Article Text

Download PDFPDF

Drug eluting stents versus bare metal stents for the treatment of extracranial vertebral artery disease: a meta-analysis
  1. Vivek H Tank1,
  2. Ritam Ghosh2,
  3. Vikas Gupta3,
  4. Nakul Sheth2,
  5. Shariyah Gordon2,
  6. Wenzhuan He2,
  7. Steven F Modica2,
  8. Charles J Prestigiacomo2,
  9. Chirag D Gandhi2
  1. 1Texas Institute for Neurological Disorders (TIND), Sherman, Texas, USA
  2. 2Department of Neurological Surgery, Rutgers New Jersey Medical School, Newark, New Jersey, USA
  3. 3Department of Neurology, University of Missouri, Columbia, Missouri, USA
  1. Correspondence to Dr Vivek H Tank, Texas Institute for Neurological Disorders, 321 North Highland Avenue, Suite 200, Sherman, TX 75092, USA; vtank{at}


Background While a growing number of reports offer evidence for the potential of drug eluting stents (DES) in treating atherosclerotic stenosis of the extracranial vertebral artery, their efficacy when compared with bare metal stents (BMS) is uncertain due to the lack of a large prospective randomized trial.

Methods A search strategy using the terms ‘stents’, ‘drug-eluting stents’, ‘atherosclerosis’, ‘vertebral artery’, and ‘vertebrobasilar insufficiency’ was employed through Medline. Five studies met the criteria for a comparative meta-analysis. The technical/clinical success, periprocedural complications, target vessel revascularization (TVR), rates of restenosis, recurrent symptoms, and overall survival were compared.

Results There was no significant difference in the technical success (OR=1.528, p=0.622), clinical success (OR=1.917, p=0.274), and periprocedural complications (OR=0.741, p=0.614) between the two groups. An OR of 0.388 for no restenosis in the BMS to DES arms (p=0.001) indicated a significantly higher restenosis rate in the BMS group relative to the DES group (33.57% vs 15.49%). When compared with the DES group, the BMS group had a significantly higher rate of recurrent symptoms (2.76% vs 11.26%; OR=3.319, p=0.011) and TVR (4.83% vs 19.21%; OR=4.099, p=0.001).

Conclusions A significantly lower rate of restenosis, recurrent symptoms, and TVR was noted in the DES group compared with the BMS group.

  • Stenosis
  • Stent
  • Neck
  • Atherosclerosis
View Full Text

Statistics from


Posterior circulation strokes carry a high rate of morbidity and mortality due to the vital structures that receive blood supply from the vertebrobasilar system. It is estimated that, of the approximately 700 000 strokes occurring annually, 20% of these affect the posterior circulation.1 Furthermore, between 9% and 20% of these cerebrovascular ischemic events occur as a result of vertebral artery stenosis (VAS). The mortality associated with vertebrobasilar circulation stroke is as high as 30%, and patients experiencing vertebrobasilar transient ischemic attacks (TIAs) have a 25–35% risk of stroke within 5 years.2

The vertebral artery is the second most common area of stenosis after the carotid bifurcation, with the stenosis most frequently occurring near the origin (V0) and proximal area of the vessel (V1). The etiology of the stenosis is similar to stenosis in the anterior circulation—namely, hypertension, smoking, hyperlipidemia, etc. However, more frequently it is the result of artery–artery embolism from the heart, aorta, or a proximal vessel such as the subclavian artery (figure 1).3

Figure 1

(A) V1–V3 are the extracranial segments of the vertebral artery. (B) Schematic of stenosis at the ostium. (C) Tortuous nature of ostium and V1 segment may place mechanical strain on stents.

When VAS occurs, symptoms include dizziness, syncope, diplopia, hemiparesis, leg weakness, and possible cerebellar dysfunction. Due to frequently present collateral circulation from the contralateral vertebral artery, treatment is usually only necessary for bilateral stenosis, severe stenosis of the dominant vertebral artery, or severe ipsilateral stenosis with vertebrobasilar insufficiency (VBI).4 Treatment starts with a regimen of antiplatelet agents such as aspirin or clopidogrel, or anticoagulant agents such as warfarin. However, when symptoms are persistent, either surgical or endovascular options need to be explored. Traditionally, open surgical vascular procedures which include endarterectomy, surgical bypass, and carotid–vertebral transposition have been shown to be technically feasible in this region.1 However, a great deal of experience is needed to perform these procedures and they have also been associated with considerable postoperative complications, such as Horner's syndrome and lymphocele.5 Furthermore, the mortality associated with some surgeries has been reported to be as high as 20%.3

Since the 1980s, endovascular approaches to VAS have been gaining popularity due to the lower rates of complications as well as high technical success.1 The two most common forms of endovascular surgery are angioplasty and stenting. Balloon angioplasty, while maintaining high technical success, can be associated with poor outcomes due to the angioarchitecture of the vertebral artery, primarily in the ostium. Due to a high level of elastin and smooth muscle in the tunica media, there is a considerable risk of elastic recoil, vessel dissection, and restenosis.6 As an alternative, endovascular stenting has been shown to have excellent immediate results and low periprocedural complications. The two main forms of stents are bare metal stents (BMS) and drug eluting stents (DES). BMS, which are older and more commonly used, have been found to have high levels of in-stent restenosis (ISR) (figure 2).6 ,7 ISR rates range from 10–67% and can be attributed to stent fracture as well as neointimal hyperplasia, which is the thickening of the tunica intima of the blood vessel resulting from a surgical procedure or injury. This is due to the aforementioned anatomical features of the vertebral artery, as well as the tortuous nature of the ostium and V1 segments, which places great mechanical strain on the stents.6 These complications with BMS lead to high rates of target vessel revascularization (TVR), which is defined by a repeat intervention of the target vessel driven by recurrent clinical symptoms with a >70% stenosis.8

Figure 2

Treatment with bare metal stents. (A) Left subclavian arteriography shows severe stenosis at the vertebral artery ostium. (B) Post-stenting angiography shows sufficient dilation of the lesion. (C) A dilated coronary balloon-expandable stent is seen on a simple radiograph immediately after treatment. (D) Follow-up angiography at 4 months after stent placement shows in-stent restenosis (arrow). (E) Deformity of the stent is seen on a simple radiograph during follow-up at 4 months (arrow).

To combat these complications, DES coated with a variety of substances such as sirolimus and paclitaxel have been used with increased efficacy in the coronary arteries. The immunosuppressant and immunomodulatory drugs released by the stents reduce the risk of ISR by limiting macrophage accumulation and smooth muscle cell proliferation around the stent (figure 3).6 However, there has not been extensive research comparing BMS and DES in the clinical treatment of VAS. Therefore, this meta-analysis was undertaken to compare BMS and DES in terms of clinical success, technical success, mortality, perioperative complications, recurrent symptoms, and TVR.8

Figure 3

Treatment with short balloon-expandable drug eluting stents. (A) Preprocedural angiograph shows significant vertebral ostial stenosis. (B) Postprocedural angiograph shows sufficient dilation of the lesion. (C) Follow-up at 5 months shows no intimal hyperplasia.

Materials and methods

A Medline search was performed using the terms ‘stents’, ‘drug-eluting stents’, ‘atherosclerosis’, ‘vertebral artery’, and ‘vertebrobasilar insufficiency’. The criteria were that the studies were published within the last 5–7 years, looked specifically at BMS and DES, and were testing for similar outcome measures of interest. Five studies met the criteria for a comparative meta-analysis.8–12 The BMS were either made of cobalt chromium or stainless steel, while the DES were either coated with sirolimus or paclitaxel.

The technical and clinical success, periprocedural complications, TVR, rates of restenosis, and recurrent symptoms and overall survival were compared between the BMS and DES groups. Technical success was defined by resolving stenosis to between 20% and 30%, while clinical success was defined by technical success as well as a resolution of symptoms. Finally, the recurrent symptoms studied included TIAs, stroke, or VBI.

Outcome measures were calculated using the Mantel–Haenszel method with fixed effect models. Heterogeneity among studies was analyzed as well. Sensitivity analysis was then performed with subgroup fixed effect model to solve the heterogeneity among studies. If all attempts were tried and the statistical heterogeneity was found to be a part of the variability of the selected sample trials other than the effect of differences among the groups of patients, the pooled data were then analyzed with random effect models. In case of significant heterogeneity, random effect models were used and heterogeneity between the studies was assessed with the Q statistic.

While calculating OR, as some events had a number of 0, to be able to calculate the OR the name of the parameter was changed. For example, when calculating the OR for total death in BMS versus DES, the total survival rate was calculated rather than the death rate.


The data were first compiled from the eligible studies which are listed in table 1. The mean values of the outcome parameters that were analyzed are shown in table 2, while the meta-analysis results are shown in table 3. The mean pretreatment stenosis was 83.8±4.2% in the DES group (n=156) and 80.12±2.7% in the BMS group (n=148); this difference was not significant. The rates of technical success, clinical success, and periprocedural complications were 98.78%, 95.77%, and 1.94%, respectively, for the DES group versus 100%, 97.96%, and 2.96%, respectively, for the BMS group. However, there was no significant difference in the technical success (OR=1.528, p=0.622), clinical success (OR=1.917, p=0.274), or periprocedural complications (OR=0.741, p=0.614) between the two cohorts. There was no periprocedural mortality, stroke, or TIA. The mean clinical and radiological follow-up times were 19.12±6.9 months and 14.23±1.5 months, respectively, for the DES arm and 26±7.6 months and 20.5±3.3 months, respectively, for the BMS group. When compared with the DES group, the BMS group had a significantly higher rate of recurrent symptoms (2.76% vs 11.26%; OR=3.319, p=0.011) and TVR (4.83% vs 19.21%; OR=4.099, p=0.001). There was no significant difference between overall survival (OR=0.655, p=0.32). An OR of 0.388 of no restenosis in the BMS to DES arms (p=0.001) indicated a significantly higher restenosis rate in the BMS group relative to the DES group (33.57% vs 15.49%). Thus, the DES group was significantly different from the BMS group in terms of restenosis rates, recurrent symptoms, and TVR rate.

Table 1

Overview of studies included

Table 2

Mean value of outcome parameters reported in eligible studies

Table 3

Meta-analysis results


VAS is an occurrence that is still not fully understood. Despite the fact that it accounts for up to 20% of posterior circulation ischemia, the natural history and prognosis of vertebral artery disease has not been as extensively delineated as it has with carotid artery disease.2 Furthermore, it can be hard to diagnose since the most common symptoms—syncope and dizziness—can be associated with a myriad of etiologies.1

VAS has been found to occur in patients who have risk factors for atherosclerosis such as diabetes, hypertension, hyperlipidemia, and smoking.8 To date, the usual sequence of treatment is to first start the patient on antiplatelet agents such as aspirin and clopidogrel, or anticoagulant agents such as warfarin. However, when symptoms are persistent, surgical options such as endarterectomy and carotid artery transposition are then considered. These open surgical techniques require considerable technical skill, and one study reported mortality and morbidity rates up to 8.4% and 13.3% respectively.13

Endovascular surgery is seen as an alluring alternative as the rates of morbidity and mortality have been reported to be as low as 3.3% and 1.5%, respectively.1 While BMS have been used with great technical and clinical success, they have also been associated with high rates of ISR on follow-up, ranging from 10% to 67%.10 In a study by Werner et al,14 the two primary reasons for restenosis were stent fracture, with a rate of 32.1%, and intimal hyperplasia, which occurred at a rate of 20.7%. It has been theorized that the anatomical nature of the vertebral artery ostium, which contains larger relative amounts of smooth muscle and elastin, leads to higher recoil and increases the mechanical stress on stents.7 In the SSYLVIA trial, restenosis after stent deployment occurred in 67% of the vertebral ostium lesions compared with only 25% of the pre-posterior inferior cerebellar artery lesions and 32% of intracranial lesions, suggesting that ostia recoil may require a more robust stent.15 The tortuous nature and continuous mobility of the subclavian–vertebral artery junction may also promote ISR due to stent fracture caused by mechanical stress, as a study by Matula et al16 reported 47.2% tortuosity of the proximal V1 segment, which may preclude safe stent placement. Finally, longer lesion lengths are also associated with higher ISR rates. A study evaluating lesion size and ISR reported ISR rates of 21%, 29%, and 50%, respectively, in lesions of <5 mm, 5–10 mm, and >10 mm.4

In order to combat the high ISR rates associated with BMS, DES coated with antiproliferative agents such as paclitaxel and sirolimus have been engineered to mitigate the macrophage accumulation and smooth muscle cell proliferation which leads to ISR.8 No significant difference was found in aspects such as periprocedural mortality, stroke, and TIA. Furthermore, there was no significant difference in technical and clinical success. However, there was a significant difference between BMS and DES in terms of recurrent symptoms, restenosis, and TVR. The BMS group had a threefold greater rate of recurrent symptoms, twice the rate of restenosis, and a fourfold greater rate of TVR. DES have been used with great efficacy in coronary arteries, which implies that they may also be preferred to BMS in the treatment of extracranial VAS. While DES have been used with great efficacy in coronary artery stenosis, an increased risk of in-stent thrombosis has also been realized as a consequence of inflammation during the healing process.17 Current guidelines recommend dual antiplatelet therapy, most commonly aspirin and clopidogrel, for 6–12 months, with a prolongation of therapy at the discretion of the physician. While recent data show that the prolongation of dual antiplatelet therapy may lead to increased bleeding complications, it may be beneficial in those at a high risk for ischemic events.17–20 Thus, more long-term research is necessary to accurately assess the implications of this possible complication and the duration of antiplatelet course in cases where DES are used within the vertebrobasilar system.

Despite our findings, there are certain limitations to our study. Since this was a retrospective meta-analysis it was difficult to find trials that were homogenous in the variables of interest. Because VAS is a rare procedure, many of the current data are varied but, in cases of significant heterogeneity, random effect models were used and assessed using the Q statistic. ISR has been reported to appear as late as 42 months in certain patients, while the routine follow-up is more commonly 6–12 months.10 This discrepancy may underestimate the reported complication rate with DES and shows that the long-term results with DES are unclear, so additional study is needed. Furthermore, ISR itself is sometimes asymptomatic, since collateral circulation often forms. Even robust occlusions may be clinically silent due to collateral flow from other arteries such as the thyrocervical trunk.10 There is also the limitation of imaging, as the monitoring of the vertebral artery ostium is limited to non-invasive imaging that may be beyond the resolution of ultrasound. In a study by Compter et al, the degree of restenosis after stenting could not be accurately graded except in cases of complete occlusion, as the small size and tortuous nature of the vertebral artery as well as stent artifact on CT angiography prevented accurate results.21 Also, MR angiography can be hampered by factors such as respiratory motion and stent artifact, all of which limits evaluation.6 Finally, human error when measuring aspects such as restenosis can play a significant role, so there remains a need for a randomized prospective study looking at the efficacy of BMS versus DES in both the short and long term.


Compared with the DES group, the BMS group had a threefold greater rate of recurrent symptoms (11.26% vs 2.76%; OR=3.319, p=0.011), twice the rate of restenosis (33.57% vs 15.49%), and a fourfold greater rate of TVR (19.21% vs 4.83%; OR=4.099, p=0.001). Further investigation is needed to determine the significance of these findings.


View Abstract


  • Correction notice This article has been corrected since it published Online First. The sixth author's name has been corrected.

  • Contributors All the authors contributed to the writing of this manuscript.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

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.