Objective Limited data exist regarding the use of antiplatelet response assays during neuroendovascular intervention. We report outcomes after carotid artery stenting (CAS) based on aspirin and P2Y12 assays.
Methods We retrospectively identified patients who had aspirin and P2Y12 assays at the time of stenting. Aspirin (325 mg) and clopidogrel (75 mg) were started 7–10 days pre-intervention. If not possible, aspirin (650 mg) and clopidogrel (600 mg) loading doses were given pre-intervention. Assays were checked on postoperative day 0/1. Outcomes included neurological ischemic sequela at 30 days, 1 and 2 years, as well as 30 day death/hemorrhage/myocardial infarction.
Results 449 patients were included. Mean P2Y12 reaction unit (PRU) values were higher in patients with an ipsilateral ischemic event (stroke/transient ischemic attack (TIA)) or stroke (alone) at 1 and 2 years than in patients with no events: ischemic event versus no event at 1 year, 252 vs 202 (p=0.008); stroke versus no stroke at 1 year, 252 versus 203(p=0.029); ischemic event versus no event at 2 years, 244 vs 203 (p=0.047); stroke versus no stroke at 2 years, 243 versus 203 (p=0.082). Ischemic event free survival (stroke/TIA, p=0.0268) and overall survival (p=0.0291) post-CAS were longer in patients with PRU ≤198 compared with an initial threshold of PRU ≤237. Mean PRU values were higher in patients who died from all causes at 30 days than in survivors (p=0.031). No correlation was found between lower PRU values and hemorrhage. Aspirin reaction units did not correlate with outcome.
Conclusions PRU ≤198 may be associated with a lower incidence of ischemic neurological sequela and death post-CAS. Prospective studies are needed to validate the relationship between antiplatelet assays and outcomes post-CAS.
Statistics from Altmetric.com
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.
To date, there is limited evidence describing the use of antiplatelet response assays in the neuroendovascular literature.1–5 Such tests are reported in the cardiac literature to guide aspirin and clopidogrel therapy in preventing ischemic sequelae after percutaneous coronary intervention (PCI).6–9 Furthermore, a ‘non-responder’ subset of patients has been identified that may be at higher risk for myocardial infarction (MI) after PCI on the basis of high platelet on treatment reactivity (OTR).6–13 Theoretical similarities between percutaneous coronary and percutaneous neurological interventions imply that these tests should play a similar role in platelet inhibition monitoring. We describe our experience using aspirin and P2Y12 assays in therapeutic antiplatelet monitoring after carotid artery stenting (CAS).
We conducted a retrospective review of our carotid stenting registry to identify patients for whom aspirin and P2Y12 response assays were recorded at the time of CAS. Patient data, including demographics, antiplatelet regimen, serologic testing, medical comorbidities, ipsilateral neurological ischemic events (transient ischemic attack (TIA) and stroke), and relevant angiographic findings were recorded. Platelet function was assessed with the VerifyNow Aspirin Assay and VerifyNow P2Y12 Assay (Accumetrics, San Diego, California, USA) in aspirin reaction units (ARU) and P2Y12 reaction units (PRU), respectively.
Dual antiplatelet regimen consisted of aspirin (325 mg daily) and clopidogrel (75 mg daily), starting 7–10 days pre-intervention. Under circumstances in which this regimen could not be achieved, loading doses of aspirin (650 mg) and clopidogrel (600 mg) were given on the day of intervention. Aspirin and P2Y12 assays were then checked on postoperative day 0 or 1 to evaluate therapeutic levels. Although therapeutic platelet inhibition thresholds are a source of debate in the current literature, our institution adopted thresholds of ARU ≤550 and PRU ≤237.2 ,9 ,14–16 Patients were maintained on dual antiplatelet therapy for 3 months after CAS, after which time aspirin monotherapy was maintained indefinitely. No adjustments to antiplatelet regimens were made in response to antiplatelet assays in this study cohort.
Patients were included in the present study if they were >18 years of age and underwent carotid stent placement in the setting of the documented dual antiplatelet regimen described above. Patients were excluded if they had: (1) contraindications to antiplatelet therapy, including intracranial hemorrhage, gastrointestinal bleeding, major systemic trauma, and primary or secondary coagulopathy; (2) abnormal laboratory parameters that would influence assay function, including platelets <92 000 µL and hematocrit <33% or >52%; (3) lack of recorded aspirin/P2Y12 responses; or (4) loss to follow-up.
Primary endpoints of interest were the prevalence of ipsilateral ischemic event (defined as stroke or TIA) and stroke (only) at 30 days, and 1 and 2 years after CAS, stratified by aspirin and P2Y12 assays. Stroke was defined as a new thromboembolic event, confirmed with neuroimaging, that resulted in permanent neurological deficit associated with an increase in the National Institutes of Health Stroke Scale (NIHSS) score of at least 2 points. TIA was defined as a transient neurological deficit associated with an increase in NIHSS score of at least 2 points that resolved within 24 h and had no correlating findings on neuroimaging. These outcomes were determined from review of inpatient and outpatient charts and imaging studies. Because all strokes occurred ipsilateral to the stented carotid artery and fell within the qualifying artery territory, they were presumed to be atheroembolic rather than cardioembolic in origin. Secondary endpoints of interest included prevalence of hemorrhage (intracranial and/or extracranial), death (all cause or stroke related), and MI at 30 days, also stratified by aspirin and P2Y12 assay results. Intracranial hemorrhage was defined as any intra- or extra-axial hemorrhage that was present on imaging after carotid revascularization. This imaging was obtained on an as needed basis prompted by clinical concern. Intracranial hemorrhage was then classified as symptomatic versus asymptomatic on the basis of an associated increase in NIHSS score of at least 2 points, correlating with findings on neuroimaging. Extracranial hemorrhage was defined as any clinically overt bleeding requiring transfusion or causing a decrease in hemoglobin of 2–3 g/dL, or any hematoma leading to a prolonged or new hospitalization.17 MI was defined as detection of a rise and fall in cardiac biomarkers with at least one value above the 99th percentile of the upper reference limit, together with evidence of myocardial ischemia based on symptomatology or electrocardiogram/echocardiogram changes.18 Serial cardiac enzyme levels were obtained for every patient after CAS in accordance with our institution's protocol. This retrospective analysis was approved by our local institutional review board.
Description of endovascular technique
All CAS procedures were performed under conscious sedation and a local anesthetic using femoral artery access. The size of the arterial sheath was contingent on the intention to use proximal verses distal embolic protection devices. Distal embolic protection devices, including Emboshield Nav 6 (Abbott Medical, Santa Clara, California, USA) or Filterwire EZ (Boston Scientific, San Jose, California, USA), were typically passed through 6 Fr Cook Shuttle long sheaths (Cook, Bloomington, Indiana, USA) or another 6 Fr guide catheter after systemic heparinization to a goal activated coagulation time of 250–300 s. Proximal embolic protection devices, including the Mo Ma Ultra (Invatec, Roncadelle, Italy) or Gore flow reversal system (WL Gore and Associates, Flagstaff, Arizona, USA), were passed through 9 Fr sheaths after heparinization. All cases were performed with embolic protection. Choice of stent was contingent on vascular anatomy, plaque morphology, and senior authors' preference. In general, no pre-stent angioplasty was performed except in cases when the lesion could not be crossed prior to angioplasty. Post-stent angioplasty was performed with a balloon sized either under (if the patient was symptomatic) or up to (if asymptomatic) the diameter of the distal internal carotid artery. Sheath removal and vascular closure were performed with Mynx (AccessClosure Inc, Mountain View, California, USA) or Angio-Seal (St Jude Medical, Minnesota, USA) closure devices.
Statistical analysis was done using commercially available software (Graphpad 6.0, Graphpad Software Inc, La Jolla, California, USA; and Microsoft Office Excel 2007, Microsoft Inc, Redmond, Washington, USA). The relationship between primary and secondary endpoints in reference to ARU and PRU values was analyzed using a two tailed Student's t test or Mann–Whitney test when appropriate. A p value <0.05 was considered statistically significant. Kaplan–Meier survival curves were generated to identify ischemic event free (stroke and/or TIA), stroke free, and overall survival after CAS in reference to ARU and PRU values. Statistical significance between curves was analyzed with Mantel–Cox log rank tests.
Between January 2008 and January 2011, 514 patients were identified in our carotid stent registry, of whom 65 were excluded based on the criteria discussed above; therefore, 449 patients were included in this study. Age ranged from 27 to 94 years (mean 70 years). Among the 449 study patients, 231 (51.4%) underwent stent placement for symptomatic disease whereas 218 (48.5%) underwent stent placement for asymptomatic disease. Medical comorbidities considered risk factors for stroke were similar between both groups (table 1). The follow-up period after CAS ranged from 6 months to 3 years (mean 18 months).
The prevalence of ischemic events (stroke or TIA) at 30 days was 2.4% (n=11). The prevalence of stroke (only) at 30 days was 1.3% (n=6) (2.2% (n=5) in the symptomatic group and 0.5% (n=1) in the asymptomatic group). The prevalence of ischemic events at 1 year was 4.9% (n=22). The prevalence of stroke at 1 year was 2.9% (n=13) (5.2% (n=12) in the symptomatic group and 0.5% (n=1) in the asymptomatic group). The prevalence of ischemic events at 2 years was 5.3% (n=24). The prevalence of stroke at 2 years was 3.1% (n=14) (5.6% (n=13) in the symptomatic group and 0.5% (n=1) in the asymptomatic group). The prevalence of death from all causes at 30 days was 1.8% (n=8). The prevalence of death related to stroke at 30 days was 0.7% (n=3). Other causes of death at 30 days included MI (n=2), intracranial hemorrhage (n=1), and sepsis (n=2). The prevalence of MI at 30 days was 2.9% (n=13). The prevalence of a 30 day hemorrhagic event was 4.7% (n=21), of which nine hemorrhages were intracranial (2.0%) (six symptomatic vs three asymptomatic) and 12 were extracranial (2.7%).
P2Y12 reaction units
Primary outcomes stratified by mean PRU are displayed in table 2. Mean PRU values were higher in patients with ischemic or stroke events at 1 and 2 years compared with patients with no events. Ischemic event free survival after CAS was longer in patients when therapeutic the PRU goal was defined as ≤198, in comparison with the original PRU goal of ≤237 (figure 1A, B). There was a strong association of longer stroke free survival after CAS in patients with PRU ≤198 (figure 1C).
Aspirin reaction units
Primary outcomes stratified by mean ARU are displayed in table 3. ARU values were not associated with statistically significant differences in primary outcomes after CAS.
P2Y12 reaction units
Secondary outcomes stratified by mean PRU are displayed in table 4. Mean PRU values were higher in patients who died from all causes or related to stroke at 30 days than in patients who survived after CAS. Mean PRU values were higher in patients who sustained any hemorrhage or isolated intracranial hemorrhage at 30 days than in patients without hemorrhagic events. Statistical association between increased platelet inhibition (ie, low PRU) and hemorrhagic sequelae could not be established. No difference in PRU values in patients with and without MI was appreciated. Overall survival after CAS was longer in patients when therapeutic PRU goal was defined as ≤198 (figure 1D).
Aspirin reaction units
Secondary outcomes stratified by mean ARU are displayed in table 5. ARU values were not associated with statistically significant differences in secondary outcomes after CAS.
To date, evidence describing the use of antiplatelet response assays in the neuroendovascular literature has been limited and largely restricted to feasibility and resistance studies.1–5 Thromboembolic events related to subtherapeutic platelet inhibition have been implied; however, these studies consist of small patient populations with heterogeneous pathologic conditions prompting the use of antiplatelet therapy.2 ,5 Therefore, there remains a gap in the neurointerventional literature that correlates aspirin and P2Y12 assays with patient outcomes after interventions. We can provisionally bridge this gap with cardiac literature that has identified a ‘non-responder’ subset of patients at risk for ischemic sequela after intervention on the basis of high platelet OTR.6–13
Several studies have suggested that patients with high OTR while receiving clopidogrel are at risk for ischemic events after percutaneous cardiac interventions.19–22 In the Gauging Responsiveness with A VerifyNow assay-Impact on Thrombosis and Safety (GRAVITAS) trial, high versus standard dose clopidogrel therapy was effective in lowering PRU values in patients with high OTR but did not show significant changes in the composite primary outcome of death–MI–stent thrombosis when the threshold for therapeutic platelet inhibition was PRU ≤237.16 When the threshold for platelet inhibition was lowered to a PRU value of ≤208 in a GRAVITAS post hoc analysis, the primary endpoint was significantly lower in patients with PRU ≤208 and found to be an independent predictor of outcome after intervention.9 This threshold adjustment and post hoc analysis were made in response to additional studies showing further protective benefits using newer more potent antiplatelets that allowed better platelet inhibition in patients who were ‘non-responders’ to clopidogrel.10–11 ,13 In long term follow-up of the Tailoring Treatment with Tirofiban in Patients Showing Resistance to Aspirin and/or Resistance to Clopidogrel (3T/2R) study, the composite endpoint described above was 8.6% in patients with OTR ≤208 versus 15.8% in patients with OTR >208.10 Furthermore, in the Testing Platelet Reactivity In Patients Undergoing Elective Stent Placement on Clopidogrel to Guide Alternative Therapy with Prasugrel (TRIGGER-PCI) study, prasugrel was used to achieve an average PRU value of 80 in clopidogrel non-responders, thus demonstrating newer therapeutics available to further improve platelet inhibition after percutaneous intervention.12 Collet et al23 reported no difference in outcomes after coronary stenting in patients with antiplatelet regimens adjusted according to response assays (therapeutic PRU threshold defined as OTR ≤235) compared with patients who were not monitored. They recommend further analysis of different threshold values as this may have an impact on procedural outcomes.
The above cardiac studies illustrate two important points. First, there has yet to be a defined optimal goal of platelet inhibition that identifies maximal therapeutic protection against ischemic events. Second, there appears to be a convincing relationship between increased platelet inhibition and reduction of primary/secondary endpoints after intervention. This is consistent with our findings of (1) lower mean PRU values in patients who remained free of neurological ischemic events, stroke, or death; and (2) longer stroke free and overall survival after CAS in patients with PRU of ≤198 in comparison with our initial cut-off of PRU ≤237. Our center's initial therapeutic PRU cut-off of ≤237 was derived from a combination of cardiac and manufacturer sources that used receiver operating characteristic curve analyses in identifying OTR levels that provided the highest prediction of cardiovascular events after intervention.6 ,20–22 We subsequently choose to evaluate the lower PRU cut-off of ≤198 because it has been identified in the neurointerventional literature available to date and is achievable in clinical practice with the above aspirin and clopidogrel regimen.2 ,5
It is important to note that mean PRU values were higher in patients with ischemic or stroke events at 1 and 2 years compared with patients without events, but not within the 30 day period. We conjecture that one of the main reasons that 30 day ischemic and stroke rates were not significantly associated with higher PRUs is because they can be influenced by periprocedural events that may not be protected by dual antiplatelet regimens—namely, technical or mechanical challenges of surgical technique within diseased anatomy. This is congruent with findings of the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST) of increased ipsilateral stroke rates within the periprocedural period in CAS patients (CAS 4.1% vs carotid endarterectomy 2.3%; p=0.03) followed by similarly low ipsilateral stroke rates after the periprocedural period (CAS 2.0% vs carotid endarterectomy 2.4%; p=0.85).24 We believe the antiplatelet regimen affords the most stroke protection during the period of stent endothelialization and reactive intimal hyperplasia associated with stenting and angioplasty, which may be prolonged for several months in some patients. For this reason, dual antiplatelet therapy is maintained for 3 months. However, no consensus currently exists on duration of dual therapy, and maintenance for up to 1 year has been described.1 ,25 We hypothesize that this explanation might account for the protective effects of antiplatelet therapy found in our therapeutic cohort at 1 and 2 years but not within the 30 day period. Of note, it is more challenging to rationalize why overall survival benefits of therapeutic OTR were seen predominately within the 30 day period despite the lack of significant immediate postoperative PRU differences in patients with ischemic/stroke events. We hypothesize that this can be related to medical comorbidities, largely cardiac disease, that benefit from therapeutic OTR. Of the patients in our cohort who died from MI (n=2), both had high OTR with a mean PRU of 321. However, due to the small number of events, our statistical analysis is underpowered and further conclusions become tenuous. We hope to illustrate an area of study that requires further attention within the neuroendovascular field.
As a consequence of the findings in our study, if high P2Y12 OTR is now identified in our patients, another loading dose of 600 mg of clopidogrel is administered and maintenance is increased to 150 mg daily. If subsequent P2Y12 assays demonstrate high OTR, patients are switched to a 60 mg loading dose of prasugrel with 5 mg daily maintenance or a 180 mg ticagrelor loading dose with 90 mg daily maintenance.
We expected a higher risk of hemorrhagic events in patients with lower OTR values—that is, maximized platelet inhibition; however, this pattern could not be established in our patient population. We found higher PRU values in patients with hemorrhagic sequelae. This lack of correlation between higher PRU values and hemorrhagic events is consistent with the cardiac literature previously cited,9 ,16 ,19–22 with the exception of the TRIGGER-PCI study.12 In that study, patients in the prasugrel arm had a mean PRU value of 80, with an event rate of 1.4%, whereas the clopidogrel arm had a mean PRU value of 241, with an event rate of 0.5%.12 Similarly, Goh et al26 correlated hyperresponse to clopidogrel as a risk factor for bleeding sequelae during neurointerventional procedures, identifying a ‘hyperresponder’ group that had a 42.8% rate of major bleeding complications when platelet inhibition was >72% (PRU ≤83). At the present time, the optimum threshold of platelet inhibition that provides the highest predictive value of hemorrhagic events after intervention remains undefined and continues to be the source of study.
This paper has several limitations. First, our hemorrhagic event numbers were heavily influenced by perioperative access complications related to the femoral artery approach. After excluding these complications, there still remained no correlation between lower PRU values and hemorrhage. Second, the antiplatelet regimen used in this study consisted of aspirin and clopidogrel. Newer antiplatelet medications, such as prasugrel and ticagrelor, have limited experience in the neuroendovascular literature to date. Of note, prasugrel is contraindicated in any patient with a history of TIA or stroke, thus impacting its use in this population.27 Ticagrelor has a drug–drug interaction, with doses of aspirin >100 mg resulting in its decreased efficacy, thereby limiting concurrent aspirin doses to 81 mg.28 Older antiplatelet medications, such as ticlopidine, have a limited safety profile with multiple known hematologic toxicities, such as neutropenia, agranulocytosis, thrombotic thrombocytopenia purpura, and aplastic anemia, requiring additional monitoring of a complete blood count with differential every 2 weeks for 3 months after initiation.29 Third, this paper is a retrospective review of a moderately sized CAS registry. Further evidence derived from larger sized, prospective studies correlating outcomes with assay response variability over time is required to validate our findings. Fourth, because of the limited number of stroke events in our asymptomatic CAS cohort (n=1), stroke outcomes were disproportionately influenced by our symptomatic CAS cohort (n=13). Further study focusing on an asymptomatic cohort and antiplatelet assays after CAS is required. Finally, we chose to look at our carotid stent registry because it represents a homogeneous population of extracranial atherosclerotic disease. Whether the ischemic and overall survival benefits of PRU values ≤198 described above are applicable to other neuroendovascular interventions, such as stenting for intracranial atherosclerotic disease or stent assisted coiling or flow diversion for aneurysm treatment, remains unknown.
In accordance with the percutaneous coronary literature, our study shows that there may be potential utility in monitoring platelet inhibition during CAS. Furthermore, we found P2Y12 OTR of ≤198 to be significant in the prevention of ischemic neurological sequelae and death after CAS. As a consequence, we titrate our antiplatelet regimen to a goal of PRU ≤198 prior to carotid stent placement. Multicenter, prospective studies are required to validate these findings.
We thank Jihnhee Yu, PhD, from the Department of Biostatistics, University at Buffalo, State University of New York, for statistical review and figure recommendations, Paul H Dressel, BFA, from the University at Buffalo Department of Neurosurgery for assistance with preparation of the illustrations, and Debra J Zimmer (also from the neurosurgery department) for editorial assistance.
Contributors Concepts and design: GCS, MSB, and AHS. Data analysis, acquisition, and interpretation: all authors. Literature research: GCS and MSB. Drafted manuscript: GCS. Statistical analysis: GCS, TMD, and SKN. Review and critical revision of the manuscript: all authors.
Competing interests LNH: research support—Toshiba; consultant—Abbott, Boston Scientific, Cordis, Micrus, and W L Gore; financial interests—AccessClosure, Augmenix, Boston Scientific, Claret Medical, Micrus, and Valor Medical; board/trustee/officer—AccessClosure and Claret Medical; speakers’ bureau—Abbott Vascular; honoraria—Boston Scientific, Cleveland Clinic, Complete Conference Management, Cordis, SCAI, University of Southern California, and VIVA Physicians. EIL: research grant, devices, and honoraria—Boston Scientific; research support—Codman and Shurtleff and ev3/Covidien Vascular Therapies; ownership interests—Intratech Medical and Mynx/AccessClosure; consultant—Codman and Shurtleff, ev3/Covidien Vascular Therapies, and TheraSyn Sensors; fees for carotid stent training—Abbott Vascular and ev3/Covidien Vascular Therapies. MM: educational grant—Toshiba. AHS: research grants—National Institutes of Health (co-investigator: NINDS 1R01NS064592-01A1, not related to present submission) and University at Buffalo; financial interests—Hotspur, Intratech Medical, StimSox, Valor Medical, and Blockade Medical; consultant—Codman and Shurtleff, Concentric Medical, Covidien Vascular Therapies, GuidePoint Global Consulting, Penumbra, Stryker Neurovascular, and Pulsar Vascular; speakers’ bureaus—Codman and Shurtleff and Genentech; serves on National Steering Committees for Penumbra 3D Separator Trial and Covidien SWIFT PRIME Trial; advisory board—Codman and Shurtleff and Covidien Vascular Therapies; honoraria—American Association of Neurological Surgeons’ courses, Annual Peripheral Angioplasty and All That Jazz Course, Penumbra, and from Abbott Vascular and Codman and Shurtleff for training other neurointerventionists in carotid stenting and for training physicians in endovascular stenting for aneurysms.
Ethics approval The institutional review board at the University at Buffalo, State University of New York, approved this study (project NSG2180712E), and a standard Health Insurance Portability and Accountability Act compliant protocol was followed.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Unpublished anonymised/de-identified data may be available. This would be on a per request basis.