Methods Study design. Study procedures

Transcription

1 Impact of prasugrel pretreatment and timing of coronary artery bypass grafting on clinical outcomes of patients with non-st-segment elevation myocardial infarction: From the A Comparison of Prasugrel at PCI or Time of Diagnosis of Non ST- Elevation Myocardial Infarction (ACCOAST) study Dariusz Dudek, MD, PhD, a Artur Dziewierz, MD, a Petr Widimsky, MD, DrSc, b Leonardo Bolognese, MD, c Patrick Goldstein, MD, d Christian Hamm, MD, e Jean-Francois Tanguay, MD, f LeRoy LeNarz, MD, g Debra L. Miller, RN, RCIS, g Eileen Brown, PhD, g Jurrien ten Berg, MD, PhD, h and Gilles Montalescot, MD, PhD i, for the ACCOAST Investigators Krakow, Poland; Prague, Czech Republic; Arezzo, Italy; Lille, Paris, France; Bad Nauheim, Germany; Quebec, Canada; IN, USA; and Nieuwegein, Netherlands Objectives We evaluated impact of timing of coronary artery bypass grafting (CABG) and prasugrel pretreatment in patients with non ST-segment elevation myocardial infarction undergoing CABG in the ACCOAST study. Methods Of 4033 enrolled patients, 314 (7.8%) underwent isolated CABG through 30 days. Primary efficacy end point for this analysis was any cardiovascular death, myocardial infarction, stroke, urgent revascularization, or glycoprotein IIb/IIIa inhibitor bailout through 30 days. Results More CABG versus percutaneous coronary intervention or medically managed patients were men, diabetic, or had peripheral arterial disease. Per randomization, 157 of 314 patients received a 30-mg prasugrel loading dose before CABG, and 157 of 314 received placebo. Patients were stratified by tertile of time from randomization to CABG: b2.98 days (n = 104), 2.98 and b6.95 days (n = 106), and 6.95 days (n = 104). Primary end point occurred in 12.5%, 4.7%, and 4.8%, respectively (b2.98 days vs other tertiles, hazard ratio [HR] = 2.80; P =.011). Similarly, the rate of all TIMI major bleeding was highest in the lowest tertile From the a Department of Interventional Cardiology, Jagiellonian University Medical College, Krakow, Poland, b Third Medical Faculty of Charles University and University Hospital Royal Vineyards, Prague, Czech Republic, c Cardiovascular and Neurological Department, Azienda Ospedaliera, Arezzo, Italy, d SAMU and Emergency Department, Lille University Hospital, Lille, France, e Kerckhoff Heart and Thorax Center, Bad Nauheim, Germany, f Montreal Heart Institute, Université de Montréal, Montreal, Quebec, Canada, g Eli Lilly and Company, Lilly Corporate Center, Indianapolis, IN, USA, h Department of Cardiology, St. Antonius Hospital, Nieuwegein, Netherlands, and i ACTION Study Group, Institut de Cardiologie, Centre Hospitalier Universitaire Pitié-Salpêtriėre (AP-HP), Université Paris 6, Paris, France. ClinicalTrials.gov Identifier NCT Conflicts of interest: Dariusz Dudek has received consulting and lecture fees from Abbott, Adamed, Adyton Medical Polska, Abiomed Europe, AstraZeneca, Biotronik, Balton, Bayer, Braun, BioMatrix, Boston Scientific, Boehringer Ingelheim, Bracco, Bristol-Myers Squibb, Comesa Polska, Cordis, Cook, Covidien Polska, DRG MedTek, Eli Lilly, EuroCor, Hammermed, GE Healthcare, GlaxoSmithKline, Inspire-MD, Iroko Cardio International, Medianet,Medtronic,TheMedicinesCompany,MerilLifeSciences,MerckSharp& Dohme, Orbus-Neich, Pfizer, Possis, ProCardia Medical, Promed, REVA Medical, Sanofi, Siemens, Solvay, Stentys, St Jude Medical, Terumo, Tyco, and Volcano. Artur Dziewierz has no conflicts of interest to declare. Petr Widimsky has received consulting and lecture fees from Eli Lilly and Daiichi Sankyo. Leonardo Bolognese has received fees for board membership from Daiichi Sankyo and Eli Lilly; consulting fees from Daiichi Sankyo; and lecture fees from Daiichi Sankyo, Eli Lilly, Menarini, Abbott, AstraZeneca, and Iroko Cardio International. Patrick Goldstein has received fees for board membership from Boehringer Ingelheim, The Medicines Company, Daiichi Sankyo, and Eli Lilly and Company; consulting fees from Boehringer Ingelheim, Bayer, Sanofi, AstraZeneca, and Eli Lilly; lecture fees from Boehringer Ingelheim, The Medicines Company, Daiichi Sankyo, Bayer, Sanofi, and AstraZeneca; payment for the development of educational presentations from Boehringer Ingelheim, and AstraZeneca; and travel support from Astra Zeneca, Bayer, Boehringer Ingelheim, Daiichi Sankyo, The Medicines Company, and Sanofi. Christian Hamm has received payment for board membership from AstraZeneca, Medtronic, and Boehringer Ingelheim; consulting and lecture fees from Medtronic, Boehringer Ingelheim, Eli Lilly, The Medicines Company, Abbott Vascular, Bayer, Sanofi, Boston Scientific, Correvio, Roche Diagnostics, Pfizer, Cordis, Daiichi Sankyo, and GlaxoSmithKline; and lecture fees from AstraZeneca and Merck. Jean-Francois Tanguay has received consulting fees from Eli Lilly, AstraZeneca, Abbott Vascular, Roche, and GlaxoSmithKline; lecture fees from Sanofi, Eli Lilly, AstraZeneca, Abbott Vascular, Pfizer, and Bristol-Myers Squibb; and grant support from Eli Lilly, AstraZeneca, Roche, Hexacath, Ikaria, Abbott Vascular, GlaxoSmithKline, Roche, and Sanofi. LeRoy LeNarz, Debra L. Miller, and Eileen Brown are employees and shareholders of Eli Lilly. Jurrien ten Berg has received fees for board membership from AstraZeneca; consulting fees from AstraZeneca, Eli Lilly, and Merck; and lecture fees from AstraZeneca and Eli Lilly. Gilles Montalescot reports receiving consulting fees from Bayer, Boehringer Ingelheim, Cardiovascular Research Foundation, Europa Organisation, the Gerson Lehrman Group, Iroko Cardio International, Lead-Up, Luminex, McKinsey & Company, Inc, Remedica, Servier, TIMI Study Group, WebMD, Wolters Kluwer Health, Bristol-Myers Squibb, AstraZeneca, Biotronik, Eli Lilly, The Medicines Company, Menarini Group, Roche, Sanofi-Aventis, Pfizer, Daiichi-Sankyo, and Medtronic; and grant support from Bristol-Myers Squibb, AstraZeneca, Biotronik, Eli Lilly, The Medicines Company, Menarini Group, Sanofi-Aventis, Pfizer, Roche, Accumetrics, Medtronic, Abbott Laboratories, Daiichi-Sankyo, Nanosphere, Inc, and Stentys. Submitted March 10, 2015; accepted July 16, Reprint requests: Dariusz Dudek, MD, PhD, Department of Interventional Cardiology, Jagiellonian University Medical College, 17 Kopernika St., , Krakow, Poland. mcdudek@cyfronet.pl Elsevier Inc. All rights reserved.

2 1026 Dudek et al American Heart Journal November 2015 (26.0% vs 10.4% and 4.8%; P b.001), but no difference in all-cause death was observed through 30 days (3.9% vs 1.9% and 1.9%; P =.30). Time from randomization to CABG (HR = 0.84 for each day delay), left main disease (HR = 1.76), region of enrollment (Non- Eastern Europe vs Eastern Europe; HR = 3.83), but not prasugrel pretreatment and baseline troponin 3 upper limit of normal, were independent predictors of combined 30-day end point of all-cause death/myocardial infarction/stroke/timi major bleeding. Conclusions In ACCOAST, early (b2.98 days) surgical revascularization carried increased risk of bleeding and ischemic complications without affecting all-cause mortality through 30 days. Baseline troponin and prasugrel pretreatment did not impact ischemic clinical outcomes. (Am Heart J 2015;170: e2.) Dual antiplatelet therapy with acetylsalicylic acid and a P2Y 12 inhibitor (clopidogrel, prasugrel, or ticagrelor) is standard of care in patients with acute coronary syndromes (ACS). 1,2 Early initiation of antiplatelet treatment in patients with non ST-segment elevation ACS may reduce the rate of ischemic events; however, it may increase the risk of bleeding, especially in patients who proceed to urgent coronary artery bypass graft (CABG) surgery after coronary angiogram. 3 8 To reduce the risk of CABG-related bleeding, postponing of CABG and discontinuation of a P2Y 12 inhibitor 5 7 days before surgery is recommended, if possible. 1 3 On the other hand, delaying surgery in patients with non ST-segment elevation myocardial infarction (NSTEMI) might increase resource use and increase the risk of recurrent ischemia/ MI in patients waiting for surgical revascularization. Prasugrel and ticagrelor were shown to be more effective than clopidogrel in terms of reduction of ischemic events in patients with ACS Interestingly, the use of either prasugrel or ticagrelor was associated with reduced mortality in patients with ACS undergoing CABG. 4,5,12 However, prasugrel use in the TRITON-TIMI 38 study was associated with an elevated risk of bleeding by TIMI adjudication, platelet transfusion, and surgical re-exploration for bleeding in comparison to clopidogrel, in which the majority of the CABG procedures were performed during the maintenance dose phase. 4 No increase of major CABG-related bleeding was observed in patients treated with ticagrelor as compared with patients treated with clopidogrel in the PLATO by PLATO adjudication criteria. 5 As the majority of patients included in PLATO and TRITON-TIMI 38 studies underwent elective CABG rather than an urgent procedure, the optimal timing of CABG after presentation with NSTEMI, and the impact of pretreatment with oral P2Y 12 inhibitors in patients undergoing early CABG, is still a matter of debate. An increased preoperative troponin level in patients with acute MI undergoing CABG has been associated with a higher incidence of major adverse cardiovascular events and hospital mortality. 13 Thus, regional differences for delaying CABG, due to entry level of troponin and reducing the risk of bleeding complications when a P2Y 12 inhibitor is utilized, are essential for understanding patient outcomes. We sought to evaluate the impact of timing of CABG and prasugrel pretreatment on clinical outcomes of patients with NSTEMI undergoing CABG and enrolled in the ACCOAST study. Methods Study design Design and main results of the ACCOAST study have been published elsewhere (ClinicalTrials.gov identifier was NCT ). 14,15 In brief, ACCOAST was a phase 3, double-blind study that randomized 4033 patients to either an initial loading dose of prasugrel 30 mg after the NSTEMI diagnosis, followed by coronary angiography with an additional dose of prasugrel 30 mg given at the time of the percutaneous coronary intervention (PCI) procedure if PCI was indicated (prasugrel pretreatment), or placebo followed by a loading dose of prasugrel 60 mg given only to patients undergoing PCI at the time of the procedure if PCI was indicated (prasugrel no-pretreatment). Key inclusion criteria included NSTEMI symptoms within 48 hours prior to study entry and elevated troponin ( 1.5 upper limit of normal [ULN]) per local lab(s), and the patient was to be scheduled for coronary angiography and PCI within 2 to 24 hours of randomization and no later than 48 hours after randomization. Key exclusion criteria included STEMI patients, a medical history contraindicating therapy with prasugrel, a history of stroke or transient ischemic attack, or a loading dose of any P2Y 12 antagonist 7daysofstudy entry. The study design and protocol were approved by the national regulatory authorities in all participating countries, and by local ethics committees/institutional review boards at all participating sites. All patients provided written informed consent. The ACCOAST study was funded by Daiichi Sankyo, Ltd, and Eli Lilly and Company. The authors are solely responsible for the design and conduct of this substudy, all substudy analyses, the drafting and editing of the manuscript, and its final contents. Study procedures Eligible patients were randomly assigned to receive either prasugrel or matching placebo with a randomization ratio of 1:1. Patients received a 30-mg loading dose of prasugrel (or placebo) immediately after randomization. In the pretreatment group (those patients that received a 30-mg loading dose of prasugrel after randomization), a second 30-mg loading dose of prasugrel was given after the coronary angiogram at the time of PCI once the indication was confirmed. In the no-pretreatment group, the approved 60-mg prasugrel loading dose was given only in patients undergoing PCI at the time of PCI. If the

3 American Heart Journal Volume 170, Number 5 Dudek et al 1027 Table I. Baseline characteristics of patients undergoing CABG only in the ACCOAST study stratified by time from randomization to CABG Time from randomization to CABG Variable b2.98 d (n = 104) 2.98 and b6.95 d (n = 106) 6.95 d (n = 104) P P P Eastern Europe, (%) Western Europe/Other, (%) Age, median, (y) Age 75 y, (%) Female, (%) Weight b60 kg, (%) BMI, median (kg/m 2 ) Current smoker (%) Arterial hypertension, (%) Hypercholesterolemia, (%) Diabetes mellitus, (%) Peripheral artery disease, (%) Carotid artery disease, (%) Previous myocardial infarction, (%) Previous CABG, (%) Previous PCI, (%) Family history of CAD, (%) Previous stroke, (%) Congestive heart failure, (%) Atrial fibrillation, (%) CRUSADE Risk Score, median, (points) Very high CRUSADE Risk Score, (%) Abnormal creatinine (%) GRACE score, median, (points) GRACE score 140, (%) Baseline troponin level, (%) b3 ULN ULN and b10 ULN ULN Baseline antithrombin therapy LMWH monotherapy UFH monotherapy b.001 b Fondaprinux monotherapy Multiple agents No antithrombin usage Left main disease, (%) Abbreviations: BMI, Body mass index; CAD, coronary artery disease; LMWH, low-molecular-weight heparin; UFH, unfractionated heparin. P value comparing 'b2.98 days' vs ' 2.98 and b6.95 days' group. P value comparing 'b2.98 days' vs ' 6.95 days' group. P value comparing ' 2.98 and b6.95 days' vs ' 6.95 days' group. Primarily Western Europe; also includes Israel and Canada. coronary angiography revealed an anatomy not amenable to PCI, the patient did not receive the second loading dose. Any subsequent thienopyridine therapy (open-label clopidogrel or ticlopidine) for patients who were medically managed or who proceeded to CABG surgery were left to the investigators discretion. Investigators were instructed to delay any surgery for at least 7 days if an antiplatelet effect was not desired; however, data on antiplatelet therapy during peri- and post-operative period were not collected for patients treated with CABG. The use of platelet function testing was prohibited during the ACCOAST study (with the exception of the blinded substudy). Study population For the purpose of this pre-specified analysis of the ACCOAST study, data on patients who were treated with CABG during the course of the study were retrieved from the database. Twenty-six patients who were initially treated with PCI and then proceeded with CABG were excluded from the analysis. Finally, 314 (7.8%) of the 4033 ACCOAST patients underwent isolated CABG during the study. These patients constituted the study group. Study outcomes The primary efficacy measure for the ACCOAST study was a composite of cardiovascular death, MI, stroke, urgent revascularization, or glycoprotein IIb/IIIa inhibitor bailout through 7 days from randomization. However, for the purpose of the present analysis, 30-day outcome data were assessed. Major safety end points of TIMI major and minor bleeding were evaluated in terms of CABG surgery-related, non-cabg surgery-related, and all

4 1028 Dudek et al American Heart Journal November 2015 Table II. Primary and secondary end points through 30 days from randomization for patients undergoing CABG only in the ACCOAST study stratified by time from randomization to CABG Time from randomization to CABG Variable b2.98 d (n = 104) 2.98 and b6.95 d (n = 106) 6.95 d (n = 104) P P P CVD/MI/stroke/UR/GP IIb/IIIa bailout, (%) CVD, (%) MI, (%) Stroke, (%) UR, (%) NE NE NE GP IIb/IIIa bailout, (%) NE NE NE All TIMI major bleeding, (%) b CABG related TIMI major bleeding, (%) b All TIMI minor bleeding, (%) All-cause death, (%) All-cause death, MI, stroke and all TIMI major bleeding, (%) b.001 b Abbreviations: CVD, Cardiovascular death; GP, glycoprotein; MI, myocardial infarction; NE, not evaluable; UR, urgent revascularization. P value comparing 'b2.98 days' vs ' 2.98 and b6.95 days' group. P value comparing 'b2.98 days' vs ' 6.95 days' group. P value comparing ' 2.98 and b6.95 days' vs ' 6.95 days' group. bleeding. Independent predictors of the composite end point of all-cause death, MI, stroke, and TIMI major bleeding were evaluated. Additionally, total chest tube volume (milliliters) and transfusion data were collected during the study. All definitions of individual end points have been published previously. 14,15 Statistical analysis Patients were stratified by tertiles of the time from randomization to CABG and randomization assignment (prasugrel pretreatment vs no-pretreatment). Continuous variables were presented as medians (interquartile ranges) or mean ± SD, and categorical variables were expressed as percentages. Patient characteristics on a continuous scale were compared using analysis of variance model. Patient characteristics on a categorical scale were compared using the Pearson χ 2 test or Fisher exact test, as appropriate. The independent predictors of delaying CABG were assessed through a multivariable assessment using a step-wise selection process starting with all variables in Table I with entry P value of.15. Multivariable analyses for composite efficacy end points and for all TIMI major bleeding used Cox proportional hazards models with variables for randomized treatment group, left main disease, high baseline troponin, Eastern European region, and time to CABG. Efficacy and safety end point comparisons were performed on the basis of time to first event, using a two sided log-rank test. The HR and a 2-sided 95% CI are provided via a Cox proportional hazards model. P values for efficacy and safety were considered statistically significant when at b.05. The inferential statistics were only provided if at least a total of 10 patients experienced the event. All statistical analysis was performed using SAS (version 9.2; SAS Institute, Cary, NC). Results Study population A total of 4033 patients with NSTEMI from 171 centers between December 2009 and November 2012 were randomized in the ACCOAST study. Of these, 314 (7.8%) patients underwent isolated CABG through 30 days the study group. The CABG-only cohort had more males, and more patients with a history of diabetes mellitus, peripheral arterial disease, and a higher GRACE score as compared with patients treated with PCI or patients scheduled for medical management (online Appendix Supplementary Table I). In contrast, patients treated with CABG were less likely to have a history of previous cardiac revascularization, both PCI and CABG. Overall cumulative hazard and time to CABG were comparable between treatment cohorts (online Appendix Supplementary Figure). Impact of CABG timing on clinical outcomes Median (interquartile range) time from randomization to CABG was 5.48 ( ) days. Baseline characteristics of CABG patients stratified by tertiles of the time from randomization to CABG (lowest tertile: b2.98 days [n = 104]; intermediate tertile: 2.98 and b6.95 days [n = 106]; highest tertile: 6.95 days [n = 104]) are shown in Table I. There was no difference in the use of aspirin at baseline and at day 7 between groups (aspirin was used in all patients). In contrast, patients from the lowest tertile were more likely to be treated with unfractionated heparin then patients from other tertiles (Table I). Region of enrollment, baseline troponin level, and left main disease tended to be associated with timing of CABG. Primary efficacy end point occurred in 12.5% of patients in the lowest, in 4.7% in the intermediate, and

5 American Heart Journal Volume 170, Number 5 Dudek et al 1029 Table III. Primary and secondary end points through 30 days from randomization for patients undergoing CABG only in the ACCOAST study stratified by prasugrel pretreatment strategy Variable Prasugrel pretreatment (n = 157) Prasugrel no-pretreatment (n = 157) Hazard ratio 95% CI P CVD/MI/stroke/UR/GP IIb/IIIa bailout, (%) ( ).29 CVD, (%) ( ).16 MI, (%) NE NE Stroke, (%) ( ).41 UR, (%) NE NE GP IIb/IIIa bailout, (%) NE NE All TIMI major bleeding, (%) ( ).07 CABG related TIMI major bleeding, (%) ( ).09 All TIMI minor bleeding, (%) ( ).46 All-cause death, (%) ( ).16 All-cause death, MI, stroke and all TIMI major bleeding ( ).23 Abbreviations: CVD, Cardiovascular death; GP, glycoprotein; MI, myocardial infarction; NE, not evaluable; UR, urgent revascularization. Hazard ratio and two-sided 95% confidence interval are from a Cox proportional hazards model with treatment as a fixed effect. Two-sided P value based on the log rank test. in 4.8% in the highest time to CABG tertile (lowest versus other tertiles, HR = 2.80; P =.011). Similarly, the rate of all TIMI major bleeding was highest in the lowest time to CABG tertile (26.0% vs 10.4% and 4.8%, P b.001), but no difference in all-cause death through 30 days was confirmed (Table II). The rate of combined end point of all-cause death, MI, stroke, and all TIMI major bleeding through 30 days was 32.7% of patients in the lowest, in 14.2% in the intermediate, and in 9.6% in the highest time to CABG tertile (P b.001 for the lowest versus the intermediate and highest tertiles). Impact of prasugrel pretreatment on clinical outcomes Of 314 patients treated with isolated CABG through 30 days, 157 patients received prasugrel pretreatment (a 30-mg loading dose at the time of diagnosis), and 157 patients did not receive prasugrel pretreatment (placebo at the time of diagnosis and no prasugrel thereafter). The groups were well balanced in terms of baseline characteristics, especially chronic kidney disease, GRACE score, and CRUSADE score (data not shown). There was no difference in the rate of the primary efficacy end point between patients with and without prasugrel pretreatment (Table III). Similarly, no difference in all-cause death was observed. However, a trend towards higher rate of all TIMI major bleeding was confirmed for patients who received prasugrel pretreatment (pretreatment vs no-pretreatment: 17.2% vs 10.2%; P =.07). A trend towards a higher rate of any transfusions was observed in the pretreatment as compared to no-pretreatment group (21.7% vs 15.3%; P = 0.11). The frequency of fresh frozen plasma administration was higher in the prasugrel pretreatment group (8.3% vs 2.5%; P =.025), but without difference in packed red blood cells (17.8% vs 12.7%; P =.16) and platelets (7.0% vs 5.1%; P =.48)usecompared with the prasugrel no-pretreatment group. Surgical exploration was left to the surgeon s decision and was infrequent. There were four total, one patient in the pretreatment group and three patients in the non-pretreatment group. A total of 247 patients (121 prasugrel pretreatment, 126 placebo) had chest tube drain volume to report. The total chest tube drain volumes for the prasugrel pretreatment group were higher as compared with the no-pretreatment group (median [interquartile range]: 800 [ ] ml vs 600 [ ] ml; P =.012). Medians of the total chest tube drain volumes were higher in patients pretreated with prasugrel as compared with non-pretreated patients among patients in the lowest and in the intermediate time to CABG tertile (Figure). However, no difference in bleeding events was confirmed when patients were stratified by prasugrel pretreatment strategy and timing of CABG (Table IV). Multivariable analysis Time from randomization to CABG was identified as an independent predictor for the primary efficacy end point (HR = 0.87 for each day of delay). Time from randomization to CABG (HR = 0.84 for each day of delay), left main disease (HR = 1.76), and region of enrollment (Non-Eastern Europe vs Eastern Europe; HR = 3.83), but not prasugrel pretreatment or baseline troponin level 3 ULN were identified as independent predictors of the combined end point of all-cause death, MI, stroke, and TIMI major bleeding through 30 days. Independent predictors for all TIMI major bleeding were prasugrel pretreatment (HR = 2.05), left main disease (HR = 2.48), region of enrollment (Non Eastern Europe vs Eastern Europe; HR = 5.91) and time from randomization to CABG (HR = 0.77 for each day of delay). Discussion The main finding of the present study is that, in patients with NSTEMI scheduled to undergo CABG, early (b2.98 days) surgical revascularization carried an increased risk of

6 1030 Dudek et al American Heart Journal November 2015 Total chest tube drain volume, ml (median; Q1, Q3) Figure P =.015 <2.98 Days (n = 38, 43) P = and <6.95 Days (n = 46, 37) Pretreatment (n = 121) No pretreatment (n = 126) P = Days (n = 37, 46) Days from first loading-dose to coronary artery bypass grafting Total chest tube drain volume by days from first loading dose to coronary artery bypass grafting by treatment. bleeding and ischemic complications. However, importantly, early surgical revascularization was not associated with increased all-cause mortality through 30 days. In the ACCOAST study, isolated CABG surgery b30 days from index event was performed in 7.8% of patients with NSTEMI. This rate is similar to that reported in the large multicenter registry of patients with NSTEMI. 16 Also, a large discrepancy in timing of CABG related to region of enrollment and preoperative level of troponin was confirmed. Differences were seen in the timing of CABG with respect to withdrawal of prasugrel, with patients in Eastern Europe showing a greater delay to CABG. The median time from randomization of 5.48 days to CABG in the ACCOAST study was longer than reported for US centers participating in the National Cardiovascular Data Registry ACTION Registry GWTG. 16 In that registry, the median time from arrival to CABG for NSTEMI patients was 72.9 hours (about 3 days). It may suggest the difference in qualification and treatment protocols for patients with acute MI scheduled for CABG in different regions and centers. Still, the optimal timing of CABG after a NSTEMI remains a matter of debate. Previous studies and registries have clearly confirmed the association of a higher acuity of illness or an increased preoperative troponin level with less favorable outcomes of patients with acute MI undergoing CABG. 13 On the other hand, the benefit of delaying cardiac surgery in such patients is not well established. In our study, baseline level of troponin was strongly associated with timing of CABG, but no impact of preoperative troponin level on ischemic and bleeding outcomes was confirmed. Despite lack of evident differences in risk profile between patients undergoing surgery early and late, the risk of bleeding and ischemic complications was higher among patients operated early (b2.98 days). In contrast, no difference in ischemic outcomes, including death between patients with NSTEMI undergoing CABG early ( 48 hours) vs late (N48 hours), was confirmed in the ACTION Registry GWTG. 16 Furthermore, delayed surgery was associated with a longer hospital stay and likely a higher resource use. 16 In line with the main results of the ACCOAST study, there was no impact of prasugrel pretreatment on the occurrence of ischemic events in patients undergoing CABG. However, in a post-hoc analysis of the TRITON- TIMI-38 study, administration of prasugrel as compared with clopidogrel administration was associated with significant reduction of all-cause mortality in patients with ACS scheduled to undergo CABG. 4 Similar benefit of ticagrelor 5,12 and abciximab 17 in patients treated with CABG was confirmed. The mortality benefit of prasugrel and ticagrelor was observed primarily in the first 30 days after CABG, 4,5,12 This may suggest that antiplatelet therapy may reduce the risk for early vein graft failure, which is predominantly thrombosis related, thus reducing the risk of ischemic events. 18 On the other hand, detailed analysis of the PLATO study confirmed that mortality reduction observed in ticagrelor-treated patients was not related to ischemic events, but rather to reduction of the risk for bleeding-related and infectionsrelated CABG deaths as compared with clopidogreltreated patients. 12 In addition, other studies have confirmed an increased risk of infections after CABG in patients receiving clopidogrel. 19,20 The mechanism by which clopidogrel may increase the risk of infections and infection-related deaths is not yet clearly defined. Increased risk of infections may be directly related to increased risk of bleedings and subsequent consequences (ie, transfusions, prolonged drainage tube placement, and prolonged length of hospitalization). This unfavorable effect of clopidogrel was somewhat confirmed by the results of our study, as the ischemic event rates for no-pretreatment groups were much lower than reported for clopidogrel in both the TRITON-TIMI-38 and the PLATO studies. Low events rate in the comparator group and short follow-up may account for lack of mortality benefit of prasugrel pretreatment observed in our study. As confirmed by the ACCOAST study, pretreatment with prasugrel as compared with no-pretreatment increases the risk of bleedings in NSTEMI patients. 15 This analysis confirms that increase in bleeding related to prasugrel pretreatment is also possible in patients treated with CABG. However, prasugrel pretreatment was not identified as an independent predictor of TIMI major bleeding in our patients. The risk of bleeding was high in patients operated early, especially b2.98 days in both prasugrel pretreated and non-pretreated patients. Importantly, a recent study from Drews et al confirmed that preoperative therapy with prasugrel (as compared to clopidogrel) and urgent/emergent surgery were predictors for the need for perioperative platelet transfusion and/or surgical re-exploration. 21 In patients operated

7 American Heart Journal Volume 170, Number 5 Dudek et al 1031 Table IV. Primary and secondary end points through 30 days from randomization for patients undergoing CABG only in the ACCOAST study stratified by prasugrel pretreatment strategy and timing of CABG Variable All (n = 314) Prasugrel pretreatment (n = 157) Prasugrel no-pretreatment (n = 157) P Primary composite components CVD/MI/stroke/UR/GP IIb/IIIa bailout (%) Time from randomization to CABG b2.98 d and b6.95 d NE 6.95 d NE All TIMI major Bleed Time from Randomization to CABG b2.98 d and b6.95 d d NE All TIMI minor Bleed Time from Randomization to CABG b2.98 d and b6.95 d NE 6.95 d NE Abbreviations: CVD, Cardiovascular death; GP, glycoprotein; MI, myocardial infarction; NE, not evaluable; UR, urgent revascularization. Two-sided P value based on the log rank test. early, the risk of bleeding may be affected also by the use of glycoprotein IIb/IIIa inhibitors and antithrombin agents in the acute phase of ACS. On the other hand, overall rate of TIMI major bleeding in patients undergoing CABG 6.95 days was very low. This confirms the guidelines recommendation of postponing CABG and discontinuing P2Y 12 inhibitors 5 to 7 days before surgery, if possible. 1 3 On the other hand, delaying surgery in patients with NSTEMI might increase resource use and increase the risk of recurrent ischemia/mi in patients waiting for surgical revascularization. Despite a higher risk of bleeding and a high-risk profile, including presence of left main disease in N50% of patients, the short-term mortality for patients undergoing surgery b2.98 days was comparable to that observed for other groups. Thus, even in the cases where left main disease or other compelling anatomical aspects are noted, in spite of presenting enzyme leak, the mortality after CABG in patients with NSTEMI remains low. Results of emergency CABG for patients presenting with NSTEMI can be compared with those of elective revascularization, even in patients with left main disease. 16,22 25 Limitations Our study has several limitations. This was a post hoc, prespecified analysis of the large, randomized, prospective ACCOAST study. We cannot exclude a possibility of selection bias, as the decision to perform CABG and the timing of CABG were at the discretion of the investigators. Sample size of the study was calculated to prove the difference in outcomes in overall and PCI study populations, but not for CABG population. Thus, the sample size was too small for adequate assessment of clinical events in CABG cohort of the ACCOAST study. A 30-mg loading dose of prasugrel was used for pretreatment, which is not currently recommended dose for the treatment of NSTEMI. However, the pharmacodynamic results for a 30-mg loading dose of prasugrel are similar to results seen in patients at steady state on a 10-mg maintenance dose. 15 In addition, some important data on CABG technique (on- vs off-pump coronary artery bypass), as well as peri- and post-procedural medications, especially open label clopidogrel/ticlopidine were not collected. Data on anatomical (SYNTAX Score) and clinical (EuroSCORE II, Society of Thoracic Surgeons) risk scores were not collected. Thus, we were unable to adjust outcome results for these risk scores. Despite these limitations, reported results are clinically important and may affect current management of patients presenting with NSTEMI proceeding to CABG and pretreated with a P2Y 12 inhibitor. On the other hand, nowadays, pre-treatment with prasugrel in patients with NSTEMI in whom coronary anatomy is not known, is not recommended. 2 Conclusions In this group of high-risk patients presenting with NSTEMI, early (b2.98 days) surgical revascularization carried an increased risk of bleeding and ischemic complications, but without impact on all-cause mortality through 30 days. No impact of baseline troponin level and prasugrel pretreatment (important factors influencing time of CABG) on ischemic clinical outcomes was confirmed. Acknowledgements The trial was sponsored by Daiichi Sankyo, Ltd, and Eli Lilly and Company.

8 1032 Dudek et al American Heart Journal November 2015 References 1. Jneid H, Anderson JL, Wright RS, et al ACCF/AHA focused update of the guideline for the management of patients with unstable angina/non-st-elevation myocardial infarction (updating the 2007 guideline and replacing the 2011 focused update): a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2012;60: Windecker S, Kolh P, Alfonso F, et al ESC/EACTS Guidelines on myocardial revascularization: The Task Force on Myocardial Revascularization of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS) Developed with the special contribution of the European Association of Percutaneous Cardiovascular Interventions (EAPCI). Eur Heart J 2014;35: Sousa-Uva M, Storey R, Huber K, et al. Expert position paper on the management of antiplatelet therapy in patients undergoing coronary artery bypass graft surgery. Eur Heart J 2014;35: doi.org/ /eurheartj/ehu Smith PK, Goodnough LT, Levy JH, et al. Mortality benefit with prasugrel in the TRITON-TIMI 38 coronary artery bypass grafting cohort: risk-adjusted retrospective data analysis. J Am Coll Cardiol 2012;60: Held C, Asenblad N, Bassand JP, et al. Ticagrelor versus clopidogrel in patients with acute coronary syndromes undergoing coronary artery bypass surgery: results from the PLATO (Platelet Inhibition and Patient Outcomes) trial. J Am Coll Cardiol 2011;57: dx.doi.org/ /j.jacc Mehta RH, Roe MT, Mulgund J, et al. Acute clopidogrel use and outcomes in patients with non-st-segment elevation acute coronary syndromes undergoing coronary artery bypass surgery. J Am Coll Cardiol 2006;48: Berger JS, Frye CB, Harshaw Q, et al. Impact of clopidogrel in patients with acute coronary syndromes requiring coronary artery bypass surgery: a multicenter analysis. J Am Coll Cardiol 2008;52: Fox KA, Mehta SR, Peters R, et al. Benefits and risks of the combination of clopidogrel and aspirin in patients undergoing surgical revascularization for non-st-elevation acute coronary syndrome: the Clopidogrel in Unstable angina to prevent Recurrent ischemic Events (CURE) Trial. Circulation 2004;110: Wiviott SD, Braunwald E, McCabe CH, et al. Prasugrel versus clopidogrel in patients with acute coronary syndromes. N Engl J Med 2007;357: Wallentin L, Becker RC, Budaj A, et al. Ticagrelor versus clopidogrel in patients with acute coronary syndromes. N Engl J Med 2009;361: Komosa A, Lesiak M, Siniawski A, et al. Significance of antiplatelet therapy in emergency myocardial infarction treatment. Postep Kardiol Inter 2014;10: Varenhorst C, Alstrom U, Scirica BM, et al. Factors contributing to the lower mortality with ticagrelor compared with clopidogrel in patients undergoing coronary artery bypass surgery. J Am Coll Cardiol 2012;60: Caceres M, Weiman DS. Optimal timing of coronary artery bypass grafting in acute myocardial infarction. Ann Thorac Surg 2013;95: Montalescot G, Bolognese L, Dudek D, et al. A comparison of prasugrel at thetimeofpercutaneouscoronaryinterventionoraspretreatmentatthe time of diagnosis in patients with non-st-segment elevation myocardial infarction: design and rationale for the ACCOAST study. Am Heart J 2011;161: Montalescot G, Bolognese L, Dudek D, et al. Pretreatment with prasugrel in non-st-segment elevation acute coronary syndromes. N Engl J Med 2013;369: Parikh SV, de Lemos JA, Jessen ME, et al. Timing of in-hospital coronary artery bypass graft surgery for non-st-segment elevation myocardial infarction patients results from the National Cardiovascular Data Registry ACTION Registry-GWTG (Acute Coronary Treatment and Intervention Outcomes Network Registry-Get With The Guidelines). JACC Cardiovasc Interv 2010;3: Lincoff AM, LeNarz LA, Despotis GJ, et al. Abciximab and bleeding during coronary surgery: results from the EPILOG and EPISTENT trials. Improve Long-term Outcome with abciximab GP IIb/IIIa blockade. Evaluation of Platelet IIb/IIIa Inhibition in STENTing. Ann Thorac Surg 2000;70: Nocerino AG, Achenbach S, Taylor AJ. Meta-analysis of effect of single versus dual antiplatelet therapy on early patency of bypass conduits after coronary artery bypass grafting. Am J Cardiol 2013; 112: Blasco-Colmenares E, Perl TM, Guallar E, et al. Aspirin plus clopidogrel and risk of infection after coronary artery bypass surgery. Arch Intern Med 2009; 169: Blais DM, Zukkoor SM, Hayes III C, et al. Bleeding outcomes associated with coronary artery bypass graft surgery and recent clopidogrel exposure. Heart Surg Forum 2013;16:E doi.org/ /hsf Drews S, Bolliger D, Kaiser C, et al. Prasugrel increases the need for platelet transfusions and surgical reexploration rates compared with clopidogrel in coronary artery bypass surgery. Thorac Cardiovasc Surg 2015;63: Khaladj N, Bobylev D, Peterss S, et al. Immediate surgical coronary revascularisation in patients presenting with acute myocardial infarction. J Cardiothorac Surg 2013;8: / Morice MC, Serruys PW, Kappetein AP, et al. Outcomes in patients with de novo left main disease treated with either percutaneous coronary intervention using paclitaxel-eluting stents or coronary artery bypass graft treatment in the Synergy Between Percutaneous Coronary Intervention with TAXUS and Cardiac Surgery (SYNTAX) trial. Circulation 2010;121: CIRCULATIONAHA Mohr FW, Morice MC, Kappetein AP, et al. Coronary artery bypass graft surgery versus percutaneous coronary intervention in patients with three-vessel disease and left main coronary disease: 5-year follow-up of the randomised, clinical SYNTAX trial. Lancet 2013;381: http: //dx.doi.org/ /s (13) Zalewska-Adamiec M, Bachorzewska-Gajewska H, Kralisz P, et al. Prognosis in patients with left main coronary artery disease managed surgically, percutaneously or medically: a long-term follow-up. Kardiol Pol 2013;71:

9 American Heart Journal Volume 170, Number 5 Dudek et al 1032.e1 Appendix Supplementary Figure Time from randomization to isolated CABG in ACCOAST.

10 1032.e2 Dudek et al American Heart Journal November 2015 Supplementary Table I. Baseline characteristics of patients enrolled in the ACCOAST study stratified by treatment modality Variable CABG only cohort (n = 314) PCI only cohort (n = 2,761) Medical management cohort (n = 932) P P Eastern Europe, (%) Western Europe/Other, (%) Age, median, (years) Age 75 years, (%) Female, (%) b.001 Weight b60 kg, (%) BMI, median (kg/m 2 ) Current smoker (%) Arterial hypertension, (%) Hypercholesterolemia, (%) Diabetes mellitus, (%) b.001 Peripheral artery disease, (%) Carotid artery disease, (%) Previous myocardial infarction, (%) Previous CABG, (%) b.001 b.001 Previous PCI, (%) Family history of CAD, (%) Previous stroke, (%) Congestive heart failure, (%) Atrial fibrillation, (%) CRUSADE Risk Score, median, (points) Very high CRUSADE Risk Score, (%) Abnormal creatinine, (%) GRACE score, median, (points) GRACE score 140, (%) Baseline troponin level, (%) b3 ULN ULN and b10 ULN ULN Left main disease, (%) b.001 b.001 Abbreviations: BMI, Body mass index; CAD, coronary artery disease. P value comparing CABG only to PCI only cohort reported from Pearson s Chi-square test when total number of patients with non-missing values of the variable in each category N10 and from Fisher s exact test otherwise. For continuous variables, P value is reported from an analysis of variance model. P value comparing CABG only to medical management cohort. Primarily Western Europe; also includes Israel and Canada.