Clinical Implications of Changes in Individual Platelet Reactivity to Aspirin Over Time in Acute Ischemic Stroke
Background and Purpose—Time-dependent changes in individual platelet reactivity have been detected in patients with coronary artery disease. Therefore, we sought to evaluate the time-dependent changes in platelet reactivity to aspirin during the acute stage after ischemic stroke and the clinical implications of variable patient responses to aspirin in acute ischemic stroke.
Methods—We conducted a single-center, prospective, observational study. The acute aspirin reaction unit (ARU) was measured after 3 hours of aspirin loading, with higher values indicating increased platelet reactivity despite aspirin therapy. The follow-up ARU was measured on the fifth day of consecutive aspirin intake. The numeric difference between the follow-up ARU and the acute ARU was defined as ΔARU and was stratified into quartiles. Early neurological deterioration was regarded as an early clinical outcome.
Results—Both the acute ARU (476±69 IU) and the follow-up ARU (451±68 IU) were measured in 349 patients in this study. Early neurological deterioration was observed in 72 patients (20.6%). Changes in aspirin platelet reactivity over time showed an approximately Gaussian distribution. The highest ΔARU quartile was independently associated with early neurological deterioration (odds ratio, 3.19; 95% confidence interval, 1.43–7.10; P=0.005) by multivariate logistic regression analysis.
Conclusions—The results of our study showed that the increase in platelet reactivity to aspirin over time is independently associated with early neurological deterioration in patients with acute ischemic stroke. In addition, during the acute stage of ischemic stroke, serial platelet reactivity assays may be more useful than a single assay for identifying the clinical implications of aspirin platelet reactivity after ischemic stroke.
Long-term aspirin use clearly reduces the risk of vascular events in patients with cardiovascular disease.1 However, studies have demonstrated the variability in the pharmacodynamic response to aspirin.2 High on-aspirin platelet reactivity (HAPR) has been detected in 5% to 45% of patients with cardiovascular disease, and this parameter may be associated with an increased risk of vascular events.3–7 However, contrasting results suggest that a limited biological response to aspirin does not correlate with clinical aspirin resistance.8,9 Platelet hyper-reactivity despite aspirin administration may be a frequent phenomenon in patients with acute ischemic stroke.10
Recently, time-dependent variability in platelet reactivity to clopidogrel has been detected in individuals with coronary artery disease.11 Thus, measuring platelet function at a single time point may not be sufficient to determine platelet reactivity in individual patients with acute ischemic stroke. Considering that the risk of recurrent stroke is the highest during the acute phase,12 it is important to investigate the influence of platelet reactivity on early outcomes of acute ischemic stroke. Particularly, serial platelet reactivity assays may be useful for evaluating the association between early outcomes and platelet reactivity during unstable periods of acute ischemic stroke.
Therefore, we sought to evaluate the individual variability in platelet reactivity of patients on aspirin during the acute stage after ischemic stroke and to determine the clinical implications of this variability in platelet reactivity in acute ischemic stroke.
This study was a single-center, prospective, observational study. The patients in this study were consecutively recruited from the Cerebrovascular Center of Chonnam National University Hospital between April 2012 and May 2013. The methods of this study were similarly described in the previous study because the same cohort was used but for a clearly different purpose and with different methods.13 Briefly, patients were included according to the following criteria: (1) presented and were evaluated within 3 days of symptom onset; (2) exhibited positive lesions on diffusion-weighted imaging; (3) were not at high risk of cardioembolism; and (4) provided written informed consent. The exclusion criteria included other etiology according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification, thrombolysis, malignant infarction, early loss to follow-up, chronic use of nonsteroidal anti-inflammatory drugs (>3 days per week for the past 3 months), history of recent hemorrhagic disorder within the past 4 weeks, coagulopathy, thrombocytopenia (<90 000 platelets/μL), low hematocrit (<29%), and chronic liver or renal disease.
This study was approved by the Institutional Review Board of Chonnam National University Hospital. Written informed consent was obtained from all the participants or their caregivers.
Demographic, clinical, and laboratory data were prospectively collected by dedicated research nurses or physicians. The following stroke risk factors were identified: age, sex, hypertension, diabetes mellitus (DM), dyslipidemia, current smoking, and a previous history of stroke or transient ischemic attack. Baseline data collected from all the patients included the National Institutes of Health Stroke Scale score and the stroke subtype, which was stratified according to the TOAST classification after complete diagnostic profiling. The National Institutes of Health Stroke Scale scores were assessed at admission and on each day of hospitalization by well-trained, dedicated stroke nurses.
Aspirin was administered to the study patients immediately after brain magnetic resonance imaging. The use of enteric-coated aspirin or intravenous antithrombotics, such as glycoprotein IIb/IIIa inhibitor, was not permitted in the study. The initial loading dose was 300 mg for all patients, followed by maintenance doses of 100 mg daily. We checked the daily medication administration during the study period. A combination of aspirin and other antiplatelet agents was administered at the discretion of the treating physicians to patients with symptomatic arterial steno-occlusion, a previous history of coronary artery disease, or previous antiplatelet treatment. The patients whose mechanism of stroke was identified as a cardioembolism were prescribed anticoagulants instead of aspirin and were not included in the study.
Platelet Function Study
The aspirin reaction unit (ARU) was measured using VerifyNow (Accumetrics, San Diego, CA). This system uses turbidimetric optical detection of platelet aggregation in whole blood with a system containing fibrinogen-coated beads. The test cartridges contain fibrinogen-coated microbeads and arachidonic acid. On activation by arachidonic acid, the platelets bind to the microbeads and aggregate. As aggregation occurs, light transmission through the sample increases.
The acute ARU (aARU) was measured after 3 hours of aspirin loading in the emergency department. The follow-up ARU (fuARU) was measured on the fifth consecutive day of aspirin intake. Both the physicians and the patients were blinded to the ARU measurement results. The aspirin nonresponder status, or HAPR, was defined as an ARU of ≥550 IU, as is commonly used. The difference between the fuARU and aARU was defined as ΔARU (fuARU minus aARU) for a given participant. The ΔARU values were stratified into quartiles: the lowest quartile included ΔARU values below <−70 IU; the second quartile ranged from −70 to −18 IU; the third quartile ranged from −18 to 16 IU; and the fourth quartile included ΔARU values >16 IU.
Early neurological deterioration (END) was regarded as an early clinical outcome in this study and was defined as an increase in the National Institutes of Health Stroke Scale score of ≥1 point (or the development of new neurological symptoms, indicating that a subcomponent of the scale that was previously scored 0 was subsequently scored ≥1 point) between admission and day 5. For sensitivity analysis, END4 was defined as an increase in the National Institutes of Health Stroke Scale score of ≥4 points. In addition, END was divided into the following 3 categories: progressive END, END with new lesions, and other causes (Method I in the online-only Data Supplement).
Detailed descriptions of the analysis are shown in Method II in the online-only Data Supplement. Briefly, we compared the aARU and fuARU values using paired t tests. In addition, we compared the baseline characteristics between the patients in each ΔARU quartile, between the patients with HAPR and those with normal on-aspirin platelet reactivity (NAPR), and between the patients with and without END. Multivariate logistic regression analysis was used to evaluate independent factors associated with END, which were adjusted for the variables with P<0.1 based on univariate analysis or for the clinically significant variables. Further analysis was performed to investigate the association between END and the ΔARU quartiles in patients treated with aspirin monotherapy. Odds ratios and 95% confidence intervals were calculated. A P value of <0.05 was considered to be statistically significant. All statistical analyses were performed using SPSS for Windows version 17 (SPSS Inc, Chicago, IL).
A total of 1027 patients with acute ischemic stroke were screened within 3 days of symptom onset during the study period. Of these patients, 620 were initially excluded based on the exclusion criteria, and after providing informed consent, 58 additional patients were excluded (Figure I in the online-only Data Supplement). Ultimately, 349 patients (men, 211; mean age, 65.7±11.3 years) who underwent both aARU and fuARU measurements were included in the study.
END and END4 were observed in 72 (20.6%) and 59 patients (16.9%), respectively. Table 1 presents the general characteristics of the patients with and without END. The frequencies of DM and relevant arterial steno-occlusion were significantly higher in the patients with END than in those without END. In addition, the use of dual antiplatelet therapy after the aARU measurement was more frequent in the patients with END than in those without END; however, this difference was not statistically significant (34.7% versus 23.8%; P=0.071).
ARU Measurement Results
The ARU measurements revealed a mean aARU of 476.3±69.16 IU and a mean fuARU of 451.3±68.46 IU (P<0.001 by paired t test). HAPR was detected in 58 patients (16.6%) (Table I in the online-only Data Supplement). In addition, patients with HAPR, as determined by the aARU, less frequently had DM and relevant arterial steno-occlusion than those without HAPR; however, these differences were not statistically significant. Individual patient analyses indicated that 39 (67.2%) of the 58 patients with HAPR based on the aARU value converted to NAPR based on the fuARU value. In contrast, 28 (9.6%) of the 291 patients with NAPR based on the aARU value converted to HAPR based on the fuARU value (Table 2). The ΔARU was not significantly different between patients receiving monotherapy (n=258) and those receiving dual therapy (n=91; mean ΔARU: −21.7±83.8 versus −34.3±80.4; P=0.212). Because even minor changes in platelet reactivity alter the response status (ie, a change in the ARU from 549–551 IU), the ARU was also analyzed as a continuous variable. Changes in platelet reactivity over time showed an approximately Gaussian distribution, with similar changes in both directions (Figure 1).
The ΔARU quartile characteristics are shown in Table II in the online-only Data Supplement. The numbers of patients with DM and previous antiplatelet users were greater in the higher ΔARU quartile than in the lower ΔARU quartile. In contrast, the number of patients receiving dual antiplatelet therapy after the aARU measurement was nonsignificantly greater in the lower than in the higher quartile. In addition, there was no difference in weight among the ΔARU quartiles.
END and the ARUs
END was more frequently observed in the patients with NAPR than in those with HAPR based on the aARU but not on the fuARU (Table 3). However, the patients with END and END4 exhibited higher ΔARU values than those without each respective outcome. Furthermore, END and END4 were more frequently observed in the patients in the fourth quartile based on ΔARU than in patients in the other quartiles (Table 3). In addition, END with new lesions was more frequently observed in the patients in the fourth quartile based on ΔARU than in patients in the other quartiles (Figure 2). However, progressive END was not significantly associated with ΔARU, although a positive trend was observed.
Multivariate logistic regression analysis showed that the fourth quartile ΔARU was significantly associated with END compared with the first quartile. The adjusted odds ratio (95% confidence interval) was 3.16 (1.42–7.02) for all subjects and was 3.36 (1.22–9.31) for the patients treated with aspirin monotherapy (Table 4). In addition, the fourth ΔARU quartile was independently associated with END4 (Table III in the online-only Data Supplement).
Time-dependent individual variability in ARU, as measured by the VerifyNow assay, was detected between aspirin loading and the fifth day of consecutive aspirin administration among patients with acute ischemic stroke. More than 60% of patients with HAPR based on the aARU exhibited NAPR on the fifth day of consecutive aspirin intake. In addition, in patients with acute ischemic stroke, END was independently associated with a time-dependent increase in the ARU, especially values in the highest quartile, but not with HAPR, based on the aARU alone. Therefore, the findings of this study suggest that serial ARU measurements, but not a single ARU measurement, can be helpful in predicting END in patients with acute ischemic stroke. Notably, this is the first study to evaluate the clinical implications of ΔARU in acute ischemic stroke.
In our study, on an individual basis, a significant number of patients exhibited discordant ARUs, as >65% of patients with HAPR exhibited a change in their aspirin responder status after 5 days of aspirin administration. In contrast, only ≈10% of patients with NAPR based on the aARU exhibited a nonresponder status based on the fuARU. These results are concordant with those of a recent study reporting variable responses to clopidogrel among patients with coronary artery disease.11 In healthy subjects, the mean and median VerifyNow values did not significantly differ when participants were tested 20 to 24 hours after the final dose of 80 mg to 325 mg of aspirin.14 However, in contrast to healthy volunteers, patient populations at risk for arterial thrombosis have pathophysiologies that evoke variable intensities and mechanisms to stimulate platelets. Accordingly, our study provides supportive evidence that variable responses to aspirin could occur in acute ischemic stroke.
Because even minor changes in platelet reactivity alter the aspirin response status (ie, a change in the ARU from 549–551 IU), the cut-off value of 550 IU for HAPR may result in the inconsistent identification of clinical significance during the acute stage of ischemic stroke. Compared with HAPR, ΔARU is more consistently associated with END among patients with acute ischemic stroke. The increase in the ARU after aspirin intake, as assessed by serial platelet function assays, may have resulted in clinical instability or deterioration in patients with acute ischemic stroke. Briefly, the lack of a decrease in platelet reactivity during the first several days after aspirin intake might represent aspirin nonresponsiveness during acute ischemic stroke. However, because of our study design, we cannot exclude the possibility that this pattern of increasing ARU for serial measurements may be an epiphenomenon of END in patients with acute ischemic stroke. A previous study reported that HAPR was related to lesion volume and clinical severity in patients with acute ischemic stroke.15 However, we did not detect a correlation between initial stroke severity and the aARU.
In addition, some END might occur irrespective of aspirin responsiveness. Briefly, the laboratory response to arachidonic acid might not explain all cases of END. Therefore, we attempted to evaluate the different mechanisms of END, although our investigation might be incomplete. As shown in Figure 2, overall END and END with new lesions were significantly associated with the high ΔARU quartiles. However, progressive END was not significantly associated with ΔARU values, although a positive trend was observed. Therefore, the different END mechanisms seem to be differently influenced by ΔARU. However, because the END mechanisms have not clearly defined thus far, further study is warranted.
A recent meta-analysis has demonstrated that patients with HAPR are at high risk for cardiovascular events.16 Although various platelet reactivity assays have not been considered accurate or consistent,17 in separate HAPR analyses using different tests, the results of 12 tests have independently shown that patients with HAPR exhibit a significantly increased risk for cardiovascular events, particularly coronary artery disease, compared with patients with NAPR.16 However, our study focused on the acute period of ischemic stroke. Few studies have focused on this period or on the platelet reactivity after aspirin administration during the first several days after acute ischemic stroke.18 The findings of our study are novel and relevant to previously unaddressed and unresolved issues.
Although the effects of aspirin begin promptly after aspirin loading, END is observed in ≈10% to 45% of patients with acute ischemic stroke. Previous studies have shown that HAPR, as determined by a single ARU measurement, may be associated with an increased risk of END and early recurrent ischemic lesions.19,20 However, this study should be cautiously interpreted because of its retrospective design and small sample size. In our study, patients with NAPR based on the aARU exhibited a higher incidence of END than those with HAPR. This finding may have resulted from an imbalance in baseline characteristics, such as a higher frequency of DM in the patients with NAPR than in those with HAPR (Table I in the online-only Data Supplement). After adjustment for covariates, HAPR based on the aARU was not independently associated with END. A single platelet reactivity measurement during the acute phase of ischemic stroke may be unreliable because of several confounding physiological/clinical factors. In addition, previous studies have shown that noncompliance, high body weight, and enteric-coated aspirin administration were associated with aspirin resistance.21–23 Our study could appropriately control for these confounders through its acute and in-hospital study design, and few included patients had a high body weight (>90 kg). However, another study of our cohort revealed that HAPR based on the aARU is independently associated with new ischemic lesions on follow-up diffusion-weighted imaging but not with END.13 Therefore, whether a single routine platelet reactivity assay after aspirin loading or during the acute period should be performed to assess patients with acute ischemic stroke remains unclear.17 Because patients in the acute stage after ischemic stroke may be more unstable and may exhibit more variable platelet function in response to antiplatelet therapy than those in the chronic or stable stage, serial platelet reactivity assays are more useful than a single assay for identifying the clinical influences of platelet reactivity after acute ischemic stroke. Furthermore, although not performed in our study, regular platelet function measurements could identify patients who are noncompliant, allowing an intervention to improve compliance to be conducted.
Previous studies have demonstrated that HAPR is associated with increased clinical severity and stroke infarct volume in patients with acute stroke who have previously received aspirin therapy.15 However, our study differed from other studies that enrolled patients who had previously received aspirin therapy. For example, the percentage of previous antiplatelet users was greater for the patients in the fourth ΔARU quartile than for patients in the other ΔARU quartiles. We hypothesized that the patients who experienced a new ischemic stroke while taking aspirin may have been experiencing aspirin failure or clinical aspirin resistance and that the lack of a decrease in ΔARU even after aspirin administration may suggest poor aspirin responsiveness. The results of our study suggested that patients with antiplatelet failure, typically involving aspirin, may have reduced therapeutic effects or delayed responses to aspirin during the first 5 days of treatment. Further prospective studies using a longer follow-up duration are required to confirm this hypothesis.
This study has several limitations. We analyzed platelet reactivity and outcomes during only the first 5 days after aspirin loading. Therefore, the individual variability in ARU over time could not be assessed in patients with ischemic stroke during the chronic or stable stages. However, although our study was not a randomized, controlled trial, because of the time limitation, we could control for several potential confounds, such as drug compliance, dose timing, and drug–drug interactions. Second, the use of the VerifyNow assay could have limitations and the selection of cut-off value for the ΔARU seemed to be highly arbitrary because these values were determined based on quartiles. The cut-off ARU value of 550 IU for aspirin resistance was not consistent in acute ischemic stroke. Furthermore, platelet function tests for aspirin resistance have not been standardized and they do not correlate well with each other.24,25 Accordingly, it seems to be highly test-specific for laboratory diagnosis of aspirin resistance. Our results suggest that the drawbacks of the single measurement performed using the VerifyNow assay could be reduced by conducting serial measurements in acute ischemic stroke. Furthermore, this study was conducted to investigate the clinical implications of ΔARU in acute ischemic stroke. Further investigation is required to assess the factors associated with the time-dependent changes in platelet reactivity to aspirin in patients with acute ischemic stroke. The lack of a consistent protocol for the use of antiplatelet agents was another limitation of our study. Although 3 reasonable initial treatment options were available for the secondary prevention of ischemic stroke, the guidelines stated that the selection of an antiplatelet agent should be individualized based on the characteristics of each patient.26 Moreover, sensitivity analysis of patients treated with aspirin monotherapy indicated that the fourth ΔARU quartile was significantly associated with END. In addition, our study has limitations that are inherent to single-center studies with relatively small sample sizes. Because of the ethnic characteristics of the patients, only 4 patients had a high body weight (>90 kg), which might be associated with aspirin resistance.21 Therefore, there is a lack of generalizability between the genetically disparate Korean population and other ethnic populations evaluated in other studies.
In conclusion, the results of our study showed that the time-dependent increase in the ARU was independently associated with END in patients with acute ischemic stroke. In addition, during the acute stage of ischemic stroke, serial platelet reactivity assays seem to be more useful than a single assay for identifying the clinical implications of platelet reactivity after ischemic stroke. However, as platelet function tests, including the VerifyNow assay, do not correlate well with each other, further study will be needed to confirm our results.
Sources of Funding
This study was supported by a grant (CRI 13041-21) from Chonnam National University Hospital Biomedical Research Institute. This study was supported by a grant of the Korea Healthcare technology R&D Project, Ministry of Health and Welfare, Republic of Korea (HI10C2020).
Guest Editor for this article was Louis Caplan, MD.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.115.009428/-/DC1.
- Received March 13, 2015.
- Revision received June 16, 2015.
- Accepted June 22, 2015.
- © 2015 American Heart Association, Inc.
- 1.↵Antiplatelet Trialists’ Collaboration. Collaborative overview of randomised trials of antiplatelet therapy–I: prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients. BMJ. 1994;308:81–106.
- Alberts MJ.
- Snoep JD,
- Hovens MM,
- Eikenboom JC,
- van der Bom JG,
- Huisman MV.
- Krasopoulos G,
- Brister SJ,
- Beattie WS,
- Buchanan MR.
- Karon BS,
- Wockenfus A,
- Scott R,
- Hartman SJ,
- McConnell JP,
- Santrach PJ,
- et al
- Eikelboom JW,
- Emery J,
- Hankey GJ.
- Richard S,
- Toussaint-Hacquard M,
- Fay R,
- Lacour JC,
- Ducrocq X,
- Lecompte T.
- Cox D,
- Maree AO,
- Dooley M,
- Conroy R,
- Byrne MF,
- Fitzgerald DJ.
- Harrison P,
- Segal H,
- Blasbery K,
- Furtado C,
- Silver L,
- Rothwell PM.
- Kernan WN,
- Ovbiagele B,
- Black HR,
- Bravata DM,
- Chimowitz MI,
- Ezekowitz MD,
- et al