Assessment of platelet function in patients with stroke using multiple electrode platelet aggregometry: a prospective observational study

Platelets play a major role in arterial thrombus formation and therefore in the pathophysiology of ischaemic stroke [1‐4]. Excessive platelet activation leads to increased thrombin generation and potentially abnormal thrombus formation [5]. Hence the importance of oral antiplatelet drugs, which are the mainstream therapy in the primary and secondary prevention of cerebrovascular disease [6, 7]. Currently, the antiplatelet (AP) drugs most widely used are aspirin and clopidogrel, and multiple electrode platelet aggregometry (MEA) has been used to study their effects on platelet function [8‐11]. MEA has also been used to investigate the effects of non-opioid analgesics [12], anticoagulants [13], antifibrinolytics [14] and temperature [15] on platelet aggregation.

Despite the effectiveness of aspirin in the primary and secondary prevention of atherothrombotic disease, patients continue to suffer recurrent thromboembolic vascular events whilst on AP treatment [16]. This recurrence has been associated with high on-treatment platelet reactivity (HPR), with up to more than 60% of subjects being reported to be resistant to antiplatelet therapy (aspirin or clopidogrel) [17, 18]. Although some studies showed a link between HPR and major adverse vascular events [19, 20], the use of platelet function analysis to detect and manage HPR continues to be debated. Most studies that investigated HPR were undertaken for patients already receiving antiplatelet therapy without assessing baseline platelet reactivity [19, 20], which may partly explain the conflicting results. We were therefore interested in whether patients have a higher platelet reactivity at baseline, which may contribute to the suboptimal response to treatment. Hence this study aimed to compare platelet function, as determined by MEA, in healthy subjects and stroke patients prior to treatment initiation and between patients naive to AP therapy and those on regular low dose aspirin or clopidogrel.

One hundred twenty two patients with suspected first-ever ischaemic stroke were recruited between May 2012 and February 2014. 52 patients were excluded. The reason for exclusion included: 18 had stroke mimics such as Bell’s palsy; sepsis and brain tumour; 8 transient ischemic attacks (TIA); 8 had a haemorrhagic stroke; 4 previous strokes; 3 cancers; 1 renal failure; 6 found to have been given loading aspirin prior to blood collection and 4 were taking NSAIDs such as Ibuprofen and Diclofenac. Leaving seventy patients (42 men and 28 women; mean age, 73.1 ± 13.3 years) with confirmed first time ischaemic stroke, these patients were used in the analysis. Their baseline characteristics are presented in Table 1. Stroke patients were matched for sex and age with 72 healthy subjects (mean age, 71.3 ± 6.9 years, p = 0.29; 40 men and 32 women, 0.59) recruited from the same population as the stroke patients. Whilst matched for age and sex the stroke group did have a significantly higher proportion of current smokers than healthy controls (22.9% versus 6.9%, p = 0.01).

Full blood count and coagulation profile results are presented in Table 2 for all groups. Of the 70 patients, 31 were thrombolysed and 39 were loaded with 300 mg aspirin. At baseline 33 stroke patients were on antiplatelet therapy (26 on 75 mg of aspirin only, 3 on 75 mg of clopidogrel only and 4 on both agents) and 37 were AP-naive.

Multiplate® analysis showed that ASPI was the only significant parameter that differed between the healthy subjects and the entire stroke group (97.6 ± 27.3 versus 81.1 ± 50.1, p = 0.02). These differences reversed when healthy subjects were only compared with AP-naive patients (97.6 ± 27.3 versus 117.4 ± 30.7, p < 0.005). In the AP-naive patients, significantly higher values were also observed for ADP and Collagen as shown in Table 3. All MEA parameters were significantly lower in patients who were on regular antiplatelet therapy at baseline as compared to healthy subjects and AP-naive patients (Table 3). All these differences continued to be significant after controlling for smoking status using ANCOVA.

Our study is the first to assess platelet reactivity using MEA in patients with stroke at baseline before loading doses of 300 mg aspirin were given. The main finding showed that platelet aggregation in AP-naive patients is higher than in healthy subjects for all pathways investigated. However, it is difficult to ascertain if this is due to an underlying high platelet reactivity or is secondary to the acute event. Future larger studies are needed to confirm our findings and investigate this assumption.

Secondly, our study detected changes between long-term aspirin users and aspirin-naive patients. Despite the proved effectiveness of AP therapy, patients continue to have recurrent vascular events while on treatment [16]. In agreement with previously reported studies [25], about 50% of patients in our study were taking one or more AP drugs at presentation but this did not prevent stroke. New approaches to improve clinical outcomes used MEA to individualise antiplatelet therapy [19] and hence highlighting the importance of our findings. Based on definition of aspirin resistance in previous studies, 33.3% of patients would be defined as aspirin-resistant according to ASPI results. As more than 60% of long term aspirin users are considered responders, the use of loading dose of aspirin in these patients may not have any additional benefit. It may indicate the need to investigate thoroughly for underlying silent risk factors such as paroxysmal atrial fibrillation. The emergence of new P2Y12 antagonists and ongoing uncertainties of platelet function testing dictate the need for larger RCT to answer these questions and investigate the clinical benefit of AP switching strategy in stroke patients based on MEA results. This supports the recommendation by the National Institute for Health and Care Excellence for future research in this area about effectiveness and safety of clopidogrel, alone or in combination with aspirin, in treating stroke and TIA [26].

Thirdly, our study also detected the effect of aspirin loading therapy on platelet aggregation (ASPI) in acute stroke. Interestingly at 3–5 days after in hospital administration of aspirin, only 11.1% were non or low responders. Therefore, there is a discrepancy of more than 20% in presumed aspirin-resistance rates between self-reported and in hospital aspirin intake. Many factors may explain this discrepancy including poor compliance, timing of testing following aspirin intake and low-dose versus high-dose aspirin. Up to 50% of patients are reported to be poorly compliant [27, 28]. Although many strategies are in place to encourage the public to recognise and act on stroke symptoms, little has been done to improve compliance. Previous studies suggested female sex, single marital status, lower education level, depression, diabetes mellitus, polypharmacy, and smoking as predictors of poor compliance [27]. Hence, more educational initiatives targeting these high-risk patients of poor compliance should be done to enforce the importance of taking AP regularly. Another possible explanation of this discrepancy is the timing of baseline sample after aspirin intake, which was variable, whereas it was 2–4 h after administering aspirin for subsequent samples. Timing of platelet function testing has been acknowledged as a factor explaining the variability in the results with serum salicylate level peaking at 1 h after administration [5, 29]. The higher aspirin dose given acutely in hospital is another potential factor but published data do not support this assumption reporting that doses between 30 and 1300 mg per day have the same clinical efficacy [18, 30]. Nonetheless some may argue that this does not equate to similar platelet inhibition as detected by platelet function tests.

Despite the availability of a wide-range of techniques to investigate platelet function, their reliability has been questioned. A previous study showed poor correlation between six platelet function tests to define aspirin-resistance [31], and this variation in methodology may be a factor in explaining the difficulties in determining the true prevalence of AP resistance. Hence it is important to link these tests to clinical endpoints to establish cut-offs of each assay. Although LTA may be considered the gold standard for the determination of platelet activity, we chose to use Multiplate® in this study because it is a standardized fully automated point-of-care test that is relatively easy to use without many of the technical limitations of the previous assays [32]. Additionally the feasibility of using Multiplate® in the context of routine hospital service has been ascertained to provide reproducible and precise results in stroke patients [33]. This potentially makes our findings easier to replicate and translate into large clinical studies to investigate clinical endpoints.

Limitations: this is a small single centre study and hence may not be generalizable to other populations. Aspirin and clopidogrel intake at baseline was self-reported by patients or their family. Although MEA results indicate two thirds of the patients were taking aspirin, confirming compliance is an important issue but was beyond the scope of this study. Another limitation is the higher number of active smokers in the stroke group as compared to healthy subjects. Conflicting results exist about the effect of smoking on platelet function testing [20, 24, 34]. However in our study, ANCOVA analysis was used to control for the effect of smoking. Another limitation is the attrition rate despite the prospective study design. Multiplate analysis at 3–5 days was not possible for all cases for various reasons (patient discharged or transferred to another hospital, death, declined further samples, difficult to bleed). Despite the strict inclusion and exclusion criteria heterogeneity of stroke presentations and patient populations are key issues to consider when investigating stroke as noticed in this study, and subgroup analysis should be interpreted with caution.

Despite these limitations our study reports a higher platelet reactivity as determined by Multiplate® in AP-naive stroke patients for all pathways as compared to healthy subjects. Many patients receiving low dose aspirin met the criteria of HPR or drug resistance but the prevalence was much lower at 3–5 days of inpatient treatment. It also reiterates the importance of compliance and probably timing of testing as potential factors affecting MEA testing. It also questions the benefits of giving aspirin in patients with laboratory evidence of response when this did not prevent recurrent stroke. Hence further work in large prospective multicentre trials is needed to confirm that the laboratory results can be translated onto clinical endpoints. For any future research the main question is whether the use of MEA to individualise AP treatment in stroke/TIA results in better clinical outcomes. We recommend establishing platelet reactivity at baseline to determine best AP therapy and then monitor their effect at 3 days to decide whether a switch is warranted. This may also help establish the true causes of HPR in stroke and the value of MEA in predicting future events.