1706 Tantry et al. JACC Vol. 46, No. 9, 2005 Aspirin Resistance November 1, 2005:1705–9 ment at the time of blood drawing. Healthy subjects had Abbreviationsand Acronyms blood drawn immediately before receiving 325 mg aspirin AA ϭ arachidonic acid and then 24 h later. ADP ϭ adenosine diphosphate Platelet function studies. LTA. The blood-citrate mixture COX-1 ϭ cyclooxygenase-1 was centrifuged at 120 g for ﬁve minutes, and PRP was GP ϭ glycoprotein LTA ϭ light-transmittance aggregometry recovered. The blood-citrate mixture was subjected to fur- MAAA ϭ arachidonic acid-induced clot strength ther centrifugation at 850 g for ﬁve minutes to recover (measurement of aspirin effect) platelet-poor plasma (PPP). Aggregation was assessed using MAﬁbrin ϭ activator-induced clot strength a Chronolog Lumi-aggregometer (model 490, Chronolog, (measurement of ﬁbrin contribution) Haverton, Pennsylvania) with the Aggrolink software pack- MAthrombin ϭ thrombin-induced clot strength (maximum clot strength) age as previously described (13). We assessed the aggrega- PCI ϭ percutaneous coronary intervention tion effect of 1 mmol/l and 2 mmol/l AA in the ﬁrst 40 PPP ϭ platelet-poor plasma patients and found no differences between the two AA PRP ϭ platelet-rich plasma concentrations (data not shown). Therefore, platelets were SAT ϭ stent thrombosis TEG ϭ thrombelastography stimulated with 1 mmol/l AA in the remaining PRP samples and platelet aggregation was expressed as the maximal percent change in light transmittance from base-exclude the inﬂuence of leukocytes and erythrocytes on line, using PPP as a reference.platelet aggregation (11,12). HAEMOSCOPE THROMBELASTOGRAPH (TEG) PLATELET MAPPING ASSAY. The recent U.S. Food and DrugMETHODS Administration-approved TEG Platelet Mapping assayPatients. The study was approved by the Investigational (Haemoscope Corporation, Niles, Illinois) relies on theReview Board of the Sinai Hospital of Baltimore. Two measurement of clot strength to enable a quantitativehundred three consecutive patients undergoing coronary analysis of platelet function. The assay uses heparin as anstenting were prospectively studied. Twenty patients with a anticoagulant to eliminate thrombin activity in the sample.history of stent thrombosis (SAT) were identiﬁed by review Reptilase and factor XIIIa (activator F) are used to generateof medical records and were contacted for the study. All a cross-linked ﬁbrin clot to isolate the ﬁbrin contribution topatients reported compliance with long-term aspirin ther- the clot strength (14). The contribution of P2Y12 receptorapy. All patients undergoing stenting received 325 mg or COX pathways to the clot formation can be measured byaspirin on the day of the procedure and daily thereafter. All the addition of an appropriate agonist, ADP or AA.patients who suffered SAT were on a maintenance dose of Therefore, AA is added to activator F to measure the degree75 mg/day clopidogrel at the time of the blood draw. of thromboxane A2-induced platelet aggregation.Seventy percent of the patients who underwent percutane- Blood was analyzed according to the manufacturer’sous coronary intervention (PCI) were not on a maintenance instructions. One milliliter of heparinized blood was trans-dose of clopidogrel before the procedure and received a 300- ferred to a vial containing kaolin and mixed by inversion.or 600-mg clopidogrel loading dose during the procedure Five hundred microliters of the activated blood was thenfollowed by a 75-mg/day maintenance dose thereafter. transferred to a vial containing heparinase and mixed toThirty percent of the PCI patients were on a 75-mg daily neutralize heparin. The neutralized blood (360 l) wasclopidogrel maintenance dose before the procedure and did immediately added to a heparinase-coated cup and assayednot receive a further loading dose. in the TEG analyzer to measure the thrombin-induced clotControl subjects. Six healthy subjects were studied before strength (MAthrombin). Heparinized blood (340 l) wasand 24 h after receiving 325 mg aspirin. added to a non-coated cup containing reptilase and activatorBlood samples. Blood was collected in a blood collection F to generate a whole-blood ﬁbrin cross-linked clot in thetube containing 3.8% trisodium citrate for LTA and con- absence of thrombin generation or platelet stimulationtaining 40 USP lithium heparin tube (United States Phar- (MAﬁbrin). A third sample (340 l) of heparinized bloodmacopeia, Rockville, Maryland) for thrombelastography was added to a non– heparinase-coated cup in the presence(TEG), ﬁlled to capacity, and then inverted three to ﬁve of the activator F and AA (1 mmol/l) to generate whole-times for gentle mixing. Blood samples were obtained blood cross-linked clot with platelet activation (MAAA).before clopidogrel loading and heparin and glycoprotein Platelet aggregation in response to AA is calculated using(GP) IIb/IIIa receptor inhibitor administration (baseline), computerized software based on the formula: % aggregationand at 24 h after the procedure in those patients not ϭ [(MAAA Ϫ MAﬁbrin)/(MAthrombin Ϫ MAﬁbrin)] ϫ 100.receiving GP IIb/IIIa inhibitors. In patients receiving GP Deﬁnition of aspirin resistance. In this study we mea-IIb/IIIa inhibitors, blood samples were evaluated at ﬁve to sured platelet response to aspirin therapy by measuring theseven days after the procedure. All patients with a history of aggregation in both PRP (LTA) and whole blood (TEG)subacute thrombosis were conﬁrmed to be on aspirin treat- with 1 mmol/l AA as the agonist. Because whole-blood
JACC Vol. 46, No. 9, 2005 Tantry et al. 1707November 1, 2005:1705–9 Aspirin ResistanceTable 1. Patient Demographics and Procedural Characteristics performed were non-emergent; 36 patients were admitted with Patients Patients unstable angina, and 11 patients had non–ST-segment eleva- Undergoing With Stent tion myocardial infarction. The remainder of the patients had PCI Thrombosis stable angina. In the patients with a history of SAT, the mean (n )302 ؍ (n )02 ؍ time from the index procedure to SAT was 23 Ϯ 16 days. TheAge (yrs) 63 Ϯ 17 65 Ϯ 11 average time from the occurrence of SAT to blood drawing forRace (caucasian) (%) 75 50Gender (male) (%) 60 50 laboratory evaluation was 218 Ϯ 204 days. The patientRisk factors/past medical history (%) demographics and procedural characteristics of patients under- Smoking 61 45 going PCI and patients with SAT are shown in Table 1. All Family history of coronary artery 53 80 patients undergoing PCI and all patients who suffered SAT disease Hypertension 73 75 reported taking aspirin 325 mg daily. Hyperlipidemia 75 75 Pre-treatment aggregation. CONTROL SUBJECTS. The Diabetes 38 60 LTA and TEG analyses were completed in all patients (n ϭ Prior myocardial infarction 16 60 223) and healthy subjects (n ϭ 6). In healthy subjects, Prior CABG graft 16 25 Prior PTCA 24 20 AA-induced aggregation by LTA was 82 Ϯ 10% and byBaseline medications (%) TEG was 86 Ϯ 14% before aspirin, and it decreased Beta-blockers 81 85 signiﬁcantly after aspirin treatment (2 Ϯ 1% by LTA and 5 ACE inhibitors 69 75 Ϯ 7% by TEG, respectively [p Ͻ 0.001]), indicating the Calcium blockers 19 25 Lipid-lowering agents 71 80 complete inhibition of COX-1 by a 325-mg aspirin dose Clopidogrel 30 100 (Figs. 1 and 2). Aspirin 100 100Procedural variables PCI PATIENTS. In patients undergoing PCI (n ϭ 203), all of Total lesion length, mm 20 Ϯ 12 24 Ϯ 6 whom claimed compliance with aspirin therapy, pre-PCI Reference vessel diameter, mm 3.1 Ϯ 0.4 2.8 Ϯ 0.3 AA-induced platelet aggregation by LTA and TEG was Number of vessels treated 1.4 Ϯ 0.6 1.3 Ϯ 0.4 low (5 Ϯ 14% and 7 Ϯ 17%, respectively). Seven PCIACE ϭ angiotensin-converting enzyme; CABG ϭ coronary artery bypass graft; patients had high aggregation by LTA and TEG andPCI ϭ percutaneous coronary intervention; PTCA ϭ percutaneous transluminalcoronary angioplasty. admitted non-compliance with aspirin therapy. In these non-compliant patients, aggregation by LTA was 73 Ϯ 21%aggregation in the TEG assay includes ﬁbrin generation and before PCI and decreased to 2 Ϯ 2% post-PCI afterthe interaction of platelets with ﬁbrin, higher values of in-hospital treatment with 325 mg aspirin (p Ͻ 0.001).aggregation are expected compared with data obtained from Similarly, aggregation by TEG was 72 Ϯ 27% before PCIPRP. Therefore, aspirin resistance is deﬁned as the occur- and 7 Ϯ 13% post-PCI after treatment with 325 mg aspirinrence of more than 20% platelet aggregation after stimula- (p Ͻ 0.001). Thus, all patients in the non-compliant grouption by 1 mmol/l AA as measured by LTA or more than were found to be aspirin sensitive.50% platelet aggregation after stimulation by 1 mmol/l AA All patients in the compliant group (n ϭ 196) wereas measured by TEG after aspirin therapy. aspirin sensitive. Pre-PCI and post-PCI aggregation byStatistical analysis. Comparisons were made within LTA was 3 Ϯ 2% and 3 Ϯ 2%, respectively (p ϭ NS).groups between baseline and post-treatment by t tests, andp Ͻ 0.05 was considered signiﬁcant. Based on the normaldistribution of data, the mean value Ϯ standard deviationwas used. A standard regression analysis was performed tocorrelate aggregation by conventional LTA with TEGmeasurements of aggregation. In the present study, wehypothesized that by using light transmittance aggregom-etry with 1 mmol/l AA as the agonist, the prevalence ofaspirin resistance is ϳ5%. Based on the formula: n ϭ 15.4 ϫp ϫ (1 Ϫ p)/w2, in which n ϭ number of patients required,p ϭ prevalence of aspirin resistance (5%), and w is thedesired width of 95% conﬁdence interval (Ϯ 6%), 200patients were required to complete the study.RESULTS Figure 1. Scatter plot showing 1 mmol/l arachidonic acid-induced platelet aggregation by light-transmittance aggregometry in individual healthyHealthy subjects and patients. Healthy subjects were subjects (n ϭ 6) (before and after aspirin therapy), compliant (n ϭ 196) and non-compliant patients (n ϭ 7) undergoing percutaneous interventions35 Ϯ 5 years old and had no modiﬁable cardiovascular risk (PCIs) (before and after PCI), and patients with a history of stentfactors. Among the patients undergoing PCI, all procedures thrombosis (SAT).
1708 Tantry et al. JACC Vol. 46, No. 9, 2005 Aspirin Resistance November 1, 2005:1705–9 whereas we analyzed AA-induced aggregation at 1.0 mmol/l. The differences in the estimate of resistance based on AA-induced aggregation between the two studies may be in part caused by the difference in the AA concentration used in two studies. However, we compared the effect of 1 mmol/l AA and 2 mmol/l AA on platelet aggregation in the ﬁrst 40 patients and found no difference in platelet aggre- gation. Therefore, on the remaining patients we used 1 mmol/l AA. Finally, the patient population studied, the dose of aspirin used, and also the strict in-hospital compli- ance measures used in our study may have affected the results.Figure 2. Scatter plot showing 1 mmol/l arachidonic acid-induced platelet Since the original study by Gum et al. (3), other inves-aggregation by thrombelastography in individual healthy subjects (n ϭ 6)(before and after aspirin therapy), compliant (n ϭ 196) and non-compliant tigators have used various methods such as the bleedingpatients (n ϭ 7) undergoing percutaneous interventions (PCIs) (before and time and cationic propyl gallate-, ADP-, epinephrine-, andafter PCI), and patients with a history of stent thrombosis (SAT). collagen-induced aggregation to assess aspirin responsive- ness (1– 6). The latter methods affect platelet aggregation byAggregation by TEG was 4 Ϯ 9% before PCI and 6 Ϯ 14% pathways that include but are not speciﬁc to COX-1 (7– 8).after PCI (p ϭ NS). Moreover, the deﬁnition of aspirin resistance also variedPATIENTS WITH SAT. In patients with SAT (n ϭ 20), depending on the assay. Recently, aspirin resistance hasAA-induced platelet aggregation by LTA was 6 Ϯ 12% and been reported by using point-of-service assays. The PFA-by TEG was 9 Ϯ 20%. One patient met the deﬁnition of 100 assay is partially dependent on epinephrine-inducedaspirin resistance; AA-induced aggregation was 54% by aggregation and thus does not solely isolate COX activityLTA and 85% by TEG in that patient. Therefore, among (6). The Ultegra assay estimated aspirin resistance by using223 patients with coronary artery disease evaluated in our cationic propyl gallate to indirectly inﬂuence AA metabo-investigation, only one patient (ϳ0.4%), who also had a lism (4,5). Thus, stimulation with a known concentration ofhistory of SAT, was resistant to aspirin treatment. AA, a direct substrate of the COX-1 enzyme, was not conducted in the latter studies.CORRELATION OF AGGREGATION MEASURED BY LTA AND In the present study, we evaluated the platelet response toTEG. The PRP aggregation measured by LTA strongly aspirin in PRP and in whole blood. Moreover, both meth-correlated with aggregation measured in whole blood by ods strongly correlated in measuring AA-induced plateletTEG (r ϭ 0.85, p Ͻ 0.001). All of the non-compliant aggregation. Finally, we evaluated the efﬁcacy of 325-mgpatients had high aggregation by both TEG and LTA aspirin therapy in patients and healthy subjects. Among the(Figs. 1 and 2). 223 patients enrolled in the study, ϳ3% were found non- compliant with aspirin therapy and were therefore pseudoDISCUSSION non-responders. This ﬁnding is concordant with previousThe present study suggests that aspirin resistance is rare in reports (15,16). In-hospital treatment of these non-compliant patients with coronary artery disease when as- compliant patients with 325 mg aspirin showed that theysessed by methods directly dependent on platelet COX-1. were indeed sensitive to aspirin therapy. Only one patientIn our study, only one of 223 patients (ϳ0.4%) was resistant who suffered an SAT was found to be resistant to aspirinto aspirin treatment. The concordant data in whole blood therapy as measured by both LTA and TEG. The latterand PRP suggest that other cells do not signiﬁcantly ﬁnding suggests that aspirin non-responsiveness may be ainﬂuence ex vivo platelet responsiveness to aspirin. risk factor for ischemic events after stenting, as has been The antiplatelet effect of aspirin is based on its irreversible reported by other investigators (1– 6).acetylation of cyclooxygenase, speciﬁcally the COX-1 isoen- The rare occurrence of aspirin resistance in the presentzyme in platelets. The COX-1 enzyme catalyzes the con- study strongly suggests that the occurrence of aspirin resis-version of AA to prostaglandin H2 and subsequent synthesis tance in published reports is overestimated. As can be seenof thromboxane A2, a potent platelet agonist. Thus the ideal with the non-compliant subgroup of patients in our study,laboratory evaluation of the antiplatelet effect of aspirin outpatient treatment and lack of compliance are importantshould include the functional response of platelets to AA or considerations in determining the prevalence of aspirinmeasurement of the levels of thromboxane B2, a stable resistance. In addition, use of COX-1–speciﬁc laboratorymetabolite of thromboxane A2. analyses such as platelet aggregation in response to AA may The earlier studies of Gum et al. (3) used Ͼ20% show the actual prevalence of aspirin resistance in studyAA-induced platelet aggregation and Ͼ70% ADP-induced populations. This hypothesis is strongly supported by a veryaggregation as criteria for deﬁning aspirin resistance. The recent study of Schwartz et al. (17), in which the prevalenceconcentration of AA used by Gum et al. (3) was 1.6 mmol/l, of the aspirin resistance was studied in 190 patients with a
JACC Vol. 46, No. 9, 2005 Tantry et al. 1709November 1, 2005:1705–9 Aspirin Resistancehistory of myocardial infarction using AA-stimulated light 3. Gum PA, Kottke-Marchant K, Welsh PA, White J, Topol EJ. A prospective, blinded determination of the natural history of aspirintransmittance aggregometry. These investigators found that resistance among stable patients with cardiovascular disease. J Am Collonly one patient was resistant to aspirin therapy. Therefore, Cardiol 2003;41:961–5.the application of speciﬁc methods directly indicating 4. Wang JC, Aucoin-Barry D, Manuelian D, et al. Incidence of aspirinCOX-1 activity in platelets is recommended to indicate the nonresponsiveness using the Ultegra Rapid Platelet Function Assay- ASA. Am J Cardiol 2003;92:1492– 4.efﬁcacy of aspirin therapy (7,8). 5. Chen W-H, Lee P-Y, Ng W, et al. Aspirin resistance is associatedStudy limitations. This study is not powered to show the with a high incidence of myonecrosis after non-urgent percutaneousrelationship of aspirin resistance and demographic variables. coronary intervention despite clopidogrel pretreatment. J Am Coll Cardiol 2004;43:1122– 6.However, the study population included both patients 6. Andersen K, Hurlen M, Arnesen H, Seljeﬂot I. Aspirin non-admitted to the hospital for PCI and patients who had responsiveness as measured by PFA-100 in patients with coronarysuffered SAT despite dual antiplatelet therapy. We did not artery disease. Thromb Res 2002;108:37– 42.measure thromboxane metabolite levels or expression of 7. Michelson AD, Cattaneo M, Eikelboom JW, et al. Aspirin resistance: position paper of the Working Group on Aspirin Resistance, Plateletactivation-dependent surface molecules, which may have Physiology Subcommittee of the Scientiﬁc and Standardization Com-further strengthened our conclusion that platelet aspirin mittee, International Society on Thrombosis and Haemostasis. Jresistance is rare. Thromb Haemost 2005;3:1309 –11. 8. Cattaneo M. Aspirin and clopidogrel: efﬁcacy, safety, and the issue ofConclusions. The result of the present study indicates that drug resistance. Arterioscler Thromb Vasc Biol 2004;24:1980 –7.aspirin resistance may be overestimated by previous reports 9. Eikelboom JW, Hankey GJ. Failure of aspirin to prevent atherothrom-using nonspeciﬁc laboratory measurements that do not bosis: potential mechanisms and implications for clinical practice.isolate aspirin’s primary target, platelet COX-1. In the Am J Cardiovasc Drugs 2004;4:57– 67. 10. Maree AO, Fitzgerald DJ. Aspirin and coronary artery disease.current study, the target of aspirin therapy was assessed by Thromb Haemost 2004;92:1175– 81.techniques using whole blood and PRP, and aspirin resis- 11. Valles J, Santos MT, Aznar J, et al. Erythrocyte promotion of platelettance was nearly absent. Measurements of platelet response reactivity decreases the effectiveness of aspirin as an antithrombotic therapeutic modality: the effect of low-dose aspirin is less than optimalto aspirin using more speciﬁc methods targeting COX-1 in patients with vascular disease due to prothrombotic effects ofenzyme activity and also a standardized deﬁnition of aspirin erythrocytes on platelet reactivity. Circulation 1998;97:350 –5.resistance are warranted to ultimately link aspirin- 12. Stafford NP, Pink AE, White AE, Glenn JR, Heptinstall S. Mecha-dependent tests to clinical outcomes in patients with coro- nisms involved in adenosine triphosphate-induced platelet aggregation in whole blood. Arterioscler Thromb Vasc Biol 2003;23:1928 –33.nary artery disease. 13. Gurbel PA, Bliden KP, Hayes MH, Yoho JA, Herzog WR, Tantry US. The relation of dosing to clopidogrel responsiveness and theReprint requests and correspondence: Dr. Paul A. Gurbel, incidence of high post-treatment platelet aggregation in patientsSinai Center for Thrombosis Research, Hoffberger Building, undergoing coronary stenting. J Am Coll Cardiol 2005;45:1392– 6. 14. Craft RM, Chavez JJ, Bresee SJ, Wortham DC, Cohen E, Carroll RC.Suite 56, 2401 West Belvedere Avenue, Baltimore, Maryland A novel modiﬁcation of the thromboelastograph assay, isolating21215. E-mail: email@example.com. platelet function, correlates with optical platelet aggregation. J Lab Clin Med 2004;143:301–9.REFERENCES 15. Glynn RJ, Buring JE, Manson JE, LaMotte F, Hennekens CH. Adherence to aspirin in the prevention of myocardial infarction. The 1. Grotemeyer KH, Scharaﬁnski HW, Husstedt IW. Two-year Physicians’ Health Study. Arch Intern Med 1994;154:2649 –57. follow-up of aspirin responder and aspirin non responder. A pilot- 16. Ferrari E, Benhamou M, Cerboni P, Marcel B. Coronary syndromes study including 180 post-stroke patients. Thromb Res 1993;71:397– following aspirin withdrawal: a special risk for late stent thrombosis. 403. J Am Coll Cardiol 2005;45:456 –9. 2. Eikelboom JW, Hirsh J, Weitz JI, Johnston M, Yi Q, Yusuf S. 17. Schwartz KA, Schwartz DE, Ghosheh K, Reeves MJ, Barber K, Aspirin-resistant thromboxane biosynthesis and the risk of myocardial Defranco A. Compliance as a critical consideration in patients who infarction, stroke, or cardiovascular death in patients at high risk for appear to be resistant to aspirin after healing of myocardial infarction. cardiovascular events. Circulation 2002;105:1650 –5. Am J Cardiol 2005;95:973–5.