Available online at www.sciencedirect.com

www.elsevier.com/locate/jocn

Journal of Clinical Neuroscience 15 (2008) 1204–1209

Clinical Study

A pilot study of resistance to aspirin in stroke patients

Daniel Bennett, Bernard Yan *, Lachlan MacGregor, David Eccleston, Stephen M. Davis

Departments of Neurology, Royal Melbourne Hospital and University of Melbourne, Grattan Street, Parkville, Victoria 3050, Australia

Received 31 August 2007; accepted 10 January 2008

Abstract

Aspirin resistance has been shown to be a significant risk factor for recurrent cardiovascular ischaemic events. However, there are alack of data correlating aspirin resistance and risk of cerebrovascular ischaemic events. This pilot study aimed to determine the prev-alence of aspirin resistance in an Australian stroke population and to correlate aspirin resistance with an increased risk of ischaemicstroke. Fifty patients treated with aspirin for 2 years were tested for aspirin resistance using the Ultegra Rapid Platelet FunctionAssay (Accumetrics, San Diego, CA, USA) on admission to Royal Melbourne Hospital for ischaemic stroke. The 2-year historyof ischaemic stroke and transient ischaemic attack (TIA) were assessed. Prevalence of aspirin resistance among our patients was30%. Univariate analysis suggested a non-significant trend towards increased rate of previous ischaemic stroke or TIA and aspirinresistance (odds ratio, OR = 3.88; 95% confidence interval 0.54–29.87; p = 0.18). This study shows that aspirin resistance is prevalentwithin the Australian ischaemic stroke population.Crown Copyright � 2008 Published by Elsevier Ltd. All rights reserved.

Keywords: Aspirin resistance; Ischaemic stroke; Transient ischaemic attack

1. Introduction

Aspirin remains the standard first-line therapy in thesecondary prevention of vascular events due to underlyingatherothrombotic conditions. The Antithrombotic Trial-ists’ Collaboration analysed patient data from 21 random-ised controlled studies and concluded that the benefit ofaspirin reduced the recurrent odds of stroke by 22% com-pared to placebo.1 However, the benefits in a stroke popu-lation are even more modest, with a relative risk reductionof only 13%.2 Therefore, a substantial proportion of pa-tients do not derive the expected benefits of strokeprevention.

Between 5% and 60% of patients treated with aspirin maybe regarded as aspirin-resistant, on various laboratory mea-sures of platelet function. This has been independently asso-ciated with an increased risk of atherothrombotic vascular

0967-5868/$ - see front matter Crown Copyright � 2008 Published by Elsevie

doi:10.1016/j.jocn.2008.01.006

* Corresponding author. Tel.: +61 3 9342 8448; fax: +61 3 9342 8427.E-mail address: bernard.yan@mh.org.au (B. Yan).

events in a wide range of cardiovascular patients.3–6 Aspirinresistance has been associated with a 4 times increased haz-ard risk (HR) of adverse events (HR = 4.1 [1.4–12.1])among stable cardiovascular patients.3 The increased riskof clinical events is expected to be similarly linked with aspi-rin resistance among ischaemic stroke patients, given thecommon underlying pathology of atherothrombosis. Ifaspirin resistance predicts an increased risk of stroke recur-rence, treatment could be individualised.

Few studies have investigated the role of anti-plateletresistance in stroke. Harrison et al. recruited 100 stroke pa-tients and showed that the incidence of aspirin resistancewas 17%, using light transmission aggregometry (LTA).7

This was comparable to the incidence in studies of myocar-dial ischaemia.8 However, this study did not address recur-rent stroke. The primary objective of our pilot study was toexamine the prevalence of aspirin resistance in Australianpatients who present with acute ischaemic stroke. The sec-ondary objective was to test the hypothesis that aspirinresistance was more common in those with previous cere-brovascular events.

r Ltd. All rights reserved.

D. Bennett et al. / Journal of Clinical Neuroscience 15 (2008) 1204–1209 1205

2. Materials and methods

2.1. Patients and study design

This study was approved by the Human Research andEthics Committee (HREC) at the Royal Melbourne Hospi-tal (Melbourne, Victoria) and all subjects provided writteninformed consent. Over a 7 month period patients present-ing consecutively to Royal Melbourne Hospital Compre-hensive Stroke Centre with acute ischaemic stroke werescreened for enrolment. Inclusion criteria were consentingage >18 and patients who had been taking aspirin for atleast 2 years and had presented with an acute ischaemicstroke (excluding transient ischaemic attack, TIA). Patientswere excluded if they had an intracerebral haemorrhage onCT scan, significant diseases or abnormalities (terminalcancer, life threatening infection) that might compromiseparticipation. Patients treated with clopidogrel in combina-tion with aspirin were also excluded.

All patients received standard treatment according toRoyal Melbourne Hospital Comprehensive Stroke Centreprotocol. After a single screening visit to establish eligibil-ity, qualified patients were enrolled in the study. Demo-graphic information (age, gender, race, vascular riskfactors) and clinical data which included National Insti-tutes of Health Stroke Scale (NIHSS), Trial of Org 10172in Acute Stroke Treatment (TOAST) scale9 and Oxford-shire Community Stroke Project Classification10 were ob-tained within 48 hours of admission. Any history of priorischaemic events was analysed retrospectively for the previ-ous 2 years while taking aspirin and included ischaemicstroke, TIA, myocardial infarction and unstable angina.A single blood sample was obtained per recruited patientto establish the presence of aspirin resistance.

2.2. Blood sampling

Whole blood was obtained via venepuncture into a 2 mLpartial fill 3.2% sodium citrate vacuum collection tubeusing a 21-gauge needle. A tourniquet was used while per-forming venepuncture. For patients with an in-dwellingcannula, blood was collected after sufficient discard (about5 mL) had been drawn to clear the line. The tubes werethen gently inverted 3–5 times and incubated at room tem-perature for at least 30 min after collection before testingbut for no longer than 4 hours. Blood was taken no lessthen 2 hours after aspirin ingestion as documented on wardcharts.

2.3. Measuring aspirin resistance

Aspirin-induced platelet inhibition was measured usingthe point-of-care device, the Ultegra Rapid PlateletFunction Assay (RPFA) developed by Accumetrics (SanDiego, CA, USA). This turbidimetric-based optical detec-tion system measures platelet induced aggregation propor-tional to changes in light transmission. Modified

disposable cartridges containing fibrinogen-coated beadsand platelet agonist are inserted into the Ultegra RPFAdevice to test specifically for aspirin responsiveness. A cit-rate anti-coagulated blood sample is then gently inverted4–5 times and placed onto the cartridge. The deviceautomatically dispenses blood, mixing it with the car-tridge reagents, thereby no blood handling is required.Fibrinogen-coated beads agglutinate in proportion tothe number of activated platelets expressing glycoproteinGP IIb/IIIa receptors, subsequently increasing light trans-mission through the sample. The result is expressed inAspirin Reaction Units (ARU), taking about 5 min to testone sample of blood. An ARU value P550 is consistentwith no platelet dysfunction whereas values <550 are con-sistent with platelet dysfunction. Aspirin resistance basedon the Ultegra RPFA is defined as an ARU P 550 in apatient taking aspirin. The threshold value of 550 ARUwas determined by correlation with adrenaline-inducedlight transmission aggregometry in aspirin-naive patientstested before and then between 2 hours and 30 hours afteraspirin ingestion.11

Patients received 100 mg aspirin dose as is standardmanagement in the Royal Melbourne Hospital Compre-hensive Stroke Centre. Aspirin resistance was recorded asa dichotomous variable, all patients with ARU value<550 and known to be taking aspirin were regarded asaspirin sensitive.

2.4. Analyses

All statistical analyses were performed using STATA 9(StataCorp LP., College Station, TX, USA). Continuousvariables are expressed as median and range. Dichotomousvariables are presented as percentages. The relative fre-quency of dichotomous baseline and clinical variablesamong aspirin-resistant, compared to aspirin-sensitive, pa-tients was analysed using odds ratios (OR) and their 95%confidence intervals (CIs) with the Fisher’s exact test calcu-lated. Logistic regression was used for the continuous var-iable of age to assess association with aspirin resistance.Statistical significance was defined as p < 0.05. However,this pilot study does not have sufficient power to providestatistically significant results because it was designed toprovide indications for future research.

3. Results

We studied 50 ischaemic stroke patients admitted to theRoyal Melbourne Hospital Comprehensive Stroke Centreduring the recruitment period (Table 1). The median ageof patients was 77 [51.4–93.6], 74% were male and 98%were Caucasian.

3.1. Prevalence of aspirin resistance

Fifteen out of 50 patients (30%) were found to be aspirinresistant. The distribution of ARU values is shown in

Table 1Comparison of patient characteristics between aspirin-resistant and aspirin-sensitive patients (n = 50)

Variable Frequency (%) or median (range)

All patients (n = 50) Aspirin resistant (n = 15) Aspirin sensitive (n = 35)

Age (yrs) 77 (51, 93) 77 (59, 92) 77 (51, 93)Gender (male) 37 (74%) 14 (93.3%) 23 (65.7%)Caucasian 49 (98%) 14 (93.3%) 35 (100.0%)ARU 509 (385, 626) 576 (550, 626) 466 (385, 547)Duration of aspirin therapy (yrs) 5 (2, 31) 6 (2, 31) 5 (2, 17)

History of ischaemic events

Past history of TIA 5 (10%) 3 (20.0%) 2 (5.7%)Past history of stroke 3 (6%) 1 (6.7%) 2 (5.7%)Past history of combined stroke + TIA 7 (14%) 4 (26.7%) 3 (8.6%)Past history of unstable angina 6 (12%) 2 (13.3%) 4 (11.4%)Past history of MI 4 (8%) 0 (0.0%) 4 (11.4%)Past history of combined MI + unstable angina 10 (20%) 2 (13.3%) 8 (22.9%)

TOAST8 classification of strokes

Large artery atherosclerosis 13 (26%) 2 (13.3%) 11 (31.4%)Cardiogenic embolic 7 (14%) 3 (20.0%) 4 (11.4%)Lacunar infarcts 9 (18%) 3 (20.0%) 6 (17.1%)Strokes of undetermined aetiology 21 (42%) 7 (46.7%) 14 (40.0%)

Clinical syndrome of stroke

NIHSS21 4.5 (2, 20) 7 (2, 14) 4 (2, 20)TACI 4 (8%) 1 (6.7%) 3 (8.6%)PACI 24 (48%) 9 (60.0%) 15 (42.9%)POCI 8 (16%) 2 (13.3%) 6 (17.1%)LACI 14 (28%) 3 (20.0%) 11 (31.4%)

Vascular risk factors

Atrial fibrillationa 15 (30%) 6 (40.0%) 9 (25.7%)Smokingb 36 (72%) 11 (73.3%) 25 (71.4%)Hypercholesterolemiac 39 (78%) 9 (60.0%) 30 (85.7%)Previous carotid endarterectomyd 3 (6%) 0 (0.0%) 3 (8.6%)Carotid stenosise 12 (24%) 2 (13.3%) 10 (28.6%)Hypertension 48 (96%) 14 (93.3%) 34 (97.1%)Diabetesf 21 (42%) 4 (26.7%) 17 (48.6%)

ARU = aspirin reaction unit, LACI = lacunar infarct, MI = myocardial infarction, NIHSS = National Institutes of Health Stroke Scale, PACI = partialanterior circulation infarct, POCI = posterior circulation infarct, TACI = total anterior circulation infarct, TIA = transient ischaemic attack,TOAST = Trial of Org 10172 in Acute Stroke Treatment.

a Known atrial fibrillation on admission or known paroxysmal atrial fibrillation.b History of smoking or current smoker.c Hypercholesterolemia defined as previous treatment with cholesterol lowering agent on admission or fasting lipids of >5.5 mmol/L cholesterol on

admission.d Previous endarterectomy of at least one internal carotid artery.e Defined as at least one carotid artery with measurement P50% on Ultrasound by Doppler or North American Symptomatic Carotid Endarterectomy

Trial (NASCET) criteria in patients receiving an angiogram.f Previous diagnosis of diabetes on admission. All patients with diabetes in this study had non insulin dependent diabetes mellitus.

1206 D. Bennett et al. / Journal of Clinical Neuroscience 15 (2008) 1204–1209

Fig. 1. Our patients had median ARU of 509 (range 385–626). All patients enrolled in the study had been takingaspirin for at least 2 years prior to admission with the med-ian duration of aspirin therapy 5 years (range 2–31). Fig. 2shows ARU values in patients who had a history of ischae-mic stroke or TIA compared to those who did not. Patientswith previous events had a higher median ARU of 550(range 471–573) compared to those without events, medianARU of 496 (range 385–626).

3.2. Aspirin resistance and recurrent ischaemic stroke

Three patients had a history of stroke and 5 had ahistory of TIA with 1 patient having both, giving a rate

of recurrent ischaemic stroke or TIA of 7/50 (14%) overthe previous 2 years. Of these 7 patients, 4/7 were aspirinresistant. Of the 43 patients without a history of ischaemicevents, 11/43 were aspirin resistant. In this sample, theodds of having a past history of either TIA or ischaemicstroke was about 4 times higher in patients with aspirinresistance compared to those without, but the differencewas not significant (odds ratio, OR = 3.88; 95% confidenceinterval, 0.54–29.87; p = 0.18).

3.3. Aspirin-resistant vs. aspirin-sensitive patients

Aspirin-resistant patients were comparable to aspirin-sensitive patients with regard to stroke syndrome and

Fig. 1. Frequency and distribution of aspirin reaction unit (ARU) valuesacross the study population.

Fig. 2. Parallel boxplots showing aspirin reaction unit (ARU) values bypast history of stroke or transient ischaemic attack (TIA).

D. Bennett et al. / Journal of Clinical Neuroscience 15 (2008) 1204–1209 1207

classification (Table 1). However, there was a trendtowards lower rates of diabetes, carotid stenosis and hyper-cholesterolemia among resistant patients (Table 2). Malegender also suggested an increased risk of being aspirin

Table 2Association between baseline and clinical variables with aspirin reactionunits (ARU) P550

Variable Odds ratio 95% CI p value

Age 1.02 0.95–1.11 0.57Gender (male) 7.30 0.87–334.10 0.08Hypertension 0.41 0.01–34.61 0.51Atrial fibrillation 1.93 0.43–8.20 0.33Smoking 1.10 0.24–5.86 1.00Hypercholesterolemia 0.25 0.05–1.28 0.06Diabetes 0.39 0.08–1.66 0.21Carotid stenosis 0.38 0.04–2.26 0.30History of ischaemic stroke + TIA 3.88 0.54–29.87 0.18History of MI + unstable angina 0.52 0.05–3.19 0.70

MI = myocardial infarction, TIA = transient ischaemic attack.

resistant (OR = 7.3; 95% confidence interval, 0.87–334.10;p = 0.08). No statistically significant differences betweenthe 2 groups were demonstrated for any of the variableswe assessed using univariate analysis (Table 2).

4. Discussion

Prospective clinical studies have shown that aspirinresistance as measured by laboratory tests is associatedwith an increased risk of atherothrombotic events. Thesestudies, however, have not analysed ischaemic stroke pa-tients in particular and have not asked whether aspirinresistance increases the risk of recurrent ischaemic stroke.Our study is the first to determine the prevalence of aspirinresistance in an Australian stroke population and todetermine if this was more common in patients with aprior history of vascular events. This would suggest thataspirin-resistant patients are at higher risk of recurrentevents, which would require testing in a prospective design.

This study suggests that aspirin resistance is prevalent inpatients presenting with acute ischaemic stroke. Aspirinresistance tended to be more common among males butdid not show a statistically significant correlation withany clinical variable. Aspirin resistance in our patients sug-gests increased risk of ischaemic stroke and TIA but not ofcardiovascular events among our population of ischaemicstroke patients.

The prevalence of aspirin resistance at 30% in our studyis comparable to results of cardiology and stroke studiesusing the Ultegra RPFA. Previous studies reported ratesof aspirin resistance of 19%, 23% and 17% respectivelyusing the Ultegra RPFA in varying patient popula-tions.5,7,8 The limited number of studies examining aspirinresistance in a stroke population indicate that aspirin resis-tance prevalence is 12% to 34%.12,13

It is difficult to compare our results in terms of increasedrisk of ischaemic stroke associated with aspirin resistancedue to the lack of literature analysing this hypothesis.Two studies have attempted to correlate aspirin resistancewith recurrent ischaemic stroke.13,14 These studies usedsmall sample sizes and observation periods similar to ourstudy. Grundman et al. conducted a retrospective studyin 53 patients taking aspirin for secondary prevention ofischaemic stroke or TIA. The rate of aspirin resistancewas 34% among patients with a recent recurrent ischaemicevent compared to 0% of patients with no recent events.Grundman’s inclusion criteria consisted of history of previ-ous cerebrovascular or cardiovascular events, which indi-cated that the population was not exclusively stroke. Thisopen criterion also suggests that asymptomatic and symp-tomatic groups may have been insufficiently comparable interms of ischaemic stroke events.

Grotemeyer et al. followed up 180 post ischaemic strokepatients over 2 years. This study found 33% of the popula-tion to be aspirin resistant with a 10-fold increase in clinicalevents among resistant patients. The major study endpointswere not specific to recurrent stroke constituting MI, vas-

1208 D. Bennett et al. / Journal of Clinical Neuroscience 15 (2008) 1204–1209

cular death and stroke. The ability to generalise the resultsof this study was reduced as all patients had ischaemicstroke in the internal carotid artery territory excludingthose with posterior circulation infarcts. These two studiesused alternative methods of measuring aspirin resistancecompared to our study.

The cause of aspirin resistance is uncertain and is prob-ably multifactorial. Biological, genetic and environmentalfactors are likely candidates. Our study provides no statis-tically significant clinical predictors of aspirin resistance onunivariate analysis. Diabetes,15,16 hypercholesterolemia,17

smoking,18 age,19 and female gender3 have been associatedwith aspirin resistance. Hypercholesterolemia appeared tobe lower among resistant patients in our study population.These findings are not expected as numerous studies sug-gest that diabetes and hypercholesterolemia predict in-creased risk of aspirin resistance.

Gum et al. provided the first evidence of a clinical asso-ciation between aspirin resistance and increased cardiovas-cular events.3 Aspirin resistance predicted increased eventsindependent of age and conventional vascular risk factors.The suggestion of lower rates of diabetes, cholesterol andincreased incidence of male gender among our resistantpopulation may be due to small sample size and chance.Further studies are needed to analyse clinical predictorsof aspirin resistance.

Our study has strengths. Because patients were testedonly after 100 mg aspirin had been given as documentedon the ward drug charts, measuring aspirin resistance onadmission was not confounded by compliance, a simplemechanism of laboratory and clinical aspirin resistance of-ten overlooked.20,21 Some previous studies have relied onpatient compliance when measuring aspirin resistance andthus potentially overestimated the numbers of aspirin-resis-tant patients.4,5 The use of a point-of-care (POC) test tomeasure aspirin resistance adds clinical relevance to our re-sults. A quick and easy method of testing aspirin resistanceis necessary if it is to become a part of standard manage-ment in secondary prevention of stroke. POC tests over-come problems encountered by earlier studies in whichaspirin resistance had been measured using the traditionalgold standard of LTA.3,14

Our study has limitations. First, our sample may haveunderestimated the true prevalence of aspirin resistanceand rate of recurrent ischaemic stroke. Patients who hadbeen treated with antiplatelet agents other then aspirin(e.g. clopidogrel) were excluded. Second, assessment ofthe patient’s history of ischaemic events was carried outretrospectively. This was a particular limitation for historyof TIA and unstable angina for which patients are not al-ways admitted to hospital and therefore rely on historyand records kept by their general practitioner. Third, com-pliance with aspirin therapy over the previous 2 years couldbe assessed only on patient history or the history from theprimary carer.

The issue still remains that if aspirin resistance does cor-relate with increased cerebral ischaemic events in a stroke

population, the mechanism of aspirin resistance is not clar-ified, nor is the appropriate management of these patients.Further studies need to be conducted to assess the mostappropriate management of aspirin-resistant patients.Eventually aspirin resistance may become routinelyscreened and managed in much the same way as bloodpressure or hypercholesterolemia as one of multiple riskfactors that optimise secondary prevention of ischaemicstroke.

Our data suggests that aspirin resistance is prevalentwithin the ischaemic stroke population and is similar tothat demonstrated in the cardiology literature. Further-more aspirin resistance tended to be more prevalent in pa-tients at an increased risk of cerebral ischaemic events.Further evaluation of the clinical consequences of aspirinresistance using the Ultegra RPFA in a larger prospectivestudy is required.

References

1. Collaboration AT. Collaborative meta-analysis of randomised trialsof antiplatelet therapy for prevention of death, myocardial infarction,and stroke in high risk patients. BMJ 2002;324:71–86.

2. Algra A, van Gijn J. Aspirin at any dose above 30 mg offers onlymodest protection after cerebral ischaemia. J Neurol Neurosurg

Psychiatry 1996;60:197–9.3. Gum PA, Kottke-Marchant K, Welsh PA, et al. A prospective,

blinded determination of the natural history of aspirin resistanceamong stable patients with cardiovascular disease. J Am Coll Cardiol

2003;41:961–5.4. Eikelboom JW, Hirsh J, Weitz JI, et al. Aspirin-resistant thrombox-

ane biosynthesis and the risk of myocardial infarction, stroke, orcardiovascular death in patients at high risk for cardiovascular events.Circulation 2002;105:1650–5.

5. Wang JC, Aucoin-Barry D, Manuelian D, et al. Incidence of aspirinnonresponsiveness using the Ultegra Rapid Platelet Function Assay-ASA. Am J Cardiol 2003;92:1492–4.

6. Mueller MR, Salat A, Stangl P, et al. Variable platelet response tolow-dose ASA and the risk of limb deterioration in patients submittedto peripheral arterial angioplasty. Thromb Haemost 1997;78:1003–7.

7. Harrison P, Segal H, Blasbery K, et al. Screening for aspirinresponsiveness after transient ischemic attack and stroke: comparisonof 2 point-of-care platelet function tests with optical aggregometry.Stroke 2005;36:1001–5.

8. Chen WH, Lee PY, Ng W, et al. Aspirin resistance is associated with ahigh incidence of myonecrosis after non-urgent percutaneous coro-nary intervention despite clopidogrel pretreatment. J Am Coll Cardiol

2004;43:1122–6.9. Adams HP, Bendixen BH, Kappelle LJ, et al. Classification of subtype

of acute ischemic stroke – definitions for use in a multicenter clinical-trial. Stroke 1993;24:35–41.

10. Bamford J, Sandercock P, Dennis M, et al. Classification and natural-history of clinically identifiable subtypes of cerebral infarction. Lancet

1991;337:1521–6.11. Coleman JL, Wang JC, Simon DI. Determination of individual

response to aspirin therapy using the accumetrics Ultegra RPFA-ASASystem. Point of Care: The Journal of Near-Patient Testing &

Technology 2004;3:77–82.12. Lee SJ, Kim TH, Kim JG, et al. The prevalence of aspirin resistance

and related factors in patients with ischemic stroke using the rapidplatelet function assay. Stroke 2007;38:524-4.

13. Grundmann K, Jaschonek K, Kleine B, et al. Aspirin non-responderstatus in patients with recurrent cerebral ischemic attacks. J Neurol

2003;250:63–6.

D. Bennett et al. / Journal of Clinical Neuroscience 15 (2008) 1204–1209 1209

14. Grotemeyer KH, Scharafinski HW, Husstedt IW. 2-Year follow-up ofaspirin responder and aspirin nonresponder – a pilot-study including180 poststroke patients. Thromb Res 1993;71:397–403.

15. Gurbel PA, Bliden KP, Chaganti S, et al. Effects of diabetes on theprevalence of aspirin resistance during low dose aspirin therapy. J Am

Coll Cardiol 2007;49:230A.16. Fateh-Moghadam S, Plockinger U, Cabeza N, et al. Prevalence of

aspirin resistance in patients with type 2 diabetes. Acta Diabetol

2005;42:99–103.17. Friend M, Vucenik I, Miller M. Research pointers: Platelet respon-

siveness to aspirin in patients with hyperlipidaemia. BMJ

2003;326:82–3.

18. Mirkhel A, Peyster E, Sundeen J, et al. Frequency of aspirin resistancein a community hospital. Am J Cardiol 2006;98:577–9.

19. Alberts MJ, Bergman DL, Molner E. Antiplatelet effect ofaspirin in patients with cerebrovascular disease. J Vasc Surg

2004;39:1142.20. Hankey GJ, Eikelboom JW. Aspirin resistance. Lancet 2006;367:

606–17.21. Cotter G, Shemesh E, Zehavi M, et al. Lack of aspirin effect: aspirin

resistance or resistance to taking aspirin? Am Heart J

2004;147:293–300.