The clinical pharmacology of antiplatelet drugs has been reviewed previously by the European Society of Cardiology (ESC) Task force and by the 8th American College of Chest Physicians (ACCP) Evidence-Based Clinical Practice Guidelines. Moreover, information on the efficacy and safety of antiplatelet drugs in the treatment and prevention of atherothrombosis is provided by collaborative meta-analyses of 287 secondary prevention trials and 6 primary prevention trials. The present document intends to provide practicing physicians with an updated instrument to guide their choice of the most suitable antiplatelet strategy for the individual patient at risk, or with different clinical manifestations, of atherothrombosis.
The clinical pharmacology of antiplatelet drugs has been reviewed previously by a European Society of Cardiology (ESC) task force1 and by the 8th American College of Chest Physicians (ACCP) Evidence-Based Clinical Practice Guidelines.2 Moreover, information on the efficacy and safety of antiplatelet drugs in the treatment and prevention of atherothrombosis is provided by collaborative meta-analyses of 287 secondary prevention trials and 6 primary prevention trials.3,4 The present document intends to provide practicing physicians with an updated instrument to guide their choice of the most suitable antiplatelet strategy for the individual patient at risk, or with different clinical manifestations, of atherothrombosis.
Platelets are vital components of normal haemostasis and key participants in atherothrombosis by virtue of their capacity to adhere to injured blood vessels and to accumulate at sites of injury.5 Although platelet adhesion and activation should be viewed as a physiological response to the fissuring or rupture of an atherosclerotic plaque, eventually contributing to its repair, uncontrolled progression of such a process through a series of self-sustaining amplification loops may lead to intraluminal thrombus formation, and vascular occlusion with transient or permanent ischaemia or necrosis. The major agonists, receptors, and effector systems participating in platelet activation are discussed and illustrated in the Supplementary material online, Figure S1. Currently available antiplatelet drugs (Table 1) interfere with some of the steps leading to platelet aggregation (Figure 1),1,2 and have a measurable impact on the risk of arterial thrombosis that cannot be dissociated from an increased risk of bleeding.1,2
Platelet activation and inhibition mechanisms and the sites of action of antiplatelet drugs. Platelet activation via multiple pathways leads to numerous responses including: shape change; dense granule secretion of ATP, 5HT, and ADP (which binds to P2Y1 and P2Y12 receptors, the latter playing a powerful role in amplification of platelet activation); α granule secretion of chemokines—which lead to leucocyte and endothelial cell activation—and coagulation factors; procoagulant changes in the platelet surface membrane supporting thrombin generation and activation of αIIbβ3 (GPIIb/IIIa) leading to platelet aggregation and outside-in signalling that further amplifies platelet activation. Release of NO and PGI2 from the endothelium reduces platelet reactivity. Adapted with permission from Storey RF. Biology and pharmacology of the platelet P2Y12 receptor. Current Pharmaceutical Design. 2006;12:1255–1259.
Mechanism of action of antiplatelet drugs
Aspirin induces a permanent functional defect in platelets, which can be detected clinically as a prolonged bleeding time. This appears to be primarily, if not exclusively, due to irreversible inactivation of a key enzyme in platelet arachidonate metabolism through acetylation of a critical serine residue near its catalytic site (Supplementary material online, Figure S2). This enzyme, cyclooxygenase (COX)-1, is responsible for the formation of prostaglandin (PG)H2, the precursor of thromboxane (TX)A2. The non-linear relationship between inactivation of platelet COX-1 and inhibition of TXA2-dependent platelet function by low-dose aspirin6 (Supplementary material online, Figure S3) has important implications: (i) a substantial reduction in platelet inhibition is associated with less than maximal inactivation of COX-1; (ii) recovery of platelet function is disproportionately rapid, occurring within 3–4 days upon drug withdrawal,6 (iii) the requirement for virtually complete and persistent inhibition of platelet COX-1 cannot be met by most traditional non-steroidal anti-inflammatory drugs (NSAIDs), allowing their COX-2-dependent cardiotoxicity to be unmasked.7 Moreover, inhibition of TXA2-dependent platelet function by aspirin leaves other platelet pathways [adenosine diphosphate (ADP)-P2Y12, thrombin-protease-activated receptor (PAR)-1] largely unaffected, thus providing a rationale for dual or triple antiplatelet therapy in high-risk settings.
Thienopyridines inhibit adenosine diphosphate (ADP)-dependent platelet function by irreversible modification of the platelet P2Y12 receptor8 through short-lived active metabolites, generated by liver cytochrome P-450 (CYP) isozymes, that form covalent bonds with critical cysteine residues within the receptor.9 The inhibition of ADP-dependent platelet function by clopidogrel is less predictable than the inhibition of TXA2-dependent platelet function by aspirin. Insufficient availability of the active metabolite of clopidogrel at conventional therapeutic doses results in incomplete inactivation of platelet P2Y12.10 Because of the linear relationship between P2Y12 inactivation and inhibition of ADP-dependent platelet aggregation, recovery of platelet function after drug withdrawal occurs linearly over 7–8 days as a function of platelet turnover.1 Genetic variation of the liver enzymes responsible for the metabolism of clopidogrel as well as drug–drug interactions [e.g. with proton pump inhibitors (PPIs)] are important determinants of the variable circulating levels of its active metabolite.11,12,13 These, in turn, are associated with variable inhibition of ADP-induced platelet aggregation and variable clinical response to clopidogrel treatment.11,12
Like ticlopidine and clopidogrel, prasugrel is a prodrug that is inactive in vitro.14 While equimolar concentrations of the active metabolites of clopidogrel and prasugrel result in similar levels of platelet inhibition in vitro, the markedly different amounts of each metabolite generated in vivo following a loading dose of clopidogrel (300 mg) or prasugrel (60 mg) result in ∼10-fold higher platelet exposure to the latter when compared with the former.14 This observation provides a pharmacokinetic basis for the faster, more profound and less variable inhibition of platelet function observed with prasugrel when compared with clopidogrel in healthy subjects14 as well as in patients with ischaemic heart disease.10,15,16 In contrast to clopidogrel, the lack of drug interaction potential and the apparent independence of CYP2C19 genetic variance result in a predictable antiplatelet response to prasugrel.15–17
Glycoprotein (GP)IIb/IIIa blockers
GPIIb/IIIa antagonists prevent fibrinogen binding to activated GPIIb/IIIa receptors and, thus, formation of fibrinogen bridges between platelets (reviewed in detail in ref. 2). Activation of GPIIb/IIIa constitutes the final common pathway of platelet aggregation. Currently, three GPIIb/IIIa blockers are available for intravenous administration: abciximab, eptifibatide, and tirofiban (reviewed in detail in ref.2).
Abciximab, a non-competitive inhibitor of GPIIb/IIIa, is the humanized chimeric Fab-fragment of the monoclonal mouse antibody 7E3. Abciximab cross-reacts with the αvβ3 integrin on endothelial cells and smooth muscle cells and with the αMβ2 integrin (CD11b/CD18) on granulocytes and monocytes.
Two small-molecule GPIIb/IIIa blockers act specifically on the αIIb-chain of GPIIb/IIIa: eptifibatide, a cyclic heptapeptide, and tirofiban, a non-peptide (‘peptidomimetic’) antagonist. Eptifibatide and tirofiban are competitive inhibitors. Their effect on platelet aggregation is closely linked to plasma concentrations.1 Owing to their short plasma half-lives, continuous infusion is needed for sustained platelet inhibition.1
Thromboxane receptor antagonists
Potent thromboxane receptor (TP) antagonists have been developed, including GR 32191, BMS-180291 (ifetroban), BM 13.177 (sulotroban), and S-18886 (terutroban). Despite the anti-thrombotic activity demonstrated in various animal species and the cardioprotective and antiatherogenic activities demonstrated in experimental models, these compounds have yielded disappointing results in phase 2 and 3 clinical trials.2 The clinical development of terutroban has been discontinued recently after an interim futility analysis revealed that superiority vs. low-dose aspirin on major vascular events in over 19 000 patients with recent cerebrovascular ischaemia was unlikely to be demonstrated.18
Reversible P2Y12 antagonists
Three direct-acting and reversible P2Y12 antagonists, ticagrelor (an oral agent), cangrelor (an intravenous agent), and elinogrel (available both as an intravenous and oral agent) are associated with rapid onset and offset of platelet inhibition.11,19,20 Unlike the thienopyridines, they do not require metabolic activation by the liver. Detailed molecular studies of ticagrelor have demonstrated that this selective P2Y12 inhibitor displays a non-competitive antagonism towards ADP-induced receptor activation, suggesting the existence of more than one ligand-binding site on P2Y12.21 Ticagrelor is rapidly absorbed and undergoes enzymatic transformation to at least one active metabolite.19 Peak plasma concentrations of ticagrelor and maximum platelet inhibition are achieved within 1–3 h after dosing. The plasma half-life of ticagrelor is 6–13 h, which dictates a twice daily regimen of administration.19
Thrombin receptor antagonists
Thrombin interacts with two platelet receptors, called PAR-1 and -4, that are activated through proteolytic cleavage (Figure 1).22 Protease-activated receptor-1 is the major human platelet receptor, exhibiting 10–100 times higher affinity for thrombin when compared with PAR-4.22 Two thrombin receptor antagonists (TRA) with PAR-1 selectivity are currently under clinical evaluation. Vorapaxar (SCH530348), a synthetic analogue of himbacine, is a potent oral TRA23,24 with half-life of 126–269 h, that inhibits platelet function for up to 4 weeks after its withdrawal.23,24 Two ongoing placebo-controlled phase 3 trials in high-risk acute coronary syndrome (ACS) patients25 and in stable patients at high cardiovascular risk26 have undergone recent changes recommended by their data and safety monitoring board. Another oral TRA, atopaxar (E5555), has recently completed phase 2 evaluation.27 This compound has a shorter half-life and faster recovery of platelet function after its withdrawal than vorapaxar.28 However, a dose-dependent increase in liver function abnormalities and QTc prolongation were noted in the dose-finding study of atopaxar.27
Dipyridamole is a pyrimidopyrimidine derivative with vasodilator and antiplatelet properties. Its mechanism of action as an antiplatelet agent has been a subject of controversy (reviewed in ref.2). The absorption of dipyridamole from conventional formulations is quite variable and may result in low systemic bioavailability of the drug. A modified-release formulation of dipyridamole with improved bioavailability has been developed and tested in association with low-dose aspirin. Dipyridamole is eliminated primarily by biliary excretion as a glucuronide conjugate and is subject to enterohepatic recirculation. A terminal half-life of 10 h has been reported. This is consistent with the b.i.d. regimen used in recent clinical studies.29,30
Cilostazol is a reversible type III phosphodiesterase inhibitor, with vasodilator and antiplatelet effects. It increases intraplatelet cAMP, reduces cellular adenosine uptake, and inhibits vascular smooth muscle cell proliferation.31 Added to a standard combination of aspirin plus clopidogrel, cilostazol 100 mg b.i.d. has been found to potentiate inhibition of ADP-induced platelet aggregation when compared with aspirin and clopidogrel.32
Patients who may benefit from different antiplatelet regimens
Single antiplatelet therapy
In the meta-analysis of the antithrombotic trialists'(ATT) collaboration,3 allocation of high-risk patients to a prolonged course of antiplatelet therapy reduced the combined outcome of non-fatal myocardial infarction (MI), non-fatal stroke or vascular death (‘serious vascular events’) by ∼25% compared with placebo. Non-fatal MI was reduced by one-third, non-fatal stroke by one-quarter, and vascular mortality by one-sixth. Absolute reductions in the risk of having a serious vascular event in different groups of high-risk patients are illustrated in Supplementary material online, Figure S4. In each of these high-risk categories, the absolute benefits substantially outweighed the absolute risks of major bleeding complications (Table 2).
Benefit/risk ratio of antiplatelet prophylaxis with aspirin in different settings
Number of patients in whom a major vascular event is avoided per 1000/year
Number of patients in whom a major GI bleeding event is caused per 1000/year
Men and women at low-cardiovascular risk
Chronic stable angina
Prior stroke or TIA
Prior myocardial infarction
aBenefits are calculated from randomized trial data reviewed in refs2,4 and depicted in Supplementary Figure 4.
bRisks of upper GI bleeding are estimated from a background rate of 1 event per 1000 per year in the general population of non-users and a relative risk of 2.0–3.0 associated with aspirin prophylaxis. Such an estimate assumes comparability of other risk factors for upper GI bleeding, such as age and concomitant use of NSAIDs, and may actually underestimate the absolute risk in an elderly population exposed to ‘primary’ prevention. The absolute excess of major extra-cranial bleeding complications in the ‘primary’ prevention trials reviewed in ref.4 ranged between 0.2 and 2.0 per 1000 patient-years. Modified from Patrono et al., Chest 20082.
It is interesting to compare the effects of low-dose aspirin in primary prevention with the well-known benefits in secondary prevention (Table 3).4 In the six primary prevention trials among 95 000 low-risk individuals, with mean follow-up 6.9 years, aspirin allocation yielded a 12% relative risk reduction in serious vascular events, from an annual rate of 0.57 to 0.51%.4 This effect was mainly due to a reduction in non-fatal MI, from 0.23 to 0.18% per year. The net effect on stroke was not significant, reflecting a small reduction in presumed ischaemic stroke and counterbalancing effects on haemorrhagic stroke and other stroke.4 There was no significant reduction in vascular mortality. Aspirin allocation increased gastrointestinal (GI) (or other extra-cranial) bleeds from 0.07 to 0.1% per year.4 The risks of serious vascular events and of major extra-cranial bleeds were predicted by the same independent risk factors (age, male gender, diabetes mellitus, current smoking, blood pressure, and body mass index) (Table 4), so individuals at higher risk of vascular complications also had a high risk of bleeding.4
Rate ratios (95% CI) associated with risk factors for selected outcomes in people with no known vascular disease in primary prevention trials
Major coronary event
Probably ischaemic stroke
Major extracranial bleed
Age (per decade)
Mean blood pressure (per 20 mmHg)b
Cholesterol (per 1 mmol/L)
Body mass index (per 5 kg/m2)
aAnalyses are stratified by trial. The relevance of male sex can therefore be assessed only in the two trials that included both men and women, so the 95% CIs for it are wide, particularly for stroke.
bMean of systolic and diastolic blood pressure. Associations with measured values are not corrected for the effects of regression dilution. Reproduced from antithrombotic trialists' (ATT) Collaboration. Lancet 20094.
While for secondary prevention the net benefits of adding aspirin to other, safer, preventive measures (e.g. statin therapy, antihypertensive therapy) would substantially exceed the bleeding hazards, irrespective of age and gender, in people without pre-existing vascular disease the benefits of adding long-term aspirin to other, safer, forms of primary prevention (e.g. statins and antihypertensive drugs) would be of similar magnitude as the hazards.4 Hence, the currently available trial results do not seem to justify general guidelines advocating the routine use of aspirin in all apparently healthy individuals above a moderate level of coronary risk, unless additional long-term benefits of antiplatelet therapy33 become established.
Ticlopidine and clopidogrel have been tested for superiority vs. aspirin in patients with a recent MI, and both drugs failed to demonstrate superiority in this setting.34,35 Both aspirin36–39 and ticlopidine40 have been shown to reduce by ∼50% the rate of MI or death in randomized controlled trials (RCTs) of patients with unstable angina.
Dual antiplatelet therapy
In the CURE study, blockade of both platelet COX-1 with aspirin and the platelet P2Y12 receptor with clopidogrel produced additive effects in patients with non-ST-elevation (NSTE)-ACS, by reducing the rate of the first primary outcome (a composite of cardiovascular death, non-fatal MI, or stroke) from 11.4 to 9.3% (RR, 0.80; 95% CI, 0.72–0.90, P < 0.001) when compared with aspirin alone, with no evidence of attenuation of the additional benefit over 12 months of treatment.41 As would be expected from more aggressive antiplatelet therapy, there were significantly more patients with major bleeding complications in the aspirin plus clopidogrel group than in the aspirin alone group (3.7 vs. 2.7%; P = 0.001), an effect in part related to the variable aspirin dose (75–325 mg daily) used in this trial41 (Supplementary material online, Table S1).
The clinical benefit of dual antiplatelet therapy vs. aspirin alone has been confirmed in patients undergoing PCI,42 in the short-term treatment of those presenting with an acute STEMI treated with fibrinolysis,43,44 and in patients with non-valvular atrial fibrillation.45 Although long-term treatment with aspirin and clopidogrel for 1 year after a STEMI is recommended by the ESC guidelines, irrespective of the acute reperfusion treatment, no studies have been performed in this regard.46
At present there is no sound evidence for defining the optimal duration of dual antiplatelet therapy after elective drug-eluting stent (DES) placement with current recommendations ranging between 3 and 12 months, depending on DES type.47 However, convincing data exist only for continuation up to 6 months. In the combined analysis of two trials performed by Park et al.,48 the use of dual antiplatelet therapy for a period longer than 12 months in patients who had received DES was not significantly more effective than aspirin monotherapy in reducing the rate of MI or cardiac death. Preliminary data from the PRODIGY trial, in which 6 months of clopidogrel was compared with 24 months of clopidogrel following coronary stent implantation, have shown a statistically significant two-fold increase in the bleeding rates without evidence of significant efficacy with the longer duration of clopidogrel and so do not support the use of long-term dual antiplatelet therapy following percutaneous coronary intervention (Valgimigli M, presented at ESC annual scientific congress, Paris, France, 30 August 2011). Several randomized studies, such as EXCELLENT, OPTIDUAL, ISAR-SAFE, and DAPT, are currently investigating this controversial issue. Until the results of these studies are available, the potential benefit of extended dual antiplatelet therapy needs to be weighed against the excess risk of major bleeding.
In contrast to the consistent finding of a favourable benefit/risk profile of dual antiplatelet therapy in ACS patients41,43,44 and those undergoing PCI,42 the same strategy has not proven advantageous when compared with clopidogrel alone in patients after a recent ischaemic stroke or TIA,49 or when compared with aspirin alone in patients at high risk for atherothrombotic events.50 Major bleeding was increased by dual antiplatelet therapy in both MATCH49 and CHARISMA.50
The potential of dual antiplatelet therapy with low-dose aspirin and new P2Y12 inhibitors in ACS has been shown by two large trials: the TRITON-TIMI 38 trial comparing prasugrel to clopidogrel in ACS patients undergoing PCI51 and the PLATO trial comparing ticagrelor to clopidogrel in high-risk ACS52 (Supplementary material online, Table S1). Both trials met their primary endpoint by demonstrating that the 12-month (ticagrelor) or 15-month (prasugrel) rate of major vascular events can be further reduced by 16% (ticagrelor) to 19% (prasugrel) over clopidogrel with a P2Y12 blocker that achieves faster, more profound and less variable inhibition of ADP-dependent platelet function than clopidogrel.51,52 Differences between the two trials are worth noting (Supplementary material online, Table S1). While the benefit of prasugrel vs. clopidogrel was largely confined to a reduction in non-fatal MI,51 the incremental benefit of ticagrelor translated into a significant reduction in vascular as well as total mortality.52 It is unknown whether the mortality reduction was entirely due to greater P2Y12 inhibition in a high-risk population, or whether reversibility of receptor binding and/or off-target effects of ticagrelor also contributed. It should be emphasized that this finding derives from a single trial, in which mortality was not the primary endpoint. P2Y12 blockade with prasugrel was associated with significant increases in both fatal and non-fatal major bleeds.51 Ticagrelor did not increase the total number of major bleeds but did increase non-fatal spontaneous [i.e. non-coronary artery bypass grafting (CABG)-related] bleeds and fatal intracranial haemorrhages.52 Coronary artery bypass grafting-related bleeds were not significantly different in PLATO, but increased in TRITON-TIMI 38. Three patient subsets appeared to be particularly prone to major bleeding with prasugrel: the elderly and the underweight (who derived no net benefit), and patients with prior cerebrovascular ischaemia (who derived net harm).51 Regardless of the mechanism(s), the increase in fatal haemorrhagic complications of prasugrel treatment may have masked any benefit of the drug on vascular mortality. Additional side effects of the novel agents include non-platelet related effects, with prasugrel associated with a higher rate of diagnosis of colonic neoplasms (13 vs. 4, P = 0.03), a finding which remains to be confirmed and may be a chance finding,51 and ticagrelor associated with more frequent episodes of dyspnoea and ventricular pauses than clopidogrel.52 No head-to-head comparison of prasugrel and ticagrelor has been performed yet.
Based on the TRITON-TIMI 38 results, prasugrel has been approved by both the FDA and EMA for the reduction in thrombotic cardiovascular events in ACS patients who undergo PCI. A black-boxed warning underscores the increased bleeding risk for patients 75 years of age or older and for patients undergoing urgent CABG. Prasugrel is contraindicated in patients with prior stroke or TIA. Because of the high probability of a false positive finding, the imbalance in newly diagnosed cancers is described in the adverse-reactions section of prasugrel's label but is not emphasized as a warning.53 However, the sponsors have a post-marketing requirement by the FDA to collect baseline and subsequent data on cancer in a large, ongoing clinical trial.53
Based on the PLATO results, ticagrelor has been approved by the EMA and FDA for the reduction in thrombotic cardiovascular events in ACS patients regardless of planned management strategy and including patients older than 75 years or with prior history of ischaemic stroke or TIA but excluding those with a history of intracranial haemorrhage. In the FDA label, a boxed warning indicates that maintenance doses of aspirin >100 mg reduce the effectiveness of ticagrelor and should be avoided.
Whereas the efficacy of GPIIb/IIIa receptor blockade in the setting of conservative treatment of ACS is marginal,54 this class of drugs substantially improved the safety of PCI in the 1990s. By preventing peri-procedural thrombotic events, GPIIb/IIIa blockade reduced the 30-day rate of major adverse cardiac events after PCI by up to 55%.55 Pooled analysis of five randomized trials including 11 612 patients yielded a relative risk reduction by GPIIb/IIIa blockade of 31% (95% CI, 23–39%) for the 30-day event rate.56 This early benefit was maintained for 3 years.57
These early studies have been recently challenged, because PCI was then performed without pre-treatment with thienopyridines (see Supplementary material online). Currently, the place of systematic GPIIb/IIIa blockade for primary PCI is uncertain within the context of concomitant antithrombotic treatment. There appears to be a role for provisional GPIIb/IIIa blockade in this setting, but efficacy and indications of this approach need further study. There is no sound evidence that upstream vs. in-cath-lab administration of GPIIb/IIIa blockers affords a beneficial effect that is clinically relevant.
Perioperative management of antiplatelet therapy
The practice of withdrawing antiplatelet therapy 7–10 days before surgery/endoscopy/biopsy is undergoing critical reappraisal. Patients with a clear indication for antiplatelet therapy have a three-fold, or greater, incidence of thrombotic events associated with discontinuation of antiplatelet drugs,58 overwhelming the estimated 1.5-fold bleeding hazard associated with perioperative drug continuation.59 Indeed, up to 10% of acute vascular events may be linked to antiplatelet therapy withdrawal.59 Moreover, the post-operative acute phase is thrombogenic, involving platelet hyper-reactivity, increased synthesis of coagulation factors, and hypo-fibrinolysis. Perioperative interruption of antiplatelet therapy should therefore be considered carefully and, if deemed necessary, carried out for the shortest possible time.60
Data from RCTs in patients undergoing ‘on pump’ cardiac surgery indicate that aspirin continuation increases reoperation rates ∼2.5-fold, without affecting transfusion rates.61 Post-operative blood loss increases as well, by ∼100 mL, but not for daily aspirin doses <325 mg.61 A comprehensive meta-analysis of non-cardiac surgery indicates that length of operation, length of hospitalization, severity of bleeding (with the exception of intracranial surgery), and need for transfusions are not increased by aspirin.59 Similarly, a randomized trial comparing 150 mg daily aspirin to placebo in patients undergoing transurethral prostatectomy showed increased post-operative blood loss (∼140 mL) with aspirin, but no significant differences in intra-operative bleeding, transfusion rates, or length of hospitalization.62
A risk-benefit assessment should be performed, based on the patient's thrombotic risk of stopping therapy against the haemorrhagic risk of continuing single or dual antiplatelet therapy (Table 5),63 particularly in patients with stents or with ACSs. Antiplatelet therapy may be withdrawn before surgery when it is prescribed for primary prevention (given the relatively low-thrombotic risk) or when bleeding in a closed space may be life-threatening or irreversible (e.g. intracranial, posterior eye).60,63 At the other extreme, patients receiving BMS within 6 weeks or DES within 6 months should not routinely discontinue dual antiplatelet therapy prior to surgery.63 When aspirin is maintained, a daily dose of 75–100 mg should be given. Resuming antiplatelet therapy is considered safe 12–24 h after adequate haemostasis has been achieved.
Proposal for perioperative antiplatelet management based on patient's risk of thrombosis vs. surgical bleeding risk
Patient's thrombotic risk
Surgical bleeding risk
Low: >9–12 months after uncomplicated ACS, DES, POBA, BMS, CABG
Medium: 7 weeks to 9–12 months after uncomplicated ACS, POBA, BMS, CABG; 7–12 months after DES, or high-risk stent
High: ≤6 weeks after ACS, POBA, BMS, CABG, or <9–12 months after their complications; ≤6 months after DES or high-risk stent
Low (transfusion usually not needed): general biopsies. skin, dental, anterior eye, minor general, minor orthopaedic, minor ENT surgery, endoscopy
Maintain low-dose aspirin
Maintain low-dose aspirin and P2Y12 blocker (if prescribed)
Maintain low-dose aspirin and P2Y12 blocker (if prescribed)
Medium (transfusion often required): cardiovascular, visceral, ENT, reconstructive, major orthopaedic, endoscopic urological surgery
Maintain low-dose aspirin
Maintain low-dose aspirin and P2Y12 blocker (if prescribed)
Maintain low-dose aspirin and P2Y12 blocker (if prescribed)
High: intracranial, spinal canal, posterior eye surgery. Possible bleed in closed space. Large expected blood loss
Withdraw aspirin for 3–5 days
Postpone elective surgery. If urgent, maintain low-dose aspirin for all but intracranial surgery. Withdraw P2Y12 blocker (if prescribed) for 5 daysa
Postpone non-vital surgery. If vital, maintain low-dose aspirin. Withdraw P2Y12 blocker (if prescribed) for 5 days.a Consider bridging with small molecule i.v. GPI
Individual characteristics that enhance bleeding risk (e.g. age, renal failure) are not considered here but are discussed elsewhere. Withdrawal of aspirin is recommended for only one, and or P2Y12 blockers for two, of the nine combinations. Because stroke patients may receive aspirin alone, aspirin + dipyridamole, or clopidogrel alone, their management is not discussed, nor are there sufficient data to make specific recommendations.
ACS, acute coronary syndrome; BMS, bare metal stent; CABG, coronary artery bypass graft; DES, drug-eluting stent; ENT, ear nose and throat; GPI, glycoprotein IIb/IIIa inhibitor. POBA, plain old balloon angioplasty.
aThe duration of P2Y12 blocker withdrawal may vary according to the individual agent; however, comparative data among the three available P2Y12 blockers are lacking.
Management of bleeding
In ACS patients, therapy-related major bleeding is associated with an increase in early and late morbidity and mortality64,65 and there is emerging evidence that this association may be in part causal.66,67 There is also solid data to indicate that bleeding is not unpredictable: there are well-characterized risk factors for bleeding, such as age, renal disease, female gender, low body weight (factors potentially associated with overdosing of antithrombotic drugs) which have led to the design of bleeding risk scores for stable68 and unstable coronary artery disease.69,70 Therefore, proper quantification of bleeding by use of appropriate standardized consensus definitions71 and identification of patients at risk through the use of bleeding risk scores are important to improve assessment and prevention. The design of algorithms to integrate the stratification of risk on the ischaemic/thrombotic side with that on the bleeding side72 is a key topic for future research.
When bleeding occurs, physicians are tempted to interrupt anti-platelet therapy. The withdrawal of platelet and coagulation inhibition may, however, re-exacerbate pro-thrombotic mechanisms with the consequence of precipitating additional ischaemic events.64,65 Withdrawal of only one antiplatelet agent in case of dual anti-platelet therapy may be more appropriate, although this suggestion is not based on randomized trial evidence.
Various effects, including the release of epinephrine and norepinephrine, angiotensin, endothelin-1, and vasopressin as well as the depletion in 2,3-diphosphoglyceric acid and nitric oxide, may explain why the transfusion of packed red blood cells is not as beneficial as expected. In fact, a restrictive transfusion strategy is associated with lesser ischaemic events, pulmonary oedema, and even mortality.73–75 Therefore, an aggressive transfusion strategy cannot be recommended and blood transfusions should be restricted to patients with a haematocrit value <25% or a Hb <80 g/L.76
Bleeding is sufficiently controlled by local measures in most cases, when it is attributable to a single local source. Few bleeding events require surgical interventions, and most of these interventions can safely be performed despite residual platelet inhibition. Antiplatelet therapy may be continued or reinstituted early after surgery. Platelet transfusions, FVIIa, aprotinin, or tranexamic acid are required only in rare cases. Such treatments, however, may carry a risk of thromboembolic events.
Aspirin once daily is recommended in clinical conditions in which antiplatelet prophylaxis has a favourable benefit/risk profile. The available evidence supports daily doses of aspirin in the range of 75–100 mg for the long-term prevention of serious vascular events in high-risk patients, including those with ACS and those undergoing PCI.77 In situations where an immediate antithrombotic effect is required (such as in ACS or in acute ischaemic stroke), a loading dose of 160–300 mg should be given in order to ensure complete inhibition of TXA2-dependent platelet aggregation.1,78 The use of intravenous formulations of aspirin is considered unnecessary, unless dictated by the inability of the patient to swallow or chew an oral formulation. It should be emphasized that the commonly used intravenous administration of 500 mg of aspirin would achieve systemic drug levels equivalent to ∼1000 mg of plain aspirin given orally, thereby producing substantial inhibition of PGI2 biosynthesis.79 These considerations should limit the use of intravenous aspirin to 80–150 mg and encourage drug companies to develop an appropriate intravenous formulation of low-dose aspirin.
No test of platelet function is routinely recommended to assess the antiplatelet effect of aspirin in the individual patient. Measurement of serum TXB2 may help detect non-compliance or less than complete inactivation of platelet COX-1 due to an interaction with other NSAIDs (see Supplementary material online).
The routine use of PPIs or cytoprotective agents is not recommended in patients taking daily doses of aspirin in the range of 75–100 mg, because of lack of large randomized trials demonstrating the efficacy of such protective strategies in reducing the risk of upper GI bleeding complications. Two randomized studies80,81 suggest that the combination of esomeprazole and low-dose aspirin is superior to clopidogrel in preventing recurrent ulcer bleeding in patients with a history of aspirin-induced ulcer bleeding.
Arthritic patients on low-dose aspirin therapy should be instructed to avoid the use of ibuprofen or naproxen because even over-the-counter doses of these NSAIDs may interfere with the antiplatelet effect of aspirin.82,83 Alternative options are represented by paracetamol or diclofenac, because they do not interfere with the platelet pharmacodynamics of low-dose aspirin.84
Clopidogrel, 75 mg daily, is an appropriate alternative for patients with coronary, cerebrovascular or peripheral arterial disease who cannot tolerate low-dose aspirin.
The publication of the CURE,41 COMMIT,44 and CLARITY-TIMI 2843 trials and the need for dual antiplatelet therapy after stenting has led to the guideline recommendation of aspirin and clopidogrel for both patients with NSTE-ACS and STEMI. While not overcoming the problem of variability in response, higher loading and maintenance doses of clopidogrel (600 vs. 300 mg loading and 150 vs. 75 mg daily for 7 days) achieve greater mean levels of platelet inhibition.85 However, when evaluated in a wide population of ACS patients intended to undergo PCI, high-dose clopidogrel was not found to reduce significantly the 30-day outcome when compared with standard dose, except in the subgroup actually undergoing PCI,86 but major bleeding was significantly increased in high- vs. low-dose clopidogrel.
Based on international recommendations,87,88 PPIs are recommended to reduce GI bleeding among patients with a history of upper GI bleeding. Proton pump inhibitors are appropriate in patients with multiple risk factors for GI bleeding who require antiplatelet therapy.88 However, PPIs might also interact with the antithrombotic effect of clopidogrel because of a pharmacokinetic interaction reducing formation of its active metabolite.11 The FDA and EMA have issued warnings concerning all or specific PPIs in terms of the potential clinical impact of this pharmacokinetic interaction. A recently published meta-analysis of observational studies and post hoc analyses of RCTs12 has reported that PPI users displayed a 41% increased risk of MACE and an 18% increased risk of death compared with non-users. The majority of patients included in the meta-analysis was treated with omeprazole, the most potent CYP2C19 inhibitor.12 The COGENT trial represents the first, randomized assessment of a fixed combination of clopidogrel (75 mg) and omeprazole (20 mg), compared with clopidogrel in 3873 ACS patients with a median follow-up of 133 days.89 One hundred and nine patients had a cardiovascular event and there was no evidence of an adverse interaction between omeprazole and clopidogrel (HR = 0.99; 95% CI, 0.68–1.44). However, due to premature termination of the trial, COGENT had limited statistical power to detect a clinically relevant interaction. These recent findings suggest the need for reconsidering the appropriateness of PPI co-administration in CHD patients treated with clopidogrel, carefully balancing the potential benefits and risks of this adjuvant therapy in the individual patient.
The main advantage of prasugrel over clopidogrel appears to be the prevention of non-fatal MI and stent thrombosis in ACS patients who undergo PCI.51 The cost of this prevention is excess bleeding. The improved efficacy of prasugrel may be exploited in the setting of STEMI referred for primary PCI or after coronary angiography in patients with NSTE-ACS undergoing PCI.47 The use of prasugrel should also be considered in patients who develop stent thrombosis despite aspirin and clopidogrel therapy. The place of prasugrel in the initial medical management of ACS remains uncertain until further ongoing studies are completed.90
A lower maintenance dose of 5 mg in the elderly (>75 years) and the underweight (<60 kg) would seem logical, but formal testing of this hypothesis is necessary before such a strategy can be recommended.90 The ongoing TRILOGY trial and other studies are designed to answer this question.
The addition of dipyridamole to aspirin has not been shown clearly to produce additional reductions in serious vascular events in an overview of 25 trials among ∼10 000 high-risk patients,3 although two trials suggested that there may be a worthwhile further reduction in stroke.29,30 The combination of low-dose aspirin and extended-release dipyridamole (200 mg, b.i.d.) is considered an acceptable option for patients with non-cardioembolic cerebral ischaemic events;91 however, there is no basis to recommend this combination in patients with ischaemic heart disease.
Ticagrelor provides superior prevention of death and recurrent MI compared with clopidogrel when used in patients with STEMI planned for primary PCI or moderate-to-high-risk NSTE-ACS and is therefore recommended in these patients other than those with prior history of intracranial haemorrhage or active pathological bleeding that cannot be controlled by local measures. There is no requirement to modify the dose of ticagrelor according to age or body weight but the concomitant use of strong CYP3A inhibitors such as ketoconazole is contraindicated since these will markedly increase the plasma levels of ticagrelor and increase the risk of bleeding and other side effects. The PEGASUS-TIMI 54 study will address the question of whether ticagrelor compared with placebo provides ongoing benefit in moderate-to-high-risk patients >1 year after MI (Clinicaltrials.gov identifier NCT01225562).
Ticagrelor can be commenced prior to obtaining the results of coronary angiography and may be administered to patients who have already received clopidogrel. If it is not tolerated as a result of adverse effects other than bleeding, maintenance therapy with either clopidogrel or prasugrel, as appropriate, may be commenced in its place. Despite the observation of greater creatinine increases in ticagrelor-treated vs. clopidogrel-treated patients in the PLATO study,52 ticagrelor showed a greater absolute risk reduction compared with clopidogrel in patients with evidence of chronic kidney disease and is therefore recommended in these patients, other than those requiring renal replacement therapy.92
The main treatment recommendations of antiplatelet drugs are summarized in Table 6.
Dual antiplatelet therapy is recommended initially for all patients undergoing elective PCI (I A); the duration of P2Y12 blocker administration varies according to type of coronary stent: at least 1 month after BMS (I B); at least 3 months (I B), 6 months (IIa B), or 1 year (IIb C) after DES
Clopidogrel 300 mg oral LD >12 h before PCI
An upstream LD of clopidogrel 300 mg (instead of delayed or no LD) is recommended (I B)
Chronic IHD patients
ASA 75–100 mg p.o. o.d. long term
Low-dose ASA is recommended in the absence of contraindications in all patients with diagnosed IHD (I A)
E.g. stable angina, elective CABG, >12 months post-ACS
Clopidogrel 75 mg p.o. o.d. long term
Long-term clopidogrel instead of ASA is recommended in case of ASA intolerance (I B)
Asymptomatic individuals at high risk (>2% pa) of CV disease
ASA 75–100 mg p.o. o.d. long term
Low-dose ASA may be considered on an individual basis, if risk remains high after optimal blood pressure control and statin therapy (IIa C)
ACS, acute coronary syndrome; ASA, aspirin; b.i.d., bis in die; BMS, bare metal stent; CABG, coronary artery bypass graft; CV, cardiovascular, DES, drug-eluting stent; GPI, glycoprotein IIbIIIa inhibitor, IHD, ischaemic heart disease; i.v., intravenous, LD, loading dose, o.d., once a day, MI, myocardial infarction, NSTE, non-ST-elevation, o.d. , once a day, pa, per annum, PCI, percutaneous coronary intervention, p.o., per os, STE, ST-elevation.
aAlthough P2Y12 blockers (namely clopidogrel) have been tested against placebo only in patients with NSTEACS with or without PCI for up to 12 months and in patients with STEMI (treated with fibrinolysis) for up to 4 weeks, a P2Y12 blocker should be considered on top of ASA in all STEMI patients for up to 12 months (IIa C).
bIndicates general statement reported for all ACS patients would apply to three separate sub-settings.
cA head to head comparison of prasugrel vs. ticagrelor is lacking.
This position paper represents an independent initiative of the Working Group on Thrombosis of the European Society of Cardiology, funded by the Working Group.
Conflict of interest: C.P. Research grant (to institution): Bayer; Speaking or consulting: AstraZeneca, Bayer, Eli Lilly, Merck, NicOx, Novartis, Sanofi-Aventis, Servier. F.A. Grant, consultant or speaker fees from Astra Zeneca, Bayer, Boehringer Ingelheim, Bristol-Myers Squibb, Daiichi-Sankyo, Lilly, Merck Sharp & Dohme, Pfizer, Servier. L.B. Honoraria and consultant fees from AstraZeneca, Eli Lilly/Daiichi Sankyo, Amgen, Esteve. J-P.C. Research/educational grants from Bristol-Myers Squibb, Sanofi-Aventis, Eli Lilly, Guerbet Medical, Medtronic, Boston Scientific, Cordis, Stago, Fondation de France, INSERM, Fédération Française de Cardiologie, and Société Française de Cardiologie; consulting fees from Sanofi-Aventis, Eli Lilly, and Bristol-Myers Squibb; and lecture fees from Bristol-Myers Squibb, Sanofi-Aventis, Eli Lilly and AstraZeneca. R.D.C. Consulting fees and honoraria from Boehringer Ingelheim, Bayer, Bristol Myers-Squibb, Merck Sharp & Dohme, Pfizer, Daiichi-Sankyo, Sanofi-Aventis; Grant support from Bristol Myers-Squibb, Boehringer Ingelheim, Sanofi-Aventis. K.H. lecture fees from AstraZeneca, Bayer, Boehringer-Ingelheim, Daiichi Sankyo, Eli Lilly, Sanofi-Aventis and The Medicines Company. S.H. Research/educational grants from Eli Lilly, AstraZeneca, Boehringer Ingelheim, Astella, Pfizer, BMS, Bayer. Consultant fees from Boehringer Ingelheim, BMS, Pfizer, Sanofi-aventis, AstraZeneca, Bayer, Eli Lilly. S.D.K. Lecture fees: AstraZeneca, Eli Lilly, Merck, The Medicines Company. F-J.N. Institutional: Speaker honoraria, consultancy fees and research grants from Lilly, Daiichi Sankyo, Medicines, Sanofi-Aventis, Bristol, Novartis, Roche. A.S. Institutional grants from Eli Lilly, AstraZeneca and Boehringer Ingelheim. Advisory board: Boehringer Ingelheim. P-G.S. Research Grant (to institution): Servier; Consulting/advisory board: Ablynx, Astellas, AstraZeneca, Bayer, Boehringer-Ingelheim, BMS, Daiichi-Sankyo-Lilly, Eisai, GSK, Medtronic, MSD, Otsuka, Pfizer, Roche, Sanofi-aventis, Servier, The Medicines Company. R.F.S. research/educational grants from AstraZeneca, Dynabyte, Eli Lilly/Daiichi Sankyo alliance, Schering-Plough/Merck, and Accumetrics; honoraria from AstraZeneca, Eli Lilly/Daiichi Sankyo alliance, Medscape, Novartis, GlaxoSmithKline, Iroko and Schering-Plough/Merck; consultant fees from AstraZeneca, Eli Lilly/Daiichi Sankyo alliance, Merck, Novartis, Sanofi-Aventis/Bristol-Myers Squibb, Eisai, The Medicines Company and Accumetrics; travel support from AstraZeneca, Eli Lilly/Daiichi Sankyo alliance, and Schering-Plough/Merck. F.V.d.W. Research grants from Merck, Boehringer Ingelheim, Sanofi-Aventis, Speaker's bureau and advisory committee of Merck, Boehringer Ingelheim, Astra Zeneca, The Medecines Company, Pfizer, Lilly. F.V. AstraZeneca, Eli Lilly and Daiichi-Sankyo. The other authors have no potential conflicts of interest to disclose.
↵† The Authors are members of a Task Force on Antiplatelet Agents of the Working Group on Thrombosis of the European Society of Cardiology.
Antithrombotic Trialists’ Collaboration. Collaborative meta-analysis of randomised trials of antiplatelet therapy for prevention of death, myocardial infarction, and stroke in high-risk patients. BMJ 2002;324:71-86. doi:10.1136/bmj.324.7329.71.
Antithrombotic Trialists’ (ATT) Collaboration. Aspirin in the primary and secondary prevention of vascular disease: collaborative meta-analysis of individual participant data from randomised trials. Lancet 2009;373:1849-1860. doi:10.1016/S0140-6736(09)60503-1.
. Do selective cyclooxygenase-2 inhibitors and traditional non-steroidal anti-inflammatory drugs increase the risk of atherothrombosis? Meta-analysis of randomised trials. BMJ 2006;332:1302-1308. doi:10.1136/bmj.332.7553.1302.
. Prasugrel achieves greater and faster P2Y12 receptor-mediated platelet inhibition than clopidogrel due to more efficient generation of its active metabolite in aspirin-treated patients with coronary artery disease. Eur Heart J 2008;29:21-30. doi:10.1093/eurheartj/ehm545.
. Cardiovascular risk in clopidogrel-treated patients according to cytochrome P450 2C19*2 loss-of-function allele or proton pump inhibitor coadministration. A systematic meta-analysis. J Am Coll Cardiol 2010;56:134-143. doi:10.1016/j.jacc.2009.12.071.
. Prasugrel: a novel thienopyridine antiplatelet agent. A review of preclinical and clinical studies and the mechanistic basis for its distinct antiplatelet profile. Cardiovasc Drug Rev 2007;25:357-374. doi:10.1111/j.1527-3466.2007.00027.x.
. Patients with poor responsiveness to thienopyridine treatment or with diabetes have lower levels of circulating active metabolite, but their platelets respond normally to active metabolite added ex vivo. J Am Coll Cardiol 2008;52:1968-1977. doi:10.1016/j.jacc.2008.07.068.
; for the TRITON-TIMI 38 Investigators. Pharmacodynamic assessment of platelet inhibition by prasugrel vs. clopidogrel in the TRITON-TIMI 38 trial. Eur Heart J 2009;30:1753-1763. doi:10.1093/eurheartj/ehp159.
; on behalf of the TAIPAD Investigators. Thromboxane antagonism with terutroban in peripheral arterial disease: the TAIPAD study. J Thromb Haemost 2010;8:2369-2376. doi:10.1111/j.1538-7836.2010.04020.x.
. Randomized double-blind assessment of the ONSET and OFFSET of the antiplatelet effects of ticagrelor versus clopidogrel in patients with stable coronary artery disease. The ONSET/OFFSET study. Circulation 2009;120:2577-2585.
. Ticagrelor binds to human P2Y12 independently from ADP but antagonizes ADP-induced receptor signalling and platelet aggregation. J Thromb Haemost 2009;7:1556-1565. doi:10.1111/j.1538-7836.2009.03527.x.
. Basic and translational research on proteinase-activated receptors: antagonism of the proteinase-activated receptor 1 for thrombin, a novel approach to antiplatelet therapy for atherothrombotic disease. J Pharmacol Sci 2008;108:433-438. doi:10.1254/jphs.08R06FM.
; for the TRA-PCI Investigators. Safety and tolerability of SCH 530348 in patients undergoing non-urgent percutaneous coronary intervention: a randomised, double blind, placebo controlled phase II study. Lancet 2009;373:919-928. doi:10.1016/S0140-6736(09)60230-0.
; on behalf of the J-LANCELOT Investigators. Double-blind, placebo-controlled phase II studies of the protease-activated receptor 1 antagonist E5555 (atopaxar) in Japanese patients with acute coronary syndrome or high-risk coronary artery disease. Eur Heart J. 2010;31:2601-2613. doi:10.1093/eurheartj/ehq320.
The ESPRIT Study Group. Aspirin plus dipyridamole versus aspirin alone after cerebral ischaemia of arterial origin (ESPRIT): randomised controlled trial. Lancet 2006;367:1665-1673. doi:10.1016/S0140-6736(06)68734-5.
. Adding cilostazol to dual antiplatelet therapy achieves greater platelet inhibition than high maintenance dose clopidogrel in patients with acute myocardial infarction: results of the adjunctive cilostazol versus high maintenance dose clopidogrel in patients with AMI (ACCEL-AMI) study. Circ Cardiovasc Interv, 2010;3:17-26. doi:10.1161/CIRCINTERVENTIONS.109.880179.
. Protective effects of aspirin against acute myocardial infarction and death in men with unstable angina: results of a veterans administration cooperative study. N Engl J Med 1983;309:396-403. doi:10.1056/NEJM198308183090703.
The RISC Group. Risk of myocardial infarction and death during treatment with low dose aspirin and intravenous heparin in men with unstable coronary artery disease. Lancet 1990;336:827-830. doi:10.1016/0140-6736(90)92336-G.
; VahanianA, CammJ, De CaterinaR, DeanV, DicksteinK, FilippatosG, Funck-BrentanoC, HellemansI, KristensenSD, McGregorK, SechtemU, SilberS, TenderaM, WidimskyP, ZamoranoJL, SilberS, AguirreFV, Al-AttarN, AlegriaE, AndreottiF, BenzerW, BreithardtO, DanchinN, Di MarioC, DudekD, GulbaD, HalvorsenS, KaufmannP, KornowskiR, LipGY, RuttenF; ESC Committee for Practice Guidelines (CPG). Management of acute myocardial infarction in patients presenting with persistent ST-segment elevation: the Task Force on the Management of ST-Segment Elevation Acute Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2008;29:2909-2945.
; on behalf of the MATCH Investigators. Aspirin and clopidogrel compared with clopidogrel alone after recent ischaemic stroke or transient ischaemic attack in high-risk patients (MATCH): randomised, double-blind, placebo-controlled trial. Lancet 2004;364:331-337.
. Low-dose aspirin for secondary cardiovascular prevention—cardiovascular risks after its perioperative withdrawal versus bleeding risks with its continuation—review and meta-analysis. J Intern Med 2005;257:399-414.
. The effect of pre-operative aspirin on bleeding, transfusion, myocardial infarction, and mortality in coronary artery bypass surgery: a systematic review of randomised and observational studies. Eur Heart J 2008;29:1057-1071.
. Baseline risk of major bleeding in non-ST-segment-elevation myocardial infarction: the CRUSADE (Can Rapid risk stratification of Unstable angina patients Suppress Adverse outcomes with Early implementation of the ACC/AHA Guidelines) Bleeding Score. Circulation 2009;119:1873-1882.
. Prognostic modelling of individual patient risk and mortality impact of ischemic and hemorrhagic complications: assessment from the acute catheterization and urgent intervention triage strategy trial. Circulation 2010;121:43-51.
; CRUSADE Investigators. The implications of blood transfusions for patients with non-ST-segment elevation acute coronary syndromes: results from the CRUSADE National Quality Improvement Initiative. J Am Coll Cardiol 2005;46:1490-1495.
; Task Force for Diagnosis and Treatment of Non-ST-Segment Elevation Acute Coronary Syndromes of European Society of Cardiology. Guidelines for the diagnosis and treatment of non-ST-segment elevation acute coronary syndromes. Eur Heart J 2007;28:1598-1660.
; American College of Chest Physicians. Antithrombotic therapy for non-ST-segment elevation acute coronary syndromes. American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th edition). Chest 2008;133:670S-707S.
; International Consensus Upper Gastrointestinal Bleeding Conference Group. International consensus recommendations on the management of patients with nonvariceal upper gastrointestinal bleeding. Ann Intern Med 2010;152:101-113.
. ACCF/ACG/AHA 2010 Expert Consensus Document on the concomitant use of proton pump inhibitors and thienopyridines: a focused update of the ACCF/ACG/AHA 2008 expert document on reducing the gastrointestinal risks of antiplatelet therapy and NSAID use. J Am Coll Cardiol 2010;56:2051-2066.