1. Incidence and Impact of Dual Antiplatelet Therapy (DAPT) Cessation on Adverse Events following Percutaneous Coronary Intervention (PCI): Results from the Real-World PARIS Registry Roxana Mehran, MD Professor of Medicine (Cardiology) and Health Evidence Policy Director of Interventional Cardiovascular Research and Clinical Trials The Icahn School of Medicine at Mount Sinai, New York, NY on behalf of PARIS Investigators PARIS Registry

2. Background and Rationale Antiplatelet agents are the cornerstone of therapy in patients with ACS and in those undergoing PCI Antiplatelet agents are the cornerstone of therapy in patients with ACS and in those undergoing PCI Current ACC/AHA guidelines 1 recommend 30 days DAPT following placement of a BMS and 1 year following placement of a DES. Current ACC/AHA guidelines 1 recommend 30 days DAPT following placement of a BMS and 1 year following placement of a DES. In patients with ACS 12 months of DAPT is recommended regardless of stent type In patients with ACS 12 months of DAPT is recommended regardless of stent type 1. Wright et al. JACC. 10 May.2011. PARIS Registry

3. DAPT Cessation and PCI: Existing Evidence Premature cessation of DAPT, within the first 6 months after PCI, has been associated with an increased risk of stent thrombosis. 1 Premature cessation of DAPT, within the first 6 months after PCI, has been associated with an increased risk of stent thrombosis. 1 Sustained DAPT (one year or longer) has been associated with lower risk for adverse events in observational studies. 2,3 Sustained DAPT (one year or longer) has been associated with lower risk for adverse events in observational studies. 2,3 Most studies involved select cohorts and limited by pre- specified or standard criteria to define DAPT status PARIS Registry 1 Schulz et al., EHJ 2009; 2 Ho et al., AHJ 2007; 3 Park et al., AJC 2006

4. DAPT Cessation and PCI: Unresolved Questions Does risk after DAPT cessation depend on the underlying context or clinical circumstances in which antiplatelet therapy is stopped (surgery vs. bleeding vs. physician-guidance)? How long does risk persist after antiplatelet therapy is withdrawn? How long does risk persist after antiplatelet therapy is withdrawn? What is the overall contribution of DAPT cessation on adverse events in the contemporary PCI era? What is the overall contribution of DAPT cessation on adverse events in the contemporary PCI era? PARIS Registry

5. PARIS – Objectives Determine the incidence of different modes of DAPT cessation after PCI. Evaluate the associations between DAPT cessation and adverse events following PCI by the underlying clinical context in which antiplatelet therapy was withdrawn. PARIS Registry

6. Study Design Multicenter, multinational, observational study Multicenter, multinational, observational study 5,031 subjects were followed for approximately 24 months post stent implantation 5,031 subjects were followed for approximately 24 months post stent implantation Included bare metal and drug-eluting stents Included bare metal and drug-eluting stents All events, including all occurrences of DAPT cessation, were adjudicated by a blinded external clinical events committee All events, including all occurrences of DAPT cessation, were adjudicated by a blinded external clinical events committee PARIS Registry

7. Enrollment Eligibility (1) Inclusion Criteria Successful stent placement in one or more lesions in native coronary arteries using an approved coronary stent, and discharged on DAPT Successful stent placement in one or more lesions in native coronary arteries using an approved coronary stent, and discharged on DAPT Diagnosis of acute coronary syndrome, stable angina, or documented silent ischemia Diagnosis of acute coronary syndrome, stable angina, or documented silent ischemia Patient was over 18 years old, provided consent, and agreed to follow-up Patient was over 18 years old, provided consent, and agreed to follow-up PARIS Registry

8. Eligibility for Enrollment (2) Exclusion Criteria Evidence of stent thrombosis during baseline procedure Evidence of stent thrombosis during baseline procedure Patient was already participating in an investigational device or drug study Patient was already participating in an investigational device or drug study PARIS Registry

9. Modes of DAPT Cessation Discontinuation Discontinuation  patients had discontinued DAPT as per recommendation of their physician who felt the patient no longer needed therapy Interruption Interruption  patients had interrupted DAPT use on a voluntary basis and as guided by a physician due to (e.g. surgery)  DAPT was then reinstituted within 14 days Disruption Disruption  patients had disrupted DAPT use due to bleeding or non- compliance. PARIS Registry

10. Paris Enrollment - Patients 5,031 patients enrolled at 15 centers in 5 countries USA [10] n=3,666, 72,9% July 1 st, 2009 and October 29 th, 2010 Columbia University Medical Center (n=927, 18,5%) Minneapolis Heart Institute Foundation (n=704, 14%) Mount Sinai Medical Center (n=555, 11%) LeBauer Cardiovascular Research Foundation/ Moses Cone Heart and Vascular Center (n=344, 6,8%) St. Luke's Hospital/ Mid-America Heart Institute (n=318, 6,3%) Geisinger Medical Center (n=276, 5,5%) Washington Adventist Hospital (n=199, 4%) University of Kentucky (n=143, 2,8%) Heart Center of Indiana/ St. Vincent's/ The Care Group (n= 125, 2,5% ) Washington Hospital Center (n=75, 1,5%)

11. PARIS Registry Paris Enrollment - Patients July 1 st, 2009 and October 29 th, 2010 Germany [1] n=720, 14,3% Italy [2] n=307, 6% France [1] n=160, 3,2% Greece [1] n=180, 3,6% EUROPE [5] n=1,367, 27,1% Charité Hospital, Germany (n=720, 14,3%) San Raffaele Hospital, Italy (n=221, 4,4%) Onassis Cardiac Surgery Center, Greece (n= 180, 3,5%) Hospital Bichat, France (n=160, 3,2%) Careggi Hospital, Italy (n=86, 1,7%) 5,031 patients enrolled at 15 centers in 5 countries

12. SitePrincipal InvestigatorTotal # Enrolled Columbia University, USADr. Giora Weisz927 Charité, GermanyDr. Bernhard Witzenbichler720 Minneapolis Heart, USADr. Tim Henry704 Mount Sinai, USADr. Annapoorna Kini553 LeBauer/Moses Cone, USADr. Thomas Stuckey/Dr. Bruce Brodie344 St. Luke's /Mid-America Heart, USADr. David Cohen318 Geisinger Medical, USADr. Peter Berger276 San Raffaele, ItalyDr. Antonio Colombo221 Washington Adventist, USADr. Fayaz Shawl199 Onassis, GreeceDr. George Dangas/Dr. Ioannis Iakovou180 Hospital Bichat, FranceDr. Gabriel Steg160 University of Kentucky, USADr. David Moliterno143 Heart Center of Indiana/St. Vincent's, USA Dr. James Hermiller125 Careggi, ItalyDr. David Antonucci86 Washington Hospital Center, USADr. Ron Waksman75 Participating Sites, Principal Investigators and Patients Enrolled

13. PARIS: Study Organization Data Management and Monitoring Data Management and Monitoring Medical Devices Consultancy, Ltd. New Zealand Medical Devices Consultancy, Ltd. New Zealand Statistical Analysis Statistical Analysis London School of Tropical Health and Hygiene: Stuart Pocock, Cono Ariti London School of Tropical Health and Hygiene: Stuart Pocock, Cono Ariti Mount Sinai School of Medicine: Usman Baber, Samantha Sartori Mount Sinai School of Medicine: Usman Baber, Samantha Sartori Project Management/Clinical Coordinating Center Project Management/Clinical Coordinating Center Mount Sinai School of Medicine: Kristin Falciglia, Maria Alu Mount Sinai School of Medicine: Kristin Falciglia, Maria Alu Funding (Investigator-Initiated) Funding (Investigator-Initiated) Bristol Myers Squib/Sanofi-Aventis Bristol Myers Squib/Sanofi-Aventis PARIS Registry

14. Executive Committee Members *Dr. Roxana Mehran, USA and **Dr. Antonio Colombo, Italy Dr. Alaide Chieffo, Italy Dr. David Cohen, USA Dr. David Moliterno, USA Dr. Gabriel Steg, France Dr. Michael Gibson, USA Dr. Mitch Krucoff, USA Dr. Bernhard Witzenbichler, USA Dr. Giora Weisz, USA Executive Committee and CEC Clinical Events Committee *Dr. Steven Marx, Columbia University, NY Dr. Jason C. Kovacic, Mount Sinai, NY Dr. Mun Hong, St. Luke’s Roosevelt Hospital, NY Dr. S. Chiu Wong, NYP Cornell, NY Dr. Bruce Darrow, Mount Sinai, NY * Chair, ** Co-chair

15. Sequence of Results PARIS Registry Baseline characteristics compared between subjects by presence or absence of any DAPT cessation over 2 years. Baseline characteristics compared between subjects by presence or absence of any DAPT cessation over 2 years. Incidence of major ischemic and bleeding adverse events calculated in overall population. Incidence of major ischemic and bleeding adverse events calculated in overall population. Association between modes of DAPT cessation on adverse events examined using Cox proportional hazards regression with DAPT cessation entered as time updated covariate. Association between modes of DAPT cessation on adverse events examined using Cox proportional hazards regression with DAPT cessation entered as time updated covariate. Time updated covariates were generated using hierarchy of ‘worst’ DAPT status defined as Recommended Discontinuation -> Interruption -> Disruption Time updated covariates were generated using hierarchy of ‘worst’ DAPT status defined as Recommended Discontinuation -> Interruption -> Disruption

16. 5,031 Patients with successful PCI with stenting enrolled at 15 sites in the US and Europe Final Study Population – 5018 Patients 13 Patients excluded from analysis (1 died prior to discharge and 12 not discharged on DAPT) Lost to follow-up (n=340) Within 2 years Available Follow-up: 4678/5018 (93.2%) Lost to follow-up (n=133) Within 365 days Available Follow-up: 4885/5018 (97.3%) Within 30 days Available Follow-up: 4972/5018 (99.1%) Lost to follow-up (n=46)

17. 2-Year Kaplan-Meier Plot of Any DAPT Cessation 60 06121824 Time since PCI, Months 50 Cumulative Incidence, % 40 30 20 10 30 Days (2.9%) One Year (23.3%) Two Years (57.3%) Incidence rates calculated over entire study population. Patients censored at last known contact, death or study end.

18. 2-Year Kaplan-Meier Plots of Any Discontinuation, Interruption and Disruption 60 06121824 Time since PCI, Months 50 Cumulative Incidence, % 40 30 20 10 Discontinuation Disruption Interruption 40.8% 14.4% 10.5% Incidence rates calculated over entire study population. Patients censored at last known contact, death or study end.

19. * *SD = standard deviation; PVD = peripheral vascular disease; MI = myocardial infarction; CABG = coronary artery bypass graft; DAPT=dual antiplatelet therapy. BMI= body mass index. P-value is for each DAPT status versus the No DAPT cessation category. Baseline Characteristics by Any DAPT Cessation Over 2 Years No DAPT Cessation (n= 2304) Recommended Discontinuation (n = 1611) Interruption (n = 412) Disruption (n = 691) Age (years)62.9 ± 11.464.7 ± 10.8565.3 ± 10.7464.8 ± 12.26 Female Gender, n (%)559 (24.3)404 (25.1)103 (25)213 (30.8) BMI (kg/m 2 )29.5 ± 5.728.7 ± 5.330.0 ± 5.529.3 ± 6.2 Hypertension, n (%)1860 (80.7)1258 (78.1)355 (86.2)536 (77.6) Previous MI, n (%)637 (27.6)329 (20.4)109 (26.5)139 (20.1) Previous CABG, n (%)374 (16.2)169 (10.5)67 (16.3)75 (10.9) Stroke, n (%)79 (3.4)52 (3.2)17 (4.1)25 (3.6) PVD, n (%)188 (8.2)120 (7.4)33 (8)51 (7.4) Current smoker, n (%)445 (34.7)315 (41.1)62 (27.7)159 (40.3) Diabetes, n (%)829 (36)456 (28.3)150 (36.4)219 (31.7) Acute Coronary Syndrome, n (%)938 (40.7)640 (39.7)140 (34)338 (48.9) BMS, n (%) 255 (11.1)325 (20.2)68 (16.5)163 (23.6) 1st generation DES, n (%)347 (15.1)194 (12.0)58 (14.1)75 (10.9) 2nd generation DES, n (%)1702 (73.9)1092 (67.8)286 (69.4)453 (65.6)

20. Overall Event Rates Over 2 Years Cumulative Incidence, % Incidence calculated as cumulative incidence from a Kaplan-Meier estimate of the time to the first occurrence of the adverse event.

21. Major Bleeding Rates Over 2 Years Cumulative Incidence, % Incidence calculated as cumulative incidence from a Kaplan-Meier estimate of the time to the first occurrence of the adverse event.

22. Impact of DAPT Cessation on Adverse Events

23. 1.00 (Ref) 0.63 (0.46, 0.86) 1.41 (0.94, 2.12) 1.50 (1.14, 1.97) On-DAPT Discontinuation Interruption Disruption 0.004 0.101 0.004 413 52 26 67 7.04 (3.31, 14.95) 2.17 (0.97, 4.88) 1.30 (0.97, 1.76) 0-7 Days 8-30 days 31+ days <0.001 0.06 0.083 7 6 54 HR (95% CI)PEvents (n) 0.250.5124816 DAPT Cessation and MACE* Hazard Ratio *Cardiac Death, Def/Prob ST, Spontaneous MI, Clinically Driven TLR. All Cox Models adjusted for age, gender, region, ACS presentation, type of stent, number of stents implanted.

24. DAPT Cessation and Cardiac Death, Def/Prob ST, Spontaneous MI 1.00 (Ref) 0.76 (0.50, 1.14) 1.05 (0.58, 1.92) 2.06 (1.49, 2.83) On-DAPT Discontinuation Interruption Disruption 0.181 0.864 <0.001 218 31 12 54 9.82 (4.57, 21.12) 2.96 (1.21, 7.24) 1.71 (1.20, 2.44) 0-7 Days 8-30 days 31+ days <0.001 0.017 0.003 7 5 42 HR (95% CI)P 0.250.512481632 Hazard Ratio All Cox Models adjusted for age, gender, region, ACS presentation, type of stent, number of stents implanted. Events (n)

25. DAPT Cessation and Spontaneous MI 1.00 (Ref) 0.92 (0.53, 1.58) 1.20 (0.55, 2.63) 2.95 (1.99, 4.38) On-DAPT Discontinuation Interruption Disruption 0.748 0.647 <0.001 116 18 7 39 0-7 Days 8-30 days 31+ days 18.25 (8.34, 39.95) 4.69 (1.71, 12.83) 2.22 (1.42, 3.46) <0.001 0.003 <0.001 7 4 28 0.250.51248163264 Hazard Ratio HR (95% CI) P All Cox Models adjusted for age, gender, region, ACS presentation, type of stent, number of stents implanted. Events (n)

26. 1.00 (Ref) 0.39 (0.11, 1.35) 0.64 (0.09, 4.82) 2.58 (1.22, 5.46) On-DAPT Discontinuation Interruption Disruption 0.137 0.664 0.013 57 3 1 10 0-7 Days 8-30 days 31+ days 15.94 (5.57, 45.58) 2.68 (0.36, 19.68) 1.35 (0.50, 3.64) <0.001 0.334 0.551 4 1 5 0.250.5124816 32 64 Hazard Ratio HR (95% CI)P DAPT Cessation and Def/Prob Stent Thrombosis All Cox Models adjusted for age, gender, region, ACS presentation, type of stent, number of stents implanted. Events (n)

27. 1.00 (Ref) 0.64 (0.36, 1.16) 1.06 (0.48, 2.34) On-DAPT Discontinuation Interruption Disruption 0.141 0.885 100 15 7 1.68 (1.05, 2.67)0.029 26 5.73 (1.39, 23.62) 3.44 (1.08, 10.98) 1.44 (0.87, 2.38) 0-7 Days 8-30 days 31+ days 0.016 0.037 0.161 2 3 21 HR (95% CI)P 0.250.512481632 Hazard Ratio DAPT Cessation and Cardiac Death All Cox Models adjusted for age, gender, region, ACS presentation, type of stent, number of stents implanted. Events (n)

28. Overall Contribution of DAPT Cessation on Adverse Events

29. Number (%) of Spontaneous MI events by DAPT Status* *Out of 180 spontaneous MI events at 2 years, 116 (64.4%) occurred while patients were ON DAPT. Spontaneous MI defined according to Universal Definition. Number of Events

30. Number (%) of Def/Prob ST events by DAPT Status* *Out of 71 ST events at 2 years, 57 (80.3%) occurred while patients were ON DAPT. ST defined by the Academic Research Consortium (ARC) Critera. Number of Events

31. Number (%) of Cardiac Death events by DAPT Status* *Out of 148 Cardiac Death events at 2 years, 100 (67.6%) occurred while patients were ON DAPT. Cardiac Death defined using ARC criteria. Number of Events

32. Number (%) of Major (BARC ≥ 3) Bleeding Events by DAPT Status* *Out of 196 Bleeding events at 2 years, 131 (66.8%) occurred while patients were ON DAPT. Major Bleeding defined as BARC ≥ 3). Number of Events

33. Attributable Risk Calculation For each time interval after Interruption or Disruption: For each time interval after Interruption or Disruption: Adjusted HR = Obs Events/Expected Events Adjusted HR = Obs Events/Expected Events # events Expected = Observed # events/adjusted HR # events Expected = Observed # events/adjusted HR (Ex: Adjusted HR for ST after Disruption = 2.58 and 10 events Observed Expected # Events = 10/2.58 = 3.9) Obs Events – Expected Events = Additional Events Statistically attributed to Interruption or Disruption Obs Events – Expected Events = Additional Events Statistically attributed to Interruption or Disruption 10 – 3.9 = 6.1 Additional ST events attributed to Disruption10 – 3.9 = 6.1 Additional ST events attributed to Disruption % of Events Attributable = # Attributable Events/Total # Events % of Events Attributable = # Attributable Events/Total # Events 6.1/71 = 8.6% of all ST statistically attributed to Disruption6.1/71 = 8.6% of all ST statistically attributed to Disruption

34. Proportion of Ischemic Adverse Events Attributable to DAPT Disruption or Interruption Attributable Risk, %

35. Limitations Observational design precludes causal inferences Follow-up phone calls to ascertain DAPT status can introduce recall bias Small number of events (n=7) occurring early (days 1-7) after disruption led to imprecise risk estimates with wide CI. Findings merit confirmation in larger studies.

36. Conclusions The impact of DAPT cessation on cardiac risk after PCI is not uniform but varies substantially by underlying mode, a novel finding with important implications for study design and practice. Relative risk for MACE due to disruption is substantial, albeit short-lived, compared to those on DAPT. The overall impact of DAPT cessation on adverse events is modest and may have been mitigated with the introduction of safer stent platforms. Findings highlight the need for uniform approaches in classifying DAPT cessation, analogous to those currently used for bleeding and MI.