Background and purpose Reported rates of in-stent restenosis after carotid artery stenting (CAS) vary, and restenosis risk factors are poorly understood. We evaluated restenosis rates and risk factors, and compared patients with ‘hostile-neck’ carotids (a history of ipsilateral neck surgery or irradiation) and atherosclerotic lesions.
Methods Demographic, clinical, and radiological characteristics of patients undergoing cervical CAS between 1995 and 2010 with at least 1 month of follow-up were reviewed. Patients with substantial (≥50%) radiographic restenosis were compared with those without significant restenosis to identify restenosis risk factors.
Results The analysis included 121 patients with 133 stented vessels; 91 (68.4%) lesions were symptomatic. Indications for stent placement included hostile-neck lesions, substantial surgical comorbidities, inclusion in a randomized carotid stenting trial, acute carotid occlusion, tandem stenosis, large pseudoaneurysm, high carotid bifurcation, and contralateral laryngeal nerve palsy. Procedures were technically successful in all but one lesion (99.2%). Perioperative stroke occurred in four cases (3.0%). Mean follow-up was 38 months (range 1–204 months), during which 23 vessels (17.3%) developed restenosis. Hostile-neck carotids (n=57) comprised 42.9% of all vessels treated and were responsible for 15 of 23 restenosis cases, resulting in a significantly higher restenosis rate than that of primary atherosclerotic lesions (26.3% vs 10.5%, p=0.017). By univariate analysis, the presence of calcified plaque was significantly associated with the incidence of in-stent restenosis (p=0.02).
Conclusions Restenosis rates after carotid angioplasty and stenting are low. Patients with a history of ipsilateral neck surgery or irradiation are at higher risk for substantial radiographic and symptomatic restenosis.
Statistics from Altmetric.com
Extracranial carotid artery stenosis is a major cause of stroke,1 and is treated with either carotid endarterectomy (CEA) or percutaneous transluminal angioplasty (PTA) and carotid artery stenting (CAS).2 ,3 Results from the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST), comparing the safety and efficacy of CEA versus CAS, found these two procedures to be equivalent in safety and efficacy.4 However, concerns remain regarding the long-term efficacy of CAS compared with CEA. Recent long-term follow-up from major trials has documented in-stent restenosis (ISR) rates ranging from 5.0% to 10.8% for CAS.5 ,6
Risk factors leading to significant ISR remain poorly understood, although a history of prior ipsilateral neck surgery or irradiation (‘hostile-neck’ lesions) is thought to confer a higher risk of ISR than in patients with purely atherosclerotic disease.7 Therefore, rates of ISR may be falsely represented or elevated by the inclusion of differing pathology. We report the ISR rates and risk factors of a large series of patients undergoing CAS procedures, comparing those with hostile-neck and atherosclerotic lesions.
A retrospective review of all patients undergoing CAS between 1995 and 2010 with at least 1 month of both clinical and radiographic follow-up was performed in order to include all periprocedural (<30-day) events and complications. Demographic, clinical, and radiological characteristics at the time of treatment and at longest follow-up were reviewed. Patients with substantial (≥50%) ISR were compared with those with <50% or no ISR to identify ISR risk factors. Composite ipsilateral stroke or death was considered a primary outcome variable. Hostile-neck carotids were defined as vessels that had undergone prior endarterectomy or those in the setting of prior neck irradiation. Given the time period over which the patient cohort was treated, additional analyses of treatment differences were carried out in patients treated prior to 2005 and in those treated thereafter.
The degree of stenosis at presentation was calculated using the North American Symptomatic Carotid Endarterectomy Trial (NASCET) method.8 Symptomatic patients (those suffering stroke, transient ischemic attack (TIA), or amaurosis fugax in the artery referable to the treated lesion) with stenosis >50% and asymptomatic patients with >60% stenosis were treated (one asymptomatic patient with a large pseudoaneurysm and stenosis of 50% was also treated). Indications for CAS included a history of ipsilateral CEA or neck irradiation, substantial surgical comorbidities, inclusion in a randomized trial, acute occlusion, tandem stenosis, large pseudoaneurysm, high carotid bifurcation, and contralateral laryngeal nerve palsy.
Patients were either pretreated with oral aspirin (325 mg) and clopidogrel (75 mg) daily for 7 days prior to the procedure or they were loaded with aspirin (650 mg) and clopidogrel (600 mg) on the day of or the day prior to the procedure. Patients presenting with acute stroke received intravenous abciximab (0.125 mg/kg) immediately after stent placement. After CAS, patients were maintained on 325 mg aspirin and 75 mg clopidogrel daily for at least 6 months, followed by daily aspirin indefinitely.
Follow-up consisted of neurological examination and at least one subsequent radiographic study with ultrasound, CT angiography, or magnetic resonance angiography. Any new neurological symptom in the follow-up period also prompted radiographic evaluation. Patients with an ISR of ≥50% on non-invasive imaging underwent catheter angiography. Symptomatic patients and those with asymptomatic angiographic ISR ≥50% were retreated with either PTA, additional stent placement, or both.
Frequencies with percentages and SDs with means were used to describe the cohort. Independent-sample t tests were used to compare mean differences for continuous variables and the χ2 or Fisher exact tests were used to test for an association between non-continuous variables. Kaplan–Meier survival curves with time to restenosis of ≥50% between the hostile-neck versus the atherosclerotic groups were compared using the log-rank test. Two-tailed p values were considered significant if p<0.05. SPSS Statistics for Windows, V.21 (IBM, Armonk, New York) was used for statistical analyses.
In total, 121 patients with 133 stented vessels met the inclusion criteria. Average stenosis was 83.2%±13.9%. Clinical and radiographic follow-up averaged 38 months (range 1–204 months). The characteristics of all lesions treated with angioplasty and stenting, as well as a comparison between hostile-neck and atherosclerotic lesions, are shown in table 1. Lesions in the atherosclerotic group were more likely to be associated with high-risk radiographic features (eg, dissection, calcification, ulceration) (p=0.02).
Table 2 shows the indications for treatment and lesion characteristics. Of 133 lesions, 91 (68.4%) were symptomatic. Symptomatic lesions were not significantly associated with higher baseline stenosis or development of ISR (p=0.46 and p=0.26, respectively). Contralateral lesions were present in 50.4% of patients, of whom 12 underwent contralateral treatment in separate procedures. Technical success was achieved in all but one vessel (99.2%), in which the lesions could not be angiographically traversed. Ten (7.5%) non-stroke procedure-related complications occurred among the 133 vessels treated (2 myocardial infarctions, 1 TIA with neurological recovery, 3 asymptomatic carotid or vertebral vessel dissections, and 4 femoral access site complications).
Periprocedural and long-term stroke and mortality rates are shown in table 3. Four (3.0%) strokes occurred among the 133 cases in the periprocedural period (<30 days), two of which were major. No deaths occurred during the perioperative period. Eight (6.0%) strokes occurred in the follow-up interval; three (2.3%) were major ipsilateral strokes. Two patients died of neurological causes during follow-up, one from contralateral internal carotid artery occlusion and subsequent hemorrhagic stroke and one from a posterior inferior cerebellar artery infarct due to unrelated vertebral stenosis. The remaining 15 deaths (11.3%) were due to other causes.
Hostile-neck carotids comprised 57 of the 133 (42.9%) vessels treated. Radiographic follow-up showed substantial (≥50%) ISR in 23 (17.3%) vessels at an average of 27.7 months after the index procedure; 15 of these 23 vessels (65.2%) belonged to the hostile-neck group. A comparison of hostile-neck and atherosclerotic carotid groups is shown in table 4 and figure 1. Hostile-neck carotids were more likely to demonstrate substantial restenosis than atherosclerotic lesions (26.3% vs 10.5%, p=0.02, Fisher exact test). Of five lesions (3.8%) with ISR that were symptomatic in the index vessel (including TIA or stroke), four were in the hostile-neck group. Overall symptomatic ISR rates were 4/57 (7.0%) in the hostile-neck group and 1/76 (1.3%) in the atherosclerotic group (p=0.16). Restenosis rates for the two lesion cohorts are shown in figure 1, in which the atherosclerotic cohort demonstrates a longer time without substantial ISR than the hostile-neck cohort (p=0.013, log-rank test). Mean time to restenosis did not differ significantly between atherosclerotic carotids and hostile-neck carotids (34.6 vs 23.9 months, p=0.42). In the univariate analysis (table 5), only the presence of a calcified plaque was significantly associated with the incidence of ISR (p=0.02).
Further analysis was carried out for lesions with at least 6 months of follow-up, of which there was a total of 119 out of 133 lesions. After the exclusion of cases with shorter follow-up, the rate of ISR remained similar (22/119 (18.5%)). The rate of ISR in hostile-neck carotids remained significantly higher than the rate of ISR for all others (14/53 (26.4%) vs 8/66 (12.1%), p=0.04).
Retreatment was carried out on 14 (10.5%) of the 133 lesions, all without periprocedural complications. Hostile-neck carotids were significantly more likely to undergo retreatment than atherosclerotic carotids (11/57 (19.3%) vs 3/76 (3.9%), p=0.004, Fisher exact test). Retreatment was performed with both PTA and stenting in nine vessels, PTA and thromboendarterectomy in one vessel, and PTA alone in four vessels. Two patients eventually developed symptomatic occlusions of the ipsilateral carotid artery 3 and 4 years, respectively, after stent placement and were treated with extracranial-intracranial bypasses. Four patients required a second retreatment procedure for further symptomatic restenosis.
Treatment types and practices were compared in patients treated from 1995 to 2004 (n=59) and those treated from 2005 and later (n=74). A significantly higher proportion of cases used a distal protection device in the group treated from 2005 and later than in the group treated in 1995–2004 (70/74 (94.6%) vs 17/59 (28.8%), p<0.001). There was no statistically significant difference in rates of pre-stent angioplasty or post-stent angioplasty (p=0.44 and p=0.58, respectively). Rates of ISR did not differ significantly between the two time periods (14/59 (23.7%) in the group treated in 1995–2004 vs 9/74 (12.2%) in those treated from 2005 and later, p=0.08).
We have identified ISR risk factors in a large population with stented carotid arteries who have long-term follow-up. Importantly, we found the presence of previous neck surgery or irradiation (hostile-neck lesions) to be significantly associated with substantial ISR and retreatment of ISR. Additionally, cumulative time without ISR over long-term follow-up was significantly shorter for patients with hostile-neck carotids. These results support the growing body of evidence that carotid stenosis arising from neck irradiation or previous surgery has a more severe natural history than primary atherosclerotic carotid disease. Restenosis rates after CAS for these differing cohorts should not be generalized together because they probably perpetuate falsely elevated values for this treatment modality as reported in the medical literature. Patient counseling for CAS should therefore distinguish between these pathologies, informed by the knowledge that restenosis after CAS for primary atherosclerotic disease is generally low. Additionally, we found the periprocedural safety of CAS similar for the three pathologies of radiation-induced stenosis, post-CEA stenosis, and primary atherosclerotic disease.
ISR has been commonly cited as a limitation of endovascular treatment of carotid stenosis. Restenosis rates between 1% and 75% have been reported.9–14 Many studies of restenosis included patients undergoing angioplasty alone and do not appear to reflect current ISR rates. A recent study found peripheral vascular disease to be an independent predictor of ISR.15 A long-term follow-up study by Baldi et al16 evaluated patients treated with CAS up to 7 years after treatment and found ISR in 7.4%. A recent secondary analysis of CREST revealed comparable rates of restenosis between CAS and CEA at 2 years (6.0% vs 6.3%, p=0.58). This analysis identified female sex, diabetes mellitus, and dyslipidemia as risk factors for restenosis, but did not analyze hostile-neck carotids as a separate variable.17 The long-term follow-up data from the Endarterectomy Versus Angioplasty in Patients with Symptomatic Severe Carotid Stenosis (EVA-3S) clinical trial demonstrated a higher rate of restenosis after CAS (12.5%) than after CEA (5%).6 However, ISR was determined only by carotid duplex ultrasonography, which may overestimate the degree of stenosis in stented vessels.18 ,19 The rate of substantial ISR in the current study (determined angiographically) probably reflects a more accurate ISR rate of 10.5% in purely atherosclerotic lesions. Finally, the long-term follow-up from the International Carotid Stenting Study (ICSS) documented a cumulative 5-year risk of severe ISR (>70%) of 10.8% in the stenting group, but no significant difference from that of the endarterectomy group (8.6%).5
Several previous CAS reports that studied ISR included patients with hostile-neck lesions. We found a significantly higher rate of substantial ISR in these patients, and the inclusion of such patients in previous reports may artificially increase the overall ISR rates reported in the literature. Several other studies have also reported higher rates of ISR in patients with hostile-neck lesions.20–23 This has been disputed by a report by Eskandari et al, in which the rate of ISR in patients with hostile-neck lesions was not significantly higher than in patients with atherosclerotic lesions (4.5% vs 2.0%, respectively).21 ,24
In our study, asymptomatic lesions with ≥50% ISR were candidates for retreatment. This threshold is lower than that of other series (70–80%).25 ,26 No patients suffered procedural or periprocedural events resulting from retreatment in our study. This is consistent with previous reports of a lower periprocedural complication rate in retreatment of previously treated carotid stenosis.27 However, the optimal treatment of asymptomatic restenosis remains debatable. The etiology of mild ISR after CAS probably results from smooth muscle proliferation and could confer a lower risk of stroke compared with that for untreated lesions.28 However, recurrent stenosis after neck radiation therapy has been shown to double the relative risk of future TIAs and strokes.29 While the optimal treatment for restenosis after CEA remains unknown, a recent report found a risk of cranial nerve injury of 5.5% in patients treated with repeat CEA.30 The timing and etiology of ISR are also variable. Early restenosis may be related to technical issues and residual stenosis at the time of the procedure. Commonly, intimal hyperplasia occurs within 18 months and is usually asymptomatic whereas late (>18 months) ISR is usually from atherosclerosis and may be more likely to cause symptoms.31
Patients presenting with symptomatic disease did not demonstrate greater average stenosis at baseline than the asymptomatic cohort, nor was there a significant difference in ISR between the two cohorts, although the small numbers make statistical comparisons less meaningful. Choi et al,32 who evaluated patients presenting with near-occlusion of the carotid artery, had similar findings although they reported a higher rate of periprocedural neurological complications (10%) than were found in the current study. Age was not found to be a significant factor for the development of ISR, although octogenarians tended to have higher ISR rates.
Patients with 1-month follow-up were included in order to capture all periprocedural events. The periprocedural stroke and death rate in the present study compares favorably with that of the NASCET trial and is similar to that of the Asymptomatic Carotid Atherosclerosis Study (ACAS).8 ,33 Demographics and other risk factors for hostile-neck and atherosclerotic carotids are similar, with high-risk lesion characteristics being the only variable significantly more common in hostile-neck carotids. Other studies evaluating the impact of premorbid risk factors on complications have found significant risk from diabetes.34 However, our results and an earlier study by Setacci et al21 found no correlation with vascular risk factors.
Our study has several limitations, including that of generalizability inherent to a single-center retrospective study. There are limitations to the use of ultrasonography as an initial screening tool in the follow-up setting, but angiographic confirmation was used in all symptomatic patients. All measurements of ISR at the time of angiographic follow-up were made by the primary endovascular neurosurgeon and were not independently adjudicated. The inclusion of patients with 1-month follow-up may dilute true rates of ISR, but the analysis of patients with at least 6 months of follow-up revealed no significant differences. Angiographic technology and techniques have evolved over the duration of this study, but subanalysis by time period showed differences only in the use of a distal protection device in the later era, whereas techniques such as pre- and post-stent angioplasty and rates of ISR did not differ.
Rates of ISR after carotid angioplasty and stenting are low, but differ between patients with atherosclerotic and hostile-neck etiologies. In this large series of CAS at a high-volume institution, nearly half the treated lesions were hostile-neck carotids, a cohort that is at higher risk for substantial radiographic and symptomatic restenosis. Counseling for patients undergoing CAS should reflect the differences between these distinct pathologies.
We acknowledge the contributions of Kristina Chapple, including all independent statistical analyses and modeling.
Contributors FCA is responsible for the conception and design of this work. KM, ASA, and MRL are responsible for primary data collection. Data interpretation and analysis was performed by KM and MRL. The article was drafted by KM and MRL. All authors critically revised the article and approved the final version of the manuscript. The study was supervised by FCA and CGMcD. All authors agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.
Competing interests None declared.
Ethics approval Ethics approval was obtained from St Joseph's Hospital and Medical Center Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.