INTRODUCTION — Rupture of an atherosclerotic plaque is the usual initiating event in an acute coronary syndrome (ACS), leading to subsequent thrombus formation. Persistent thrombotic occlusion results in acute myocardial infarction (MI). (See "The role of plaque rupture in acute coronary syndromes".)

Platelets play an important role in this process. Aggregating platelets form the core of the growing thrombotic mass, with upstream and/or downstream propagation of fibrin and red blood cell-rich clot [1]. Platelet-rich thrombi are more resistant to clot lysis than red blood cell-rich thrombi and, if lysis occurs, platelet-rich thrombi promote the development of reocclusion [2].

The actions of antiplatelet agents, the evidence that they are beneficial in patients with a non-ST elevation ACS (unstable angina or non-ST elevation MI), and recommendations for the use of these agents in this setting will be reviewed here. The efficacy of antiplatelet agents in ST elevation MI (STEMI) is discussed separately. (See "Antiplatelet agents in acute ST elevation myocardial infarction".)

PLATELETS IN ACUTE CORONARY SYNDROMES — Platelet adhesion, activation, and aggregation are stimulated during an ACS. Intimal injury due to plaque rupture exposes collagen and von Willebrand factor, to which circulating platelets adhere. Following adhesion, multiple metabolic pathways are stimulated within the platelet, resulting in the production and release of thromboxane A2 (TXA2), ADP, and other substances from platelet granules. These platelet products stimulate further platelet recruitment, activation, and vasoconstriction; they also lead to platelet aggregation by activating the glycoprotein IIb/IIIa (GP IIb/IIIa) complex, which binds platelets to one another through linkage with fibrinogen molecules (figure 1). (See "Congenital and acquired disorders of platelet function", section on 'Normal platelet function' and "The role of platelets in coronary heart disease".)

CLASSIFICATION OF ANTIPLATELET AGENTS — Antiplatelet agents can interfere with a number of platelet functions, including aggregation, release of granule contents, and platelet-mediated vascular constriction. (See "Effect of cardiac drugs on platelet function".) They can be classified according to their mechanism of action (algorithm 1):

• Aspirin and related compounds (nonsteroidal antiinflammatory drugs and sulfinpyrazone) block cyclooxygenase (prostaglandin H synthase), the enzyme that mediates the first step in the biosynthesis of prostaglandins and thromboxanes (including TxA2) from arachidonic acid (algorithm 2).
• Dipyridamole inhibits phosphodiesterase-mediated breakdown of cyclic AMP, which prevents platelet activation by multiple mechanisms.
• The thienopyridines clopidogrel, ticlopidine, prasugrel and ticagrelor block the binding of ADP to a specific platelet receptor P2Y12, thereby inhibiting activation of the GP IIb/IIIa complex and platelet aggregation [3] (figure 2).
• GP IIb/IIIa antibodies and receptor antagonists inhibit the final common pathway of platelet aggregation (the crossbridging of platelets by fibrinogen binding to the GP IIb/IIIa receptor) and may also prevent adhesion to the vessel wall.

ASPIRIN — Aspirin has been used in a variety of vascular disorders including primary and secondary prevention of coronary heart disease, transient ischemic attack, and stroke, and in the acute therapy of patients with an acute coronary syndrome (ACS). (See "Benefits and risks of aspirin in secondary and primary prevention of cardiovascular disease".) Among patients who tolerate aspirin, many are also treated with clopidogrel. (See 'Thienopyridines and ticagrelor' below.)

The antiplatelet activity of aspirin appears to be mediated principally through inhibition of the synthesis of thromboxane A2 (TxA2) [4]. TxA2 is released by platelets in response to a number of agonists, amplifying the platelet response that leads to aggregation. Aspirin irreversibly acetylates and inactivates cyclooxygenase (prostaglandin G/H synthase), which catalyzes the first step of the conversion of arachidonic acid to thromboxane A2 (graph 1). (See "Benefits and risks of aspirin in secondary and primary prevention of cardiovascular disease", section on Mechanism of action.)

Platelets do not synthesize new enzyme. As a result, the functional defect induced by aspirin persists for the life of the platelet.

Clinical trials — The Antithrombotic Trialists' Collaboration reviewed the effect of antiplatelet therapy, mostly aspirin, in 12 trials of over 5000 patients with a non-ST elevation ACS (table 1) [5]. Antiplatelet therapy produced a significant 46 percent reduction in the combined end point of subsequent nonfatal myocardial infarction (MI), nonfatal stroke, or vascular death (8.0 versus 13.3 percent).

The range of findings can be illustrated by the following clinical trials that evaluated different types of therapy.

VA Cooperative Study — The Veterans Administration (VA) Cooperative Study was a multicenter, double-blind, randomized trial that compared aspirin (325 mg/day) to placebo in 1266 men with a non-ST elevation ACS (rest angina or new onset angina occurring with minimal physical activity). Aspirin lowered the incidence of death or acute MI by 51 percent (5 versus 10.1 percent) [6]. Although therapy was discontinued after 12 weeks, the mortality rate remained 43 percent lower after one year follow-up in the aspirin group.

Canadian multicenter trial — A double-blind randomized Canadian multicenter trial compared four regimens in 555 patients with a non-ST elevation ACS (rest or crescendo angina): aspirin (325 mg/day), sulfinpyrazone (200 mg four times daily), combination therapy, or placebo [7]. Treatment was initiated within eight days after hospitalization and continued for 18 months. Aspirin administration resulted in a 71 percent reduction in mortality and a 51 percent reduction in the combined end point of death or nonfatal MI when compared to placebo (graph 1).

The reduction in the frequency of events from 17 to 8 percent means that the absolute benefit was approximately 9 percent. There was no observable benefit with sulfinpyrazone, given alone or in combination with aspirin.

RISC trial — The RISC trial randomly assigned 796 patients with a non-ST elevation ACS to low-dose aspirin (75 mg/day), heparin (five days of intermittent intravenous infusion), or placebo [8]. Compared to placebo, aspirin was associated with a highly significant reduction in the combined end point of acute MI or death at 5, 30, and 90 days of treatment (graph 2). The absolute benefit at 90 days was approximately 12 percent (6.8 versus 18.8 percent). Prolonged follow-up showed that the benefit of low-dose aspirin was maintained after one year of therapy [9]. (See "Anticoagulant therapy in unstable angina and acute non-ST elevation myocardial infarction", for a review of the role of heparin in this disorder).

Effect of prior aspirin use — Aspirin use in the preceding seven days before an NSTEMI tends to decrease the severity of the ACS (less NSTEMI, more UA) [10-12]. Prior aspirin use has also been associated with a reduction in in-hospital mortality and with a reduced frequency of heart failure or pulmonary edema.

Paradoxically, aspirin use was a risk factor for a worse outcome in the analysis that lead to the development of the TIMI risk score in non-ST elevation ACS (calculator 1) [13] and in some other nonrandomized comparisons [11,12]. It is likely that this association reflects higher rates of aspirin use among patients with more severe underlying vascular disease, rather than a harmful effect of aspirin. (See "Risk stratification after unstable angina or non-ST elevation myocardial infarction", section on 'TIMI risk score'.)

Dose — An initial loading dose of 162 to 325 mg of uncoated aspirin should be given as soon as possible to any patient thought to have an ACS [14-16]. At this dose, aspirin produces a rapid antithrombotic effect due to immediate and almost complete inhibition of thromboxane A2 production. The first tablet should be chewed or crushed, including enteric coated preparations, to establish a high blood level quickly. More rapid absorption occurs with non-enteric coated formulations.

Aspirin (75 to 162 mg once a day) should be continued indefinitely for secondary prevention unless the patient cannot tolerate aspirin because of gastrointestinal intolerance or hypersensitivity. For such patients, clopidogrel (75 mg/day) is an equally effective alternative [5,17].

The long-term clinical benefit derived from aspirin appears to be relatively independent of dose. In the Antithrombotic Trialists' Collaboration, the most widely tested regimen in the secondary prevention trials was "medium dose" aspirin (75 to 325 mg/day). Neither higher aspirin doses nor other antiplatelet drugs were more effective than daily aspirin in this dose range [5]. There was insufficient evidence to confirm the efficacy of doses below 75 mg/day.

A dose of 75 to 162 mg/day may provide optimal efficacy while limiting toxicity. In the Antithrombotic Trialists' Collaboration overview, aspirin at a dose of 75 to 150 mg/day was as effective as a dose of 150 to 325 mg/day [5]. (See "Benefits and risks of aspirin in secondary and primary prevention of cardiovascular disease", section on Recommendations and dosing.)

Duration of therapy — Aspirin therapy is associated with long-term benefit in patients with an acute MI [5,18,19]. The Antiplatelet Trialists' Collaboration concluded that, in patients with a prior MI, antiplatelet therapy (mostly aspirin) was associated with a 25 percent reduction in vascular events, with an absolute benefit of 3.6 vascular events prevented per 100 patients at two years (13.5 versus 17.0 percent) (table 1) [5].

Side effects — Aspirin therapy may be associated with gastrointestinal intolerance or bleeding, allergy (primarily manifested as bronchospasm or asthma), and worsening of preexistent bleeding. Contraindications to aspirin include active bleeding, hemophilia, active retinal bleeding, severe untreated hypertension, an active peptic ulcer, or another serious source of gastrointestinal bleeding.

Gastrointestinal side effects such as dyspepsia and nausea are infrequent with the low doses used in this setting. Patients who develop gastrointestinal side effects should be treated with a proton pump inhibitor and are usually able to tolerate a lower and probably still effective dose (80 to 100 mg of aspirin per day) [14]. Enteric-coated aspirin may be of some benefit, but it does not protect against gastrointestinal bleeding. (See "Benefits and risks of aspirin in secondary and primary prevention of cardiovascular disease", section on Enteric-coated aspirin.)

A more detailed discussion of gastrointestinal bleeding in ACS patients is found below. (See 'Gastrointestinal bleeding' below.)

In patients who cannot tolerate aspirin because of gastrointestinal intolerance or hypersensitivity, equally effective alternatives are clopidogrel (75 mg/day) and ticlopidine (250 mg twice daily) [5,17,20]. The 2007 ACC/AHA guidelines on unstable angina (UA) and non-ST-segment myocardial infarction (UA/NSTEMI) and the 2008 American College of Chest Physicians (ACCP) guideline on antithrombotic therapy for non-ST-elevation acute coronary syndromes recommended clopidogrel in such patients [14,15] because it is as least as effective as and has fewer side effects than ticlopidine [21,22]. Sulfinpyrazone is not recommended [4].

Aspirin resistance — The subject of resistance/nonresponse to aspirin is discussed separately. (See "Nonresponse and resistance to aspirin and clopidogrel".)

Use with warfarin — Both the ACC/AHA and ACCP guidelines state that low-dose aspirin can be given with warfarin in patients who have an indication for warfarin therapy (eg, deep vein thrombosis, atrial fibrillation) [14-16]. (See "Chronic anticoagulation after myocardial infarction".)

THIENOPYRIDINES AND TICAGRELOR — The thienopyridines clopidogrel, ticlopidine, and prasugrel and the cyclopentyltriazolopyrimidine ticagrelor, block the adenosine diphosphate receptor P2Y12 on platelets. This issue is discussed in detail elsewhere. (See "Platelet biology", (figure 2).

Most evidence suggests that the antiplatelet benefits of these agents are additive to those of aspirin. Clopidogrel is preferred to ticlopidine in most situations due to its better side effect profile. The most common important side effect associated with thienopyridine therapy is bleeding. (See 'Bleeding risk' below.)

However, three limitations to the use of clopidogrel (delayed onset of action, large between individual variability in platelet response, and irreversibility of its inhibitory effect on platelets) have led to the development of other thienopyridines including prasugrel and ticagrelor [23]. Both prasugrel and ticagrelor lead to more intense platelet inhibition than clopidogrel. (See "Platelet biology", section on 'Adenosine diphosphate (ADP)',).

Prasugrel is superior to clopidogrel in terms of efficacy, but has a higher rate of major bleeding. Ticagrelor has been studied in one randomized trial and is not yet approved for use. The use of these agents in patients with ACS who receive stents is discussed elsewhere. (See "Coronary artery stent thrombosis: Prevention and management", section on 'Prevention with combined antiplatelet therapy'.)

Clopidogrel therapy — In patients with a non-ST elevation ACS, clopidogrel has been evaluated in combination with aspirin. Clopidogrel (75 mg/day) is also an alternative in patients who cannot tolerate aspirin because of gastrointestinal intolerance or hypersensitivity.

The efficacy of clopidogrel was established in the CURE trial, which randomly assigned 12,562 patients who presented within 24 hours after the onset of a non-ST elevation ACS to aspirin alone (75 to 325 mg/day) or with clopidogrel (300 mg loading dose immediately followed by 75 mg/day); the great majority were at increased risk because of ECG changes (mostly ST depression ≥1 mm or T wave inversion ≥2 mm) or elevated cardiac enzymes [24]. The primary end point was cardiovascular death, myocardial infarction, or stroke.

At an average follow-up of nine months, combination therapy was associated with reductions in the primary end point (9.3 versus 11.4 percent for aspirin alone), which was largely due to fewer myocardial infarctions (5.2 versus 6.7 percent) (graph 3). Clopidogrel was also associated with modest reductions in the incidence of severe or refractory in-hospital ischemia, in-hospital heart failure, and need for a revascularization procedure, and with a significant increase in major bleeding (3.7 versus 2.7 percent) but not in life-threatening bleeding or hemorrhagic stroke.

Clopidogrel therapy produced a similar relative risk reduction in patients who were treated medically or underwent revascularization [24] and in low, intermediate, and high risk patients as defined by the TIMI risk score (calculator 1) [25]. The high risk patients derived the greatest absolute benefit, since they had the highest baseline risk.

A subsequent analysis from CURE evaluated the time course of benefit of clopidogrel [26]. Evidence of benefit began to emerge within 24 hours and gradually increased in magnitude during the first 30 days (4.3 versus 5.4 percent incidence of the primary end point, relative risk 0.79). The benefit continued to increase from 31 days to one year (5.2 versus 6.3 percent incidence of new events, relative risk 0.82) (graph 4). There was no significant excess of late life-threatening bleeding, but there was a small excess of major bleeds (5 per 1000) that was much smaller than the total cardiovascular benefit at one year (22 per 1000).

Efficacy in PCI — A subset of 2658 patients in CURE (PCI-CURE) underwent a percutaneous coronary intervention (PCI) at the discretion of the physician; stenting was performed in 82 percent [27]. The patients were pretreated with aspirin plus either clopidogrel or placebo for a median of six days prior to intervention; all received aspirin and clopidogrel for four weeks after intervention. The primary end point (cardiovascular death, myocardial infarction, or urgent target vessel revascularization at 30 days after PCI) was significantly reduced with clopidogrel (4.5 versus 6.4 percent for placebo); the benefit persisted until the end of follow-up (average eight months) (graph 5) and was seen in both stented and nonstented patients. There was no significant difference in major bleeding episodes between the two groups. (See "Coronary arteriography and revascularization for unstable angina or non-ST elevation acute myocardial infarction".)

There were two concerns about the design of PCI-CURE. First, pretreatment was given for a median of six days, whereas most patients with a non-ST elevation ACS in the United States undergo angiography within 4 to 48 hours followed by revascularization, if appropriate. Second, GP IIb/IIIa inhibitors were only used in 21 percent of patients undergoing PCI in PCI-CURE [28]. A post hoc nonrandomized analysis from the TARGET trial, which evaluated the efficacy of different GP IIb/IIIa inhibitors, found that event rates were lower in the 93 percent of patients pretreated with clopidogrel [29]. In PCI-CURE, combination therapy was not associated with an increase in major or life-threatening bleeding.

The CREDO trial addressed the limitations of PCI-CURE in more than 2100 patients undergoing elective PCI (89 percent with stenting, all bare metal): approximately one-third had an ACS and 40 percent were treated with a GP IIb/IIIa inhibitor [30]. The patients were randomly assigned to pretreatment with placebo or clopidogrel at a loading dose of 300 mg given 3 to 24 hours before the procedure; after the procedure, all patients received clopidogrel (75 mg/day) for 28 days; clopidogrel or placebo, according to original randomization, was then given for the remainder of one year. All patients received aspirin for the duration of the study, and more than 40 percent also received concomitant therapy with an intravenous GP IIb/IIIa inhibitor

The following findings were noted:

• A significant benefit at 28 days was seen in patients who had received a clopidogrel loading dose but, in a prespecified subset analysis, only if the drug was given at least six hours before the procedure (graph 6). A subsequent subgroup analysis showed that the benefit of the clopidogrel loading dose was first significant 15 hours before PCI and was greatest at 24 hours before PCI (graph 7) [31]. The absolute risk reduction in patients with ≥15 hours of clopidogrel treatment was 6.2 percent compared to placebo pretreatment in the same time interval (3.5 versus 9.7 percent, adjusted odds ratio 0.72). The risk reduction with prolonged clopidogrel pretreatment was greatest in patients who did not receive a GP IIb/IIIa inhibitor.

The benefit of clopidogrel pretreatment was primarily due to fewer acute thrombotic complications (death, MI, or urgent target vessel revascularization) and was independent of whether or not the patient had an ACS, underlying diabetes, or received concurrent therapy with a GP IIb/IIIa inhibitor [30,31].

• A significant long-term benefit was noted with prolonged daily clopidogrel therapy (12 months versus 1 month) (graph 8).

Dose — The standard clopidogrel regimen in patients with a non-ST elevation ACS, which was used in CURE and CREDO, has been a 300 mg loading dose followed by a maintenance dose of 75 mg/day [24,27,30]. However, as noted above, the benefit in CREDO was only seen in patients who had been treated at ≥15 hours before PCI [31].

As a result, we suggest a 600 mg loading dose in patients who will undergo angiography and possible revascularization the same day. Compared to the 300 mg loading dose, the 600 mg loading dose appears to produce maximal antiplatelet activity within two to three hours, to have a greater antiplatelet effect, to diminish the likelihood of clopidogrel resistance, and may, based upon limited data, improve outcomes. The data supporting these conclusions are presented separately. (See "Antithrombotic therapy for intracoronary stent implantation: Clinical trials", section on Timing and dose and section on Clopidogrel resistance.)

Duration of therapy — The 2007 ACC/AHA UA/NSTEMI guidelines recommended that clopidogrel therapy (75 mg/day) be continued for at least 1 month and ideally up to 12 months in patients with a non-ST elevation ACS who are treated medically [15]. The duration of therapy is the same for patients who receive bare metal stents, while all patients with drug-eluting stents should receive clopidogrel for a minimum of one year.

Recommendations for clopidogrel use beyond guideline recommendations have come from concerns resulting from both the observation of a clustering of adverse events in the initial 90 days after stopping clopidogrel, as well as from the development of very late stent thrombosis. (See "Coronary artery stent thrombosis: General issues".)

The frequency and timing of both all-cause mortality or acute MI after clopidogrel cessation was evaluated in a retrospective cohort study of ACS patients treated with stents (1569, 63 percent received bare metal stents) and without (1568) stents [32]. The mean duration of clopidogrel therapy was 278 and 302 days, respectively. In multivariate analysis, including adjustment for duration of clopidogrel treatment, the first 90-day interval after stopping clopidogrel treatment was associated with a small risk of adverse events in both medical and PCI groups (1.31 and 0.57 per 1000 patient-days, respectively) that was significantly higher than in the interval of 91 to 180 days (incidence rate ratios 1.82 and 1.98). This early incidence of adverse events was consistent whether patients took clopidogrel for three, six, or nine or more months.

The development of very late coronary artery stent thrombosis in patients with DES has led some experts to suggest longer term clopidogrel use, until data are available defining the optimal duration of clopidogrel therapy. We agree and suggest continuing dual antiplatelet therapy indefinitely after DES placement in some patients. This is particularly relevant when the antithrombotic benefit appears to exceed the risk for bleeding. (See "Coronary artery stent thrombosis: General issues", section on Summary and recommendations.)

A similar duration is required when stenting is combined with radiation (brachytherapy) for in-stent restenosis, since brachytherapy retards the normal coverage of the stent with endothelial cells, and with drug-eluting stents for the same reason.

Some physicians recommend continuing clopidogrel beyond one year or even indefinitely as long as the drug is well tolerated and expense is not an issue based upon evidence of continued benefit over time in CURE (figure 2), PCI-CURE (graph 1), and CREDO (graph 2) [24,27,30]. This approach is primarily considered in patients with more severe vascular disease (eg, prior MI or cerebrovascular event or peripheral vascular disease).

Controlled data in favor of clopidogrel use for longer than one year in non-ST elevation ACS are lacking. However, some support comes from the CAPRIE trial, which compared clopidogrel to aspirin (but not the combination) in patients with atherosclerosis not limited to acute coronary syndromes [17]. Clopidogrel therapy was associated with a significant reduction in fatal or nonfatal MI; the magnitude of this benefit increased progressively over the three year period of the study [33]. Support also comes from subgroup analyses from CHARISMA, where benefit was seen in patients with prior MI or stroke [34].

The recommendations for prolonged clopidogrel plus aspirin therapy in non-ST elevation ACS should not be confused with the lack of overall benefit from combined antiplatelet therapy in the CHARISMA trial. The patient populations were different, as CHARISMA consisted of patients with stable clinical cardiovascular disease or multiple atherothrombotic risk factors [35]. (See "Secondary prevention of cardiovascular disease: Risk factor reduction", section on 'CHARISMA trial'.)

Clopidogrel resistance — The subjects of resistance/nonresponse to clopidogrel, along with methods to determine in vitro platelet function following treatment with clopidogrel, are discussed separately. (See "Nonresponse and resistance to aspirin and clopidogrel".)

Ticlopidine therapy — The effect of ticlopidine on platelet function begins 24 to 48 hours after the initiation of therapy and peaks three to six days later. The multicenter STAIG trial evaluated the use of ticlopidine in patients with a non-ST elevation ACS [36]. This multicenter study randomly assigned 652 patients with crescendo angina, new angina, or angina at rest to ticlopidine (250 mg twice daily) or conventional therapy that did not include aspirin. Ticlopidine significantly reduced the occurrence of both nonfatal MI and the combined end point of vascular death or nonfatal MI by 46 percent at six months. A subgroup analysis of 489 patients with transient ischemic electrocardiographic changes at rest demonstrated a more profound benefit (65 percent reduction) in the combined end point [37].

No trials have compared ticlopidine to aspirin or assessed the role of combined therapy. As mentioned previously, side effects have limited its use compared to clopidogrel. (See "Antithrombotic therapy for intracoronary stent implantation: General use", section on 'Ticlopidine'.)

Prasugrel therapy — Prasugrel has a more rapid onset of action and is able to achieve higher degrees of platelet inhibition than clopidogrel, while having a comparable rate of significant bleeding [38]. Furthermore, prasugrel effectively suppresses platelet activity in a larger numbers of patients than clopidogrel, since 20 to 25 percent of patients appear to be clopidogrel-resistant.

The TRITON-TIMI 38 trial directly compared prasugrel to clopidogrel in 13,608 moderate- to high-risk ACS patients undergoing PCI, including 10,074 with non-ST elevation ACS [39]. Prasugrel was given with a loading dose of 60 mg and maintenance dose of 10 mg/day while clopidogrel was given with a 300 mg loading dose and a 75 mg/day maintenance dose; the median duration of therapy was 14.5 months. For patients with unstable angina or NSTEMI, the coronary anatomy had to be know before randomization. Thus, in the majority of cases both clopidogrel and prasugrel were given after coronary angiography.

Long-term aspirin therapy was recommended in all patients in addition to thienopyridine therapy.

At 15 month follow-up, the following findings were noted; the overall data are given since the benefits were similar in non-ST elevation and ST elevation syndromes:

• The primary efficacy end point (cardiovascular death, nonfatal MI, or nonfatal stroke) occurred significantly less often in patients treated with prasugrel (9.9 versus 12.1 percent; hazard ratio [HR] 0.81; 95% CI 0.73-0.90) irrespective of whether or not they underwent PCI with stenting [39,40]. This was driven primarily by a significant reduction in nonfatal MI (7.4 versus 9.7 percent).
• Among patients who sustained a nonfatal MI or stroke, the likelihood of a recurrent primary end point was significantly reduced with prasugrel compared to clopidogrel (10.8 versus 15.4 percent), as was the likelihood of cardiovascular death following the non-fatal first event (3.7 versus 7.1 percent) [41]. This benefit was found in patients with either procedure-related or non-procedure related MIs and was seen irrespective of MI size or timing [42].
• The safety end point of a major bleeding event not associated with CABG occurred significantly more often in patients treated with prasugrel (2.4 versus 1.8 percent; HR 1.32; 95% CI 1.03-1.68). This difference was attributable to an increase in bleeding events with prasugrel after (but not before) the first three days [43]. The rate of life-threatening bleeding was also significantly increased with prasugrel.
• Post hoc analysis identified three predictors of less net clinical efficacy and greater absolute levels of bleeding with prasugrel, resulting in less net clinical benefit: a history of stroke or transient ischemic attack; age ≥75 years of age; and body weight ≤60 kg. In those with stroke or transient ischemic attack, the use of prasugrel was associated with harm.

These high risk bleeding groups are similar to high risk bleeding groups previously identified in fibrinolytic and IIb/IIIa antagonists trials and underscores the need for careful risk-benefit assessment in this population.

• The rate of definite or probable stent thrombosis (ARC definition) was significantly reduced in the prasugrel group (1.1 versus 2.4 percent) of patients who received both drug eluting and bare metal stents. (See "Coronary artery stent thrombosis: General issues", section on 'Definitions'.)

In a prespecified subset analysis of TRITON-TIMI 38, patients with (n = 3,146) and without (n = 10,462) diabetes were evaluated [44]. The primary endpoint was significantly reduced with prasugrel among patients with (12.2 versus 17 percent) and without diabetes (9.2 versus 10.6 percent), attributable to fewer episodes of myocardial infarction. The finding of a similar rate of major bleeding in patients with diabetes who received either clopidogrel or prasugrel used was unexpected (2.6 versus 2.5 percent respectively).

One important concern about generally applying the results of TRITON-TIMI 38 is that most patients with unstable angina or NSTEMI received clopidogrel after coronary angiography. As current we (and others) recommend that clopidogrel be give soon after presentation, we believe that the trial results should only be applied to patients who did not receive precatheterization therapy with clopidogrel. (See 'Summary and recommendations' below.)

Prasugrel appears to be an important advance in antiplatelet therapy. With appropriate patient selection for prasugrel treatment, one can expect significant benefits with respect to subacute stent thrombosis with an acceptable risk of bleeding. Based on the results of TRITON-TIMI 38, patients in whom prasugrel should be considered include those with STEMI and those with NSTEMI in whom a decision has been made to withhold thienopyridine therapy until after diagnostic coronary angiography.

Bleeding risk — The most important common side effect associated with thienopyridine therapy is bleeding. Much of the available data comes from studies of clopidogrel. Three important issues include: the rate with combination therapy with aspirin; issues related to bleeding in the subset of patients who require coronary artery bypass graft surgery (CABG), and gastrointestinal bleeding. The latter is discussed separately. (See 'Gastrointestinal bleeding' below.)

The combination of clopidogrel plus aspirin in CURE was associated with a significant increase in major (3.7 versus 2.7 percent with aspirin alone), minor (5.1 versus 2.4 percent), and gastrointestinal bleeding (1.3 versus 0.7 percent), but not life-threatening bleeding events [24]. (See "Antithrombotic therapy for intracoronary stent implantation: Clinical trials", section on 'Thienopyridines'.)

There is a concern for an increased perioperative bleeding risk with clopidogrel plus aspirin in the subset of patients with a non-ST elevation ACS who require CABG. The likelihood of life-threatening or major bleeding within seven days of CABG was nonsignificantly increased in patients in the CURE trial who had received clopidogrel within the five days prior to CABG (9.6 versus 6.3 percent with placebo) but not in those who had discontinued clopidogrel ≥5 days prior to CABG (4.4 versus 5.3 percent) [45]. There was also a nonsignificant trend toward more reoperations for bleeding in the patients who had taken clopidogrel within the five days prior to CABG (4.1 versus 2.3 percent). (See "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Bleeding'.)

The relative rate of major bleeding in patients taking clopidogrel within five days of CABG, compared to those who have not, has also been evaluated in several observational studies [46-50]. The following studies, which vary in their definition of major bleeding, illustrate the range of findings:

• The largest experience comes from a report of 2858 patients enrolled in the CRUSADE registry who underwent CABG after a non-ST elevation ACS [48]. All patients had clopidogrel stopped at the time of catheterization. There were 739 patients who received clopidogrel within five days of CABG. These patients were significantly more likely to have any red cell transfusion (65 versus 57 percent in those in whom clopidogrel was stopped more than five days before CABG) or to receive more than four units of blood (28 versus 18 percent). As in CURE, there was no increase in bleeding when CABG was performed more than five days after cessation of clopidogrel. The rate of reoperation for bleeding was not evaluated in this study.
• In a retrospective cohort analysis of patients who had an ACS within seven days of CABG, 298 matched pairs were selected based on exposure or no exposure to clopidogrel within 5 days prior to surgery [49]. Patients exposed to clopidogrel were significantly more likely to require reoperation for assessment of bleeding (6.4 versus 1.7 percent and more likely to sustain major bleeding (35 versus 26 percent).
• In a single center, observational study of 4794 patients (62 percent with recent ACS) who underwent CABG, multiple parameters of bleeding were compared between the 332 patients who received clopidogrel within five days of surgery and the 4462 who did not [50]. After multivariate adjustment for clinical factors, clopidogrel use was not significantly associated with reoperation for bleeding (odds ratio [OR] 1.24, 95% CI 0.63-2.41) or a bleeding composite consisting of reoperation, red cell transfusion, or hematocrit drop >15 percent (OR 1.23, 95% CI 0.72-2.10). Recent clopidogrel use was modestly associated with a greater rate of transfusion (OR 1.40, 95% CI 1.04-1.89), but more weakly than other factors such as which surgeon performed the operation.

Timing and concerns about early CABG — Any excess bleeding risk due to recent clopidogrel use in the minority of patients who will require CABG must be weighed against the potentially deleterious effect of delaying or not administering clopidogrel therapy [51]. Many experts give a loading dose of clopidogrel with aspirin at presentation in patients with a non-ST elevation ACS. This is consistent with the 2007 ACC/AHA guidelines for the management of patients with unstable angina/non-ST-elevation myocardial infarction, which states that non-elective surgery may be performed on patients who have received clopidogrel within five to seven days of surgery by experienced surgeons if the incremental bleeding risk is considered acceptable [15].

The preference for treating at presentation, even though some patients will require CABG, is based upon the following observations [52]:

• In a post-hoc, subgroup analysis of the 1539 patients who received CABG in the ACUITY trial (which compared bivalirudin alone or with GP IIb/IIIa inhibitors to UFH or enoxaparin with GP IIb/IIIa inhibitors in nearly 14,000 patients with non-ST-elevation acute coronary syndromes), outcomes were compared between those who received clopidogrel (n=773) and those who did not (n=747) [53]. Clopidogrel was given between randomization and two hours after angiography and a five day wash-out period before CABG was recommended. (See "Anticoagulant therapy in unstable angina and acute non-ST elevation myocardial infarction", section on 'ACUITY trial'.)

Those patients who received clopidogrel had significantly fewer adverse, composite ischemic events (death, myocardial infarction, or unplanned revascularization) than those who did not at 30 days (12.7 versus 17.3 percent). The rates of non-CABG related and post CABG bleeding were not significantly different.

• Among patients who underwent PCI in CREDO, the absolute benefit obtained from at least six hours of pretreatment (almost 2 percent reduction in death, MI, or urgent target vessel revascularization at 28 days) is greater than the added risk in the minority of patients who will require CABG [45]. (See 'Efficacy in PCI' above.)

• The frequency of requiring CABG has fallen to 6 to 8 percent in more recent studies [54,55] and is likely to fall further with the increasing use of drug-eluting stents. These stents can be safely used for multivessel disease and even unprotected left main disease and they markedly reduce the need for target vessel revascularization to levels not much higher than that following CABG (as demonstrated by the ARTS II registry). (See "Bypass surgery versus percutaneous intervention in the management of stable angina pectoris: Clinical trials", section on 'ARTS II registry' and "Use of drug-eluting intracoronary stents".)

• The increasing experience of cardiac surgeons in performing CABG in patients treated with clopidogrel, should reduce the bleeding risk [52,56,57]. In the CLARITY-TIMI 28 trial of clopidogrel therapy in patients with an ST elevation MI, there was no increase in major bleeding within 30 days overall, including those patients in whom CABG was performed during the initial hospitalization within five days of discontinuing clopidogrel [56]. (See "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Bleeding'.)

Ideally patients should be off clopidogrel therapy for five to seven days if CABG is performed, as is recommended in the 2007 ACC/AHA guidelines for the management of unstable angina/NSTEMI [14,15,54,58].

The decision whether to perform CABG within five days of clopidogrel use in an individual patient requires an analysis of the risks and benefits. Local institutional and surgeon preferences need to be taken into account. In situations where local preferences argue against early clopidogrel, initiation of intravenous glycoprotein IIb/IIIa inhibitor in the emergency room, and deferral of clopidogrel administration may be a reasonable therapeutic strategy [59]. (See 'GP IIb/IIIa inhibitors' below.)

Non-bleeding adverse effects — Three important non-bleeding side effects of thienopyridine use have been identified: neutropenia, thrombotic thrombocytopenia purpura/hemolytic uremic syndrome (TTP-HUS), both of which are primarily described with ticlopidine, and hypersensitivity, which is more common with clopidogrel. These issues are discussed separately. (See "Antithrombotic therapy for intracoronary stent implantation: General use".)

Ticagrelor therapy — Ticagrelor differs from the thienopyridines clopidogrel and prasugrel in that it binds reversibly rather than irreversibly with the P2Y12 platelet receptor and has a more rapid onset of action than clopidogrel [23]. It has been assigned to a new chemical class of antiplatelet agents, the cyclopentyltriazolopyrimidines. Similar to prasugrel, treatment with ticagrelor leads to more intense platelet inhibition than clopidogrel. Ticagrelor is not yet available for clinical use (figure 2).

The efficacy and safety of ticagrelor were evaluated in the PLATO trial in which 18,624 patients with ACS were randomly assigned to either ticagrelor (180 mg loading dose followed by 90 mg twice daily) or clopidogrel (300 to 600 mg loading dose followed by 75 mg daily) intended for one year [60]. Treatment was started as soon as possible after hospital admission (median of five hours); in many patients this was after diagnostic coronary angiography. All patients received chronic aspirin therapy. The final diagnosis of STEMI, NSTEMI, or UA was present in 38, 43 and 17 percent of patients respectively. The intent to perform PCI was mandated for patients with STEMI, but not for those with NSTEMI or UA.

At 12 months the composite primary end point (death from vascular causes, MI, or stroke) occurred significantly less often in patients receiving ticagrelor (9.8 percent versus 11.7 percent with clopidogrel, hazard ratio 0.84, 95% CI 0.77-0.92). There was no significant difference in the rates of major bleeding in the two groups (11.6 versus 11.2 percent respectively), but ticagrelor was associated with a significantly increased risk of major bleeding not related to CABG (4.5 versus 3.8 percent). In addition, the secondary end points of the rates of myocardial infarction (5.8 vs. 6.9 percent), death from vascular causes (4.0 vs. 5.1 percent), and death from any cause (4.5 vs. 5.9 percent) were significantly lower in the ticagrelor group.

Summary — Evidence from TRITON-TIMI 38 and PLATO, as well as other studies, provides strong support for the general idea that the level of platelet inhibition is correlated with both efficacy and bleeding outcomes [39]. Agents with higher levels of platelet inhibition, such as prasugrel, have lower cardiovascular event rates but higher rates of bleeding.

Thienopyridine therapy decreases the long-term incidence of adverse cardiovascular outcomes in patients with non-ST elevation ACS, including patients managed with both conservative and invasive strategies. Clopidogrel is the best studied and for that reason is preferred in most patients. The findings in TRITON-TIMI 38 suggest that prasugrel will become an important part of the antiplatelet strategy in patients with ACS. It is it is best targeted at those patients with NSTEMI who have undergone diagnostic coronary angiography and in whom a decision has been made to perform PCI. These patients should not at high risk for bleeding. We prefer prasugrel instead of clopidogrel for these patients if they are assessed to be at high risk for stent thrombosis. While there is no data on the efficacy or safety of pre-PCI loading with prasugrel, use in that setting may be considered as long as the usual tradeoff between the increased efficacy and increased bleeding risk (particularly if the patient is found to need CABG) is taken into account. (See "Coronary artery stent thrombosis: General issues", section on 'Risk factors'.)

Ticagrelor has been studied in one outcomes trial and is not yet approved for use and its role in patients with non-ST elevation ACS remains to be determined.

GP IIb/IIIa INHIBITORS — More potent inhibition of platelet aggregation may be of importance in patients with a non-ST elevation ACS. This can be achieved by the use of an intravenous GP IIb/IIIa inhibitor, which inhibits the final common pathway of platelet aggregation, the crossbridging of platelets secondary to fibrinogen binding to the activated GP IIb/IIIa receptor. (See "Platelet biology", section on 'Platelet activation'.)

The trials evaluating the effect of GP IIb/IIIa inhibitors will be briefly reviewed here, but are discussed in detail separately. Only the intravenous agents are beneficial; the oral GP IIb/IIIa inhibitors have been ineffective and may increase mortality. (See "Antithrombotic therapy for intracoronary stent implantation: Clinical trials" and "Clinical trials of platelet glycoprotein IIb/IIIa receptor inhibitors in coronary heart disease: Intravenous agents".)

Abciximab — Most of the initial clinical trials in patients with non-ST elevation ACS were conducted with abciximab. The EPIC and CAPTURE trials, for example, were performed in patients who required percutaneous transluminal coronary angioplasty (PTCA), while the GUSTO 4-ACS trial evaluated patients who did not undergo revascularization.

• The EPIC trial reported the outcome of a subset of 489 patients with a non-ST elevation ACS who were randomly assigned to receive abciximab or placebo immediately before PTCA [61]. At 30 days, abciximab reduced the primary end point of death, MI, or urgent repeat revascularization by 62 percent compared to placebo (4.8 versus 12.8 percent, respectively), primarily due to a decrease in death and MI (graph 9). The benefit was maintained up to three years after the procedure (graph 10) [62].
• Similar benefits from therapy given before PTCA were noted in the CAPTURE trial, as the 30-day incidence of the primary end point (death, MI, urgent revascularization) was significantly lower in the abciximab-treated patients (11.3 versus 15.9 percent for placebo) [63]. The benefit was primarily due to a lower rate of MI before and during PTCA [64]. Abciximab also had a short-term benefit, reducing the incidence of recurrent ischemia in the six hours after PTCA [65].
• The GUSTO 4-ACS trial evaluated the role of abciximab in 7800 patients with a non-ST elevation ACS receiving heparin and aspirin who did not undergo revascularization (in contrast to EPIC and CAPTURE [66]. The patients were randomly assigned to abciximab bolus plus a 24 hour infusion, abciximab bolus plus 48 hour infusion, or placebo. The primary end point of death or MI at 30 days was the same in the three groups (8.2 and 9.1 versus 8 percent for placebo), as was the rate of revascularization (30 percent in each group).

The lack of survival benefit of abciximab in the GUSTO 4-ACS trial persisted at one year [67]. No subgroup benefited from abciximab, but patients who had a negative serum troponin or elevated serum CRP at baseline had increased mortality in association with abciximab.

The negative findings from GUSTO 4-ACS are different from those observed in other trials with intravenous GP IIb/IIIa inhibitors, in which benefit was noted in high-risk patients who received medical therapy alone. This includes patients with elevated serum troponin I or T which, according to the current definition, identifies patients with an NSTEMI [68,69], a TIMI risk score ≥4 (calculator 1) [70], or continuing ischemia or other high-risk features (table 2) [15].

Other GP IIb/IIIa inhibitors — A number of trials have shown benefit from other GP IIb/IIIa inhibitors in patients with non-ST elevation ACS treated with aspirin and heparin. Among patients being treated with medical therapy alone, benefit has been demonstrated with eptifibatide (PURSUIT) (graph 11) [71], tirofiban (PRISM, PRISM-PLUS, and RESTORE) (graph 12 and graph 13 and graph 14) [72-74], and lamifiban (PARAGON A) (graph 15) [75]. Among patients undergoing percutaneous coronary intervention, benefit has been demonstrated with eptifibatide and tirofiban (PURSUIT, PRISM-PLUS, and ADVANCE) [71,73,76,77].

An analysis of data from 12,296 patients in CAPTURE, PURSUIT, and PRISM-PLUS showed that intravenous GP IIb/IIIa inhibitors added to heparin and aspirin significantly reduced the rate of death or nonfatal MI during drug infusion (2.5 versus 3.8 percent for placebo) and during the first 48 hours after angioplasty (4.9 versus 8 percent) [59]. In the PURSUIT trial, the benefit of eptifibatide was equivalent in those undergoing early PCI or being treated medically [76]. However, from the viewpoint of improving mortality, the benefit of these drugs may be limited to patients with diabetes (see 'Diabetes' below [78].

One other finding in PRISM-PLUS deserves mention. Patients treated with tirofiban without heparin had an increase in mortality compared to heparin alone, while tirofiban plus heparin was beneficial compared to heparin alone [73].

Timing — The optimal time of initiation of GP IIb/IIIa inhibitor therapy in patients with an NSTEMI who are treated with an early invasive strategy was addressed in the ACUITY Timing and EARLY ACS trials, the former of which was a subrandomization of the ACUITY trial of moderate to high risk ACS patients. (See "Antithrombotic therapy for intracoronary stent implantation: Clinical trials", section on 'Bivalirudin'.)

In the ACUITY Timing trial, 9207 non-ST elevation ACS patients scheduled to undergo an invasive strategy were randomly assigned to either upstream GP IIb/IIIa inhibitor (eptifibatide or tirofiban at the investigator's preference) or deferred GP IIb/IIIa inhibitor (either eptifibatide or abciximab at the investigator's preference) using the following dosing schedules [79]:

• For patients assigned to upstream therapy (ie, initiated at the time of diagnosis), the dose of eptifibatide was a 180 mcg/kg bolus plus 2.0 mcg/kg per min infusion and that of tirofiban was 0.4 mcg/kg per min bolus for 30 minutes followed by 0.1 mcg/kg per min infusion.
• For patients assigned to deferred therapy (ie, initiated in the catheterization laboratory), the dose of eptifibatide was 180 mcg/kg bolus plus 2.0 mcg/kg per min infusion with a second bolus given in 10 minutes or abciximab 0.25 mg/kg bolus plus 0.125 mcg/kg infusion, with a maximum of 10 mcg/min.

The primary outcome was assessment of noninferiority of deferred GP IIb/IIIa inhibitor use compared with upstream administration for the prevention of composite ischemic events. The following findings were noted at 30 days:

• The composite primary end point of death, MI, or unplanned revascularization occurred more often in patients in the deferred group (7.9 versus 7.1 percent; relative risk 1.12, 95% CI 0.97-1.29) and therefore the criteria for noninferiority for deferred therapy was not met.
• Deferred use resulted in significantly reduced 30 day rates of major bleeding (4.9 versus 6.1 percent). The lower rate of bleeding with deferred therapy probably reflects selective use. With deferred therapy, a GP IIb/IIIa inhibitor would only be given to those undergoing PCI, while all patients would be treated with upstream therapy.

In EARLY ACS, 9492 non-ST elevation high risk ACS patients scheduled to undergo an invasive strategy were randomly assigned to either early (after randomization) eptifibatide or placebo with provisional use of eptifibatide after angiography (delayed) [80]. High risk was defined as meeting two or more of the following criteria: ischemic changes on electrocardiography, a level of troponin or creatine kinase MB that was above the upper limit of normal, and an age of 60 years of greater. Early aspirin use (initial dose of 162 to 325 mg orally) was mandated; early clopidogrel use was encouraged by not mandated. Antithrombin therapy was used in over 94 percent of patients.

The following dosing schedules for eptifibatide were used:

• For patients assigned to upstream therapy (ie, initiated at the time of diagnosis), the dose of eptifibatide was two 180 mcg/kg boluses administered 10 minutes apart plus 2.0 mcg/kg per min infusion
• For patients assigned to provisional therapy (ie, initiated in the catheterization laboratory), the dose of eptifibatide was 180 mcg/kg bolus plus 2.0 mcg/kg per min infusion. Provisional therapy (given to about 26 percent of patients) could be administered at the investigator's discretion on the basis of clinical evidence and angiographic findings. Patients who were randomly assigned to eptifibatide received placebo as provisional therapy.

The early use of eptifibatide was associated with the following findings:

• There was no significant reduction in the rate of the primary composite outcome of death from any cause, myocardial infarction, recurrent ischemia requiring urgent revascularization, or thrombotic bailout at 96 hours (9.3 versus 10.0 percent; odds ratio [OR] 0.92, 95% CI 0.80 to 1.06).
• There was no significant reduction in the rate of the secondary composite outcome of death or MI at 30 days (11.2 versus 12.3 percent; OR 0.89, 95% CI 0.79 to 1.01).
• The likelihood of TIMI major hemorrhage was significantly increased (2.6 versus 1.8 percent; OR 0.015, 95% CI 1.07 to 1.89).

These two large trials of similar but not identical design, did not demonstrate a significant benefit of early compared with delayed use of glycoprotein IIb/IIIa inhibitor, even in high risk patients with non-STEMI ACS who are scheduled to undergo early PCI. However, a benefit in the primary composite outcome of as much as twenty percent cannot be excluded. A trend toward a reduction in ischemic events was noted but, as expected, GP IIb/IIIa use was associated with a significantly increased risk of bleeding. (See 'Bleeding risk' above.)

Diabetes — From the viewpoint of reducing mortality, intravenous GP IIb/IIIa inhibitors are particularly beneficial in diabetic patients. In a meta-analysis of six randomized trials that enrolled 6458 diabetic patients, therapy with the GP IIb/IIIa inhibitors was associated with a significant reduction in overall 30-day mortality (4.6 versus 6.2 percent, odds ratio 0.74) and in 30-day mortality in those undergoing a PCI (1.2 versus 4 percent, odds ratio 0.3) compared to placebo [78]. In contrast, there was no survival benefit with GP IIb/IIIa inhibitors in the 23,070 nondiabetic patients (3 percent in both groups). (See "Treatment of acute myocardial infarction in diabetes mellitus".)

Prior CABG — Patients with prior CABG who present with a non-ST elevation ACS have a worse prognosis compared to those without a prior CABG. (See "Classification of unstable angina and non-ST elevation myocardial infarction", section on 'Postrevascularization angina'.)

The role of eptifibatide in these patients was evaluated in the PURSUIT trial of almost 11,000 patients, 1134 of whom (12 percent) had had a prior CABG [81]. The patients with prior CABG had significantly increased mortality rates at 30 days and six months compared to patients without prior CABG (hazard ratio 1.45 and 1.32, respectively). Among prior CABG patients, eptifibatide therapy was associated with a trend toward a lower 30-day rate of death or MI that was similar in magnitude to that seen in patients without a prior CABG.

Serum troponins — An interesting observation from the PRISM, CAPTURE, and PARAGON B trials and a meta-analysis is that the benefit from GP IIb/IIIa inhibition primarily occurred in the subset of patients who had elevations in troponin T and/or I [68,69,82,83]. In PRISM, this benefit was seen whether patients managed were medically or with revascularization (graph 16) [69]. The same pattern of benefit limited to patients with elevated troponins was also noted in the ISAR-REACT 2 trial of patients also treated with clopidogrel [84]. (See 'Use with clopidogrel' below.)

Increases in serum troponin T and I are considered to be a surrogate marker for thrombus formation, since they are associated with complex lesion characteristics and visible thrombus [85]. Such patients may be most likely to benefit from a GP IIb/IIIa inhibitor.

Meta-analysis — In order to more fully understand the efficacy and safety of intravenous GP IIb/IIIa inhibitors given with aspirin and heparin in patients with a non-ST elevation ACS, a meta-analysis examined individual data from six randomized trials of 31,400 patients who did not undergo early (<48 hours) revascularization during study drug infusion [68].

A limitation to applying the following findings to current practice is that the patients were not treated with clopidogrel. The limited data on the role of GP IIb/IIIa inhibitors in patients who are treated with clopidogrel is presented in the next section.

• GP IIb/IIIa inhibitors were associated with a significant reduction in the combined end point (death or MI) at five days (5.7 versus 6.9 percent for placebo/control, odds ratio [OR] 0.84, 95 percent confidence interval [CI] 0.77 to 0.93) and at 30 days (10.8 versus 11.8 percent, OR 0.91, 95% CI 0.85-0.98); a reduction was also seen for the individual end points of death and MI.
• The benefit of therapy was limited to the 38 percent of patients in whom PCI or CABG was performed within 30 days (OR for death or MI 0.89, 95% CI 0.80-0.98). In the other patients who did not undergo coronary revascularization, the OR for death or MI was not significantly reduced (OR 0.95, 95% CI 0.86-1.05).
• The benefit of therapy also appeared to be limited to patients with a positive serum troponin T or I concentration (≥0.1 ng/mL) (OR 0.85, 95% CI 0.71-1.03). In contrast, there was no benefit in those with concentrations <0.1 ng/mL.
• Gender may be important. GP IIb/IIIa inhibitors significantly decreased the combined end point at 30 days in men (10.4 versus 12.6 percent, OR 0.81, 95% CI 0.75-0.89); in comparison, there was a significant increase in the end point in women (11.5 versus 10.4, OR 1.14, 95% CI 1.01-1.30). This difference in outcome disappeared when troponin status was included, since men more often presented with positive baseline troponins.
• GP IIb/IIIa inhibitors were associated with an increased risk of major bleeding complications compared to placebo/control (2.4 versus 1.4 percent, OR 1.64, 95% CI 1.36-1.97).

Use with clopidogrel — The above meta-analysis was published before the current use of clopidogrel in patients with a non-ST elevation ACS [68], raising the possibility that GP IIb/IIIa inhibitors might not be of benefit in patients already treated with clopidogrel.

This issue was directly addressed in the ISAR-REACT 2 trial of 2022 patients with a non-ST elevation ACS undergoing PCI with stenting who were treated with clopidogrel (600 mg loading dose at least two hours before PCI) and were then randomly assigned to abciximab or placebo in the catheterization laboratory. Abciximab therapy was associated with a significant reduction in the primary end point of death, MI, or urgent target vessel revascularization at 30 days and one year. However, on subgroup analysis, the benefit was only seen in the approximately one-half of patients with an elevated serum troponin concentration. (See "Antithrombotic therapy for intracoronary stent implantation: Clinical trials", section on ISAR-REACT 2 trial.)

Combined use with enoxaparin — Although most of the trials have evaluated the safety and efficacy of the GP IIb/IIIa agents in patients receiving unfractionated heparin, these drugs are also safe and of benefit when used with enoxaparin, a low molecular weight heparin. Enoxaparin may be associated with improved outcomes compared to unfractionated heparin, particularly in high-risk patients. (See "Anticoagulant therapy in unstable angina and acute non-ST elevation myocardial infarction", section on 'Meta-analyses of enoxaparin versus UFH'.)

Bleeding risk — Major bleeding can occur after the administration of a GP IIb/IIIa inhibitor as a consequence of inhibition of platelet function or thrombocytopenia. The mechanism and frequency of thrombocytopenia, as well as recommendations for measurement of platelet count, are discussed separately. (See "Thrombocytopenia induced by glycoprotein IIb/IIIa inhibitors".)

In randomized, placebo-controlled clinical trials of the GP IIb/IIIa inhibitors, abciximab had a 0 to 7 percent excess risk of major bleeding compared to placebo, while tirofiban and eptifibatide had a 0 to 2 percent excess risk. (See "Clinical trials of platelet glycoprotein IIb/IIIa receptor inhibitors in coronary heart disease: Intravenous agents".)

However, most of these trials were completed before the stent era in which the administration of both aspirin and clopidogrel have become routine. Data on the rates of major bleeding in patients with non-ST elevation ACS treated with aspirin and clopidogrel are available from the ISAR-REACT 2 trial and the CRUSADE initiative:

• In ISAR-REACT 2, all patients received pretreatment with aspirin and 600 mg of clopidogrel [84]. The rates of major bleeding for patients receiving abciximab or placebo were not significantly different (4.5 and 4.0, respectively).
• In an analysis of 32,600 patients with a non-ST elevation ACS in the CRUSADE initiative, the rate of major bleeding, defined as a drop in hematocrit of ≥12 percent, need for transfusion, or intracranial hemorrhage, was significantly higher in patients who received GP IIb/IIIa inhibitors (10.1 versus 6.8 percent) [86]. Some of this increase in risk may have been due to clopidogrel, since the rate of clopidogrel use was higher in the patients treated with a GP IIb/IIIa inhibitor (63 versus 39 percent in those who did not receive a GP IIb/IIIa inhibitor).

Subsets of patients at increased risk for GP IIb/IIIa-induced major bleeding have been identified:

• Patients with moderate to severe renal insufficiency are at higher risk for bleeding after administration of GP IIb/IIIa inhibitors [87,88]. The potential risk does not interfere with the efficacy of these drugs, in part since such patients are often at higher risk for ischemic complications. Abciximab is preferred for severe renal insufficiency, although eptifibatide may be used at a modified dose for moderate renal insufficiency.
• Elderly patients are at increased risk of bleeding after PCI, with a rate of 9 percent noted in a series of octogenarians [89]. The risk of bleeding requiring transfusion or major bleeding did not appear to be influenced by whether or not a GP IIb/IIIa inhibitor was given.
• In the report from the CRUSADE initiative cited above, women had a significantly higher rate of major bleeding than men among those treated with GP IIb/IIIa inhibitors (15.7 versus 7.3 percent) and also among those not treated (8.5 versus 5.4 percent) with these drugs [86]. Treated women were much more likely than men to receive excess GP IIb/IIIa doses, defined as the absence of dose adjustment for a reduced creatinine clearance (46 versus 17 percent). The smaller muscle mass in women contributes to excessive dosing.

Thus, the use of GP IIb/IIIa inhibitors in elderly patients and women requires careful weight and renal function-based dose adjustment, and should be limited to circumstances when GP IIb/IIIa drugs are clearly needed [90].

Summary — Based upon the evidence cited above, GP IIb/IIIa inhibitor therapy is warranted in most patients treated with PCI and in high-risk patients treated with medical therapy alone.

OTHER DRUGS — Neither sulfinpyrazone nor dipyridamole appear to confer additional benefit in patients with a non-ST elevation ACS [5,7]. Thus, neither drug is recommended [14,91].

GASTROINTESTINAL BLEEDING — A history of or active gastrointestinal bleeding presents a challenge to the care of the patient with ACS. The potential reductions in tissue perfusion and hematocrit associated with bleeding can worsen the ACS [92]. On the other hand, the hemodynamic instability and stress associated with the MI, as well as the vigorous antithrombotic therapies (including multiple antiplatelet agents), may promote bleeding.

The risk of gastrointestinal bleeding for a patient taking aspirin alone has been relatively well studied. The risk in patients with ACS who have received dual or triple antiplatelet therapy, with or with out antithrombotic therapy, has been less well studied.

The rate of upper gastrointestinal bleeding in patients receiving aspirin, clopidogrel and enoxaparin therapy at the same time or within seven days of stopping enoxaparin was 2.7 percent in an observational study of 666 ACS patients [93]. Prior peptic ulcer, cardiogenic shock, and the lack of proton pump inhibitor treatment were significant risk factors.

In a review of over 3100 patients who underwent primary PCI for acute ST elevation MI in the PAMI trials, gastrointestinal bleeding developed in 2.3 percent [94]. The major risk factor was age greater than 70. The affected patients had significant increases in hospital stay, in-hospital mortality (9.8 versus 2.8 percent percent, adjusted odds ratio 4.0) and six-month mortality (14.1 versus 4.6 percent, adjusted odds ratio 2.8). Almost all patients were treated with aspirin, but most were not treated with a thienopyridine.

If gastrointestinal bleeding occurs in the peri-ACS period, we make every effort to continue aspirin and clopidogrel unless the bleeding is life-threatening, particularly if stenting has been performed. We also obtain an urgent consultation with a gastroenterologist for endoscopic evaluation and therapy. Such patients should be maintained on a proton pump inhibitor to minimize the risk of recurrent bleeding [95]. (See "NSAIDs (including aspirin): Secondary prevention of gastroduodenal toxicity", section on 'With continued low-dose aspirin'.)

The substantially worse outcomes in patients who develop gastrointestinal bleeding after PCI [94] raises the question of whether patients at increased bleeding risk would benefit from prophylaxis with a proton pump inhibitor. This issue is discussed separately. (See "Overview of the management of unstable angina and acute non-ST elevation myocardial infarction", section on 'Gastrointestinal prophylaxis'.)

The discussion of the long term primary prevention of gastrointestinal bleeding is found elsewhere. (See "Benefits and risks of aspirin in secondary and primary prevention of cardiovascular disease", section on Aspirin associated bleeding.)

SUMMARY AND RECOMMENDATIONS

Summary — Platelets play a central role in the development of non-ST elevation ACS and antiplatelet agents have been shown to improve clinical outcomes. Evidence strongly supports the early initiation of dual antiplatelet therapy with aspirin and either clopidogrel or GP IIb/IIIa inhibitor in all patients with non-ST elevation ACS, whether they are managed by a conservative or early invasive strategy.

We feel that the data are sufficient to recommend GP IIb/IIIa inhibitors in addition to thienopyridine and aspirin in most high risk ACS patients (eg, markedly elevated troponin, recurrent ischemic discomfort, dynamic electrocardiographic changes, or hemodynamic instability) undergoing an invasive strategy. This strategy may also be appropriate for high risk patients managed with medical therapy. For those stable patients undergoing an invasive strategy who have not received a GP IIb/IIIa inhibitor, the addition of one at the time of percutaneous coronary intervention is appropriate in most cases. (See 'GP IIb/IIIa inhibitors' above.)

For patients referred for urgent CABG, the thienopyridine should be held for five days if possible. (See 'Timing and concerns about early CABG' above.)

Available evidence supports the daily use of aspirin indefinitely and thienopyridine for at least one year after non-ST elevation ACS. (See 'Aspirin' above and 'Thienopyridines and ticagrelor' above.)

Early therapy

• For nearly all patients with non-ST elevation ACS, we recommend dual antiplatelet therapy (Grade 1A). We suggest aspirin plus clopidogrel as the preferred regimen when they can be given as soon as possible after presentation (Grade 2B).

• - For those patients undergoing an early invasive approach in whom a decision is made to withhold clopidogrel until after coronary angiography and who are not at high risk of bleeding (age <75 years, weight ≥60 kilograms, or those without prior transient ischemic attack or stroke), we suggest prasugrel as opposed to clopidogrel (Grade 2B).
• - Aspirin plus a GPIIb/IIIa inhibitor is a reasonable alternative and is the preferred strategy in patients known to need CABG where the surgeon will not operate with clopidogrel on board. (See 'Clopidogrel therapy' above.)
• - The first aspirin tablet should contain 162 to 325 mg and should be chewed.
• - The loading dose of clopidogrel is 300 mg. However, if the patient is going to catheterization the same day, we suggest clopidogrel 600 mg.
• - For those patients in whom prasugrel is given after diagnostic coronary angiography, the loading dose is 60 mg.
• - For those patients with a history of gastrointestinal bleeding, drugs which reduce the risk of recurrent bleeding (eg, proton pump inhibitors) should be given. (See "Overview of the management of unstable angina and acute non-ST elevation myocardial infarction", section on 'Gastrointestinal prophylaxis'.)

• For those high risk patients scheduled for an early invasive approach in whom a decision is made to withhold thienopyridine until after diagnostic coronary angiography, we recommend initiation of a GP IIb/IIIa inhibitor as soon as possible after presentation (Grade 1A). Tirofiban and eptifibatide are continued for 18 hours after PCI, or 12 hours if abciximab is chosen. (See 'GP IIb/IIIa inhibitors' above.)
• For very high risk patients (eg, markedly elevated troponin, recurrent ischemic discomfort, dynamic electrocardiographic changes, or hemodynamic instability) undergoing an invasive approach, we suggest the addition of a GP IIb/IIIa inhibitor to thienopyridine and aspirin (Grade 2B). (See 'Use with clopidogrel' above.)

• - The GP IIb/IIIa inhibitor infusion should be initiated at presentation and continued through the PCI and following PCI for 12-18 depending on the agent.
• - Eptifibatide is generally preferred because its cost is substantially lower than abciximab and its optimal dosing regimen is more clearly defined than with tirofiban. (See "Antithrombotic therapy for intracoronary stent implantation: General use".)
• - In patients who have received aspirin, clopidogrel 300 mg at least six hours (or 600 mg at least two hours) before the procedure and anticoagulant therapy with bivalirudin, a GP IIb/IIIa inhibitor may be withheld before the catheterization.

• For very high risk patients (eg, elevated troponin, recurrent ischemic discomfort, dynamic electrocardiographic changes, or hemodynamic instability) who will be managed with a conservative approach, we suggest adding a GP IIb/IIIa inhibitor (Grade 2B). (See 'GP IIb/IIIa inhibitors' above.)

• - Either eptifibatide or tirofiban are preferred.
• - The GP IIb/IIIa inhibitor infusion should be initiated at presentation and continued for 48 to 72 hours. Dose adjustment may be necessary in the presence of chronic kidney disease.

Long-term therapy

We recommend indefinite aspirin for all patients with non-ST elevation ACS (Grade 1A). (See "Benefits and risks of aspirin in secondary and primary prevention of cardiovascular disease", section on Non-ST elevation ACS.)

• - For patients who have not undergone stenting we use aspirin 75 to 162 mg/day. We use a higher dose of aspirin (162 to 325 mg daily) for one month in patients with a bare metal stent, for three months for those who received a sirolimus-eluting stent, and for six months for those who received a paclitaxel-eluting stent. (See "Benefits and risks of aspirin in secondary and primary prevention of cardiovascular disease", section on non-ST elevation MI.)
• - Clopidogrel (75 mg daily) and ticlopidine (250 mg twice daily) have equal efficacy to aspirin for long-term therapy and are alternatives in patients who cannot tolerate aspirin because of major gastrointestinal intolerance or hypersensitivity. We prefer clopidogrel due to its better side effect profile. (See 'Ticlopidine therapy' above.)

• We recommend the duration of clopidogrel (75 mg/day) or prasugrel (10 mg for patients ≥60 kg or 5 mg for patients <60 kg) be at least one year in all patients with non-ST elevation ACS (Grade 1A). (See 'Thienopyridines and ticagrelor' above.)
• Until data are available defining the optimal duration of clopidogrel therapy in this setting, we suggest continuing clopidogrel indefinitely after DES if the patient is tolerating the medication without bleeding (Grade 2C). If prasugrel was used, it should be continued for 15 months. (See "Coronary artery stent thrombosis: Prevention and management".)