Validation and comparison of drug eluting stent to bare metal stent for restenosis rates following vertebral artery ostium stenting: A single-center real-world study

Abstract

Background and purpose

While drug-eluting stents (DES) have been widely applicated in coronary stenosis, uncertainty persists concerning the relative performance and clinical benefit in patients undergoing vertebral artery stenting when compared with a bare metal stent (BMS). We sought to compare in-stent restenosis (ISR) rates of DES and BMS in the treatment of vertebral artery ostium (VAO) stenosis.

Materials and Methods

This study analyzed a single-center prospective cohort. Over 1.5-year period (January 2014-June 2015), 137 consecutive patients underwent VAO stenting involving deployment of 76 DES and 74 BMS. Patient demographics, comorbidities, stenosis severity, stent diameters and lengths, periprocedural complications, imaging and duplex ultrasonography follow-up and recurrent symptoms were assessed.

Results

Technical success was achieved in all patients. Mean VAO stenosis at presentation were 82.4 ± 7.2% in the DES group and 83.3 ± 7.5% in the BMS group and were reduced to 12.5 ± 4.5% and 11.3 ± 4.0%. Mean stent diameter was 3.53 ± 0.40 mm in DES and 5.05 ± 0.40 mm in BMS (p < 0.0001). Mean follow-up was 12.3 months for DES and 11.7 months for BMS. The use of DES was associated with significant lower ISR rates compared with BMS (18.4% vs. 31.1%; OR = 2.628, p = 0.021). Recurrent symptoms rates were similar in DES vs. BMS (2.6% vs 2.7%, p = 0.680). Stent type and stent diameter were independent risk factors for ISR (P = 0.026).

Conclusion

Our results suggest superior efficacy of deploying DES for the treatment of VAO stenosis with lower ISR rates as compared to BMS, but do not support significant differences in periprocedural risk and recurrent symptoms rate.

Introduction

Atherosclerotic vertebrobasilar occlusive disease is a prevalent but under-recognized source of stroke. Due to slow blood flow velocity and curved vessel anatomy, the origin/proximal portion of vertebral artery is susceptible to atherosclerosis, which makes it the most common site of large artery stenosis of the vertebrobasilar system.¹

Unlike carotid artery stenosis, the optimal treatment of vertebral artery ostium (VAO) stenosis remains disputable. Best medical treatment yields unsatisfactory outcomes with the incidence rate of stroke or mortality is up to 5%–11% at 1 year.² Meanwhile, open surgery is not recommended due to technical difficulty and high rates of perioperative complications. Percutaneous angioplasty with stenting seems by contrast to be a promising modality. However, the high incidence of in-stent restenosis (ISR), varies from 16.4 to 70%^3–7 when bare metal stents (BMS) are used, greatly encumbering its application.

In the past decades, anti-proliferative drug-eluting stents (DES) coated with sirolimus or paclitaxel (first-generation DES) have shown superior results to BMS in coronary arteries, especially in aorto-ostial lesions,⁸ a vascular bed that is similar to VAO. To minimize the risk of restenosis post VAO angioplasty, first-generation DES was originally adopted by Ko et al circa 2000.⁹ A rising body of literature supports the technology from this time onwards. Although two meta-analyses have shown that DES has a significantly lower restenosis rate compared to BMS in VAO,^10,11 debate continues due to the retrospective nature of most original studies and the paucity of randomized data in both the short and long term. In fact, as there are no dedicated DES for VAO, larger sizes of DES are unavailable on the market (more than 4 mm in diameter), making direct head-to-head comparisons between DES and BMS virtually impossible. Therefore, this prospective, single-center, real-world study was conducted to evaluate the ISR and symptom recurrence rates of BMS versus DES for symptomatic severe VAO stenosis.

Materials and methods

Study design and clinical definitions

Between January 2014 to June 2015, data were prospectively collected in a consecutive series of patients that underwent stent deployment of the VAO at our institution. Clinical and angiographic data, details of endovascular procedures, periprocedural adverse events and follow-up information were recorded. Technical success was defined as post-interventional residual stenosis <30%. ISR was defined as a diameter loss ≥50% in the stent on vascular imaging (CT angiography or DSA) or as peak systolic velocity (PSV) ≥170 cm/s, end diastolic velocity (EDV) ≥45 cm/s, and PSV ratio ≥2.7, according to published criteria for ISR.¹² The study has been registered in Clinicaltrials.gov. Written informed consent was obtained from all patients and this prospective study was approved by our Institutional Review Board.

Patient selection

All patients were evaluated by the same neurologist at admission. Indications for stenting in our institution were as follows: (1) VAO stenosis ≥70%, defined by the Vertebral Artery Stenting Trial (VAST criteria),¹³ without other stenotic lesion in the intracranial vertebrobasilar territory; (2) Patients suffered from posterior circulation ischemic symptoms during the past 6 months, such as vertigo, gait disturbance or stroke; (3) No new infarction was found on MRI 3 days before procedure; (4) Normal or only mild neurological deficit, defined as modified Rankin scale ≤2. Exclusion criteria were the following: (1) VAO stenosis induced by non-atherosclerotic lesions; (2) Posterior circulation symptoms were induced by disease of perforating artery or small vessels; (3) Patients who were not amenable to endovascular treatment.

Interventional technique

Aspirin (100 mg/d) and clopidogrel (75 mg/d) were prescribed for all patients at least 5 days prior to the procedure. Local anesthesia was adopted in all cases in favor of monitoring neural function during procedure. Access was obtained through a transfemoral approach, after which systemic heparinization was induced to achieve an activated coagulation time of 250 to 300 s. Then, a 6-F guiding catheter was advanced into the subclavian artery proximal to the VAO with a 0.035-inch micro guidewire and the lesion was crossed with a 0.014-inch micro guidewire. Undersized balloon pre-dilatation (2.0 or 2.5 mm) was performed only if there was critical stenosis limiting stent delivery. The stent type was within the individual responsibility of each neurointerventionalist (each performed ≥20 VAO stenting per year for more than 5 years) according to the diameter of target vessel. Stents were chosen depending on the diameter of the distal normal vessel, and the length of the stent must cover 3 mm of the normal lumen on either side of the lesion. It should be noted that when the original diameter of the VAO was > 4.0 mm, only BMSs were available. The BMSs used in this trial included Blue (Cordis Corp, Netherlands) and Express (Boston Scientific, USA). The DESs used in this trial included Xience V (Abbott, USA), Endeavor (Medtronic, Ireland) and Firebird (Microport Medical, China). Distal protection devices were not used in any cases. Post-stenting angiography was performed immediately to measure the residual stenosis. After the procedure, DES patients and BMS patients received double antiplatelet regimen (aspirin 100 mg daily, clopidogrel 75 mg daily) for 12 months and 3 months respectively, followed by aspirin monotherapy indefinitely.

Follow-up

All the patients underwent MRI before and 3 days after procedure if no contraindications to MRI were present. Each scan included diffusion-weighted imaging (DWI) and fluid-attenuated inversion recovery (FLAIR) sequences. Acute periprocedural ischemic brain lesions were defined as hyperintense DWI signals on postprocedural MRI that were not present on preprocedural MRI, namely fresh infarctions.

Carotid and vertebral artery DUS was arranged before discharge and at 361,224 months. To ensure the accuracy of the measurement, all ultrasound examinations were conducted by two physicians who had extensive experience in vascular ultrasound (at least 3 years of clinical experience of more than 3000 vascular cases per year). Repeated CTA or DSA were recommended if ISR was detected combined with recurrent symptoms. Clinical status was assessed at the 1st month and during the ultrasound visits. In addition to the regular follow up, some patients were also followed when they feel out of sort or were convenient to clinical visit at any time.

Statistical analysis

Continuous variables were presented as the mean ± standard deviation, and categorical variables were presented as percentages. Independent sample t-test and Chi-square test were used to compare continuous variables and categorical variables between BMS and DES groups. Logistic regression analysis was used to investigate the risk factors for ISR after stenting, and cumulative ISR rates were computed by the Kaplan-Meier method. All statistical analyses were performed by the SPSS software (Version 23.0, IBM, USA). P < 0.05 was considered statistically significant.

Results

Baseline information and periprocedural managements of BMS and DES groups

From January 2014 to June 2015, 150 VAO stenting sessions including 76 DES and 74 BMS were performed in 137 consecutive patients. Between the BMS and DES groups, no differences in baseline information were identified (Table 1). The detailed distributions of stent types were summarized, including Xience V (n = 68), Endeavor (n = 3) and Firebird (n = 5) in DES group and Blue (n = 44) and Express (n = 30) in BMS group (Table 1). However, balloon pre-dilatation was performed more often among BMS patients (P = 0.035). The mean diameter of BMSs was larger than DESs (P < 0.0001), whereas there was no significantly difference with respect to the stent length (Table 2).

Table 1.

Baseline characteristics of DES and BMS groups.

	DES (n = 76)	BMS (n = 74)	P-value
Age (years)	64.84 ± 10.38	64.30 ± 8.39	0.661
Male, n (%)	62 (81.6)	64 (86.5)	0.846
Hypertension, n (%)	60 (78.9)	52 (70.3)	0.474
Diabetes mellitus, n (%)	26 (34.2)	23 (31.1)	0.461
Coronary heart disease, n (%)	24 (31.6)	17 (23.0)	0.122
Hyperlipidemia, n (%)	12 (15.8)	10 (13.5)	0.853
Smoking, n (%)	42 (55.3)	44 (59.5)	0.691
Associated with anterior circulation symptom, n (%)	15 (19.7)	13 (17.6)	0.724
Associated with ≥70% ICA stenosis, n (%)	54 (71.1)	55 (74.3)	0.526
Associated with ≥70% contralateral VAO stenosis, n (%)	30 (39.5)	26 (35.1)	0.377
Preprocedural stenosis degree (%)	82.4 ± 7.2	83.3 ± 7.5	0.821
DES types, n (% of DES)
Xience V	68 (89.5)	n/a
Endeavor	3 (4.0)	n/a
Firebird	5 (6.6)	n/a
BMS types, n (% of BMS)
Blue	n/a	44 (59.5)
Express	n/a	30 (40.5)

Table 2.

Procedural details and periprocedural complications.

	DES	BMS	P-value
Procedural details
Right side, n (%)	35 (46.1)	33 (44.6)	0.451
Balloon dilatation, n (%)	7 (9.2)	20 (27.0)	0.035
Stent diameter (mm)	3.53 ± 0.40	5.05 ± 0.40	<0.0001
Stent length (mm)	13.29 ± 2.30	15.51 ± 1.43	0.427
Residual stenosis degree (%)	12.5 ± 4.5	11.3 ± 4.0	0.396
Postprocedural complications
Fresh infarctions, n (%)	5 (6.6)	3 (4.1)	0.459
Hematoma in puncture site, n (%)	0	1 (1.4)	0.465
Pulmonary infection, n (%)	1 (1.3)	0	0.318

Periprocedural complications

Technical success was achieved in all the 150 stenting sessions. No stroke or TIA occurred in the vertebrobasilar territory during the first 30 days. Mean VAO stenosis at presentation were 82.4 ± 7.2% in DES and 83.3 ± 7.5% in BMS respectively and were reduced to 12.5 ± 4.5% and 11.3 ± 4.0% after procedure. Postprocedural MRI showed that new infarctions of ipsilateral vertebrobasilar territory occurred in 5 DES patients and 3 BMS patients; local hematoma occurred in 1 BMS patients; and 1 DES patient developed pulmonary infection after the procedure (Table 2).

ISR during the follow up period

All patients underwent clinical and DUS follow-up for at least 6 months. The mean follow-up time was 12.3 months (range: 6–24 months) for DES cases and 11.7 months (range: 6–24 months) for BMS cases. During the follow up, ISR occurred in 14 cases (18.4%) among the DES group, which was significantly lower than the BMS group (23 cases, 31.1%, P = 0.021, Figure 1). Among all the 37 ISR patients, posterior circulation ischemic events recurred in 4 cases during the follow up (2 patients complained of recurrent dizziness, 2 patients complained of recurrent vertigo), and the recurrence rate was insignificantly higher than the non-ISR patients (10.8% vs. 7.1%, p = 0.251, Table 3). The 4 patients were subsequently managed with a prolongation of dual antiplatelet therapy (n = 1), drug-coated balloon angioplasty (n = 1) and ordinary balloon angioplasty (n = 2).

Figure 1.

Cumulative ISR rates of BMS and DES group.

The cumulative ISR rates of the DES and BMS groups calculated by the Kaplan-Meier method. The 2-year ISR rates in the DES and BMS groups were 18.4% and 31.1%, respectively. The cumulative ISR rate of BMS group was significantly higher than that of DES group (P = 0.021).

Table 3.

Symptom recurrence among all patients during the follow up.

Symptom recurred during follow up	ISR cases (n = 37)	Non-ISR cases (n = 113)	P-value
Non-symptom	33	105
TIA	4	8	0.251
Stroke	0	0

Risk factors for ISR after stenting

On univariable analysis, BMS, Stent diameter >4 mm and previous history of coronary heart disease were associated with higher ISR rate. The further multivariable logistic analysis indicated that BMS and Stent diameter >4 mm were independent risk factors for ISR (Table 4).

Table 4.

Logistic analysis for risk factors of ISR after stenting.

Univariate analysis	Non-ISR N = 113		ISR N = 37	P-value
Age, n (%)
<70	75 (77.3)		22 (22.7)	0.284
≥70	38 (71.7)		15 (28.3)
Gender, n (%)
Male	97 (77.0)		29 (23.0)	0.204
Female	16 (66.7)		8 (33.3)
Hypertension, n (%)
N	27 (71.1)		11 (28.9)	0.307
Y	86 (76.8)		26 (23.2)
Diabetes mellitus, n (%)
N	79 (78.2)		22 (21.8)	0.165
Y	34 (69.4)		15 (30.6)
Hyperlipidaemia, n (%)
N	95 (76.0)		30 (24.0)	0.422
Y	18 (62.0)		7 (38.0)
Coronary heart disease, n (%)
N	78 (71.6)		31 (28.4)	0.049
Y	35 (85.4)		6 (14.6)
Smoking, n (%)
N	49 (74.2)		17 (25.8)	0.465
Y	64 (76.2)		20 (23.8)
Procedure side, n (%)
L	66 (80.5)		16 (19.5)	0.078
R	47 (69.1)		21 (30.9)
Stent type, n (%)
DES	62 (81.6)		14 (18.4)	0.036
BMS	51 (68.9)		23 (31.1)
Stent diameter, n (%)				0.036
≤4	62 (81.6)		14 (18.4)
>4	51 (68.9)		23 (31.1)
			95% CI
Multivariate analysis	P	OR	Lower limit	Upper limit
Stent type	0.026	2.628	1.124	6.148
Stent diameter	0.026	2.628	1.124	6.148

Discussion

This study of VAO stenosis patients treated with DES or BMS showed that VAO stenting is a relatively safe procedure. Significant difference was detected in restenosis rates between DES and BMS. However, recurrence of symptoms is rare even after ISR is detected and not different between patients with and without ISR.

In the past decades, vertebral artery angioplasty and stenting has become a universally applicable method for symptomatic severe VAO stenosis.¹⁴ In spite of the relatively low stroke or death rate, the high rate of ISR has raised questions about long-term prognosis after stenting.¹⁵ A systematic review concerning 300 cases of VAOS indicated the ISR rate was 26% after 12 months follow up but most of the restenosis were asymptomatic.¹⁶ By contrast, the study by Langwieser et al showed that the recurrent ischemic events developed in 83% patients with ISR after VAOS.¹⁷ In our study, recurrence of symptoms among ISR patients was insignificantly higher than the non-ISR patients (10.8% vs. 7.1%, P = 0.251), which is consistent with the systematic review. It is possible that severe VAO stenosis results in compromised cerebral hemodynamics as well as sudden symptoms. In this acute state, stenting can easily and safely restore perfusion and relieve symptoms. It also buys time for collateral circulation development to prevent symptom recurrence, even restenosis occurs as a result of intima hyperplasia or stent fracture. In our experience, the Xience V stent should be the first choice for a DES as it is widely used and easy to obtain, but also more flexible and easier to deliver in the setting of VAO with less vascular injury (Figure 2). Ultimately however, dedicated trials that specifically compare different DES are needed.

Figure 2.

(a, b) Placement of a Blue (Cordis Corp, Netherlands) Bare-Metal Stent 5-15mm and (c, d) placement of an Xience V (Abbott, USA) Drug-Eluting Stent 4–12mm at similar vertebral artery ostial stenoses.

ISR after VAO angioplasty/stenting may be caused by many factors. It has been suggested that the anatomic structure and histological features of VAO are directly associated with the high ISR rate. Similar to the coronary artery and renal artery, VAO contains a high concentration of elastin and smooth muscle that facilitates the rate of elastic recoil,^3,18,19 which most likely lead to stent deformation or fracture. Moreover, the atheromatous plaques located at the VAO are usually smooth and stiff, rarely accompanied with ulceration. These characters may have increased the difficulty of stenting and the probability of postprocedural residual stenosis.¹⁹

Vessel diameter may be also associated with ISR after VAOS. It was reported that smaller diameter lead to more restenosis after stent implantation.²⁰ As slender vessel often generates greater elastic recoil and are more likely to develop ISR under the same level of intimal hyperplasia. Ledermann et al found that the ISR rate was 36% among patients with vessel diameter < 4.5 mm after invention; while for patients with vessel diameter >4.5 mm, the ISR rate was only 12%.²¹ In this study, we did not observe the same difference as there was an inseparable bias between the different stent sizes that can be obtained in DES and BMS. And the relationship of vessel diameter or stent size on ISR might be interfered by stent type. Even though DES were used in significantly smaller VAO in our study, their ISR rate was still significantly lower compared to BMS.

In cardiology, it is suggested that BMS is associated with high ISR rate as the intimal hyperplasia induced by iatrogenic injury of vascular wall.^22,23 On the other hand, DES effectively inhibits the migration and proliferation of vascular smooth muscle cells, thus limiting the intimal hyperplasia and reducing the likelihood of ISR.²⁴ Following the success of DES in coronary artery stenosis, more and more investigators have adopted DES for the treatment of VAO stenosis.

Although many studies have shown improved restenosis rates with DES, uncertainty persists concerning the relative clinical benefit in patients undergoing vertebral artery stenting when compared with BMS. In literatures, the ISR rate of DES ranged from 4.3% to 21%, which was significantly lower than that of BMS.^25–28 Ogilvy et al reported a series of patients with VAO stenosis treated by 35 BMS and 15 DES. Long-term follow up found ISR rate of DES group was significantly lower than BMS group (17% vs. 38%).²⁹ Stayman’s follow up data also showed the DES could significantly reduce the ISR rate of VAOS, compared with BMS.³⁰ Chen et al. found the DES could not only reduce the long-term ISR rate of VAO stenosis, but could also effectively prevent the posterior circulation stroke.²⁵ Lu’s retrospective study also reported the similar results.³¹ However, it was also reported that the ISR rate of DES could be as high as 63%.³² Moreover, some other studies failed to identify the significant difference between the ISR rates of DES and BMS.^1,33 In general, more researches have indicated the advantage of DES over BMS in treatment of VAO stenosis, which is consistent with our study.

We believe there is a difference in periods prone to restenosis between DES and BMS. The survival analysis curve (Figure 1) showed that during the first 6 months of follow-up, no significant difference of the accumulative ISR rates between the two groups. However, during the 7th to the 12th month, the ISR rate of BMS group was significantly higher than that of DES. Then the growth rate of restenosis in BMS flattened out and was lower than DES. So, longer observation is needed for further confirmation.

For other factors that have been described as contributing to ISR, we did not observe significant differences. Edgell et al found smoking and aging over 67 years were independent risk factors for ISR after VAOS.¹ In our study, we found smoking or advanced age would not significantly increase the ISR rate, nor the history of hypertension, diabetes mellitus, coronary heart disease, and hyperlipidaemia.

Even though DES used has been widespread with significantly reduced ISR in the post-procedural period, there is growing concern about their long-term safety.³⁴ Concerning long-term complications include late stent thrombosis, cytotoxic effects on distal perfusion areas such as myocardium, and others. It is believed that the antimitotic agents in DES such as sirolimus or paclitaxel and their derivatives induce chronic low-grade inflammation, endothelial dysfunction, and delayed healing.^35,36 However, most of the evidence derives from animal models or human coronary autopsy. Lack of powerful in vivo imaging modalities hampers further understanding of pathophysiological changes. Complications following DES placement are multifactorial, and operator experience may play an important role. Further study of this domain is needed to improve the clinical outcomes of patients with VAO stenosis.

Strengths of our study lie within the prospective design. As a single-center real-world study, we made the protocol according to our clinical practice. To increase the feasibility of follow-up, DSA/CTA was not performed mandatorily during follow up. Instead, Duplex Ultrasound was chosen as the routine technique for the follow-up for its non-invasiveness, as well as convenience and low price in our country. Though some researchers consider that it is operator dependent, it was also reported to be a reliable method for the evaluation of VAO stenosis under uniform and strict criteria.^37,38

The primary limitation of the present study is the huge insuperable bias between the different stent sizes that can be obtained in DES and BMS. The maximum size available for DES was 4 mm making it impossible to conduct a randomized controlled study. Dedicated DES with high radial strength, fatigue resistance and appropriate sizes are needed for VAO stenosis. However, the well performances of off-label coronary DES, as our study shows, diminish the enthusiasm of stent manufacturers. It is difficult to organize additional randomized controlled studies to identify the best treatment strategy for VAO stenosis.

Conclusions

DES for the treatment of symptomatic severe VAO stenosis is both technically feasible and clinically safe. Our single-center real-world study confirms the efficacy of deploying DES for the treatment of VAO stenosis with lower ISR rate as compared to BMS but does not support significant differences in periprocedural risk and recurrent symptoms rate. Longer-term observation and further randomized controlled trial are necessary to compare the clinical benefit of DES and BMS in treating VAO stenosis, but currently we recommend DES as an ideal alternative measure for the future treatment of VAO stenosis.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors thank the support of the National Key Research and Development Project (2016YFC1301703) and the Beijing Scientific and Technologic Project (D161100003816002).

ORCID iDs

Long Li https://orcid.org/0000-0003-2807-5728

Yao Feng https://orcid.org/0000-0001-7923-8158