Article Text

Download PDFPDF

Original research
Tailoring platelet inhibition according to multiple electrode aggregometry decreases the rate of thrombotic complications after intracranial flow-diverting stent implantation
  1. Ismail Oran1,
  2. Celal Cinar1,
  3. Halil Bozkaya1,
  4. Mehmet Korkmaz2
  1. 1Department of Radiology, Ege University Medical School, Izmir, Turkey
  2. 2Department of Radiology, Dumlupinar University Medical School, Kutahya, Turkey
  1. Correspondence to Professor Ismail Oran, Department of Radiology, Ege University Hospital, Izmir 35100, Turkey; ismailoran{at}


Background and purpose Pretreatment with dual antiaggregant drugs is accepted as a standard step in intracranial stent implantation. The aim of this study was to determine whether tailored antiaggregant medication based on platelet reactivity testing with multiple electrode aggregometry (Multiplate) yields superior outcomes after intracranial flow-diverting stent (FDS) implantation compared with standard clopidogrel treatment.

Methods We retrospectively analyzed the following data from 100 consecutive patients: endovascular procedure characteristics, antiaggregant medications, procedural variables, and perioperative complications after FDS implantation for intracranial aneurysm. Patients were divided into two groups: uniform treatment with clopidogrel (untailored, early phase) and tailored treatment based on the results of aggregometry (late phase). Statistical comparisons included the Fisher exact test to compare categorical variables between the standard and aggregometry groups and the Mann–Whitney U test to compare ADP test values within the aggregometry group between groups receiving tailored or untailored treatment.

Results In the aggregometry group (68 patients, 71 procedures) there were 17 (25%) clopidogrel-resistant patients, according to a cut-off value of 468 area under the aggregation curve; 12 underwent FDS implantation under tailored antiaggregant medication. In the standard treatment group (32 patients, 33 procedures) there were 3 (9.1%) spontaneous thrombotic events and 1 (3.3%) technical hemorrhagic complication. In the aggregometry group there were 2 (2.8%) spontaneous hemorrhagic events and 1 (1.4%) technical ischemic complication. In the aggregometry group, thrombotic complications and morbidity were lower than in the standard (no test) group (p<0.03).

Conclusions Tailoring platelet reactivity according to multiple electrode aggregometry decreases the rate of thrombotic complications after intracranial FDS implantation.

  • Aneurysm
  • Complication
  • Drug
  • Flow Diverter
  • Platelets

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.

Flow-diverting stents (FDS) offer a new alternative for the treatment of intracranial aneurysm morphologies previously considered difficult to treat, including giant, wide-necked, fusiform, and blister. Since FDS contain about fivefold more metal than standard intracranial stents, they are traditionally implanted under dual antiplatelet therapy with aspirin and clopidogrel. About one-quarter of the population (range 4–44%) demonstrate inadequate clopidogrel responses.1 In patients non-responsive or resistant to clopidogrel, novel P2Y12 antagonists (prasugrel and ticagrelor) as well as old P2Y12 antagonists (ticlopidine) may be used instead. Recently, several studies have shown that P2Y12-specific platelet function assays can predict both ischemic and hemorrhagic complications during neurointerventional procedures,2–6 and therefore may be useful for tailoring the degree of P2Y12 receptor inhibition.

The aim of our study was to analyze the clinical impact of platelet function-guided antiaggregant therapy compared with standard-dose clopidogrel treatment in patients after FDS implantation for intracranial aneurysm treatment. Platelet response to antiaggregant medications was measured using multiple electrode aggregometry.



Data were collected retrospectively on patients in whom FDS were implanted to treat intracranial aneurysm since January 2010. Patients with unruptured aneurysms were included if their aneurysms were giant, wide-necked, fusiform-dissecting, or recurrent/remnant (after intrasaccular coiling); among ruptured aneurysms, fusiform-dissecting or blister aneurysms were included. Collected data included the presence of subarachnoid hemorrhage (SAH), evidence of mass effect, aneurysm size, applied endovascular strategy, perioperative antiplatelet regimen, procedure-related technical complications, and hemorrhagic and thrombotic events occurring up to day 30. Some patients constituting our study group were also subjects in two previous articles published by the same team.7 ,8

In patients with unruptured aneurysms, the vascular anatomy of the aneurysm and parent vessels, and aneurysm-related parenchymal changes were assessed by cranial MRI/MR angiography and/or CT angiography. Ruptured aneurysms were assessed by CT angiography and digital subtraction angiography.

Impedance aggregometry

Platelet function was measured by using multiple electrode aggregometry (Multiplate analyzer, Verum Diagnostica/F Hoffmann-La Roche, Germany). There is a single-use test cell on the device which incorporates two independent impedance electrodes. During analysis, 300 µL of saline and 300 µL of patient blood already mixed with 25 µg/mL hirudin are pipetted into the test cell. After adding agonists (for the ASPI test, stimulation with 0.5 mM arachidonic acid and, for the ADP test, stimulation with 6.4 µM ADP), real-time recording of the test result is completed within 6 min. Arachidonic acid and ADP activate platelets, increase adhesion and aggregation over the electrode surface and thus change impedance of the electrodes. The impedance change is transformed to arbitrary aggregation units and plotted against time. The area under the aggregation curve (AUC) is used to quantify the aggregation response or expressed in units (U) (1 U corresponds to 10 AUC). The results represent the mean value of the two determined AUC values driven by two different electrodes (internal control). Pretreatment with aspirin and clopidogrel blocks adhesion and aggregation and lowers the values in responsive patients. A strong agreement has been found between the Multiplate analyzer and light transmission aggregometry, which is considered the standard method for assessing platelet reactivity.9 ,10 It is therefore accepted as a reliable, handy, rapid tool to monitor antiplatelet therapy in clinical practice and for many clinical investigations in the literature.


All endovascular procedures were performed under general anesthesia and systemic heparinization, which was usually initiated before placement of the guiding catheter to maintain the activated clotting time 2–3-fold above the baseline (initial heparin bolus 5000–10 000 IU). All patients received a loading dose of 600 mg clopidogrel (or, in clopidogrel-resistant patients, 1 g ticlopidine) and 300 mg aspirin 8–12 h before the procedure. Typically, antiaggregant pretreatment was administered at midnight the night before a morning endovascular procedure, along with simple sedatives to aid sleep if needed. During the first part of our study (standard group) we had no opportunity to assess platelet function while, in the remaining patients (aggregometry group), platelet aggregation was measured immediately before the endovascular procedure.

According to the consensus opinion of the Working Group on High-On-Treatment Platelet Reactivity, platelet aggregation >468 AUC (47 U) following clopidogrel pretreatment (normal range in the absence of antiaggregants is 569–1130 AUC as reported by the manufacturer) was considered indicative of clopidogrel non-response or hyporesponse (resistance).11 Aspirin non-responsiveness or hyporesponsiveness was defined as aggregation >500 AUC (50 U) (normal range in the absence of aspirin is 706–1148 AUC). In the standard group, FDS implantation followed the standard protocol. In the aggregometry group, however, according to the test results, the following procedures and loading doses were changed: in some clopidogrel-resistant patients, following ingestion of the alternative P2Y12 receptor antagonist ticlopidine at a loading dose of 1 g (reloading), the embolization procedure was postponed for at least 4 h. Other clopidogrel-resistant patients underwent the same scheduled embolization after 1 g of ticlopidine via nasogastric tubing under the protection of bridging intravenous tirofiban (a short half-life platelet glycoprotein IIb–IIIa receptor inhibitor) in a dose of 0.25–1 mg (aggressive loading). For the remaining hyporesponding patients, the heparin intravenous dose was increased during and after the procedure without a change in the antiaggregant medication (unchanged loading). In general, very high ADP values in patients with a stable aneurysm led to postponed procedures, high ADP values in patients with an aneurysm with a history of SAH or a recently symptomatic aneurysm led to modified loading, and limited hyporesponse to clopidogrel (ADP test values near the upper level of therapeutic window, 468 AUC) and a full response to aspirin were considered compatible with the standard procedure. The reloading and aggressive loading groups together constituted the ‘antiaggregant tailored’ group. After intravenous tirofiban bridging loading accompanied by peroral ticlopidine loading via a nasogastric tube, the ADP test was repeated 30 min and 4 h later. Following stent implantation, all patients were prescribed dual antiplatelet treatment with 100–300 mg aspirin (for 6 months) and 75 mg clopidogrel (or 2×250 mg ticlopidine) daily (for 3 months).

If MRI detected an objective mass effect such as perianeurysmal edema and/or neurological examination revealed cranial nerve palsy due to aneurysm compression, patients immediately received an intravenous steroid at the beginning of the procedure (methylprednisolone 80–250 mg) and were thereafter prescribed a peroral steroid (prednisolone) starting at 1–2 mg/kg and gradually decreasing the dose within 1 month.

Endovascular procedure

The parent artery was catheterized telescopically with a 6 Fr long sheath (Shuttle; Cook Medical, Bloomington, Indiana, USA or Destination; Terumo, Japan) and 6 Fr distal access guide catheter (Fargo/FargoMax, Balt Extrusion, Montmorency, France). All patients were implanted with Pipeline FDS (Covidien/ev3, Irvine, California, USA). During implantation, care was taken to obtain full expansion of the stent. In case of insufficient expansion, an intracranial balloon (Hyperglide, Covidien/ev3) was used to dilate the stent. Some procedures necessitated wire exchange maneuvers.

Statistical analysis

Statistical analysis was performed using the SPSS V.15.0 software. The Fisher exact test was used to analyze categorical variables in the standard and aggregometry groups. Within the aggregometry group, the Mann–Whitney U test was used to compare the ADP values in the tailored and untailored groups. A p value of < 0.05 was taken to indicate statistical significance.


Between January 2010 and June 2013, 100 consecutive patients with intracranial aneurysm undergoing FDS implantation in the interventional radiology department were enrolled into this retrospective study; the first 32 were allocated to the standard group (no platelet function test) while clopidogrel and aspirin response was tested by multiple electrode aggregometry in the remaining 68. There were 33 loadings in the standard group (one patient had bilateral aneurysms treated in a separate elective session) and 71 loadings in the aggregometry group (three patients had bilateral aneurysms treated separately). All 33 procedures in the standard group and 70 of 71 in the aggregometry group were performed after loading with clopidogrel; the remaining patient was loaded with ticlopidine. Baseline anatomical, clinical, and procedural characteristics were well balanced between the standard and aggregometry groups (table 1).

Table 1

Initial comparison of the two study groups

In the aggregometry group, 17 of 68 patients (25%) were considered clopidogrel non-responsive/hyporesponsive based on the cut-off value (>468 AUC). This predetermined cut-off value thus defined the upper quartile of platelet activity in our patient population. Five of these 17 patients were responsive to aspirin and had only limited clopidogrel resistance (ie, ADP test values were very close to the cut-off value of 468 AUC); their endovascular procedures proceeded without changes in the antiplatelet regimen, while the remaining 12 (17.6%) required changes. FDS implantation was postponed in three patients; the waiting period was 4 h for one patient and 1 day for another patient, while the third patient's ruptured aneurysm was occluded suboptimally with coils and the FDS was eventually implanted 1 month later. All three patients were reloaded with ticlopidine before FDS implantation.

The remaining 9 of the 12 resistant patients underwent FDS implantation in the same scheduled endovascular session (no postponement). We applied an aggressive loading regimen in these patients, which consisted of intravenous tirofiban infusion (0.25–1 mg, mean 0.69 mg) and peroral ticlopidine via nasogastric tubing. Tirofiban infusion reduced the median ADP value (n=9) from 719 (597–883) (mean 742.3) to 209 (52–651) (mean 322.4) after 30 min. The median ADP value (n=9) was finally measured as 289 (190–451) (mean 304) after 4 h, reflecting predominantly the effect of ticlopidine (tirofiban has a plasma half-life of <2 h). Overall, the ‘tailored antiaggregant’ regimen applied in cases of clopidogrel resistance (n=12) reduced the median ADP value from 827 (597–1164) (mean 828.3±209.4) to 310 (120–541) (mean 297.6±129.2). The ‘untailored antiaggregant’ group (n=56, loading=59) had a median ADP value of 186 (46–560) (mean 250.7±138.5). There was no significant difference between ADP values in the untailored group and the tailored tirofiban and tailored final ADP values (figure 1).

Figure 1

Box plot representation of ADP test values among the four subgroups of the aggregometry group. The significant difference between the untailored and tailored initial groups (p<0.001, Mann–Whitney U test) disappeared after a period of 30 min (under the predominant effect of tirofiban, tailored tirofiban) and 4 h (under the effect of tirofiban and ticlopidine, tailored final).

In the standard group (32 patients, 33 procedures) there were three (9.1%) spontaneous thrombotic events. An aneurysm in the embryonic posterior cerebral artery coming from the large-necked giant posterior communicating artery and adjacent anterior choroidal artery occluded spontaneously after 8 h in one patient, resulting in permanent hemiparesia. Another patient with a moderately sized distal internal carotid artery aneurysm experienced spontaneous carotid artery occlusion after 18 h, also causing permanent hemiparesia. The last patient, with a recanalized basilar tip aneurysm, developed acute stent thrombosis affecting the basilar, bilateral superior cerebellar and posterior cerebral arteries; although all these branches and the stent itself were cleared successfully following intravenous tirofiban, the patient developed posterior fossa ischemia and acute hydrocephalus and died 7 days later. Additionally, there was one (3%) technical hemorrhagic complication during the procedure, wire rupture of the distal internal carotid artery harboring a fusiform aneurysm, managed with parent artery occlusion. The patient became mildly hemiparetic. Overall, one death (3.1%) and three permanent morbidities (9.3%) occurred in the standard group.

In the aggregometry group (68 patients, 71 procedures) there were two (2.8%) spontaneous hemorrhagic events; neither was in the tailored antiaggregant group. One patient with a fusiform distal aneurysm located in the P2–P3 segment of the posterior cerebral artery experienced a huge fatal contralateral cerebellar hematoma 24 h following the procedure. The other, with a moderate distal internal carotid artery aneurysm, developed ipsilateral parenchymal hematoma 18 days later, which was also fatal. Although the ADP test results of these two patients were lower than the cut-off value for hyper-responsiveness to clopidogrel (cut-off 188 AUC; patient values 186 and 122 AUC), the statistical power did not support the increased likelihood of spontaneous hemorrhagic complications in hyper-responsive patients (p>0.12). Additionally, there was one (1.4%) fatal technical ischemic complication; in a patient with a giant ruptured vertebrobasilar junction aneurysm located on a fenestration, repeated balloon dilation within the insufficiently expanded stent resulted in brain stem infarction. In total, the aggregometry group included three deaths (4.4%) and no morbidity.

Among the various types of complication (thrombotic, hemorrhagic, technical), only the difference in thrombotic complications reached statistical significance (p<0.03); the aggregometry group had fewer thrombotic complications than the standard (no test) group. In terms of mortality, morbidity, and morbimortality, only morbidity differed significantly (p<0.03); morbidity in the aggregometry group was lower than in the standard (no test) group (table 2).

Table 2

Final comparison of the two groups


The results of our study demonstrate that, compared with uniform pretreatment with clopidogrel, a strategy incorporating multiple electrode aggregometry to tailor antiaggregant medication reduces thrombotic complications as well as permanent morbidity in FDS implantations to treat intracranial aneurysm.

Concern arose over a decade ago surrounding the clinical implications of poor platelet inhibition, as indicated by aggregometry, in clopidogrel-treated coronary intervention patients.12 In 2010, the Working Group on High-On-treatment Platelet Reactivity provided a consensus definition of high platelet reactivity to ADP for the most commonly used methods.11 Clopidogrel non-response or hyporesponse (resistance) is defined as >468 AUC and hyper-response is <188 AUC for the Multiplate on-site platelet function analyzer, so the therapeutic window is between 188 and 468 AUC (19 and 47 U).

Among many neurointerventional procedures, intracranial/supra-aortic stent implantation constitutes one of the main reasons for the use of antiaggregants. The intracranial FDS, indeed, necessitates dual antiaggregant pretreatment because the stent contains about fivefold more metal (more thrombogenic) than standard intracranial stents. The prevalence13 and clinical impact2 ,3 of clopidogrel non-response measured by rapid platelet function tests (Multiplate and VerifyNow) in patients with neuroendovascular procedures was first reported in 2008. These studies have strongly suggested a correlation of insufficient platelet inhibition by clopidogrel with an increased risk of thromboembolic events in supra-aortic stent placement for various reasons.2 ,3 This correlation has been further supported by Fifi et al,4 while Koerner et al14 and Nordeen et al15 found no correlation between insufficient clopidogrel response and thromboembolic event rate.

There may also be a connection between clopidogrel hyper-response and bleeding complications. A recent study by Goh et al5 demonstrated (using VerifyNow) the association of clopidogrel hyper-response with increased hemorrhagic complications following intracranial stenting. This study population included 15 FDS implantations in a total of 47 stenting procedures.

Cerebral aneurysm treatment by FDS constitutes a significant logical change in the field of endovascular treatment of cerebral aneurysms. The clinical results of Pipeline (Covidien/ev3) and Silk (Balt Extrusion) stents, two widely used examples of FDS, have recently been published in the neuroendovascular literature. A retrospective case series of FDS-treated patients reported a wide range of perioperative thromboembolic and hemorrhagic complications, with the risk of cerebral infarction up to 14% and the risk of intracerebral hemorrhage up to 11%.6 Assessment of the effect of modification of dose or type of antiaggregant may be crucial to lower the frequencies of thromboembolic and ischemic complications.

There is only one previous study exploring platelet reactivity measured by point-of-care function test (VerifyNow) devoted to the FDS (Pipeline) used in the intracranial circulation.6 Of note, three of their four major perioperative complications occurred in patients who had markedly elevated (one thromboembolic) or markedly decreased (two hemorrhagic) test values shortly before or at the time of the complication. These authors concluded that pre-procedure platelet reactivity outside the therapeutic window was the strongest independent predictor of perioperative thromboembolic and hemorrhagic complications after FDS placement.

By using on-site multiple electrode aggregometry (Multiplate analyzer), we adjusted antiaggregant interventions to maintain pre-procedural ADP test values below the upper limit of the therapeutic window in a group of patients (n=68) undergoing FDS implantation for aneurysm. Another group of patients (n=32) underwent the same treatment following uniform clopidogrel pretreatment without testing. We compared the rate of complications between these two groups and analyzed the clinical impact of the platelet function-guided strategy compared with standard dose clopidogrel treatment. We found that tailoring antiaggregants to the patient's P2Y12 receptor according to the aggregometry eradicated thrombotic complications and reduced permanent morbidity in our study population undergoing FDS implantation.

The proportion of clopidogrel hyporesponders in our cohort (25%) is similar to that reported by Delgado Almandoz et al6 (26% and 21%, according to the different cut-off values of VerifyNow). In general, the proportion of high on-treatment platelet reactivity after clopidogrel treatment (clopidogrel resistance) is estimated at 20–30%. Increasing the dose of clopidogrel and using another P2Y12 antagonist are well-known pre-procedural antiaggregant strategies in case of clopidogrel resistance in neuroendovascular procedures.2–6 ,13–15 These two strategies, however, appear to be insufficient to lower spontaneous (not technique-related) thrombotic complications6; this might be due to insufficient monitoring of platelet reactivity immediately after each change of dose and type of medication. This is the first study to explore the beneficial effect of tailoring antiaggregant medication guided by a rapid platelet function test (compared with standard medication) on the clinical results of FDS implantation for intracranial aneurysm. Our results suggest that platelet reactivity should be measured, even after each antiaggregant medication adjustment, to ensure that the final platelet reactivity is below the upper limit of the therapeutic window just before the procedure.

The mode of antiaggregant administration is one of the notable aspects of our study. Two methods of premedication with antiaggregant agents have been traditionally used: one is dosing the drugs for several days (typically 100–300 mg/day aspirin and 75 mg/day clopidogrel for at least 5–7 days) (continuous premedication) until a steady drug level is achieved, and the other is to load a high dose (typically 300 mg aspirin, 300–600 mg clopidogrel) of the drugs (high-dose loading) immediately before the procedure (at least 4 h, ideally 12–24 h). If high-dose loading is performed, therapeutic levels of platelet inhibition can be achieved much more rapidly, sometimes within hours, depending on the dose. In cases of continuous premedication lasting 5–7 days, the duration of compromised platelet function is recognized as a potential risk for hemorrhagic events in patients with intracranial aneurysm. There may also be concern among both patients and physicians regarding hemorrhagic events related to aneurysm, especially in patients with a history of SAH.8 Indeed, among patients with ischemic cerebrovascular disease, the risk of life-threatening bleeding is significantly higher (1.7% for all bleeding events and 0.7% for intracranial bleeding events per year) in the dual antiplatelet treatment group than in the antiplatelet monotherapy group.16 We therefore prefer a high-dose single-loading strategy for our patients.

The reloading strategy is another distinctive aspect of our study. After discovery of clopidogrel resistance, one can postpone the endovascular procedure followed by reloading with another P2Y12 receptor antagonist and the procedure is performed later (as in three of the 12 patients). This first option appears to be relatively safe for stable patients. We continued the procedure (no postponement) in the remaining nine patients after reloading with another P2Y12 receptor antagonist via a nasogastric tube under the protection of bridging tirofiban infusion. Repeating the ADP test 30 min and 4 h later confirmed a sufficient level of platelet inhibition (figure 1), which was initially due predominantly to the effect of the tirofiban itself, but later to the concomitant effect of tirofiban and ticlopidine. This second option may be feasible in patients with a history/suspicion of SAH and intradural aneurysm presenting with abrupt/intense symptoms (evidence of sudden enlargement or prone to rupture) because postponement may expose the patient to the hypothetical ‘erosive’ effect of aspirin loaded previously.

Evidence regarding the use of alternative new drugs such as prasugrel and ticagrelor have been mixed and have raised strong doubts with respect to increased hemorrhagic complication.15 ,17 We prefer ticlopidine (an old P2Y12 receptor antagonist) for reloading in case of clopidogrel resistance. Although the hematological profile of ticlopidine is poor, its short period of usage and close laboratory follow-up of hematologic profile make the drug preferable in many patients.

The limitations of this study are its single-center retrospective design, limited patient numbers, and heterogeneous nature of aneurysms (ruptured and unruptured). Another important limitation is lack of blinding. Ideally, to compare the clinical impact of the platelet function-guided strategy with the standard dose of clopidogrel in FDS implantation to treat intracranial aneurysm, the study design should have randomized patients to either group. However, we felt that the evidence linking the risk of thrombosis and platelet resistance to P2Y12 inhibitors made such a blinded design potentially unethical because it would require denying appropriate treatment to clopidogrel-resistant patients. We therefore constituted the control (no test) group from our archive prior to the acquisition of platelet function testing equipment. Nevertheless, a double-blinded multicenter design with a higher number of patients would be the ideal means of evaluating whether the platelet function-guided strategy using rapid platelet assays is better than the standard therapy. The theoretical rationale for tailoring platelet reactivity according to aggregometry is that confining platelet reactivity within the therapeutic window may prevent both thrombotic and hemorrhagic complications. The lack of antiaggregant intervention adjustment in hyper-responsive patients is another limitation of the current study. Finally, the temporal effect of operator experience (learning curve) on outcomes should be commented upon, since the standard antiplatelet therapy arm was early in the use of FDS in our study.


Our results suggest that rapid platelet function assays may be useful to guide adjustments to antiplatelet therapy. In our sample a tailored antiplatelet regime decreased the rate of complications following FDS implantation.


View Abstract


  • Contributors IO and CC participated in the conception and design of the study. HB and MK analyzed and interpreted the data. IO and CC treated the patients in the study. IO wrote the article. All authors advised, reviewed, and approved the manuscript.

  • Patient consent Informed consent was provided by patients and/or their relatives before endovascular therapy.

  • Ethics approval The study was approved by the Institutional Review Board.

  • Competing interests None.

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

  • Data sharing statement No unpublished data from the study have been given to anybody.