- Email: [email protected]
Monitoring Platelet Function to Reduce the Risk of Ischemic and Bleeding Complications
Monitoring Platelet Function to Reduce the Risk of Ischemic and Bleeding Complications Matthew J. Price, MD Patients with acute coronary syndromes and patients who undergo coronary stent implantation frequently receive dual antiplatelet therapy with aspirin and a thienopyridine. A wide variability exists in the inhibitory response to the thienopyridine clopidogrel among individuals, but this variability can be detected by a variety of platelet function assays. Several studies have demonstrated an association between high platelet reactivity post clopidogrel and subsequent cardiac events. To be adopted into routine clinical practice, platelet function assays must not only be able to identify at-risk patients, but they must also determine the appropriate thresholds for nonresponsiveness that provide appropriate sensitivity and specificity for subsequent adverse events. Prospective, randomized trials should determine whether therapeutic interventions guided by platelet function tests can safely and effectively reduce thrombotic events in at-risk patients. The use of platelet function monitoring to guide the timing of surgery after the discontinuation of antiplatelet therapy also merits further study. © 2009 Published by Elsevier Inc. (Am J Cardiol 2009;103[suppl]: 35A–39A)
Dual antiplatelet therapy with aspirin and a thienopyridine is a cornerstone of the management of patients undergoing percutaneous coronary intervention (PCI) and coronary stent implantation. This approach was adopted after trials such as the Stent Antithrombotic Regimen Study (STARS) demonstrated a marked reduction in thrombotic events within 30 days after PCI in patients treated with aspirin and ticlopidine (0.5%) compared with patients who received aspirin alone (3.6%) or aspirin plus warfarin (2.7%).1 Since that time, however, it has become well established that the platelet inhibitory response to clopidogrel varies greatly among individuals.2,3a Several monitoring assays are currently available that can measure platelet reactivity after clopidogrel exposure. Yet, to be clinically useful, a platelet function assay must fulfill several criteria: (1) the test must accurately characterize low- and high-risk patients; (2) the incidence of the cardiac outcome “at risk” must be great enough to provide the assay with a sufficient positive predictive value; (3) the assay must be practical for use in a clinical setting; and (4) the identification of at-risk individuals should result in a change in therapy that improves outcome. A large body of data has established that there is a wide variability in interindividual responses to clopidogrel, as
Cardiac Catheterization Laboratory, Division of Cardiovascular Diseases, Scripps Clinic, La Jolla, California, USA. Statement of author disclosure: Please see the Author Disclosures section at the end of this article. Address for reprints: Matthew J. Price, MD, Cardiac Catheterization Laboratory, Division of Cardiovascular Diseases, Scripps Clinic, 10666 North Torrey Pines Road, Maildrop S1056, La Jolla, California 92037. E-mail address: [email protected]. 0002-9149/09/$ – see front matter © 2009 Published by Elsevier Inc. doi:10.1016/j.amjcard.2008.11.022
discussed by other authors in this supplement.3b–3d The effects of 600-mg loading doses and 75–mg/day maintenance doses of clopidogrel on platelet aggregation are illustrated in Figure 1.2,3a After 2 hours of exposure to 600 mg of clopidogrel in clopidogrel-naive individuals, plasma samples from some individuals exhibited a marked reduction in platelet aggregation whereas others did not. A 900-mg loading dose does not appear to increase the magnitude of platelet inhibition or decrease the variability in response among individuals,4 probably because of limited intestinal absorption.5 However, increasing the daily maintenance dose of clopidogrel from 75 to 150 mg/day appears to increase inhibition in patients with high residual reactivity after standard dosing, although the effects are not uniform and a considerable number of patients continue to show inadequate platelet inhibition.6,7
Platelet Function Monitoring in Clinical Practice Given the range of responses to clopidogrel, is it possible to determine whether a particular platelet function test can help risk stratify patients? One approach is to perform a receiver-operating characteristic (ROC) curve analysis, which can assess the ability of a diagnostic test to discriminate between people with and without disease or, in this case, between patients with and without adverse outcomes. ROC curve analysis can also be used to define a potentially optimal threshold or cutoff point (by maximizing sensitivity and specificity) that can be subsequently verified in a larger patient population. Although only a few studies have performed such analyses with platelet function assays, the www.AJConline.org
36A
The American Journal of Cardiology (www.AJConline.org) Vol 103 (3A) February 2, 2009
Figure 1. Variability in clopidogrel response. (A) Change in adenosine-diphosphate–induced platelet aggregation 75-mg chronic dosing. (B) Maximal aggregation 5 mol/L ADP (%) after 600-mg loading dose. Cath ⫽ catheterization. (Adapted with permission from Circulation2 and J Am Coll Cardiol.3a)
Figure 2. Platelet mapping and ischemic outcomes after percutaneous coronary intervention (PCI) in patients receiving chronic clopidogrel therapy (N ⫽ 100; n ⫽ 75 for elective PCI). End point was cardiovascular death/myocardial infarction/target vessel revascularization/cardiovascular accident/rehospitalization for ischemia at 1 year. Findings included an overall event rate of 23%, a positive predictive value (predefined high platelet reactivity) of 73%, and a negative predictive value (predefined high platelet reactivity) of 91%. High reactivity was demonstrated at ⬎70% aggregation by the TEG Platelet Mapping (Haemoscope, Niles, IL) system (2 mol/L adenosine diphosphate). The optimal cutoff by receiver-operating characteristic was not reported. AUC ⫽ area under the curve; LTA ⫽ light transmittance aggregometry; TEG ⫽ thromboelastography. (Adapted with permission from J Am Coll Cardiol.10)
results of these studies highlight the potential benefits and pitfalls of platelet function monitoring in daily practice. Angiolillo and colleagues6 examined the relationship between adenosine diphosphate (ADP)-induced platelet reactivity measured by light transmittance aggregometry and the incidence of the composite end point of cardiovascular death, ST-segment elevation myocardial infarction (MI), non–ST-segment elevation MI, or stroke in stable patients with coronary artery disease and diabetes mellitus who were being treated with maintenance clopidogrel therapy.6 The overall 2-year event rate was 19.6%. The optimal cutoff of reactivity by ROC analysis was 62% maximal aggregation, which provided a sensitivity of 46% and a specificity of 84%, resulting in a moderately low positive predictive value of 41% but a negative predictive value of 86%. Vasodilator-stimulated phosphoprotein (VASP) analysis is another approach used to measure clopidogrel responsive-
ness, although its clinical utility is limited because the technique is based on flow cytometry. Blindt et al8 examined the association between the VASP platelet reactivity index post clopidogrel and subsequent stent thrombosis in 99 high-risk patients. The patients with angiographically confirmed stent thrombosis had a higher mean VASP platelet reactivity index than did patients without thrombosis; the optimal cutoff by ROC curve analysis was a VASP platelet reactivity index of 48%. Interestingly, Bonello et al9 showed in a small randomized study of patients undergoing PCI that adjusting the clopidogrel loading dose, according to VASP platelet monitoring to achieve a VASP platelet reactivity index of ⬍50%, was safe and may be associated with a lower rate of major adverse cardiac events over a 30-day follow-up period. Bliden and coworkers10 examined the ability of the Thromboelastograph (TEG) Hemostasis Analyzer with Plate-
Price/Monitoring Platelet Function to Reduce the Risk of Ischemic and Bleeding Complications
let Mapping Assay (Haemonetics, Braintree, MA) to discriminate between patients with and without subsequent ischemic outcomes after PCI (Figure 2). The overall rate of major adverse cardiovascular events (cardiovascular death, nonfatal MI, target vessel revascularization, stroke, non– target vessel revascularization, rehospitalization with ischemia) at 1 year of follow-up was 23%. The optimal cutoff of this assay calculated by ROC was not reported, but the investigators’ predefined cutoff for high on-treatment reactivity (⬎70% aggregation by the TEG platelet mapping system using 2-mol ADP) provided a positive predictive value of 73% and a negative predictive value of 91% for the combined end point. A prospective study was performed by myself and others11 that examined the association between postclopidogrel platelet reactivity measured by the point-of-care VerifyNow–P2Y12 assay (Accumetrics, San Diego, CA) and thrombotic outcomes in patients undergoing drug-eluting stent (DES) implantation. The primary study end point was a composite of cardiovascular death, nonfatal MI, and definite/probable stent thrombosis at 6-month follow-up. By ROC analysis, the optimal cutoff of posttreatment reactivity with the VerifyNow P2Y12 assay was 235 P2Y12 reaction units; this cutoff was consistent with the boundary of the upper tertile of reactivity in the population studied. This threshold for nonresponsiveness provided a sensitivity of 78% and a specificity of 69%. However, because the event rate in the overall population was only 2.3%, the positive predictive value of the assay was held to only 5.8%. These preliminary studies illustrate the challenge for any platelet function assay to adequately stratify patients for risk of a rarely occurring event. A low positive predictive value will be observed with any platelet function test used to “screen” for the risk of stent thrombosis because it is infrequent, and a limited response to clopidogrel is not uncommon. For example, assume a generous stent thrombosis rate of 2% in a given population and that a given platelet function assay has an excellent sensitivity and specificity of 85%. For this hypothetical platelet function test, the positive predictive value for stent thrombosis would be only 10% because many patients exhibit high reactivity post clopidogrel (or have a low clopidogrel response) but will not go on to experience a thrombotic event. However, because of the high morbidity and mortality associated with stent thrombosis, such a screening test may still be useful (although perhaps not cost-effective) if a therapeutic intervention (eg, providing more aggressive platelet inhibition) in a nonresponder can be performed safely. As an analogy, air bags are required in new cars, although traffic fatalities are rare relative to the amount of miles traveled per individual per year. Although the positive predictive value of the platelet function assay in Price et al11 was low, we found the negative predictive value of the optimal cutoff of the VerifyNow P2Y12 assay was high (99%). This suggests that platelet function monitoring may prove most useful in identifying patients who are at low rather than high risk for stent
37A
thrombosis by providing the clinician with confidence regarding the effectiveness of the patient’s antiplatelet regimen and the decision to place a DES. If nonresponsiveness to antiplatelet therapy is associated with an increased risk of thrombotic events, it would seem logical to monitor in vitro responses to clopidogrel when it is administered in conjunction with stent implantation and to take appropriate action to minimize the risk of thrombotic events in low responders. According to the 2005 American College of Cardiology/American Heart Association/Society for Cardiovascular Angiography and Interventions (ACC/ AHA/SCAI) guidelines for PCI, “In patients in whom subacute thrombosis may be catastrophic or lethal . . . platelet aggregation studies may be considered and the [maintenance] dose of clopidogrel increased to 150 mg per day if ⬍50% inhibition of platelet inhibition is demonstrated.”12 This recommendation is a clarion call to the importance of platelet reactivity in relation to patient outcomes. However, the recommendation (1) is based on a very minimal amount of data on the appropriate threshold for “nonresponsiveness” with respect to the end point of stent thrombosis; (2) is derived from studies with relatively small sample sizes; (3) is focused on an end point (stent thrombosis) that, given its low incidence, will be associated with a poor positive predictive value for almost any test; (4) suggests that platelet reactivity should be measured using light transmittance aggregometry, which is not widely available nor clinically practical; and (5) suggests an antiplatelet dosing regimen that has not been adequately studied. So where do we go from here?
Platelet Monitoring: Goalposts for the Future First, each point-of-care or automated platelet function assay must be tested in a prospective fashion to determine whether the assay can identify at-risk patients within a particular patient population (eg, patients with acute coronary syndromes) for a particular adverse outcome (eg, periprocedural MI) through techniques such as ROC curve analysis (Figure 3).11 Appropriate cutoffs for optimal sensitivity and specificity that can be used in clinical practice should be determined. Because the incidence of stent thrombosis is low (a meta-analysis of 3,445 patients studied for as long as 3 years after implantation of paclitaxel stents showed it to be ⬍2%13), and the frequency of poor clopidogrel responsiveness is high, no test is likely to provide a high positive predictive value. Prospective, preferably randomized trials must therefore determine the safety, efficacy, and cost-effectiveness of more aggressive antiplatelet therapy in patients identified as being at high risk for thrombotic events. Several studies that fulfill these criteria are planned or ongoing. The Assessment of Dual AntiPlatelet Therapy with Drug-Eluting Stents (ADAPT-DES) thrombosis study is a large, 11,000-patient, prospective registry study that will
38A
The American Journal of Cardiology (www.AJConline.org) Vol 103 (3A) February 2, 2009
Figure 3. The receiver-operating characteristic curve analysis of the VerifyNow–P2Y12 assay (Accumetrics, San Diego, CA): 6-month incidence of cardiovascular death, myocardial infarction, or stent thrombosis after drug-eluting stent placement (area under the curve, 0.71; p ⫽ 0.03). The optimal cutoff was 235 P2Y12 reaction units (the upper tertile was defined by P2Y12 reaction units ⬎231). The overall event rate was 2.3%; sensitivity, 78%; specificity, 69%, positive predictive value, 5.8%; and negative predictive value, 99%. (Adapted with permission from Eur Heart J.11)
relate the results of the VerifyNow P2Y12 assay to outcomes, including a prespecified end point of stent thrombosis.14 ADAPT-DES will assess the ability of the VerifyNow P2Y12 assay to discriminate between patients with and without various cardiac events after PCI and determine the appropriate thresholds of “nonresponsiveness” for each of these events. The Gauging Responsiveness with a VerifyNow P2Y12 Assay–Impact on Thrombosis and Safety (GRAVITAS) trial will examine whether tailored antiplatelet therapy in patients who have high postclopidogrel reactivity, as determined by the VerifyNow P2Y12 assay, can safely reduce major adverse cardiovascular events after DES implantation.15 A cost-effectiveness analysis is also planned as part of this trial. Platelet function monitoring may also play an important role in reducing bleeding events in patients receiving antiplatelet therapy who require surgery. Recent exposure to clopidogrel before cardiac surgery increases periprocedural bleeding and the need for transfusions, surgical reexploration, and resource utilization.16 –18 In a small study based on light transmittance aggregometry, Chen and colleagues19 examined the relationship between platelet aggregation and bleeding during coronary bypass grafting and found that patients who had greater clopidogrel-induced platelet inhibition— especially those with ⬎60% inhibition—required a significantly higher number of platelet and red blood cell transfusions. Because variability in the response to clopidogrel is so common, the recovery of platelet function after the discontinuation of clopidogrel may also be variable. We used the VerifyNow–P2Y12 assay to examine the time to platelet functional recovery in healthy volunteers who had been receiving clopidogrel daily for ⱖ2 weeks.11 Change in P2Y12-mediated platelet reactivity from baseline (before clopidogrel exposure) was measured each day after clopidogrel discontinuation. By day 3 after discontinuation, ⬎50% of the subjects had recovered platelet function; at
day 5 after discontinuation, 4% of these subjects actually had persistent inhibition. This observation argues against a uniform approach to patients exposed to clopidogrel who need surgery. Potential applications of platelet function monitoring in this setting would be for patients with surgical anatomy who received a loading dose before diagnostic angiography and patients receiving maintenance clopidogrel therapy who require cardiac or noncardiac surgery. Platelet function monitoring in this context may allow a reduction in the number of days that patients receiving maintenance therapy must wait for cardiac surgery, a reduction in the duration of time that patients with a recently placed DES are off antiplatelet agents before noncardiac surgery, and an adequate prediction of the duration of bridging therapy required to reduce the risk of stent thrombosis in patients who have recently received a DES. Such a tailored approach to clopidogrel discontinuation requires further examination through prospective studies.
Conclusion The wide interindividual variability in the response to clopidogrel has been established by various platelet function assays. Although several studies have demonstrated a relation between nonresponsiveness and cardiac events, the sample sizes of these studies are relatively small. Moreover, the studies have differed with respect to the method of platelet function testing used, the definition/threshold of nonresponsiveness used, and the specific outcomes examined. Regardless of the ability of a given platelet function test to discriminate between patients who may or may not have a specific adverse outcome, the positive predictive value of the test will be low when an infrequent but important outcome, such as stent thrombosis, is studied in the setting of an antiplatelet agent such as clopidogrel that has a wide range of responses among individuals. Prospective
Price/Monitoring Platelet Function to Reduce the Risk of Ischemic and Bleeding Complications
studies must determine the optimal threshold for “nonresponsiveness,” and then this threshold must be validated in a large population. Finally, to be adopted into daily clinical practice, randomized trials must show that tailoring antiplatelet therapy in response to the findings of platelet function tests improves outcomes without significantly increasing bleeding.
8.
9.
Author Disclosures The author who contributed to this article has disclosed the following industry relationships. Matthew J. Price, MD, has served as a consultant to Accumetrics, Inc and Volcano Corporation; and has been a speaker for Boston Scientific Corporation, Cardis Corporation, and The Medicines Company. 1. Leon MB, Baim DS, Popma JJ, Gordon PC, Cutlip DE, Ho KK, Giambartolomei A, Diver DJ, Lasorda DM, Williams DO, Pocock SJ, Kuntz RE, for the Stent Anticoagulation Restenosis Study Investigators. A clinical trial comparing three antithrombotic-drug regimens after coronaryartery stenting. N Engl J Med 1998;339:1665–1671. 2. Hochholzer W, Trenk D, Frundi D, Blanke P, Fischer B, Andris K, Bestehorn HP, Buttner HJ, Neumann FJ. Time dependence of platelet inhibition after a 600-mg loading dose of clopidogrel in a large, unselected cohort of candidates for percutaneous coronary intervention. Circulation 2005;11:2560 –2564. 3a.Serebruany VL, Steinhubl SR, Berger PB, Malinin AI, Bhatt DL, Topol EJ. Variability in platelet responsiveness to clopidogrel among 544 individuals. J Am Coll Cardiol 2005;45:246 –251. 3b.Bhatt DL. Role of antiplatelet therapy across the spectrum of patients with coronary artery disease. Am J Cardiol 2009;103(suppl):11A– 19A. 3c.Angiolillo DJ. Variability in responsiveness to oral antiplatelet therapy. Am J Cardiol 2009;103(suppl):27A–34A. 3d.Angiolillo DJ, Bhatt DL, Gurbel PA, Jennings LK. Advances in antiplatelet therapy: agents in clinical development. Am J Cardiol 2009;103(suppl):40A–51A. 4. Price MJ, Coleman JL, Steinhubl SR, Wong GB, Cannon CP, Teirstein PS. Onset and offset of platelet inhibition after high-dose clopidogrel loading and standard daily therapy measured by a point-of-care assay in healthy volunteers. Am J Cardiol 2006;98:681– 684. 5. von Beckerath N, Taubert D, Pogatsa-Murray G, Schömig E, Kastrati A, Schömig A. Absorption, metabolization, and antiplatelet effects of 300-, 600-, and 900-mg loading doses of clopidogrel: results of the ISAR-CHOICE (Intracoronary Stenting and Antithrombotic Regimen: Choose Between 3 High Oral Doses for Immediate Clopidogrel Effect) Trial. Circulation 2005;112:2946 –2950. 6. Angiolillo DJ, Bernardo E, Sabate M, Jimenez-Quevedo P, Costa MA, Palazuelos J, Hernandez-Antolin R, Moreno R, Escaned J, Alfonso F, et al. Impact of platelet reactivity on cardiovascular outcomes in patients with type 2 diabetes mellitus and coronary artery disease. J Am Coll Cardiol 2007;50:1541–1547. 7. Angiolillo DJ, Costa MA, Shoemaker SB, Desai B, Bernardo E, Suzuki Y, Charlton RK, Zenni MM, Guzman LA, Bass TA. Functional
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
39A
effects of high clopidogrel maintenance dosing in patients with inadequate platelet inhibition on standard dose treatment. Am J Cardiol 2008;101:440 – 445. Blindt R, Stellbrink K, de Taeye A, Muller R, Kiefer P, Yagmur E, Weber C, Kelm M, Hoffmann R. The significance of vasodilatorstimulated phosphoprotein for risk stratification of stent thrombosis. Thromb Haemost 2007;98:1329 –1334. Bonello L, Camoin-Jau L, Arques S, Boyer C, Panagides D, Wittenberg O, Simeoni MC, Barragan P, Dignat-George F, Paganelli F. Adjusted clopidogrel loading doses according to vasodilator-stimulated phosphorylation index decrease rate of major adverse cardiovascular events in patients with clopidogrel resistance: a multicenter randomized prospective study. J Am Coll Cardiol 2008;51:1404 –1411. Bliden KP, DiChiara J, Tantry US, Bassi AK, Chaganti SK, Gurbel PA. Increased risk in patients with high platelet aggregation receiving chronic clopidogrel therapy undergoing percutaneous coronary intervention: is the current antiplatelet therapy adequate? J Am Coll Cardiol 2007;49:657– 666. Price MJ, Endemann S, Gollapudi RR, Valencia R, Stinis CT, Levisay JP, Ernst A, Sawhney NS, Schatz RA, Teirstein PS. Prognostic significance of post-clopidogrel platelet reactivity assessed by a point-ofcare assay on thrombotic events after drug-eluting stent implantation. Eur Heart J 2008;29:992–1000. Smith SC Jr, Feldman TE, Hirshfeld JW Jr, Jacobs AK, Kern MJ, King SB III, Morrison DA, O’Neill WW, Schaff HV, Whitlow PL, et al. ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/SCAI Writing Committee to Update the 2001 Guidelines for Percutaneous Coronary Intervention). J Am Coll Cardiol 2006;47:e1– e121. Ellis SG, Colombo A, Grube E, Popma J, Koglin J, Dawkins KD, Stone GW. Incidence, timing, and correlates of stent thrombosis with the polymeric paclitaxel drug-eluting stent: a TAXUS II, IV, V, and VI meta-analysis of 3,445 patients followed for up to 3 years. J Am Coll Cardiol 2007;49:1043–1051. Assessment of Dual AntiPlatelet Therapy With Drug Eluting Stents (ADAPT-DES). Available at: http://clinicaltrials.gov/ct2/show/ NCT00638794. Accessed November 29, 2008. GRAVITAS: Gauging Responsiveness With A VerifyNow AssayImpact On Thrombosis And Safety. Available at: http://clinicaltrials. gov/ct2/show/NCT00645918. Accessed November 29, 2008. Leong JY, Baker RA, Shah PJ, Cherian VK, Knight JL. Clopidogrel and bleeding after coronary artery bypass graft surgery. Ann Thorac Surg 2005;80:928 –933. Kapetanakis EI, Medlam DA, Boyce SW, Haile E, Hill PC, Dullum MK, Bafi AS, Petro KR, Corso PJ. Clopidogrel administration prior to coronary artery bypass grafting surgery: the cardiologist’s panacea or the surgeon’s headache? Eur Heart J 2005;26: 576 –583. Mehta RH, Roe MT, Mulgund J, Ohman EM, Cannon CP, Gibler WB, Pollack CV Jr, Smith SC Jr, Ferguson TB, Peterson ED. Acute clopidogrel use and outcomes in patients with non-ST-segment elevation acute coronary syndromes undergoing coronary artery bypass surgery. J Am Coll Cardiol 2006;48:281–286. Chen L, Bracey AW, Radovancevic R, Cooper JR Jr, Collard CD, Vaughn WK, Nussmeier NA. Clopidogrel and bleeding in patients undergoing elective coronary artery bypass grafting. J Thorac Cardiovasc Surg 2004;128:425– 431.
Dual antiplatelet therapy with aspirin and a thienopyridine is a cornerstone of the management of patients undergoing percutaneous coronary intervention (PCI) and coronary stent implantation. This approach was adopted after trials such as the Stent Antithrombotic Regimen Study (STARS) demonstrated a marked reduction in thrombotic events within 30 days after PCI in patients treated with aspirin and ticlopidine (0.5%) compared with patients who received aspirin alone (3.6%) or aspirin plus warfarin (2.7%).1 Since that time, however, it has become well established that the platelet inhibitory response to clopidogrel varies greatly among individuals.2,3a Several monitoring assays are currently available that can measure platelet reactivity after clopidogrel exposure. Yet, to be clinically useful, a platelet function assay must fulfill several criteria: (1) the test must accurately characterize low- and high-risk patients; (2) the incidence of the cardiac outcome “at risk” must be great enough to provide the assay with a sufficient positive predictive value; (3) the assay must be practical for use in a clinical setting; and (4) the identification of at-risk individuals should result in a change in therapy that improves outcome. A large body of data has established that there is a wide variability in interindividual responses to clopidogrel, as
Cardiac Catheterization Laboratory, Division of Cardiovascular Diseases, Scripps Clinic, La Jolla, California, USA. Statement of author disclosure: Please see the Author Disclosures section at the end of this article. Address for reprints: Matthew J. Price, MD, Cardiac Catheterization Laboratory, Division of Cardiovascular Diseases, Scripps Clinic, 10666 North Torrey Pines Road, Maildrop S1056, La Jolla, California 92037. E-mail address: [email protected]. 0002-9149/09/$ – see front matter © 2009 Published by Elsevier Inc. doi:10.1016/j.amjcard.2008.11.022
discussed by other authors in this supplement.3b–3d The effects of 600-mg loading doses and 75–mg/day maintenance doses of clopidogrel on platelet aggregation are illustrated in Figure 1.2,3a After 2 hours of exposure to 600 mg of clopidogrel in clopidogrel-naive individuals, plasma samples from some individuals exhibited a marked reduction in platelet aggregation whereas others did not. A 900-mg loading dose does not appear to increase the magnitude of platelet inhibition or decrease the variability in response among individuals,4 probably because of limited intestinal absorption.5 However, increasing the daily maintenance dose of clopidogrel from 75 to 150 mg/day appears to increase inhibition in patients with high residual reactivity after standard dosing, although the effects are not uniform and a considerable number of patients continue to show inadequate platelet inhibition.6,7
Platelet Function Monitoring in Clinical Practice Given the range of responses to clopidogrel, is it possible to determine whether a particular platelet function test can help risk stratify patients? One approach is to perform a receiver-operating characteristic (ROC) curve analysis, which can assess the ability of a diagnostic test to discriminate between people with and without disease or, in this case, between patients with and without adverse outcomes. ROC curve analysis can also be used to define a potentially optimal threshold or cutoff point (by maximizing sensitivity and specificity) that can be subsequently verified in a larger patient population. Although only a few studies have performed such analyses with platelet function assays, the www.AJConline.org
36A
The American Journal of Cardiology (www.AJConline.org) Vol 103 (3A) February 2, 2009
Figure 1. Variability in clopidogrel response. (A) Change in adenosine-diphosphate–induced platelet aggregation 75-mg chronic dosing. (B) Maximal aggregation 5 mol/L ADP (%) after 600-mg loading dose. Cath ⫽ catheterization. (Adapted with permission from Circulation2 and J Am Coll Cardiol.3a)
Figure 2. Platelet mapping and ischemic outcomes after percutaneous coronary intervention (PCI) in patients receiving chronic clopidogrel therapy (N ⫽ 100; n ⫽ 75 for elective PCI). End point was cardiovascular death/myocardial infarction/target vessel revascularization/cardiovascular accident/rehospitalization for ischemia at 1 year. Findings included an overall event rate of 23%, a positive predictive value (predefined high platelet reactivity) of 73%, and a negative predictive value (predefined high platelet reactivity) of 91%. High reactivity was demonstrated at ⬎70% aggregation by the TEG Platelet Mapping (Haemoscope, Niles, IL) system (2 mol/L adenosine diphosphate). The optimal cutoff by receiver-operating characteristic was not reported. AUC ⫽ area under the curve; LTA ⫽ light transmittance aggregometry; TEG ⫽ thromboelastography. (Adapted with permission from J Am Coll Cardiol.10)
results of these studies highlight the potential benefits and pitfalls of platelet function monitoring in daily practice. Angiolillo and colleagues6 examined the relationship between adenosine diphosphate (ADP)-induced platelet reactivity measured by light transmittance aggregometry and the incidence of the composite end point of cardiovascular death, ST-segment elevation myocardial infarction (MI), non–ST-segment elevation MI, or stroke in stable patients with coronary artery disease and diabetes mellitus who were being treated with maintenance clopidogrel therapy.6 The overall 2-year event rate was 19.6%. The optimal cutoff of reactivity by ROC analysis was 62% maximal aggregation, which provided a sensitivity of 46% and a specificity of 84%, resulting in a moderately low positive predictive value of 41% but a negative predictive value of 86%. Vasodilator-stimulated phosphoprotein (VASP) analysis is another approach used to measure clopidogrel responsive-
ness, although its clinical utility is limited because the technique is based on flow cytometry. Blindt et al8 examined the association between the VASP platelet reactivity index post clopidogrel and subsequent stent thrombosis in 99 high-risk patients. The patients with angiographically confirmed stent thrombosis had a higher mean VASP platelet reactivity index than did patients without thrombosis; the optimal cutoff by ROC curve analysis was a VASP platelet reactivity index of 48%. Interestingly, Bonello et al9 showed in a small randomized study of patients undergoing PCI that adjusting the clopidogrel loading dose, according to VASP platelet monitoring to achieve a VASP platelet reactivity index of ⬍50%, was safe and may be associated with a lower rate of major adverse cardiac events over a 30-day follow-up period. Bliden and coworkers10 examined the ability of the Thromboelastograph (TEG) Hemostasis Analyzer with Plate-
Price/Monitoring Platelet Function to Reduce the Risk of Ischemic and Bleeding Complications
let Mapping Assay (Haemonetics, Braintree, MA) to discriminate between patients with and without subsequent ischemic outcomes after PCI (Figure 2). The overall rate of major adverse cardiovascular events (cardiovascular death, nonfatal MI, target vessel revascularization, stroke, non– target vessel revascularization, rehospitalization with ischemia) at 1 year of follow-up was 23%. The optimal cutoff of this assay calculated by ROC was not reported, but the investigators’ predefined cutoff for high on-treatment reactivity (⬎70% aggregation by the TEG platelet mapping system using 2-mol ADP) provided a positive predictive value of 73% and a negative predictive value of 91% for the combined end point. A prospective study was performed by myself and others11 that examined the association between postclopidogrel platelet reactivity measured by the point-of-care VerifyNow–P2Y12 assay (Accumetrics, San Diego, CA) and thrombotic outcomes in patients undergoing drug-eluting stent (DES) implantation. The primary study end point was a composite of cardiovascular death, nonfatal MI, and definite/probable stent thrombosis at 6-month follow-up. By ROC analysis, the optimal cutoff of posttreatment reactivity with the VerifyNow P2Y12 assay was 235 P2Y12 reaction units; this cutoff was consistent with the boundary of the upper tertile of reactivity in the population studied. This threshold for nonresponsiveness provided a sensitivity of 78% and a specificity of 69%. However, because the event rate in the overall population was only 2.3%, the positive predictive value of the assay was held to only 5.8%. These preliminary studies illustrate the challenge for any platelet function assay to adequately stratify patients for risk of a rarely occurring event. A low positive predictive value will be observed with any platelet function test used to “screen” for the risk of stent thrombosis because it is infrequent, and a limited response to clopidogrel is not uncommon. For example, assume a generous stent thrombosis rate of 2% in a given population and that a given platelet function assay has an excellent sensitivity and specificity of 85%. For this hypothetical platelet function test, the positive predictive value for stent thrombosis would be only 10% because many patients exhibit high reactivity post clopidogrel (or have a low clopidogrel response) but will not go on to experience a thrombotic event. However, because of the high morbidity and mortality associated with stent thrombosis, such a screening test may still be useful (although perhaps not cost-effective) if a therapeutic intervention (eg, providing more aggressive platelet inhibition) in a nonresponder can be performed safely. As an analogy, air bags are required in new cars, although traffic fatalities are rare relative to the amount of miles traveled per individual per year. Although the positive predictive value of the platelet function assay in Price et al11 was low, we found the negative predictive value of the optimal cutoff of the VerifyNow P2Y12 assay was high (99%). This suggests that platelet function monitoring may prove most useful in identifying patients who are at low rather than high risk for stent
37A
thrombosis by providing the clinician with confidence regarding the effectiveness of the patient’s antiplatelet regimen and the decision to place a DES. If nonresponsiveness to antiplatelet therapy is associated with an increased risk of thrombotic events, it would seem logical to monitor in vitro responses to clopidogrel when it is administered in conjunction with stent implantation and to take appropriate action to minimize the risk of thrombotic events in low responders. According to the 2005 American College of Cardiology/American Heart Association/Society for Cardiovascular Angiography and Interventions (ACC/ AHA/SCAI) guidelines for PCI, “In patients in whom subacute thrombosis may be catastrophic or lethal . . . platelet aggregation studies may be considered and the [maintenance] dose of clopidogrel increased to 150 mg per day if ⬍50% inhibition of platelet inhibition is demonstrated.”12 This recommendation is a clarion call to the importance of platelet reactivity in relation to patient outcomes. However, the recommendation (1) is based on a very minimal amount of data on the appropriate threshold for “nonresponsiveness” with respect to the end point of stent thrombosis; (2) is derived from studies with relatively small sample sizes; (3) is focused on an end point (stent thrombosis) that, given its low incidence, will be associated with a poor positive predictive value for almost any test; (4) suggests that platelet reactivity should be measured using light transmittance aggregometry, which is not widely available nor clinically practical; and (5) suggests an antiplatelet dosing regimen that has not been adequately studied. So where do we go from here?
Platelet Monitoring: Goalposts for the Future First, each point-of-care or automated platelet function assay must be tested in a prospective fashion to determine whether the assay can identify at-risk patients within a particular patient population (eg, patients with acute coronary syndromes) for a particular adverse outcome (eg, periprocedural MI) through techniques such as ROC curve analysis (Figure 3).11 Appropriate cutoffs for optimal sensitivity and specificity that can be used in clinical practice should be determined. Because the incidence of stent thrombosis is low (a meta-analysis of 3,445 patients studied for as long as 3 years after implantation of paclitaxel stents showed it to be ⬍2%13), and the frequency of poor clopidogrel responsiveness is high, no test is likely to provide a high positive predictive value. Prospective, preferably randomized trials must therefore determine the safety, efficacy, and cost-effectiveness of more aggressive antiplatelet therapy in patients identified as being at high risk for thrombotic events. Several studies that fulfill these criteria are planned or ongoing. The Assessment of Dual AntiPlatelet Therapy with Drug-Eluting Stents (ADAPT-DES) thrombosis study is a large, 11,000-patient, prospective registry study that will
38A
The American Journal of Cardiology (www.AJConline.org) Vol 103 (3A) February 2, 2009
Figure 3. The receiver-operating characteristic curve analysis of the VerifyNow–P2Y12 assay (Accumetrics, San Diego, CA): 6-month incidence of cardiovascular death, myocardial infarction, or stent thrombosis after drug-eluting stent placement (area under the curve, 0.71; p ⫽ 0.03). The optimal cutoff was 235 P2Y12 reaction units (the upper tertile was defined by P2Y12 reaction units ⬎231). The overall event rate was 2.3%; sensitivity, 78%; specificity, 69%, positive predictive value, 5.8%; and negative predictive value, 99%. (Adapted with permission from Eur Heart J.11)
relate the results of the VerifyNow P2Y12 assay to outcomes, including a prespecified end point of stent thrombosis.14 ADAPT-DES will assess the ability of the VerifyNow P2Y12 assay to discriminate between patients with and without various cardiac events after PCI and determine the appropriate thresholds of “nonresponsiveness” for each of these events. The Gauging Responsiveness with a VerifyNow P2Y12 Assay–Impact on Thrombosis and Safety (GRAVITAS) trial will examine whether tailored antiplatelet therapy in patients who have high postclopidogrel reactivity, as determined by the VerifyNow P2Y12 assay, can safely reduce major adverse cardiovascular events after DES implantation.15 A cost-effectiveness analysis is also planned as part of this trial. Platelet function monitoring may also play an important role in reducing bleeding events in patients receiving antiplatelet therapy who require surgery. Recent exposure to clopidogrel before cardiac surgery increases periprocedural bleeding and the need for transfusions, surgical reexploration, and resource utilization.16 –18 In a small study based on light transmittance aggregometry, Chen and colleagues19 examined the relationship between platelet aggregation and bleeding during coronary bypass grafting and found that patients who had greater clopidogrel-induced platelet inhibition— especially those with ⬎60% inhibition—required a significantly higher number of platelet and red blood cell transfusions. Because variability in the response to clopidogrel is so common, the recovery of platelet function after the discontinuation of clopidogrel may also be variable. We used the VerifyNow–P2Y12 assay to examine the time to platelet functional recovery in healthy volunteers who had been receiving clopidogrel daily for ⱖ2 weeks.11 Change in P2Y12-mediated platelet reactivity from baseline (before clopidogrel exposure) was measured each day after clopidogrel discontinuation. By day 3 after discontinuation, ⬎50% of the subjects had recovered platelet function; at
day 5 after discontinuation, 4% of these subjects actually had persistent inhibition. This observation argues against a uniform approach to patients exposed to clopidogrel who need surgery. Potential applications of platelet function monitoring in this setting would be for patients with surgical anatomy who received a loading dose before diagnostic angiography and patients receiving maintenance clopidogrel therapy who require cardiac or noncardiac surgery. Platelet function monitoring in this context may allow a reduction in the number of days that patients receiving maintenance therapy must wait for cardiac surgery, a reduction in the duration of time that patients with a recently placed DES are off antiplatelet agents before noncardiac surgery, and an adequate prediction of the duration of bridging therapy required to reduce the risk of stent thrombosis in patients who have recently received a DES. Such a tailored approach to clopidogrel discontinuation requires further examination through prospective studies.
Conclusion The wide interindividual variability in the response to clopidogrel has been established by various platelet function assays. Although several studies have demonstrated a relation between nonresponsiveness and cardiac events, the sample sizes of these studies are relatively small. Moreover, the studies have differed with respect to the method of platelet function testing used, the definition/threshold of nonresponsiveness used, and the specific outcomes examined. Regardless of the ability of a given platelet function test to discriminate between patients who may or may not have a specific adverse outcome, the positive predictive value of the test will be low when an infrequent but important outcome, such as stent thrombosis, is studied in the setting of an antiplatelet agent such as clopidogrel that has a wide range of responses among individuals. Prospective
Price/Monitoring Platelet Function to Reduce the Risk of Ischemic and Bleeding Complications
studies must determine the optimal threshold for “nonresponsiveness,” and then this threshold must be validated in a large population. Finally, to be adopted into daily clinical practice, randomized trials must show that tailoring antiplatelet therapy in response to the findings of platelet function tests improves outcomes without significantly increasing bleeding.
8.
9.
Author Disclosures The author who contributed to this article has disclosed the following industry relationships. Matthew J. Price, MD, has served as a consultant to Accumetrics, Inc and Volcano Corporation; and has been a speaker for Boston Scientific Corporation, Cardis Corporation, and The Medicines Company. 1. Leon MB, Baim DS, Popma JJ, Gordon PC, Cutlip DE, Ho KK, Giambartolomei A, Diver DJ, Lasorda DM, Williams DO, Pocock SJ, Kuntz RE, for the Stent Anticoagulation Restenosis Study Investigators. A clinical trial comparing three antithrombotic-drug regimens after coronaryartery stenting. N Engl J Med 1998;339:1665–1671. 2. Hochholzer W, Trenk D, Frundi D, Blanke P, Fischer B, Andris K, Bestehorn HP, Buttner HJ, Neumann FJ. Time dependence of platelet inhibition after a 600-mg loading dose of clopidogrel in a large, unselected cohort of candidates for percutaneous coronary intervention. Circulation 2005;11:2560 –2564. 3a.Serebruany VL, Steinhubl SR, Berger PB, Malinin AI, Bhatt DL, Topol EJ. Variability in platelet responsiveness to clopidogrel among 544 individuals. J Am Coll Cardiol 2005;45:246 –251. 3b.Bhatt DL. Role of antiplatelet therapy across the spectrum of patients with coronary artery disease. Am J Cardiol 2009;103(suppl):11A– 19A. 3c.Angiolillo DJ. Variability in responsiveness to oral antiplatelet therapy. Am J Cardiol 2009;103(suppl):27A–34A. 3d.Angiolillo DJ, Bhatt DL, Gurbel PA, Jennings LK. Advances in antiplatelet therapy: agents in clinical development. Am J Cardiol 2009;103(suppl):40A–51A. 4. Price MJ, Coleman JL, Steinhubl SR, Wong GB, Cannon CP, Teirstein PS. Onset and offset of platelet inhibition after high-dose clopidogrel loading and standard daily therapy measured by a point-of-care assay in healthy volunteers. Am J Cardiol 2006;98:681– 684. 5. von Beckerath N, Taubert D, Pogatsa-Murray G, Schömig E, Kastrati A, Schömig A. Absorption, metabolization, and antiplatelet effects of 300-, 600-, and 900-mg loading doses of clopidogrel: results of the ISAR-CHOICE (Intracoronary Stenting and Antithrombotic Regimen: Choose Between 3 High Oral Doses for Immediate Clopidogrel Effect) Trial. Circulation 2005;112:2946 –2950. 6. Angiolillo DJ, Bernardo E, Sabate M, Jimenez-Quevedo P, Costa MA, Palazuelos J, Hernandez-Antolin R, Moreno R, Escaned J, Alfonso F, et al. Impact of platelet reactivity on cardiovascular outcomes in patients with type 2 diabetes mellitus and coronary artery disease. J Am Coll Cardiol 2007;50:1541–1547. 7. Angiolillo DJ, Costa MA, Shoemaker SB, Desai B, Bernardo E, Suzuki Y, Charlton RK, Zenni MM, Guzman LA, Bass TA. Functional
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
39A
effects of high clopidogrel maintenance dosing in patients with inadequate platelet inhibition on standard dose treatment. Am J Cardiol 2008;101:440 – 445. Blindt R, Stellbrink K, de Taeye A, Muller R, Kiefer P, Yagmur E, Weber C, Kelm M, Hoffmann R. The significance of vasodilatorstimulated phosphoprotein for risk stratification of stent thrombosis. Thromb Haemost 2007;98:1329 –1334. Bonello L, Camoin-Jau L, Arques S, Boyer C, Panagides D, Wittenberg O, Simeoni MC, Barragan P, Dignat-George F, Paganelli F. Adjusted clopidogrel loading doses according to vasodilator-stimulated phosphorylation index decrease rate of major adverse cardiovascular events in patients with clopidogrel resistance: a multicenter randomized prospective study. J Am Coll Cardiol 2008;51:1404 –1411. Bliden KP, DiChiara J, Tantry US, Bassi AK, Chaganti SK, Gurbel PA. Increased risk in patients with high platelet aggregation receiving chronic clopidogrel therapy undergoing percutaneous coronary intervention: is the current antiplatelet therapy adequate? J Am Coll Cardiol 2007;49:657– 666. Price MJ, Endemann S, Gollapudi RR, Valencia R, Stinis CT, Levisay JP, Ernst A, Sawhney NS, Schatz RA, Teirstein PS. Prognostic significance of post-clopidogrel platelet reactivity assessed by a point-ofcare assay on thrombotic events after drug-eluting stent implantation. Eur Heart J 2008;29:992–1000. Smith SC Jr, Feldman TE, Hirshfeld JW Jr, Jacobs AK, Kern MJ, King SB III, Morrison DA, O’Neill WW, Schaff HV, Whitlow PL, et al. ACC/AHA/SCAI 2005 guideline update for percutaneous coronary intervention: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (ACC/AHA/SCAI Writing Committee to Update the 2001 Guidelines for Percutaneous Coronary Intervention). J Am Coll Cardiol 2006;47:e1– e121. Ellis SG, Colombo A, Grube E, Popma J, Koglin J, Dawkins KD, Stone GW. Incidence, timing, and correlates of stent thrombosis with the polymeric paclitaxel drug-eluting stent: a TAXUS II, IV, V, and VI meta-analysis of 3,445 patients followed for up to 3 years. J Am Coll Cardiol 2007;49:1043–1051. Assessment of Dual AntiPlatelet Therapy With Drug Eluting Stents (ADAPT-DES). Available at: http://clinicaltrials.gov/ct2/show/ NCT00638794. Accessed November 29, 2008. GRAVITAS: Gauging Responsiveness With A VerifyNow AssayImpact On Thrombosis And Safety. Available at: http://clinicaltrials. gov/ct2/show/NCT00645918. Accessed November 29, 2008. Leong JY, Baker RA, Shah PJ, Cherian VK, Knight JL. Clopidogrel and bleeding after coronary artery bypass graft surgery. Ann Thorac Surg 2005;80:928 –933. Kapetanakis EI, Medlam DA, Boyce SW, Haile E, Hill PC, Dullum MK, Bafi AS, Petro KR, Corso PJ. Clopidogrel administration prior to coronary artery bypass grafting surgery: the cardiologist’s panacea or the surgeon’s headache? Eur Heart J 2005;26: 576 –583. Mehta RH, Roe MT, Mulgund J, Ohman EM, Cannon CP, Gibler WB, Pollack CV Jr, Smith SC Jr, Ferguson TB, Peterson ED. Acute clopidogrel use and outcomes in patients with non-ST-segment elevation acute coronary syndromes undergoing coronary artery bypass surgery. J Am Coll Cardiol 2006;48:281–286. Chen L, Bracey AW, Radovancevic R, Cooper JR Jr, Collard CD, Vaughn WK, Nussmeier NA. Clopidogrel and bleeding in patients undergoing elective coronary artery bypass grafting. J Thorac Cardiovasc Surg 2004;128:425– 431.