Endovascular therapy versus medical treatment for symptomatic intracranial artery stenosis

Abstract

Background

Intracranial atherosclerotic stenosis (ICAS) is an arterial narrowing in the brain that can cause stroke. Endovascular therapy and medical management may be used to prevent recurrent ischaemic stroke caused by ICAS. However, there is no consensus on the best treatment for people with ICAS.

Objectives

To compare the safety and efficacy of endovascular therapy (ET) plus conventional medical treatment (CMT) with CMT alone for the management of symptomatic ICAS.

Search methods

We searched the Cochrane Stroke Group Trials Register (30 August 2019), Cochrane Central Register of Controlled Trials (CENTRAL: to 30 August 2019), MEDLINE Ovid (1946 to 30 August 2019), Embase Ovid (1974 to 30 August 2019), Scopus (1960 to 30 August 2019), Science Citation Index Web of Science (1900 to 30 July 2019), Academic Source Complete EBSCO (ASC: 1982 to 30 July 2019), and China Biological Medicine Database (CBM: 1978 to 30 July 2019). We also searched the following trial registers: ClinicalTrials.gov, WHO International Clinical Trials Registry Platform, and Stroke Trials Registry. We also contacted trialists and researchers where additional information was required.

Selection criteria

Randomised controlled trials (RCTs) comparing ET plus CMT with CMT alone for the treatment of symptomatic ICAS. ET modalities included angioplasty alone, balloon‐mounted stent, and angioplasty followed by placement of a self‐expanding stent. CMT included antiplatelet therapy in addition to control of risk factors such as hypertension, hyperlipidaemia, and diabetes.

Data collection and analysis

Two review authors independently screened trials to select potentially eligible RCTs and extracted data. Any disagreements were resolved by discussing and reaching consensus decisions with the full team. We assessed risk of bias and applied the GRADE approach to assess the quality of the evidence. The primary outcome was death of any cause or non‐fatal stroke of any type within three months of randomisation. Secondary outcomes included any‐cause death or non‐fatal stroke of any type more than three months of randomisation, ipsilateral stroke, type of recurrent event, death, restenosis, dependency, and health‐related quality of life.

Main results

We included three RCTs with 632 participants who had symptomatic ICAS with an age range of 18 to 85 years. The included trials had high risks of performance bias and other potential sources of bias due to the impossibility of blinding of the endovascular intervention and early termination of the trials. Moreover, one trial had a high risk of attrition bias because of the high rate of loss of one‐year follow‐up and the high proportion of participants transferred from endovascular therapy to medical management. The quality of evidence ranged from low to moderate, downgraded for imprecision.

Compared to CMT, ET probably results in a higher rate of 30‐day death or stroke (risk ratio (RR) 3.07, 95% confidence interval (CI) 1.80 to 5.24; 3 RCTs, 632 participants, moderate‐quality evidence), 30‐day ipsilateral stroke (RR 3.54, 95% CI 1.98 to 6.33; 3 RCTs, 632 participants, moderate‐quality evidence), 30‐day ischaemic stroke (RR 2.52, 95% CI 1.37 to 4.62; 3 RCTs, 632 participants, moderate‐quality evidence), and 30‐day haemorrhagic stroke (RR 15.53, 95% CI 2.10 to 115.16; 3 RCTs, 632 participants, low‐quality evidence). ET was also likely associated with a worse outcome in one‐year death or stroke (RR 1.69, 95% CI 1.21 to 2.36; 3 RCTs, 632 participants, moderate‐quality evidence), one‐year ipsilateral stroke (RR 2.28, 95% CI 1.52 to 3.42; 3 RCTs, 632 participants, moderate‐quality evidence), one‐year ischaemic stroke (RR 2.07, 95% CI 1.37 to 3.13; 3 RCTs, 632 participants, moderate‐quality evidence), and one‐year haemorrhagic stroke (RR 10.13, 95% CI 1.31 to 78.51; 2 RCTs, 521 participants, low‐quality evidence). There were no significant differences between ET and CMT in 30‐day transient ischaemic attacks (TIA) (RR 0.52, 95% CI 0.11 to 2.35, P = 0.39; 2 RCTs, 181 participants, moderate‐quality evidence), 30‐day death (RR 5.53, 95% CI 0.98 to 31.17, P = 0.05; 3 RCTs, 632 participants, low‐quality evidence), one‐year TIA (RR 0.82, 95% CI 0.32 to 2.12; 2 RCTs, 181 participants, moderate‐quality evidence), one‐year death (RR 1.20, 95% CI 0.50 to 2.86, P = 0.68; 3 RCTs, 632 participants, moderate‐quality evidence), and one‐year dependency (RR 1.90, 95% CI 0.91 to 3.97, P = 0.09; 3 RCTs, 613 participants, moderate‐quality evidence). No data on restenosis and health‐related quality of life for meta‐analysis were available from the included trials. Two RCTs are ongoing.

Authors' conclusions

This systematic review provides moderate‐quality evidence showing that ET, compared with CMT, in people with recent symptomatic severe intracranial atherosclerotic stenosis probably does not prevent recurrent stroke and appears to carry an increased hazard. The impact of delayed ET intervention (more than three weeks after a qualifying event) is unclear and may warrant further study.

Plain language summary

Endovascular therapy versus medical treatment for symptomatic intracranial artery stenosis

Question

Is endovascular therapy (ET) plus conventional medical treatment (CMT) beneficial for symptomatic intracranial atherosclerotic stenosis compared to CMT alone?

Background

Narrowing of blood vessels in the brain (intracranial atherosclerotic stenosis: ICAS) is a common cause of stroke worldwide. The main treatment choices are catheter treatments (endovascular therapy: ET) and conventional medical treatment (CMT). However, it is unclear which treatment is best. We reviewed trials that compared ET and CMT for ICAS that had recently caused stroke symptoms.

Search date

The search was completed on 30 August 2019.

Study characteristics

We included three randomised controlled trials (studies in which participants are assigned to one of two or more treatment groups using a random method) with a total of 632 participants who had recent symptoms from ICAS. Two trials were carried out across multiple centres and compared ET with metal devices (stents) with CMT alone. One trial took place in a single Chinese centre and compared different types of ET with CMT alone.

Key results

ET was associated with a higher rate of death or further stroke at both early and late review. There were no major differences in the rates of mini‐stroke and death or dependency in the long term.

Quality of the evidence

The quality of evidence was low to moderate due to performance bias and early termination of trials.

Summary of findings

Summary of findings 1. Endovascular therapy compared to conventional medical treatment for symptomatic intracranial artery stenosis.

Endovasular therapy compared to conventional medical treatment for symptomatic intracranial artery stenosis
Patient or population: symptomatic intracranial artery stenosis Setting: hospital Intervention: endovascular therapy Comparison: conventional medical treatment
Outcomes			Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	No. of participants (studies)	Quality of the evidence (GRADE)	Comments
			Risk with CMT	Risk with ET
Death or stroke		Short‐term follow‐up: mean 30 days	Study population		RR 3.07 (1.80 to 5.24)	632 (3 RCTs)	⊕⊕⊕⊝ Moderatea
			51 per 1000	156 per 1000 (92 to 267)
		Long‐term follow‐up: mean 12 months	Study population		RR 1.69 (1.21 to 2.36)	632 (3 RCTs)	⊕⊕⊕⊝ Moderatea
			143 per 1000	242 per 1000 (173 to 338)
Ipsilateral stroke		Short‐term follow‐up: mean 30 days	Study population		RR 3.54 (1.98 to 6.33)	632 (3 RCTs)	⊕⊕⊕⊝ Moderatea
			41 per 1000	147 per 1000 (82 to 262)
		Long‐term follow‐up: mean 12 months	Study population		RR 2.28 (1.52 to 3.42)	632 (3 RCTs)	⊕⊕⊕⊝ Moderatea
			92 per 1000	211 per 1000 (140 to 316)
Type of recurrent event	Transient ischaemic attack	Short‐term follow‐up: mean 30 days	Study population		RR 0.52 (0.11 to 2.35)	181 (2 RCTs)	⊕⊕⊕⊝ Moderatea
			46 per 1000	24 per 1000 (5 to 108)
		Long‐term follow‐up: mean 12 months	Study population		RR 0.82 (0.32 to 2.12)	181 (2 RCTs)	⊕⊕⊕⊝ Moderatea
			92 per 1000	75 per 1000 (29 to 195)
	Ischaemic stroke	Short‐term follow‐up: mean 30 days	Study population		RR 2.52 (1.37 to 4.62)	632 (3 RCTs)	⊕⊕⊕⊝ Moderatea
			41 per 1000	104 per 1000 (57 to 191)
		Long‐term follow‐up: mean 12 months	Study population		RR 2.07 (1.37 to 3.13)	632 (3 RCTs)	⊕⊕⊕⊝ Moderatea
			92 per 1000	191 per 1000 (127 to 289)
	Haemorrhagic stroke	Short‐term follow‐up: mean 30 days	Study population		RR 15.53 (2.10 to 115.16)	632 (3 RCTs)	⊕⊕⊝⊝ Lowa,b
			0 per 1000	0 per 1000 (0 to 0)
		Long‐term follow‐up: mean 12 months	Study population		RR 10.13 (1.31 to 78.51)	521 (2 RCTs)	⊕⊕⊝⊝ Lowa,b
			4 per 1000	39 per 1000 (5 to 301)
Death		Short‐term follow‐up: mean 30 days	Study population		RR 5.53 (0.98 to 31.17)	632 (3 RCTs)	⊕⊕⊝⊝ Lowa,b
			3 per 1000	18 per 1000 (3 to 99)
		Long‐term follow‐up: mean 12 months	Study population		RR 1.20 (0.50 to 2.86)	632 (3 RCTs)	⊕⊕⊕⊝ Moderatea
			29 per 1000	34 per 1000 (14 to 82)
Dependency		Long‐term follow‐up: mean 12 months	Study population		RR 1.90 (0.91 to 3.97)	613 (3 RCTs)	⊕⊕⊕⊝ Moderatea
			33 per 1000	63 per 1000 (30 to 131)
Restenosis (≥ 50%)			‐	‐	‐	‐	‐	No data for analysis
Health‐related quality of life			‐	‐	‐	‐	‐	No data for analysis
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; CMT: conventional medical treatment; ET: endovascular therapy; RCT: randomised controlled trial; RR: risk ratio
GRADE Working Group grades of evidence High quality: We are very confident that the true effect lies close to that of the estimate of the effect. Moderate quality: We are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low quality: Our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low quality: We have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

^aDowngraded one level due to a low incidence of adverse events and small sample size.
^bDowngraded one level due to the wide confidence interval which difference is more than 10.

Background

Description of the condition

Stroke, including ischaemic and haemorrhagic stroke, is the second‐leading cause of death worldwide. It is the third‐leading cause of disability and causes great economic hardships for families and society (Banerjee 2017; Benjamin 2017; WHO 2017). Intracranial atherosclerotic stenosis (ICAS) represents an advanced stage of intracranial atherosclerotic disease (ICAD) with narrowing of vessel lumen by 50% to 90% (Qureshi 2009). ICAS is suggested to be the most common cause of stroke worldwide (Arenillas 2011; Bang 2014; Chatterjee 2015; Gorelick 2008; Holmstedt 2013; Qureshi 2014). The middle cerebral arteries are the most common site for ICAS, followed by the basilar artery and the internal carotid arteries. The vertebral arteries, the posterior cerebral arteries, and anterior cerebral arteries are less commonly affected, while the cerebellar and communicating arteries are rarely involved (Resch 1970; Van der Kolk 2015; Wityk 1996). ICAD often occurs concomitantly with systemic atherosclerosis, involving extracranial, coronary, or peripheral arteries (Manzano 2012).

ICAS is more prevalent in Asia, Africa, South America, and the Middle East (Liu 1996; Sacco 1995; Wang 2014: White 2005), than in Europe or North America (Gorelick 2008; Sacco 1995). It causes 30% to 50% of ischaemic strokes in Asia, compared with 8% to 10% of ischaemic strokes in North America (Feldmann 1990; Leung 1993; Solberg 1972; White 2005; Wong 2006; Wong 2007). The precise reason for this remains unknown; however, genetic susceptibility and environmental factors are thought to play an important role. The high prevalence of ICAS amongst African‐Americans may be partly attributed to the disproportionately high rate of hypertension, diabetes mellitus, and hyperlipidaemia, which are well‐known risk factors for atherosclerotic disease (Carson 2011; Ritz 2014; Waddy 2009). On the other hand, the white population in the USA and Europe has a higher rate of extracranial atherosclerotic disease (Stevens 2008). Based on clinical manifestation, ICAS can be classified into asymptomatic and symptomatic, with symptomatic ICAS correlating with a high clinical recurrence rate of stroke (Chimowitz 2005).

ICAS can be diagnosed using invasive neuroimaging, namely catheter angiography, which is the gold standard for accurately measuring degree of stenosis, differentiation of occlusion from severe stenosis, and collateral flow evaluation. Non‐invasive modalities include transcranial Doppler (TCD), magnetic resonance angiography (MRA), and computed tomographic angiography (CTA). These offer safer, accessible, and less expensive methods of evaluating intracranial circulation (Feldmann 2007; Hirai 2002; Hou 2009; Sloan 2004). More recently, high‐resolution magnetic resonance imaging (MRI), intravascular ultrasonography, and optical coherence tomography (OCT) allow visualisation of the submillimetre structure of the intracranial arterial wall. These modalities help in characterisation of plaque morphology and identification of high‐risk plaque components, such as intraplaque haemorrhage, thin or ruptured fibrous cap, and high lipid core scores (Aoki 1995; Chu 2004; Hatsukami 2000; Patel 2013; Saam 2005; Turan 2015).

Description of the intervention

Contemporary treatment for ICAS consists of medical and endovascular treatments, and rarely surgical therapy. Medical therapy includes antiplatelet therapy, optimal glycaemic and blood pressure control, statin therapy, and lifestyle modifications. Despite advances in medical management, the risk of recurrent stroke remains quite high. In the Warfarin‐Aspirin Symptomatic Intracranial Disease (WASID) trial, the risk of recurrent stroke was 20.4% in the aspirin group and 17% in the warfarin group during the average follow‐up time of 1.8 years (Kasner 2006). In recent years, the preference for aggressive medical management (i.e. dual antiplatelet therapy with aspirin and clopidogrel combined with intensive risk factor management) in people with symptomatic ICAS has increased, supported by the relatively lower complication rates reported in several trials (Chimowitz 2015; Kasner 2006; Turan 2010). Surgical therapy is rarely used for ICAS management. The most commonly used surgical procedure is extracranial‐to‐intracranial bypass surgery (EC‐IC bypass). However, the EC‐IC bypass trial failed to show any benefit of EC‐IC bypass over medical management. The trial demonstrated that EC‐IC bypass surgery was associated with worse outcomes for middle cerebral artery stenosis compared to medical management alone (EC/IC Bypass Study Group 1985).

Endovascular therapy, namely percutaneous transluminal angioplasty and stenting (PTAS), has been proposed for ICAS management since the 1980s (Derdeyn 2007). The middle cerebral arteries, basilar artery, internal carotid arteries, and intracranial vertebral arteries are the major target vessels for PTAS (Derdeyn 2014; Zaidat 2015). PTAS was considered as a minimally invasive approach to prevent recurrent ischaemia in symptomatic ICAS. PTAS was noted to have potential efficacy and an acceptable periprocedural morbidity in initial studies (Aoki 1995; Derdeyn 1998; Feldmann 2007; Zacharatos 2010). The results of the first randomised controlled trial, Stenting versus Aggressive Medical Therapy for Intracranial Arterial Stenosis (SAMMPRIS), showed no advantage of PTAS over medical treatment for ICAS management (Chimowitz 2011). Subsequently, there has been a decrease in endovascular treatment for ICAS (Chimowitz 2015; Derdeyn 1999). Practice guidelines allowed for the investigational use of angioplasty, with or without stenting, for severe ICAS (70% to 99% stenosis) of a major intracranial artery and actively progressing symptoms in spite of aggressive medical management (An 2002). Studies from Asia, however, demonstrated promising outcomes for endovascular therapy (Derdeyn 2014; Hankey 2002). The safety and efficacy of endovascular therapy compared with medical treatment for treating ICAS therefore remains unclear.

How the intervention might work

Endovascular therapy can be broadly categorised into angioplasty alone, balloon‐mounted stenting, and angioplasty followed by placement of a self‐expanding stent. Angioplasty alone is the simplest method for achieving revascularisation by dilating the stenotic vascular lumen with an endovascular balloon. The advantages of angioplasty alone include the low procedural risk and the potential for the lesion to remodel after angioplasty (Chatterjee 2015). The major disadvantage is the vessel recoil and flow‐limiting dissection of the plaque and vessel, which could be avoided by stenting (Connors 2014). With the use of a balloon‐mounted stent, the stent delivery balloon, together with a microguide wire, is navigated through the stenosis. Stent dilation and deployment is achieved by inflation of the balloon (Berkefeld 2009). Balloon‐mounted stents were first designed for coronary arteries, and not for intracranial use, thus it was often difficult to deliver them through the tortuous intracranial vasculature, commonly resulting in the distortion of the regional anatomy and sometimes causing vascular trauma. Delayed in‐stent restenosis is an undesirable and notable issue for balloon‐mounted bare metal stents, which could possibly be reduced by drug‐eluting stents (Gupta 2006; Natarajan 2010). Compared to self‐expanding stents, balloon‐expandable stents have the advantage of rapid exchange single‐step systems that do not require the more complex exchange length guidewires (Jiang 2007; Kurre 2012). Moreover, the lesion does not need to be navigated more than once with balloon‐mounted stents, thus reducing risk of embolic strokes and haemorrhagic (wire perforation) complications (Chimowitz 2011; Derdeyn 2013; Marks 2012). While using self‐expanding stents, the lesion is first angioplasted with a balloon followed by positioning and deployment of self‐expanding stent across the lesion. The angioplasty balloon and stent are less rigid than a balloon‐mounted stent, causing less distortion of anatomy and traumatic injuries with greater technical success (Chatterjee 2015). Each of the above methods has its own technical advantages and disadvantages.

Why it is important to do this review

Stroke is the second‐leading cause of death worldwide after ischaemic heart disease (WHO 2017). ICAS is one of most common causes of stroke (Bang 2014; Holmstedt 2013; Qureshi 2014), particularly high‐grade symptomatic ICAS (generally > 70% stenosis) (Benjamin 2017; Huang 2014). The results of the SAMMPRIS study showed no advantages of endovascular therapy over medical treatment for ICAS management (Chimowitz 2011). The SAMMPRIS trial was criticised on its study design, including the lack of a lead‐in phase, the inexperience of the operators, and poor patient selection (Abou‐Chebl 2012; Gao 2015; Gao 2016; Luo 2018; Tsivgoulis 2016). The final results of a revised randomised controlled trial, comparing endovascular therapy and medical treatment, will be published in the near future (Gao 2015). Recently, several studies have shown favourable outcomes after stenting versus medical treatment. A prospective study from Hong Kong of 65 participants with ICAS treated with Wingspan stent reported a 30‐day periprocedural stroke or death rate of 6.1%, with no stroke up to one year. There is decreased likelihood of recurrent stroke following Wingspan stenting amongst Asians compared to the white population (Yu 2014). This review drew together the latest evidence to determine the safety and efficacy of endovascular therapy versus conventional medical treatment for management of symptomatic ICAS. The conclusions may help clinicians make more informed decisions and prompt more targeted and localised study designs for researchers.

Objectives

To compare the safety and efficacy of endovascular therapy (ET) plus conventional medical treatment (CMT) with CMT alone for the management of symptomatic intracranial artery stenosis (ICAS).

Methods

Criteria for considering studies for this review

Types of studies

We included all randomised controlled trials (RCTs) that compared ET plus CMT with CMT alone in the management of symptomatic ICAS.

Types of participants

We included adults (aged over 18) with symptomatic ICAS (≥ 50%, measured by digital subtraction angiography) related to atherosclerotic factors, where the stenosis was located in at least one major intracranial artery (internal carotid artery, vertebral artery, middle cerebral artery, or basilar artery). Participants with ICAS with a transient ischaemic attack (TIA) or stroke attributable to the territory of the stenotic artery were defined as symptomatic. A TIA was defined as a transient episode of neurological dysfunction (focal weakness or language disturbance, transient monocular blindness, or required assistance in walking) caused by focal brain or retinal ischaemia that lasted for at least 10 minutes but resolved within 24 hours (Easton 2009). We accepted the diagnosis of TIA and stroke made by experienced researchers.

Types of interventions

We compared ET plus CMT with CMT alone. The following ET modalities were acceptable: angioplasty alone, balloon‐mounted stent, and angioplasty followed by placement of a self‐expanding stent. CMT included antiplatelet therapy in addition to control of risk factors such as hypertension, hyperlipidaemia, and diabetes.

Types of outcome measures

Primary outcomes

• Safety outcomes: short‐term death or stroke

We defined 'short‐term' as the periprocedural period, or mean follow‐up time less than or equal to three months after randomisation. Stroke was identified in the vascular territory of the stenosed vessel, either ischaemic or haemorrhagic. We defined death or stroke as a composite of death of any cause or non‐fatal stroke of any type in any territory.

Secondary outcomes

• Death or stroke (long term; more than three months)

• Ipsilateral stroke (same territory as the index stenosis)

• Type of recurrent event (TIA, ischaemic stroke, haemorrhagic stroke)

• Death

• Restenosis (≥ 50%) of the involved vessel documented by conventional cerebral angiography

• Dependency: modified Rankin Scale or equivalent

• Health‐related quality of life

We defined 'long‐term' as mean follow‐up time more than three months after randomisation. With the exception of death or stroke, we evaluated all secondary outcomes at both short and long term.

Search methods for identification of studies

See details of the 'Specialized register' at Cochrane Stroke. We searched for trials in all languages and arranged for the translation of relevant articles where necessary.

Electronic searches

We searched the Cochrane Stroke Group Trials Register (30 August 2019) and the following electronic databases.

• Cochrane Central Register of Controlled Trials (CENTRAL) in the Cochrane Library (to 30 August 2019) (Appendix 1)

• MEDLINE Ovid (from 1946 to 30 August 2019) (Appendix 2)

• Embase Ovid (from 1974 to 30 August 2019) (Appendix 3)

• Science Citation Index Web of Science (from 1900 to 30 July 2019) (Appendix 4)

• Scopus (from 1960 to 30 August 2019) (Appendix 5)

• Academic Source Complete EBSCO (ASC; from 1982 to 30 July 2019) (Appendix 6)

• China Biological Medicine Database (CBM; from 1978 to 30 July 2019) (Appendix 7)

We developed the MEDLINE search strategy with the help of the Cochrane Stroke Information Specialist, and adapted it for searching the other databases. We combined the search strategies deployed with adaptations of the Highly Sensitive Search Strategy designed by Cochrane for identifying RCTs and controlled clinical trials, as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

We searched the following ongoing trials registers.

• US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov/; searched 30 August 2019) (Appendix 8)

• World Health Organization (WHO) International Clinical Trials Registry Platform (who.int/ictrp/en/; searched 30 August 2019) (Appendix 9)

• Stroke Trials Registry (www.strokecenter.org/trials/; searched 31 July 2019) (Appendix 10)

We also contacted trialists and researchers where we required additional information.

Searching other resources

In an effort to identify further published, unpublished, and ongoing trials, we:

• checked the bibliographies of included studies and any relevant systematic reviews identified for further references to relevant trials (searched 31 August 2019);

• contacted experts/trialists/organisations in the field to obtain additional information on relevant trials; and

• conducted a search of the grey literature using the Canadian Coordinating Office for Health Technology Assessment's (CCOHTA's) resource (searched 31 August 2019).

Data collection and analysis

Selection of studies

Two review authors (JL and XW) independently screened the titles and abstracts of the references identified as a result of the search and excluded obviously irrelevant reports. We retrieved the full‐text articles for the remaining references and the same two review authors (JL and XW) independently screened the full‐text articles, identifying studies for inclusion and recording the reasons for exclusion of the ineligible studies. Any disagreements were resolved through consensus decision made by the full review team. We collated multiple reports of the same study so that each study, not each reference, was the unit of interest in the review. We recorded the selection process and completed a PRISMA flow diagram (Liberati 2009).

Data extraction and management

Two review authors (JL and XW) independently extracted data and recorded details from the included studies using a standardised data extraction form. We extracted all relevant data of interest from the included studies, as follows: methods, characteristics of participants, interventions, primary and secondary outcomes, and time points reported. Any disagreements between the two review authors on data extraction were resolved by discussion and consensus decision made by the full review team. We contacted the original study authors for key information absent from the full text. For dichotomous data, we extracted the number of participants experiencing the event and the total number of participants in each arm of the trial. For continuous data, we extracted the mean value and standard deviation (SD) for the changes in each arm of the trial, along with the total number in each group.

Assessment of risk of bias in included studies

Two review authors (JL and XW) independently assessed the risk of bias for each included study using the criteria outlined in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Any disagreements between the two review authors were resolved through discussion with the full review team. We assessed risk of bias according to the following seven domains.

• Random sequence generation

• Allocation concealment

• Blinding of participants and personnel

• Blinding of outcome assessment

• Incomplete outcome data

• Selective outcome reporting

• Other possible bias

We graded the risk of bias for each domain as high, low, or unclear and provided information from the study report together with a justification for our judgement in the 'Risk of bias' tables.

Measures of treatment effect

For dichotomous data (e.g. ischaemic stroke, recurrent ischaemic stroke, death), we calculated effect sizes as risk ratios (RRs) with 95% confidence intervals (CIs) (Boissel 1999). For continuous data (e.g. time from qualifying event to intervention, degree of stenosis of symptomatic qualifying artery), we planned to convert the data into standardised mean differences (SMDs) and present 95% CIs, as it was assumed that study authors would have used different measurement scales. If continuous outcome data were recorded using the same measurement scale, we would convert data into mean differences (MDs) and present with 95% CIs. In the case of missing data, such as SDs, we would have obtained these using the calculations in Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We would have extracted both change scores (i.e. change from baseline) and final values. We planned to incorporate studies with change‐from‐baseline outcomes into a meta‐analysis with studies with final measurement outcomes by using the (unstandardised) MD method in Review Manager 5 (Review Manager 2014).

Unit of analysis issues

We assessed the level of randomisation for all included studies. For studies with non‐standard designs (e.g. cluster‐randomised trials, multiple‐arm studies), we managed the data according to the recommendations in Chapter 9 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).

Dealing with missing data

We conducted a complete‐case analysis as far as possible on an intention‐to‐treat basis for all outcomes based on the methods described in Chapter 16 of the Cochrane Handbook for Systematic Reviews of Interventions for dealing with missing data (Higgins 2011). In the case of missing data on dropouts, withdrawals, and outcome measures, we attempted to contact the original study authors or sponsors to request the required information, where appropriate. When this was not possible, we considered both best‐case and worst‐case scenarios for dichotomous data (e.g. ischaemic stroke, intracranial haemorrhage, death), as sensitivity analyses.

Assessment of heterogeneity

We utilised the I² statistic to measure heterogeneity amongst the trials in each analysis. We considered an I² greater than or equal to 60% as evidence of moderate to substantial levels of heterogeneity (Higgins 2011). When heterogeneity was found, we attempted to explore the potential reasons for it by subgroup characteristics.

Assessment of reporting biases

Funnel plots are useful for measuring reporting bias, but are of limited power when there are fewer than 10 studies involved. We therefore planned that if there were 10 or more included studies in this systematic review, we would use funnel plots to assess the potential existence of small‐study bias and reporting bias. If fewer than 10 studies were included, we would not perform the funnel plots for outcomes; instead we planned to assess reporting bias qualitatively on the basis of the characteristics of the included studies.

Data synthesis

We performed a meta‐analysis on the results when we found at least two studies suitable for inclusion. When a meta‐analysis was not feasible due to an insufficient number of studies, we provided a narrative description of the study results. Where appropriate, we pooled data from studies which were sufficiently similar using Review Manager 5 (Review Manager 2014). We used a random‐effects model to analyse outcomes, or a fixed‐effect model if there was little evidence of heterogeneity.

Subgroup analysis and investigation of heterogeneity

We planned that when there was a sufficient number of studies, we would perform subgroup analyses for the primary outcome by:

• different types of endovascular therapy (i.e. angioplasty alone, balloon‐mounted stent, and self‐expandable stent);

• ethnicity (e.g. Asian versus white versus African versus Hispanic people);

• timing from qualifying event to randomisation (i.e. time interval within three weeks versus time interval more than three weeks);

• lesion location (i.e. internal carotid artery, middle cerebral artery, basilar artery versus intracranial vertebral artery).

We evaluated differences between subgroups using the 'test for subgroup differences' in Review Manager 5 (Review Manager 2014).

Sensitivity analysis

We planned that when heterogeneity of results was substantial, we would conduct a sensitivity analysis based on the primary outcome by excluding studies:

• with inadequate allocation concealment;

• in which the assessment of outcomes was not blinded;

• which were unpublished;

• in which loss to follow‐up was not reported or was more than 10%;

• in which the funder played an important role that could have affected the primary outcome.

GRADE and 'Summary of findings' table

We created a 'Summary of findings' table using the following outcomes: death or stroke (short term: less than three months follow‐up; long term: more than three months follow‐up), ipsilateral stroke (same territory as the index stenosis), type of recurrent event, death, dependency (modified Rankin Scale), restenosis (≥ 50%) of the involved vessel documented by cerebral angiography, health‐related quality of life. We used the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness, and publication bias) to assess the quality of a body of evidence as it relates to the studies that contribute data to the meta‐analyses for the prespecified outcomes (Atkins 2004). We used the methods and recommendations described in Section 8.5 and Chapter 12 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), employing GRADEpro GDT software (GRADEpro GDT). We justified all decisions to downgrade the quality of studies using footnotes, and made comments to aid the reader's understanding of the review where necessary.

Results

Description of studies

See Characteristics of included studies, Characteristics of excluded studies, and Characteristics of ongoing studies.

Results of the search

After removal of 6888 duplicates from the original 26,906 references, we screened a total of 20,018 references. We excluded 20,010 references by title or abstract in level one screening. We reviewed eight studies in full text in level two screening, and included three studies (Chimowitz 2011; Miao 2012; Zaidat 2015). We identified two ongoing studies (Cui 2016; Gao 2015). We excluded three completed studies because the majority of participants had extracranial atherosclerotic stenosis (Compter 2015; Coward 2007; Markus 2017). The PRISMA flow diagram is shown in Figure 1.

1.

Study flow diagram.

Included studies

All participants had intracranial atherosclerosis with more than 70% of stenosis. The age of participants was 18 to 85 years; the average age ranged from 53.4 to 61.8 years across the three trials. The majority of participants were male (62.3%). There were 318 participants enrolled in the ET group and 314 participants enrolled in the CMT group.

Of the three included studies, one was a prospective, randomised, controlled, single‐centre clinical comparing ET plus CMT versus CMT alone for symptomatic severe middle cerebral artery stenosis (≥ 70%) (Miao 2012). Two studies, SAMMPRIS and VISSIT (Vitesse Stent Ischemic Therapy) were prospective, randomised, controlled, multicentre clinical trials comparing ET plus aggressive medical treatment with aggressive medical treatment alone for symptomatic ICAS (≥ 70% stenosis) including internal carotid artery, middle cerebral artery, intracranial vertebral artery, and basilar artery (Chimowitz 2011; Zaidat 2015).

There were some differences in the subgroups of the interventions used in the included studies. First, the types of endovascular therapy were different. Participants enrolled in the ET group of SAMMPRIS were treated with Wingspan stent, which was a self‐expandable intracranial stent (Chimowitz 2011). Participants enrolled in the ET group of VISSIT were treated with the PHAROS Vitesse stent, which was a balloon‐expandable intracranial stent (Zaidat 2015). Participants of the last trial were treated with various types of endovascular treatment, for example primary angioplasty, coronary balloon‐expandable stent (Coroflex or Coroflex Blue), or Wingspan stent (Miao 2012). Second, the time interval of intervention after the qualifying event was different. One trial with 70 participants included ET more than three weeks from the qualifying event (Miao 2012), whilst the other two trials (562 participants) included ET less than three weeks from the qualifying event (Chimowitz 2011; Zaidat 2015). Finally, one trial was conducted in China recruiting 70 Chinese participants (Miao 2012), whereas the other two trials enrolled 71.4% (401/562) white participants (Chimowitz 2011; Zaidat 2015).

The study endpoints were similar in the three studies. Detailed data of any stroke or death, ischaemic stroke caused by culprit lesions, intracranial haemorrhage, disable or fatal stroke or all‐cause death at 30 days and one year could be acquired and analysed.

Excluded studies

We excluded three studies that included both extra‐ and intracranial atherosclerotic stenosis, where separate data for analysis could not be obtained (Compter 2015; Coward 2007; Markus 2017).

Ongoing studies

Two studies are ongoing (Cui 2016; Gao 2015). Only study protocols have been published, and we were unable to obtain more information for analysis.

Risk of bias in included studies

Information related to risk of bias is presented in Characteristics of included studies, Figure 2, and Figure 3.

2.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

3.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

We deemed allocation concealment for the three trials to be adequate with a low risk of bias.

Blinding

Due to the nature of ET, blinding of participants and operators/personnel was not possible. The risk of performance bias was thus high in all three RCTs (Chimowitz 2011; Miao 2012; Zaidat 2015). The risk of bias from blinding of outcome assessment was low because the investigators were not aware of the treatment assignment, and detection and reporting of outcome events was also blinded.

Incomplete outcome data

The risk of attrition bias was low in Chimowitz 2011 and Zaidat 2015, but high in Miao 2012 (see Figure 2 and Figure 3). Miao 2012 reported a 30% (49/70) loss in one‐year follow‐up, and 19.4% of participants in the ET group transferred into CMT group, which resulted in a high risk of attrition bias in spite of using an intention‐to‐treat analysis of the results.

Selective reporting

We did not use funnel plots to assess the potential existence of small‐study bias and reporting bias because fewer than 10 included studies were analysed. We assessed reporting bias qualitatively on the basis of the characteristics of the included studies. We assessed the risk of bias to be low because the protocols for the included trials were available. The prespecified outcomes were reported in the published protocol for two studies (Chimowitz 2011; Zaidat 2015). Although one trial did not publish a study protocol, we obtained the registered information from the Chinese Clinical Trial Registration (ChiCTR‐TRC‐12001989) website. The outcomes of this study were also prespecified in the website (Miao 2012).

Other potential sources of bias

All three studies were terminated early. The final results would not have changed in Chimowitz 2011 and Zaidat 2015 due to the superiority of CMT over ET in the interim analysis. However, in Miao 2012, the final results might have been different if recruitment had continued because the interim analysis showed equipoise between CMT and ET with regard to the safety and efficacy of the treatments. We therefore assessed the risk of other bias as low in Chimowitz 2011 and Zaidat 2015 and high in Miao 2012.

Effects of interventions

See: Table 1

See Table 1 for the main comparison: ET plus CMT versus CMT alone for symptomatic ICAS in adults.

The primary and secondary outcomes available for these three trials were outcome data at 30 days and one year after randomisation, respectively. We therefore defined 30‐day outcomes as short‐term outcomes and one‐year outcomes as long‐term outcomes in order to reduce statistic heterogeneity.

Primary outcomes

Safety outcomes: short‐term death or stroke

Three RCTs with 632 participants were available and analysed for safety outcomes, namely 30‐day death or stroke (Chimowitz 2011; Miao 2012; Zaidat 2015). Meta‐analysis showed ET was related to a significantly higher rate of 30‐day death or stroke compared with CMT (3 RCTs, 632 participants, risk ratio (RR) 3.07, 95% confidence interval (CI) 1.80 to 5.24, P < 0.001; moderate‐quality evidence: Analysis 1.1) with no heterogeneity (I² = 0%).

1.1. Analysis.

Comparison 1: Primary outcomes: safety outcomes, Outcome 1: Short‐term death or stroke

Subgroup analysis

Short‐term death or stroke with different types of endovascular therapy

We included two RCTs with a total of 562 participants: Chimowitz 2011 included 451 participants treated with Wingspan stent, a self‐expandable stent, and Zaidat 2015 included 111 participants treated with PHAROS Vitesse stent, a balloon‐mounted stent. We excluded Miao 2012 because multiple types of ET, including primary angioplasty, balloon‐mounted stent, and self‐expandable stent, were used but could not be distinguished from one another. ET was associated with a high risk of 30‐day death or stroke compared to CMT in both subgroups (self‐expandable stent: RR 2.57, 95% CI 1.39 to 4.76, P = 0.003, 1 RCT, 451 participants, Analysis 1.2; balloon‐mounted stent: RR 5.18, 95% CI 1.61 to 16.68, P = 0.006, 1 RCT, 111 participants, Analysis 1.2). There was no significant intergroup difference (I² = 7.2%).

1.2. Analysis.

Comparison 1: Primary outcomes: safety outcomes, Outcome 2: Short‐term death or stroke with all types of endovascular therapy

Short‐term death or stroke in various ethnicity

Data were not available for subgroup analysis related to multiple ethnicities. Chimowitz 2011 and Zaidat 2015 enrolled participants from multiple ethnicities including white, Asian, black, and Hispanic. They did not report separate outcomes by ethnicity in both ET and CMT groups. We were not able to obtain individual patient data for these studies after requests to the original study authors. Miao 2012 enrolled only Asian participants and showed that there was no significant difference (RR 2.84, 95% CI 0.12 to 67.36, P = 0.52, 1 RCT, 70 participants).

Short‐term death or stroke under different time interval from qualifying event to performance

Data from three RCTs were available. Two trials, SAMMPRIS and VISSIT (562 participants), performed ET within three weeks after the qualifying event, with the mean time interval being seven and nine days in the SAMMPRIS and VISSIT trials, respectively (Chimowitz 2011; Zaidat 2015). One trial with 70 participants performed ET more than three weeks after the qualifying event, with the mean time interval being 261.95 days (Miao 2012). The analysis showed that ET was associated with a worse outcome compared with CMT in the subgroup of the time interval of intervention less than three weeks (RR 3.08, 95% CI 1.80 to 5.29, P < 0.001, 2 RCTs, 562 participants, Analysis 1.3). However, in the subgroup of the time interval of intervention more than three weeks, there was no significant difference between ET and CMT (RR 2.84, 95% CI 0.12 to 67.36, P = 0.52, 1 RCT, 70 participants, Analysis 1.3). There was no significant intergroup difference (I² = 0%).

1.3. Analysis.

Comparison 1: Primary outcomes: safety outcomes, Outcome 3: Short‐term death or stroke under different time from qualifying event to performance

Short‐term death or stroke at different lesion location

Data were not available for subgroup analysis by lesion location. Chimowitz 2011 and Zaidat 2015 enrolled participants with various intracranial lesions including internal carotid artery, middle cerebral artery, basilar artery, and vertebral artery of the intracranial segment. They did not report separated outcomes by single lesion location. Miao 2012 enrolled participants with middle cerebral artery stenosis, which demonstrated no significant difference between ET and CMT (RR 2.84, 95% CI 0.12 to 67.36, P = 0.52, 1 RCT, 70 participants).

Secondary outcomes

Short‐term outcomes

Data from the included trials were available for analysing five 30‐day outcomes including 30‐day ipsilateral stroke, transient ischaemic attack (TIA), ischaemic stroke, haemorrhagic stroke, and death. Chimowitz 2011 did not report the event of TIA. Meta‐analysis showed that ET was associated with a worse outcome compared with CMT in 30‐day ipsilateral stroke (RR 3.54, 95% CI 1.98 to 6.33, P < 0.001, 3 RCTs, 632 participants, moderate‐quality evidence; Analysis 2.1); 30‐day ischaemic stroke (RR 2.52, 95% CI 1.37 to 4.62, P = 0.003, 3 RCTs, 632 participants, moderate‐quality evidence; Analysis 2.2); and 30‐day haemorrhagic stroke (RR 15.53, 95% CI 2.10 to 115.16, P = 0.007, 3 RCTs, 632 participants, low‐quality evidence; Analysis 2.3). There were no significant differences between ET and CMT in 30‐day TIA (RR 0.52, 95% CI 0.11 to 2.35, P = 0.39, 2 RCTs, 181 participants, moderate‐quality evidence; Analysis 2.4) and 30‐day death (RR 5.53, 95% CI 0.98 to 31.17, P = 0.05, 3 RCTs, 632 participants, low‐quality evidence; Analysis 2.5). There was no heterogeneity amongst studies in terms of these 30‐day outcomes.

2.1. Analysis.

Comparison 2: Secondary outcomes, Outcome 1: Short‐term ipsilateral stroke

2.2. Analysis.

Comparison 2: Secondary outcomes, Outcome 2: Short‐term ischaemic stroke

2.3. Analysis.

Comparison 2: Secondary outcomes, Outcome 3: Short‐term haemorrhagic stroke

2.4. Analysis.

Comparison 2: Secondary outcomes, Outcome 4: Short‐term TIA

2.5. Analysis.

Comparison 2: Secondary outcomes, Outcome 5: Short‐term death

Long‐term outcomes

Similarly, data from the included trials were available for analysing seven one‐year outcomes including one‐year death or stroke, ipsilateral stroke, TIA, ischaemic stroke, haemorrhage stroke, death, and dependency. The outcome of one‐year TIA was not available from SAMMPRIS (Chimowitz 2011). Dependency was indicated as greater than 3 on the modified Rankin Scale.

Meta‐analysis showed that ET was associated with a worse outcome compared with CMT in one‐year death or stroke (RR 1.69, 95% CI 1.21 to 2.36, P = 0.002, 3 RCTs, 632 participants, moderate‐quality evidence; Analysis 2.6); one‐year ipsilateral stroke (RR 2.28, 95% CI 1.52 to 3.42, P < 0.001, 3 RCTs, 632 participants, moderate‐quality evidence; Analysis 2.7); one‐year ischaemic stroke (RR 2.07, 95% CI 1.37 to 3.13, P < 0.001; Analysis 2.8); and one‐year haemorrhagic stroke (RR 10.13, 95% CI 1.31 to 78.51, P = 0.03, 2 RCTs, 521 participants, low‐quality evidence; Analysis 2.9). There were no significant differences between ET and CMT in one‐year TIA (RR 0.82, 95% CI 0.32 to 2.12, P = 0.68, 2 RCTs, 181 participants, moderate‐quality evidence; Analysis 2.10); one‐year death (RR 1.20, 95% CI 0.50 to 2.86, P = 0.68, 3 RCTs, 632 participants, moderate‐quality evidence; Analysis 2.11); and one‐year dependency (RR 1.90, 95% CI 0.91 to 3.97, P = 0.09, 3 RCTs, 613 participants, moderate‐quality evidence; Analysis 2.12). For the above analysis, I² was between 0 and 43%, therefore no moderate to substantial heterogeneity was observed.

2.6. Analysis.

Comparison 2: Secondary outcomes, Outcome 6: Long‐term death or stroke

2.7. Analysis.

Comparison 2: Secondary outcomes, Outcome 7: Long‐term ipsilateral stroke

2.8. Analysis.

Comparison 2: Secondary outcomes, Outcome 8: Long‐term ischaemic stroke

2.9. Analysis.

Comparison 2: Secondary outcomes, Outcome 9: Long‐term haemorrhagic stroke

2.10. Analysis.

Comparison 2: Secondary outcomes, Outcome 10: Long‐term TIA

2.11. Analysis.

Comparison 2: Secondary outcomes, Outcome 11: Long‐term death

2.12. Analysis.

Comparison 2: Secondary outcomes, Outcome 12: Long‐term dependency

With regard to restenosis of the involved vessel determined by conventional cerebral angiography, we did not perform meta‐analysis due to lack of relevant data in the CMT group. Only Zaidat 2015 reported the result of restenosis for stent. The results described 26.5% of risk with at least 50% restenosis and 2.9% of risk with at least 70% restenosis at one‐year follow‐up. We did not assess health‐related quality of life because we would not extract any data from the included studies.

Discussion

Summary of main results

The current systematic review included three RCTs with 632 participants. Meta‐analysis showed that CMT was significantly superior to ET in participants with symptomatic and degree of stenosis ≥ 70% intracranial atherosclerotic stenosis in terms of 30‐day and one‐year death or stroke, ipsilateral stroke, ischaemic stroke, and haemorrhagic stroke. However, there were no significant differences in terms of 30‐day and one‐year TIA and death as well as long‐term dependency. Overall, the quality of the evidence for the major outcomes assessed by GRADE ranged from low to moderate (GRADEpro GDT), mainly due to early termination of the trials and small sample sizes. Moreover, the sample size of the current review was small based on the calculation of sample size reported in the previous systematic review (Cruz 2006), which stated that 950 participants per group was necessary under alpha of 0.05, power of 80%, and an expected relative risk reduction of 25% as well as less than 7% perioperative risk of death or stroke.

Overall completeness and applicability of evidence

Overall, the quality of evidence for major outcomes ranged from low to moderate, which partially limits its application to guide clinical decision. Nonetheless, the results from this review represent the highest level of current evidence. Furthermore, patient selection for ET has been revised in the post‐SAMMPRIS trials. Trials with modified inclusion criteria have reported a promising perioperative risk of 2% and 4.3%, respectively (Gao 2016; Miao 2015). The ongoing RCTs have adopted the revised inclusion criteria (Cui 2016; Gao 2015). The safety and efficacy of endovascular treatment for selected participants needs to be clarified by future studies. A major limitation of this review is the limited number trials that could be included for meta‐analysis. Consequently, the tests of heterogeneity have a very low power and do not provide reliable results. Substantial heterogeneity might exist despite the tests being negative.

The time interval from the qualifying event to intervention may be a factor affecting the outcomes of endovascular treatment. According to the subgroup analysis of this review, endovascular treatment performed more than three weeks after the qualifying events appears to be safer than when performed less than three weeks. Several studies supported this result. Miao 2015 and Gao 2016 showed a 4.3% and 2% rate of short‐term death or stroke, respectively, when endovascular treatment was performed more than three weeks after the qualifying events. The cause of the increased risk for ET in the early period of stroke may relate to instability of plaque and intracranial microcirculation. This in turn may be causative of the snowplowing effect of plaque and reperfusion haemorrhage leading to the increased risk of death or stroke (Luo 2018). Future studies are needed to confirm this speculation.

Quality of the evidence

The quality of evidence of the included RCTs was generally low or moderate. The included trials had a high risk of performance bias because blinding of the endovascular intervention was not possible. Moreover, all trials were terminated early after interim analysis. Miao 2012 has a high risk of attrition bias because the rate of loss of one‐year follow‐up was as high as 30%, and 19.4% of participants transferred from endovascular therapy to medical management.

Potential biases in the review process

We tried to minimise bias at every stage of the review's development; however, we could not avoid all bias during our review. Two review authors independently assessed studies for eligibility and extracted the data. Any discrepancies were resolved through discussion with a third review author, or if necessary, by consulting the Cochrane statistician. We did not perform a funnel‐plot analysis as we only identified three eligible RCTs.

Agreements and disagreements with other studies or reviews

Based on the previous Cochrane Review published in 2006 (Cruz 2006), which did not support ET over CMT alone with low‐level data, we confirmed this result with high‐level data from randomised trials. The results from the present study showed that CMT alone is superior to ET, which was also similar to another meta‐analysis by Zhang 2017 that extracted data from two randomised trials and two retrospective cohort studies. Pooled analysis showed that ET and CMT alone had comparable risks of event‐free survival and recurrent TIA, whilst ET had a higher risk of short‐term stroke.

However, there are some interesting results from several recent observational studies. The following studies suggested that ET would benefit selected patients who had symptomatic intracranial atherosclerosis with cerebral hypoperfusion and those who suffered at least two failures of best medical management (Abualhasan 2019; Padalia 2018). The WEAVE (Wingspan Stent System Post Market Surveillance) trial was a post‐market surveillance trial permitted by the US Food and Drug Administration in order to reassess the perioperative safety of the Wingspan stent in on‐label patients who have symptomatic intracranial atherosclerosis with 70% to 90% degree of stenosis, baseline modified Rankin Scale less than 3 and at least two failures of best medical management. The result was favourable for ET, with a 2.6% of rate perioperative complications (Alexander 2019). Moreover, the stent was performed more than eight days after the last event.

Authors' conclusions

Implications for practice.

This systematic review provides moderate‐quality evidence indicating that conventional medical treatment probably carries a lower risk of early recurrent stroke and death than endovascular therapy (ET) when used to treat recently symptomatic severe intracranial atherosclerotic stenosis. Longer‐term outcomes also tend to favour CMT. Further information on treatment decisions is limited by the relatively small sample size.

Implications for research.

There are continuing debates that ET may be of benefit for a specific subgroup of patients, such as those patients who received ET three weeks after the qualifying event. Additional high‐level data from prospective high‐quality randomised controlled trials meeting accepted standards are therefore warranted.

History

Protocol first published: Issue 2, 2019
Review first published: Issue 8, 2020

Appendices

Appendix 1. CENTRAL search strategy

1. MeSH descriptor: [Vascular Surgical Procedures] this term only
2. MeSH descriptor: [Endovascular Procedures] this term only
3. MeSH descriptor: [Angioplasty] this term only
4. MeSH descriptor: [Angioplasty, Balloon] this term only
5. MeSH descriptor: [Angioplasty, Balloon, Laser‐Assisted] this term only
6. MeSH descriptor: [Cerebral Revascularization] this term only
7. MeSH descriptor: [Blood Vessel Prosthesis] this term only
8. MeSH descriptor: [Blood Vessel Prosthesis Implantation] this term only
9. MeSH descriptor: [Stents] this term only
10. MeSH descriptor: [Dilatation] this term only
11. MeSH descriptor: [Catheterization] this term only
12. (angioplasty or stent* or pta or revasculari?ation or recanali?ation or catheter* or dilatation):ti,ab,kw
13. #1 or #2 or #3 or #4 or #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12
14. MeSH descriptor: [Intracranial Arterial Diseases] explode all trees
15. MeSH descriptor: [Intracranial Arteriosclerosis] this term only
16. MeSH descriptor: [Vertebrobasilar Insufficiency] this term only
17. MeSH descriptor: [Cerebral Arteries] explode all trees
18. MeSH descriptor: [Basilar Artery] this term only
19. MeSH descriptor: [Vertebral Artery Dissection] this term only
20. #17 or #18 or #19
21. MeSH descriptor: [Arterial Occlusive Diseases] explode all trees
22. MeSH descriptor: [Arteriosclerosis] this term only
23. MeSH descriptor: [Constriction, Pathologic] this term only
24. #21 or #22 or #23
25. #20 and #24
26. ((intracranial or intra‐cranial or cerebral arter* or basilar arter* or vertebral arter* or vertebrobasilar or vertebro basilar) near/5 (stenos?s or isch?emia or insufficien* or arteriosclerosis or atherosclerosis or occlus* or obstruct* or plaque*)):ti,ab,kw
27. #14 or #15 or #16 or #25 or #26
28. #13 AND #27
29. MeSH descriptor: [Carotid Artery Diseases] this term only and with qualifier(s): [surgery ‐ SU, therapy ‐ TH]
30. MeSH descriptor: [Carotid Stenosis] this term only and with qualifier(s): [surgery ‐ SU, therapy ‐ TH]
31. MeSH descriptor: [Vertebrobasilar Insufficiency] this term only and with qualifier(s): [surgery ‐ SU, therapy ‐ TH]
32. MeSH descriptor: [Brain Ischemia] this term only and with qualifier(s): [surgery ‐ SU, therapy ‐ TH]
33. MeSH descriptor: [Cerebral Arteries] this term only and with qualifier(s): [surgery ‐ SU]
34. MeSH descriptor: [Basilar Artery] this term only and with qualifier(s): [surgery ‐ SU]
35. MeSH descriptor: [Vertebral Artery] this term only and with qualifier(s): [surgery ‐ SU]
36. #29 or #30 or #31 or #32 or #33 or #34 or #35
37. #28 or #36 in Trials

Appendix 2. MEDLINE Ovid search strategy

1. vascular surgical procedures/
2. endovascular procedures/ or angioplasty/ or angioplasty, balloon/ or angioplasty, balloon, laser‐assisted/
3. cerebral revascularization/ or Blood Vessel Prosthesis/ or Blood Vessel Prosthesis Implantation/
4. stents/ or dilatation/ or catheterization/
5. (angioplasty or stent$ or pta or revasculari?ation or recanali?ation or catheter$ or dilatation).tw.
6. or/1‐5
7. exp intracranial arterial diseases/ or intracranial arteriosclerosis/ or vertebrobasilar insufficiency/
8. exp cerebral arteries/ or basilar artery/ or vertebral artery/
9. exp arterial occlusive diseases/ or arteriosclerosis/ or constriction, pathologic/
10. 8 and 9
11. ((intracranial or intra‐cranial or cerebral arter$ or basilar arter$ or vertebral arter$ or vertebrobasilar or vertebro basilar) adj5 (stenos?s or isch?emia or insufficien$ or arteriosclerosis or atherosclerosis or occlus$ or obstruct$ or plaque$)).tw.
12. 7 or 10 or 11
13. 6 and 12
14. *Carotid Artery Diseases/su, th [Surgery, Therapy]
15. *Carotid Stenosis/su, th [Surgery, Therapy]
16. *Vertebrobasilar Insufficiency/su, th [Surgery, Therapy]
17. *Brain Ischemia/su, th [Surgery, Therapy]
18. *Cerebral Arteries/su [Surgery]
19. *Basilar Artery/su [Surgery]
20. *Vertebral Artery/su [Surgery]
21. or/14‐20
22. 13 or 21
23. randomized controlled trial.pt.
24. controlled clinical trial.pt.
25. randomized.ab.
26. placebo.ab.
27. randomly.ab.
28. trial.ab.
29. groups.ab.
30. 23 or 24 or 25 or 26 or 27 or 28 or 29
31. 22 and 30

Appendix 3. Embase Ovid search strategy

1. endovascular surgery/ or vascular surgery/
2. angioplasty/ or bare metal stenting/ or carotid angioplasty/ or carotid artery stenting/ or patch angioplasty/
3. cerebral revascularization/
4. blood vessel prosthesis/ or artery prosthesis/
5. balloon dilatation/ or stent/ or exp metal stent/ or exp self expanding stent/
6. (angioplasty or stent$ or pta or revasculari?ation or recanali?ation or catheter$ or dilatation).tw.
7. or/1‐6
8. cerebral artery disease/ or brain atherosclerosis/ or vertebrobasilar insufficiency/
9. exp brain artery/ or vertebral artery/
10. peripheral occlusive artery disease/ or atherosclerosis/ or arteriosclerosis/ or atherosclerotic plaque/ or brain atherosclerosis/ or carotid atherosclerosis/
11. 9 and 10
12. ((intracranial or intra‐cranial or cerebral arter$ or basilar arter$ or vertebral arter$ or vertebrobasilar or vertebro basilar) adj5 (stenos? s or isch?emia or insufficien$ or arteriosclerosis or atherosclerosis or occlus$ or obstruct$ or plaque$)).tw.
13. 8 or 11 or 12
14. 7 and 13
15. carotid artery disease/su, th [Surgery, Therapy]
16. carotid artery obstruction/su, th [Surgery, Therapy]
17. vertebrobasilar insufficiency/su, th [Surgery, Therapy]
18. brain ischemia/su, th [Surgery, Therapy]
19. brain artery/su [Surgery]
20. basilar artery/su [Surgery]
21. vertebral artery/su [Surgery]
22. or/15‐21
23. 14 or 22
24. Randomized Controlled Trial/ or "randomized controlled trial (topic)"/
25. randomisation/
26. Controlled clinical trial/ or "controlled clinical trial (topic)"/
27. control group/ or controlled study/
28. clinical trial/ or "clinical trial (topic)"/ or phase 1 clinical trial/ or phase 2 clinical trial/ or phase 3 clinical trial/ or phase 4 clinical trial/
29. Crossover Procedure/
30. Double Blind Procedure/
31. Single Blind Procedure/ or triple blind procedure/
32. placebo/ or placebo effect/
33. (random$ or RCT or RCTs).tw.
34. controlled adj5 (trial$ or stud$)).tw.
35. (clinical$ adj5 trial$).tw.
36. ((control or treatment or experiment$ or intervention) adj5 (group$ or subject$ or patient$)).tw.
37. (quasi‐random$ or quasi random$ or pseudo‐random$ or pseudo random$).tw.
38. ((control or experiment$ or conservative) adj5 (treatment or therapy or procedure or manage$)).tw.
39. ((singl$ or doubl$ or tripl$ or trebl$) adj5 (blind$ or mask$)).tw.
40. (cross‐over or cross over or crossover).tw.
41. (placebo$ or sham).tw.
42. trial.ti.
43. (assign$ or allocat$).tw.
44. controls.tw.
45. or/24‐44
46. 23 and 45

Appendix 4. Science Citation Index Web of Science search strategy

1. TS=(angioplasty OR stent* OR pta OR revascularization OR revascularisation OR recanalisation OR recanalization OR recanaliation OR catheter* OR dilatation)
2. TS=(intracranial OR intra‐cranial OR cerebral arter* OR basilar arter* OR vertebral arter* OR vertebrobasilar OR vertebro basilar) AND TS=(stenosis OR stenosis OR ischemia OR ischaemia OR insufficien* OR arteriosclerosis OR atherosclerosis OR occlus* OR obstruct* OR plaque*)
3. #1 AND #2
4. TS=(randomized OR placebo OR randomly OR trial OR groups)
5. #3 AND #4

Appendix 5. Scopus search strategy

1. INDEXTERMS("endovascular procedures" OR angioplasty OR "vascular surgical procedures" OR "cerebral revascularization" OR "Blood Vessel Prosthesis" OR "Blood Vessel Prosthesis Implantation" OR "balloon dilatation" OR "stents" OR "dilatation" OR "catheterization" )
2. TITLE‐ABS ( angioplasty OR stent* OR pta OR revasculari?ation OR recanali?ation OR catheter* OR dilatation )
3. #1 OR #2
4. INDEXTERMS("intracranial arterial diseases" OR "intracranial arteriosclerosis" OR "vertebrobasilar insufficiency")
5. INDEXTERMS("cerebral arteries" OR "basilar artery" OR "vertebral artery" )
6. INDEXTERMS("arterial occlusive diseases" OR arteriosclerosis OR "constriction, pathologic")
7. #5 AND #6
8. TITLE‐ABS ( intracranial OR intra‐cranial OR "cerebral arter*" OR "basilar arter*" OR "vertebral arter*" OR vertebrobasilar OR "vertebro basilar" ) W/5 ( stenos?s OR isch?emia OR insufficien* OR arteriosclerosis OR atherosclerosis OR occlus* OR obstruct* OR plaque* )
9. #4 OR #7 OR #8
10. #3 AND #9
11. TITLE‐ABS ( randomized OR placebo OR randomly OR trial OR groups )
12. INDEXTERMS ( "randomized controlled trial" OR "controlled clinical trial" )
13. #11 OR #12
14. #10 AND #13

Appendix 6. Academic Source Complete EBSCO search strategy

1. SU endovascular procedures OR SU angioplasty OR SU vascular surgical procedures OR SU cerebral revascularization OR SU Blood Vessel Prosthesis OR SU Blood Vessel Prosthesis Implantation OR SU balloon dilatation OR SU stents OR SU dilatation OR SU catheterization
2. angioplasty OR stent* OR pta OR revascularization OR revascularisation OR recanalisation OR recanalization OR recanaliation OR catheter* OR dilatation
3. #1 OR #2
4. SU intracranial arterial diseases OR SU intracranial arteriosclerosis OR SU vertebrobasilar insufficiency
5. SU cerebral arteries OR SU basilar artery OR SU vertebral artery
6. SU arterial occlusive diseases OR SU arteriosclerosis OR SU constriction, pathologic
7. #5 AND #6
8. (intracranial OR intra‐cranial OR cerebral arter* OR basilar arter* OR vertebral arter* OR vertebrobasilar OR vertebro basilar) AND (stenosis OR stenosis OR ischemia OR ischaemia OR insufficien* OR arteriosclerosis OR atherosclerosis OR occlus* OR obstruct* OR plaque*)
9. #4 OR #7 OR #8
10. #3 AND #9
11. SU randomized controlled trial OR SU controlled clinical trial OR AB randomized OR AB placebo OR AB randomly OR AB trial OR AB groups
12. #10 AND #11

Appendix 7. China Biological Medicine Database search strategy

1. ((((((((("血管内操作"[不加权:不扩展]) OR "血管成形术"[不加权:扩展]) OR "血管外科手术"[不加权:不扩展]) OR "脑血管重建术"[不加权:不扩展]) OR "人工血管"[不加权:不扩展]) OR "血管假体植入"[不加权:不扩展]) OR "血管成形术, 气囊"[不加权:不扩展]) OR "支架"[不加权:不扩展]) OR "扩张术"[不加权:不扩展]) OR "导管插入术"[不加权:不扩展]
2. 血管 OR 血运
3. 重建 OR 再建 OR 再生 OR 再造 OR 重造 OR 重生 OR 成形
4. (#3) AND (#2)
5. 导管 OR 扩张 OR 膨胀 OR 支架 OR PTA
6. (#5) OR (#4) OR (#1)
7. (("颅内动脉疾病"[不加权:扩展]) OR "颅内动脉硬化"[不加权:不扩展]) OR "椎底动脉供血不足"[不加权:不扩展]
8. (("脑动脉"[不加权:扩展]) OR "基底动脉"[不加权:不扩展]) OR "椎动脉"[不加权:不扩展]
9. (("动脉闭塞性疾病"[不加权:扩展]) OR "动脉硬化"[不加权:不扩展]) OR "缩窄, 病理性"[不加权:不扩展]
10. (#9) AND (#8)
11. 颅内动脉 OR 脑动脉 OR 基底动脉 OR 椎动脉 OR 椎基底动脉
12. 狭窄 OR 缺血 OR 供血不足 OR 动脉粥样硬化 OR 动脉硬化 OR 粥样硬化 OR 阻塞 OR 闭塞 OR 梗阻 OR 斑块
13. (#12) AND (#11)
14. (#13) OR (#10) OR (#7)
15. (#14) AND (#6)
16. (((((("颈动脉疾病/外科学"[加权:不扩展]) OR "颈动脉狭窄/外科学"[加权:不扩展]) OR "椎底动脉供血不足/外科学"[加权:不扩展]) OR "脑缺血/外科学"[加权:不扩展]) OR "脑动脉/外科学"[加权:不扩展]) OR "基底动脉/外科学"[加权:不扩展]) OR "椎动脉/外科学"[加权:不扩展]
17. ("随机对照试验"[不加权:扩展]) OR "临床对照试验"[不加权:扩展]
18. (((("随机"[摘要:智能]) OR "安慰剂"[摘要:智能]) OR "试验"[摘要:智能]) OR "分组"[摘要:智能]) OR "对照"[摘要:智能]
19. (#18) OR (#17)
20. (#16) OR (#15)
21. (#20) AND (#19)
22. ((#30) AND (#19)) AND ( 随机对照试验[文献类型])
23. 动物[特征词] NOT 人类[特征词]
24. #22 NOT #23

Appendix 8. ClinicalTrials.gov search strategy

1. (angioplasty OR stent OR surgery OR revascular* OR recanal* OR catheter* OR dilatation ) | Interventional Studies | intracranial artery OR cerebral artery OR vertebral OR basilar OR vertebrobasilar

Appendix 9. WHO ICTRP search strategy

1. intracranial AND angioplasty OR intracranial AND stent OR intracranial AND PTA OR intracranial AND revascularization OR intracranial AND revascularisation OR intracranial AND recanalisation OR intracranial AND recanalization OR intracranial AND recanaliation OR intracranial AND catheterisation OR intracranial AND dilatation OR intracranial AND endovascular procedures OR intracranial AND surgical OR intracranial AND blood vessel prosthesis
2. cerebral arteries AND angioplasty OR cerebral arteries AND stent OR cerebral arteries AND PTA OR cerebral arteries AND revascularization OR cerebral arteries AND revascularisation OR cerebral arteries AND recanalisation OR cerebral arteries AND recanalization OR cerebral arteries AND recanaliation OR cerebral arteries AND catheterisation OR cerebral arteries AND dilatation OR cerebral arteries AND endovascular procedures OR cerebral arteries AND surgical OR cerebral arteries AND blood vessel prosthesis
3. vertebral AND angioplasty OR vertebral AND stent OR vertebral AND PTA OR vertebral AND revascularization OR vertebral AND revascularisation OR vertebral AND recanalisation OR vertebral AND recanalization OR vertebral AND recanaliation OR vertebral AND catheterisation OR vertebral AND dilatation OR vertebral AND endovascular procedures OR vertebral AND surgical OR vertebral AND blood vessel prosthesis
4. basilar AND angioplasty OR basilar AND stent OR basilar AND PTA OR basilar AND revascularization OR basilar AND revascularisation OR basilar AND recanalisation OR basilar AND recanalization OR basilar AND recanaliation OR basilar AND catheterisation OR basilar AND dilatation OR basilar AND endovascular procedures OR basilar AND vascular surgical procedures OR basilar AND blood vessel prosthesis

Appendix 10. Stroke Trials Registry search strategy

1. angioplasty
2. stent
3. stents
4. PTA
5. revascularization
6. revascularisation
7. recanalisation
8. catheter
9. catheters
10. dilatation
11. endovascular procedures
12. vascular surgical procedures
13. Blood Vessel Prosthesis

Data and analyses

Comparison 1. Primary outcomes: safety outcomes.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Short‐term death or stroke	3	632	Risk Ratio (M‐H, Fixed, 95% CI)	3.07 [1.80, 5.24]
1.2 Short‐term death or stroke with all types of endovascular therapy	2	562	Risk Ratio (M‐H, Fixed, 95% CI)	3.08 [1.80, 5.29]
1.2.1 Self‐expandable stent versus CMT	1	451	Risk Ratio (M‐H, Fixed, 95% CI)	2.57 [1.39, 4.76]
1.2.2 Balloon‐mounted stent versus CMT	1	111	Risk Ratio (M‐H, Fixed, 95% CI)	5.18 [1.61, 16.68]
1.3 Short‐term death or stroke under different time from qualifying event to performance	3	632	Risk Ratio (M‐H, Fixed, 95% CI)	3.07 [1.80, 5.24]
1.3.1 Time interval within 3 weeks	2	562	Risk Ratio (M‐H, Fixed, 95% CI)	3.08 [1.80, 5.29]
1.3.2 Time interval more than 3 weeks	1	70	Risk Ratio (M‐H, Fixed, 95% CI)	2.84 [0.12, 67.36]

Comparison 2. Secondary outcomes.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Short‐term ipsilateral stroke	3	632	Risk Ratio (M‐H, Fixed, 95% CI)	3.54 [1.98, 6.33]
2.2 Short‐term ischaemic stroke	3	632	Risk Ratio (M‐H, Fixed, 95% CI)	2.52 [1.37, 4.62]
2.3 Short‐term haemorrhagic stroke	3	632	Risk Ratio (M‐H, Fixed, 95% CI)	15.53 [2.10, 115.16]
2.4 Short‐term TIA	2	181	Risk Ratio (M‐H, Fixed, 95% CI)	0.52 [0.11, 2.35]
2.5 Short‐term death	3	632	Risk Ratio (M‐H, Fixed, 95% CI)	5.53 [0.98, 31.17]
2.6 Long‐term death or stroke	3	632	Risk Ratio (M‐H, Fixed, 95% CI)	1.69 [1.21, 2.36]
2.7 Long‐term ipsilateral stroke	3	632	Risk Ratio (M‐H, Fixed, 95% CI)	2.28 [1.52, 3.42]
2.8 Long‐term ischaemic stroke	3	632	Risk Ratio (M‐H, Fixed, 95% CI)	2.07 [1.37, 3.13]
2.9 Long‐term haemorrhagic stroke	2	521	Risk Ratio (M‐H, Fixed, 95% CI)	10.13 [1.31, 78.51]
2.10 Long‐term TIA	2	181	Risk Ratio (M‐H, Fixed, 95% CI)	0.82 [0.32, 2.12]
2.11 Long‐term death	3	632	Risk Ratio (M‐H, Fixed, 95% CI)	1.20 [0.50, 2.86]
2.12 Long‐term dependency	3	613	Risk Ratio (M‐H, Fixed, 95% CI)	1.90 [0.91, 3.97]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Chimowitz 2011.

Study characteristics
Methods	Multicentre, prospective, randomised, controlled, open‐label clinical trial with 3‐year follow‐up
Participants	Anticipated 764 participants, but actually randomised 451 participants who had TIA or non‐disabling ischaemic stroke (mRS ≤ 3) and intracranial stenosis with 70% to 99% degree of stenosis. Symptomatic qualifying arteries included internal carotid artery, middle cerebral artery, vertebral artery, and basilar artery.
Interventions	ET plus CMT CMT
Outcomes	Primary endpoints any stroke or death within 30 days after enrolment any stroke or death within 30 days after a revascularisation procedure of the qualifying lesion during follow‐up ischaemic stroke in the territory of the qualifying artery beyond 30 days Secondary endpoints disabling stroke any stroke or death myocardial infarction major non‐stroke haemorrhage (i.e. major systemic haemorrhage, subdural haemorrhage, epidural haemorrhage) functional outcome at the end of follow‐up measured by the Rankin Scale and BI cognitive outcome at end of follow‐up measured by the MoCA
Funding source	Funded by the US National Institute of Neurological Disorders and Stroke
Notes	ClinicalTrials.gov identifier: NCT00576693
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The sequence of treatment assignments are stratified by centre and produced at the Statistical Coordinating centre using a pseudo‐random number generator with randomly permuted block sizes"
Allocation concealment (selection bias)	Low risk	Quote: "Patients are randomized (1:1) to aggressive medical management alone or to PTAS plus aggressive medical management"
Blinding of participants and personnel (performance bias) All outcomes	High risk	Treatment is dramatically different, which could be identified.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "All potential endpoints are adjudicated by independent panels of neurologists and cardiologists who are unaware of treatment assignments"
Incomplete outcome data (attrition bias) All outcomes	Low risk	Incomplete outcome data adequately resolved and unlikely to seriously alter the results with ITT analysis.
Selective reporting (reporting bias)	Low risk	All of the study's outcomes were prespecified in the published protocol.
Other bias	Low risk	Although the study was stopped early, the final results would not have changed due to the superiority of CMT over ET, and the probability of a reversal is unlikely according to an interim analysis.

Miao 2012.

Study characteristics
Methods	Single‐centre prospective, randomised, controlled clinical trial with 1‐year follow‐up
Participants	Anticipated 184 participants but actually randomised 70 participants aged 25 to 75 years old with TIA or stroke in the territory of unilateral MCA stenosis with 70% to 99% degree of stenosis, stenotic length < 10 mm, and distal vessel diameter > 2 mm. The mRS score was less than 3, and the NIHSS score was less than 15.
Interventions	ET plus CMT CMT
Outcomes	The primary endpoint was ipsilateral stroke, TIA, or death at 30 days and 1 year.
Funding source	Funded by the Research Fund of Capital Medical Development (2007–2069)
Notes	Clinical Trial Registration URL: www.chictr.org/en/ Unique identifier: ChiCTR‐TRC‐12001989
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The randomization numbers were produced by Randlist software (Version 1.2 \Seed: [1497369600]; http://www.randomisation.net), 4 numbers as a block..."
Allocation concealment (selection bias)	Low risk	Quote: "Patients were randomized to either the PTAS or medical group"
Blinding of participants and personnel (performance bias) All outcomes	High risk	Treatment is dramatically different, which could be identified.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Investigators who did not participate in the enrolment of the patients and were blinded to the treatment assignment followed..."
Incomplete outcome data (attrition bias) All outcomes	High risk	The rate of loss to 1‐year follow‐up was 30% (49/70), which is high, and 19.4% participants in the ET group refused endovascular therapy.
Selective reporting (reporting bias)	Low risk	All of the study's outcomes were prespecified in the registered information.
Other bias	High risk	The study was stopped early; according to an interim analysis, the results showed the safety and efficacy of CMT and ET was similar. However, the final results could have changed if recruitment had continued.

Zaidat 2015.

Study characteristics
Methods	Multicentre, randomised, controlled trial with 1‐year follow‐up
Participants	Anticipated 250 participants, but actually randomised 112 participants aged 18 to 85 years who had at least 1 symptomatic neurovascular lesion (70% to 99%) stenosis with a hard TIA or stroke attributable to the territory of the lesion within the past 30 days and mRS score of 3 or less. Specific locations of intracranial qualifying arteries were not reported.
Interventions	ET plus CMT CMT
Outcomes	Primary endpoints stroke in the same territory (distal to the target lesion) as the presenting event within 12 months of randomisation hard TIA in the same territory (distal to the target lesion) as the presenting event from day 2 through month 12 postrandomisation Secondary endpoints stent success, defined as deployment of the PHAROS Vitesse stent across the target lesion with residual stenosis 0% to 20% percentage of stent group participants with any (symptomatic or asymptomatic) in‐stent stenosis 70% or higher confirmed by angiogram at 12 months percentage of stent group participants with symptomatic in‐stent restenosis 70% or higher confirmed by angiogram at 12 months percentage of medical therapy group participants with interventional procedure (e.g. angioplasty or stent) at 12 months comparison of NIHSS scores between the 2 treatment arms comparison of mRS scores between the 2 treatment arms
Funding source	Initiated and funded by Micrus Endovascular
Notes	ClinicalTrials.gov identifier: NCT00816166
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Patients meeting the final enrolment criteria after angiogram were randomly assigned 1:1 to either treatment with medical therapy alone or to medical therapy and PHAROS Vitesse neurovascular stent by a telephonic interactive voice response system (BioClinica Inc, Newtown, PA, USA). Randomization was stratified by 2 factors: enrolment site and age (18‐55 versus 56‐85 years)..."
Allocation concealment (selection bias)	Low risk	Quote: "Patients meeting the final enrolment criteria after angiogram were randomly assigned 1:1 to either treatment with medical therapy alone or to medical therapy and PHAROS Vitesse neurovascular stent by a telephonic interactive voice response system (BioClinica Inc, Newtown, PA, USA). Randomization was stratified by 2 factors: enrolment site and age (18‐55 versus 56‐85 years)..."
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote: "The current trial was not double‐blinded due to the lack of feasibility of masking the stent group..."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "... end point assessment by an independent neurologist who was not involved in the procedure reduced potential bias..."
Incomplete outcome data (attrition bias) All outcomes	Low risk	Incomplete outcome data adequately resolved and unlikely to seriously alter the results with ITT analysis.
Selective reporting (reporting bias)	Low risk	All of the study's outcomes were prespecified in the published protocol.
Other bias	Low risk	Although the study was stopped early, the final results would not have changed due to the superiority of CMT over ET, and the probability of a reversal is unlikely according to an interim analysis. The industry funding source might be a potential source of bias.

BI: Barthel Index
ET: endovascular therapy
CMT: conventional medical treatment
ITT: intention‐to‐treat
MCA: middle cerebral artery
MoCA: Montreal Cognitive Assessment
mRS: modified Rankin Scale
NIHSS: National Institutes of Health Stroke Scale
PTAS: percutaneous transluminal angioplasty and stenting
TIA: transient ischaemic attack

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Compter 2015	The trial recruited 115 participants who had vertebral atherosclerotic stenosis, but only 16.5% of participants had intracranial vertebral stenosis, and we could not extract detailed data related to intracranial atherosclerotic stenosis to analyse.
Coward 2007	The trial only enrolled 16 participants who had vertebral atherosclerosis including extracranial and intracranial artery to compare endovascular treatment versus best medical treatment alone. We could not extract detailed data related to intracranial atherosclerotic stenosis to analyse.
Markus 2017	The trial recruited 179 participants who had vertebral atherosclerotic stenosis, but only 17.3% of participants had intracranial vertebral stenosis, and we could not extract detailed data related to intracranial atherosclerotic stenosis to analyse.

Characteristics of ongoing studies [ordered by study ID]

Cui 2016.

Study name	Angioplasty and stenting for patients with symptomatic intracranial atherosclerosis
Methods	Ongoing, multicentre, randomised trial in 3 high‐volume centres
Participants	Anticipated 539 patients aged 40 to 70 years who have at least 3 TIAs or 1 ischaemic stroke with 70% to 99% degree of intracranial carotid artery or MCA under maximal dosage of medical management as well as with cerebral hypoperfusion in the territory of target vessels
Interventions	PTAS plus aggressive medical management performing beyond a time interval of 3 weeks Aggressive medical management
Outcomes	Primary outcome: the incidence of ipsilateral stroke (including haemorrhagic or ischaemic stroke) or death at 30 days after randomisation Secondary outcomes incidence of recurrent ischaemic stroke in the stenting‐involved vascular territory at 30 days and 3 and 12 months after randomisation rate of participants with residual stenosis < 30% at 3 and 12 months after randomisation incidence of in‐stent restenosis at 12 months after randomisation incidence of cerebral parenchymal, subarachnoid, or intraventricular haemorrhage at 30 days and 3 and 12 months after randomisation incidence of myocardial infarction or major non‐stroke haemorrhage (epidural or subdural haemorrhage or haemorrhage in major organs) at 30 days after randomisation C‐NIHSS: 28; mRS: 29; and CSQoL30 at 30 days and 3 and 12 months after randomisation
Starting date	April 2016
Contact information	ClinicalTrials.gov identifier: NCT02689037 and more detail can be obtained from the published protocol: DOI: 10.1136/bmjopen‐2016‐012175
Notes	This study was financially supported by Health Research Fund from People’s Liberation Army (No. 12MA100). The study sponsor and funders play no role in study design, collection, management, analysis, and interpretation of data, writing of the report, and the decision to submit the report for publication.

Gao 2015.

Study name	China Angioplasty and Stenting for Symptomatic Intracranial Severe Stenosis (CASSISS)
Methods	Ongoing, multicentre, randomised trial in 8 high‐volume centres
Participants	Anticipated 380 participants aged 30 to 80 years who have TIA or stroke within the past 12 months attributed to 70% to 99% stenosis of a major intracranial artery as well as with the occlusion of the terminal cortical branches or haemodynamic compromise (perforator occlusion excluded) in the territory of target vessels
Interventions	PTAS plus aggressive medical management performing beyond a time interval of 3 weeks Aggressive medical management
Outcomes	Primary endpoints stroke or death within 30 days after enrolment any stroke, death in the territory of the symptomatic intracranial artery beyond 30 days through 12 months Secondary endpoints disabling stroke or death beyond 30 days through 36 months in both arms complication rates associated with stenting procedures restenosis (> 50%) related to stenting within a follow‐up of 36 months any stroke, severe TIA, cardiovascular events related to stenting or medical therapy within a follow‐up of 36 months NIHSS, mRS, and BI assessment within a follow‐up of 36 months compliance rate of participants with regular medical therapy within a follow‐up of 36 months survival rate in both arms at the follow‐up of 12 and 36 months, respectively
Starting date	March 2014
Contact information	ClinicalTrials.gov identifier: NCT01763320 and more detail can be obtained from the published protocol: DOI: 10.1177/1591019915581778
Notes	This study was funded by the National Health and Family Planning Commission of the People's Republic of China (2011BAI08B04).

BI: Barthel Index
C‐NIHSS: Chinese version of National Institutes of Health Stroke Scale
MCA: middle cerebral artery
mRS: modified Rankin Scale
NIHSS: National Institutes of Health Stroke Scale
PTAS: percutaneous transluminal angioplasty and stenting
TIA: transient ischaemic attack

Differences between protocol and review

We added the index of type of recurrent event such as transient ischaemic attack, ischaemic stroke, and haemorrhage stroke in this review for better presentation of our analysis, which was different from the protocol (Wang 2019).

Contributions of authors

Dr Tao Wang and Dr Jichang Luo contributed equally to this review.

Draft the protocol: Tao Wang, Xue Wang, Jichang Luo, Vikram Jadhav, Liqun Jiao
Develop search strategy: Xue Wang, Tao Wang, Na Zhao
Search for trials: Xue Wang, Jichang Luo
Obtain copies of relevant references: Xue Wang, Jichang Luo, Tao Wang, Kun Yang
Trial selection: Jichang Luo, Xue Wang
Data extraction and data entry: Jichang Luo, Kun Yang, Peng Gao
Analysis, write the final review, and interpret the results: Kun Yang, Tao Wang, Peng Gao, Yan Ma, Vikram Jadhav, Na Zhao, Liqun Jiao

Declarations of interest

Tao Wang: none known
Jichang Luo: none known
Xue Wang: none known
Kun Yang: none known
Vikram Jadhav: none known
Peng Gao: none known
Yan Ma: none known
Na Zhao: none known
Liqun Jiao: none known

Joint first author

Joint first author

New