Citation
Print
INTRODUCTION
Endovascular coil embolization for the treatment of unruptured intracranial aneurysms (UIA) is relatively safe and effective compared with other surgical treatments such as aneurysm clipping, which has been increasingly performed since its introduction into clinical practice3,23). Particularly, stent-assisted coil embolization (SAC) procedures, which result in increased packing density and decreased rates of coil migration into the parent artery, have gained widespread use as an effective treatment for UIA, and are particularly useful for wide-necked aneurysms9,22,29). Despite advances in technique and management in recent years, this procedure is associated with thromboembolic complications12).
Dual antiplatelet premedication protocol of 100 mg of aspirin and 75 mg of clopidogrel daily for at least seven days prior to SAC has been widely used for decreasing thromboembolic risk13,20). Despite the routine use of dual antiplatelet premedication, thromboembolic complications continue to be encountered17). Recent studies have resulted in a growing body of neurointervention literature indicating that individual variability in patient response to antiplatelet therapy, especially to clopidogrel, may be associated with some of the thromboembolic complications encountered after endovascular procedures1). Several studies have reported a 7% to 40% incidence of thromboembolic events after SAC, in patients with high on-treatment platelet reactivity (HTPR)11).
In cardiology, many publications suggest that the variability of platelet response in patients treated with clopidogrel is correlated with the cytochrome P450 genotype, diabetes, and chronic kidney disease2,15,28). However, in the neuro-intervention literature, there is no established link between the variability of clopidogrel response and cerebral infarction.
Therefore, the purpose of our study was to analyze the variability of clopidogrel responses according to the duration of post-SAC clopidogrel regimens, and to assess any possible correlation between the variability of clopidogrel responses and thromboembolic or hemorrhagic complications.
MATERIALS AND METHODS
Study design
This retrospective study included patients from a prospectively collected database between January 2014 and June 2015. Total of 202 patients with UIA were treated in a single center. Inclusion criteria were as follows : patients underwent a SAC procedure due to UIA with dome to neck ratio >2, had a risk of coil prolapse, had a need for flow diversion, and had out-patient department follow-up data for more than one year after the procedure. In addition, patients who had measured antiplatelet response assays one day before the procedure and one month after the procedure were included. Exclusion criteria were as follows : patients who were treated with additional coil packing with stent assistance at a previous procedure site, treated with a flow diverter, such as Pipeline (EV3, Irvine, CA, USA), or failed of stent deployment. After inclusions and exclusions, 47 patients were enrolled into this study. This study was approved by our institutional review board.
Patient data
Patient clinical data, laboratory data, and preoperative drug-use history were obtained by a retrospective chart review. Clinical data consisted of age, sex, diabetes mellitus (DM) status, blood pressure, history of smoking, alcohol use, history of stroke, history of heart disease, and body mass index (BMI). Laboratory data consisted of glomerular filtration rate, platelet count, cholesterol levels, triglyceride levels, high-density lipoprotein levels, and low-density lipoprotein levels7,15,18,21). These factors were defined as follows : DM and hypertension (HTN) were defined by medical treatment after diagnosis, patients were considered smokers only if they currently smoked, alcoholism was defined as the consumption of more than 14 standard drinks per week or four drinks per day, history of stroke was determined by a previous cerebral infarction or intracranial hemorrhage, and history of heart disease included myocardial infarction, angina, valve disease, congenital heart disease, or previous heart operations. BMI was checked at admission and laboratory data were checked in the evening of the admission day, one day before the procedure. In addition, because clopidogrel is well known to be metabolized by the cytochrome P450 enzyme in the liver, we also checked whether patients were using other drugs that may compete with clopidogrel during metabolism, which includes statins, metformin, proton pump inhibitors (PPI), histamine-2 receptor antagonists, selective serotonin reuptake inhibitors (SSRI), calcium channel blockers (CCB), β-blocker, angiotensin-converting enzyme inhibitors (ACE inhibitor), angiotensin II receptor blockers (ARB), diuretics, and warfarin19,27,30).
Premedication protocol and antiplatelet response assay
Our protocol of antiplatelet premedication called for at least seven days of dual antiplatelet medication with daily doses of 100 mg of aspirin and 75 mg of clopidogrel. On the day prior to the coiling procedure, we checked the pre-procedural VerifyNow P2Y12 Assays (Accumetrics, San Diego, CA, USA), which determined pre-procedural P2Y12 reaction unit (pre-PRU) for clopidogrel24). An abnormal antiplatelet response was defined as ≥240 PRUs5). After the procedure, the antiplatelet regimen remained the same as before the procedure, regardless of the antiplatelet response assay value. In addition, the follow-up antiplatelet response assay (post-PRU) was assessed 30 days after the procedure, with the same protocol as above. In general, three months of dual antiplatelet therapy is given and then changed to monotherapy, but for some patients the duration of clopidogrel use may change depending on the symptoms. PRU variability was defined as the decreased value from pre-PRU and post-PRU.
Procedure
Endovascular treatment was performed under general anesthesia, by using a biplane angiographic unit (Artis zee, Siemens, Erlangen, Germany). All patients in this study were treated with SAC. The choice of stent was determined by vessel structure or the clinician’s individual policy. During the procedure, systemic anticoagulatives with heparin were administered from the beginning of the procedure. Heparin was given as a continuous infusion throughout the procedure to maintain an activated clotting time that was 2–3 times longer than the baseline clotting time. To evaluate the thromboembolic risk associated with the procedure, the size and location of the aneurysm was measured by an angiography before the procedure, and the degree of occlusion and thrombus formation in the stent were also immediately measured by the post-procedural angiography. Aneurysm occlusion status was determined by the Raymond Roy occlusion classification25).
Follow-up
Follow-up brain magnetic resonance imaging (MRI) scans were performed twice : one day, and one year after the procedure. Thromboembolic infarction was defined as a symptomatic ischemic event, such as transient ischemic attack, or asymptomatic ischemic event with evidence of infarction on diffusion-weighted imaging occurring after discharge. In addition, we analyzed the vascular territory of the ischemic attack according to vessel location of the inserted stent. Hemorrhagic complications were classified as major or minor according to the Thrombolysis in Myocardial Infarction (TIMI) bleeding criteria16). All MRIs and angiographies were reviewed by two independent, board-certified radiologists. If disagreement occurred, a decision was made by consensus after additional review of the MRIs and angiographies.
Statistics
Statistical analyses were performed using SPSS version 21.0 (IBM Corp., Armonk, NY, USA). Baseline characteristics of clopidogrel response were assessed and compared between the two groups (responsive vs. hypo-responsive) using the chi-square test or Fisher’s exact test in categorical variables, and Student’s t-test or Mann-Whitney test were used for parametric variables. To assess the degree of PRU variation and thromboembolic infarction or hemorrhagic complication, univariate analysis of clinical, radiological, laboratory findings, and drug-use history was performed using a chi-square test or Fisher’s exact test, and a Student’s t-test or Mann-Whitney test. Variables with a p value of <0.10 in univariate analysis were then chosen for multivariate analysis using a logistic regression model. In all analyses, p values ≤0.05 were defined as statistically significant.
RESULTS
Responsive group vs. hypo-responsive group
Among the 47 patients who had undergone SAC, 10 patients (21.3%) were included in the hypo-responsive group, and 37 (78.7%) were in the responsive group. Mean values of pre-PRU were 271.10±26.52 in the hypo-responsive group, and 181.27±44.4 in responsive group (p<0.001). Overall mean value of PRU variability was 128.55±9.55 (171.00±59.30 in the hypo-responsive group vs. 117.08±62.94 in the responsive group). In the hypo-responsive group, PRU was significantly more variable than in the responsive group (p=0.019). No factor was significantly different between the two groups except the pre-PRU and PRU variability (Table 1).
PRU variability
According to mean value of PRU variability, patients were divided into a “large variation” group (PRU≥128.55, n=27 [57.4%]) and a “small variation” group (PRU<128.55, n=20 [42.6%]). In univariate analysis, comparison of the two groups showed that HTN and ACE inhibitor or ARB usage was significantly more frequent in the Large variation group (p=0.026 and p=0.004, respectively). In addition, pre-PRU and serum platelets count were significantly higher in Large variation group (p=0.008, and p=0.001, respectively). In multivariate analysis, only pre-PRU and serum platelets were significant correlated with PRU variation count (p=0.005, odds ratio [OR] 1.034, 95% confidence interval [CI] 1.010–1.059; p=0.004, OR 1.043, 95% CI 1.013–1.073, respectively). No other variables showed a statistically significant correlation (Table 2).
Thromboembolic infarction, hemorrhagic complication
Among radiologic factors, the mean size of aneurysms was 4.2±0.32 mm. Twenty-six (55.3%) aneurysms were located on the internal carotid artery, 11 (23.4%) on the anterior cerebral artery, seven (14.9%) aneurysms on the middle cerebral artery, and three (6.4%) aneurysms on the posterior circulation. No thrombi accumulated in the stent during the procedure. Total occlusion occurred in 38 cases (80.9%), residual neck of the aneurysm was evident in four cases (8.5%), and incomplete occlusion occurred in five cases (10.6%). Thirty-five patients (74.5%) received a Solitaire stent (Medtronic, Dublin, Ireland), seven (14.9%) obtained a Low-profile Visualized Intraluminal Support stent (LVIS; Microvention, Tustin, CA, USA), and five patients (10.6%) received an Enterprise stent (Codman Neurovascular, Raynham, MA, USA).
Of the 47 patients, thromboembolic events occurred in four patients (8.5%) during the follow-up period. However, there was no evidence that this correlated with the procedure, because these events occurred outside of the treated vessel in three out of four cases (75%). Only one event (2.1%) was correlated with the procedure. This patient was a 77-year-old woman, who underwent the procedure on an anterior communicating artery with an aneurysm 4.4 mm in size. Pre-PRU was 194, post-PRU was 131, and the PRU variability was 63. One-year follow-up MRI revealed a dot-like infarction in the left frontal lobe with no symptoms (Fig. 1). Therefore, there was no significant factor correlated with a thromboembolic event statistically, due to low incidence.
In addition, there was no major bleeding complication according to the TIMI bleeding criteria. There were only 18 (38.3%) minor bleeding complications. Sixteen patients (34%) had purpura and two patients (4.3%) had epistaxis and purpura. In univariate analysis correlated with hemorrhagic complication, pre-PRU and taking a statin were statistically significant (p=0.002 and p=0.016, respectively). In multivariate analysis, only pre-PRU was significantly correlated with hemorrhagic complication (p=0.033, OR 0.983, 95% CI 0.968–0.999) (Table 3).
DISCUSSION
Following procedural neuro-interventions, the most critical complication is thromboembolic infarction. Many studies have been done to prevent such complications12,26), but there are a few reports that demonstrate a correlation between the variability of antiplatelet response during antiplatelet drug medication and thromboembolic complications following a neurointervention4,6). Therefore, we investigated the variability of antiplatelet response during medication to determine any correlations with complications of endovascular treatment. In our study, PRU variability was greater in the Hypo-responsive group than in the Responsive group. This variability is in contrast to the initial clopidogrel response, as the Responsive group showed greater initial clopidogrel response. At one month after the clopidogrel regimen, overall clopidogrel response was slightly lower in the Hypo-responsive group than in the Responsive group. Whilst the initial clopidogrel response was significantly different between the two groups, it seems that due to a compensatory mechanism related to the higher PRU variability in the Hypo-responsive group, the final response rate was not significantly different (Fig. 2).
In the coronary intervention literature, some studies have reported on the etiology of variable responses to treatment with clopidogrel. Among affective factors correlated with clopidogrel response, the cytochrome P-450 enzyme was established to be the most important factor, correlated with genetic factors2). In addition, some studies have reported that drugs which are metabolized by the cytochrome P-450 enzyme, such as statins, PPIs, CCBs, β-blockers and/or SSRIs, affect clopidogrel response, as well as patient clinical factors such as DM, renal failure, and/or BMI7,15,18,19,30). In our study, pre-PRU and serum platelet count were important predictive factors, as in other studies10). In the Hypo-responsive group, the reason for higher PRU variability could be a higher potential for increased response, as the initial clopidogrel response was low. In other words, a higher number of initially unresponsive platelets, which account for the low initial clopidogrel response, could suggest a higher probability of platelets becoming reactive as the ratio of initially unresponsive platelets is higher in the Hypo-response group. In addition, serum platelets count has shown differences in the variability of clopidogrel response. It is possible that if there were higher initial platelet counts, it would result in a higher clopidogrel response rate, due to increased drug receptors.
In regard to any correlation with complications, thromboembolic complication was not correlated with pre-PRU, PRU variability, or post-PRU. However, many studies have reported that HTPR after one-week preparation is correlated with thromboembolic infarction. Kim et al.14) reported that acute thromboembolic infarction after endovascular treatment is a procedure-related complication. However, because delayed thromboembolic complication affects antiplatelet response, antiplatelet drug adjustment could decrease or delay thromboembolic complication in patients treated with SAC13). In the current study, it is thought that statistical significance was lost due to too small of a sample size, with only one thromboembolic infarction case.
In contrast, hemorrhagic complication was correlated with a few factors, demonstrated in 18 cases of hemorrhagic complications. Patients with high pre-PRU values had less frequent hemorrhagic complications. However, there was no correlation with PRU variability or post-PRU. It is thought that the difference between the two groups disappears because platelets that have not reacted are converted to react after sufficient medication. In addition, it can be predicted that although thromboembolic complication was not statistically significant in this study, this may be significant in a larger study. Therefore, determining clopidogrel response after early-phase treatment is important for preventing complications. If this were determined, no additional response test would be necessary.
Limitation
This study has several limitations. First, this study has a small sample size and is retrospective in nature. Therefore, there might be a selection bias or compounding effects, such as minimal, transient symptoms that may have been missed in this analysis. To reduce this limitation, we performed MRIs regularly after one year and compared them with MRIs taken one day after the procedure. In addition, in order to exclude thromboembolic infarction caused by the procedure, like cracking the atherosclerotic plaque and dislodging thrombi due to use of wires or catheters, MRI was performed one day after the procedure14). Second, among variables correlated with clopidogrel response, there was no significant factor correlated with thromboembolic complication, perhaps due to small sample size. However, since many studies have reported that the clopidogrel hypo-response is associated with thromboembolic infarction, our results would have been similar if the sample size of our study were larger8). In addition, because the results of hemorrhagic complication were similar to those of the previous studies, this assumption is further supported6). Third, of the factors associated with clopidogrel response reported in cardiology, only the pre-PRU and serum platelet counts were significant in the multivariate analysis. It is thought that this is also perhaps due to the small study size. Despite reduced statistical power, probably due to small sample size, pre-PRU and serum platelet count was prominently significant, indicating their statistical strength. In order to obtain optimal guidelines, well planned and larger prospective studies are necessary in the future.
CONCLUSION
Variability of clopidogrel response during the pre-operative medication regimen was related to the serum platelet count and the initial clopidogrel response. Thromboembolic and hemorrhagic complications were not related to the variability of clopidogrel response or clopidogrel response after one month of medication, but hemorrhagic complications were associated with the initial clopidogrel response. Therefore, patients undergoing SAC with UIA are recommended to be tested for initial clopidogrel response; no further tests would be necessary.
XML Download