Angioplasty for intracranial artery stenosis

Abstract

Background

Intracranial artery stenosis causes up to 10% of all ischaemic strokes. The rate of recurrent vascular ischaemic events is very high. Angioplasty with or without stent placement is a feasible procedure to dilate the vessel affected. However, its safety and efficacy have not been systematically studied.

Objectives

To determine the efficacy and safety of angioplasty combined with best medical treatment compared with best medical treatment alone in patients with acute ischaemic stroke or transient ischaemic attack (TIA) resulting from intracranial artery stenosis for preventing recurrent ischaemic strokes, death, and vascular events.

Search methods

We searched the Cochrane Stroke Group Trials Register (last searched March 2006). In addition we searched the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library Issue 1, 2006), MEDLINE (1966 to March 2006), EMBASE (1980 to February 2006) and Science Citation Index (1945 to March 2006). To identify further published, unpublished and ongoing trials we searched reference lists of relevant articles and contacted authors and experts in the field.

Selection criteria

Randomised or otherwise controlled studies comparing best medical care plus angioplasty of the intracranial cerebral arteries, with or without stent placement, with best medical care alone. Studies were only included if data for clinical significant endpoints such as recurrent ischaemic stroke, haemorrhagic stroke and death were available.

Data collection and analysis

Two review authors selected trials for inclusion, and independently assessed trial quality and extracted data. Calculation of relative treatment effects with subgroup analysis was done if possible.

Main results

No randomised controlled trials were found. There were 79 articles of interest consisting of open‐label case series with three or more cases. The safety profile of the procedure showed an overall perioperative rate of stroke of 7.9% (95% confidence intervals (CI) 5.5% to 10.4%), perioperative death of 3.4% (95% CI 2.0% to 4.8%), and perioperative stroke or death of 9.5% (95% CI 7.0% to 12.0%). No comments can be made on the effectiveness of the procedure.

Authors' conclusions

At present there is insufficient evidence to recommend angioplasty with or without stent placement in routine practice for the prevention of stroke in patients with intracranial artery stenosis. The descriptive studies show that the procedure is feasible although carries a significant morbidity and mortality risk. Evidence from randomised controlled trials is needed to assess the safety of angioplasty and its effectiveness in preventing recurrent stroke.

Plain language summary

Angioplasty for intracranial artery stenosis

There is insufficient evidence to support the use of angioplasty for intracranial artery stenosis. Narrowing of the arteries inside the skull is a significant cause of stroke. Medical treatment for prevention consists of the control of risk factors such as high blood pressure, diabetes, and high cholesterol. Blood thinners are also used, but none has been demonstrated to be superior to another. Angioplasty, a procedure for opening narrowed arteries by means of a balloon or stent, is feasible but its safety and efficacy is not known. This review found no randomised controlled trials and no evidence to support the use of this procedure in routine practice. More research is needed to establish the role of this procedure in the treatment of this disease.

Background

Intracranial artery stenosis secondary to atherosclerosis is a significant cause of ischaemic stroke, accounting for 5% to 10% of all ischaemic strokes (Gorelick 1993; Hass 1968; Sacco 1995; Wityk 1996). Although the annual risk of stroke in patients with intracranial artery stenosis was estimated at 3% to 15% (Bogousslavsky 1986; Chimowitz 1995; Corston 1984; Craig 1982; EC/IC Bypass 1985; Hinton 1979; Marzewski 1982; Moufarrij 1986; Pessin 1987; WASID 1998; Wechsler 1986) a recent clinical trial showed a risk of recurrent vascular events at 15% to 17% (Chimowitz 2005; Kasner 2006). Patients with severe stenosis of the vertebral artery, the basilar artery, or both, are at particularly high risk of recurrent stroke despite antithrombotic therapy (Chimowitz 1995; Chimowitz 2005; Thijs 2000; WASID 1998). Retrospective studies suggested that chronic anticoagulation with warfarin was the best therapeutic alternative (Benesch 2000; Chimowitz 1995; WASID 1998); however, a randomised trial showed no benefit of warfarin over aspirin for the prevention of stroke or vascular events. Moreover, its use was associated with higher rates of adverse events such as bleeding (Chimowitz 2005). Despite claims of ballon angioplasty low complication rate associated with improved experience and technology (Gomez 2001), the reported rate of stroke or death during or after angioplasty ranges from 3% to 40%. Even fewer data exist on the evaluation of cerebral artery stenting. Although there have been several large series undertaken, these have included rather heterogeneous groups of patients with variable follow up (Abou‐Chebl 2005; Alazzaz 2000; Barakate 2001; Boulos 2005; Callahan 1997; Chow 2005; Connors 1999; Ferguson 1993; Gomez 2000b;Gomez 2000a; Mori 2000; Qureshi 2000; Rasmussen 2000). In this systematic review of all randomised or otherwise controlled studies, we aimed to compare the effectiveness of angioplasty with or without stent placement and medical care with medical care alone in patients with symptomatic or asymptomatic intracranial artery stenosis.

Objectives

The objective of this review was to determine the safety and effectiveness of percutaneous transluminal angioplasty with or without stent placement combined with best medical care compared to best medical care alone in preventing recurrent stroke in patients with either symptomatic or asymptomatic intracranial artery stenosis.

Methods

Criteria for considering studies for this review

Types of studies

We sought all randomised or quasi‐randomised trials comparing best medical care plus angioplasty of the intracranial cerebral arteries, with or without stent placement, with best medical care alone. Studies were only included if data for clinically significant endpoints such as ischaemic stroke, haemorrhagic stroke and death were available. Studies were assessed for comparability between both treatment assignment groups for major factors including age, gender, type of clinical presentation (asymptomatic versus stroke versus transient ischaemic attack), stroke location, vessel involved, time to randomisation, and time of follow up. Valid endpoints included ischaemic stroke, haemorrhagic stroke, death. An assessment was made of all complications related to the procedure and other endpoints such as length of stay and readmissions to hospital.

Types of participants

Trials including patients with asymptomatic or symptomatic intracranial cerebral artery stenosis caused by atherosclerosis were considered eligible for inclusion in this review. Intracranial stenosis was defined as stenosis greater than 50% only by conventional angiography. Symptomatic patients were defined as those people that presented with a transient ischaemic attack or an ischaemic stroke in the territory of the stenosed vessel. Studies including patients treated with angioplasty or stent placement or both as a salvage or rescue procedure during an acute ischaemic event with acute vascular occlusion were excluded.

Types of interventions

Angioplasty was defined as the percutaneous insertion of an endovascular catheter with a balloon designed to inflate at areas of vascular stenosis to dilate the vascular lumen. Stent placement refers to the deployment of a metallic mesh of a tubular device designed to dilate an area of vascular stenosis.

Trials with the following characteristics of the interventions were included. 
 (1) The active treatment group involved intracranial artery angioplasty with or without stent placement on a background of best medical treatment that included antithrombotic therapy. 
 (2) The control group involved medical treatment alone based on antithrombotic therapy that included antiplatelet agents or anticoagulant agents alone or in combination. Antithrombotic therapy is defined as the use of antiplatelet agents or anticoagulant agents, alone or in combination.

We planned to examine the patterns of medical care between the two arms of a study to determine any potential for confounding. The areas of medical care considered included use of antithrombotic agents, use of anticoagulation, specialist nursing, medical care or both, and variation in follow up. If such discrepancies existed between the arms, the study was considered poorly controlled and therefore was excluded from this review.

Types of outcome measures

We planned an intention‐to‐treat analysis. We sought the number of patients allocated to each treatment arm in each trial. Outcomes were analysed at 30 days and at one year after randomisation. A perioperative outcome was defined as an endpoint occurring during the procedure or within 30 days of the procedure. We divided the endpoints into primary and secondary outcomes as follows.

Primary outcomes

• Stroke in the vascular territory of the stenosed vessel, either ischaemic or haemorrhagic after randomisation. First stroke, either ischaemic or haemorrhagic occurring after randomisation (for asymptomatic patients).

• Death

• Stroke or death

Secondary outcomes

• Restenosis of the involved vessel documented by conventional cerebral angiography. Restenoses were classified as: (1) symptomatic, (2) asymptomatic.

• Other perioperative complications. These were classified as: (1) major; those requiring additional therapy, prolonged hospital stay or causing death, (2) minor; hematoma, wound infection, nerve palsy.

• Hospital length of stay

• Re‐admission

Strokes were classified, if possible, as disabling or non‐disabling. Disabling stroke was defined by the Glasgow Outcome Score (GOS) of 2 or 3 (Jennett 1975), a modified Rankin score (MRS) 3, 4, or 5 (Rankin 1957), or a Barthel index less than 80 out of 100 (Mahoney 1965). In the studies that do not provide information using a valid scale, we estimated the proportion of disabling stroke by seeking further detail if necessary. The GOS is often misquoted, with the scale inverted, and care was taken interpreting data given in this form. Death was considered a poor outcome. Restenosis was defined as stenosis greater than 50% from the original lumen diameter.

Search methods for identification of studies

See: 'Specialized register' section in Cochrane Stroke Group

We searched the Cochrane Stroke Group Trials Register, which was last searched by the Review Group Co‐ordinator in March 2006. In addition, we searched the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library Issue 1, 2006), MEDLINE (1966 to March 2006) (Appendix 1), EMBASE (1980 to February 2006) (Appendix 2) and Science Citation Index (1945 to March 2006).

Reference search

We searched the reference lists of all relevant articles for publications of relevant clinical trials.

Personal communication

We contacted a selection of specialists, both medical and surgical, enquiring about relevant published, unpublished, pending or recently commenced trials not already identified. We received information from Marc Chimowitz and Adnan Qureshi.

Data collection and analysis

The review authors (SCF and ALD) inspected the titles identified by the searches. Both review authors independently examined abstracts of references with titles of interest to determine their relevance. Where it was not clear from the abstract if the trial was relevant, or if no abstract was available, we obtained a copy of the article. We tried to contact the study author if further clarification was necessary. Any study deemed relevant was critically appraised using the checklist developed by Fowkes and Fulton (Fowkes 1991). Review authors discussed limitations of studies with a view to inclusion. Both review authors independently extracted and cross‐checked data from the included studies on to a pro‐forma. We specifically sought the following information.

• Method of randomisation and whether the randomising physician was blind to the treatment allocation.

• Method of measuring outcome and whether the adjudication of outcome was done independently, blindly, or both, to treatment allocation.

• Number of exclusions and losses to follow up.

Any disagreements were resolved by discussion. We planned to calculate the relative treatment effects by using a fixed‐effect model, and expressing odds ratios with 95% confidence intervals. We decided to use the Peto method to calculate odds ratios. Subgroup analyses were planned to stratify participants by the presence or absence of symptoms prior to the procedure, gender, degree of stenosis, and angioplasty alone or in combination with stent placement. To perform these analyses we looked for the presence or absence of overlap between the confidence intervals of the summary estimates of the groups considered. When in doubt, because of a small visual overlap in the confidence intervals, we used the significance test recommended by Deeks (Deeks 2001). We planned to perform sensitivity analyses to assess the effects on outcomes according to the degree of concealment of allocation, blinded outcome assessor or not, and degree of dropout. Heterogeneity would be sought by using the chi‐squared test. If heterogeneity was found we planned to discuss it and consult a statistician. We used The Cochrane Collaboration software Review Manager (RevMan 4.2).

Results

Description of studies

No completed or ongoing randomised controlled trials, or studies with comparative data, were found. Seventy‐nine articles with anecdotal case series were found, and are detailed in Table 1. These comprise a total of 1999 cases, spread among series of different sizes.

1. Anecdotal data on intracranial angioplasty (rate of events).

First author	Year	No of Cases	Average age	Periop stroke	Periop death	Periop stroke/death	Other periop complic	Mean f/u months	Stroke/death 1 year
Higashida 1993a	1993a	18		33.33	16.67	33.33	0.00		0.00
Higashida 1993b	1993b	8		37.50	12.50	37.50	0.00		25.00
Ferguson 1993	1993	102	63	7.84	0.98	8.82	0.00		0.00
Clark 1995	1995	17	68	11.76	0.00	11.76	17.65	22	5.88
McKenzie 1996	1996	11		0.00	0.00	0.00	27.27		0.00
Terada 1996	1996	12	61	16.67	0.00	16.67	0.00	16	16.67
Mori 1997	1997	35	63	5.71	2.86	5.71	0.00	16	8.57
Takis 1997	1997	10	65	40.00	0.00	40.00	50.00	9	20.00
Kubis 1997	1997	3	56	0.00	0.00	0.00	0.00		0.00
Callahan 1997	1997	15	63	13.33	6.67	13.33	0.00		0.00
Houdart 1997	1997	3	56	0.00	0.00	0.00	0.00	24	0.00
Yokote 1997	1997	17	64	11.76	0.00	11.76	5.88	14	0.00
Touho 1998	1998	43	65	2.33	0.00	2.33	6.98		0.00
Hyodo 1998	1998	38		0.00	0.00	0.00	10.53	40	18.42
Mandai 1998	1998	11	58	9.09	0.00	9.09	0.00	21	0.00
Kasita 1998	1998	74		5.41	1.35	6.76	0.00		13.51
Hacein‐Bey 1998	1998	12	67	8.33	8.33	8.33	25.00	12	33.33
Mori 1998	1998	42	59	4.76	0.00	4.76	4.76		14.29
Marks 1999	1999	23		4.35	4.35	4.35	4.35	35	8.70
Mori 1999	1999	11		0.00	0.00	0.00	0.00		0.00
Hyodo 1999	1999	47	63	8.51	0.00	8.51	10.64	44	12.77
Connors 1999	1999	70		4.29	1.43	4.29	30.00		0.00
Nomura 1999	1999	5	58	0.00	0.00	0.00	0.00	27	0.00
Morris 1999	1999	3	44	0.00	0.00	0.00	33.33	4	0.00
Terada 1999	1999	11		0.00	0.00	0.00	0.00		0.00
Horowitz 1999	1999	3	68	33.33	33.33	33.33	0.00		0.00
Eckard 1999	1999	8		0.00	0.00	0.00	37.50	53	37.50
Gomez 2000a	2000a	12	62	0.00	0.00	0.00	0.00	6	0.00
Mori 2000	2000	10	68	0.00	0.00	0.00	0.00	13	0.00
Alazzaz 2000	2000	16	60	12.50	0.00	12.50	18.75	13	0.00
Rasmussen 2000	2000	8	65	12.50	12.50	12.50	25.00		0.00
Nahser 2000	2000	20	59	10.00	10.00	10.00	0.00	8	5.00
Phatouros 2000	2000	2	67	0.00	0.00	0.00	0.00	6	0.00
Ramee 2001	2001	15	68	0.00	0.00	0.00	0.00		6.67
Terada 2001	2001	45		8.89	2.22	8.89	0.00	29	6.67
Yano 2001	2001	3	56	0.00	0.00	0.00	0.00	7	0.00
Levy 2001	2001	11	63	36.36	27.27	36.36	0.00	4	9.09
Barakate 2001	2001	11	66	18.18	0.00	18.18	27.27	18	9.09
Miao 2002	2002	7		14.29	0.00	14.29	14.29	3	0.00
Nakahara 2002	2002	2	59	0.00	0.00	0.00	0.00	13	0.00
Lee 2002	2002	10	48	0.00	0.00	0.00	10.00	34	20.00
Lylyk 2002	2002	34	54	8.82	2.94	8.82	5.88	5	2.94
Gress 2002	2002	25		28.00	8.00	28.00	0.00		0.00
Zhang 2003	2003	48		6.25	2.08	6.25	4.17		0.00
Jiang 2003	2003	42		9.52	2.38	9.52	2.38	8	0.00
Gupta 2003	2003	18	66	50.00	16.67	50.00	11.11		0.00
Uda 2003	2003	5	64	0.00	0.00	0.00	0.00	12	0.00
Levy 2003	2003	10	74	30.00	0.00	30.00	0.00	15	10.00
Liu 2004	2004	46	63	2.17	2.17	2.17	2.17	8	0.00
SSYLVIA 2004	2004	61	63	6.56	1.64	6.56	1.64	12	9.84
Rochemont 2004	2004	18	66	5.56	5.56	5.56	0.00	12	0.00
Tsuura 2004	2004	18	62	0.00	5.56	11.11	0.00	12	0.00
Matsumaru 2004	2004	62	66	1.61	3.23	4.84	4.84	43	25.81
Harakuni 2004	2004	62	64	4.84	1.61	12.90	9.68	46	16.13
Hauth 2004	2004	4	61	0.00	25.00	25.00	75.00		0.00
Kim 2004	2004	14	58	14.29	0.00	14.29	0.00	11	7.14
Tsumoto 2005	2005	16	67.5	0.00	0.00	0.00	0.00		0.00
Chow 2005	2005	39		25.64	5.13	28.21	0.00	13	0.00
Kessler 2005	2005	16	62	6.25	6.25	25.00	6.25	6	0.00
Henkes 2005	2005	15	64	6.67	0.00	6.67	6.67		0.00
Yu 2005	2005	18	69	5.56	5.56	11.11	16.67	26	16.67
Qureshi 2005	2005	24	61	4.17	4.17	12.50	16.67	20	12.50
Weber 2005	2005	22	67	9.09	4.55	13.64	4.55	9	0.00
Lee 2005b	2005b	7	62	0.00	0.00	0.00	14.29	10	0.00
Kim 2005	2005	17	64	5.88	0.00	5.88	11.76	17	5.88
Lee 2005a	2005a	17	59	5.88	5.88	17.65	70.59	10	0.00
Marks 2005	2005	36	62	2.78	5.56	8.33	33.33	52	22.22
Yoon 2005	2005	24	55	4.17	4.17	8.33	50.00	20	4.17
Abou‐Chebl 2005	2005	8	65	0.00	0.00	0.00	12.50	11	0.00
Boulos 2005	2005	13	65	0.00	0.00	0.00	0.00		0.00
Lylyk 2005	2005	104	67	3.85	1.92	5.77	6.73	6	1.92
Straube 2005	2005	9	64	0.00	0.00	0.00	22.22		0.00
Suh 2005	2005	35	53	0.00	2.86	8.57	8.57	22	0.00
Bose 2005	2005	45	66	4.44	2.22	4.44	20.00	6	8.89
Guimaraens 2005	2005	8	65	0.00	0.00	0.00	0.00		0.00
Hatano 2005	2005	15	68	0.00	0.00	0.00	0.00		6.67
Jiang 2006	2006	175	51	5.70	2.00	6.86	4.00	21	4.00
Total / Mean		1999		7.9% (5.5 to 10.4)	3.4% (2.0 to 4.8)	9.5% (7.0 to 12.0)	9.9% (6.4 to 13.4)		5.6% (3.7 to 7.6)

Risk of bias in included studies

No studies had methods suitable for inclusion in a meta‐analysis of controlled trials. The available data consisted of case series and anecdotal reports (Table 1). The patient characteristics, procedural methods and medical management made this population very heterogeneous. The participants were all selected from a population with an estimated high risk for recurrent stroke and death. Additionally, these studies defined success and good outcome differently. No attempt to quantify the effect of the treatment was made, as no comparison groups were found. Although these reports showed that angioplasty with or without stent placement can be performed, the associated complication rate and survival ranged from 0% to 50%. No comment can be made regarding the efficacy of the procedure since these reports might be subject to reporting bias. In fact, the series with fewer participants had the lower rate of complications reported.

Effects of interventions

No relevant trials were found in the Cochrane Stroke Group Trials Register or the Cochrane Central Register of Controlled Trials (CENTRAL). A total of 5884 references were retrieved from the electronic databases searches of which 79 were articles of interest comprising a total 1999 participants. Citation tracking identified further articles but no relevant trials. An ISCI search showed no further citations as all had already been retrieved by the EMBASE and MEDLINE searches. The rate for perioperative stroke was 7.95% (95% confidence intervals (CI) 5.5% to 10.4%) (range 0% to 50%), perioperative death 3.4% (95% CI 2.0% to 4.8%) (range 0% to 33%), perioperative stroke or death 9.5% (95% CI 7.0% to 12.0%) (range 0% to 50%), and other perioperative complications (such as groin hematoma and arterial dissection) 9.88% (95% CI 6.4% to 13.4%) (range 0% to 75%). Additionally, in those studies with follow up of at least one year, the risk of stroke or death was 5.6% (95% CI 3.7% to 7.6%) (range 0% to 50%).

Discussion

The studies retrieved consist of open label case reports and case series providing descriptive data and, therefore, no meta‐analytic approach was pursued. These data suggest that angioplasty with or without stent placement of the intracranial vessel is a feasible procedure but no comments can be made with regards to the effectiveness of the procedure. In addition, it is clear that the rate of complications can be very high. The evidence from the most recent randomised trial comparing warfarin to aspirin for the prevention of vascular events in patients with intracranial artery stenosis clearly indicates a very high annual risk of both stroke (20%) and the combined endpoint stroke, myocardial infarction and vascular death (23%) (Chimowitz 2005; Kasner 2006). Nevertheless, it cannot be assumed that angioplasty is the treatment of choice considering the high rate of complications reported to date in open label series that did not use randomisation, concealment of allocation, or blinded outcome assessment. Furthermore, two important facts from the WASID trial (Chimowitz 2005; Kasner 2006) need to be considered: first, the risk of recurrent stroke is highest within the first two weeks from the initial stroke; and second, the risk of stroke is highest among those patients with stenosis greater than 70%. Our findings and those from the WASID trial suggest that before adopting angioplasty with or without stent placement for intracranial stenosis in routine clinical practice, randomised controlled trials on this intervention should target the population at highest risk of recurrent stroke, such as those patients within two weeks from their initial event and those with stenosis greater than 70%, as an invasive, high‐risk intervention may not be warranted in patients without these findings. Therefore, results from well‐controlled trials are required before adopting this treatment for routine practice.

Authors' conclusions

Implications for practice.

At the present time there is insufficient evidence to recommend angioplasty with or without stent placement for stroke prevention in patients with transient ischaemic attacks or stroke secondary to intracranial artery stenosis in routine clinical practice.

Implications for research.

A controlled clinical trial comparing angioplasty with or without stent placement with best medical treatment is needed. Considering the average rate of complications found in the case series, such a trial must target the population at highest risk of recurrent stroke. In fact, assuming that the rate of stroke in this population is 20% per year and considering the point estimates (with its confidence intervals) for complication rates from angioplasty observed in this review, a trial comparing angioplasty with best medical treatment and alpha of 0.05, power of 80% and an expected relative risk reduction of 25%, will require 950 participants per treatment arm and a perioperative rate of stroke or death of less than 7%. A similar trial with an expected relative risk reduction of 50% will require 220 participants per treatment arm and a perioperative rate of stroke or death less than 3%.

What's new

Date	Event	Description
16 July 2008	Amended	Converted to new review format.

Appendices

Appendix 1. MEDLINE search strategy

The following search strategy, using controlled vocabulary (/) and free text terms (.tw), was used for MEDLINE and was modified for the other databases.

MEDLINE (Ovid) 1966 to March 2006 
 1. exp angioplasty/ 
 2. vascular surgical procedures/ 
 3. cerebral revascularization/ or blood vessel prosthesis implantation/ 
 4. balloon dilatation/ or stents/ or dilatation/ or catheterization/ 
 5. (angioplasty or stent$ or PTA or revasculari#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 cerebral arter$ or basilar arter$ or vertebral arter$ or vertebrobasilar) adj5 (stenos$s or isch?emia or insufficiency or arteriosclerosis or atherosclerosis or occlus$)).tw. 
 12. 7 or 10 or 11 
 13. 6 and 12 
 14. *brain ischemia/su or *vertebrobasilar insufficiency/su or exp *cerebral arteries/su or *basilar artery/su or *vertebral artery/su 
 15. 13 or 14 
 16. limit 15 to human

Appendix 2. EMBASE search strategy

EMBASE (Ovid) 1980 to February 2006 
 1. angioplasty/ or percutaneous transluminal angioplasty/ 
 2. exp microsurgery/ or endovascular surgery/ 
 3. balloon catheter/ or balloon dilatation/ 
 4. artery prosthesis/ or artery catheterization/ or artery dilatation/ 
 5. blood vessel catheterization/ or blood vessel prosthesis/ 
 6. stent/ 
 7. (angioplasty or stent$ or pta or revasculari#ation or catheter$ or dilatation).tw. 
 8. or/1‐7 
 9. exp brain artery/ and stenosis/ 
 10. brain atherosclerosis/ or exp occlusive cerebrovascular disease/ or vertebrobasilar insufficiency/ or cerebral artery disease/ 
 11. exp brain artery/ and (artery occlusion/ or exp arteriosclerosis/) 
 12. ((intracranial or cerebral arter$ or basilar arter$ or vertebral arter$ or vertebrobasilar) adj5 (stenos#s or isch?emia or insufficiency or arteriosclerosis or atherosclerosis or occlus$)).tw. 
 13. or/9‐12 
 14. 8 and 13 
 15. (intracranial adj10 (angioplasty or stent$)).tw. 
 16. 14 or 15 
 17. limit 16 to human

Contributions of authors

Salvador Cruz‐Flores: designed the review, ran the searches, assessed papers, conducted the analyses, and drafted the protocol and review. 
 Alan Diamond: assessed papers, drafted and edited the text.

Declarations of interest

Dr Cruz‐Flores received honoraria for being part of the Speakers Bureau for Bristol Myers Squibb.

Edited (no change to conclusions)