Surgeon’s 30-Day Outcomes Supporting the Carotid Revascularization Endarterectomy versus Stenting Trial

Importance  While the Carotid Revascularization Endarterectomy Versus Stenting Trial (CREST) has been widely accepted as a landmark trial establishing an equivalent risk of major adverse events following carotid endarterectomy (CEA) or carotid artery stenting (CAS), the applicability of these findings to single centers has been questioned owing to the rigid selection criteria for investigators in the study. Although refuted by the findings of a subsequent study, a substudy of CREST established a higher periprocedural stroke rate for CAS when the surgeon was a vascular surgeon.

Objective  To present our 30-day results of stroke, death, myocardial infarction, and composite major adverse events to determine if a single vascular surgeon’s outcomes at our hospital are consistent with the results of CREST.

Design, Setting, and Participants  A retrospective analysis of patients with high-grade carotid artery stenosis treated with CEA or CAS by a vascular surgeon at our institution from September 9, 2005, through December 17, 2012, was performed. A χ² analysis was used to compare the incidence of specific high-risk patient characteristics in each group. The Fisher exact test was used to compare the risks of stroke, death, myocardial infarction, and composite major adverse events between CEA and CAS. These results were then compared with those reported in CREST.

Results  A total of 182 cases (94 CAS and 88 CEA) performed by a single vascular surgeon were included for analysis. While in CREST the periprocedural risk of stroke was higher following CAS (4.1% vs 2.3%, P = .01) and the risk of myocardial infarction was higher following CEA (2.3% vs 1.1%, P = .03), there was no significant difference in the incidence of these outcomes between the 2 treatment modalities in our study. When compared with CREST, our rates of myocardial infarction, stroke, death, and composite adverse events (CEA, 4.5% vs 3.4%; P = .79; CAS, 5.2% vs 4.3%; P >.99) were no different.

Conclusions and Relevance  Similar to CREST, the 30-day risk of composite major adverse events was equivalent for the 2 treatment modalities. We attribute our comparable incidence of perioperative stroke with CAS and CEA to improved patient selection. We excluded most patients older than 80 years and those with complex anatomy from consideration for CAS. Our results confirm those of CREST and demonstrate that both CEA and CAS can be performed safely by a vascular surgeon in properly selected patients.

Carotid artery disease is a prevalent, easily recognizable, and preventable cause of ischemic stroke. A number of major studies^1,2 have demonstrated the efficacy of carotid endarterectomy (CEA) in stroke prevention among patients with symptomatic carotid artery disease. Several randomized trials have assessed the role of carotid artery stenting (CAS) compared with conventional CEA.^3-5 The results from the recently completed, multicenter, randomized clinical Carotid Revascularization Endarterectomy versus Stenting Trial (CREST)¹ were the most revealing, with no significant short- or long-term differences in a combined primary endpoint of stroke, death, and myocardial infarction (MI) between patients who received CEA or CAS with distal protection. However, the arduous training program required of the surgeons and interventionalists selected to participate has raised concern regarding the generalizability of these results in the population at large. Furthermore, in the lead-in phase of this trial, the periprocedural risk of stroke following CAS was higher than with CEA when performed by vascular surgeons compared with interventional cardiologists and neuroradiologists.

While the findings reported in CREST and other subsequent trials are encouraging, it is unknown whether these results are attainable by a single surgeon performing both interventions. Thus, the true applicability of these studies to individual centers and surgeons needs validation. Accordingly, we reviewed our outcomes of CAS and CEA in patients with high-grade carotid artery stenosis to determine whether a single surgeon’s (M.B.M.) experience paralleled the findings published in CREST and to establish the safety of CEA and CAS when performed by experienced vascular surgeons.

A retrospective observational analysis of patients with high-grade carotid artery stenosis treated by CEA or CAS was conducted at Johns Hopkins Bayview Medical Center between 2005 and 2012. If patients were considered suitable candidates for either CEA or CAS, they were randomized before intervention according to the protocols outlined in CREST or the Asymptomatic Carotid Trial and were included in this study for analysis. Patients in our investigation who were not candidates for CEA (and thus did not meet criteria for inclusion in CREST) were then randomized to one of several CAS-only trials (CAPTURE II [Carotid RX ACCULINK/RX ACCUNET Post-Approval Trial to Uncover Unanticipated or Rare Events], CHOICE [Carotid Stenting For High Surgical-Risk Patients; Evaluating Outcomes Through The Collection Of Clinical Evidence], SAPPHIRE W [Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy], or FREEDOM [Flow Reversal System and GORE Embolic Filter Extension Study for the Ongoing Collection of Patient Outcomes]).

Although some patients included in this analysis were randomized in CREST or the Asymptomatic Carotid Trial, many others participated in nonrandomized trials; therefore, we did not believe this study represented a true randomization. The electronic medical record was referenced to obtain basic demographic information as well as the operative notes and results of follow-up studies. High-grade stenosis was defined as more than 70% stenosis by carotid artery duplex. Only patients who received CEA or CAS performed by a single vascular surgeon at our institution were included for evaluation. Of note, while this surgeon was a CREST contributor, he was not subjected to any training requirements. Patients who received a revision or a second surgical procedure for carotid disease were excluded. The primary outcomes of this analysis were 30-day (periprocedural) major adverse events (MAEs), including stroke, MI, and death. Approval for this study was obtained from the Johns Hopkins Hospital Institutional Review Board.

All patients received the usual preoperative aspirin, statin, and β-adrenergic blockade. Neurovascular monitoring was performed with intraoperative somatosensory evoked potentials. An intraluminal shunt was placed during CEA if changes to the somatosensory evoked potentials were appreciated. Anticoagulation was established with intravenous heparin. The plaque was elevated and dissection continued into the internal carotid artery until a smooth taper was achieved. Patch angioplasty with Dacron or bovine pericardium was then performed with running suture. A completion duplex was performed on all patients.

All patients received clopidogrel bisulfate (Plavix) and acetylsalicylic acid (aspirin) 5 days before their procedure. A duplex-guided micropuncture technique was used with local anesthesia for transfemoral access. An activated clotting time of 250 to 300 seconds was maintained with intravenous heparin. Carotid and cerebral angiograms were performed to confirm the degree and extent of stenosis as well as carotid tortuosity and baseline cerebral perfusion. Patients with severe aortic atherosclerosis or tortuosity (type III arch) were approached via a transcervical carotid artery cutdown. Carotid artery stenting was not performed if the patient had less than 80% asymptomatic stenosis measured by North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria. The use of a combination of stents, embolic protection devices, and flow reversal systems depended on patient enrollment in the aforementioned stent trials. All patients received angioplasty with a 2- to 3-mm balloon, followed by stent placement. Poststent ballooning was performed only in selected patients with severe residual stenosis. Completion cerebral angiogram was performed after retrieval of the filter.

A thorough neurovascular examination was performed by the surgical staff at completion of the procedure. Patients were then monitored overnight in a surgical intensive care unit with hourly neurologic assessments. Baseline electrocardiograms were performed as part of the preoperative evaluation. Admission electrocardiograms were obtained on arrival to the intensive care unit. Cardiac enzymes were not ordered routinely but were reserved for patients who experienced chest pain or any other signs of angina. Patients were discharged when the standard criteria were met and were then evaluated in the clinic for follow-up. Stroke was defined by any change in neurologic examination findings confirmed by independent assessment by the neurology staff and substantiated by radiographic findings on magnetic resonance imaging. Myocardial infarction was defined by an elevation in cardiac enzymes with or without concomitant changes on electrocardiogram.

Stata SE, version 12.1 (StataCorp), was used for data analysis. Continuous variables were evaluated by t test and are reported as mean (SD). Dichotomous variables were evaluated for significance with the Fisher exact test owing to the small sample size of our study. Pearson χ² analysis was used for analysis of the data published in CREST. The incidence of each outcome (MI, stroke, or death) as well as a composite of any of these MAEs was compared between the CEA and CAS groups as well as between our institution and CREST to determine if there were any differences. Values were deemed significant at P < .05.

From September 9, 2005, through December 17, 2012, a total of 182 consecutive patients who received CEA (48.4%) or CAS (51.6%) by a single vascular surgeon at the Johns Hopkins Bayview Medical Center were analyzed. The mean (SD) ages in the CAS and CEA groups were 67.8 (10) years and 66.8 (10) years, respectively (P = .46); 67.0% of the patients in the CEA group were male compared with 47.9% in the CAS group (P = .01). Symptomatic disease was present in 34.0% (32 of 94) and 26.1% (23 of 88) of those who received CAS and CEA, respectively (P = .25). Hyperlipidemia (P < .001) and chronic obstructive pulmonary disease (P = .01) were more common in the endarterectomy cohort (see Table 1 for detailed patient demographics).

There were 2 strokes (2.3%) and 1 MI (1.1%) in the CEA group. The CAS group experienced 1 stroke (1.1%), 1 death (1.1%), and 2 MIs (2.1%). The overall mortality in those who received intervention was 0.55% (1 of 182). In all, there was no significant difference in the occurrence of MI, stroke, death, or combined MAEs at our institution between the CEA and CAS groups (Table 2). These results were then compared with those observed in CREST. Again, there were no significant differences in the incidence of MI, stroke, death, or combined MAEs between our institution and CREST for either CEA or CAS (Table 3).

Subgroup analysis of those with symptomatic carotid artery disease at our institution demonstrated no difference in periprocedural outcomes between CEA and CAS (Table 4). The incidences of periprocedural MI, stroke, and death in this cohort were similar to those outlined in CREST (Table 5).

The efficacy and durability of CEA in reducing the risk of ischemic stroke in patients with significant carotid artery disease has been demonstrated in several large randomized clinical trials.^1,2 The criteria for surgical intervention have been firmly established by these studies and have been further substantiated by longer-term follow-up.⁶ In some high-risk patient populations, such as those with surgically inaccessible lesions, those with radiation-induced stenosis, those receiving intervention for re-stenosis after a previous repair, and those with prohibitively severe cardiac morbidity, angioplasty with stenting has been proposed as a safer, less invasive technique for managing clinically significant carotid disease.^7,8 The effect of this paradigm shift has been recently investigated in several large trials and meta-analyses.^3-5,9

CREST has emerged as the most recent landmark study, to our knowledge, investigating the safety and efficacy of CAS in contrast to the criterion standard of CEA. It revealed similar rates of the composite outcome of MI, stroke, and death between the 2 groups in the periprocedural (hazard ratio [HR], 1.18; 95% CI, 0.82-1.68; P = .38) and 4-year periods (HR, 1.11; 95% CI, 0.81-1.51; P = .51).³ There was, however, a greater incidence of stroke in the endovascular group and MI in the CEA group at both time intervals.³ In addition, the lead-in phase of CREST demonstrated a higher event rate when CAS was performed by a surgeon (7.7%) compared with a cardiologist (3.9%) or neuroradiologist (1.6%).¹⁰ A substudy analysis of CREST, however, failed to identify any difference in composite outcomes (stroke, MI, or death) when an experienced vascular surgeon performed CAS compared with the other identified operators.¹¹ Last, in patients with symptomatic carotid disease, CREST demonstrated a greater incidence of periprocedural stroke in those who received CAS compared with CEA (P = .04).³

Owing to the stringent training requirements in CREST, the applicability of the findings has been questioned with respect to extrapolating these results across the broad spectrum of practitioners who perform CAS. We therefore reviewed outcomes following CAS and CEA performed at a single institution by a single surgeon and compared these results with those outlined in CREST. We demonstrated that CEA and CAS had equivalent rates of stroke, MI, death, and combined MAEs in the periprocedural period when performed by an experienced vascular surgeon. Similarly, subgroup analysis of patients who received CEA or CAS for symptomatic disease demonstrated no differences in periprocedural outcomes (Table 4).

The cause of the aforementioned discrepancies is most likely multifactorial, but more rigorous selection criteria for stenting as well as a greater understanding of the technical nuances of endovascular techniques since the initiation of CREST are important potential contributors. Multiple studies investigating CAS have demonstrated an age-related effect on outcomes, with relatively younger patients faring slightly better after CAS.^3,12-15 This has been attributed to vessel tortuosity and more abundant calcification in those of advanced age; therefore, we attempt to avoid endovascular intervention in this population when possible.^16,17

It has also been posited that carotid disease in elderly persons is especially problematic owing to the length of the lesion. While the deployment of multiple stents could circumvent this problem, this technique has been associated with an increased risk of thromboembolic events. Therefore, we deploy only 1 carotid stent and do not perform subsequent angioplasty for residual stenosis.

If contraindications for endarterectomy preclude open carotid repair, however, alterations to the standard endovascular technique can be used to bypass the vessel tortuosity and atherosclerotic debris common in this population. In selected patients, we use a supraclavicular incision, perform a common carotid artery cut-down with local anesthesia, and access the artery after a purse-string suture has been fashioned.

In those with no anatomic or health-related considerations, we minimize the occurrence of periprocedural stroke by using embolic protection devices when technically possible and flow reversal systems for tight stenoses.

This study represents a single-center retrospective study of CAS vs CEA. Analysis was restricted to the perioperative period; thus, commentary regarding longer-term follow-up was omitted. While the 2 cohorts were similar regarding their baseline level of comorbid conditions and the incidence of symptomatic vs asymptomatic disease, they were not recruited in an identical fashion. Patients who were deemed appropriate for either traditional open repair or endovascular therapy were randomized to treatment. If there was a contraindication noted to CEA, the patient received CAS after enrollment in one of the ongoing endovascular trials. Thus, in the CAS cohort, there were slight differences in the procedure regarding embolic protection and stent use. In addition, while CREST excluded patients with complex anatomy and other pertinent medical issues, those patients were included for analysis in our single-surgeon experience. Last, routine cardiac enzyme evaluation and intracranial imaging were not used unless symptoms were present in the postoperative period.

Carotid endarterectomy and CAS should be viewed as complementary approaches to patients with clinically significant carotid artery stenoses. While appropriate patient selection for each intervention is paramount, CEA and CAS can be performed safely with similarly low risks of perioperative stroke, MI, and death by experienced vascular surgeons. When treating older patients with tortuous carotid arteries and heavily calcified aortic arches, the risk of postoperative stroke with stenting should be strongly considered.

Accepted for Publication: April 24, 2014.

Corresponding Author: Mahmoud B. Malas, MD, MHS, Division of Vascular Surgery and Endovascular Therapy, Johns Hopkins Bayview Medical Center, 4940 Eastern Ave, Baltimore, MD 21401 (bmalas1@jhmi.edu).

Published Online: November 12, 2014. doi:10.1001/jamasurg.2014.1762.

Author Contributions: Dr Malas had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Grimm, Arhuidese, Malas.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Grimm, Arhuidese, Beaulieu, Qazi, Malas.

Critical revision of the manuscript for important intellectual content: Grimm, Arhuidese, Beaulieu, Perler, Freischlag, Malas.

Statistical analysis: Grimm, Arhuidese, Malas.

Obtained funding: Qazi.

Administrative, technical, or material support: Arhuidese, Qazi, Malas.

Study supervision: Perler, Freischlag, Malas.

Conflict of Interest Disclosures: None reported.

Previous Presentation: This study was presented as a poster at the 2014 Annual Meeting of the Southern Association for Vascular Society; January 17, 2014; Palm Beach, Florida.

Disclaimer: Dr Freischlag is the editor of JAMA Surgery and had no role in the review or editorial decision process of this article.