Article Text
PDF
Abstract
Objective To determine whether drug-eluting (DE) stents offer any advantage over bare-metal (BM) stents in terms of restenosis rates and to identify indications for vertebral artery stenting (VAS) based on the current literature.
Methodology A retrospective analysis was performed for patients who underwent extracranial VAS for symptomatic stenosis. Patients were divided into two groups based on stent use (DE and BM).
Results A total of 28 stents were implanted in 24 patients. Of these, 13 were DE and 15 BM. Technical success was achieved in all patients. There were procedure-related complications of retroperitoneal hematoma in some patients (3.6%). The median clinical and imaging follow-up times were 26 and 21 months, respectively, for the BM arm and 13 and 12 months, respectively, for the DE group. The mean stenosis percentage was 80.7±10.9 for the BM group and 87.0±9.5 for the DE group. The mean stent diameter was 4.7±0.5 for the BM group and 3.3±0.4 for the DE group (p<0.0001). There was no statistical difference in the restenosis rates among the two groups, 4/15 (26.7%) for BM and 4/13 (30.8%) for DE stents.
Conclusions Extracranial VAS is a relatively safe procedure. No differences could be elicited in the restenosis rates between BM and DE stents. While VAS has been reported for symptomatic and asymptomatic disease, a benefit of VAS has only been demonstrated for patients with symptoms clearly attributable to the diseased artery. The benefit of stenting in asymptomatic disease or stenting to improve perfusion to the anterior circulation is speculative.
- Brain
- MRI
- drug
- thrombectomy
- catheter
Statistics from Altmetric.com
Introduction
Endovascular treatment for extracranial vertebral artery stenosis is not novel.1 2 Stenting of vertebral artery origin disease has been performed for symptomatic and asymptomatic disease.3–5 With the advent of drug-eluting stents (DES) for treatment of coronary vascular disease it was inevitable that this technology would also be applied to cerebral circulation. There have been reports of use of DES for treating vertebral origin disease with some evidence of reduced restenosis rates,3 6–9 but no large series have demonstrated a clear advantage over bare-metal stents. Further, the need for long-term dual antiplatelet treatment associated with the use of DES in vertebral arteries has not been investigated. The indications of vertebral artery stenting (VAS) also continue to evolve. Our goal was primarily to determine whether DES offered a clear advantage over bare-metal stents in the treatment of extracranial vertebral disease and secondarily to identify indications for VAS that are supported by evidence.
Materials and methods
A retrospective analysis of patients who had undergone stent placement for symptomatic atherosclerotic disease involving the extracranial vertebral arteries was performed under an institutional review board approved protocol. A total of 24 consecutive patients over a 3-year period were included. Demographic, procedural and follow-up data were collected for all patients.
Inclusion criteria
All patients had neurological symptoms referable to the diseased vertebral artery with or without MRI correlation. All patients, with one exception, had undergone an unsuccessful trial of single or dual antiplatelet treatment. The one patient who was treated without a trial of antiplatelet treatment had severe bilateral vertebral artery disease and unstable symptoms.
Technique
Procedures were performed with conscious sedation and after at least 3–5 days of dual antiplatelet preparation with aspirin and clopidogrel. Weight-based intraprocedure heparin was administered to maintain an activated clotting time around or above 250 s. Of the 28 stents, 25 were deployed through a transfemoral approach while radial access was utilized in 2 procedures and brachial access in 1 procedure. Pre-stent angioplasty was performed in 3 of 28 procedures and post-stent dilatation was performed in only 1 procedure. All stent procedures were performed without the use of distal protection. Post-procedure recovery included an overnight stay in the intensive or intermediate care unit and then either discharge to home or further observation in the neurology ward.
Results
A total of 28 stents were implanted in 24 patients, with 4 patients receiving bilateral stents. Of the 28 stents, 13 (46.4%) were drug eluting and 15 (53.6%) bare metal. The median imaging follow-up for the entire cohort was 15 months (IQR 7.3–36.7) and the median clinical follow-up was 25.5 months (IQR 6.5–37.5).
The contralateral vertebral artery origin demonstrated atherosclerotic disease in 20 of the 24 patients. The four patients with no disease on the contralateral side had the ipsilateral (diseased) vertebral artery terminating in an isolated posterior inferior cerebellar artery with corresponding neurological symptoms.
Procedural success as determined by the ability to accurately deploy the stent was achieved in all procedures without any technical complications. There were no intraprocedural or post-procedure neurological complications observed either clinically or on imaging. A retroperitoneal hemorrhage requiring transfusion was observed in one patient, without any permanent sequelae. All patients remained on single or dual antiplatelet treatment after the procedure. Tables 1 and 2 summarize the stent and patient characteristics for the bare-metal and drug-eluting groups, respectively. There was one case of stent kinking with subsequent complete occlusion.
Bare-metal stents (n=15)
Drug-eluting stents (n=13)
For all patients, restenosis was seen in 8 out of 28 stents (28.6%). There was no statistically significant difference between the bare-metal and drug-eluting stent groups in terms of restenosis. Two patients (7%) had complete occlusion (one in each group) of the stent on follow-up. One of these occluded at 3 months and the other at 55 months. None of the restenosis or reocclusion events were symptomatic. There were no preprocedure or demographic factors that could be significantly correlated to restenosis. Table 3 gives a comparison of the two groups. The only statistically significant difference was in the stent diameter between the BM and the DE groups, with the mean diameter of the DE stents being smaller than the BM stents (p<0.0001), however the stent diameter did not have an impact on the restenosis rate. This was probably because the maximum available diameter of the DES was 4 mm and any vessel with a larger normal diameter was treated with a bare-metal stent. All patients demonstrated neurological improvement without recurrent symptoms in the same vascular territory.
Comparison of bare-metal and drug-eluting stents
Discussion
Extracranial vertebral artery atherosclerotic stenosis typically involves the ostium of the vessel and can be associated with vertebrobasilar insufficiency. Vertebrobasilar insufficiency in general is an underdiagnosed clinical condition, partly because of its association with non-specific symptoms, such as dizziness, and syncope that also may be caused by lesions in other arterial territories.10 Up to 20% of all ischemic strokes occur in the posterior circulation.11 Demographically, Caucasian men have been shown to be more likely to have atherosclerotic lesions at the origin of the vertebral arteries from the subclavian arteries. Patients with atherosclerosis at this site often have associated carotid, coronary and peripheral vascular disease.12 13
Up to 20% of patients with extracranial proximal vertebral artery disease have lesions in the V1 segment of the vertebral artery (VA).11 Detailed analysis of this data demonstrated that the most likely cause of posterior circulation ischemic events was artery-to-artery embolism from VA origin in 24% of the population studied, and this number increased to almost 50% when possible artery-to-artery embolism cases were added. Hemodynamic transient ischemic attack (TIA) due to bilateral vertebral disease was a less common, but significant, mechanism in posterior circulation ischemia, accounting for 16% of the study population.11 14
Management of posterior circulation ischemic disease is an evolving science. The currently accepted primary approach to vertebral artery origin stenosis is a trial of medical treatment with antiplatelet agents. If there is a failure of medical treatment, surgical solutions are sought.10 Open surgical bypass or endarterectomy for vertebral artery origin disease has complication rates approaching 20% in the literature.15 16 However, some newer series with improved results and lower restenosis rates have been recently published.17 18 Vertebral artery angioplasty and stenting has emerged as a less invasive option, with good technical results and low complication rates.
The currently utilized indications for endovascular treatment of vertebral artery stenosis are:
Symptomatic (ie, transient ischemic attack or non-disabling ischemic stroke in the vertebrobasilar (VB) system) significant bilateral VA stenoses (causing 60% diameter reduction).
Symptomatic unilateral significant stenosis of a dominant VA.
Significant unilateral lesions of a non-dominant VA when symptoms indicated ischemia in the ipsilateral posterior inferior cerebellar artery.
Significant stenosis in an asymptomatic patient who needed collateral support (eg, concurrent carotid artery occlusion).
It has been suggested that the ostium of the vertebral artery has a high concentration of elastin in its composition, thereby increasing the recoil rates.19 This has resulted in very high restenosis rates with angioplasty alone in historical series. The use of balloon-mounted stents has resulted in an improvement in long-term patency over angioplasty.
The local anti-inflammatory effects of DES that reduce the incidence of restenosis in coronary arteries when compared to bare-metal stents may also be applicable to vertebral artery stenosis.20 21 Restenosis rates have ranged from 3.4% to 48% with bare-metal stents in various studies.3 4 22–25 Various factors have been described as contributing to restenosis in bare-metal stents: diabetes mellitus, smoking, small diameter of stent, small vertebral artery, tortuous V1 segment and long stenosis (>10 mm). In contrast, use of cobalt chromium stents, overdilatation of vessels and short stenosis were associated with lower restenosis rates.23–26
Restenosis rates with DES have been lower, ranging from 5.3% to 13% in most studies. In two studies, however, there has been a higher rate. A 63% restenosis rate was identified in the study by Lugmayr et al, however this was a small study of seven patients.27 The study by Werner et al showed a fairly high incidence of stent failure with a 21.6% restenosis rate related to stent kinking/fracture; we also had one case where stent kinking led to complete occlusion. Overall the stent kinking/fracture rate was 39%.9 In the study by Park et al,8 significant (>50%) restenosis rate was 21%, with an additional 25% showing mild restenosis (<50%).
Though small series have shown improved restenosis rates with DES,7 it is as yet unclear whether this provides any clinical benefit. Short-term follow-up in most series of bare mental stenting (BMS) have shown a majority of the restenoses to be asymptomatic. Another gap in the current knowledge is the long-term patency rate. Most series have reported short-term angiographic and longer-term clinical follow-up. The mean angiographic follow-up periods in the literature reviewed ranged from 6 months to 39 months. One of the issues with monitoring of the vertebral artery ostium is the limitation of available non-invasive imaging techniques. This region is beyond the resolution of currently available ultrasound equipment. CT angiography (CTA) of the vertebral artery origin is limited by the tortuosity of this vessel and also beam-hardening artifacts related to the thickness of the shoulder region on CTA. With magnetic resonance angiography, respiratory motion and stent artifact can limit evaluation. Hence, unlike with carotid stenosis, follow-up of vertebral stenosis after stent placement remains difficult. As most restenoses are diagnosed around the 6–12 month period, a repeat angiogram in this time frame appears to be the consensus approach of most practitioners. Hence the incidence of long-term complications within the stented region such as restenosis and fracture presently remain unclear. Restenosis and reocclusion as long as 42 months from initial DES placement has been reported, raising the issue of long-term restenosis/reocclusion and duration of antiplatelet treatment.8 However, many restenoses remain clinically asymptomatic, possibly related to collateral development or resolution of the underlying embolic source at the stenotic site; hence, a consensus on the appropriate monitoring interval is not available. We have had experience of two patients with complete occlusion of the stent on follow-up angiographic imaging. In both cases, robust collateral circulation via the thyrocervical trunk had resulted in maintaining patency of the distal vertebral segment and hence the intracranial flow. As observed in other neurological events related to cerebral vascular occlusion or stenosis, it is possible that it is the severity of the vertebral origin stenosis that results in neurological deficits as well as the suddenness with which it occurs. Stenting in the acute state by restoring perfusion may allow symptomatic improvement related to reversible ischemia. A restenosis or reocclusion graduated over time allows the collateral circulation to take over and prevent a recurrence of symptoms.
Stent occlusion rates with BMS have ranged between 0.8% to 14%.3 23 24 26 With bare-metal stents, delayed occlusion is usually a consequence of restenosis related to neointimal hyperplasia and vascular smooth muscle proliferation that progresses to occlusion. With DES, occlusion mechanisms are likely related to inhibition of re-endothelialization due to the local antiproliferative effect of the active in-stent drug resulting in progressive thrombosis.28 One patient with occlusion of DES at the vertebral ostium was reported in the series by Park et al.8 However, it remains unclear whether the risk of occlusion of DES over the intermediate term is the same in the cardiac population.29 In the cardiac population, DES are managed with long-term antiplatelet treatment to reduce the risk of restenosis.28 Most series describing vertebral artery stents have continued dual antiplatelet treatment in patients with DES for 3–6 months, however the duration of continued treatment is controversial.7 30–32 An interesting study by Rathore et al33 showed stenosis characteristics to be different between DES and BMS in the coronary arteries with in-stent restenosis (ISR) more likely to be diffuse with BMS and focal with DES. It is unclear whether this applies to vertebral DES.
Vertebral ostial stents are also subject to shear stresses related to the tortuosity of the vessel origin as well as flow in the subclavian artery that subjects the stent to deforming forces. Stent fractures on follow-up have been reported in two series with DES6 9 and one series with BMS.26 The incidence of fractures is different depending on the stent architecture, with open cell designs more likely to fracture as compared to closed cell designs. The mobility of the subclavian artery–vertebral ostium is also a contributor to stress on the stent. Use of coronary stents may also increase the incidence of fractures. Fracture incidence also appears to be higher with longer stents and with longer stent protrusion into the subclavian artery. Fractures have also been reported with DES.34 In one series reviewing DES, there was a 39% incidence of stent deformation and one fracture. There was 21.6% restenosis rate related to stent kinking and deformation rather than intimal hyperplasia.9
Most practitioners consider that VAS may be performed safely without a distal embolic protection device. The anatomy of the plaque at the vertebral artery appears to be different from carotid artery plaque, with a more fibrotic plaque as opposed to an atheromatous (fatty) plaque. This could account for the lower incidence of distal emboli and hence intraprocedural embolic complications are quite low in vertebral artery ostial stenting series. A few small series have shown the utility of a distal protection device, however the majority of the studies (without a distal protection device) have shown low procedural complication rates. The use of a distal protection device seems to be a sensible idea to reduce the risk of emboli; however, it also adds another layer of complexity to the procedure. Also, the smaller lumen of the vessel and the often tortuous nature of vertebral artery origins can be challenging to navigate, thereby rendering deployment of a distal protection device quite difficult.
Procedural complications with VAS have been low in a number of case series.3 4 6 7 35 Incidence of embolic complications is fairly low. Access through the brachial artery or radial artery has also been shown to be safe in some series of VAS.
There remain three controversial issues regarding VAS, outlined below.
Stenting of asymptomatic patients
These patients maybe treated because of the perceived need for the VB system to provide hemodynamic or collateral support.4 22 However, there is no real evidence supporting this hypothesis and no long-term data on the impact of asymptomatic vertebral artery stenosis on posterior fossa strokes or for improving flow to the anterior circulation. A thorough neurological assessment is important and, in the face of clearly absent symptoms, intervention is not warranted based on the current data.
Stenting for symptomatic vertebral artery ostium stenosis with a normal contralateral VA
The presumed pathophysiology is artery-to-artery embolism from an in situ thrombus at the stenotic lesion, with one series reporting this to be a factor in almost 25% of the patients.3 4 22 36 However, again while theoretically this mechanism is appealing, the evidence supporting this ‘indication’ is slim. Even though thromboembolism may be a risk factor in posterior fossa strokes, we do not know if antiplatelet treatment alone could be sufficient in preventing these embolic strokes.
Stenting without a trial of medical treatment
There are no clear evidence-based guidelines regarding the role of medical treatment, such as risk factor modification and antiplatelet treatment versus stenting. For patients with unstable symptoms secondary to high-grade stenosis in the dominant vertebral artery, stenting in the acute situation may be warranted. In these patients there is a relatively clear correlation between hypoperfusion as the pathological mechanism and hence reperfusion as the possible solution. For all other symptomatic but otherwise stable patients there is no evidence that stenting is better than medical treatment and, disregarding a trial of non-invasive, medical measures, may not be justified.
Conclusions
In our experience and opinion, and based on the available literature, vertebral artery origin stenting is a relatively safe procedure that should be performed in carefully selected patients with clear correlation between the angiographic lesion, the patients symptoms, the presumed pathophysiological mechanism (hypoperfusion vs embolic) and in stable patients after a trial of medical management. From our data, the use of drug-eluting stents did not result in reduced restenosis, however our sample size may be too small to elicit a difference and further studies with regard to the role of DES and stents in general versus medical treatment are required to expand the indications of vertebral artery origin stenting.
References
Footnotes
Competing interests None declared.
Ethics approval This study was conducted with the approval of the West Virginia University Hospital Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.