Article Text

Download PDFPDF

Original research
Outcome of intracranial flow diversion according to the antiplatelet regimen used: a systematic review and meta-analysis
  1. Anna Podlasek1,2,
  2. Abdul Aziz Al Sultan3,
  3. Zarina Assis4,
  4. Nima Kashani4,
  5. Mayank Goyal4,5,
  6. Mohammed A Almekhlafi3,5
  1. 1 Stroke, Southend University Hospital NHS Foundation Trust, Westcliff-on-Sea, UK
  2. 2 Neuroscience and Vascular Simulation, Anglia Ruskin University, Chelmsford, UK
  3. 3 Clinical Neuroscience, University of Calgary Cumming School of Medicine, Calgary, Alberta, Canada
  4. 4 Radiology, University of Calgary Cumming School of Medicine, Calgary, Alberta, Canada
  5. 5 Radiology, Hotchkiss Brain Institute, Calgary, Alberta, Canada
  1. Correspondence to Dr Mohammed A Almekhlafi, Clinical Neurosciences, University of Calgary Cumming School of Medicine, Calgary, AB T2N 4N1, Canada; mohammed.almekhlafi1{at}


Background Thromboembolic complications are not uncommon in patients undergoing neurointerventional procedures. The use of flow diverting stents is associated with higher risks of these complications despite current dual antiplatelet regimens.

Objective To explore contemporary evidence on the safety of emerging dual antiplatelet regimens in flow diverting stenting procedures.

Methods We performed a systematic review and meta-analysis to identify relevant articles in electronic databases, and relevant references. Studies reporting the complications and mortality of flow diverting stenting procedures using acetyl salicylic acid (ASA) + ticagrelor or ASA + prasugrel compared with ASA + clopidogrel were included.

Results Of 452 potentially relevant studies, we identified 49 studies (2526 patients) which reported the safety of ticagrelor or prasugrel for pooled analysis, and five studies (1005 patients) for meta-analysis. The pooled overall mortality in all studies was 2.14%, ischemic complications 6.89%, and hemorrhagic complications 3.68%. The use of ticagrelor or prasugrel was associated with a lower risk of mortality compared with clopidogrel (RR=4.57, 95% CI 1.23 to 16.99; p=0.02). Considering ischemic events, ASA + clopidogrel was as safe as ASA + prasugrel (RR=0.55, 95% CI 0.11 to 2.74; p=0.47) and ASA + ticagrelor (RR=0.74, 95% CI 0.32 to 1.74; p=0.49). ASA +ticagrelor was not associated with a higher risk of hemorrhagic complications (RR=0.92, 95% CI 0.27 to 3.16; p=0.89).

Conclusions Evidence suggests that dual antiplatelet regimens including ticagrelor or prasugrel are safe for patients undergoing flow diversion procedures. Regimens using ticagrelor were associated with better survival than those using clopidogrel in the included studies.

  • flow diverter

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Endovascular therapy using flow diversion has become the preferred approach for treating complex intracranial aneurysms.1 2 The approach is considered physiologic because it aims to divert blood flow away from the aneurysm along the anatomical course of the vessel. Consequently, the shear stress on the aneurysm wall is reduced3 and intra-aneurysm flow stasis and thrombosis occur. This phenomenon is affected by the amount of metal surface area coverage provided by the stent. The pore density of flow diverters, rather than porosity, seems to be a critical factor for modulating device efficacy.3 Moreover, flow diverting stents provide a base for the development of endothelial and neointimal tissue across the aneurysm neck that is proportional to the amount of metal surface area coverage and the stent material.4–9 This high metal surface area makes these stents more susceptible to thrombus formation compared with regular stents.

The risk of neurological complications following endovascular treatment of unruptured aneurysms is estimated to be 4.1–7.4%.10 Thromboembolic complications are the most common cause of morbidity after these procedures, especially when flow diverting stents (FDS) are used.11 12 The 6-month incidence of in-stent thrombosis has ranged between 3.5% and 16%.13 14 In the meta-analysis by Bond et al, the incidence of new MR diffusion lesions following FDS placement was 67% (95% CI 46% to 85%).15 This indicates that current prophylactic antiplatelet therapies for FDS procedures may not be sufficient. Currently, prophylactic dual antiplatelet therapy (DAPT) regimens primarily include acetyl salicylic acid (ASA) and clopidogrel, which are started 3–14 days before the procedure; clopidogrel is stopped 3–12 months postprocedurally, whereas ASA is continued indefinitely.16 However, other antiplatelet agents, such as prasugrel and ticagrelor, are becoming increasingly used in neurointerventional procedures, supported by evidence from cardiac literature.17 This experience cannot be fully translated to neurointerventional procedures owing to differences in stents designs as well as vessel size (both parent vessel and side branches) and consequently, changes in the hemodynamic environment.18 Despite encouraging reports of better efficacy of these agents compared with clopidogrel, ASA and clopidogrel remain the first-line choice for DAPT—based on a recent neurointerventional practice patterns survey.19 To date, there is no evidence on the safety and efficacy of the newer antiplatelet regimens with flow diverting stents. We conducted a systematic review of current literature to summarize evidence on the safety of alternative antiplatelet regimens in FDS procedures.


Search strategy, study selection, and eligibility criteria

The study was performed in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines.20 We systematically searched electronic databases including PubMed/Medline, Google Scholar, and Cochrane up to February 2019. The following keywords were used in combination or individually using the Boolean operators ‘OR’ and ‘AND’: ‘intracranial’, ‘cerebral’, “carotid’, ‘basilar’, ‘Neuroendovascular’, ‘Neurointerventional’, ‘aneurysm’, ‘flow diverting’, ‘flow diversion’, ‘ticagrelor’, ‘prasugrel’ NOT ‘coronary’, ‘myocardial’. The articles were selected in two stages. First, the titles and abstracts identified by the above searches were screened for relevant studies. Second, the full texts of these shortlisted articles were downloaded and assessed for eligibility based on the inclusion criteria. The reference lists of publications were then hand-searched for additional relevant studies. This process was carried out by three assessors independently (AAAS, AP, MAA). Any differences were resolved by consensus.

Studies reporting on patients with ruptured and unruptured aneurysms who underwent FDS procedures were included. Exclusion criteria were studies published in abstract form only, review articles, and meta-analyses, guidelines, technical notes, studies in languages other than English, in vitro/animal studies, unspecified antiplatelet regimen or duration, lack of uniform antiplatelet regimen protocol, studies of <10 patients, and lack of follow-up data on complications. In cases of overlapping patient population, only the series with the largest number of patients or the most detailed data was included. The Risk Of Bias In Non-randomized Studies - of Interventions (ROBINS-I)21 tool was used to assess the individual risk of bias of each study.

Data extraction

The main exposure variables were antiplatelet regimens (clopidogrel vs ticagrelor or prasugrel). One assessor (AP) extracted data on demographics and anatomical variables, antiplatelet regimen, hemorrhagic and ischemic complications, mortality, and platelet function testing.

Outcome measures

Study characteristics and extracted variables were summarized using standard descriptive statistics. Continuous variables were expressed as means and SD or medians and ranges, and categorical variables were expressed as frequencies or percentages. The primary outcome was the incidence of postprocedural thromboembolic and hemorrhagic complications. Secondary outcomes were mortality and the use of platelet inhibition testing.

Statistical analysis

We calculated the overall estimate of complications across eligible studies based on the proportion of participants who developed complications and the total number of patients undergoing a flow diverting stent procedure. This was followed by a random-effects meta-analysis model of proportions to estimate the pooled rates of ischemic and hemorrhagic complications and mortality with each antiplatelet regimen. Tests of heterogeneity were conducted with the Q statistic that is distributed as a Χvariate (assumption of homogeneity of effect sizes). The extent of between-study heterogeneity was assessed with the I2 statistic, with interpretation as follows: I2 <25% indicates no heterogeneity, I2<50% indicates low heterogeneity, I2<75% indicates moderate heterogeneity, and I2≥75% indicates high heterogeneity.22 23 Publication bias was assessed using funnel plots.24 25

P values were two-tailed, with values <0.05 considered statistically significant. All analyses were implemented in Statistica 13.1 software.


Literature search results

The initial search yielded 452 articles, which were screened by title and abstract. Of these, 97 articles were eligible and underwent full-text review (figure 1). However, the DAPT regimen was not always well-described and only 54 studies were finally included. Five studies11 12 26–28 (total of 1005 patients) with a control comparison group were included in the meta-analysis while the remainder (49 studies) were used to generate pooled estimates (figure 1). The study characteristics are summarized in table 1.

Table 1

PICO model for studies with ticagrelor and prasugrel included in the meta-analysis

Figure 1

PRISMA flow chart for study selection. DAPT, dual antiplatelet therapy.

Characteristics of the controlled studies included

The five studies included for meta-analysis described 1005 patients, with both ruptured12 26 27 and unruptured28 29 aneurysms. A total of 832 (82.8%) patients received clopidogrel + ASA, while 137 (13.6%) patients received ticagrelor + ASA, and 36 (3.6%) patients received prasugrel + ASA.

Procedural outcomes

The primary outcome of ischemic or hemorrhagic complications did not differ between the three antiplatelet regimens. Prasugrel-based DAPT was associated with a lower risk of ischemic complications than the clopidogrel regimen but the difference did not reach significance (RR=0.55, 95% CI 0.11 to 2.74; p=0.47) (figure 2). Ticagrelor-based DAPT also appeared to be associated with a lower risk of ischemic complications that was not significant (RR=0.74, 95% CI 0.32 to 1.74; p=0.49) (figure 2) and the risk of hemorrhagic complications was similar to that of clopidogrel-based regimen (RR=0.92, 95% CI 0.27 to 3.16; p=0.89) (figure 3). However, patients who received a clopidogrel-based regimen had higher mortality risk than the alternative regimens (RR=4.57, 95% CI 1.23 to 16.99; p=0.023). This risk difference was driven by better survival in the ticagrelor-based regimens (RR=5.38, 95% CI 1.10 to 26.24; p=0.038) as the mortality risk with clopidogrel was not significantly different from that of prasugrel (RR=3.21, 95% CI 0.31 to 33.35; p=0.33) (figure 4) As this difference in mortality appeared to be driven by one study,26 we conducted a sensitivity analysis excluding the study by Atallah et al.26 The difference in mortality became insignificant for both overall regimens versus that for the clopidogrel-based one (RR=3.24, 95% CI 0.57 to 18.44; p=0.18), and for the ticagrelor-based regimen (RR=2.93, 95% CI 0.42 to 20.50; p=0.28) (online supplementary figure 1).

Figure 2

Ischemic complications. Forest plot of the risk ratio and 95% CIs for procedure-induced complications in neurovascular intervention between different DAPT regimens. The size of the middle square corresponding to each study is proportional to the sample size. The horizontal line shows the corresponding 95% CI of the risk ratio (UL, upper limit; LL, lower limit). The combined estimate is based on a random-effects model.

Figure 3

Hemorrhagic complications. Forest plot of the risk ratio and 95% CIs for procedure-induced complications in neurovascular intervention between different DAPT regimens. The size of the middle square corresponding to each study is proportional to the sample size. The horizontal line shows the corresponding 95% CI of the risk ratio (UL, upper limit; LL, lower limit). The combined estimate is based on a random-effects model. DAPT, dual antiplatelet therapy.

Figure 4

Mortality. Forest plot of the risk ratio and 95% CIs for procedure-induced complications in neurovascular intervention between different DAPT regimens. The size of the middle square corresponding to each study is proportional to the sample size. The horizontal line shows the corresponding 95% CI of the risk ratio (UL, upper limit; LL, lower limit). The combined estimate is based on a random-effects model. DAPT, dual antiplatelet therapy.

Antiplatelet testing was performed for all patients only in three studies; all used the VerifyNow system.11 12 26 The study by Atallah et al switched clopidogrel hyporesponders (defined as inhibition <30%) to ticagrelor, and to prasugrel if the response was again not satisfactory.26 The study by Moore et al defined clopidogrel hyporesponders as those with platelet inhibition <50% or P2Y12 reaction unit (PRU)>208, and those patients were switched to ticagrelor.30 The study by Adeeb et al used three different methods to obtain platelet inhibition rate and either switched hyporesponders to ticagrelor or gave a loading dose of clopidogrel before the procedure.12 Two other studies included selective antiplatelet testing at the discretion of the institution involved.27 28

Single-arm cohorts

We identified 49 single-arm studies with 2526 patients, which reported on flow diversion procedure complications and stated the drugs used for DAPT (table 2). We classified these studies into three groups: C - DAPT with ASA + clopidogrel (20 studies with 732 patients), CPT – DAPT with clopidogrel, prasugrel or ticagrelor but did not report the individual outcome of each DAPT regimen (15 studies with 988 patients), O – DAPT with clopidogrel and/or other agents (14 studies with 806 patients) like abciximab (five studies), tirofiban (three studies), ticlopidine (five studies), eptifibatide (one study). The overall mortality in the 49 studies was 54/2526 (2.14%), ischemic complications were 174/2526 (6.89%), and hemorrhagic complications were 93/2526 (3.68%). There was a significant difference for mortality (p=0.01) and hemorrhagic complications (p<0.001) between different antiplatelet therapy regimens (DAPT based on clopidogrel, DAPT with clopidogrel, ticagrelor or prasugrel, DAPT with other agents) reported in the literature. No difference between ischemic complications was present (0.89) (table 3).

Table 2

Literature reports on flow diversion procedures complications.

Table 3

Incidence of deaths, ischemic and hemorrhagic complications in pooled analysis of 49 studies and meta-analysis

The antiplatelet testing was performed for all patients (n=1473) in 24 studies and for some patients in eight studies (n=312). Among those 32 studies, two studies reported an increased incidence of hemorrhagic complications if PRU<60 and an increased incidence of thromboembolic complications if PRU>240.31 32 One study found that all patients were among the expected antiplatelet inhibition rate.33 The remaining studies did not report complications based on antiplatelet testing results, but 20 studies reported the action taken after abnormal results: two studies reported postponing procedures,34 35 three reported modification of the loading dose,36–38 eight reported switching to prasugrel/ticagrelor regimens,14 39–45 and seven studies reported switching to an alternative regimen.46–52 Another eight studies (335 patients) did not conduct antiplatelet testing at all and nine studies with 406 patients did not report the data on its usage (table 2).


In this meta-analysis, antiplatelet therapy regimens based on ticagrelor or prasugrel appear to be safe and effective alternatives to clopidogrel-based regimens.

Clopidogrel resistance or non-responsiveness is an emerging concern that is estimated to affect 4–50% of patients.10 19 A meta-analysis conducted by Shim et al shows that in studies of flow diverters, thromboembolic events occurred more frequently in hyporesponders than in responders (RR=3.09, 95% CI 1.81 to 5.27).53 However, the benefit of routine testing for antiplatelet resistance before neurointerventional procedures is unproven and P2Y12 is an often-imprecise measure.54 This stems from the great variations in the testing technologies and vague definition of cut-off values. This practice will probably remain controversial until the level of evidence improves through more rigorous testing and standardized reporting.16 Moreover, the GRAVITAS study randomized 2214 patients after coronary intervention with high reactivity to clopidogrel during treatment (PRU >230) and showed no differences in 6-month mortality between a high (600+150 mg) and a low (75 mg) dose of clopidogrel.55 Our meta-analysis suggests that the mortality rate might be higher in clopidogrel groups than in those receiving ticagrelor or prasugrel. However, this needs to be interpreted with caution given the small sample size of some of the studies that influenced this mortality difference.

No published studies have assessed the efficacy of either prasugrel or ticagrelor in neurovascular procedures. However, there is ample evidence on this topic from cardiology literature. The PLATO trial showed better efficacy of ticagrelor versus clopidogrel in patients with coronary artery disease.56 The TRITON-TIMI trial showed significantly reduced rates of ischemic events with prasugrel versus clopidogrel, but with an increased risk of major bleeding, including fatal bleeding (for prasugrel vs clopidogrel: life-threatening bleeding 1.4% vs 0.9%; p=0.01, non-fatal bleeding 1.1% vs 0.9%; HR=1.25; p=0.23, and fatal bleeding 0.4% vs 0.1%; p=0.002).57 Our analysis suggests that ticagrelor is a safe and effective agent for the prevention of thromboembolic complications following FDS compared with clopidogrel. However, ticagrelor remains significantly more expensive than clopidogrel and is currently reserved for patients who are intolerant to clopidogrel or for those who develop ischemic events while compliant with clopidogrel.11 Another option for those patients is to be switched to a higher ASA dose (>150 mg) or to use DAPT for at least 6 months as suggested by a recent meta-analysis.58 Our meta-analysis had scarce data on hemorrhagic complications. Two studies11 12 showed insignificant increase in intracerebral hemorrhage among patients receiving ticagrelor.

Administration of intra-arterial abciximab is an effective and safe rescue strategy for the management of acute intraprocedural thromboembolic complications during a flow diverting procedure. A dosing strategy of (1) 5 mg increments or (2) 0.125 mg/kg IA bolus (half cardiac dosing) can provide high rates of recanalization with low rates of hemorrhagic complications and long-term morbidity.59 A protocol of anticoagulation with tirofiban during flow diversion showed an encouraging safety profile.60 This protocol provides a reasonable alternative to pretreatment with ASA and clopidogrel and is useful in patients with ruptured aneurysms or when the use of a stent is not expected.61

Limitations and further directions

Our study has potential limitations that need to be considered. The results are based mainly on retrospective, descriptive studies, which may hold the risk of both selection and publication bias. The techniques and devices used were different. Prevention of thromboembolic events is complicated by the variable response to antiplatelet therapy, unpredictable patient compliance with medication, potential drug interactions, and the presence of comorbidities.18 Moreover, the ASA dosage and the length of the follow-up period differ between studies. Finally, the baseline characteristics and rupture status varied among the included studies.

The limited scientific evidence on the alternatives to clopidogrel antiplatelet agents could be attributed to the relatively short time which has elapsed since these agents started to be used.19 Moreover, only in 2014 did we get confirmation of the evidence showing that DAPT is better than ASA alone in preventing thromboembolic complications in vascular diseases.60 Thus, ASA and clopidogrel remain the first-line therapy.19 Based on the literature and results of our study, future clinical practice might tailor antiplatelet therapy towards an alternative to clopidogrel regimens according to antiplatelet response testing.


Ticagrelor- and prasugrel-based antiplatelet regimes appear to be a safe alternative to clopidogrel in flow diverting procedures. Prospective, standardized randomized clinical trials are needed to identify the most effective and safest antiplatelet regimen in patients undergoing neuroendovascular procedures with flow diverting stents.


This study is a systematic review and meta-analysis and no human participant procedure was involved. Informed consent and ethical approval may not be essential for this study.



  • Contributors AP, MAA, AAAS took part in planning, conduct and reporting, MG, ZA, NK took part in reporting.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Patient consent for publication Not required.