CI-1 The effect of cangrelor versus clopidogrel on periprocedural outcomes in a pooled analysis of patient-level data Christian W. Hamm for the CHAMPION Executive Committees and Investigators Christian W. Hamm for the CHAMPION Executive Committees and Investigators

CI-2 Disclosures: C. Hamm Honoraria for lectures: AstraZeneca, Bayer, Boehringer Ingelheim, Daiichi Sankyo, Essex, GSK, Lilly, Sanofi Aventis, Correvio, Pfizer, Roche, The Medicines Company. Honoraria for advisory board activities: AstraZeneca, Bayer, and Boehringer Ingelheim. This presentation includes off-label and/or investigational uses of drugs, including clopidogrel and cangrelor. The CHAMPION PHOENIX, PLATFORM and PCI trials were funded by The Medicines Company.

CI-3 Antiplatelet Therapy during PCI ► Antiplatelet therapy is a critical part of PCI. ► In the era of aspirin and unfractionated heparin, intravenous glycoprotein IIb/IIIa inhibition reduced ischaemic events, but significantly increased bleeding. ► ADP receptor (P2Y 12 ) antagonism with oral agents was also shown to reduce ischaemic events in PCI and especially ACS. ► However, available oral agents are limited by long duration of action and bioavailability, which might be a liability:  if given prior to coronary angiography and emergent CABG is necessary,  in situations where absorption may be problematic,  in patients who are intubated, nauseated, or shock. Harrington RA, et al. PURSUIT. NEJM 1998 Desai N and Bhatt DL. Periprocedural Antiplatelet Therapy. JACC Intervention 2010

CI-4 Cangrelor ► Cangrelor is an intravenous ADP receptor antagonist with fast-onset, potent, and reversible P2Y 12 inhibition, and a half-life of 3 to 6 minutes. Cangrelor has been studied in >25,000 PCI patients in three trials: CHAMPION PHOENIX CHAMPION PLATFORM CHAMPION PCI ► Neither CHAMPION PLATFORM nor CHAMPION PCI met its primary endpoint but evidence of cangrelor effect was demonstrated on ST, n IDR, Q-MI, and in CHAMPION PLATFORM on mortality. There was no excess in GUSTO severe or moderate bleeding. ► The potential efficacy signals led to the CHAMPION PHOENIX trial which demonstrated a 21% reduction in thrombotic events compared with clopidogrel (p=0.0055). Harrington RA, et al. CHAMPION PCI. NEJM 2009 Bhatt DL, et al. CHAMPION PLATFORM. NEJM 2009 Bhatt DL, et al. CHAMPION PHOENIX. NEJM 2013 White HD, et al. Meta-Analysis of CHAMPION PCI and PLATFORM. AHJ 2012

CI-5 Study Objectives Prespecified, pooled analysis of patient-level data of:  CHAMPION-PHOENIX,  CHAMPION-PLATFORM, and  CHAMPION-PCI comparing cangrelor with clopidogrel control for prevention of thrombotic complications during and after PCI. Harrington RA, et al. CHAMPION PCI. NEJM 2009 Bhatt DL, et al. CHAMPION PLATFORM. NEJM 2009 Bhatt DL, et al. CHAMPION PHOENIX. NEJM 2013 White HD, et al. Meta-Analysis of CHAMPION PCI and PLATFORM. AHJ 2012

CI-6 The CHAMPION Executive Committees Deepak L. Bhatt, M.D., M.P.H. (Co-Principal Investigator) VA Boston Healthcare System, Brigham and Women's Hospital, and Harvard Medical School Boston, MA Robert A. Harrington, M.D. (Co-Principal Investigator) Department of Medicine, Stanford University, Stanford, CA C. Michael Gibson, M.S., M.D. Beth Israel Deaconess Medical Center, Division of Cardiology, Boston, MA Christian W. Hamm, M.D. Kerckhoff Heart and Thorax Center, Bad Nauheim, Germany Kenneth W. Mahaffey, M.D. Duke Clinical Research Institute, Durham, NC Matthew J. Price, M.D. Scripps Clinic and Scripps Translational Science Institute, La Jolla, CA Ph. Gabriel Steg, M.D. INSERM U-698, Université Paris-Diderot, and Hôpital Bichat, Assistance-Publique- Hôpitaux de Paris, Paris, France Gregg W. Stone, M.D. Columbia University Medical Center and the Cardiovascular Research Foundation, New York, NY Harvey D. White, D.Sc. Auckland City Hospital, Auckland, New Zealand

CI-7 The CHAMPION Executive Committees (continued) A. Michael Lincoff, MD Cleveland Clinic, Cleveland, OH Charles Pollack, MD Pennsylvania Hospital, University of PA Health System, Philadelphia, PA Neal Kleiman, MD Methodist Hospital, Houston, TX Gilles Montalescot, MD Groupe Hospitalier Pitie-Salpetriere, Paris, France Shaun Goodman, MD Canadian Heart Research Centre, Toronto, Canada

CI-8 The CHAMPION DSMB Frans Van de Werf, M.D. (Chair) Universitair Ziekenhuis Gasthuisberg, Belgium David P. Faxon, M.D. Brigham & Women’s Hospital, Dept. of Medicine, Boston, MA E. Magnus Ohman, M.D. Duke University Medical Center, Durham, NC Freek W.A. Verheugt, M.D. Heartcenter University Medical Center, Amsterdam W. Douglas Weaver, M.D. Henry Ford Hospital, Detroit, MI Jan G.P. Tijssen, Ph.D. (Statistician) Department of Cardiology, Academic Medical Center-University of Amsterdam, The Netherlands

CI-9 CHAMPION Trials Study Designs 12 hours0 PCI ~30’ Cangrelor bolus then infusion Clopidogrel 600 mg oral CHAMPION PHOENIX n=10,942 mITT SA / NSTE-ACS / STEMI P2Y 12 naïve Placebo or clopidogrel before or after PCI CHAMPION PCI n=8667 mITT SA / NSTE-ACS / STEMI Placebo or clopidogrel before PCI CHAMPION PLATFORM n=5301 mITT SA / NSTE-ACS P2Y 12 naïve Placebo or clopidogrel after PCI Cangrelor bolus then infusion Clopidogrel 600 mg oral Clopidogrel 600 mg oral Clopidogrel 600 mg oral Clopidogrel 600 mg or 300 mg oral Clopidogrel 600 mg oral OR Randomised, Double Blind, Controlled Trials of patients undergoing PCI

CI-10 Study Endpoints ► Primary efficacy endpoint: Death/MI/IDR/ST at 48 hours  Modified Intent-to-Treat (mITT) analysis (patients actually got study drug and PCI) ► Secondary endpoints:  Stent Thrombosis at 48 hours  Death/MI/IDR at 48 hours ► MI was defined according to the universal definition ► Efficacy endpoints also examined at 30 days ► Primary safety endpoint: GUSTO Severe Bleeding at 48 hours MI, myocardial infarction; IDR, ischaemia-driven revascularization; ST, stent thrombosis Harrington RA, et al. CHAMPION PCI. NEJM 2009 Bhatt DL, et al. CHAMPION PLATFORM. NEJM 2009 Bhatt DL, et al. CHAMPION PHOENIX. NEJM 2013 Thygesen K, et al. Universal definition of myocardial infarction. Circulation 2007.

CI-11 Demographics, mITT Cangrelor (N= 12,475) Clopidogrel (N= 12,435) Age, years63 Male72% Diabetes mellitus29%30% Patient Type Stable angina31% NSTE-ACS57%58% STEMI11%12% Loading Dose* 300 mg clopidogrel11% 600 mg clopidogrel89% Timing of clopidogrel Before PCI start56% During PCI13%14% Within 1 hour post PCI31%30% *Figures pertain to clopidogrel placebo loaded according to stratification at the time of randomisation.

CI-12 Primary Efficacy Outcomes at 48 Hours, mITT Cangrelor (N=12,475) Clopidogrel (N=12,435) OR (95% CI)P-value Death/MI/IDR/ST 473/12,459 (3.8%) 579/12,422 (4.7%) 0.81 ( ) Secondary Efficacy Outcomes at 48 Hours, mITT Stent thrombosis 62/12,459 (0.5%) 105/12,422 (0.8%) 0.59 ( ) Death/MI/IDR 446/12,459 (3.6%) 543/12,422 (4.4%) 0.81 ( ) MI 387/12,459 (3.1%) 453/12,422 (3.6%) 0.85 ( ) Q-wave MI 19/12,459 (0.2%) 36/12,422 (0.3%) 0.53 ( ) IDR 66/12,459 (0.5%) 92/12,422 (0.7%) 0.71 ( ) Death 33/12,459 (0.3%) 45/12,422 (0.4%) 0.73 ( ) Steg GS, Bhatt DL, Hamm CW et al…. Harrington RA. Lancet 2013 at

CI-13 Death/ MI/ IDR/ Stent Thrombosis within 48 Hours No. patients at risk Hours from randomization Cangrelor:12,47512,05312,04012,03312,02112,00612,00211,99411,985 Clopidogrel:12,43511,90311,89711,89111,88211,87411,86611,85311,843 Clopidogrel Cangrelor Log-rank p value= % 4.7% Event rate (%) Steg GS, Bhatt DL, Hamm CW et al…. Harrington RA. Lancet 2013 at

CI-14 Stent Thrombosis within 48 Hours Steg GS, Bhatt DL, Hamm CW et al…. Harrington RA. Lancet 2013 at Event rate (%) No. patients at risk Hours from randomization Cangrelor:12,47512,42012,40612,40312,39512,38712,38412,37712,371 Clopidogrel:12,43512,32712,31912,31812,30812,30612,30412,29712,291 Log-rank p value= % 0.8% Clopidogrel Cangrelor

CI-15 Death/ MI/ IDR within 48 Hours Steg GS, Bhatt DL, Hamm CW et al…. Harrington RA. Lancet 2013 at Event rate (%) Cangrelor Clopidogrel Patient at risk Hours from randomization Cangrelor:12,47512,08112,06812,06112,04812,03312,02912,02112,012 Clopidogrel:12,43511,93911,93311,92711,91811,91011,90211,89011,879 Log-rank p value= % 3.6%

CI-16 Death / MI / IDR / STOR (95% CI) p value p for interaction PLATFORM 0.72 (0.53, 0.97) PCI 0.90 (0.72, 1.14) PHOENIX 0.79 (0.67, 0.93) Pooled 0.81 (0.71, 0.91) ST PLATFORM 0.31 (0.11, 0.85) PCI 0.73 (0.33, 1.59) PHOENIX 0.62 (0.43, 0.90) Pooled 0.59 (0.43, 0.80) Death / MI / IDR PLATFORM 0.72 (0.53, 0.97) PCI 0.90 (0.72, 1.14) PHOENIX 0.80 (0.67, 0.95) Pooled 0.81 (0.71, 0.92) Death / QMI / IDR PLATFORM 0.55 (0.33, 0.93) PCI 0.66 (0.42, 1.05) PHOENIX 0.76 (0.53, 1.11) Pooled 0.68 (0.52, 0.87) Summary of Clinical Efficacy: Pooled Analysis Comparator betterCangrelor better Steg GS, Bhatt DL, Hamm CW et al…. Harrington RA. Lancet 2013 at

OR [95% CI]P [Int] Overall0.81 (0.71, 0.91) Age >= (0.52, 0.91)0.219 Age < (0.73, 0.96) Male0.85 (0.74, 0.99)0.18 Female0.71 (0.56, 0.89) Ethnicity: White0.80 (0.70, 0.92)0.683 Ethnicity: Non-white0.87 (0.60, 1.28) United States0.82 (0.68, 0.98)0.816 ROW0.80 (0.67, 0.94) Stable Angina0.77 (0.64, 0.93)0.866 NSTE-ACS0.82 (0.68, 0.99) STEMI0.84 (0.55, 1.27) Weight >= (0.71, 0.92)0.927 Weight < (0.52, 1.30) Biomarker Positive0.79 (0.62, 1.02)0.962 Biomarker Negative0.79 (0.68, 0.91) Diabetic Yes0.87 (0.69, 1.09)0.446 Diabetic No0.78 (0.67, 0.91) Insulin-Dependent Diabetes: Yes0.63 (0.41, 0.98)0.247 Insulin-Dependent Diabetes: No0.83 (0.72, 0.94) Prior MI0.72 (0.57, 0.92)0.25 No Prior MI0.85 (0.74, 0.99) Subgroups: Death, MI, IDR, Stent Thrombosis Clopidogrel betterCangrelor better

OR [95% CI]P [Int] Prior TIA/Stroke0.80 (0.48, 1.34)0.972 No Prior TIA/stroke0.81 (0.71, 0.92) History of PAD0.57 (0.39, 0.86)0.082 No PAD0.83 (0.73, 0.95) History of CHF0.68 (0.46, 1.02)0.375 No history of CHF0.83 (0.72, 0.94) Clopidogrel 300 mg0.84 (0.62, 1.14)0.772 Clopidogrel 600 mg0.80 (0.70, 0.92) Bivalirudin0.83 (0.65, 1.06)0.622 Heparin0.77 (0.66, 0.90) # vessels =10.84 (0.73, 0.96)0.214 # vessels ≥20.69 (0.53, 0.90) Drug-Eluting Stent0.85 (0.72, 1.00)0.231 Bare Metal Stent0.72 (0.59, 0.89) Aspirin ≤100 mg0.83 (0.66, 1.04)0.689 Aspirin >100 mg0.79 (0.67, 0.92) Clopidogrel Load before PCI Start0.83 (0.70, 0.97)0.735 Clopidogrel Load after PCI Start0.79 (0.65, 0.97) Cangrelor infusion ≤ 129 minutes0.82 (0.70, 0.96)0.753 Cangrelor infusion > 129 minutes0.79 (0.65, 0.96) Prior Thienopyridine Use0.80 (0.55, 1.16)0.948 No Prior Thienopyridine Use0.81 (0.71, 0.92) GP IIb/IIIa Use at Peri-procedural0.74 (0.55, 1.01)0.552 No GP IIb/IIIa Use at Peri-procedural0.82 (0.72, 0.94) Subgroups: Death, MI, IDR, Stent Thrombosis (continued) Clopidogrel betterCangrelor better

CI-19 Efficacy Outcomes at 30 Days, MITT Cangrelor (N=12,407) Clopidogrel (N=12,357) OR (95% CI)P Value Death/MI/IDR/ST (primary endpoint) 657/12,407 (5.3%) 748/12,357 (6.1%) 0.87 ( ) Stent thrombosis 113/12,407 (0.9%) 162/12,357 (1.3%) 0.69 ( ) Death/MI/IDR 631/12,407 (5.1%) 710/12,357 (5.7%) 0.88 ( ) MI 418/12,407 (3.4%) 487/12,357 (3.9%) 0.85 ( ) Q-wave MI 31/12,407 (0.2%) 51/12,357 (0.4%) 0.60 ( ) IDR 153/12,407 (1.2%) 178/12,357 (1.4%) 0.85 ( ) Death 137/12,407 (1.1%) 141/12,357 (1.1%) 0.97 ( ) Steg GS, Bhatt DL, Hamm CW et al…. Harrington RA. Lancet 2013 at

CI-20 Non-CABG Bleeding at 48 Hours, Safety Bleeding Scale Cangrelor (N=12,565) Clopidogrel (N=12,5425) OR (95% CI)P Value GUSTO Severe28 (0.2%)23 (0.2%)1.22 (0.70, 2.11) GUSTO Moderate76 (0.6%)56 (0.4%)1.36 (0.96, 1.92) GUSTO Severe + Moderate 103 (0.8%)79 (0.6%)1.30 (0.97, 1.75) TIMI Major32 (0.3%)28 (0.2%)1.14 (0.69, 1.90) TIMI Minor77 (0.6%)51 (0.4%)1.51 (1.06, 2.15) TIMI Major + Minor109 (0.9%)79 (0.6%)1.38 (1.03, 1.85) Any Blood Transfusion90 (0.7%)70 (0.6%)1.29 (0.94, 1.76) ACUITY Major534 (4.2%)353 (2.8%)1.53 (1.34, 1.76)< ACUITY w/out hematoma169 (1.3%)123 (1.0%)1.38 (1.09, 1.74) Steg GS, Bhatt DL, Hamm CW et al…. Harrington RA. Lancet 2013 at

CI-21 OR [95% CI]P [Int] Overall 1.30 (0.97, 1.75) Age >= (0.58, 1.51) Age < (1.09, 2.32) Male 0.97 (0.62, 1.51) Female 1.65 (1.11, 2.46) Ethnicity: White 1.21 (0.87, 1.67) Ethnicity: Non-white 1.83 (0.92, 3.62) United States 1.12 (0.76, 1.65) ROW 1.60 (1.01, 2.53) Stable Angina 0.91 (0.44, 1.88) NSTE-ACS 1.47 (1.02, 2.12) STEMI 1.20 (0.60, 2.41) Weight >= (0.88, 1.67) Weight < (0.95, 3.70) Biomarker Positive 1.40 (0.95, 2.08) Biomarker Negative 1.21 (0.77, 1.90) Diabetic No 1.56 (0.97, 2.52) Diabetic Yes 1.17 (0.80, 1.70) Insulin-Dependent Diabetes: Yes 2.65 (0.95, 7.38) Insulin-Dependent Diabetes: No 1.21 (0.89, 1.65) Prior MI 1.13 (0.53, 2.40) No Prior MI 1.34 (0.97, 1.84) Subgroups: GUSTO Severe/Moderate bleeding at 48 hours (Safety population) Clopidogrel betterCangrelor better

CI-22 OR [95% CI]P [Int] Prior TIA/Stroke0.69 (0.22, 2.17)0.251 No Prior TIA/stroke1.36 (1.01, 1.85) History of PAD1.55 (0.73, 3.30)0.647 No PAD1.28 (0.93, 1.77) History of CHF2.19 (0.94, 5.09)0.198 No history of CHF1.21 (0.88, 1.67) Clopidogrel 300 mg4.02 (1.13, 14.27)0.056 Clopidogrel 600 mg1.20 (0.88, 1.62) Bivalirudin1.14 (0.60, 2.15)0.780 Heparin1.26 (0.88, 1.81) # vessels =11.60 (1.15, 2.22)0.011 # vessels ≥20.59 (0.29, 1.20) Drug-Eluting Stent1.09 (0.74, 1.63)0.307 Bare Metal Stent1.52 (0.93, 2.51) Aspirin ≤100 mg1.17 (0.61, 2.24)0.538 Aspirin >100 mg1.48 (1.04, 2.10) Clopidogrel Load before PCI Start1.13 (0.75, 1.68)0.251 Clopidogrel Load after PCI Start1.61 (1.02, 2.54) Cangrelor infusion ≤ 129 minutes1.42 (0.97, 2.07)0.692 Cangrelor infusion > 129 minutes1.25 (0.77, 2.03) Prior Thienopyridine Use0.82 (0.40, 1.66)0.506 No Prior Thienopyridine Use1.44 (1.04, 1.99) Peri-procedural GP IIb/IIIa Use0.80 (0.47, 1.35)0.022 No Peri-procedural GP IIb/IIIa Use1.68 (1.17, 2.42) Subgroups: GUSTO Severe/Moderate bleeding at 48 hours (continued) Clopidogrel betterCangrelor better

CI-23 Limitations ► The definition of myocardial infarction differed among the three trials, but was standardized for this analysis. ► Stent Thrombosis was adjudicated by angiographic core lab review in CHAMPION PHOENIX only. ► Follow-up is limited to 30 days as this was the only data available for CHAMPION PHOENIX. ► The comparator arm differed among the three trials  Different timing of administration of clopidogrel – before vs. after PCI  Different loading doses of clopidogrel - 300mg vs. 600mg  Different clopidogrel status - clopidogrel naïve vs. pretreated

CI-24 Conclusions ► In this pooled analysis of 25,000 undergoing PCI, intravenous ADP receptor antagonism with cangrelor significantly reduced the composite of death/MI/IDR/ST at 48 hours, with a 19% odds reduction (p=0.0007). ► The key secondary endpoint of stent thrombosis was also significantly reduced, with a 41% odds reduction (p=0.0008). ► The benefit was sustained through 30 days. ► There was no excess in GUSTO severe bleeding or transfusions. ► Intravenous cangrelor may be an attractive option across the full spectrum of PCI: stable angina, NSTEMI, STEMI.

CI-25 For full details, please go to: INSERT PDF OF TITLE and AUTHORS from LANCET PDF

CI-26 Thank you!

CI-27 Limitations ► The definition of myocardial infarction differed among the three trials.  The MI component used in the pooled analysis was aligned with the Universal Definition of MI which was protocol defined for CHAMPION PHOENIX and programmed post hoc for CHAMPION PLATFORM and CHAMPION PCI based on CEC adjudicated results.  As clinical practice has evolved, time from admission to angiography has shortened substantially, creating important challenges in the definition of recurrent MI in the setting of very early PCI for ACS. These problems were addressed by the UDMI.  These challenges may have confounded the MI outcome adjudication and potentially obscured the detection of treatment effects in CHAMPION PLATFORM and CHAMPION PCI prompting the design of CHAMPION PHOENIX. It appeared logical therefore to use the UDMI definition in the primary analysis. ► CHAMPION PLATFORM and PCI only reported ST related to IDR and did not use the ARC definition of stent thrombosis. Similarly, the collection of intraprocedural stent thrombosis was not done in those trials.

CI-28 Limitations ► Follow-up is limited to 30 days as this was the only data available for CHAMPION PHOENIX.  However, it is unlikely that the effect of a short-term infusion (typically 2 hours) would have significantly extended beyond 30 days as most of the effect of cangrelor in the three trials appears to emerge in the first 6 hours. ► The comparator arm differed among the three trials  Different timings of administration of clopidogrel – before vs. after PCI  Different loading doses of clopidogrel – 300mg vs. 600mg  Different clopidogrel status – clopidogrel naïve vs. pretreated

CI-29 Summary of Treatment Emergent Adverse Events Adverse Event Cangrelor (N=12,565) Clopidogrel (N=12,542) Patients with any AEs2,900 (23.1%)2,745 (21.9%) Patients with any SAEs281 (2.2%)270 (2.2%) Patients discontinuing due to AE74 (0.6%)51 (0.4%) Dyspnoea143 (1.1%)48 (0.4%) Patients who died298 (2.4%)323 (2.6%) Fatal bleeding † 8 (0.1%)9 (0.1%) Steg GS, Bhatt DL, Hamm CW et al…. Harrington RA. Lancet 2013 at AE=adverse event; SAE=serious adverse event. *p-value based on the Chi-square test. † Includes patients who died within 30 days from dosing start time.

CI-30 BACKUP SLIDES

CI-31 Cangrelor ► Direct platelet P2Y 12 receptor antagonist ► ATP analogue MW=800 Daltons ► Parenteral administration ► T1/2 = 3 to 6 minutes ► Offset = 60 minutes N N N N NH S CF 3 OHO H O O P O O P P O O O Cl O O O S 4Na + Angiolillo DJ, Schneider DJ, Bhatt DL, et al. Pharmacodynamic effects of cangrelor and clopidogrel: the platelet function substudy from the cangrelor versus standard therapy to achieve optimal management of platelet inhibition (CHAMPION) trials. J Thromb Thrombolysis 2012;34:44-55.

CI-32 CHAMPION PCI | PLATFORM ► PCI | all comers PCI | 58% ACS | on clopidogrel allowed | clopidogrel 600mg administered at the start of PCI in the control arm ► PLATFORM | all comers PCI | 65% ACS | clopidogrel naïve | clopidogrel 600mg administered at the end of PCI in the control arm 12 hours0 PCI ~25’ Cangrelor 30  g/kg then 4  g/kg/min CHAMPION PCI N = 9000 SA/UA/ACS/STEMI On clopidogrel allowed CHAMPION PLATFORM N = 6400 SA/UA/ACS No clopidogrel allowed Harrington RA, et al. NEJM 2009 Bhatt DL, et al. NEJM 2009 Clopidogrel 600 mg oral

CI-33 CHAMPION PHOENIX Lessons from CHAMPION PCI | PLATFORM Trial DesignImplementation Patient population All comers PCIInclusion of patients with normal cardiac markers at baseline | est. 65% trial population P2Y 12 inhibitor naïvePatients not pre-treated with P2Y 12 inhibitor within 7 days prior to angiogram Endpoint definitions MI definition 1 UDMI | Central lab to assure consistency of CKMB mass assay globally | angiographic core lab to consistently assess evidence of ischaemia Stent thrombosis definition 2 ARC Definition includes occurrence associated with IDR | Death | MI | also intra-procedural stent thrombosis measured by angiographic core lab 3 1. Thygesen K, Alpert JS, and White HD, on behalf of the Joint ESC/ACCF/WHF Task Force for the Redefinition of Myocardial Infarction. Universal definition of myocardial infarction. Eur Heart J. 2007;28: Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in coronary stent trials: a case for standardized definitions. Circulation 2007;115: Brener SJ, Cristea E, Kirtane AJ, et al. Intra-Procedural Stent Thrombosis. J Am Coll Cardiol Intv 2013;6:36–43.

CI-34 CHAMPION PHOENIX Study Design 1 Randomization occurred once suitability for PCI was confirmed either by angiography or STEMI diagnosis. Double blind study medication was administered as soon as possible following randomization. 2 Study drug Infusion (cangrelor or matching placebo) was continued for 2-4 hours at the discretion of the treating physician. At the end of the infusion patients received a loading dose of clopidogrel or matching placebo and were transitioned to maintenance clopidogrel therapy. 3 Clopidogrel loading dose (or matching placebo) was administered as directed by the investigator. At the time of patient randomization, a clopidogrel loading dose of 600 mg or 300 mg was specified by the investigator. MITT=modified intent-to-treat; NSTE-ACS=non-ST-elevation acute coronary syndrome; PCI=percutaneous coronary intervention; SA=stable angina; STEMI=ST-elevation MI. 12 to 4 hours0 Cangrelor 2 bolus & infusion (30ug/kg; 4ug/kg/min) Clopidogrel 600 mg oral CHAMPION PHOENIX N = 10,900 MITT SA/ NSTE-ACS/ STEMI Patients requiring PCI 1 P2Y 12 inhibitor naïve OR Placebo 3 oral (right before PCI or right after, per physician) Placebo 2 bolus & infusion Placebo oral PCI ~30’ OR Clopidogrel 3 (600 mg or 300 mg oral, per physician) Rand

CI-35 Universal Definition of MI Thygesen K, Alpert JS, and White HD, on behalf of the Joint ESC/ACCF/WHF Task Force for the Redefinition of Myocardial Infarction. Universal definition of myocardial infarction. Eur Heart J. 2007;28: Better discrimination of MI with consideration of multiple criteria CKMB elevations | ischemic symptoms | angiographic evidence | ECG changes Diagnosis of MI from various perspectives Type 1 | spontaneous MI related to ischaemia Type 2 | MI secondary to ischaemia | change in O 2 demand/supply Type 3 | sudden unexpected cardiac death Type 4 | associated with coronary angioplasty | stents Type 4a | MI associated with PCI Type 4b | MI associated with Stent Thrombosis Type 5 | MI associated with CABG

CI-36 Definition of ST Angiographic Evidence: ARC ST (Academic Research Consortium) 1 ► Acute (<24 hours post-procedure) ► Subacute stent thrombosis (>24 hours and ≤30 days)  Definite from probable stent thrombosis  Adjudicated by the CEC IPST (Intra-procedural stent thrombosis) ► New or worsening thrombus related to the stent or ► Abrupt closure due to thrombosis 1. Cutlip DE, Windecker S, Mehran R, et al. Clinical end points in coronary stent trials: a case for standardized definitions. Circulation 2007;115(17):

CI-37 Patient Disposition Enrolled patients with Stable Angina/NSTE-ACS/STEMI Indicated for PCI (N=25,384) 1:1 Randomization ITT CANGRELOR (N=12,707) ITT CLOPIDOGREL (N=12,677) No PCI (N=198) Did Not Receive Study Drug (N=142) No PCI / No Study Drug (N=232) No PCI (N=204) Did Not Receive Study Drug (N=135) No PCI / No Study Drug (N=242) MITT CANGRELOR (N=12,475) MITT CLOPIDOGREL (N=12,435) Withdrew Consent (N=2) 48 hour follow-up (N=12,459) 48 hour follow-up (N=12,422) Lost to follow-up (N=49)Lost to follow-up (N=62) 30 day follow-up (N=12,407) 30 day follow-up (N=12,357) PCI Withdrew Consent (N=3) Lost to follow-up (N=13)Lost to follow-up(N=11) Withdrew Consent (N=3) Steg GS, Bhatt DL, Hamm CW et al…. Harrington RA. Lancet 2013 at

CI-38 Study Objectives ► Prespecified, pooled analysis of patient-level data of CHAMPION- PHOENIX, CHAMPION-PLATFORM, and CHAMPION-PCI comparing cangrelor with clopidogrel control for prevention of thrombotic complications during and after PCI. ► The main differences between the trials related to the:  Timing of clopidogrel administration  Loading dose of clopidogrel  Differences in the population – e.g. no STEMI in PLATFORM  Outcome definitions Harrington RA, et al. CHAMPION PCI. NEJM 2009 Bhatt DL, et al. CHAMPION PLATFORM. NEJM 2009 Bhatt DL, et al. CHAMPION PHOENIX. NEJM 2013 White HD, et al. Meta-Analysis of CHAMPION PCI and PLATFORM. AHJ 2012

CI-39 CHAMPION Countries USA Australia Spain Poland Germany Netherlands Belgium India New Zealand Austria Czech Rep. Lithuania Ukraine Thailand South Africa Russia Georgia South Korea Bulgaria Brazil Slovakia Argentina