- Email: [email protected]
Medical and endovascular treatments of symptomatic intracranial stenosis. A Bayesian network meta-analysis
Journal of Clinical Neuroscience 63 (2019) 84–90
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical study
Medical and endovascular treatments of symptomatic intracranial stenosis. A Bayesian network meta-analysis q S. Vidale a,⇑, E. Agostoni b, G. Grampa a, A. Consoli c,d, D. Consoli e a
Department of Neurology & Stroke Unit, Sant’Anna Hospital, Como, Italy Department of Neurology, Niguarda Ca’ Granda Hospital, Milan, Italy c Service de Neuroradiologie Diagnostique et Thérapeutique, Hopital Foch, Paris, France d Interventional Neurovascular Unit, Careggi University Hospital, Florence, Italy e Department of Neurology, ‘‘G. Jazzolino” Hospital, Vibo Valentia, Italy b
a r t i c l e
i n f o
Article history: Received 8 April 2018 Accepted 28 January 2019
Keywords: Network meta-analysis Symptomatic intracranial stenosis Treatment
a b s t r a c t Intracranial stenosis is a well-established stroke risk factor with an increase of stroke recurrence or TIA up to 12.6% at 1 year. Treatments are different: medical and endovascular. We performed a multiple treatment comparison analysis to detect the best treatment in reducing the risk of stroke recurrence. We searched in Medline, Embase, Cochrane Central Register of Controlled Trials databases between 1979 and October 2017. Inclusion criteria were prospective randomized trials that evaluated patients with TIA or stroke due to intracranial stenosis and treated with different medical therapies and/or endovascular procedures. Primary endpoint was the recurrence of TIA or stroke in the territory of intracranial stenosis, while secondary endpoint was represented by any stroke or vascular death. Multiple treatment comparison meta-analysis based on a Bayesian fixed and random effects Poisson model was performed. Seven trials were included with a total of 1337 patients. At multiple treatment comparison, no significant differences between treatments were observed for both primary (median fixed effect standard OR: 0.40; 95%CI: 0.02–1.07) and secondary endpoints (median random effect standard OR: 1.17; 95%CI: 0.32–1.92). Treatment with aspirin alone ranked with high values both for primary and secondary endpoints (surface under the cumulative ranking curve of 70% and 82%, respectively). In patients with symptomatic intracranial stenosis, no differences between treatments were observed. However, aspirin alone was more effective than stenting in the reduction of TIA or stroke recurrences, with a better safety profile than oral anticoagulants. Ó 2019 Elsevier Ltd. All rights reserved.
1. Introduction Intracranial arterial stenosis is a well-established risk factor for ischemic stroke with different values of risk between races [1–4]. This condition is also associated with a high risk of recurrent stroke
q The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, a worldwide licence to the Publishers and its licensees in perpetuity, in all forms, formats and media (whether known now or created in the future), to i) publish, reproduce, distribute, display and store the Contribution, ii) translate the Contribution into other languages, create adaptations, reprints, include within collections and create summaries, extracts and/or, abstracts of the Contribution, iii) create any other derivative work(s) based on the Contribution, iv) to exploit all subsidiary rights in the Contribution, v) the inclusion of electronic links from the Contribution to third party material where-ever it may be located; and, vi) licence any third party to do any or all of the above. ⇑ Corresponding author at: Department of Neurology & Stroke Unit, Via Napoleona, 60, 22100 Como, Italy. E-mail address: [email protected] (S. Vidale).
https://doi.org/10.1016/j.jocn.2019.01.040 0967-5868/Ó 2019 Elsevier Ltd. All rights reserved.
[5,6]. Indeed, patients with transient ischemic attacks (TIA) or strokes with severe intracranial stenosis (70–99% of the arterial diameter) are at particular high risk for recurrent stroke in the territory of atherosclerotic artery, up to 12.6% at 1 year, despite the best medical management or control of the common vascular risk factors (i.e., blood hypertension, diabetes mellitus, dyslipidemia. . .) [7]. For these reasons, prompt and effective treatments have been applied in the past in two principal directions: a preventive strategy to reduce the risk of atherosclerotic thromboembolic process by the medical approach and a direct action on arterial stenosis by endovascular procedures.
1.1. Rationale The treatment options for the symptomatic intracranial stenosis include antiplatelet and anticoagulant agents, even if the efficacy and safety of these last therapies are debatable and aspirin
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90
appeared to be safer than warfarin in a previous randomized controlled trial [5,8]. Despite the best medical treatment, some patients suffered of a recurrent stroke, failing to respond to the pharmacological therapies [9]. For this reason, in the last decades several types of endovascular devices have been applied and the angioplasty with stenting has been proposed as an alternative treatment to the medical therapy alone. However, previous trials showed different results, without definitive conclusions. In literature, different randomized clinical trials have been published with comparative effects between two therapeutical interventions (i.e., endovascular and antiplatelets) but without a direct comparison with other therapies (i.e., endovascular versus anticoagulants). In other words, head-to-head randomized trials comparing some treatments of interest are lacking and for this reason a new comprehensive analysis might be useful to identify the best therapeutical option.
1.2. Objective Aim of this multiple treatment meta-analysis was to determinate the comparative effectiveness and safety of current different therapeutical approaches to symptomatic intracranial stenosis, considering both medical treatment and endovascular procedures and comparing these therapies directly and indirectly.
2. Methods We conducted this meta-analysis following the guidance coming from the extension statement of the PRISMA for reporting of systematic reviews incorporating network meta-analyses [10].
2.1. Literature search and eligibility criteria We searched the published literature using Medline, EMBASE and the Cochrane Trial Databases up to October 1st, 2017. Keywords of this search were ‘‘intracranial stenosis” or ‘‘intracranial atherosclerosis”, ‘‘stent” or ‘‘endovascular”, ‘‘medical therapy” or different combinations of ‘‘antiplatelet” or ‘‘antithrombotic” or ‘‘anticoagulant”. Specific Boolean syntax of the search is described in the Supplemental Material. We included in the analysis only prospective randomized trials that evaluated patients with TIA or stroke caused by an intracranial stenosis and treated with different medical therapies and/or endovascular procedures. Cohort studies, letters, comments, editorials, case reports, proceeding, personal communication and studies not providing quantitative outcome were excluded.
2.2. Data collection process Data were extracted independently by two reviewers with a consultation of a third reviewer in the case of disagreements. We extracted data concerning the population, the study design and the main outcomes. For our purposes, we consider two main endpoints. Primary endpoint was considered as the occurrence of TIA or ischemic stroke interesting the territory of the stenotic intracranial artery during the follow up. Secondary endpoint was represented by the occurrence of any stroke (ischemic or hemorrhagic) or vascular death. We conducted also additional analyses for stroke or major bleeding or all cause death only, and for the composite of these outcomes. Different risks of bias were reported using the recommendations of the Cochrane Handbook for Systematic Reviews of Intervention.
85
2.3. Patient involvement No patient was directly involved in this study, because we performed a meta-analysis of previous trials. 2.4. Statistical analysis The network meta-analysis was based on a Bayesian random effects Poisson regression model, which preserves randomised treatment comparisons within trials. Using a full Bayesian evidence network, all indirect comparisons are taken into account to arrive at a single, integrated estimate of the effect of all included treatments based on all included studies. The model considers numbers of patients with primary and secondary outcomes, as previously described, accumulating also years to estimate rate ratios. We calculated relative effects both for direct and indirect comparisons. We included in the analysis also a meta-regression to assess the potential effects of some modifiers: age, male gender, active smoking, time between event to randomization, follow-up time and statin use. A graphical representation of the network was assessed using a specific geometry and providing the specification of the nodes by the randomised intervention. Consistency was mainly assessed by the comparison of the conventional network meta-analysis model. Finally, heterogeneity for main studies was assessed calculating Cochrane’s Q and I2 values. We performed also a sensitivity analysis to determine the robustness of the observed outcomes. Doi and funnel plots were representative of the levels of publication bias for groups of randomised trials. One of the advantages of network metaanalysis is that it can provide information about the ranking of all evaluated interventions for the studied outcome. Probabilities are often estimated for a treatment being ranked at a specific place (first, second, etc.,) according to the outcome. We provided rankograms for primary and secondary endpoints as graphs. The cumulative rankograms present the probabilities that a treatment would be among the different best treatments. The surface under the cumulative ranking curve (SUCRA), a simple transformation of the mean rank, was used to provide a hierarchy of the treatments and accounts both for the location and the variance of all relative treatment effects. The larger the SUCRA value, the better the rank of the treatment. We provided SUCRA values for primary and secondary endpoints, as well as for additional analyses. WinBUGS for Microsoft Excel was used for the multiple treatment metaanalysis, integrated with an Excel-based tool software (NetMetaXL) [11]. 3. Results We included in this network meta-analysis 7 randomised clinical trials [8,12–17]. One of these trials was considered only for secondary outcome, because data concerning the recurrent strokes interesting the territory of stenotic artery were not available [14]. Other studies were excluded because without subgroup identification of intracranial atherosclerosis or primary cerebrovascular event [18–21]. The complete history of PRISMA flow diagram was represented in the Fig. 1. The studies recruited a total of 1337 patients. Four trials compared clopidogrel and aspirin treatments with the same therapy plus endovascular procedure. In two trials, aspirin was compared with warfarin (OA), while in one trial dual antiplatelet therapy (aspirin + clopidogrel) was matched to aspirin. Details of single studies were reported in Table 1. A total of 325 subjects were treated with stenting, while 310 patients were recruited in the group of dual antithrombotic therapy (aspirin plus clopidogrel). The largest group was the aspirin treatment with 399 patients, while the smallest group
86
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90
Fig. 2. Network geometry of randomised trials for primary endpoint. The line thickness is representative of the population weight of randomised trials included in the network meta-analysis ASA: aspirin; CLOP: clopidogrel; OA: oral anticoagulant.
3.1. Primary endpoint
Fig. 1. PRISMA flow diagram.
was represented by the anticoagulated one (303 subjects). We observed no significant differences in the size of these treatment groups, neither for age (p: 0.32).
The network geometry for this outcome is represented in Fig. 2. We calculated that 22.8% patients of the stenting group had a TIA or stroke in the territory of intracranial stenosis versus 16% of aspirin-treated patients. The lowest rate of primary events was observed in the warfarin-treated group (11.5%), followed by the aspirin plus clopidogrel group with 12.2%. In the direct comparison analysis, the OA versus aspirin group was the largest one, while antiplatelet monotherapy versus dual therapy was the smallest group. At multiple treatment comparison, we observed no significant differences between groups (median fixed effect standard with OR: 0.40–95%CI: 0.02–1.07). The single ORs of direct and indirect comparisons between treatments varied between 0.26 in favour of OA on stent treatment and 0.78 in favour of OA on aspirin therapy. Direct and indirect relative estimates are presented in Table 2A and the forest plot with different ORs is shown in Fig. 3A. We performed also a sensitivity analysis, excluding 2 stud-
Table 1 Characteristics of RCT included in the network meta-analysis.
Total size Mean age (yrs) Gender (M) Stenting Anticoagulants Statins (I/C) Antihypertensive drugs Hypoglycemic drugs Time from qualifying event to randomization (days; I/ C) Mean follow-up time (months) *
SAMMPRIS [13]
VISSIT [12]
Mohammadian [14]
VAST* [15]
CLAIR [16]
WASID [8]
Martì-Fabregas [17]
451 60 272 U
111 62 73 U
63 69 38 U
115 65 85 U
100 56 76
569 63 350
28 67 19
202/195 200/203 105/103 7/7
29/32 43/49 25/20 12/15
4/6 24/20 24/17 /
49/55 40/40 / 27/22
21/16 34/34 17/14 2.5/3.2
U 184/163 189/179 / 18/16
U 8/4 12/10 5/5 25/23
32.4
12
15.2
36
0.25
20
23
Referred to the entire sample; I: interventional group; C: control group.
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90 Table 2 A) Direct and indirect estimates (with low and high 95%CI values) for primary endpoint; B) Direct and indirect estimates (with low and high 95%CI values) for secondary endpoint. A) OA 0.78 (0.12–6.01) – –
– ASA 0.55 (0.08–3.42) 0.61 (0.15–3.05)
0.42 (0.04–5.83) – ASA + CLOP –
0.26 (0.02–3.37) – 0.26 (0.02–3.37) Stent
– ASA + CLOP – 0.93 (0.12–7.21)
– 1.17 (0.00–78.10) OA –
– – 0.83 (0.01–74.12) Stent
B) ASA 0.25 (0.03–11.01) 0.30 (0.01–3.08) 0.23 (0.01–6.03)
Direct estimates are represented in the lower cells, while indirect estimates in the higher. Comparisons are represented with the first treatment on the left bottom. ASA: aspirin; CLOP: clopidogrel; OA: oral anticoagulant.
ies with higher level of bias risk [14,17]. We did not observe a significant difference of OR when compared with the main analysis (OR: 0.41; 95%CI: 0.08–1.04). The rankogram showed a higher probability for OA in the first rank with a following reduction (from 64% to 7%), while stent treatment presented an inverse trend (from 2% to 76%). The dual antiplatelet treatment (aspirin plus clopidogrel) has the higher value at rank 3 (63%). Different values of treatment ranking are presented in Supplemental Material. The larger SUCRA value was observed for OA (SUCRA: 80%), followed by aspirin treatment (SUCRA: 70%), dual antiplatelet therapy (SUCRA: 39%) and finally stenting therapy (SUCRA: 11%). The total rankogram is showed in Supplemental Material. Inconsistency results varied between OR: 0.9895 for the comparison of aspirin plus clopidogrel vs aspirin alone and OR: 0.5919 for the indirect comparison of OA vs stent. The graph representing the ratio between consistency and inconsistency models is showed in Supplemental Material. Considering the heterogeneity of the trials included in this network meta-analysis, the Cochrane’s Q was 0.58 with an I2 value of 0% (p: 0.80). In the meta-regression analysis, we observed a significant influence of statin use only (S.E.: 0.028; p: 0.011). Results for each considered variables are presented in Supplemental Material. 3.2. Secondary endpoint We did not observe significant differences in the direct and indirect estimates between treatments. ORs varied significantly with lower value observed in the comparison between aspirin and stent (OR: 0.23; 95%CI: 0.01 – 6.03) and the higher value in the indirect estimate of dual antiplatelet therapy versus OA (OR: 1.17; 95%CI: 0.00–78.10). Other direct estimates, as well as the indirect estimates, are represented in the Table 2B and the forest plot was shown in Fig. 3B. The median Random effect standard was 1.41 (95%CI: 0.32–1.92). The rankogram showed a relative homogeneous risk probability between treatments with the exception for aspirin (Supplemental Material). This treatment had the highest value in rank 1 (63%) and a concomitant larger SUCRA value (SUCRA: 82%). Dual antiplatelet therapy, OA and stent treatment showed respectively a SUCRA of 41%, 41% and 37%. The Cochrane’s Q value was 5.86 with an I2: 66% (p: 0.05). 3.3. Additional analyses Considering only stroke as main outcome, we did not observe significant differences between the direct and indirect comparisons of different treatments. The higher difference was detected for the double antiplatelet therapy versus stenting (OR: 0.335; 95%CI: 0.020–1.662), while OA and ASA showed a nearly similarity
87
in their comparison (OR: 0.806; 95%CI: 0.058–10.530). The double antiplatelet treatment showed the best cumulative ranking with a SUCRA value of 72%, followed by OA (SUCRA: 58%), ASA (SUCRA: 48%) and finally by stenting (SUCRA: 22%). In the analysis for major bleeding, treatment with only ASA showed better OR when compared with OA or stenting with values of 0.236 and 0.244, respectively. Results of each comparison of treatments are presented in the Supplemental Material. Antiplatelet monotherapy showed the highest SUCRA value (84%), while OA had the lowest one (23%). Considering all cause death, the comparison between OA and stent showed an OR of 0.257 (95%CI: 0.021–3.248). A similar result was observed in the comparison between ASA and stent (OR: 0.324; 95%CI: 0.06–1.598). Other results were presented in Supplemental Material. Cumulative rankogram showed a best positioning of OA (SUCRA: 81%), followed by ASA (SUCRA: 70%). Stent ranked as last with SUCRA value of 11%. When we considered the composite of the three previous outcomes, we observed the superiority of ASA treatment versus stent (OR: 0.584; 95%CI: 0.046–13.071) or OA (OR: 0.683; 95%CI: 0.053– 8.711). The ranking of different treatments showed ASA with the higher SUCRA value (62%) and followed by the double antiplatelet therapy (SUCRA: 51%).
4. Discussion In this network meta-analysis concerning the treatment of intracranial stenosis in patients with stroke or TIA, we observed a higher risk of stroke recurrences in the territory of stenotic intracranial artery after stenting than medical treatment, even if this difference was not statistically significant. The direct comparison between aspirin and dual antiplatelet therapy did not show differences for efficacy of the treatments, as well as between antithrombotics and anticoagulants. Despite the fact that the head-to-head comparisons between treatments were not statistically significant, aspirin alone showed the best SUCRA value for all composite outcomes with 62%, followed by the double antiplatelet therapy (51%). For this reason, aspirin or aspirin plus clopidogrel appear to be the more effective treatments to reduce the risk of a recurrent stroke in the territory of intracranial stenosis, balanced with potential side-effects. This observation is in line with the current indications coming from the international guidelines [22]. In this analysis, we calculated a total NNT of 10 in the comparison of stenting versus dual antiplatelet therapy, while this value reduced to 3 in the comparison of angioplasty versus aspirin alone. The highest value of NNT was calculated comparing dual antiplatelet to aspirin alone (NNT: 33). Concomitant therapies (i.e. statins, antihypertensive drugs. . .) did not significantly differ between groups of the included trials. In particular, we calculated a total of 284 patients treated with statins in the endovascular group compared with 288 subjects in the medical group (Table 1). However, in the meta-regression analysis we observed that the statin use could have a potential effect on this primary endpoint. Considering the secondary endpoint, we did not observe significant difference between treatments, even if we observed a difference between aspirin and stenting groups of about 14.7%. A very recent meta-analysis showed similar results for the stenting treatment, but the other medical treatments were not included in the pooled analysis [23]. Despite the hope that endovascular treatment would be more beneficial in the management of intracranial stenosis than medical therapy, increasing the blood flow similarly to extracranial carotid artery, the effects of this procedure are not the same. Indeed, while the revascularization of extracranial arteries with surgery or stenting is more effective than best medical treatment, in this network meta-analysis similar evidences did not occur. A
88
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90
Fig. 3. Forest plots for primary (A) and secondary (B) endpoints.
possible explanation of this difference could be the site and size of the arterial stenosis. At the same time, for trials comparing stenting with antiplatelet therapy the variations in the type of devices (bare-metal stent or balloon-expandable and self-expanding stents), the differences of protocols and skill levels of professional performing endovascular procedures could influence the outcomes. The treatment ranking probabilities for this network meta-analysis showed a superiority of OA for the primary endpoint, but we observed a total better ranking for aspirin (SUCRA: 70% and 82% for primary and secondary outcomes, respectively). When we considered also additional analyses, we observed best ranking for double antiplatelet treatment for only stroke, but the therapy with ASA alone provided the best benefit/risk ratio in the composite outcomes (stroke + major bleeding + all cause death). Finally, for all trials the heterogeneity is significant only for sec-
ondary endpoint, without bias of publication for both primary and secondary endpoints, as represented in the funnel plot (Supplemental Material). Most of the trials presented low risks of bias for all items, as reported in the Fig. 4, with the exception of two studies [14,17]. For this reason we conducted a sensitivity analysis, excluding these trials, but the results did not differ significantly from the main analysis. We observed differences between the time from qualifying events to randomization. However, most of treated patients with stenting were randomised within two weeks, without differences with medical groups and the metaregression analysis did not show significant influence of this variable on primary endpoint. The main limitation of this network meta-analysis is the inability to perform subgroup analyses due to limited data in the included studies. In particular, some individual risk factors (i.e.
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90
89
In conclusion, in our multiple treatment meta-analysis, the antithrombotic therapy seems to be more effective and safer than stenting and in particular aspirin alone with better results for all composite outcomes. Oral anticoagulants could represent an effective alternative but with a lower profile in secondary endpoint and in some additional analyses. Considering the heterogeneity of study population and treatments, more trials with bettercontrolled randomization are needed. SV, EA, AC, and DC designed the study. SV and AC performed the literature search, data analysis, and statistical modelling. They contributed equally to the study. SV drafted the manuscript. EA, AC, GG and DC performed a critical review of the manuscript. SV is the guarantor. Data sharing statement N/A. Declarations of interest None. Funding source None. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.jocn.2019.01.040. References
Fig. 4. Risk of bias summary for each trial.
atrial fibrillation, diabetes mellitus, hemodynamic parameters. . .) and characteristics (i.e. age, gender. . .) could help to better investigate the profile of patient that best benefit from some treatments than others. In particular, it could be of great interest the comparison of treatment effect between race, because in Asian people an intracranial arterial stenosis could increase the stroke risk up to 30–50%, a value higher than in people living in Western Countries [1–4]. Another crucial point was the differentiation between anterior and posterior circulation. However, the results coming from sub-analysis in the SAMMPRIS study did not show significant differences concerning the primary outcome between anterior and posterior circulation. Finally, no data are available reporting tandem stenosis which are pathological conditions with an incidence between 2 and 13% for anterior circulation [24]. This review was not registered in PROSPERO and this could be another limitation. Strengths are represented by the clear definition of primary and secondary endpoints and the multiple comparison of different treatment with the concomitant reduction of bias introducing the use of random effect that attenuates the heterogeneity between study populations and methodology. Even if some other randomised trials are ongoing and final results will be expected in the next years [25], current literature and the evidence-based medicine are lacking of specific guidelines for the prevention of TIA and strokes in symptomatic intracranial atherosclerotic disease. However, a previous statement defined the main items and purposes to begin a new guideline project [26].
[1] Gorelick PB, Wong KS, Bae HJ, Pandey DK. Large artery incranial occlusive disease: a large worldwide burden but a relatively neglected frontier. Stroke 2008;39:2369–79. [2] Arenillas JF. Intracranial atherosclerosis: current concepts. Stroke 2011;42 (Suppl):S20–3. [3] Holmstedt CA, Turan TN, Chimowitz MI. Atherosclerotic intracranial arterial stenosis: risk factors, diagnosis, and treatment. Lancet Neurol 2013;12:1106–14. [4] Wong LK. Global burden of intracranial atherosclerosis. Int J Stroke 2006;1:158–9. [5] Kasner SE, Chimowitz MI, Lynn MJ, et al. Predictors of ischemic stroke in the territory of a symptomatic intracranial arterial stenosis. Circulation 2006;113:555–63. [6] Zaidat OO, Klucznik R, Alexander MJ, et al. The NIH registry on use of the Wingspan stent for symptomatic 70–99% intracranial arterial stenosis. Neurology 2008;70:1518–24. [7] Chaturvedi S, Turan TN, Lynn MJ, et al. Do patient characteristics explain the differences in outcome between medically treated patients in SAMMPRIS and WASID? Stroke 2015;46:2565–7. [8] Chimowitz MI, Lynn MJ, Howlett-Smith H, et al. Comparison of warfarin and aspirin for symptomatic intracranial arterial stenosis. N Engl J Med 2005;352:1305–16. [9] Thijs VN, Albers GW. Symptomatic intracranial atherosclerosis: outcome of patients who fail antithrombotic therapy. Neurology 2000;55:490–7. [10] Hutton B, Salanti G, Caldwell DB, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann Intern Med 2015;162:777–84. [11] Brown S, Hutton B, Clifford T, et al. A Microsoft-Excel-based tool for running and critically appraising network meta-analyses – an overview and application of NetMetaXL. Syst Rev 2014;3:110. [12] Zaidat OO, Fitzsimmons BF, Woodward BK, et al. Effect of balloon-expandable intracranial stent vs medical therapy on risk of stroke in patients with symptomatic intracranial stenosis. JAMA 2015;313:1240–8. [13] Derdeyn CP, Chimowitz MI, Lynn MJ, et al. Aggressive medical treatment with or without steniting in high-risk patients with intracranial artery stenosis (SAMMPRIS): the final results of randomised trial. Lancet 2014;383:333–41. [14] Mohammadian R, Pashapour A, Sharifipour E, et al. A comparison of stent implant versus medical treatment for severe symptomatic intracranial stenosis: a controlled clinical trial. Cerebrovasc Dis Extra 2012;2:108–20.
90
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90
[15] Compter A, van der Worp HB, Schonewille WJ, et al. Stenting versus medical treatment in patients with symptomatic vertebral artery stenosis: a randomised open-label phase 2 trial. Lancet Neurol 2015;14:606–14. [16] Wong KS, Chen C, Fu J, et al. Clopidogrel plus aspirin versus aspirin alone for reducing embolisation in patients with acute symptomatic cerebral or carotid artery stenosis (CLAIR study): a randomised, open-label, blinded-endpoint trial. Lancet Neurol 2010;9:489–97. [17] Martì-Fabregas J, Cocho D, Martì-Vilalta JL, et al. Aspirin or anticoagulants in stenosis of the middle cerebral artery: a randomized trial. Cerebrovasc Dis 2006;22:162–9. [18] Wong KS, Chen C, Ng PW, et al. Low-molecular-weight heparin compared with aspirin for the treatment of acute ischaemic stroke in Asian patients with large artery occlusive disease: a randomised study. Lancet Neurol 2007;6:407–13. [19] Wang Y, Wang Y, Zhao X, et al. Clopidogrel with aspirin in acute minor stroke or transient ischemic attack. N Engl J Med 2013;369:11–9. [20] Bhatt DL, Fox KAA, Hacke W, et al. Clopidogrel and aspirin versus aspirin alone fot he prevention of atherothrombotic events. N Engl J Med 2006;354:1706–17. [21] Diener HC, Bogousslavsky J, Brass LM, et al. Aspirin and clopidogrel compared with clopidogrel alone after recent ischemic stroke or transient ischemic attack in high-risk patients (MATCH): randomised, double-blind, placebocontrolled trial. Lancet 2004;364:331–7.
[22] Kernan WN, Ovbiagele B, Black HR, Bravata DM, Chimowitz MI, Ezekowitz MD, Schwamm LH, Wilson JA, et al. on behalf of the American heart association stroke council, council on cardiovascular and stroke nursing, council on clinical cardiology, and council on peripheral vascular disease. Guidelines for the prevention of stroke in patients with stroke and transient ischemic attack: a guideline for healthcare professionals from the American Heart Association/ American Stroke Association. Stroke 2014;45. [23] Zhang Q, Dong K, Song H. Comparison of stent versus medical therapy for symptomatic patients with intracranial atherosclerotic stenosis: a metaanalysis. J Neurol Sci 2017;372:272–8. [24] Rouleau PA, Huston 3rd J, Gilbertson J, et al. Carotid artery tandem lesions: frequency of angiographic detection and consequences for endarterectomy. AM J Neuroradiol 1999;20:621–5. [25] Cui XP, Lin M, Mu JS, et al. Angioplasty and stenting for patents with symptomatic intracranial atherosclerosis: study protocol of a randomised controlled trial. BMJ Open 2016;6:e012175. [26] Turan TN, Gronseth GS, Chimowitz M, et al. Protocol for proposed guideline project: prevention of stroke in patients with symptomatic large vessel intracranial atherosclerotic disease. 2015. https://www.aan.com/ uploadedFiles/Website_Library_Assets/Documents/2.Clinical_Guidelines/3. Browse_By_Status/4.Guidelines_Under_Development/Atherocslerosis-protocolfor-public-comment-final-7-9-2015.pdf.
Contents lists available at ScienceDirect
Journal of Clinical Neuroscience journal homepage: www.elsevier.com/locate/jocn
Clinical study
Medical and endovascular treatments of symptomatic intracranial stenosis. A Bayesian network meta-analysis q S. Vidale a,⇑, E. Agostoni b, G. Grampa a, A. Consoli c,d, D. Consoli e a
Department of Neurology & Stroke Unit, Sant’Anna Hospital, Como, Italy Department of Neurology, Niguarda Ca’ Granda Hospital, Milan, Italy c Service de Neuroradiologie Diagnostique et Thérapeutique, Hopital Foch, Paris, France d Interventional Neurovascular Unit, Careggi University Hospital, Florence, Italy e Department of Neurology, ‘‘G. Jazzolino” Hospital, Vibo Valentia, Italy b
a r t i c l e
i n f o
Article history: Received 8 April 2018 Accepted 28 January 2019
Keywords: Network meta-analysis Symptomatic intracranial stenosis Treatment
a b s t r a c t Intracranial stenosis is a well-established stroke risk factor with an increase of stroke recurrence or TIA up to 12.6% at 1 year. Treatments are different: medical and endovascular. We performed a multiple treatment comparison analysis to detect the best treatment in reducing the risk of stroke recurrence. We searched in Medline, Embase, Cochrane Central Register of Controlled Trials databases between 1979 and October 2017. Inclusion criteria were prospective randomized trials that evaluated patients with TIA or stroke due to intracranial stenosis and treated with different medical therapies and/or endovascular procedures. Primary endpoint was the recurrence of TIA or stroke in the territory of intracranial stenosis, while secondary endpoint was represented by any stroke or vascular death. Multiple treatment comparison meta-analysis based on a Bayesian fixed and random effects Poisson model was performed. Seven trials were included with a total of 1337 patients. At multiple treatment comparison, no significant differences between treatments were observed for both primary (median fixed effect standard OR: 0.40; 95%CI: 0.02–1.07) and secondary endpoints (median random effect standard OR: 1.17; 95%CI: 0.32–1.92). Treatment with aspirin alone ranked with high values both for primary and secondary endpoints (surface under the cumulative ranking curve of 70% and 82%, respectively). In patients with symptomatic intracranial stenosis, no differences between treatments were observed. However, aspirin alone was more effective than stenting in the reduction of TIA or stroke recurrences, with a better safety profile than oral anticoagulants. Ó 2019 Elsevier Ltd. All rights reserved.
1. Introduction Intracranial arterial stenosis is a well-established risk factor for ischemic stroke with different values of risk between races [1–4]. This condition is also associated with a high risk of recurrent stroke
q The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, a worldwide licence to the Publishers and its licensees in perpetuity, in all forms, formats and media (whether known now or created in the future), to i) publish, reproduce, distribute, display and store the Contribution, ii) translate the Contribution into other languages, create adaptations, reprints, include within collections and create summaries, extracts and/or, abstracts of the Contribution, iii) create any other derivative work(s) based on the Contribution, iv) to exploit all subsidiary rights in the Contribution, v) the inclusion of electronic links from the Contribution to third party material where-ever it may be located; and, vi) licence any third party to do any or all of the above. ⇑ Corresponding author at: Department of Neurology & Stroke Unit, Via Napoleona, 60, 22100 Como, Italy. E-mail address: [email protected] (S. Vidale).
https://doi.org/10.1016/j.jocn.2019.01.040 0967-5868/Ó 2019 Elsevier Ltd. All rights reserved.
[5,6]. Indeed, patients with transient ischemic attacks (TIA) or strokes with severe intracranial stenosis (70–99% of the arterial diameter) are at particular high risk for recurrent stroke in the territory of atherosclerotic artery, up to 12.6% at 1 year, despite the best medical management or control of the common vascular risk factors (i.e., blood hypertension, diabetes mellitus, dyslipidemia. . .) [7]. For these reasons, prompt and effective treatments have been applied in the past in two principal directions: a preventive strategy to reduce the risk of atherosclerotic thromboembolic process by the medical approach and a direct action on arterial stenosis by endovascular procedures.
1.1. Rationale The treatment options for the symptomatic intracranial stenosis include antiplatelet and anticoagulant agents, even if the efficacy and safety of these last therapies are debatable and aspirin
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90
appeared to be safer than warfarin in a previous randomized controlled trial [5,8]. Despite the best medical treatment, some patients suffered of a recurrent stroke, failing to respond to the pharmacological therapies [9]. For this reason, in the last decades several types of endovascular devices have been applied and the angioplasty with stenting has been proposed as an alternative treatment to the medical therapy alone. However, previous trials showed different results, without definitive conclusions. In literature, different randomized clinical trials have been published with comparative effects between two therapeutical interventions (i.e., endovascular and antiplatelets) but without a direct comparison with other therapies (i.e., endovascular versus anticoagulants). In other words, head-to-head randomized trials comparing some treatments of interest are lacking and for this reason a new comprehensive analysis might be useful to identify the best therapeutical option.
1.2. Objective Aim of this multiple treatment meta-analysis was to determinate the comparative effectiveness and safety of current different therapeutical approaches to symptomatic intracranial stenosis, considering both medical treatment and endovascular procedures and comparing these therapies directly and indirectly.
2. Methods We conducted this meta-analysis following the guidance coming from the extension statement of the PRISMA for reporting of systematic reviews incorporating network meta-analyses [10].
2.1. Literature search and eligibility criteria We searched the published literature using Medline, EMBASE and the Cochrane Trial Databases up to October 1st, 2017. Keywords of this search were ‘‘intracranial stenosis” or ‘‘intracranial atherosclerosis”, ‘‘stent” or ‘‘endovascular”, ‘‘medical therapy” or different combinations of ‘‘antiplatelet” or ‘‘antithrombotic” or ‘‘anticoagulant”. Specific Boolean syntax of the search is described in the Supplemental Material. We included in the analysis only prospective randomized trials that evaluated patients with TIA or stroke caused by an intracranial stenosis and treated with different medical therapies and/or endovascular procedures. Cohort studies, letters, comments, editorials, case reports, proceeding, personal communication and studies not providing quantitative outcome were excluded.
2.2. Data collection process Data were extracted independently by two reviewers with a consultation of a third reviewer in the case of disagreements. We extracted data concerning the population, the study design and the main outcomes. For our purposes, we consider two main endpoints. Primary endpoint was considered as the occurrence of TIA or ischemic stroke interesting the territory of the stenotic intracranial artery during the follow up. Secondary endpoint was represented by the occurrence of any stroke (ischemic or hemorrhagic) or vascular death. We conducted also additional analyses for stroke or major bleeding or all cause death only, and for the composite of these outcomes. Different risks of bias were reported using the recommendations of the Cochrane Handbook for Systematic Reviews of Intervention.
85
2.3. Patient involvement No patient was directly involved in this study, because we performed a meta-analysis of previous trials. 2.4. Statistical analysis The network meta-analysis was based on a Bayesian random effects Poisson regression model, which preserves randomised treatment comparisons within trials. Using a full Bayesian evidence network, all indirect comparisons are taken into account to arrive at a single, integrated estimate of the effect of all included treatments based on all included studies. The model considers numbers of patients with primary and secondary outcomes, as previously described, accumulating also years to estimate rate ratios. We calculated relative effects both for direct and indirect comparisons. We included in the analysis also a meta-regression to assess the potential effects of some modifiers: age, male gender, active smoking, time between event to randomization, follow-up time and statin use. A graphical representation of the network was assessed using a specific geometry and providing the specification of the nodes by the randomised intervention. Consistency was mainly assessed by the comparison of the conventional network meta-analysis model. Finally, heterogeneity for main studies was assessed calculating Cochrane’s Q and I2 values. We performed also a sensitivity analysis to determine the robustness of the observed outcomes. Doi and funnel plots were representative of the levels of publication bias for groups of randomised trials. One of the advantages of network metaanalysis is that it can provide information about the ranking of all evaluated interventions for the studied outcome. Probabilities are often estimated for a treatment being ranked at a specific place (first, second, etc.,) according to the outcome. We provided rankograms for primary and secondary endpoints as graphs. The cumulative rankograms present the probabilities that a treatment would be among the different best treatments. The surface under the cumulative ranking curve (SUCRA), a simple transformation of the mean rank, was used to provide a hierarchy of the treatments and accounts both for the location and the variance of all relative treatment effects. The larger the SUCRA value, the better the rank of the treatment. We provided SUCRA values for primary and secondary endpoints, as well as for additional analyses. WinBUGS for Microsoft Excel was used for the multiple treatment metaanalysis, integrated with an Excel-based tool software (NetMetaXL) [11]. 3. Results We included in this network meta-analysis 7 randomised clinical trials [8,12–17]. One of these trials was considered only for secondary outcome, because data concerning the recurrent strokes interesting the territory of stenotic artery were not available [14]. Other studies were excluded because without subgroup identification of intracranial atherosclerosis or primary cerebrovascular event [18–21]. The complete history of PRISMA flow diagram was represented in the Fig. 1. The studies recruited a total of 1337 patients. Four trials compared clopidogrel and aspirin treatments with the same therapy plus endovascular procedure. In two trials, aspirin was compared with warfarin (OA), while in one trial dual antiplatelet therapy (aspirin + clopidogrel) was matched to aspirin. Details of single studies were reported in Table 1. A total of 325 subjects were treated with stenting, while 310 patients were recruited in the group of dual antithrombotic therapy (aspirin plus clopidogrel). The largest group was the aspirin treatment with 399 patients, while the smallest group
86
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90
Fig. 2. Network geometry of randomised trials for primary endpoint. The line thickness is representative of the population weight of randomised trials included in the network meta-analysis ASA: aspirin; CLOP: clopidogrel; OA: oral anticoagulant.
3.1. Primary endpoint
Fig. 1. PRISMA flow diagram.
was represented by the anticoagulated one (303 subjects). We observed no significant differences in the size of these treatment groups, neither for age (p: 0.32).
The network geometry for this outcome is represented in Fig. 2. We calculated that 22.8% patients of the stenting group had a TIA or stroke in the territory of intracranial stenosis versus 16% of aspirin-treated patients. The lowest rate of primary events was observed in the warfarin-treated group (11.5%), followed by the aspirin plus clopidogrel group with 12.2%. In the direct comparison analysis, the OA versus aspirin group was the largest one, while antiplatelet monotherapy versus dual therapy was the smallest group. At multiple treatment comparison, we observed no significant differences between groups (median fixed effect standard with OR: 0.40–95%CI: 0.02–1.07). The single ORs of direct and indirect comparisons between treatments varied between 0.26 in favour of OA on stent treatment and 0.78 in favour of OA on aspirin therapy. Direct and indirect relative estimates are presented in Table 2A and the forest plot with different ORs is shown in Fig. 3A. We performed also a sensitivity analysis, excluding 2 stud-
Table 1 Characteristics of RCT included in the network meta-analysis.
Total size Mean age (yrs) Gender (M) Stenting Anticoagulants Statins (I/C) Antihypertensive drugs Hypoglycemic drugs Time from qualifying event to randomization (days; I/ C) Mean follow-up time (months) *
SAMMPRIS [13]
VISSIT [12]
Mohammadian [14]
VAST* [15]
CLAIR [16]
WASID [8]
Martì-Fabregas [17]
451 60 272 U
111 62 73 U
63 69 38 U
115 65 85 U
100 56 76
569 63 350
28 67 19
202/195 200/203 105/103 7/7
29/32 43/49 25/20 12/15
4/6 24/20 24/17 /
49/55 40/40 / 27/22
21/16 34/34 17/14 2.5/3.2
U 184/163 189/179 / 18/16
U 8/4 12/10 5/5 25/23
32.4
12
15.2
36
0.25
20
23
Referred to the entire sample; I: interventional group; C: control group.
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90 Table 2 A) Direct and indirect estimates (with low and high 95%CI values) for primary endpoint; B) Direct and indirect estimates (with low and high 95%CI values) for secondary endpoint. A) OA 0.78 (0.12–6.01) – –
– ASA 0.55 (0.08–3.42) 0.61 (0.15–3.05)
0.42 (0.04–5.83) – ASA + CLOP –
0.26 (0.02–3.37) – 0.26 (0.02–3.37) Stent
– ASA + CLOP – 0.93 (0.12–7.21)
– 1.17 (0.00–78.10) OA –
– – 0.83 (0.01–74.12) Stent
B) ASA 0.25 (0.03–11.01) 0.30 (0.01–3.08) 0.23 (0.01–6.03)
Direct estimates are represented in the lower cells, while indirect estimates in the higher. Comparisons are represented with the first treatment on the left bottom. ASA: aspirin; CLOP: clopidogrel; OA: oral anticoagulant.
ies with higher level of bias risk [14,17]. We did not observe a significant difference of OR when compared with the main analysis (OR: 0.41; 95%CI: 0.08–1.04). The rankogram showed a higher probability for OA in the first rank with a following reduction (from 64% to 7%), while stent treatment presented an inverse trend (from 2% to 76%). The dual antiplatelet treatment (aspirin plus clopidogrel) has the higher value at rank 3 (63%). Different values of treatment ranking are presented in Supplemental Material. The larger SUCRA value was observed for OA (SUCRA: 80%), followed by aspirin treatment (SUCRA: 70%), dual antiplatelet therapy (SUCRA: 39%) and finally stenting therapy (SUCRA: 11%). The total rankogram is showed in Supplemental Material. Inconsistency results varied between OR: 0.9895 for the comparison of aspirin plus clopidogrel vs aspirin alone and OR: 0.5919 for the indirect comparison of OA vs stent. The graph representing the ratio between consistency and inconsistency models is showed in Supplemental Material. Considering the heterogeneity of the trials included in this network meta-analysis, the Cochrane’s Q was 0.58 with an I2 value of 0% (p: 0.80). In the meta-regression analysis, we observed a significant influence of statin use only (S.E.: 0.028; p: 0.011). Results for each considered variables are presented in Supplemental Material. 3.2. Secondary endpoint We did not observe significant differences in the direct and indirect estimates between treatments. ORs varied significantly with lower value observed in the comparison between aspirin and stent (OR: 0.23; 95%CI: 0.01 – 6.03) and the higher value in the indirect estimate of dual antiplatelet therapy versus OA (OR: 1.17; 95%CI: 0.00–78.10). Other direct estimates, as well as the indirect estimates, are represented in the Table 2B and the forest plot was shown in Fig. 3B. The median Random effect standard was 1.41 (95%CI: 0.32–1.92). The rankogram showed a relative homogeneous risk probability between treatments with the exception for aspirin (Supplemental Material). This treatment had the highest value in rank 1 (63%) and a concomitant larger SUCRA value (SUCRA: 82%). Dual antiplatelet therapy, OA and stent treatment showed respectively a SUCRA of 41%, 41% and 37%. The Cochrane’s Q value was 5.86 with an I2: 66% (p: 0.05). 3.3. Additional analyses Considering only stroke as main outcome, we did not observe significant differences between the direct and indirect comparisons of different treatments. The higher difference was detected for the double antiplatelet therapy versus stenting (OR: 0.335; 95%CI: 0.020–1.662), while OA and ASA showed a nearly similarity
87
in their comparison (OR: 0.806; 95%CI: 0.058–10.530). The double antiplatelet treatment showed the best cumulative ranking with a SUCRA value of 72%, followed by OA (SUCRA: 58%), ASA (SUCRA: 48%) and finally by stenting (SUCRA: 22%). In the analysis for major bleeding, treatment with only ASA showed better OR when compared with OA or stenting with values of 0.236 and 0.244, respectively. Results of each comparison of treatments are presented in the Supplemental Material. Antiplatelet monotherapy showed the highest SUCRA value (84%), while OA had the lowest one (23%). Considering all cause death, the comparison between OA and stent showed an OR of 0.257 (95%CI: 0.021–3.248). A similar result was observed in the comparison between ASA and stent (OR: 0.324; 95%CI: 0.06–1.598). Other results were presented in Supplemental Material. Cumulative rankogram showed a best positioning of OA (SUCRA: 81%), followed by ASA (SUCRA: 70%). Stent ranked as last with SUCRA value of 11%. When we considered the composite of the three previous outcomes, we observed the superiority of ASA treatment versus stent (OR: 0.584; 95%CI: 0.046–13.071) or OA (OR: 0.683; 95%CI: 0.053– 8.711). The ranking of different treatments showed ASA with the higher SUCRA value (62%) and followed by the double antiplatelet therapy (SUCRA: 51%).
4. Discussion In this network meta-analysis concerning the treatment of intracranial stenosis in patients with stroke or TIA, we observed a higher risk of stroke recurrences in the territory of stenotic intracranial artery after stenting than medical treatment, even if this difference was not statistically significant. The direct comparison between aspirin and dual antiplatelet therapy did not show differences for efficacy of the treatments, as well as between antithrombotics and anticoagulants. Despite the fact that the head-to-head comparisons between treatments were not statistically significant, aspirin alone showed the best SUCRA value for all composite outcomes with 62%, followed by the double antiplatelet therapy (51%). For this reason, aspirin or aspirin plus clopidogrel appear to be the more effective treatments to reduce the risk of a recurrent stroke in the territory of intracranial stenosis, balanced with potential side-effects. This observation is in line with the current indications coming from the international guidelines [22]. In this analysis, we calculated a total NNT of 10 in the comparison of stenting versus dual antiplatelet therapy, while this value reduced to 3 in the comparison of angioplasty versus aspirin alone. The highest value of NNT was calculated comparing dual antiplatelet to aspirin alone (NNT: 33). Concomitant therapies (i.e. statins, antihypertensive drugs. . .) did not significantly differ between groups of the included trials. In particular, we calculated a total of 284 patients treated with statins in the endovascular group compared with 288 subjects in the medical group (Table 1). However, in the meta-regression analysis we observed that the statin use could have a potential effect on this primary endpoint. Considering the secondary endpoint, we did not observe significant difference between treatments, even if we observed a difference between aspirin and stenting groups of about 14.7%. A very recent meta-analysis showed similar results for the stenting treatment, but the other medical treatments were not included in the pooled analysis [23]. Despite the hope that endovascular treatment would be more beneficial in the management of intracranial stenosis than medical therapy, increasing the blood flow similarly to extracranial carotid artery, the effects of this procedure are not the same. Indeed, while the revascularization of extracranial arteries with surgery or stenting is more effective than best medical treatment, in this network meta-analysis similar evidences did not occur. A
88
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90
Fig. 3. Forest plots for primary (A) and secondary (B) endpoints.
possible explanation of this difference could be the site and size of the arterial stenosis. At the same time, for trials comparing stenting with antiplatelet therapy the variations in the type of devices (bare-metal stent or balloon-expandable and self-expanding stents), the differences of protocols and skill levels of professional performing endovascular procedures could influence the outcomes. The treatment ranking probabilities for this network meta-analysis showed a superiority of OA for the primary endpoint, but we observed a total better ranking for aspirin (SUCRA: 70% and 82% for primary and secondary outcomes, respectively). When we considered also additional analyses, we observed best ranking for double antiplatelet treatment for only stroke, but the therapy with ASA alone provided the best benefit/risk ratio in the composite outcomes (stroke + major bleeding + all cause death). Finally, for all trials the heterogeneity is significant only for sec-
ondary endpoint, without bias of publication for both primary and secondary endpoints, as represented in the funnel plot (Supplemental Material). Most of the trials presented low risks of bias for all items, as reported in the Fig. 4, with the exception of two studies [14,17]. For this reason we conducted a sensitivity analysis, excluding these trials, but the results did not differ significantly from the main analysis. We observed differences between the time from qualifying events to randomization. However, most of treated patients with stenting were randomised within two weeks, without differences with medical groups and the metaregression analysis did not show significant influence of this variable on primary endpoint. The main limitation of this network meta-analysis is the inability to perform subgroup analyses due to limited data in the included studies. In particular, some individual risk factors (i.e.
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90
89
In conclusion, in our multiple treatment meta-analysis, the antithrombotic therapy seems to be more effective and safer than stenting and in particular aspirin alone with better results for all composite outcomes. Oral anticoagulants could represent an effective alternative but with a lower profile in secondary endpoint and in some additional analyses. Considering the heterogeneity of study population and treatments, more trials with bettercontrolled randomization are needed. SV, EA, AC, and DC designed the study. SV and AC performed the literature search, data analysis, and statistical modelling. They contributed equally to the study. SV drafted the manuscript. EA, AC, GG and DC performed a critical review of the manuscript. SV is the guarantor. Data sharing statement N/A. Declarations of interest None. Funding source None. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi.org/10.1016/j.jocn.2019.01.040. References
Fig. 4. Risk of bias summary for each trial.
atrial fibrillation, diabetes mellitus, hemodynamic parameters. . .) and characteristics (i.e. age, gender. . .) could help to better investigate the profile of patient that best benefit from some treatments than others. In particular, it could be of great interest the comparison of treatment effect between race, because in Asian people an intracranial arterial stenosis could increase the stroke risk up to 30–50%, a value higher than in people living in Western Countries [1–4]. Another crucial point was the differentiation between anterior and posterior circulation. However, the results coming from sub-analysis in the SAMMPRIS study did not show significant differences concerning the primary outcome between anterior and posterior circulation. Finally, no data are available reporting tandem stenosis which are pathological conditions with an incidence between 2 and 13% for anterior circulation [24]. This review was not registered in PROSPERO and this could be another limitation. Strengths are represented by the clear definition of primary and secondary endpoints and the multiple comparison of different treatment with the concomitant reduction of bias introducing the use of random effect that attenuates the heterogeneity between study populations and methodology. Even if some other randomised trials are ongoing and final results will be expected in the next years [25], current literature and the evidence-based medicine are lacking of specific guidelines for the prevention of TIA and strokes in symptomatic intracranial atherosclerotic disease. However, a previous statement defined the main items and purposes to begin a new guideline project [26].
[1] Gorelick PB, Wong KS, Bae HJ, Pandey DK. Large artery incranial occlusive disease: a large worldwide burden but a relatively neglected frontier. Stroke 2008;39:2369–79. [2] Arenillas JF. Intracranial atherosclerosis: current concepts. Stroke 2011;42 (Suppl):S20–3. [3] Holmstedt CA, Turan TN, Chimowitz MI. Atherosclerotic intracranial arterial stenosis: risk factors, diagnosis, and treatment. Lancet Neurol 2013;12:1106–14. [4] Wong LK. Global burden of intracranial atherosclerosis. Int J Stroke 2006;1:158–9. [5] Kasner SE, Chimowitz MI, Lynn MJ, et al. Predictors of ischemic stroke in the territory of a symptomatic intracranial arterial stenosis. Circulation 2006;113:555–63. [6] Zaidat OO, Klucznik R, Alexander MJ, et al. The NIH registry on use of the Wingspan stent for symptomatic 70–99% intracranial arterial stenosis. Neurology 2008;70:1518–24. [7] Chaturvedi S, Turan TN, Lynn MJ, et al. Do patient characteristics explain the differences in outcome between medically treated patients in SAMMPRIS and WASID? Stroke 2015;46:2565–7. [8] Chimowitz MI, Lynn MJ, Howlett-Smith H, et al. Comparison of warfarin and aspirin for symptomatic intracranial arterial stenosis. N Engl J Med 2005;352:1305–16. [9] Thijs VN, Albers GW. Symptomatic intracranial atherosclerosis: outcome of patients who fail antithrombotic therapy. Neurology 2000;55:490–7. [10] Hutton B, Salanti G, Caldwell DB, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann Intern Med 2015;162:777–84. [11] Brown S, Hutton B, Clifford T, et al. A Microsoft-Excel-based tool for running and critically appraising network meta-analyses – an overview and application of NetMetaXL. Syst Rev 2014;3:110. [12] Zaidat OO, Fitzsimmons BF, Woodward BK, et al. Effect of balloon-expandable intracranial stent vs medical therapy on risk of stroke in patients with symptomatic intracranial stenosis. JAMA 2015;313:1240–8. [13] Derdeyn CP, Chimowitz MI, Lynn MJ, et al. Aggressive medical treatment with or without steniting in high-risk patients with intracranial artery stenosis (SAMMPRIS): the final results of randomised trial. Lancet 2014;383:333–41. [14] Mohammadian R, Pashapour A, Sharifipour E, et al. A comparison of stent implant versus medical treatment for severe symptomatic intracranial stenosis: a controlled clinical trial. Cerebrovasc Dis Extra 2012;2:108–20.
90
S. Vidale et al. / Journal of Clinical Neuroscience 63 (2019) 84–90
[15] Compter A, van der Worp HB, Schonewille WJ, et al. Stenting versus medical treatment in patients with symptomatic vertebral artery stenosis: a randomised open-label phase 2 trial. Lancet Neurol 2015;14:606–14. [16] Wong KS, Chen C, Fu J, et al. Clopidogrel plus aspirin versus aspirin alone for reducing embolisation in patients with acute symptomatic cerebral or carotid artery stenosis (CLAIR study): a randomised, open-label, blinded-endpoint trial. Lancet Neurol 2010;9:489–97. [17] Martì-Fabregas J, Cocho D, Martì-Vilalta JL, et al. Aspirin or anticoagulants in stenosis of the middle cerebral artery: a randomized trial. Cerebrovasc Dis 2006;22:162–9. [18] Wong KS, Chen C, Ng PW, et al. Low-molecular-weight heparin compared with aspirin for the treatment of acute ischaemic stroke in Asian patients with large artery occlusive disease: a randomised study. Lancet Neurol 2007;6:407–13. [19] Wang Y, Wang Y, Zhao X, et al. Clopidogrel with aspirin in acute minor stroke or transient ischemic attack. N Engl J Med 2013;369:11–9. [20] Bhatt DL, Fox KAA, Hacke W, et al. Clopidogrel and aspirin versus aspirin alone fot he prevention of atherothrombotic events. N Engl J Med 2006;354:1706–17. [21] Diener HC, Bogousslavsky J, Brass LM, et al. Aspirin and clopidogrel compared with clopidogrel alone after recent ischemic stroke or transient ischemic attack in high-risk patients (MATCH): randomised, double-blind, placebocontrolled trial. Lancet 2004;364:331–7.
[22] Kernan WN, Ovbiagele B, Black HR, Bravata DM, Chimowitz MI, Ezekowitz MD, Schwamm LH, Wilson JA, et al. on behalf of the American heart association stroke council, council on cardiovascular and stroke nursing, council on clinical cardiology, and council on peripheral vascular disease. Guidelines for the prevention of stroke in patients with stroke and transient ischemic attack: a guideline for healthcare professionals from the American Heart Association/ American Stroke Association. Stroke 2014;45. [23] Zhang Q, Dong K, Song H. Comparison of stent versus medical therapy for symptomatic patients with intracranial atherosclerotic stenosis: a metaanalysis. J Neurol Sci 2017;372:272–8. [24] Rouleau PA, Huston 3rd J, Gilbertson J, et al. Carotid artery tandem lesions: frequency of angiographic detection and consequences for endarterectomy. AM J Neuroradiol 1999;20:621–5. [25] Cui XP, Lin M, Mu JS, et al. Angioplasty and stenting for patents with symptomatic intracranial atherosclerosis: study protocol of a randomised controlled trial. BMJ Open 2016;6:e012175. [26] Turan TN, Gronseth GS, Chimowitz M, et al. Protocol for proposed guideline project: prevention of stroke in patients with symptomatic large vessel intracranial atherosclerotic disease. 2015. https://www.aan.com/ uploadedFiles/Website_Library_Assets/Documents/2.Clinical_Guidelines/3. Browse_By_Status/4.Guidelines_Under_Development/Atherocslerosis-protocolfor-public-comment-final-7-9-2015.pdf.