumulative incidence of all death and stroke (all stroke to 30 days plus ipsilateral stroke 31 to 360 days) in symptomatic patients with either pure anatomic or pure physiologic risk (p = 0.1502), or in asymptomatic patients with either pure anatomic or pure physiologic risk (p = 0.2446) continued to show nonsignificant trend (Figs. 2Aand 2B).
Predictors of adverse outcomes at 360 days were advanced age, renal insufficiency, history of congestive heart failure, including class III symptoms, requirement for coronary artery bypass grafting, age >80 years, arrhythmias, history of aortic or mitral valve stenosis, history of MI, and neurological symptomatic status (Fig. 3).Predictors of new stroke included age >80 years, previous cerebrovascular accidents, or transient ischemic attacks.
The 30-day results of CASES-PMS have demonstrated that the safe and effective treatment of carotid artery stenosis may be accomplished using the PRECISE nitinol stent and the ANGIOGUARD XP emboli capture guidewire in both symptomatic and asymptomatic patients who are at high risk for endarterectomy (10). Prospective follow-up of this large series of 1,492 patients to 360 days indicates that safety and efficacy is maintained with an interim incidence of additional deaths from all causes of approximately 6.6% between 31 and 360 days. The overall incidence of nonneurological death, which accounts for the vast majority of adverse events between 31 and 360 days, is consistent with baseline characteristics of these patients, of which over 25% were over age 80 and the majority of whom had either congestive heart failure or advanced left ventricular dysfunction. Between 31 and 360 days, the incidence of ipsilateral minor or major stroke was 1.0%, and the incidence of target vessel revascularization was 0.6%, all supporting the durability of carotid artery stenting.
The MAE rate, namely death, MI, and stroke at 30 days plus all-cause death and ipsilateral stoke from 31 to 360 days was acceptably low at 12.5%. This event rate is similar to the SAPPHIRE randomized stent cohort (14) and other pivotal registries of high-risk carotid artery stenting (3,11,12). The ARCHeR (ACCULINK for Revascularization of Carotids in High-Risk Patients) trial results demonstrated a 1-year composite primary end point (30-day death/stroke/MI plus ipsilateral stroke at 1 year) of 9.6%, with an ipsilateral stroke rate of 1.3% between 30 days and 1 year (3). In the BEACH (Boston Scientific EPI: A Carotid Stenting Trial for High-Risk Surgical Patients) trial, the primary end point (30-day stroke/death/or Q-wave MI plus ipsilateral stroke or neurological death to 1 year) occurred in 8.9 % (11). An 11.6% death, stroke, and MI rate was seen in the CABERNET (Carotid Artery Revascularization Using the Boston Scientific EPI Filter Wire EZ and the EndoTex NexStent) trial at 1 year (12). The CASES-PMS study did differ from other carotid stent registries in that the study was conducted expressly under an investigational device exemption and included site monitoring, full adjudication of all events, and assessment of neurological status by an independent neurologist. This led to a complete end point assessment and accuracy with a very high compliance rate.
Overall, outcomes in this study were similar between operators with a wide range of experience in carotid stenting following a formal training program and physicians in SAPPHIRE who were experienced in carotid stenting. Subgroup analysis by physician experience was not powered to conclude statistical significance between training levels, but was strictly observational. This lack of difference between training groups could also conceptually be secondary to the volume definitions being set too low as to allow for real differences in skill sets. The early skill definitions were set arbitrarily without any data. There is some data to suggest that skill sets for carotid stenting continue to increase beyond the set number of 25. Verzini et al. (15) attempted to define the number of cases necessary in order to perform safe CAS with a 2% threshold rate for major stroke, as achieved for CEA at their vascular surgery center. It was only after the first 195 CAS procedures that the risk of major stroke remained stable at less than 2%. Lin et al (16). reported the treatment outcomes of 200 consecutive CAS procedures, dividing the series into 4 sequential time periods of 50 interventions. A significant reduction in the 30-day stroke and death rate from 8% to 0% (p < 0.05) was seen between the first and last 50 procedures.
Based on the perioperative stroke or death rate in standard-risk symptomatic patients from NASCET (North American Symptomatic Carotid Endarterectomy Trial) (17) and standard-risk asymptomatic patients from ACAS (Asymptomatic Carotid Atherosclerosis Study) (18), the 6% and 3% targets have been adopted as the threshold for acceptable perioperative event rates in high-risk patients treated with CAS. Perioperative stroke or death rates in symptomatic patients treated in CASES-PMS were well under the 6% threshold in patients with pure anatomic high-risk factors (2.6%), but higher in the symptomatic physiologic risk population (7.4%). In asymptomatic patients, perioperative stroke or death was also below the 3.0% threshold in patients with anatomic high-risk factors (2.7%), but this was not seen in the physiologic high-risk subgroup (5.8%). Therefore, even in this population of patients at high risk for surgery, carotid stenting has demonstrated acceptable perioperative risk in symptomatic and asymptomatic patients with anatomic risk factors compared with CEA in a standard-risk population. At 1 year, there were no significant differences in the cumulative incidence of death or stroke between anatomic and physiologic high-risk patients for both the symptomatic and asymptomatic subpopulations.
CASES-PMS was a prospective registry and as such was not randomized. Comparisons were made to an OPC that was based on the previously completed SAPPHIRE trial with similar inclusion and exclusion criteria. As such, in the absence of a randomized concurrent endarterectomy surgical control group, the potential for confounding differences exist. However, the very large size of this registry and high adherence to protocol would make this potential low.
With the formalized training program utilized in this study, physicians with varied experience in carotid stenting can achieve short- and longer-term results similar to the highly experienced SAPPHIRE Investigators. One-year outcomes with CASES-PMS showed little additional neurological morbidity and mortality beyond an acceptable 30-day outcome previously reported, and largely reflects the natural history of a large cohort of patients with a high prevalence of advanced cardiovascular disease at baseline. The 1-year stroke or death rate did not significantly differ between patients with anatomic and physiologic risk factors for both symptomatic and asymptomatic stenoses.
For a complete list of the investigators who participated in CASES-PMS, please see the online version of this article.
Dr. Strickman has received lecture fees from Cordis, Boston Scientific, and Gore. Dr. Kumar has received honoraria from Cordis Corporation for proctoring carotid artery stenting training. Dr. Foster has received honoraria for course director activities. Dr. Donohoe has since retired from Cordis Corporation. Ms. Britto is a full-time employee of Cordis Corporation. Dr. Ansel has a consulting agreement and is on the advisory board for Cordis Corporation, Abbott, and Boston Scientific, and receives honoraria for proctoring carotid artery stenting training and providing regional education center training at his institution.
- Abbreviations and Acronyms
- carotid artery stenting
- carotid endarterectomy
- major adverse events
- myocardial infarction
- National Institutes of Health Stroke Scale
- objective performance criterion
- Received August 3, 2009.
- Revision received February 17, 2010.
- Accepted February 25, 2010.
- American College of Cardiology Foundation
- Safian R.D.,
- Bacharach J.M.,
- Ansel G.M.,
- et al.
- Ouriel K.,
- Wholey M.H.,
- Fayad P.,
- et al.
- White C.,
- Iyer S.S.,
- Hopkins L.N.,
- et al.
- Safian R.D.,
- Bresnahan J.F.,
- Jaff M.R.,
- et al.
- Gray W.A.,
- Yadav J.S.,
- Verta P.,
- et al.,
- CAPTURE Trial Collaborators
- Iyer S.S.,
- White C.J.,
- Hopkins L.N.,
- et al.