Strategies to Reduce Bleeding Risk in Acute Coronary Syndromes and Percutaneous Coronary Intervention: New and Emerging Pharmacotherapeutic Considerations James C. Coons,1,* and Taylor Miller2 1
University of Pittsburgh School of Pharmacy, Pittsburgh, Pennsylvania; 2Department of Pharmacy, UPMC-Presbyterian Hospital, Pittsburgh, Pennsylvania
Bleeding is a well-recognized complication among patients who undergo percutaneous coronary intervention for acute coronary syndromes. Patients who bleed in this setting have higher morbidity, mortality, and resource use. Several bleeding avoidance strategies have been identified including the use of certain pharmacologic interventions; however, the antithrombotic landscape for this patient population is evolving rapidly. Numerous practical issues are related to the appropriate selection and dosing of antiplatelets and anticoagulants, as well as their combinations, that may have an impact on bleeding. Thus we reviewed the recent evidence that describes the use of these agents and provide recommendations on the appropriate use of antiplatelets and anticoagulants in the context of current practice and guideline recommendations. Opportunities exist to reduce bleeding complications through the adoption of bleeding avoidance strategies and appropriate use of antiplatelets and anticoagulants. Pharmacist expertise is critical in the appropriate selection, dosing, and monitoring of these medications to improve patient safety as it relates to bleeding potential. Clinical outcomes evaluating combinations of antiplatelets and anticoagulants, particularly with novel anticoagulants, are ongoing as are studies that incorporate use of genotype and phenotype into antithrombotic decision making. KEY WORDS acute coronary syndrome, anticoagulation, antiplatelets, adverse drug reactions, bleeding. (Pharmacotherapy 2014;34(9):973–990) doi: 10.1002/phar.1447
Bleeding is an anticipated complication of left heart catheterization and a well-described adverse effect associated with the various combinations of antiplatelets and anticoagulants used in the management of acute coronary syndromes (ACS) and percutaneous coronary intervention (PCI). Bleeding estimates in these settings range from 1–10% but may approach 20% when dual antiplatelet therapy (DAPT) is combined with anticoagulation.1, 2 The variability in the incidence of bleeding can be explained by several factors including bleed definitions, patient-specific variables, concomitant therapies, and timing of event reporting. Significant heterogeneity *Address for correspondence: James C. Coons, University of Pittsburgh School of Pharmacy, 727 Salk Hall, 3501 Terrace Street, Pittsburgh, PA 15261; e-mail: [email protected]. Ó 2014 Pharmacotherapy Publications, Inc.
among bleeding definitions exists, and these differences have been well outlined in a consensus report.3 To harmonize these differences, the Bleeding Academic Research Consortium proposed a standardized definition and end-point reporting that facilitates a more critical examination of bleeding to better establish the relative safety and efficacy of vascular interventions.3 Impact of Bleeding on Outcomes Regardless of the exact definition used, the presence of bleeding has been shown to predict increased morbidity (nonfatal myocardial infarction [MI], stent thrombosis, and stroke) and mortality.3, 4 Recent data from the CathPCI Registry showed that 12.1% of all in-hospital mortality can be explained by major bleeding complications after PCI.5 The number needed to
974
PHARMACOTHERAPY Volume 34, Number 9, 2014
harm (NNH) for bleeding-related in-hospital mortality ranged from 16–117 depending on bleed risk and site of bleeding. Both access-site and non–access-site bleeds were associated with increased in-hospital mortality.5 Although the risk of mortality is highest soon after the bleeding event, the risk does persist over time.6 Several studies and registry data demonstrate a 2- to 10-fold excess relative risk (RR) of death at both 30 days and 1 year for patients who bleed.3 In fact, post-PCI bleeding confers the same risk of 1-year mortality as post-PCI MI.3 Even “nuisance” bleeding, defined as easy bleeding, bruising, and nose or gum bleeds, has been associated with decreased quality of life and a trend toward higher rehospitalizations.7 Nuisance bleeding is reported by more than a third of patients after acute MI and drug-eluting stent (DES) placement, and it adversely affects medication adherence with DAPT and quality of life.7 Other mechanisms by which bleeding itself may portend worse outcomes, beyond the risks of stent thrombosis and reinfarction due to drug discontinuation, include ischemia due to hypotension, shock, and/or anemia and transfusion-associated proinflammatory and prothrombotic responses.8 In addition to affecting clinical outcomes, bleeding also increases resource use (transfusions and length of stay) and costs.3 In-hospital major bleeding was found to be the most important factor that increased costs for patients undergoing early invasive management of ACS and represented 8% of total hospital costs.9 Each bleeding episode has been estimated at $6000– $8000 and is more costly than thrombotic complications on a per patient basis due to the frequency of their occurrence.9 Predictors of Bleeding and Risk Estimation Several risk factors have been identified as contributors to excess bleeding and should serve as a foundation for careful antiplatelet and anticoagulant selection and dosing.8 Independent predictors of bleeding include advanced age, female sex, history of renal insufficiency, lower body weight (LBW), history of bleeding, lower mean arterial pressure, use of invasive procedures, and antithrombotic therapy such as glycoprotein IIb/IIIa inhibitors (GPI) and fibrinolytics.8, 10 In fact, inappropriate selection and excess dosing of antithrombotics in patients with renal failure have been associated with major bleeding.11 The U.S. Food and Drug Administration (FDA)-approved antithrombotics
that have renal-dependent elimination include enoxaparin, fondaparinux, eptifibatide, tirofiban, bivalirudin, dabigatran, rivaroxaban, and apixaban. Use of the Cockcroft-Gault equation to estimate glomerular filtration rate and dose antithrombotics in patients with ACS is recommended and appears to be especially important in females, patients with LBW, and those with advanced age.12 However, dosage adjustments based on these methods are largely derived from pharmacokinetic analyses and preclinical data.13 Clinical experiences with antithrombotics in the context of randomized controlled trials are limited because renal dysfunction was an exclusion criterion for most studies. Therefore, the true risk of bleeding with these medications in outcomes-based trials is not well characterized. The CRUSADE Quality Improvement Initiative was used to derive and validate a bleeding risk score to predict in-hospital major bleeding.14 The risk score was based on predictors of bleeding that can be determined at the time of hospital admission: hematocrit, creatinine clearance (Clcr), heart rate, sex, signs of heart failure, prior vascular disease, diabetes mellitus, and systolic blood pressure.14 The CRUSADE bleeding risk score may be used to better assess baseline risk prior to the initiation of pharmacotherapy and guide the intensity of management as it relates to bleeding. Use of specific antithrombotic regimens are not part of the risk score itself, and there are no data describing its use in a prospective manner to direct pharmacotherapy decision making.14 More recently, a bleeding risk–prediction algorithm was developed that integrates six baseline factors (female sex, advanced age, elevated serum creatinine concentration, elevated white blood cell count, anemia, and ACS presentation) as well as one treatmentrelated variable (use of heparin plus a GPI as opposed to bivalirudin monotherapy).15 The use of bivalirudin monotherapy, rather than heparin plus GPI, was associated with significantly less non–coronary artery bypass grafting (CABG)related Thrombolysis in Myocardial Infarction (TIMI) major bleeding.15 Bleeding Avoidance Strategies Risk assessment of a patient’s baseline risk of bleeding can facilitate the prognostication of outcome as well as the “personalization” of appropriate management. The inclusion of treatment variables, such as the decision to use bivalirudin, also reinforces the notion that bleeding
BLEEDING IN ACS AND PCI: PHARMACOTHERAPY UPDATE Coons and Miller is an iatrogenic complication with actionable opportunities to mitigate or prevent its development. In fact, several potential bleeding avoidance strategies (BAS) have been identified in relation to PCI outcomes including pharmacologic (e.g., bivalirudin, short-duration GPI), procedural (e.g., earlier sheath removal, radial artery access), and technologic (vascular closure devices).16 Temporal trends in PCI-related bleeds from registry data demonstrate an overall reduction in major vascular complications that appears to be driven by changes in pharmacotherapy. For example, bivalirudin use increased by ~ 20% (with concomitant reduction in unfractionated heparin [UFH] plus GPI) between 2005 and 2009.16, 17 Overall, pre-PCI risk assessment has been shown to predict post-PCI bleeding, and the use of BAS is associated with significantly lower bleeding risks. It is a “risktreatment paradox,” however, that patients with the highest risk of bleeding are the least likely to receive such therapies including bivalirudin.4 It is anticipated that greater uptake of BAS into PCI practice may lead to less bleeding overall. Pharmacotherapeutic Considerations with Antiplatelets and Anticoagulants Despite the identification of current BAS, the antithrombotic landscape for treatment of ACS and PCI continues to evolve rapidly. Numerous pragmatic issues are related to the appropriate use and dosing of antiplatelets and anticoagulants including novel oral anticoagulants (NOACs), and their combinations that may have an impact on bleeding. Guideline updates from organizations such as the American College of Cardiology/American Heart Association/Society for Cardiovascular Angiography and Interventions (ACC/AHA/SCAI) reflect recent practice advances based on published literature.18 However, the evidence supporting pharmacotherapy in this setting continues to increase substantially. Table 1 summarizes the pharmacologic considerations that have been demonstrated to attenuate bleeding risk in the ACS/PCI setting. Appendix 1 defines clinical trial and quality improvement initiative acronyms.
975
meta-analyses, including the Antithrombotic Trialists’ Collaboration, demonstrated that the antiischemic efficacy of ASA does not appear to be dose dependent.26 Conversely, the risk of bleeding depends on dose and is significant for an excess of nonfatal bleeding.19 In fact, ASA doses greater than or equal to 325 mg/day have been shown to double the risk of gastrointestinal bleeding compared with doses less than 325 mg/ day.27 Data from a meta-analysis of 31 clinical trials explored the risk of bleeding across daily ASA doses categorized as follows: low (less than 100 mg), moderate (100–200 mg), and high (more than 200 mg).28 The authors found that bleeding rates were lowest for the low-dose group, but, more importantly, they observed a frequent excess of minor gastrointestinal, stroke, and total bleeding events even among those patients who received moderate doses of ASA. This finding further underscores the significance of using ASA doses less than 100 mg.28 The PLATO trial reflects continued uncertainty in ASA dosing, at least in the United States, where ~ 54% of patients received a median ASA dose greater than or equal to 300 mg/day compared with ~ 2% for patients enrolled from the rest of the world.29 More recent data, however, provide evidence to address this controversy. The landmark study CURRENT-OASIS 7 randomized patients with ACS who were planned for PCI in a 2 9 2 factorial design to double-dose or standard-dose clopidogrel (double-blind) as well as higher dose or lower dose ASA (open label).19 Table 1 lists the dosing definitions and primary results. The investigators reported no significant difference in major bleeding or the composite ischemic outcomes at 30 days, regardless of ASA dosing.19 These data also demonstrated no significant difference in the composite ischemic outcome between clopidogrel doses for the overall population; however, subgroup analysis of those patients who underwent primary PCI revealed significantly lower rates of the primary outcome and stent thrombosis.30 The higher clopidogrel dose resulted in a significant excess of bleeding at 30 days in both in the overall (Table 1) and PCI-only populations.28, 30 Prasugrel and Ticagrelor
Aspirin and Clopidogrel Dosing The optimal maintenance dose of aspirin (ASA) for the treatment of patients presenting with ACS and requiring PCI has not been well established historically. Observational data and
Prasugrel is a third-generation thienopyridine, whereas ticagrelor is a nonthienopyridine derivative. Both are potent P2Y12 inhibitors that have added to the evolving therapeutic armamentarium for treating the ischemic com-
Use SD (75 mg/day) for maintenance vs DD (600 mg on day 1, 150 mg/day on days 2–7, then 75 mg/day)
Avoid in patients with history of TIA/CVA; use clopidogrel or ticagrelor Use 5 mg/day maintenance dose for medically treated patients with ACS ≥ 75 yrs of age or body weight < 60 kg; alternatively, consider use of clopidogrel or ticagrelor Avoid pretreatment with prasugrel among patients with NSTE ACS planned for catheterization Avoid early routine precatheterization laboratory (“upstream”) use of PCI in favor of delayed provisional use at time of PCI
Consider bivalirudin as part of bleeding avoidance strategy when selecting anticoagulation
Clopidogrel
Prasugrel
Bivalirudin
GPI
Use lower dose (81 mg/day) post-PCI vs higher dose (≥ 300 mg/day)
Recommendations
Aspirin
Treatment
HORIZONS-AMI24 (n=3602)
EARLY-ACS23 (n=9492)
ACCOAST22 (n=4033)
TRILOGY ACS21 (n=9326)
TRITON-TIMI 3820 (n=13,608)
Early, routine use vs delayed, provisional use; primary outcome: 9.3% vs 10% (OR 0.92 [95% CI 0.80–1.06], p=0.23); TIMI major bleed: 2.6% vs 1.8% (p=0.015); TIMI minor bleed: 3.6% vs 1.7% (p
University of Pittsburgh School of Pharmacy, Pittsburgh, Pennsylvania; 2Department of Pharmacy, UPMC-Presbyterian Hospital, Pittsburgh, Pennsylvania
Bleeding is a well-recognized complication among patients who undergo percutaneous coronary intervention for acute coronary syndromes. Patients who bleed in this setting have higher morbidity, mortality, and resource use. Several bleeding avoidance strategies have been identified including the use of certain pharmacologic interventions; however, the antithrombotic landscape for this patient population is evolving rapidly. Numerous practical issues are related to the appropriate selection and dosing of antiplatelets and anticoagulants, as well as their combinations, that may have an impact on bleeding. Thus we reviewed the recent evidence that describes the use of these agents and provide recommendations on the appropriate use of antiplatelets and anticoagulants in the context of current practice and guideline recommendations. Opportunities exist to reduce bleeding complications through the adoption of bleeding avoidance strategies and appropriate use of antiplatelets and anticoagulants. Pharmacist expertise is critical in the appropriate selection, dosing, and monitoring of these medications to improve patient safety as it relates to bleeding potential. Clinical outcomes evaluating combinations of antiplatelets and anticoagulants, particularly with novel anticoagulants, are ongoing as are studies that incorporate use of genotype and phenotype into antithrombotic decision making. KEY WORDS acute coronary syndrome, anticoagulation, antiplatelets, adverse drug reactions, bleeding. (Pharmacotherapy 2014;34(9):973–990) doi: 10.1002/phar.1447
Bleeding is an anticipated complication of left heart catheterization and a well-described adverse effect associated with the various combinations of antiplatelets and anticoagulants used in the management of acute coronary syndromes (ACS) and percutaneous coronary intervention (PCI). Bleeding estimates in these settings range from 1–10% but may approach 20% when dual antiplatelet therapy (DAPT) is combined with anticoagulation.1, 2 The variability in the incidence of bleeding can be explained by several factors including bleed definitions, patient-specific variables, concomitant therapies, and timing of event reporting. Significant heterogeneity *Address for correspondence: James C. Coons, University of Pittsburgh School of Pharmacy, 727 Salk Hall, 3501 Terrace Street, Pittsburgh, PA 15261; e-mail: [email protected]. Ó 2014 Pharmacotherapy Publications, Inc.
among bleeding definitions exists, and these differences have been well outlined in a consensus report.3 To harmonize these differences, the Bleeding Academic Research Consortium proposed a standardized definition and end-point reporting that facilitates a more critical examination of bleeding to better establish the relative safety and efficacy of vascular interventions.3 Impact of Bleeding on Outcomes Regardless of the exact definition used, the presence of bleeding has been shown to predict increased morbidity (nonfatal myocardial infarction [MI], stent thrombosis, and stroke) and mortality.3, 4 Recent data from the CathPCI Registry showed that 12.1% of all in-hospital mortality can be explained by major bleeding complications after PCI.5 The number needed to
974
PHARMACOTHERAPY Volume 34, Number 9, 2014
harm (NNH) for bleeding-related in-hospital mortality ranged from 16–117 depending on bleed risk and site of bleeding. Both access-site and non–access-site bleeds were associated with increased in-hospital mortality.5 Although the risk of mortality is highest soon after the bleeding event, the risk does persist over time.6 Several studies and registry data demonstrate a 2- to 10-fold excess relative risk (RR) of death at both 30 days and 1 year for patients who bleed.3 In fact, post-PCI bleeding confers the same risk of 1-year mortality as post-PCI MI.3 Even “nuisance” bleeding, defined as easy bleeding, bruising, and nose or gum bleeds, has been associated with decreased quality of life and a trend toward higher rehospitalizations.7 Nuisance bleeding is reported by more than a third of patients after acute MI and drug-eluting stent (DES) placement, and it adversely affects medication adherence with DAPT and quality of life.7 Other mechanisms by which bleeding itself may portend worse outcomes, beyond the risks of stent thrombosis and reinfarction due to drug discontinuation, include ischemia due to hypotension, shock, and/or anemia and transfusion-associated proinflammatory and prothrombotic responses.8 In addition to affecting clinical outcomes, bleeding also increases resource use (transfusions and length of stay) and costs.3 In-hospital major bleeding was found to be the most important factor that increased costs for patients undergoing early invasive management of ACS and represented 8% of total hospital costs.9 Each bleeding episode has been estimated at $6000– $8000 and is more costly than thrombotic complications on a per patient basis due to the frequency of their occurrence.9 Predictors of Bleeding and Risk Estimation Several risk factors have been identified as contributors to excess bleeding and should serve as a foundation for careful antiplatelet and anticoagulant selection and dosing.8 Independent predictors of bleeding include advanced age, female sex, history of renal insufficiency, lower body weight (LBW), history of bleeding, lower mean arterial pressure, use of invasive procedures, and antithrombotic therapy such as glycoprotein IIb/IIIa inhibitors (GPI) and fibrinolytics.8, 10 In fact, inappropriate selection and excess dosing of antithrombotics in patients with renal failure have been associated with major bleeding.11 The U.S. Food and Drug Administration (FDA)-approved antithrombotics
that have renal-dependent elimination include enoxaparin, fondaparinux, eptifibatide, tirofiban, bivalirudin, dabigatran, rivaroxaban, and apixaban. Use of the Cockcroft-Gault equation to estimate glomerular filtration rate and dose antithrombotics in patients with ACS is recommended and appears to be especially important in females, patients with LBW, and those with advanced age.12 However, dosage adjustments based on these methods are largely derived from pharmacokinetic analyses and preclinical data.13 Clinical experiences with antithrombotics in the context of randomized controlled trials are limited because renal dysfunction was an exclusion criterion for most studies. Therefore, the true risk of bleeding with these medications in outcomes-based trials is not well characterized. The CRUSADE Quality Improvement Initiative was used to derive and validate a bleeding risk score to predict in-hospital major bleeding.14 The risk score was based on predictors of bleeding that can be determined at the time of hospital admission: hematocrit, creatinine clearance (Clcr), heart rate, sex, signs of heart failure, prior vascular disease, diabetes mellitus, and systolic blood pressure.14 The CRUSADE bleeding risk score may be used to better assess baseline risk prior to the initiation of pharmacotherapy and guide the intensity of management as it relates to bleeding. Use of specific antithrombotic regimens are not part of the risk score itself, and there are no data describing its use in a prospective manner to direct pharmacotherapy decision making.14 More recently, a bleeding risk–prediction algorithm was developed that integrates six baseline factors (female sex, advanced age, elevated serum creatinine concentration, elevated white blood cell count, anemia, and ACS presentation) as well as one treatmentrelated variable (use of heparin plus a GPI as opposed to bivalirudin monotherapy).15 The use of bivalirudin monotherapy, rather than heparin plus GPI, was associated with significantly less non–coronary artery bypass grafting (CABG)related Thrombolysis in Myocardial Infarction (TIMI) major bleeding.15 Bleeding Avoidance Strategies Risk assessment of a patient’s baseline risk of bleeding can facilitate the prognostication of outcome as well as the “personalization” of appropriate management. The inclusion of treatment variables, such as the decision to use bivalirudin, also reinforces the notion that bleeding
BLEEDING IN ACS AND PCI: PHARMACOTHERAPY UPDATE Coons and Miller is an iatrogenic complication with actionable opportunities to mitigate or prevent its development. In fact, several potential bleeding avoidance strategies (BAS) have been identified in relation to PCI outcomes including pharmacologic (e.g., bivalirudin, short-duration GPI), procedural (e.g., earlier sheath removal, radial artery access), and technologic (vascular closure devices).16 Temporal trends in PCI-related bleeds from registry data demonstrate an overall reduction in major vascular complications that appears to be driven by changes in pharmacotherapy. For example, bivalirudin use increased by ~ 20% (with concomitant reduction in unfractionated heparin [UFH] plus GPI) between 2005 and 2009.16, 17 Overall, pre-PCI risk assessment has been shown to predict post-PCI bleeding, and the use of BAS is associated with significantly lower bleeding risks. It is a “risktreatment paradox,” however, that patients with the highest risk of bleeding are the least likely to receive such therapies including bivalirudin.4 It is anticipated that greater uptake of BAS into PCI practice may lead to less bleeding overall. Pharmacotherapeutic Considerations with Antiplatelets and Anticoagulants Despite the identification of current BAS, the antithrombotic landscape for treatment of ACS and PCI continues to evolve rapidly. Numerous pragmatic issues are related to the appropriate use and dosing of antiplatelets and anticoagulants including novel oral anticoagulants (NOACs), and their combinations that may have an impact on bleeding. Guideline updates from organizations such as the American College of Cardiology/American Heart Association/Society for Cardiovascular Angiography and Interventions (ACC/AHA/SCAI) reflect recent practice advances based on published literature.18 However, the evidence supporting pharmacotherapy in this setting continues to increase substantially. Table 1 summarizes the pharmacologic considerations that have been demonstrated to attenuate bleeding risk in the ACS/PCI setting. Appendix 1 defines clinical trial and quality improvement initiative acronyms.
975
meta-analyses, including the Antithrombotic Trialists’ Collaboration, demonstrated that the antiischemic efficacy of ASA does not appear to be dose dependent.26 Conversely, the risk of bleeding depends on dose and is significant for an excess of nonfatal bleeding.19 In fact, ASA doses greater than or equal to 325 mg/day have been shown to double the risk of gastrointestinal bleeding compared with doses less than 325 mg/ day.27 Data from a meta-analysis of 31 clinical trials explored the risk of bleeding across daily ASA doses categorized as follows: low (less than 100 mg), moderate (100–200 mg), and high (more than 200 mg).28 The authors found that bleeding rates were lowest for the low-dose group, but, more importantly, they observed a frequent excess of minor gastrointestinal, stroke, and total bleeding events even among those patients who received moderate doses of ASA. This finding further underscores the significance of using ASA doses less than 100 mg.28 The PLATO trial reflects continued uncertainty in ASA dosing, at least in the United States, where ~ 54% of patients received a median ASA dose greater than or equal to 300 mg/day compared with ~ 2% for patients enrolled from the rest of the world.29 More recent data, however, provide evidence to address this controversy. The landmark study CURRENT-OASIS 7 randomized patients with ACS who were planned for PCI in a 2 9 2 factorial design to double-dose or standard-dose clopidogrel (double-blind) as well as higher dose or lower dose ASA (open label).19 Table 1 lists the dosing definitions and primary results. The investigators reported no significant difference in major bleeding or the composite ischemic outcomes at 30 days, regardless of ASA dosing.19 These data also demonstrated no significant difference in the composite ischemic outcome between clopidogrel doses for the overall population; however, subgroup analysis of those patients who underwent primary PCI revealed significantly lower rates of the primary outcome and stent thrombosis.30 The higher clopidogrel dose resulted in a significant excess of bleeding at 30 days in both in the overall (Table 1) and PCI-only populations.28, 30 Prasugrel and Ticagrelor
Aspirin and Clopidogrel Dosing The optimal maintenance dose of aspirin (ASA) for the treatment of patients presenting with ACS and requiring PCI has not been well established historically. Observational data and
Prasugrel is a third-generation thienopyridine, whereas ticagrelor is a nonthienopyridine derivative. Both are potent P2Y12 inhibitors that have added to the evolving therapeutic armamentarium for treating the ischemic com-
Use SD (75 mg/day) for maintenance vs DD (600 mg on day 1, 150 mg/day on days 2–7, then 75 mg/day)
Avoid in patients with history of TIA/CVA; use clopidogrel or ticagrelor Use 5 mg/day maintenance dose for medically treated patients with ACS ≥ 75 yrs of age or body weight < 60 kg; alternatively, consider use of clopidogrel or ticagrelor Avoid pretreatment with prasugrel among patients with NSTE ACS planned for catheterization Avoid early routine precatheterization laboratory (“upstream”) use of PCI in favor of delayed provisional use at time of PCI
Consider bivalirudin as part of bleeding avoidance strategy when selecting anticoagulation
Clopidogrel
Prasugrel
Bivalirudin
GPI
Use lower dose (81 mg/day) post-PCI vs higher dose (≥ 300 mg/day)
Recommendations
Aspirin
Treatment
HORIZONS-AMI24 (n=3602)
EARLY-ACS23 (n=9492)
ACCOAST22 (n=4033)
TRILOGY ACS21 (n=9326)
TRITON-TIMI 3820 (n=13,608)
Early, routine use vs delayed, provisional use; primary outcome: 9.3% vs 10% (OR 0.92 [95% CI 0.80–1.06], p=0.23); TIMI major bleed: 2.6% vs 1.8% (p=0.015); TIMI minor bleed: 3.6% vs 1.7% (p
