2009 Focused Updates: ACC/AHA Guidelines for the Management of Patients With ST-Elevation Myocardial Infarction (Updating the 2004 Guideline and 2007 Focused Update) and ACC/AHA/SCAI Guidelines on Percutaneous Coronary Intervention (Updating the 2005 Guideline and 2007 Focused Update)

A Report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines

1.1 Methodology and Evidence Review

Late-breaking clinical trials presented at the 2007 and 2008 annual scientific meetings of the ACC, AHA, Transcatheter Cardiovascular Therapeutics, the European Society of Cardiology, and the 2009 annual scientific sessions of the ACC were reviewed by the standing guideline writing committee along with the parent Task Force and other experts to identify those trials and other key data that may impact guideline recommendations. On the basis of the criteria/considerations noted above, recent trial data and other clinical information were considered important enough to prompt a focused update of the ACC/AHA 2004 Guidelines for the Management of Patients With ST-Elevation Myocardial Infarction and the ACC/AHA 2005 Guidelines for Percutaneous Coronary Intervention, inclusive of their respective 2007 focused updates (2–5).

The ST-elevation myocardial infarction (STEMI) and percutaneous coronary intervention (PCI) writing groups together considered the following studies: Two meta-analyses, “A Comparison of Abciximab and Small Molecule Glycoprotein IIb/IIIa Inhibitors in Patients Undergoing Primary Percutaneous Coronary Intervention,” (6) and “Benefits From Small Molecule Administration as Compared With Abciximab Among Patients With ST-Segment Elevation Myocardial Infarction Treated With Primary Angioplasty,” (7) FINESSE (Facilitated PCI in Patients With ST-Elevation Myocardial Infarction) (8), the HORIZONS-AMI (Harmonizing Outcomes With Revascularization and Stents in Acute Myocardial Infarction) (9), BRAVE-3 (Bavarian Reperfusion Alternatives Evaluation-3) (10), MULTISTRATEGY (Multicentre Evaluation of Single High-Dose Bolus Tirofiban Versus Abciximab With Sirolimus-Eluting Stent or Bare Metal Stent in Acute Myocardial Infarction Study) (11), ON-TIME 2 (Ongoing Tirofiban in Myocardial Infarction Evaluation) (12), TRITON-TIMI 38 (Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel–Thrombolysis in Myocardial Infarction) (13), TRANSFER-AMI (Trial of Routine ANgioplasty and Stenting after Fibrinolysis to Enhance Reperfusion in Acute Myocardial Infarction) (14), CARESS-in-AMI (Combined Abciximab Reteplase Stent Study in Acute Myocardial Infarction) (15), NICE-SUGAR (Normoglycemia in Intensive Care Evaluation—Survival Using Glucose Algorithm Regulation) (16), TAPAS (Thrombus Aspiration during Percutaneous coronary intervention in Acute myocardial infarction Study) (17), and EXPIRA (Thrombectomy With Export Catheter in Infarct-Related Artery During Primary Percutaneous Coronary Intervention) (18). Additionally, the PCI writing group considered the CARE (Cardiac Angiography in Renally Impaired Patients) (19), FAME (Fractional Flow Reserve versus Angiography for Multivessel Evaluation) study (20), SYNTAX (Synergy Between Percutaneous Intervention With Taxus and Cardiac Surgery) (21), Early ACS (Early versus Delayed, Provisional Eptifibatide in Acute Coronary Syndromes) (22), and TIMACS (Timing of Intervention in Patients With Acute Coronary Syndromes) studies (23). When considering the new data for this focused update, the writing group faced the task of weighing evidence from studies that had enrolled large numbers of subjects outside North America. Although noting that practice patterns and the rigor applied to data collection, as well as the genetic makeup of subjects, may influence the observed magnitude of a treatment's effect, the writing group believed the data were relevant to the formulation of recommendations for management of STEMI and PCI in North America. The writing group also notes that the AHA/ACCF and the Heart Rhythm Society have published updated recommendations for the standardization and interpretation of the electrocardiogram with a separate section on acute ischemia/infarction (24).

To provide clinicians with a comprehensive set of data, whenever possible, the exact event rates in various treatment arms of clinical trials are presented to permit calculation of the absolute risk difference and number needed to treat (NNT) or harm; the relative treatment effects are described either as odds ratio, relative risk (RR), or hazard ratio (HR) depending on the format used in the original publication. Along with all other statistical point estimates, the confidence interval (CI) for those statistics are added when available.

Consult the full-text or executive summary versions of the ACC/AHA 2004 Guidelines for the Management of Patients With ST-Elevation Myocardial Infarction or the ACC/AHA/SCAI 2005 Guidelines for Percutaneous Coronary Intervention, as well as their respective 2007 focused updates, for policy on clinical areas not covered by the present focused update (2–5). Unchanged recommendations from previous iterations of the guidelines are not listed in this document and remain current policy. Individual recommendations updated in this focused update will be incorporated into future revisions of the full-text guidelines.

1.2 Organization of Committee and Relationships With Industry and Other Entities

For this focused update, all members of the 2004 STEMI guideline, 2007 STEMI focused update, 2005 PCI guideline, and 2007 PCI focused update writing committees were invited to participate; those who agreed (referred to as the 2009 Focused Update Writing Group) were required to disclose all relationships with industry and other entities relevant to the data under consideration. The policies used for relationships with industry were those in effect at the initial meeting of this committee, which included disclosure of relationships 12 months prior to initiation and a chair with no relevant relationships except in a situation where more than one chair is named. In this circumstance, one chair will have no relevant relationships and the other may have relationships. Each recommendation required a confidential vote by the writing group members before and after external review of the document. Any writing group member with a relationship with industry relevant to the recommendation was recused from voting on that recommendation. The PCI writing group included 2 representatives from SCAI.

1.3 Document Review and Approval

This document was reviewed by 3 official reviewers nominated by the ACCF and 4 official reviewers nominated by the AHA, 1 official reviewer nominated by the SCAI, 6 reviewers from the ACCF Interventional Council, 2 reviewers from the ACCF Imaging Council, and 22 content reviewers. All reviewer information on relationships with industry and other entities was collected and distributed to the writing committee and is published in Appendix 3. This document was approved for publication by the governing bodies of the ACCF, the AHA, and the SCAI (specifically, the PCI portion of the guideline).

2.1 Glycoprotein IIb/IIIa Receptor Antagonists

In considering the use of intravenous glycoprotein (GP) IIb/IIIa receptor antagonists for STEMI, the writing group noted that much of the evidence favoring the use of these agents was established in the era before dual oral antiplatelet therapy and largely by placebo-controlled comparisons. Contemporary management of STEMI patients involves a complex array of antithrombotics, including dual oral antiplatelet therapy (aspirin [acetylsalicylic acid; ASA] plus a thienopyridine) and an anticoagulant. There is a paucity of trials adequately powered for assessment of clinical end points that have reevaluated the current relative role of intravenous GP IIb/IIIa receptor antagonists with respect to other pharmacological therapy in STEMI patients. Accordingly, a reevaluation of the value of GP IIb/IIIa antagonists in STEMI is appropriate, but the ability to draw definitive conclusions is limited.

At least 3 trials evaluated GP IIb/IIIa antagonists as adjuncts to oral antiplatelet therapy in the setting of primary PCI. The findings of these trials question whether GP IIb/IIIa antagonists provide significant additional benefit to STEMI patients who have received dual-antiplatelet therapy before catheterization. In the BRAVE-3 study, 800 patients presenting within 24 hours of a STEMI were pretreated with 600 mg of clopidogrel and then randomly assigned in a double-blind manner to receive either abciximab or placebo in the intensive care unit before being sent for PCI (10). The primary end point was infarct size measured by single photon emission computed tomography before hospital discharge. At 30 days, the composite of death, recurrent myocardial infarction (MI), stroke, or urgent revascularization of the infarct-related artery was not significantly different in the 2 groups (abciximab 5%, placebo 3.8%; 95% CI 0.7 to 2.6; P=0.4). There was no significant difference in infarct size or major bleeding.

ON-TIME 2 was a randomized, placebo-controlled, multicenter European trial that included 491 patients receiving high-dose tirofiban and 493 receiving placebo within a median of 76 minutes from onset of symptoms (12). Patients receiving high-dose tirofiban (25 mcg/kg bolus followed by 0.15 mcg/kg per min for 18 hours) at first medical contact before transport for primary PCI were also treated with unfractionated heparin (UFH; 5000 U), clopidogrel (600 mg), and ASA. Patients in the high-dose tirofiban group had improved ST-segment resolution (primary end point) before and 1 hour after PCI (P=0.003) compared with those receiving placebo (NNT=100). However, there was no significant difference in Thrombolysis In Myocardial Infarction (TIMI) grade 3 flow or blush grade and no significant difference in major bleeding or minor bleeding. There was no significant difference in death, recurrent MI, or urgent target-vessel revascularization (TVR) between the tirofiban and placebo groups at 30 days (25).

In the HORIZONS-AMI trial (9), patients undergoing primary PCI for STEMI were randomized to treatment with UFH plus a GP IIb/IIIa receptor antagonist (abciximab or double-bolus eptifibatide) or to bivalirudin alone with provisional IIb/IIIa. Aspirin and a thienopyridine were administered before catheterization. (See the full discussion of the trial under Section 4, Recommendations for the Use of Parenteral Anticoagulants.) Seven hundred fifty-seven of the 1661 patients who received UFH received a double bolus of eptifibatide and infusion, whereas 53 of 1661 in the bivalirudin arm received eptifibatide. At 30 days, rates of major bleeding and total adverse events were higher among patients treated with GP IIb/IIIa antagonists and heparin than among those given bivalirudin alone.

Two meta-analyses of randomized trials were published that compared small-molecule GP IIb/IIIa antagonists with abciximab in STEMI patients undergoing primary PCI (6,7). In each case, there was no statistically significant difference in 30-day mortality, reinfarction, or major TIMI bleeding, and there was no significant difference in death or reinfarction at 8 months between groups. There was also no statistically significant difference in postprocedural TIMI flow grade 3 or ST-segment resolution. On the basis of these studies, the present writing group judged that the totality of evidence indicates that the various GP IIb/IIIa antagonists demonstrate similar effectiveness in the setting of primary PCI.

MULTISTRATEGY was an open-label, multicenter, randomized European trial with a 2-by-2 factorial design that randomized 745 STEMI patients undergoing primary PCI to high-dose bolus tirofiban versus abciximab infusion and sirolimus-eluting stent versus bare-metal stent (BMS) (11). The prespecified primary end points were the achievement of 50% resolution of ST-segment elevation at 90 minutes after PCI, powered for noninferiority, and the rate of major adverse cardiac events (MACE) at 8 months, powered for superiority. All patients received ASA at the usual doses, clopidogrel 300 mg orally then 75 mg per day, and UFH. There was a similar rate of at least 50% ST-segment resolution at 90 minutes after primary PCI with abciximab and tirofiban (RR 1.020; 97.5% CI 0.958 to 1.086; P=0.001 for noninferiority). Rates of MACE, including all-cause death, clinical reinfarction, or TVR, and hemorrhagic (major and minor bleeding) complications were similar. The incidence of severe or moderate thrombocytopenia was more common with abciximab than with tirofiban (4.0% versus 0.8%, P=0.004).

In an analysis of the predictors of stent thrombosis after primary PCI in acute MI presented at the 2009 ACC Scientific Sessions, titled “Predictors of Stent Thrombosis After Primary Angioplasty in Acute Myocardial Infarction: The HORIZONS-AMI Trial,” (69) there was no significant difference in the 1-year rate of stent thrombosis with the heparin plus GP IIb/IIIa receptor antagonists compared with eptifibatide and abciximab (3.6% versus 2.8%, P=0.93), which suggests that eptifibatide has the same impact as abciximab on stent thrombosis incidence.

One investigation, FINESSE, addressed the issue of timing of GP IIb/IIIa antagonist administration. This double-blind, randomized, placebo-controlled study of 2453 patients with STEMI explored the use of pre-PCI treatment with a half-dose fibrinolytic agent plus abciximab, pre-PCI abciximab alone, and abciximab at the time of PCI (8). The primary end point was the composite of death due to all causes, ventricular fibrillation that occurred more than 48 hours after randomization, cardiogenic shock, and congestive heart failure during the first 90 days after randomization. The results of the trial are discussed in Section 5.1, Triage and Transfer for PCI. This trial showed no benefit (and a tendency toward excess bleeding) with prehospital abciximab compared with abciximab at the time of PCI. The writing group concluded there was no benefit of administration of abciximab before primary PCI, alone or in combination with reteplase. On the basis of this trial and ON-TIME 2, the writing group concluded that the use of GP IIb/IIIa antagonists before primary PCI is of uncertain benefit.

Given the results of the studies cited above, the writing group concluded that in the setting of dual-antiplatelet therapy with UFH or bivalirudin as the anticoagulant, current evidence indicates that adjunctive use of a GP IIb/IIIa antagonist can be useful at the time of primary PCI but cannot be recommended as routine therapy. These agents might provide more benefit in selective use, for example, for the patient with a large thrombus burden or for patients who have not received adequate thienopyridine loading.

3.1 Thienopyridines

Since the publication of the last guidelines (4,5), evidence has emerged about prasugrel, a thienopyridine that achieves greater inhibition of platelet aggregation than clopidogrel (27). The pivotal trial for prasugrel, TRITON-TIMI 38, focused on patients with ACS who were referred for PCI.

TRITON-TIMI 38 randomly assigned 13 608 patients with moderate- to high-risk ACS, 3534 of whom had STEMI, to receive prasugrel (6813 patients received a 60-mg loading dose and a 10-mg daily maintenance dose) or clopidogrel (6795 patients received a 300-mg loading dose and a 75-mg daily maintenance dose) for an average follow-up of 14.5 months. Aspirin was prescribed within 24 hours of PCI. Clinical end points were assessed at 30 and 90 days and then every 3 to 15 months (27).

Prasugrel was associated with a significant 2.2% absolute reduction and a 19% relative reduction in the primary efficacy end point, a composite of the rate of death due to cardiovascular causes (including arrhythmia, congestive heart failure, shock, and sudden or unwitnessed death), nonfatal MI, or nonfatal stroke during the follow-up period. The primary efficacy end point occurred in 9.9% of patients receiving prasugrel and 12.1% of patients receiving clopidogrel (HR for prasugrel versus clopidogrel 0.81; 95% CI 0.73 to 0.90; P<0.001). A significant reduction in the primary end point was seen in the prasugrel group by the first prespecified time point, which was 3 days (4.7% in the prasugrel group versus 5.6% in the clopidogrel group; HR 0.82; 95% CI 0.71 to 0.96; P=0.01), and persisted throughout the follow-up period. From Day 3 to the end of the study, the primary end point had occurred in 5.6% of patients receiving prasugrel and in 6.9% of patients receiving clopidogrel (HR 0.80; 95% CI 0.70 to 0.93; P=0.003). Prasugrel decreased cardiovascular death, MI, and stroke by 138 events (NNT=46) (27). The rate of MI with subsequent death due to cardiovascular causes was also reduced in the prasugrel group (P=0.02). The difference in the primary end point was largely related to the difference in rates of nonfatal MI (7.3% for prasugrel versus 9.5% for clopidogrel; HR 0.76; 95% CI 0.67 to 0.85; P<0.001). There were no significant differences in the 2 treatment groups in the rates of stroke or of death due to cardiovascular causes not preceded by recurrent MI (at 15 months, the nonfatal stroke rate was 1.0% for both prasugrel and clopidogrel; HR for prasugrel=1.02; CI 0.71 to 1.45; P=0.93; the rate of deaths due to cardiovascular causes not preceded by recurrent MI was 2.1% for prasugrel versus 2.4% for clopidogrel; HR 0.89; CI 0.70 to 1.12; P=0.31). There were significant reductions in the rates of ischemic events in the prasugrel group compared with the clopidogrel group: Rates of MI were 7.4% for prasugrel versus 9.7% for clopidogrel (P<0.001); urgent TVR rates were 2.5% for prasugrel versus 3.7% for clopidogrel (P<0.001); and rates of stent thrombosis were 1.1% for prasugrel versus 2.4% for clopidogrel (HR 0.48; 95% CI 0.36 to 0.64; P<0.001).

Prasugrel was associated with a significant increase in the rate of bleeding, notably, TIMI major hemorrhage, which was observed in 2.4% of patients taking prasugrel and in 1.8% of patients taking clopidogrel (HR for prasugrel versus clopidogrel 1.32; 95% CI 1.03 to 1.68, P=0.03), which represented an increase in the relative rate of major bleeding of 32%. From the standpoint of safety, prasugrel was associated with an increase of 35 TIMI major and non–coronary artery bypass graft bleeds (number needed to harm=167) (27). Also, greater rates of life-threatening bleeding were evident in the prasugrel group than in the clopidogrel group: 1.4% versus 0.9%, respectively (HR for prasugrel 1.52; 95% CI 1.08 to 2.13; P=0.01), which included nonfatal bleeding (1.1% versus 0.9%; HR for prasugrel 1.25; 95% CI 0.87 to 1.81; P=0.23) and fatal bleeding (0.4% versus 0.1%; HR for prasugrel 4.19; 95% CI 1.58 to 11.11; P=0.002). In the few patients who underwent coronary artery bypass graft (CABG), TIMI major bleeding through 15 months was also greater with prasugrel than with clopidogrel (13.4% versus 3.2%, respectively; HR for prasugrel 4.73; 95% CI 1.90 to 11.82; P<0.001) (27). Despite the increase in bleeding, the net clinical-benefit end point, which included all-cause mortality, ischemic events, and major bleeding events, favored prasugrel (27).

Prasugrel showed superior efficacy in major prespecified subgroups in the overall ACS population. The benefit tended to be greater among the 3146 patients with diabetes (12.2% of whom had the primary end point in the prasugrel group versus 17.0% in the clopidogrel group; HR 0.70; 95% CI 0.58 to 0.85; P<0.001) than among the 10 462 patients without diabetes (9.2% of whom had the primary end point in the prasugrel group versus 10.6% in the clopidogrel group; HR 0.86; 95% CI 0.76 to 0.98; P=0.02). The rate of definite or probable stent thrombosis was significantly reduced in the prasugrel group compared with the clopidogrel group, as noted (27).

A post hoc analysis suggested there were 3 subgroups of ACS patients who did not have a favorable net clinical benefit (defined as the rate of death due to any cause, nonfatal MI, nonfatal stroke, or non–CABG-related nonfatal TIMI major bleeding) from the use of prasugrel or who had net harm: Patients with a history of stroke or transient ischemic attack (TIA) before enrollment had net harm from prasugrel (HR 1.54; 95% CI 1.02 to 2.32; P=0.04), patients 75 years of age and older had no net benefit from prasugrel (HR 0.99; 95% CI 0.81 to 1.21; P=0.92); and patients with a body weight of less than 60 kg had no net benefit from prasugrel (HR 1.03; 95% CI 0.69 to 1.53; P=0.89). In both treatment groups, patients with at least 1 of these risk factors had higher rates of bleeding than those without them (27). A pharmacokinetic analysis showed greater exposure to the active metabolite of prasugrel for patients who weighed less than 60 kg and who were 75 years old or older (38).

The US Food and Drug Administration (FDA) approved prasugrel in July 2009 and incorporated the aforementioned subgroup findings into its labeling by citing a contraindication against prasugrel use in patients with a history of TIA or stroke and active pathological bleeding. The FDA further recommends that consideration be given to lowering the maintenance dose of prasugrel to 5 mg in patients who weigh less than 60 kg, with a note that the effectiveness and safety of the 5-mg dose have not been studied prospectively to date. The FDA labeling information includes a general warning against the use of prasugrel in patients older than 75 years of age because of concerns of an increased risk of fatal and intracranial bleeding and uncertain benefit, except in high-risk situations (patients with diabetes or a history of prior MI), in which case its effect appears to be greater and its use may be considered (37).

In focusing specifically on patients with STEMI, the primary composite end point of cardiovascular death, nonfatal MI, or nonfatal stroke was significantly reduced in patients assigned to prasugrel at 30 days compared with patients who received clopidogrel (6.5% versus 9.5%; HR 0.68; 95% CI 0.54 to 0.87; P=0.0017), and this trend persisted to 15 months (HR 0.79; 95% CI 0.65 to 0.97; P=0.0221) (13). Furthermore, in the STEMI group, the key secondary end point of cardiovascular death, MI, or urgent TVR was significantly reduced with prasugrel at 30 days (P=0.0205) and 15 months (P=0.0250) (13). At 30 days and 15 months, the individual end points of cardiovascular death and MI, as well as stent thrombosis, were reduced with prasugrel (13).

The interaction testing for efficacy and safety showed no significant difference in bleeding risk regardless of the type of ACS (e.g., UA/NSTEMI versus STEMI). Thus, the STEMI results for efficacy and safety are consistent with the main results of the trial. In a post hoc analysis of patients with anterior MI, event rates at 15 months for the primary end point were lower with prasugrel (9.8% for prasugrel versus 16.3% for clopidogrel; HR 0.57; 95% CI 0.42 to 0.78; P=0.0003). In patients with nonanterior MI, treatment effects did not differ for the primary end point (10.1% for prasugrel versus 9.9% for clopidogrel; HR 1.02; 95% CI 0.78 to 1.34; P=0.8749). The test for heterogeneity of the effect of prasugrel was significant (P=0.0053), which suggests that the benefit might vary by the location of the MI. Data were consistent in both the primary and secondary PCI subgroups (13).

The writing group weighed the current data regarding the use of thienopyridine therapy in patients who remain hospitalized after STEMI and are candidates for CABG and retained the 2007 focused update recommendation of empiric discontinuation of clopidogrel therapy for at least 5 days and at least 7 days in patients receiving prasugrel before planned CABG (2,27,30).

Platelet function testing to determine the degree of platelet inhibition (39) may be used, and if platelet function has normalized, CABG may be performed at an earlier time. Additionally, other strategies of platelet inhibition (GP IIb/IIIa receptor antagonists) may be used if recurrent ischemia is a concern during the waiting period for CABG. Ultimately, the patient's clinical status will determine the risk-to-benefit ratio of CABG compared with awaiting restoration of platelet function.

The results of TRITON-TIMI 38 influenced dosing recommendations for loading and chronic thienopyridine therapy with prasugrel. Sixty milligrams of prasugrel is now recommended as a loading dose for primary PCI in STEMI. For secondary PCI in those patients who have recurrent ischemia or other reasons for planned intervention during their course of treatment, 60 mg of prasugrel may be given after the coronary anatomy has been identified (to avoid dosing those patients who require CABG) either before, during, or within 1 hour of PCI (27). Furthermore, 10 mg of prasugrel may be used in addition to ASA for chronic dual-antiplatelet therapy (27).

Determination of patient groups that should be considered for continuation of dual-antiplatelet treatment beyond 12 months is based on patient-level factors (e.g., age, history of bleeding) and lesion characteristics (e.g., bifurcation, small-diameter vessel) (28).

In previous studies of patients with prior stroke or TIA, use of dual-antiplatelet therapy has been associated with an increased risk of adverse outcomes, notably intracranial bleeding, compared with single-antiplatelet therapy. In the MATCH (Management of Atherothrombosis With Clopidogrel in High-Risk Patients With TIA or Stroke) trial (40) in which patients with prior stroke or TIA and additional risk factors (n=7599) were allocated to clopidogrel 75 mg or combination therapy with clopidogrel 75 mg plus ASA 75 mg per day, there was no significant benefit of combination therapy compared with clopidogrel alone in reducing the primary outcome of the composite of ischemic stroke, MI, vascular death, or rehospitalization due to ischemic events, or any of the secondary outcomes. The risk of major hemorrhage was significantly increased in the combination-therapy group compared with those given clopidogrel alone, with a 1.3% absolute increase in life-threatening bleeding. Although clopidogrel plus ASA is recommended over ASA alone for patients with ACS (41–43), the results of MATCH do not suggest a similar risk-benefit ratio for stroke and TIA survivors. The AHA/American Stroke Association's Guidelines for Prevention of Stroke in Patients With Ischemic Stroke or Transient Ischemic Stroke contain a Class III recommendation for the use of ASA in combination with clopidogrel in patients with prior stroke or TIA (44). On the other hand, a post hoc analysis from the CHARISMA (Clopidogrel for High Atherothrombotic Risk and Ischemic Stabilization, Management, and Avoidance) trial, which included 9478 patients, suggested that patients with documented prior MI, ischemic stroke, or symptomatic peripheral artery disease derive benefit from dual-antiplatelet therapy with clopidogrel plus ASA (45). Although MATCH and CHARISMA did not involve STEMI patients, the writing group recommended weighing the benefits and risks of prescribing clopidogrel and ASA in patients with a recent history of TIA or stroke. Given prasugrel's greater tendency to cause intensive inhibition of platelet aggregation in general and the findings of increased levels of bleeding compared with clopidogrel in this population, the use of prasugrel as part of a dual-antiplatelet therapy regimen in patients with prior stroke or TIA is contraindicated (37).

3.2 Proton Pump Inhibitors and Dual-Antiplatelet Therapy for ACS

Proton pump inhibitors (PPIs) are often prescribed prophylactically when clopidogrel is started, to prevent gastrointestinal complications such as ulceration and related bleeding (59) due to dual-antiplatelet therapy, in particular ASA and clopidogrel (32). Coupled with concern about the gastrointestinal precautions, there has been increased emphasis on the prevention of premature discontinuation of dual-antiplatelet therapy, particularly in patients who have received a stent (BMS or DES), for whom 12 months of antiplatelet therapy is recommended (28). PPI medications^⁎) have been found to interfere with the metabolism of clopidogrel (34).

Although there are studies that show a pharmacodynamic interaction on ex vivo platelet function testing, to date there are no convincing randomized clinical trial data for an important clinical drug–drug interaction. Retrospective claims-based reports suggesting clinical harm, some detailed below, may be confounded by different baseline characteristics and lack of compliance data. There have been retrospective reports of adverse cardiovascular outcomes (e.g., readmission for ACS) when the antiplatelet regimen of clopidogrel and ASA is accompanied by PPIs, assessed as a group, compared with the use of this regimen without a PPI (32–34,60). In a retrospective cohort study from the Veterans Affairs' medical records and pharmacy database, concomitant clopidogrel and PPI therapy (with omeprazole, rabeprazole, lansoprazole, or pantoprazole) at any time point during follow-up of 8205 patients discharged for ACS was associated with an increased risk of death or rehospitalization for ACS (32). Other post hoc study analyses (50,61) and a retrospective data analysis from the NHLBI Dynamic Registry (62) in which PPIs were assessed as a class in combination with a clopidogrel and an ASA regimen have not found an effect of PPI therapy on the clinical effect of clopidogrel in ACS patients, after ACS, or in a general post-PCI population, respectively (50,61,62).

Some studies have suggested that adverse cardiovascular outcomes with the combination of clopidogrel and a PPI are explained by the individual PPI, in particular the use of a PPI that inhibits CYP450 2C19, which includes omeprazole, lansoprazole, and rabeprazole. The PPI omeprazole notably has been reported to significantly decrease the inhibitory effect of clopidogrel on platelet aggregation (63,64). One study reported that the PPI pantoprazole was not associated with recurrent MI among patients receiving clopidogrel, possibly because of its lack of inhibition of CYP450 2C19 (33).

Other studies have examined the thienopyridine agent prescribed with the PPI. One open-label drug study evaluated the effects of the PPI lansoprazole on the pharmacokinetics and pharmacodynamics of prasugrel and clopidogrel in healthy subjects given single doses of prasugrel (60 mg) and clopidogrel (300 mg) with and without concurrent lansoprazole 30 mg per day. The data suggest that inhibition of platelet aggregation was reduced in patients who took the combination of clopidogrel and lansoprazole, whereas it was unaffected after a prasugrel dose (56).

Another study (35) assessed the association of PPIs with the pharmacodynamics and clinical efficacy of clopidogrel and prasugrel, based on populations from 2 randomized trials, the PRINCIPLE (Prasugrel In Comparison to Clopidogrel for Inhibition of Platelet Activation and Aggregation) TIMI-44 trial (65) and the TRITON-TIMI 38 trial (27). The findings indicated that first, PPI treatment attenuated the pharmacodynamic effects of clopidpgrel and, to a lesser extent, those of prasugrel. Secondly, PPI treatment did not affect the clinical outcome of patients given clopidogrel or prasugrel. This finding was true for all PPIs that were studied, including omeprazole and pantoprazole.

The FDA communication concerning an ongoing safety review of clopidogrel bisulfate (66) advises that healthcare providers avoid the use of clopidogrel in patients with impaired CYP2C19 function due to known genetic variation or due to drugs that inhibit CYP2C19 activity. The FDA notes there is no evidence that other drugs that reduce stomach acid, such as H2 blockers or antacids, interfere with the antiplatelet activity of clopidogrel.

Further research with thienopyridines and PPI combinations, particularly drugs that are not dependent on CYP450 2C19, is needed. Consideration may be given to the use of H2 antagonists as an alternative to PPIs in the setting of dual-antiplatelet therapy, although they cannot be relied on to protect as well as PPIs, and there are few data about their use with ASA (59). The FAMOUS (Famotidine for the Prevention of Peptic Ulcers in Users of Low-Dose Aspirin) trial, a phase II, double-blind, randomized, placebo-controlled trial, found that among patients with a history of coronary heart disease, diabetes mellitus, or cerebrovascular disease who were taking low-dose ASA, 12 weeks of famotidine 20 mg twice daily (n=204) compared with placebo twice daily (n=200) was beneficial in reducing the incidence of peptic ulcer or esophagitis during follow-up endoscopy at 12 weeks. The rate of occurrence of a gastric ulcer at endoscopy at 12 weeks was 3.4% in the famotidine group versus 15% in the placebo group (P=0.0002), duodenal ulcer occurred in 0.5% versus 8.5% (P=0.0045), and erosive esophagitis was seen in 4.4% versus 19% (P<0.0001), respectively. Of note, in the famotidine group, clopidogrel use was 19% and dipyridamole use was 6% (67). The writing committee concluded that additional data, notably randomized controlled clinical trial data that have been peer reviewed and published, are needed before an official recommendation can be made about the use of dual antiplatelet therapy with PPIs in the setting of ACS.

4.1 Parenteral Anticoagulants

Parenteral anticoagulants include intravenous UFH, bivalirudin, enoxaparin, and fondaparinux. Bivalirudin was briefly cited in the 2007 STEMI focused update. The HORIZONS-AMI trial, which was reported subsequently, was a prospective, open-label, randomized, multicenter, international trial that included 3602 patients with STEMI undergoing primary PCI. Patients were randomized to treatment with UFH plus a GP IIb/IIIa receptor antagonist or to bivalirudin alone (with provisional abciximab or double-bolus eptifibatide). The primary efficacy end point was a composite of net adverse clinical events, including major bleeding plus MACE, a composite of cardiovascular death, reinfarction, TVR for ischemia, and stroke within 30 days. Bivalirudin alone resulted in a lower incidence of net adverse clinical events at 30 days (9.2% versus 12.1%; RR 0.76; 95% CI 0.63 to 0.92; P=0.005; NNT=34) and at 1 year (15.7% versus 18.3%, HR 0.84; 95% CI 0.71 to 0.98; P=0.3) (9). The difference was driven by a significant decrease in major bleeding complications with bivalirudin at 30 days (4.9% versus 8.3%, P=0.001; number needed to harm=33) and 1 year (5.8% versus 9.2%, P=0.001). There was a statistically significant 1% increase in stent thrombosis (n=17) within the first 24 hours with bivalirudin but no subsequent difference (1.3% versus 0.3%, P<0.001). More deaths at 30 days occurred after major bleeding (n=26) than after reinfarction (n=10) or definite stent thrombosis (n=5) (9). Treatment with bivalirudin resulted in significantly lower 30-day rates of death due to cardiac causes (1.8% versus 2.9%; RR 0.62; 95% CI 0.40 to 0.95; P=0.03) and death due to all causes (2.1% versus 3.1%; RR 0.66; 95% CI 0.44 to 1.00; P=0.047 compared with UFH plus GP IIb/IIIa inhibitors). At 1 year, MACE rates were identical, but there was a decrease in all-cause mortality with bivalirudin (3.4% versus 4.8%, P=0.03) (68).

Concerns about the trial include its open-label design and the administration of UFH before randomization in 66% of patients in the bivalirudin arm and 76% of patients in the UFH plus GP IIb/IIIa receptor antagonist arm. Only 615 patients received bivalirudin monotherapy, and only 60% of patients in the trial received a 600-mg clopidogrel loading dose. Major bleeding as defined in the publication included hematomas of 5 cm, intracranial hemorrhage, and bleeding that required surgery. Additionally, the study put forth a composite primary end point that combined efficacy and safety. Although there were no statistically significant interactions at 30 days between the treatment assignment and preprocedural UFH use or clopidogrel loading dose with respect to MACE or major bleeding, the occurrence of an increase in early stent thrombosis with bivalirudin and the excess bleeding with UFH and GP IIb/IIIa inhibitors may be related to the degree of platelet inhibition and antithrombin activity associated with these treatment doses.

A preliminary report suggested that the use of bivalirudin alone (P=0.005) and a lower loading dose of clopidogrel (300 versus 600 mg; P=0.01) were independent predictors of acute and subacute stent thrombosis rates, respectively (69). Probability values for secondary end points may not have been adjusted for multiple looks.

Therefore, the writing group now considers bivalirudin useful for primary PCI in STEMI whether or not the patient received pretreatment with UFH. The risk of acute stent thrombosis associated with bivalirudin appeared to be mitigated by the prior use of UFH and the risk of subacute stent thrombosis by the use of a 600-mg loading dose of clopidogrel. These data should be confirmed by prospective studies.

5.1 Triage and Transfer for PCI

6.1 Intensive Glucose Control

As detailed in the 2004 STEMI guideline and the 2007 UA/NSTEMI guideline revision, randomized trial evidence supported the use of insulin infusion to control hyperglycemia (3,97). A recently published randomized clinical trial of intensive versus conventional glucose control in critically ill patients raised uncertainty regarding the optimal level to target when achieving glucose control. NICE-SUGAR, a large, international randomized trial (n=6104) of adults admitted to the intensive care unit with either medical or surgical conditions, compared intensive glucose control (target glucose range 81 to 108 mg/dL) with conventional glucose control (to achieve a glucose level less than 180 mg/dL, with reduction and discontinuation of insulin if the blood glucose level dropped below 144 mg/dL) (16). Time-weighted glucose levels achieved were 115±18 mg/dL in the intensive glucose control group versus 144±23 mg/dL in the conventional glucose control group. The risk of death was increased at 90 days in the intensive glucose control group by 2.6% (27.5% versus 24.9%; odds ratio 1.14; 95% CI 1.02 to 1.08; P=0.02; number needed to harm=38). The result remained the same after adjustment for potential confounders. There were significantly more episodes of treatment-related hypoglycemia in the intensely managed group (6.8% versus 0.5%, P=0.001), although the contribution of hypoglycemia to excess mortality is uncertain (94,98). Overall, the hospital course and proximate causes of death were similar in the 2 groups. Excess deaths in the intensive-control group were predominantly due to cardiovascular causes (absolute difference 5.8%; P=0.02). More patients in the intensive-control group than in the conventional-control group were treated with corticosteroids.

Because NICE-SUGAR enrolled critically ill medical and surgical patients, the degree to which its results can be extrapolated to the management of patients with STEMI is unclear. Although recent data from a small, mechanistic clinical trial (28,29,98) suggest that glucose control may reduce inflammation and improve left ventricular ejection fraction in patients with acute myocardial infarction (AMI), whether it will improve patient outcomes remains uncertain.

A consensus statement by the American Association of Clinical Endocrinologists and the American Diabetes Association (99) summarized that “although hyperglycemia is associated with adverse outcomes after AMI, reduction of glycemia per se, and not necessarily the use of insulin, is associated with improved outcomes. It remains unclear, however, whether hyperglycemia is a marker of underlying health status or is a mediator of complications after AMI. Noniatrogenic hypoglycemia has also been associated with adverse outcomes and is a predictor of higher mortality” (p 1120) (99).

There is a clear need for a well-designed, definitive randomized trial of target-driven glucose control in STEMI that has meaningful clinical end points to determine glucose treatment thresholds and glucose targets. Until such a trial is completed, and based on the balance of current evidence (99–101), the writing group concluded that it was prudent to change the recommendation for the use of insulin to control blood glucose in STEMI from a Class I to a Class IIa recommendation (Level of Evidence: B) and to recommend treatment for hyperglycemia greater than 180 mg/dL while avoiding hypoglycemia.

7.1 Thrombus Aspiration

Since the publication of the last STEMI and PCI focused updates, 2 new trials of manual thrombus aspiration have been published. TAPAS was a single-center, unblinded, randomized clinical trial that compared 2 catheter-based reperfusion strategies in 1071 patients with STEMI (17,102). Before coronary angiography, patients were randomized to manual thrombus aspiration before PCI (55.1% by direct stenting, 28.6% by balloon angioplasty followed by stenting, 10.1% by PCI without thrombus aspiration) or conventional PCI with balloon angioplasty followed by stenting. BMS were implanted in 92% of PCI procedures. All patients were treated with ASA, 600 mg of clopidogrel, UFH, and abciximab unless contraindicated. TIMI myocardial blush grade 0 or 1 occurred in 17.1% of patients with thrombus aspiration and 26.3% of those with conventional PCI (P<0.001). Complete resolution of ST-segment elevation occurred in 56.6% and 44.2%, respectively (P<0.001). Death, reinfarction, and TVR rates at 30 days were not significantly different (6.8% versus 9.4%) (17). However, at 1 year, rates of cardiac death (3.6% versus 6.7%, P=0.02) and cardiac death or nonfatal reinfarction (5.6% versus 9.9%, P=0.009) were lower with thrombus aspiration. Low myocardial blush grade and incomplete ST-segment resolution were associated with clinical events (102).

EXPIRA was a smaller (n=175) randomized clinical trial that also compared thrombus aspiration with conventional PCI, but only in patients with TIMI flow 0/1 (18). TIMI myocardial blush grade of 2 or more (88% versus 60%, P<0.001) and 90-minute ST-segment resolution greater than 70% (64% versus 39%, P<0.001) occurred more frequently in the thrombus aspiration group. Infarct size measured by contrast-enhanced magnetic resonance imaging in 75 patients at 3 months was significantly reduced only in the thrombus aspiration group.

Both of these trials, as well as a meta-analysis by Bavry et al (103) and a large pooled analysis of randomized trials (104), support the use of aspiration thrombectomy for STEMI. In developing a recommendation for the role of routine aspiration thrombectomy, however, it is noteworthy that TAPAS was a study of routine thrombus aspiration versus no thrombus aspiration rather than a study of routine thrombus aspiration versus selective thrombus aspiration. It is not known whether a strategy of selective thrombus aspiration in patients with a large thrombus burden might be superior to no thrombus aspiration or equivalent to routine thrombus aspiration. Clinically, it is reasonable to assume that this strategy can be useful in STEMI patients with short ischemic times and large thrombus burden. It may not be helpful in STEMI patients with long ischemic times, side branches with small infarct territories, or lesions with low thrombus burden.

8.1 Stent Selection for STEMI

Primary PCI is generally the preferred reperfusion strategy for patients with STEMI (110). Compared with balloon angioplasty, routine BMS implantation during primary PCI decreases risk for TVR and possibly reduces MI rates but does not reduce mortality rates (111,112).

Two-year data from the Massachusetts registry (113) from 1221 propensity score–matched pairs of DES and BMS patients demonstrated a reduction in mortality and TVR rates with DES in primary PCI, and an analysis from the New York State registry (114) found a reduction in mortality rates but not TVR rates with DES. These reports were limited to patients treated before 2005, so they represent the earliest experience with DES, in which selection bias may have influenced stent choice and off-label use may have been pursued more cautiously. Additionally, duration of clopidogrel therapy was longer in the DES group.

More recently, several relatively small randomized clinical trials have shown an inconsistent efficacy for DES over BMS in primary PCI. Three meta-analyses of these trials have concluded that there were no differences in death, MI, or stent thrombosis rates, but TVR rates were decreased with DES (115–117). Variably included were 12 studies that differed in trial design, inclusion criteria, end-point definitions, stent types, duration of clopidogrel treatment, and type of follow-up (angiographic versus clinical). They were limited by sample size and duration of follow-up and by usually requiring angiographic documentation of stent thrombosis, which may have underestimated its true incidence.

The HORIZONS-AMI trial randomized, in a 3-to-1 ratio, 3006 patients to DES or BMS (9,68,105). There was no difference in the 12-month composite safety end point of death, reinfarction, stroke, or stent thrombosis. The rates of ischemia-driven TVR and target-lesion revascularization were significantly lower in the DES group (5.8% versus 8.7% and 4.5% versus 7.5%, respectively; NNT=33 at 1 year), as was the 13-month binary restenosis rate (10.0% versus 22.9%).

In summary, there appears to be no difference between BMS and DES in mortality or MI rates and no difference in stent thrombosis risk. The major advantage of DES over BMS is a small reduction in TVR rates. Given cost considerations, it could be argued that selective use of DES to prevent restenosis and TVR in high-risk patients (i.e., patients with diabetes) and in high-risk lesions (longer and smaller-diameter stents) could be recommended (118), as it has been for elective PCI. The greatest challenge in selecting patients for DES implantation, however, is determining in an emergency situation whether the patient is a candidate for prolonged thienopyridine therapy. As with elective procedures, DES should be avoided in the presence of financial barriers to continuing prolonged dual-antiplatelet therapy, social barriers that may limit patient compliance, or medical issues that involve bleeding risks or the need for invasive or surgical procedures in the following year that would interrupt antiplatelet therapy.

Angiography in Patients With Chronic Kidney Disease

Patients with chronic kidney disease (CKD) with or without diabetes who undergo angiography are at high risk for a contrast-induced nephropathy (CIN). At issue is the selection of a contrast agent to minimize this risk. The 2007 UA/NSTEMI guideline recommended that in patients with chronic kidney disease undergoing angiography, “isosmolar contrast agents are indicated and are preferred (Level of Evidence: A)” (p e112) (97). In patients with CKD or CKD and diabetes mellitus who are undergoing angiography, isosmolar contrast material was shown to lessen the rise in creatinine. This was based on evidence up to mid-2007 that suggested that isosmolar agents also reduced the risk of CIN in both a moderate-sized randomized clinical trial (RECOVER [Renal Toxicity Evaluation and Comparison Between Visipaque (Iodixanol) and Hexabrix (Ioxaglate) in Patients With Renal Insufficiency Undergoing Coronary Angiography]) that compared iodixanol with ioxaglate (119) and a meta-analysis of 16 smaller, earlier clinical trials (120).

However, in mid-2007, a major US randomized trial of contrast agents in patients with CAD and an estimated glomerular filtration rate of 20 to 59 mL/min who were undergoing angiography, the CARE study, was published. CARE compared the low-osmolar agent iopamidol and the isosmolar agent iodixanol and found no difference in the primary end point (serum creatinine increase of 0.5 mg/dL or higher over baseline) between iopamidol (4.4%) and iodixanol (6.7%, P=0.39) (19).

Since then, several larger randomized trials have been published that reported no difference in CIN when iodixanol was compared with various other low-osmolar contrast media (LOCM) (19,121–123). These and other randomized trials comparing isosmolar iodixanol with LOCM have been summarized in 2 mutually supportive and complementary meta-analyses involving 16 trials in 2763 patients (124) and 25 trials in 3260 patients (125), respectively. When more recent trials were combined with the older studies, trends in CIN favoring iodixanol were no longer significant (summary RR 0.79; 95% CI 0.56 to 1.12; P=0.29; summary RR 0.80; CI 0.61 to 1.04; P=0.10, respectively) (124,125). However, subanalyses showed variations in relative renal safety by specific LOCM: A reduction in CIN was observed when iodixanol was compared with ioxaglate, the only ionic LOCM (RR 0.58, CI 0.37 to 0.92, P=0.02 (124)), and with iohexol, a nonionic LOCM (RR 0.19 to 0.38, P<0.01) (124,125), but no difference was noted in comparisons of iodixanol with iopamidol, iopromide, or ioversal (124), and a single trial favored iomeprol (123). A pooled comparison of iodixanol with all nonionic LOCM other than iohexol indicated equivalent safety (RR 0.97; CI 0.72 to 1.32, P=0.86) (125). Results were consistent regardless of ancillary preventive therapies (hydration, acetylcysteine), route of administration (intravenous or intra-arterial), age, sex, dose, or preexisting CKD or diabetes. Of further interest, findings were similar in the 8 studies (n=1793 patients) performed in the setting of coronary angiography (124).

These more recent observations indicate that the CIN risk of contrast media cannot be attributed to osmolarity alone, but that ionicity and other and unknown characteristics of specific agents may play a role. Thus, the updated evidence base suggests that the recommended choices of contrast media during coronary angiography be expanded to either isosmolar media or LOCM other than ioxaglate or iohexol.

10.1 Fractional Flow Reserve

Coronary angiography is often performed in clinical situations in which preprocedural functional testing has not been obtained. Additionally, in the setting of multivessel disease, the need to treat individual stenosis is often difficult to determine. Although revascularization of ischemia-producing lesions improves patient outcomes, the clinical benefits of revascularization of stenotic but non–ischemia-producing lesions are less clear. Intraprocedural assessment of the functional significance of individual stenosis may help define the optimal revascularization strategy.

The objective of the FAME trial (20) was to compare clinical outcomes after PCI on the basis of conventional angiographic determination of lesion severity versus fractional flow reserve (FFR) combined with angiography in patients with multivessel disease. This prospective, randomized, multicenter trial included 1005 patients selected from 1905 screened patients at 20 medical centers who were randomized to either angiography-guided or FFR-guided (for lesions with FFR less than or equal to 0.80) PCI. Before randomization, lesions that required PCI were prespecified on the basis of the angiographic appearance. Patients assigned to angiography-guided PCI had all identified lesions treated with DES, whereas those assigned to FFR-guided PCI had only identified lesions with an FFR of 0.80 or less treated with DES. The primary end point of the trial was the rate of death, nonfatal MI, and repeat revascularization at 1 year.

No difference was evident in the number of intended lesions to be treated per patient (2.7±0.9 versus 2.8±1.0, P=0.34) in the angiography- and FFR-guided groups, respectively. In the FFR group, 37% of lesions had an FFR greater than 0.80. Evaluation of ischemia, as defined by an FFR less than 0.80, resulted in fewer lesions receiving stents (2.7±1.2 versus 1.9±1.3, P<0.001). At 1 year, the composite event rate was 18.3% in the angiography-guided group compared with 13.2% in the FFR-guided group (P=0.02).

The results of the FAME trial suggest that identification of ischemia-producing lesions by use of systematic assessment of FFR in patients undergoing multivessel PCI is associated with improved clinical outcomes compared with angiographic assessment alone. Further evidence is needed regarding the added value of assessing FFR in lesions with greater than 90% stenosis.

11.1 Unprotected Left Main Coronary Artery Disease

Although listed as a Class III indication in the 2003 guideline on the management of chronic stable angina (145), PCI of an unprotected left main coronary artery has increased in frequency (146). Early studies (listed in Appendix 6) involved short follow-up periods, which gave CABG a disadvantage, because the apparent benefits of surgery over PCI in other settings have not typically been fully evident until 1 to 5 years after the procedure. In the ACC/AHA/SCAI 2005 guideline update for PCI, the performance of PCI for left main CAD is given a Class III, Level of Evidence C recommendation if the patient is eligible for CABG and a Class IIa, Level of Evidence B recommendation for patients who are not eligible for CABG. Thus, it has been recommended that CABG still be considered the standard of care for left main CAD (147–149).

Several studies comparing CABG to PCI, however, indicated that the advantage of CABG consists primarily of fewer repeat revascularizations (139,149–155). The study by Brener et al (156) indicates no significant mortality difference between PCI and CABG after 3 years of follow-up. Longer-term follow-up is needed.

The present focused update specifically addresses the findings of SYNTAX, an unblinded, randomized clinical trial that assigned patients with 3-vessel and/or left main CAD to an initial treatment strategy of CABG or PCI (21). The primary prespecified end point for the 1800 enrolled patients was the composite of death, stroke, and myocardial revascularization determined at 12 months. Prespecified stratification occurred for diabetes mellitus and left main CAD. Ninety-seven percent of CABG patients received at least 1 arterial graft.

In SYNTAX, for the subgroup with left main CAD, there were no significant differences in the incidence of the composite end point (death, MI, stroke, or repeat revascularization) between the 2 groups (PCI 15.8% versus CABG 13.7%, P=0.44), although rates of repeat revascularization were higher (11.8% versus 6.5%, P=0.02) and rates of stroke were lower (0.3% versus 2.7%, P=0.01) in the PCI group (21). Left main stented patients with limited CAD (lower SYNTAX score) displayed a trend toward fewer adverse events at 12 months than did similar patients assigned to CABG. Specifically, MACE in patients with isolated left main CAD occurred in 8.5% with CABG versus 7.1% with PCI, and in 13.2% with CABG versus 7.5% with PCI in patients with left main CAD and disease of 1 other vessel. In contrast, MACE with CABG versus PCI were numerically less frequent in patients with disease of the left main coronary artery and disease of 2 other vessels (14.4% versus 19.8%) and in patients with both left main CAD and 3 other vessels involved (15.4% versus 19.3%) (21).

These data from a post hoc subgroup analysis in SYNTAX must be interpreted with caution for several reasons. The number of patients with isolated left main (or left main plus single vessel) CAD was relatively small, and the differences in outcomes were not statistically significant. Furthermore, SYNTAX reported outcomes at 1 year, and longer follow-up is needed before left main PCI (in patients who are otherwise surgical candidates) should become standard clinical practice. Moreover, because the overall study did not reach its primary end point, subset analyses are less robust; because noninferiority was not proven in this cohort, specific information for each subgroup is of an observational nature and is hypothesis-generating.

On the basis of the evidence in aggregate, prior to and within the present focused update, the writing group has modified the class of recommendation for PCI to unprotected left main coronary from Class III to Class IIb, now citing the Level of Evidence as B. The writing group noted 3 important caveats in classifying unprotected left main CAD as a Class IIb indication. First, patients undergoing PCI in cohort or randomized studies represent merely a subset of all patients with left main CAD. Because only certain left main coronary lesions are amenable to PCI, the Class IIb indication is intended to apply only to those left main lesions that are suitable for PCI. The primary conclusion of SYNTAX is that PCI failed to be shown to be noninferior to CABG in left main and triple-vessel disease. Because patients in SYNTAX with left main and 2- or 3-vessel disease compared with patients with left main and no other vessel or 1-vessel disease had higher rates of MACE, it is recommended that PCI to left main lesions be limited to patients without significant multivessel disease. Second, because of the narrow margin for error, operators undertaking PCI of left main coronary lesions should be experienced and backed by highly competent support staff and surgeons (157). Although routine use of intravascular ultrasound has been advocated by some authors for the evaluation of left main lesions, there is no definitive evidence at present that this technique improves outcomes (138,157). Finally, not all left main lesions respond equally well to PCI. Bifurcation lesions are technically more challenging (138) and have higher rates of restenosis (140,141,143,144). In contrast, results of PCI of ostial or mid-body left main coronary lesions more closely approximate the results of CABG, even with respect to the need for subsequent procedures (142). The best case for PCI as an alternative to CABG for left main CAD is in ostial and mid-body lesions without additional multivessel disease.

The writing group discussed the previous Class IIa recommendation for follow-up between 2 and 6 months with coronary angiography. They focused on the inability of angiography to predict a situation that might be prone to acute, sudden stent thrombosis, as well as the risk associated with angiography in a patient who has undergone placement of a left main stent. In view of these factors, the writing group decided that the Class IIa recommendation for angiographic follow-up should be omitted from the guidelines.

12.1 Timing of Angiography

A routine invasive strategy in UA/NSTEMI patients with high-risk features has been associated with improved outcomes, but the optimal timing of intervention has not been well established. Early intervention might prevent ischemic events that could occur while the patient awaits a delayed procedure. Alternatively, with intensive antithrombotic therapy with a delay for up to a few days, procedure-related complications might be avoided by intervening on a more stable, “passivated” plaque. Although one study has suggested greater benefit with relatively early intervention (106), the evidence base for a definitive recommendation on timing is weak. Thus, the question of when to intervene in UA/NSTEMI has not been answered conclusively. Given this uncertainty, the TIMACS investigators (23) undertook a large, multicenter randomized trial to determine whether a strategy of early coronary angiography and intervention was superior to a delayed strategy in patients with UA/NSTEMI assigned to an invasive approach.

TIMACS randomly assigned 3031 non–ST-elevation ACS patients to routine early intervention (coronary angiography within 24 hours) or to delayed intervention (coronary angiography at 36 hours or more). The primary outcome was the composite of death, MI, or stroke at 6 months, and a prespecified secondary outcome was death, MI, or refractory ischemia (23).

Coronary angiography was performed at a median of 14 hours in the early-intervention group and 50 hours in the delayed-intervention group. At 6 months, 9.7% of patients in the early-intervention group experienced a primary outcome versus 11.4% in the delayed-intervention group (HR 0.85; 95% CI 0.68 to 1.06; P=0.15). Death, MI, or refractory ischemia was reduced by 28% in favor of early intervention (9.6% versus 13.1%; HR 0.72; 95% CI 0.58 to 0.89; P=0.002). Prespecified analyses showed that early intervention improved the primary outcome in the one third of patients at highest risk (HR 0.65; 95% CI 0.48 to 0.88), as determined by a GRACE (Global Registry of Acute Coronary Events) risk score greater than 140, but not in the two thirds at low to intermediate risk (HR 1.14; 95% CI 0.82 to 1.58; Pfor heterogeneity=0.01). There were no safety issues related to early intervention; major bleeding occurred in 3.1% of patients in the early invasive group and 3.5% in the delayed invasive group.

Overall in TIMACS, early intervention trended to be superior to delayed intervention in preventing the composite of death, MI, or stroke (primary end point), but the difference was not statistically significant (23). However, early intervention reduced the composite of death, MI, or refractory ischemia (the secondary end point) and, in high-risk patients, was superior to a delayed invasive strategy. The trial was underpowered to discern a clinically meaningful 15% advantage to early invasive therapy for the primary end point, with recruitment stopped at 3000 because of recruitment and funding challenges. This provided a power of 80% to detect a risk reduction of 28% in the primary end point. Subgroup analysis (high-risk subset) of this overall negative trial was not robust and must be viewed cautiously.

Taken together with the earlier ISAR-COOL (Intracoronary Stenting With Antithrombotic Regimen Cooling Off) study (106), the favorable secondary end-point results, the subgroup analysis in patients at higher risk, and the lack of a safety issue with early therapy, TIMACS suggests the following conclusions: an early invasive strategy within 12 to 24 hours (median 14 hours) is preferred in high-risk patients and may be chosen in patients at low to intermediate risk at the physician's or institution's preference (e.g., efficiency and cost savings), whereas a more delayed approach may be beneficial in low- to intermediate-risk patients (23). In contrast, results from the recent ABOARD (Angioplasty to Blunt the rise Of troponin in Acute coronary syndromes Randomized for an immediate or Delayed intervention) trial (160) indicate that an immediate invasive strategy (median time 1.1 hour) in UA/NSTEMI is not associated with further incremental benefit.

Typically, early versus delayed angiography is defined with reference to a 12- to 48-hour time window. The ISAR-COOL study (106) supports an earlier compared with a more delayed time to angiography, but the data supporting this general timing suggestion are limited.

12.2 Timing of GP IIb/IIIa Receptor Antagonist Therapy in UA/NSTEMI Patients Undergoing Angiography

The optimal timing of initiation of GP IIb/IIIa receptor antagonist therapy in patients with UA/NSTEMI (i.e., whether to administer therapy upstream on presentation or later at the time of angiography/PCI) and the optimal application of this therapy (i.e., whether routine, selective, or provisional) have not been resolved. The 2007 ACC/AHA Guidelines for the Management of Patients With UA/NSTEMI recommend that patients with definite or likely UA/NSTEMI selected for an invasive approach should receive ASA and either clopidogrel or a GP IIb/IIIa receptor antagonist before angiography (Class I, Level of Evidence: A) (97). They further state that it is reasonable to initiate both clopidogrel and a GP IIb/IIIa receptor antagonist, especially in the setting of delays to angiography, high-risk features, or recurrent ischemic discomfort (Class IIa, Level of Evidence: B). The 2007 European Society of Cardiology guidelines recommend early dual-antiplatelet therapy with ASA and clopidogrel (Class I), with the addition of a GP IIb/IIIa receptor antagonist for those patients with the specific high-risk features of an elevated troponin level, ST-segment depression, or diabetes (Class IIa) (163).

The EARLY ACS (Early Glycoprotein IIb/IIIa Inhibition in Patients With Non–ST-Segment Elevation Acute Coronary Syndrome) trial (22) tested the hypothesis that a strategy of early, routine administration of the GP IIb/IIIa receptor antagonist eptifibatide would be superior to delayed, provisional administration in reducing ischemic complications among high-risk patients. EARLY ACS enrolled 9492 non–ST-segment elevation ACS (UA/NSTEMI) patients who presented within 24 hours of an episode of ischemic rest discomfort and who were assigned to an invasive treatment strategy no sooner than the next calendar day (amended later to at least 12 hours but less than 96 hours after randomization). Eligibility required at least 2 of the following features: ST-segment depression or transient ST-segment elevation, an elevated biomarker (creatine kinase-MB or troponin), and age of 60 years or older (amended to include patients 50 to 59 years with known vascular disease). The key primary end point was all-cause death, MI, recurrent ischemia that required urgent revascularization, or thrombotic bailout at 96 hours. The key secondary efficacy end point was all-cause death or MI within 30 days. Safety end points included major hemorrhage and transfusions through 120 hours after randomization. The study was powered to detect 22.5% and 15% relative reductions in the primary and key secondary end points, respectively.

The primary end point occurred in 9.3% of patients in the early therapy arm versus 10.0% of patients in the provisional GP IIb/IIIa therapy arm (odds ratio 0.92; 95% CI 0.80 to 1.06; P=0.23). Secondary end-point event rates were 11.2% versus 12.3% (odds ratio 0.89; CI 0.79 to 1.01; P=0.08). Early, routine eptifibatide administration occurred at the cost of a greater risk of TIMI major hemorrhage (2.6% versus 1.8%, P=0.02). Moderate and less severe bleeding also occurred more commonly. Rates of red cell transfusion were 8.6% and 6.7%, respectively (P=0.001).

EARLY ACS represents a large, carefully executed trial with potentially important implications; however, these results are best taken in the context of previous major trials. As a single trial, EARLY ACS does not establish the superiority of early versus delayed, provisional eptifibatide in non–ST-elevation ACS. A trend toward fewer recurrent ischemic complications was noted at 30 days, but this was counterbalanced by more frequent episodes of bleeding and need for transfusions. Given the use of eptifibatide at PCI (39% of patients in the delayed, provisional group), EARLY ACS does not contradict the benefit of GP IIb/IIIa therapy over placebo in UA/NSTEMI in previous studies. Rather, its findings relate specifically to the timing of such therapy and selective versus routine use.

In another similar study, ACUITY (Acute Catheterization and Urgent Intervention Triage strategY), superiority of early GP IIb/IIIa therapy also was not found, but investigators could not exclude as much as a 29% benefit with GP IIb/IIIa therapy nor show noninferiority of delayed administration. In addition, drug exposure before angiography was much shorter (4 hours), which might substantially diminish the opportunity for differential efficacy (161).

The results of the EARLY ACS study are similar to a meta-analysis of 6 prior large, randomized trials of GP IIb/IIIa therapy versus placebo in non–ST-elevation ACS in which an invasive strategy was not mandated, which showed a relative reduction in death/MI of 9% (CI 2% to 16%) (162). In those patients who underwent PCI, the RR reduction was a more robust 23% (CI 8% to 36%). This finding also was consistent with EARLY ACS. Both studies found no benefit in troponin-negative patients. The investigators ascribed the less than expected benefit of early GP IIb/IIIa therapy in EARLY ACS to convergence of eptifibatide use in the 2 arms at the time of PCI and more aggressive contemporary cotherapies compared with earlier studies, including frequent use of clopidogrel, low-molecular-weight heparin, and statins. A further caution is the lack of follow-up for several years.

Despite a lack of clarity from the overall and subgroup results, an argument can be made against the routine upstream use of GP IIb/IIIa therapy in all non–ST-elevation ACS patients intended for an invasive strategy. In particular, those with a normal baseline troponin level and those over the age of 75 years, in whom there was no evidence for benefit but who showed an increased risk of bleeding, might be excluded. On the other hand, findings in those with a positive troponin at baseline and those with diabetes, although not definitive in EARLY ACS alone, trend positively and are in line with previous results. The EARLY ACS trial showed no significant benefit in the composite outcome comparing early versus delayed eptifibatide as defined by the study. Thus, at this time, a high-risk group that would clearly benefit from the early administration of eptifibatide upstream before cardiac catheterization has not been identified. Early GP IIb/IIIa therapy in patient groups continues to appear reasonable if they are judged clinically to be at high risk of thrombotic events relative to bleeding risk.