Optional or optimal? off-label stenting for intracranial atherosclerotic stenosis: A scoping review

Abstract

Background

Intracranial atherosclerotic stenosis is a major cause of ischemic stroke. In addition to the Wingspan stent system, several self-expanding stents have been used off-label to treat intracranial atherosclerotic stenosis lesions. The purpose of this review is to assess the existing data on the off-label use of self-expanding stents in intracranial atherosclerotic stenosis, to highlight methodological limitations in current study designs, and thus providing strategies and precautions for clinical practice.

Methods

The PubMed, EMBASE, and the Cochrane Library databases were systematically searched for relevant articles published up to April 2022. In addition to the meta analysis of Enterprise, Neuroform EZ and closed cell stent respectively, we used a narrative synthesis to summarize and discuss the appropriate strategies and precautions for the use of each stent.

Results

We identified 17 studies (1091 patients with 1124 lesions) reporting 6 types of off-label self-expanding stents. The most common endpoints reported were incidence of short-term complications (range: 0–15.8%, median: 3.8%), long-term complications (range: 0–12.0%, median: 0%). Potential risks include infeasibility of stenting hard lesions or tortuous vessels, stent migration, and in-stent thrombosis. Less is known about the conditions that are appropriate for an optimal stent (e.g., open-cell, close-cell, hybrid cell). There was considerable heterogeneity across studies with regards to study populations and study designs.

Conclusions

The potential risks and benefits should be carefully considered when using off-label stents for intracranial atherosclerotic stenosis, particularly given the current evidence power. As a potential option for the Wingspan stent, based on device's approval only, a tailored approach with lesion-specific devices could be beneficial in certain patients.

Introduction

Intracranial atherosclerotic stenosis (ICAS) is a major cause of ischemic stroke, especially in African, Asian, and Hispanic populations. ¹ Both stenting and aggressive medical therapy (aggressive risk factor modification and antiplatelet therapy) have been introduced for the treatment of ICAS.^2–5 Although aggressive medical therapy has been recommended in the national guidelines of various regions, a certain proportion of patients treated exclusively with drugs remain at a high risk of stroke recurrence, especially in patients with severe stenosis. ⁶ The Wingspan Stent System was the only Food and Drug Administration (FDA) approved stent for ICAS. However, due to the results of previous SAMMPRIS trail, the Wingspan stenting (on-label) was not shown to be superior to aggressive medical treatment for ICAS. ³ One of the explanations is attributed to the device-related complications and technical infeasibility, e.g., perforator branch occlusion, procedure-related subarachnoid hemorrhage, etc.^3,7,8 In this regard, several self-expanding stents (SESs) were evaluated as alternatives to the Wingspan stent for satisfying the device needs of individualized endovascular procedures. On the other hand, unlike the SAMMPRIS study, recently published Wingspan trials (WEAVE/WOVEN, CASSISS) have yielded very good results.^5,9,10 Therefore, it is still worth considering and evaluating whether SESs can still be used as an optimal device option in the treatment of ICAS. However, there is a lack of quantitative analysis of the current data and evaluation of the evidence power. In addition, the stenting procedure, perioperative assessment, and outcomes varied across studies. Therefore, the purpose of this study is to synthesize the existing data, analyze potential weaknesses in study designs, identify gaps in knowledge, and propose the appropriate strategies and precautions for the currently off-label use of SESs in ICAS.

Methods

We conducted this review according to PRISMA Extension for Scoping Reviews (PRISMA-ScR). ¹¹ The protocol for this review has been registered in the Open Science Framework repository (https://osf.io/7kcep/).

Search strategy

Data for this review were identified by systematic searches of PubMed, EMBASE, and the Cochrane Library (from January 2000 to April 2022) by two independent reviewers, as well as references cited by relevant articles. We used the following keywords: “intracranial arteriosclerosis”, “intracranial atherosclerotic disease”, “intracranial atherosclerotic stenosis”, “intracranial stenosis”, “intracranial artery stenosis”, “ICAS”, “ICAD”, “Constriction, Pathologic”, “endovascular therapy”, “endovascular treatment”, “stent,” and “angioplasty”. The reference lists of retrieved articles were examined for additional studies of potential relevance. The detailed search strategy used is described in the Supplemental File 1.

Study selection

Inclusion criteria:

• Patients. Papers reporting patients receiving off-label self-expanding stenting for ICAS (≥50% luminal stenosis), off-label stent is defined as a stent that is not approved by the FDA for ICAS treatment.

• Intervention. Off-label self-expanding stenting.

• Comparison. There was no need for comparison because the purpose of this study was to evaluate the prognosis of stenting and not to compare outcomes with any other therapy.

• Outcome. At least one of the following outcome measures was reported: (a) Short-term complications: Any stroke, transient ischemic attack (TIA) or death within 30 days; (b) long-term complications: Ischemic stroke or TIA in the territory of the qualifying artery beyond 30 days, mortality beyond 30 days; (c) in-stent restenosis (ISR); (d) symptomatic in-stent restenosis (sISR).

• Study type. Randomized controlled trials, observational trials (cohort study, case-control study, case series report with sample size ≥ 10).

• Publication date and language. January 2000, to April 2022, English.

Exclusion criteria.

• Studies containing overlapping patients (chose the most recent study with the largest number of events to avoid information duplication).

• Studies with mixed interventions in which it was not possible to extract data on stenting for ICAS.

• Articles included patients with etiologies other than atherosclerosis, such as moyamoya disease or complete occlusion of the cerebral artery.

• Non-peer-reviewed articles, conference abstracts, reviews, books, and animal studies.

• Articles published in languages other than English.

Data extraction and quality assessment

A standard electric form was used to extract the following data from the included studies by two independent reviewers with at least 10 years’ experience in endovascular therapy: (a) Study characteristics: Author, publication year, study design, duration; (b) patient: number of patients/lesions, gender, age; (c) intervention: Types of stent, average imaging follow-up time, average clinical follow-up time; (d) outcome: Short- and long-term complications, ISR, and sISR. The authors were contacted for missing information when necessary. Two reviewers assessed the quality of each included study in accordance with the methodological index for non-randomized studies (MINORS). ¹² A 13-item scale was introduced for narrative assessment of the weaknesses of the study design of each included study.

Statistical analysis

Available evidence was summarized, using meta-analyses when possible. Otherwise, a narrative synthesis of the data was performed. The incidence was calculated with 95% confidence intervals (CI). The incidence was transformed according to the Shapiro-Wilk test, then a random-effects model or a fixed-effects model was used to summarize it. Publication bias was assessed quantitatively using Egger's test, which is considered significant if the p value is < 0.1. The I² statistic was used to assess study heterogeneity. I² values of 0–25%, 26–50%, 51–75%, and >75% represent low, medium, moderate, and high heterogeneity, respectively. ¹³ All the statistical analyses were conducted by R software (version 4.1.3).

Results

Search results

The initial database search included 4118 studies. After 1394 duplicates were removed, 2724 articles remained for the title and abstract screening. Then, 2475 articles were excluded based on title and abstract. After reviewing the full text among the 249 articles, 17 articles were included (Figure 1).

Figure 1.

Flow diagram nof study selection process.

Quality assessment of included studies

The mean MINORS score for the overall quality of the 17 studies was 9.9 (SD = 1.7) (Table 1). On a 13-point scale for the assessment of study design, the mean score of included studies was 7.6 (SD = 1.8) (Table 2).

Table 1.

Summary of the methodological index for non-randomized studies (MINORS) score of the included studies.

No	Study	1	2	3	4	5	6	7	8	9	10	11	12	Total score
1	Vajda et al.(2012) 15	2	2	2	2	0	1	0	0	/	/	/	/	9
2	Feng et al.(2015) 16	2	2	2	2	0	2	2	0	/	/	/	/	12
3	Lee et al.(2016) 17	2	2	2	2	0	2	1	0	/	/	/	/	11
4	Wang X et al.(2016) 18	2	2	2	2	0	1	1	0	/	/	/	/	10
5	Huang et al.(2019) 19	2	2	2	1	0	0	1	0	/	/	/	/	8
6	Erbahceci Salik et al.(2019) 14	2	2	2	2	0	1	0	0	/	/	/	/	9
7	B Sun et al.(2021) 20	2	2	2	1	0	1	1	0	/	/	/	/	9
8	Cui et al.(2021) 21	2	2	2	2	0	2	1	0	/	/	/	/	11
9	Ari et al.(2022) 22	2	0	2	2	0	2	2	0	/	/	/	/	10
10	Du et al.(2018) 23	2	2	2	2	0	1	1	0	/	/	/	/	10
11	Xu et al.(2019) 24	2	2	2	1	0	0	1	0	/	/	/	/	8
12	H. Li et al.(2022) 25	2	2	2	2	0	1	1	0	/	/	/	/	10
13	Duan et al.(2015) 26	2	2	2	2	0	2	2	0	/	/	/	/	12
14	Cao et al.(2020) 27	2	2	2	2	2	2	2	0	/	/	/	/	14
15	O. Buonomo et al.(2021) 28	2	2	2	1	0	0	0	0	/	/	/	/	7
16	Meyer L et al.(2020) 29	2	2	2	2	0	0	0	0	/	/	/	/	8
17	Tang et al.(2021) 30	2	2	2	2	0	2	1	0	/	/	/	/	11

/: not comparative study; 0: not reported; 1: reported but inadequate; 2: reported and adequate.

Table 2.

Summary of the methodologic and reporting quality of the included studies.

No	Study	Control	Blinded	Sample screening	Symptom before stenting	Time from onset to stenting	HR-MRI before stenting	Perfusion examination before stenting	Mori type	Follow-up		Complications		ISR	sISR	Total
										Clinical	Imaging	Short	Long
1	Vajda et al. (2012) 15	0	0	0	1	1	0	0	0	0	1	1	1	1	1	7
2	Feng et al. (2015) 16	0	0	0	1	1	0	0	0	1	1	1	1	1	1	8
3	Lee et al. (2016) 17	0	0	0	0	0	0	0	0	1	1	1	1	1	1	6
4	Wang X et al. (2016) 18	0	0	0	1	1	0	1	1	1	1	1	1	1	1	10
5	Huang et al. (2019) 19	0	0	0	1	1	1	1	1	1	0	1	0	0	0	7
6	Erbahceci Salik et al. (2019) 14	0	0	1	1	1	0	0	0	1	1	1	1	1	1	9
7	B Sun et al. (2021) 20	0	0	1	1	1	0	1	1	1	0	1	0	0	0	7
8	Cui et al. (2021) 21	0	0	1	1	1	0	1	1	1	1	1	1	1	1	11
9	Ari et al. (2022) 22	0	0	0	1	1	0	0	1	1	1	1	1	1	1	9
10	Du et al. (2018) 23	0	0	0	0	1	0	1	1	1	1	1	1	1	1	9
11	Xu et al. (2019) 24	0	0	0	1	1	0	0	0	1	0	1	0	0	0	4
12	H Li et al. (2022) 25	0	0	0	1	1	0	0	0	1	1	1	1	1	1	8
13	Duan et al. (2015) 26	0	0	0	1	1	0	1	0	1	1	1	1	1	1	9
14	Cao et al. (2020) 27	0	1	0	1	0	0	0	0	1	1	1	1	1	1	8
15	O Buonomo et al. (2021) 28	0	0	0	1	0	0	0	0	0	1	1	1	0	0	4
16	Meyer L et al. (2020) 29	0	0	0	1	1	0	0	0	1	1	1	1	1	0	7
17	Tang et al. (2021) 30	0	0	0	1	1	0	0	0	1	1	1	1	1	0	7

0: No; 1: Yes; ISR: in-stent restenosis; HR-MRI: high-resolution magnetic resonance imaging; sISR: symptomatic in-stent restenosis.

Characteristics of included studies

Seventeen retrospective, single-arm studies were included. All studies included patients undergoing elective stent implantation other than rescue stenting. A total of 6 kinds of SESs were identified, the Enterprise stent was used in 9 studies, the Neuroform EZ stent in 3 studies, and the Solitaire stent in 2 studies. The Neuroform Atlas stent, the Acclino flex stent, and the low-profile visualized intraluminal support (LVIS) stent were used in 1 study each. (Table 3)

Table 3.

Baseline characteristics of the included studies.

Stent type	Study	Region	Design	Duration	Patients/lesions	Male/Female	Age (years) (mean ± SD or median）	Average imaging follow-up time (months)
Enterprise	Vajda et al. (2012) 15	Germany	R	200711-201101	189/209	132/57	64	6.9
	Feng et al. (2015) 16	China	R	200907-201308	44/44	32/12	60.45 ± 9.07	11.5*
	Lee et al. (2016) 17	China	R	201301-201407	24/30	20/4	61.8 ± 10.3	15.4
	Wang X et al. (2016) 18	China	R	201206-201401	60/62	42/18	56.8 ± 8.0	6
	Huang et al. (2019) 19	China	R	201408-201811	68/70	46/22	59.43 ± 9.74	NA
	Erbahceci Salik et al. (2019) 14	Turkey	R	201207-201712	68/68	56/12	62 ± 7	22
	B Sun et al. (2021) 20	China	R	201706-202004	104/105	60/44	58.61 ± 9.32	NA
	Cui et al. (2021) 21	China	R	201505-201908	130/130	86/44	59.2 ± 8.5	8.6
	Ari et al. (2022) 22	Turkey	R	201201-201903	25/25	15/10	61.6 ± 8.19	14.3
Neuroform EZ	Du et al. (2018) 23	China	R	201611-201801	45/46	29/16	65 ± 10.8	7.3
	Xu et al. (2019) 24	China	R	201601-201710	71/72	54/17	58.9 ± 8.2	NA
	H. Li et al. (2022) 25	China	R	201801-202006	70/70	36/34	62.5 ± 1.25	≤6
Solitaire	Duan et al. (2015) 26	China	R	201011-201403	44/44	31/13	62.50 ± 11.72	9.3
	Cao et al. (2020) 27	China	R	201501-201705	32/32	25/7	57.34 ± 8.47	NA
Neuroform Atlas	O. Buonomo et al. (2021) 28	Italy	R	201902-202006	10/10	7/3	68.46 ± 8.44	3-6
Acclino flex	Meyer L et al. (2020) 29	Germany	R	201401-201903	76/76	51/25	69*	110 days*
LVIS	Tang et al. (2021) 30	China	R	201707-202011	31/31	21/10	58 ± 9.7	11.43

*: median; R: retrospective; NA: not available; LVIS: low-profile visualized intraluminal support; SES: self-expanding stent.

Outcomes

In 14 studies, stenosis was noted before the procedure, ranging from 65.4 ± 0.8% to 92.0 ± 6.0%. After stenting, the mean residual stenosis ranged from 11.2 ± 11.8% to 52.2 ± 10.4%.

For all SESs, quantitative analysis cannot be performed due to differences in stent types and study designs. In 9 studies, 712 patients with 743 lesions were treated with the Enterprise stent. The incidence of short-term complications was 6.4% (95%CI, 4.8%–8.6%). The incidence of long-term complications ranged from 0% to 12.0%, which were not bundled together, since some papers look at a time window of 30 days to 1 year, while others study up to 2 or even 3 years. The incidence of ISR and sISR during the follow-up (range from 6 to 22 months) was 10.9% (95%CI, 5.0%–16.8%) and 2.0% (95%CI, 0.6%–3.3%), respectively (Figure 2 Tables 4, 5). Technical failure of the Enterprise stent was reported in one case. ¹⁴

Figure 2.

Outcomes of meta-analysis. (a) Short-term complications of enterprise stent; (b) ISR of enterprise stent; (c) sISR of enterprise stent; (d) short-term complications of closed cell stent; (e) ISR of closed cell stent; (f) sISR of closed cell stent; (g) short-term complications of Neuroform EZ stent; (h) ISR of Neuroform EZ stent; (i) sISR of Neuroform EZ stent. ISR: in-stent restenosis; sISR: symptomatic in-stent restenosis.

Table 4.

Summary of adverse events after SES implantation.

Stent type	Author	Publication year	Number of patients/lesions	Short-term complications(%)	ISR (%)	sISR(%)
Enterprise	Vajda et al. 15	2012	189/209	7.4	24.7	2.3
	Feng et al. 16	2015	44/44	9.1	6.8	4.5
	Lee et al. 17	2016	24/30	12.5	5.0	0
	Wang X et al. 18	2016	60/62	3.3	13.3	11.1
	Huang et al. 19	2019	68/70	4.4	NA	NA
	Erbahceci Salik et al. 14	2019	68/68	1.5	3.3	0
	B Sun et al. 20	2021	104/105	6.7	NA	NA
	Cui et al. 21	2021	130/130	3.8	14.4	3.4
	Ari et al. 22	2022	25/25	8.0	8.0	8.0
Neuroform EZ	Du et al. 23	2018	45/46	2.2	0	0
	Xu et al. 24	2019	71/72	2.8	NA	NA
	H. Li et al. 25	2022	70/70	1.4	9.5	0
Solitaire	Duan et al. 26	2015	44/44	9.1	11.4	4.5
	Cao et al. 27	2020	32/32	3.1	6.3	0
Neuroform Atlas	O. Buonomo et al. 28	2021	10/10	0	NA	NA
Acclino flex	Meyer L et al. 29	2020	76/76	15.8	25	NA
LVIS	Tang et al. 30	2021	31/31	3.2	9.5	NA

NA : not available; SES: self-expanding stent; ISR: in-stent restenosis; sISR: symptomatic in-stent restenosis; LVIS: low-profile visualized intraluminal support.

Table 5.

Long-term complications after SES implantation.

Stent type	Author	Publication year	Number of patients/lesions	Long-term complications(%)	Average clinical follow-up time (months)
Enterprise	Vajda et al. 15	2012	189/209	2.3	NA
	Feng et al. 16	2015	44/44	0	25.6*
	Lee et al. 17	2016	24/30	0	15.8
	Wang X et al. 18	2016	60/62	8.3	6.2
	Huang et al. 19	2019	68/70	NA	1
	Erbahceci Salik et al. 14	2019	68/68	0	NA
	B Sun et al. 20	2021	104/105	NA	1
	Cui et al. 21	2021	130/130	4.8	27.2
	Ari et al. 22	2022	25/25	12.0	18
Neuroform EZ	Du et al. 23	2018	45/46	0	8.8
	Xu et al. 24	2019	71/72	NA	1
	H. Li et al. 25	2022	70/70	0	≤12
Solitaire	Duan et al. 26	2015	44/44	4.5	25.5
	Cao et al. 27	2020	32/32	0	24.1
Neuroform Atlas	O. Buonomo et al. 28	2021	10/10	0	NA
Acclino flex	Meyer L et al. 29	2020	76/76	6.7	110 days*
LVIS	Tang et al. 30	2021	31/31	0	15

NA: not available; *: median.

The closed cell stent (Enterprise, Solitaire, Acclino flex, and LVIS) were used in 13 studies, containing 895 patients with 926 lesions. The incidence of short-term complications was 4.7% (95%CI, 3.3%–6.1%). The incidence of long-term complications (from 0% to 12%), was not combined because the duration of clinical follow-up varied. The incidence of ISR and sISR during the follow-up (range from 6 to 22 months) was 11.5% (95%CI, 6.9%–16.1%) and 2.5% (95%CI, 0.7%–5.4%), respectively (Figure 2, Tables 4, 5).

The Neuroform EZ stents were used for 186 patients with 188 lesions in 3 studies. The incidence of short-term complications was 2.1% (95%CI, 0.3%–4.9%). The incidence of long-term complications during the clinical follow-up (8.8 months, ≤ 12months) was 0%. The incidence of ISR and sISR during the imaging follow-up (7.3 months, ≤ 6 months) was 3.9% (95%CI, 0.0%–13.1%) and 2.0% (95%CI, 0.0%–2.5%), respectively (Figure 2, Tables 4, 5).

Study heterogeneity and publication bias

In the meta-analysis of the Enterprise stent, the heterogeneity of short-term complications, ISR, and sISR was 0%, 82%, and 39%, respectively. We found no significant publication bias for short-term complications (Egger's test, P = 0.22), long-term complications (Egger's test, P = 0.11), ISR (Egger's test, P = 0.70), and sISR (Egger's test, P = 0.11) of the Enterprise stent in this meta-analysis.

In the meta-analysis of the closed cell stent, the heterogeneity of short-term complications, ISR and sISR was 38%, 75%, and 61%, respectively. Potential publication bias was found in short-term complications(Egger's test, P = 0.01), but was not detected in ISR(Egger's test, P = 0.48) and sISR(Egger's test, P = 0.92).

In the meta-analysis of the Neuroform EZ stent, the heterogeneity of short-term complications, ISR and sISR was 0%, 73%, and 0%, respectively. No significant publication bias was found for short-term complications (Egger's test, P = 0.83). Publication bias was not assessed for ISR and sISR due to there were only two included studies.

Discussion

In this review, the outcome data of 17 included studies were analyzed to assess the strength of the evidence for the off-label use of SESs for ICAS. After stenting, the stenosis of arteries was significantly improved (before stenting: 65.4 ± 0.8%–92.0 ± 6.0%, after stenting, 11.2 ± 11.8%–52.2 ± 10.4%). The incidence of adverse events up to 30 days of Enterprise stent implantation (6.4%, 95% CI 4.8–8.6%) was much lower than in the SAMMPRIS and VISSIT trials (14.7% and 24.1%, respectively), higher than in the CASSISS trial (5.1%) and WEAVE/WOVEN cohorts (2.6%).^3–5,9,10

With increasing clinical experience, neurointerventionalists are beginning to recognize that individual selection of the appropriate stent for a particular lesion has an important impact on perioperative complications. However, most of the data from the included studies could not be combined quantitatively at this time. We discussed the use of individual stent narratively.

Enterprise

The Enterprise stent (Codman, USA) is a self-expanding closed-cell stent that was initially designed for stent-assisted coiling of wide-necked cerebral aneurysms. ⁴¹ The stent can be recaptured and deployed after a maximum of 70% of the stent has been released from the catheter. Although we conducted a quantitative synthesis of the results of the enterprise stent, the results should be treated with caution because of the insufficient quality of the included studies. In our meta-analysis of 743 lesions, the technical success rate was 99.9% (742/743). As the author reported, only in one case did stent deployment fail due to inadequate balloon dilatation. The reason could be poor stent navigation or apposition in the segment with a high degree of residual stenosis after balloon dilatation. ¹⁴ In an in vitro test, the radial force per unit of Enterprise stent at 50% compression was less than Wingspan and Solitaire and greater than Neuroform ³¹ (Table 6). Even the reported technical success rate was high, the Enterprise stent may not be suitable for patients with hard lesions due to its low radial force (Table 7). In cases of severe calcification, the Enterprise stent cannot be readily or fully opened, resulting in poor adhesion to the vessel wall or distortion of the stent.^16,33 A rare complication related to a poor stent attachment is stent migration, as reported in previous studies.^35–37 In the studies included in this meta-analysis, two cases of migration were observed, one during the stenting procedure and another during follow-up.^14,15 The closed-cell design of the Enterprise stent is also a double-edged sword: on the one hand, compared to open-cell stents, closed-cell stents carry a lower risk of vessel dissection or rupture if in-stent balloon dilation or re-catheterization is required during the procedure ¹⁵ ; on the other hand, due to the bending stiffness of the closed-cell stents, open-cell stents could conform better to vascular tortuosity than closed-cell stents. ³⁴

Table 6.

The radial force of including SESs in in vitro models.

Types	Radial force at 50% compression 31	Radial force at 15% compression* 31	Perpendicular radial force 32	Circumferential radial force 32
Wingspan	1	1	NA	NA
Enterprise	3	4	2	2
Neuroform	4	2	3	1
Solitaire	2	3	NA	NA
Neuroform Atlas	NA	NA	NA	NA
Acclino flex	NA	NA	NA	NA
LVIS	NA	NA	1	3

“1-4”: ranking; “*”: chronic outward force/outward radial force; “NA”: not available; SES: self-expanding stent; LVIS: low-profile visualized intraluminal support.

Table 7.

Summary of current clinical experience with the use of each SES.

Stent type (characteristic)	Possible benefits	Potential risks
Enterprise(Closed-loop design)	Less risk of ISR 15	Not suitable for calcified lesions16,33
		Less flexibility to fit tortuous vessels (unsegmented closed-cell design) 34
		Risk of stent migration14,15,35–37
Neuroform EZ(Open-cell design)	Less risk of perforator occlusion related to stenting and ISR23,25
	Used for tortuous and small vessels, lesions close to or across a bifurcation 23
Solitaire(Retrievable design)	Rescue intraprocedural in-stent thrombosis 38	Risk of being thrombogenic (13 mm long tail) 27
		Not suitable for hard lesions 39
Neuroform Atlas(Hybrid design with closed cells)	Less intracranial exchange steps of microcatheter 40
Acclino flex(Retrievable design)	Less exchange maneuvers 29
LVIS(Full length visibility)	More accurate delivery and deployment 30

LVIS: low-profile visualized intraluminal support; SES: self-expanding stent; ISR: in-stent restenosis.

Neuroform EZ

The Neuroform EZ stent (Stryker, USA) is a self-expanding open-cell nitinol stent designed for stent-assisted embolization of wide-necked intracranial aneurysms. ⁴² The radiopaque bumpers on the preloaded delivery stent can assist with system visualization during placement. ²³ The Neuroform EZ stent has a larger cell size than the Wingspan, which has been questioned to increase the risk of embolism in distant branches due to intraprocedural plaque rupture by stenting.^31,43 However, this “large-cell risk” has not been demonstrated in real-world practice. ⁴⁴ Nevertheless, the open-cell architecture of the stent could promote stent flexibility and conformability in tortuous anatomy (e.g., lesions near or across a bifurcation) ²³ (Table 7).

In an in vitro study, the Neuroform EZ stent was found to have barely half the chronic radial force of the Wingspan stent in a vessel 85% of its reported diameter but was still stronger than the Solitaire stent and Enterprise stents. ³¹ The main advantage of the Neuroform EZ stent has been perceived as the appropriate radial force it provides, which could reduce the risk of penetrating stroke (due to the “snow-plow” effect) without significantly increasing the risk of ISR (due to intimal hyperplasia).^23,25

Solitaire

The Solitaire stent (EV3, USA) is a laser-cut, self-expanding, closed-loop nitinol stent, including Solitaire AB and Solitaire FR.

One study found that stroke and mortality rates were relatively optimal when the Solitaire AB stent was placed in perforator-bearing arteries (e.g., MCA-M1, BA). ²⁷ One explanation is that, compared to other SESs, the high porosity (greater than Enterprise and Neuroform) of the Solitaire AB stent may protect the orifices of the perforating arteries. The major concern with using a low radial force stent such as Solitaire is whether the lower radial force can prevent elastic vessel recoil and late lumen loss. Of note, two studies with small sample sizes reported that the ISR rate of Solitaire stents was relatively optimal (6.3%, 11.4%), and the sISR rate was 0% and 4.5%, possibly due to the lower radial force of the Solitaire stent reducing chronic stimulation of intimal hyperplasia. However, the low radial force might make the stent unsuitable for old, hard, or calcified lesions. ³⁹ In addition, the 13mm tail of the stent may also be thrombogenic. ²⁷ It is worth noting that the Solitaire stent can be retrieved after complete deployment, which may serve as a rescue procedure in the event of intraprocedural in-stent thrombosis ³⁸ (Table 7).

Neuroform atlas

The Neuroform Atlas stent (Stryker) is a self-expanding nitinol stent by low-profile laser-cut that is designed to provide the coil support and wall apposition needed to perform effective stent-assisted coiling procedures. ⁴⁵ It has a unique hybrid design with closed cells proximally to facilitate microcatheter re-crossing and open cells distally to allow for better anchoring and wall apposition. All sizes of the Neuroform Atlas Stent are deliverable through a 0.165-inch inner diameter microcatheter (e.g., Excelsior SL-10, XT-17 microcatheter), avoiding the need for intracranial microcatheter exchange and associated complications ⁴⁰ (Table 7). In addition, the deployment accuracy of the stent is very high due to less foreshortening. ⁴⁶ To date, only one study has reported the use of the Neuroform Atlas stent in the treatment of ICAS, but the results were promising. In this retrospective study of 10 patients, no short- or long-term complications were observed, and no trend toward ISR was noted at the 3-month follow-up with CTA. ²⁸ Therefore, the advantages and disadvantages of this stent, such as its good conformability and lack of re-sheathability, are worthy of exploration in future studies.

Acclino flex stent

The Acclino Flex Stent (Acandis GmbH, Germany) is a laser-cut self-expanding nitinol stent with a closed-cell design that received Conformité Européenne (CE) certification for the treatment of cerebral aneurysms in April 2014. ⁴⁷ The Acclino flex stent has an optimized asymmetric cell design for improved vessel wall apposition. It can be delivered using a NeuroSpeed balloon catheter, which simplifies stent implantation by avoiding wire changes ²⁹ (Table 7). During stent implantation, three radiopaque markers on each side, as well as a central marker on the transport wire, can help indicate the position and degree of expansion of the stent. The closed cell design of the Acclino stent allows it to be retrieved, re-sheathed, and re-deployed even when the stent has employed up to 90% of its full length. ⁴⁷ All of these features contribute to the deployment accuracy of stenting. The use of the Acclino flex stent for ICAS has been investigated in a multicenter study involving 76 patients. However, the rate of any stroke or death within 30 days following stenting was 14.5% (11/76), which is similar to the SAMMPRIS stent group (14.7%). In addition, ISR was found in 25% of patients (15/60) during a follow-up period of 110 days, resulting in additional PTA in 7 cases. ²⁹ Therefore, the use of this stent for ICAS still needs to be explored under the premise of reducing periprocedural problems and maintaining long-term stent patency.

Low-profile visualized intraluminal support

The LVIS stent system (Microvention, USA) is a self-expanding, retrievable nitinol stent system originally used for the endovascular treatment of cerebral aneurysms. ⁴⁸ Due to its closed-cell design, the LVIS stent can be retrieved up to 80% of its length. As a braided stent, it is equipped with 4 radiopaque markers and 2 radiopaque helical strands to ensure good visibility of the stent. The full-length visibility allows for easier navigation via microcatheters, making delivery and deployment more accurate ³⁰ (Table 7). The LVIS stent may have a higher risk of in-stent stenosis due to its larger metal surface area (12%–23%) and smaller cell size (0.9 mm). ⁴⁹ In a recent study of patients with ICAS treated with LVIS stents, the ISR rate was 9.5% (2/21) after an 11.43-month follow-up period. ³⁰ In a systematic review of nine studies of LVIS stents for intracranial aneurysms, the restenosis rate was 1.4% (95%CI 0.2%–2.5%). ⁵⁰ The lower ISR rate may be due to the lower radial force of the LVIS stent. Notably, despite the high metal coverage, no perforator stroke occurred in this study. The effect of LVIS stent on the penetrating artery still needs to be further evaluated.

Planning and commissioning of future study

There are several methodological limitations and prospects worth pointing out that would provide support for the design and planning of future studies: (a) Meticulous preoperative assessment, especially radiological examinations (e.g., high-resolution magnetic resonance imaging (HR-MRI), perfusion studies, etc.), could be crucial for the interventionalist to understand the etiology of the lesions. Therefore, further work is needed to develop a consensus on routine preoperative evaluation of patients with ICAS. (b) Due to the inclusion criteria and the small ICAS population in which off-label stents are used, hardly any controlled prospective studies comparing different SESs can be performed at present. Nevertheless, relatively long-term follow-up could help to gain clinical experience for individualized therapy. (c) Because of the limited power of current evidence, further studies evaluating the detailed physical characteristics (e.g., different types, different sizes, etc.) of these SESs are needed to assist clinicians in decision-making, as there is no reliable animal model for ICAS. (d) None of the current studies reported the device approval processes, which could be a dangerous proposition.

Limitations of this review

(a) Because of the retrospective nature of included studies, recall bias is an inherent limitation. (b) The inclusion criteria of these studies are not similar, the results similarly cannot be easily compared to one another. (c) The population of most studies included in this review is from Asia, so the population-specific experience may not be universally applicable. (d) We only included studies written in English and therefore acknowledge a publication bias. Other international studies may have been excluded.

Conclusions

The potential risks and benefits should be carefully considered when using off-label stents for ICAS, particularly given the current evidence power. As a potential option for the Wingspan stent, based on device's approval only, a tailored approach with lesion-specific devices could be beneficial in certain patients.