On behalf of the TRILOGY ACS Investigators Prasugrel versus clopidogrel for patients with UA/NSTEMI medically managed after angiographic triage — Results from the TRILOGY ACS Trial Stephen D. Wiviott, MD, Harvey D. White, MB, ChB, DSc, E. Magnus Ohman, MB, ChB, Keith A. A. Fox, MB, ChB, Paul W. Armstrong, MD, Dorairaj Prabhakaran, MD, DM, MSc, Gail Hafley, MS, William E. Boden, MD, Christian Hamm, MD, Peter Clemmensen, MD, DMSc, Jose C. Nicolau, MD, PhD, Alberto Menozzi, MD, PhD, Witold Ruzyllo, MD, Petr Widimsky, MD, DSc, Ali Oto, MD, Jose Leiva-Pons, MD, Gregory Pavlides, MD, Matthew T. Roe, MD, MHS, and Deepak L. Bhatt, MD, MPH Identifier: NCT

Authors and Disclosures* The TRILOGY ACS Trial was funded by Eli Lilly & Company and Daiichi Sankyo. Stephen D. Wiviott—grant/research support from Eli Lilly & Company, AstraZeneca,Merck, Eisai; Consulting fees/honoraria from Eli Lilly & Company, Daiichi Sankyo, AstraZeneca, BMS, Sanofi-Aventis, Eisai. Petr Widimsky—consulting fees/honoraria from Lilly, Ali Raif Ilac Sanayi, Bayer, Daiichi Sankyo, Medtronic, Boehringer Ingelheim, Abbott, AstraZeneca, Sanofi. Deepak L. Bhatt—honoraria and travel expenses from the Duke Clinical Research Institute; serving as a board member for Medscape Cardiology, Boston VA Research Institute, Society of Chest Pain Centers; grant funding from Amarin, AstraZeneca, Bristol-Myers Squibb, Eisai, Ethicon, Medtronic, Sanofi-Aventis, The Medicines Company; payment for developing educational presentations from WebMD; serving on clinical trial steering committees for the Duke Clinical Research Institute; serving as an editor for the American College of Cardiology and chief medical editor for Slack Publications. Gail Hafley and Witold Ruzyllo—nothing to report. *For all other authors, see MT Roe et al, NEJM 2012

Angiography Background  The proportion of ACS (UA/NSTEMI) patients worldwide who are managed medically without revascularization (PCI or CABG) is 40–60%. This includes 2 distinct sets of patients: triaged to medical therapy after angiography triaged to medical therapy after angiography for whom angiography is not performed for whom angiography is not performed  Prasugrel, a thienopyridine P2Y 12 inhibitor, improved ischemic outcomes in ACS patients undergoing PCI in the TRITON-TIMI 38 trial, with an increase in major bleeding. Wiviott SD et al NEJM 2007

Inclusion Criteria (Main Trial)  Randomization within 10 days of a UA/NSTEMI event NSTEMI: CK-MB or troponin > ULN NSTEMI: CK-MB or troponin > ULN UA: ST depression > 1 mm in 2 or more leads UA: ST depression > 1 mm in 2 or more leads  Medical management strategy decision determined  Angiography not required, but if performed, had to be done before randomization, and evidence of coronary diseasein a major vessel (1 lesion > 30% or prior PCI/CABG)  At least 1 of 4 enrichment criteria: Age > 60 years Age > 60 years Diabetes mellitus Diabetes mellitus Prior MI Prior MI Prior revascularization (PCI or CABG) Prior revascularization (PCI or CABG)

Study Design Roe Mt et al NEJM All patients were on aspirin, and low-dose aspirin (< 100 mg) was strongly recommended. For patients < 60 kg or ≥ 75 years, 5 mg MD of prasugrel was given. Adapted from Chin CT et al. Am Heart J 2010;160:16-22.e1. Medically Managed UA/NSTEMI Patients Clopidogrel 1 75 mg MD Clopidogrel 1 75 mg MD Prasugrel 1 5 or 10 mg MD Minimum Rx Duration: 6 months; Maximum Rx Duration: 30 months Primary Efficacy Endpoint: CV Death, MI, Stroke Randomization Stratified by: Age, Country, Prior Clopidogrel Treatment (Primary analysis cohort — Age < 75 years) Randomization Stratified by: Age, Country, Prior Clopidogrel Treatment (Primary analysis cohort — Age < 75 years) Clopidogrel mg LD + 75 mg MD Clopidogrel mg LD + 75 mg MD Prasugrel 1 30 mg LD + 5 or 10 mg MD Prasugrel 1 30 mg LD + 5 or 10 mg MD Medical Management Decision ≤ 72 hrs (No prior clopidogrel given) — 4% of total Medical Management Decision ≤ 72 hrs (No prior clopidogrel given) — 4% of total Medical Management Decision ≤ 10 days (Clopidogrel started ≤ 72 hrs in-hospital OR on chronic clopidogrel) — 96% of total Medical Management Decision ≤ 10 days (Clopidogrel started ≤ 72 hrs in-hospital OR on chronic clopidogrel) — 96% of total Median Time to Enrollment = 4.5 Days 9326 patients in 8 regions, 52 countries (Primary: 7243 patients < 75 years old)

Primary Efficacy Endpoint and TIMI Major Bleeding Through 30 Months (Age < 75 years; 7243) HR (95% CI): 0.91 (0.79, 1.05) P = 0.21 HR (95% CI): 1.31 (0.81, 2.11) P = 0.27 Endpoint (%) Roe MT et al NEJM 2012

Overall TRILOGY ACS Results: Summary (Age < 75 years)  No statistical differences in cardiovascular events or major bleeding  Lower risk multiple recurrent ischemic events suggested with prasugrel using the pre-specified Andersen-Gill model (HR = 0.85, 95% CI: 0.72–1.00, P = 0.04)  Significant interaction with treatment and time (HR for > 12 mos = 0.64, 95% CI: 0.48–0.86, Interaction P = 0.02) Roe MT et al NEJM 2012

Objectives of TRILOGY-ACS Prespecified Angiography Sub-Study  Within the TRILOGY ACS trial, to: Assess clinical characteristics and outcomes of subjects triaged to medical therapy with or without preceding angiography Assess clinical characteristics and outcomes of subjects triaged to medical therapy with or without preceding angiography Assess effects of prasugrel vs. clopidogrel in these two groups and whether there is any differential effect based on how subjects entered the trial Assess effects of prasugrel vs. clopidogrel in these two groups and whether there is any differential effect based on how subjects entered the trial

Statistical Considerations: Angiographic Cohort  Pre-specified analysis limited to the primary cohort of patients < 75 years of age (N=7243)  Stratified by pre-randomization variable Triaged after angiography: 3085 (43%) Triaged after angiography: 3085 (43%) Triaged without angiography: 4158 (57%) Triaged without angiography: 4158 (57%)  For angiography vs no angiography comparisons — p-values unadjusted  For prasugrel vs clopidogrel — p-value adjusted for clopidogrel stratum

ITT Population N = 9326 Primary Population Age < 75 N = 7243 Triaged after Angiography* N = 3085 (43%) Randomized to Prasugrel N = 1524 Randomized to Clopidogrel N = 1561 Trigaged without Angiography* N = 4158 (57%) Randomized to Prasugrel N = 2096 Randomized to Clopidogrel N = 2062 Age ≥ 75 N = 2083 Prespecified Analysis of Angiographic Cohort *For angiography vs no angiography comparisons — p-values unadjusted, for prasugrel vs clopidogrel — p-value adjusted for clopidogrel stratum

Baseline Characteristics Age < 75 Years (N = 7243) Angiography (N = 3085) No Angio (N = 4158) P-Value Age—yr62 (56–68) 63 (57–68) Female sex—% < Body weight < 60 kg—% < Disease classification—% < NSTEMI Unstable angina Medical history—% Diabetes mellitus Current/recent smoking < Prior myocardial infarction Prior PCI < Prior CABG2111.3< Baseline risk assessment GRACE risk score112 (99–124)117 (102–131)< Creatinine clearance—mL/min86 (68–109)77 (59–98)<

Regional Differences in Angiography Pre-randomization

Baseline Characteristics: Angiographic Results (>50% Stenosis) Notes: 1. Non-obstructive = 30 - <50% stenosis 2. LM disease - 6.2% of subjects

Incidence of Outcomes by Angiography Status (Age < 75 years) P < P = 0.09

Primary Efficacy Endpoint to 30 Months (Age < 75 years) P interaction = % vs 14.9% P = HR (95% CI): 0.77 (0.61, 0.98) Angio N=3085 No Angio N= % vs 16.7% P = HR (95% CI): 1.01 (0.84, 1.20)

Evaluation of All Ischemic Events over Time* (Age < 75 years) AngiographyNo Angiography PrasClopPrasClop ≥ 1 event ≥ 2 events –7 events HR (CI) 0.75 (0.58–0.98) 0.91 (0.74–1.11)  Lower risk of multiple recurrent ischemic events suggested with prasugrel in the angiography group using the pre-specified Andersen-Gill model * Pre-specified evaluation of all CV death, MI, or stroke events by treatment P interaction = 0.26

P interaction = % vs 10.3% P = HR (95% CI): 0.74 (0.55, 1.00) Angio No Angio 9.2% vs 10.6% P = HR (95% CI): 1.00 (0.79, 1.26) Myocardial Infarction

Stroke P interaction = % vs 2.4% P = HR (95% CI): 0.30 (0.13,0.71) Angio No Angio 2.2% vs 2.0% P = HR (95% CI): 1.03 (0.58,1.83)

P interaction = % vs 5.2% P = HR (95% CI): 0.91 (0.61,1.34) Angio No Angio 8.4% vs 7.9% P = HR (95% CI): 0.93 (0.73,1.20) CV Death

P interaction = % vs 5.2% P = HR (95% CI): 0.91 (0.61,1.34) Angio No Angio 8.4% vs 7.9% P = HR (95% CI): 0.93 (0.73,1.20) CV Death P interaction = 0.85 Angio: HR (95% CI): 0.98 (0.69,1.38) No Angio: HR (95% CI): 0.94 (0.75,1.18) All-Cause Death

TIMI Major Bleeding P interaction = % vs 1.4% P = HR (95% CI): 1.84 (0.93, 3.63) Angio No Angio 1.6% vs 1.5% P = HR (95% CI): 0.92 (0.47, 1.83)

TIMI Major Bleeding P interaction = % vs 1.4% P = HR (95% CI): 1.84 (0.93, 3.63) Angio No Angio 1.6% vs 1.5% P = HR (95% CI): 0.92 (0.47, 1.83) P interaction = 0.65 Angio: HR (95% CI): 1.68 (1.00, 2.83) No Angio: HR (95% CI): 1.42 (0.83, 2.41) TIMI Major or Minor Bleeding

Limitations  Reason for pursuing strategy not fully known: Angiography Angiography Medical Therapy Medical Therapy  Cannot make direct comparisons between angiography and no angiography and infer causality  Subgroup analysis of a neutral trial Prespecified Prespecified Pre-randomization variable Pre-randomization variable Large sample size Large sample size  Interaction testing is underpowered

Conclusions  Substantial differences in baseline characteristics exist among patients triaged for medical therapy with or without angiography from TRILOGY-ACS. Geographically, subjects from North America, Western Europe, Australasia, South Africa, and the Mediterranean region had higher rates of angiography pre-randomization Geographically, subjects from North America, Western Europe, Australasia, South Africa, and the Mediterranean region had higher rates of angiography pre-randomization Patients with angiography more often were enrolled with NSTEMI, and had prior history of PCI or CABG Patients with angiography more often were enrolled with NSTEMI, and had prior history of PCI or CABG  Patients with angiography had lower composite endpoint event (CVDeath, MI, or Stroke), particularly CV death.

Conclusions  Overall, in the TRILOGY ACS Trial prasugrel did not reduce cardiovascular events among patients managed medically for ACS.  When treated with prasugrel compared to clopidogrel, patients triaged to medical therapy following angiography tended to have: Lower rates of the combined endpoint of CVD/MI/CVA Lower rates of the combined endpoint of CVD/MI/CVA Lower rates of MI, CVA alone, and recurrent ischemic events Lower rates of MI, CVA alone, and recurrent ischemic events A trend to higher rates of TIMI major bleeding. A trend to higher rates of TIMI major bleeding.  Though hypothesis generating, these results are consistent with previous trials and suggest that when angiography is performed and coronary disease is confirmed, the benefits and risks of intensive antiplatelet therapy exist whether medical therapy or PCI is elected.