Bivalirudin is Superior to Heparin for ACS Patients

Bivalirudin is Superior to Heparin for ACS Patients Gregg W. Stone, MD Columbia University Medical Center New. York-Presbyterian Hospital Cardiovascular Research Foundation

Disclosures • None

Comparison 1 • Bivalirudin vs. Heparin + routine GPI in STEMI

Meta-analysis: UFH ± Abciximab in Primary PCI 8 RCTs, 3, 949 pts with STEMI w/i 12 undergoing primary (7) or rescue (1) PCI randomized to UFH + abciximab vs. UFH alone OR [95%CI] = 0. 63 [0. 54, 0. 73] P<0. 001 OR [95%CI] = 0. 72 [0. 55, 0. 94] P=0. 01 De Luca G et al. JAMA 2005; 293: 1759– 1765

CADILLAC (n=2, 082) Subacute thrombosis: Impact of abciximab 30 -day thrombosis (%) Heparin alone Heparin + abciximab 2. 5 2 1. 9 P=0. 03 1. 5 1 0. 8 1. 0 P=0. 01 0. 5 0 0 Balloon angioplasty Bare metal stent Stone GW et al. NEJM 2002; 346: 957– 66

4 RCTs of Abciximab vs. Control in Primary PCI (N=3, 266): Safety Events OR [95%CI) = 1. 45 (1. 04– 2. 02) OR [95%CI) = 1. 74 (1. 11– 2. 72) OR [95%CI) = 2. 13 (0. 78– 5. 76) OR [95%CI) = 0. 75 (0. 26– 2. 13) Kandzari DE et al. AHJ 2004; 147: 457– 62

HORIZONS-AMI + EUROMAX Pooled 30 -day results (n=5, 800) Heparin ± GPI (N=2911) Bivalirudin (N=2889) Relative risk (95% CI) P Value 90 (3. 1%) 69 (2. 4%) 0. 77 (0. 57 -1. 05) 0. 10 Cardiac 85 (2. 9%) 59 (2. 0%) 0. 70 (0. 50 -0. 97) 0. 03 Non‑cardiac 5 (0. 2%) 10 (0. 4%) 2. 01 (0. 69 -5. 88) 0. 19 Reinfarction 42 (1. 4%) 53 (1. 8%) 1. 27 (0. 85 -1. 90) 0. 24 Stent thrombosis (ARC) 40 (1. 4%) 60 (2. 1%) 1. 51 (1. 01 -2. 24) 0. 04 Acute 6 (0. 2%) 36 (1. 2%) 6. 04 (2. 55 -14. 31) <0. 0001 Sub‑acute 34 (1. 2%) 25 (0. 9%) 0. 74 (0. 44 -1. 23) 0. 24 MACE (death, MI, IDR, or stroke) 161 (5. 5%) 163 (5. 6%) 1. 02 (0. 83 -1. 26) 0. 85 Protocol major bleeding 226 (7. 8%) 120 (4. 2%) 0. 53 (0. 43 -0. 66) <0. 0001 Blood transfusion 110 (3. 8%) 62 (2. 1%) 0. 57 (0. 42 -0. 77) 0. 0002 Acquired thrombocytopenia 77 (2. 9%) 37 (1. 4%) 0. 48 (0. 33 -0. 71) 0. 0002 346 (11. 9%) 253 (8. 8%) 0. 74 (0. 63 -0. 86) <0. 0001 Death NACE (MACE or major bleeding) Breslow-Day test for study heterogeneity non-significant: P ≥ 0. 11 for all variables Stone GW et al. JACC 2015; 65: 27– 38

Comparison 2 • Bivalirudin vs. Heparin + bailout GPI in STEMI

Bivalirudin vs. Heparin Monotherapy During Primary PCI in STEMI: Three major RCTs EUROMAX HEAT PPCI BRIGHT N centers 65 1 82 N patients 2, 198 1, 812 2, 194 - Bivalirudin 1, 089 905 735 - Heparin 460 907 - Heparin + GPI 649 729 rand 730 Heparin mono bolus non-rand - 60 IU/kg 70 IU/kg 100 IU/kg Low or high dose Median 4 hrs No High dose Median 3 hrs 7. 9% vs. 25. 4% 13. 5% vs. 15. 5% 4. 4% vs. 5. 6% Prasugrel/ticagrelor 59% 89% 0% Radial 47% 81% 79% Bival infusion GPI bailout, Biv vs. Hep

EUROMAX: Treatment According to Routine GPI Use N = 2, 198 R 1: 1 UFH (91. 3%*)/LMWH ± GPI Per standard practice (n=1, 109) 58. 5% (n=649) Routine GPI Note: Non-randomized, non-stratified * Median 60 U/kg received in both arms 41. 5% (n=460) No GPI BIVALIRUDIN (provisional GPI only) (n=1, 089) 3. 9% (n=42) Routine GPI* 96% (n=1047) No GPI * Protocol Deviation 25. 4% (n=117) Bailout GPI Zeymer U et al. EHJ 2014: on-line 7. 9% (n=83) Bailout GPI

Death or major bleeding (%) EUROMAX: Primary Endpoint Death or Major Bleeding A. Heparin + Routine GPI B. Heparin Only + Bailout GPI C. Bivalirudin 12 10 Log Rank Overall: A vs. B: A vs. C: B vs. C: P values 0. 003 0. 18 0. 04 0. 0006 9. 8% 8 7. 4% 6 5. 1% 4 2 0 0 5 10 15 20 25 30 Days from Randomization Patients at risk: A. Heparin + Routine GPI 649 598 586 577 563 445 B. Hep Only + Bailout GPI 460 426 415 412 407 395 320 C. Bivalirudin 1089 1038 1024 1020 1007 899 791 Zeymer U et al. EHJ 2014: on-line

Euro. Max A prolonged high-dose bivalirudin infusion may safely reduce acute stent thrombosis Heparin ± GPI (n=1109) Bivalirudin + 0. 25 mg/kg/hr infusion* (n=670)‡ Bivalirudin + 1. 75 mg/kg/hr infusion* (n=244)§ Acute ST 2 (0. 2%) 11 (1. 6%) 1 (0. 4%) Major bleeding 57 (6. 0%) 16 (2. 4%) 7 (2. 9%) Data on a post-PCI infusion is not available for 35 patients * Median [95%CI] infusion duration was 4. 5 [4. 2, 4. 9] hours ‡ 659 of these received at least 2 hours infusion post-PCI § 191 of these received at least 2 hours infusion post-PCI Clemmensen P et al. JACC Int 2015; 8: 214– 20

Death, Re. MI, CVA, TLR (%) HEAT PPCI: 30 -Day MACE (n=1, 812) 10% 9% No. at risk Heparin Bivalirudin 8% 7% 6% Heparin 5% 4% 3% Bivalirudin Heparin 2% 1% 0% MACE 0 5 10 8. 7% 15 5. 7% 20 25 857 830 856 828 Days 907 905 871 853 866 844 862 835 Shahzad A et al. Lancet 2014 P 0. 01 30

HEAT PPCI: Stent Thrombosis ARC definite or probable stent thrombosis Bivalirudin (n=905) Heparin (n=907) Definite or probable 24 (3. 4%) 6 (0. 9%) - Definite 23 (3. 3%) 5 (0. 7%) - Probable 1 (0. 1%) - Acute 20 (2. 9%) 6 (0. 9%) - Subacute 4 (0. 6%) 0 (0%) 2 -3 x higher than in other trials Shahzad A et al. Lancet 2014

Why were the rates of acute stent thrombosis higher in HEAT PPCI than in any other bivalirudin STEMI trial? Median activated clotting time post bolus 400 357 HORIZONS-AMI ACT (seconds) 350 300 250 264 251 ~239 secs w/o GPI 200 HEAT PPCI 224 150 100 50 0 Bivalirudin HEAT PPCI: Actalyke XL MAX‐ACT system HORIZONS-AMI: Hemotec and Hemochron Heparin

HEAT PPCI: ACT* and GPI bailout Bivalirudin arm (n=915) N Measure ACT 5 -15 mins after bolus 806 (88%) 251 [229, 285] sec ACT end-procedure 771 (84%) 246 [229, 270] sec Bivalirudin rebolus anytime** 12. 7% GPI bailout 13. 5% ~25% <229 seconds; rebolus rate should have been ~25% *Actalyke XL MAX‐ACT system **By protocol, rebolus for ACT <225 seconds Shahzad A et al. Lancet 2014

1. 2. 3. 4. 5. Limited external validity Implausible effect size Unequal allocation of resources Lack of blinding Numerous examples where singlecenter trials were “reversed” by multiple multicenter trials “What works in one location may not in others due differences in baseline risk, intercurrent care, opportunity cost, and the fidelity with which the intervention can be delivered. ” “Single-center studies are valuable hypothesis-generating investigations. However, the evidence shows that. …they have typically not been confirmed by larger multicenter studies…. the history of this field argues that two beneficial RCTs are necessary with at least one being a confirmatory trial. Accordingly, we should be very careful in taking single-center trials as a demonstration of a biological truth and should avoid giving them undue weight or value. . …practice guidelines should rarely, if ever, be based on evidence from single-center trials. ” Bellomo R et al. Crit Care Med 2009; 37: 3114 -9

Limitations of HEAT PPCI • Suboptimal use of bivalirudin • PCI performed in only 82% of patients – not a typical primary PCI cohort • Study investigators did the data monitoring • CEC was at the study center – not independent • Data analysis was performed at the study center – not independent • Reinfarction diagnosed by a “common sense” definition – new thrombus qualifies! • Stent thrombosis not adjudicated by a core lab

BRIGHT: Study flow 2, 194 pts with AMI randomized at 82 centers in China Aspirin and clopidogrel 86. 2 % STEMI 13. 8% NSTEMI 79% radial Randomization (1: 1: 1) Bivalirudin alone N=735 Biv 0. 75 mg/kg bolus + 1. 75 mg /kg/h infusion (0. 3 mg/kg bolus if ACT< 225 s). Bailout GPI permitted. Biv infusion (1. 75 mg/kg/h) continued for at least 30 min post PCI (median 3 h). 4. 4% bailout tirofiban. UFH alone N=729 Heparin 100 U/kg bolus + additional dose if ACT <200 s. Bailout GPI permitted. ACT goal = 250 -300. 5. 6% bailout tirofiban. UFH + Tirofiban N=730 Heparin 60 U/kg bolus. Tirofiban 10μg/kg bolus + 0. 15 μg/kg/min infusion for 18 -36 h. ACT goal = 200 -250. Follow-up at 30 days, 6 months and 1 year Primary endpoint: NACE, including MACCE (all-cause death, re. MI, TVR or stroke) and bleeding events at 30 days. Han Y et al. Submitted.

BRIGHT: Primary and Major Secondary Endpoint Events at 30 Days Bivalirudin (n=735) Primary endpoint 18 P<0. 001 17. 0 16 Heparin (n=729) Heparin + Tirofiban (n=730) Biv vs. Hep, p=0. 009 RR (95%CI) 0. 67 (0. 50 -0. 90), NNT=23. 1 Biv vs. Hep+Tiro, p<0. 001 RR (95%CI) 0. 52 (0. 39 -0. 69), NNT=12. 3 Hep vs. Hep+Tiro, p=0. 04 RR (95%CI) 0. 78 (0. 61 -0. 99), NNT=26. 2 13. 2 14 12. 3 P<0. 001 12 (%) 10 P=0. 74 8. 8 7. 5 8 5. 0 6 5. 8 4. 9 4. 1 4 2 0 NACE MACCE Han Y et al. Submitted. Any Bleeding

BRIGHT: Stent Thrombosis at 30 Days Stent thrombosis (%) 1. 2 Bivalirudin (n=735) Heparin (n=729) Heparin + Tirofiban (n=730) P=0. 77 0. 9% 1. 0 0. 7% 0. 8 0. 6% 0. 4 0. 2 0 Bivalirudin Heparin + Tirofiban (n=735) (n=729) (n=730) Han Y et al. Submitted.

Comparison 3 • Bivalirudin vs. Heparin GPI in STEMI in the “real world”

Impact of Bleeding Avoidance Strategies NCDR Cath. PCI Registry 2004 -2008: PCI in 1, 522, 935 pts Manual compression alone, closure devices, bivalirudin, or both were used in 35%, 24%, 23%, and 18% of pts, respectively. Propensity-adjusted major bleeding Major bleeding (%) 8 6 4 Heparin + manual compression (n=508, 455) Heparin + vascular closure devices (n=205, 606) Bivalirudin + manual compression (n=172, 471) Bivalirudin + vascular closure devices (n=130, 378) Adj OR (95%CI) = 0. 77 (0. 73 – 0. 80) NNT = 148 2. 7 2. 5 Adj OR (95%CI) = 0. 67 (0. 63 – 0. 70) NNT = 118 1. 9 2 23%↓ 33%↓ 0 All pts Marso SP et al. JAMA. 2010; 303: 2156 -64 Adj OR (95%CI) = 0. 38 (0. 35 – 0. 42) NNT = 70 1. 0 62%↓

Impact of Bivalirudin in Radial Procedures NCDR Cath. PCI Registry 2009 -2012: PCI in 501, 017 pts Femoral access with closure devices, radial access with heparin, and radial access with bivalirudin were used in 76%, 13%, 11%, and 18% of pts. Major bleeding: 3. 0 2. 5 2. 7% 2190 (0. 6) Non-access bleeding Access bleeding 2. 5% 168 (0. 3) 2. 0 P<0. 0001 1. 8% 114 (0. 2) 1. 5 1. 0 0. 5 0 Propensity-adjusted All pts Radial + heparin Radial + bival OR (95% CI) NNT 0. 96 (0. 88 -1. 05) 1607 0. 79 (0. 72 -0. 86) 138 High bleeding risk pts Radial + heparin Radial + bival 1. 00 (0. 90 -1. 11) 1393 0. 79 (0. 70 -0. 88) 68 0. 4 8186 (2. 1) 1405 (2. 2) 889 (1. 6) Femoral access Radial access with VCD + + heparin + bivalirudin 0. 6 1. 0 0. 8 1. 2 Adjusted OR Lower than femoral + Higher than femoral + bivalirudin + VCD Radial access/bivalirudin combination was used less frequently in pts at highest risk for bleeding Baklanov DV et al. Circ Cardiovasc Interv. 2013; 6: 347 -53

Anticoagulation Regimens During PCI N = 458, 448 PCI pts 2004 -2008 at 299 hosps (Premier Perspective Database, ~1/5 th of all US hosp discharges; bival in 41%) In-hospital events, propensity adjusted Bleeding + Transfusion Mortality OR (95% CI) Comparator Bivalirudin monotherapy (n=156, 064) OR (95% CI) P Value Comparator OR (95% CI) P Value 0. 51 (0. 48, 0. 55) <0. 0001 Bivalirudin monotherapy (n=156, 064) 0. 59 (0. 54, 0. 65) <0. 0001 Bivalirudin + GPI (n=33, 566) 0. 96 (0. 87, 1. 06) Bivalirudin + GPI (n=33, 566) 0. 82 (0. 72, 0. 94) 0. 004 Heparin alone (n=85, 870) 0. 71 (0. 66, 0. 76) <0. 0001 Heparin alone (n=85, 870) 0. 88 (0. 82, 0. 96) 0. 003 0 Comparator Better 1 Heparin + GPI Better (n=182, 948) 0. 37 2 0 Comparator Better Wise GR et al. J Interv Cardiol 2012; 25: 278– 88 1 Heparin + GPI Better (n=182, 948) 2

SCAAR (31, 438 STEMIs) All primary PCI patients who received a stent in Sweden from January 2007 to July 2014 ARC Definite Stent Thrombosis n Bivalirudin alone Heparin + GPIIb/IIIa 16, 860 3, 246 11, 392 ST day 0 -30 0. 8% (n=142) 0. 9% (n=30) 0. 9% (n=103) ST day 0 -1 0. 3% (n=55) 0. 3% (n=9) 0. 2% (n=25) ST day 2 -30 0. 5% (n=87) 0. 7% (n=21) 0. 7% (n=78) c/o David Erlinge

The MATRIX Trial (NCT 01433627) ACC Monday, March 16 th ~8400 pts with STEMI and NSTEACS Femoral access Radial access PCI Bivalirudin with or without a 4 -hour post-PCI infusion (low dose or high dose) PI: Marco Valgimigli Unfractionated heparin + bailout GPI