A Prospective, Randomized Comparison of Paclitaxel- eluting TAXUS Stents vs. Bare Metal Stents During Primary Angioplasty in Acute Myocardial Infarction – One Year Results – HORIZONS AMI Trial Gregg W. Stone MD For the HORIZONS-AMI Investigators, TCT 2008

Background ● No consensus exists regarding the safety and efficacy of drug- eluting stents in pts with STEMI undergoing primary PCI –TLR and restenosis rates tend to be lower in STEMI vs. elective PCI patients because of less plaque burden and non viable myocardium –The safety of implanting DES in ruptured plaques with thrombus has been questioned ● Outcomes from registry studies of DES vs. BMS in STEMI have been conflicting, and no large-scale randomized trials have been performed Stone GW. TCT 2008.

CABG – Primary PCI – Medical Rx Aspirin, thienopyridine R 1:1 3,000 pts eligible for stent randomization R 3:1 Paclitaxel-eluting TAXUS stentBare metal EXPRESS stent HORIZONS AMI Trial Design Harmonizing Outcomes with Revascularization and Stents in AMI 3 UFH + GP IIb/IIIa inhibitor (abciximab or eptifibatide) Bivalirudin monotherapy (± provisional GP IIb/IIIa) 3,602 pts with STEMI with symptom onset ≤12 hours Emergent angiography, followed by triage to… Clinical FU at 30 days, 6 months, 1 year, and then yearly through 5 years; angio FU at 13 months Clinical FU at 30 days, 6 months, 1 year, and then yearly through 5 years; angio FU at 13 months Stone GW. NEJM 2008;358:

Stent Randomization Hypotheses ● In patients with STEMI undergoing primary PCI, the use of paclitaxel-eluting TAXUS stents rather than bare metal EXPRESS stents will be: –Efficacious, as evidenced by reduced rates of ischemia-driven target lesion revascularization at 1-year and angiographic binary restenosis at 13 months; and –Safe, with non-inferior rates of the composite measure of death, reinfarction, stent thrombosis or stroke at 1-year Stone GW. TCT 2008.

Clinical Inclusion Criteria ● STEMI >20 mins and <12 hours in duration –ST-segment elevation of  1 mm in  2 contiguous leads; or –Presumably new left bundle branch block; or –True posterior MI with ST depression of  1 mm in  2 contiguous anterior leads –Patients with cardiogenic shock, left main disease, etc., were not excluded ● Age ≥18 years ● Written, informed consent Stone GW. NEJM 2008;358:

Principal Clinical Exclusion Criteria ● Contraindication to any of the study medications ● Prior administration of thrombolytic therapy, bivalirudin, GP IIb/IIIa inhibitors, LMWH or fondaparinux for the present admission (prior UFH allowed) ● Current use of coumadin ● History of bleeding diathesis or known coagulopathy (including HIT), or will refuse blood transfusions ● History of intracerebral mass, aneurysm, AVM, or hemorrhagic stroke; stroke or TIA within 6 months or any permanent neurologic deficit; GI or GU bleed within 2 months, or major surgery within 6 weeks; recent or known platelet count <100,000 cells/mm 3 or hgb <10 g/dL ● Planned elective surgical procedure that would necessitate interruption of thienopyridines during the first 6 months post enrollment Stone GW. NEJM 2008;358:

Angiographic Inclusion Criteria ● The presence of least 1 acute infarct artery target vessel* in which: –a) ALL significant lesions are eligible for stenting with study stents, and –b) ALL such lesions have a visually estimated reference diameter ≥2.25 mm and ≤4.0 mm ● Expected ability to deliver the stent(s) to all culprit lesions (absence of excessive proximal tortuosity or severe calcification) ● Expected ability to fully expand the stent(s) at all culprit lesions (absence of marked calcification) *Arteries containing multiple lesions may be randomized if all lesions are study stent eligible; multiple vessels may be randomized if all lesions in each vessel are study stent eligible Stone GW. TCT 2008.

Angiographic Exclusion Criteria ● Bifurcation lesion definitely requiring implantation of stents in both the main vessel + side branch ● Infarct related artery is an unprotected left main ● >100 mm of study stent length anticipated ● Infarction due to stent thrombosis, or infarct lesion at the site of a previously implanted stent ● High likelihood of CABG within 30 days anticipated Stone GW. TCT 2008.

Study Medications (i) ● Unfractionated heparin –60 U/kg IV*; subsequent boluses titrated by nomogram to ACT secs; terminated at procedure end unless prolonged antithrombin needed ● Bivalirudin –Bolus 0.75 mg/kg IV**, infusion 1.75 mg/kg/h, not titrated to ACT; terminated at procedure end unless prolonged antithrombin needed (0.25 mg/kg/hr infusion) ● Glycoprotein IIb/IIIa inhibitors –Routine use in UFH arm; recommended only for giant thrombus or refractory no reflow in bivalirudin arm –Abciximab or double bolus eptifibatide as per investigator discretion – dosing per FDA label, renal adjusted; continued for 12  (abciximab) or  (eptifibatide) * If pre randomization UFH administered, ACT is checked first ** If pre randomization UFH administered, started 30’ after last bolus Stone GW. NEJM 2008;358:

Study Medications (ii) ● Aspirin –324 mg chewed non enteric coated or 500 mg IV in the ER, followed by mg/day in-hospital and mg/day as out patient indefinitely ● Thienopyridines –Clopidogrel 300 mg or 600 mg loading dose (per investigator discretion) in the ER followed by 75 mg PO QD for at least 6 months (1 year or longer recommended) – Ticlopidine load + daily dose permissible if clopidogrel is unavailable or patient is allergic ● Other –Beta blockers: IV pre procedure followed by PO QD in the absence of contraindications; ACE inhibitors for HTN, CHF or LVEF 100 mg/dl Stone GW. NEJM 2008;358:

2 Primary Stent Endpoints (at 12 Months) and Major Secondary Endpoint (at 13 Months) Binary angiographic restenosis 2) Composite Safety MACE = All cause death, reinfarction, stent thrombosis (ARC definite or probable)**, or stroke 1) Ischemia-driven TLR* 11 * Related to randomized stent lesions (whether study or non study stents were implanted); ** In randomized stent lesions with ≥1 stent implanted (whether study or non study stents) Stone GW. TCT 2008.

Statistical Methodology ● Second randomization stratification i.Results from the first randomization ii.Presence of medically treated diabetes mellitus iii.Presence of any lesion >26 mm in length (requiring overlapping stents by protocol) iv.U.S. vs. non-U.S. site ● Primary analysis conducted in the ITT cohort using Kaplan- Meier methodology, with the groups compared by log-rank ● 1-sided α=0.025 for NI; 2-sided α=0.05 for Sup Stone GW. TCT 2008.

Power Analysis ● With 2,850 pts randomized 3:1* Assumed event rates One YearTestEXPRESSTAXUS  Power Ischemic TLRSuperiority9.0%5.0%-95% Composite Safety MACE Noninferiority7.5% 3.0%80% Assumed event rates 13 MonthsTest EXPRESS TAXUS  Power RestenosisSuperiority26.0%15.6%-96% ● With angiographic FU in 1,125 randomized pts (analyzable) * Assumed 5% withdrew or lost to FU at 1 year → 3000 randomized Stone GW. TCT 2008.

Horizons Enrollment - Centers ● 3,602 pts randomized at 123 centers in 11 countries between March 25th, 2005 and May 7th, 2007 USA (57) (1) Spain (6) UK (2) Norway (2) Norway Poland (9) Germany (16) Austria (5) (3) Netherlands Italy (2) Argentina (12) Israel (10) Stone GW. TCT 2008.

R 1:1 HORIZONS AMI Trial Design Harmonizing Outcomes with Revascularization and Stents in AMI UFH + GPI (n=1802) Bivalirudin (n=1800) Withdrew Withdrew Lost to FU Lost to FU pts with STEMI Randomized 1-Year FU TAXUS N=2257 N=2186 (96.9%) EXPRESS BMS N=749 N=715 (95.5%) R 3:1 Primary Medical Rx193 Primary CABG 62 Deferred PCI 2 Index PCI, not eligible - PTCA only119 - Stented pts eligible for stent rand. 93.1% of all stented pts were randomized Stone GW. TCT 2008.

Baseline Characteristics (i) TAXUS (N=2257) EXPRESS (N=749) Age (years)59.9 [52.4, 69.4]59.3 [51.8, 69.2] Male77.0%76.0% Diabetes16.1%15.2% Hypertension51.2%51.9% Hyperlipidemia42.2%41.1% Current smoking46.3%51.9% Prior MI9.1%10.9% Prior PCI9.5%7.7% Prior CABG2.2%1.9% *P=0.009 * Stone GW. TCT 2008.

Baseline Characteristics (ii) TAXUS (N=2257) EXPRESS (N=749) Weight (kg)80 [71, 90] Killip class %8.0% Anterior MI42.2%44.7% LVEF (%), site50 [44, 59]50 [43, 58] Symptoms – PCI, hrs3.7 [2.7, 5.5]3.8 [2.7, 5.8] Femoral a. access93.6%92.9% Venous access8.5%8.0% Closure device30.1%28.8% Aspiration catheter11.4%10.7% Stone GW. TCT 2008.

Study Drugs TAXUS (N=2257) EXPRESS (N=749) Aspirin at home22.7%20.5% Aspirin load pre PCI97.0%97.2% Thienopyridine at home2.1%2.5% Thienopyridine loading dose98.9%98.3% - clopidogrel 300 mg34.2%35.5% - clopidogrel 600 mg63.3%61.3% - clopidogrel other1.2%1.3% - ticlopidine0.5%0.3% UFH pre randomization65.2%65.8% UFH as the procedural antithrombin49.8%50.1% Bivalirudin administered50.7%50.9% GP IIb/IIIa inhibitor administered52.0%51.5% Stone GW. TCT 2008.

Procedural Data (Site Reported) TAXUS (N=2257, L=2495) EXPRESS (N=749, L=815) N lesions treated1.1 ± ≥ 2 lesions treated11.1%9.0% - ≥ 2 vessels treated4.5%3.1% Direct stenting attempted30.4%33.7% Stent target lesion: LAD, LCX, RCA, LM, SVG 40.1%, 14.6%, 45.1%, 0.3%, 0.3% 42.4%, 15.9%, 41.3%, 0.4%, 0.4% N stents implanted1.5 ± ± 0.7 Total stent length**30.8 ± ± 14.9 Max balloon dia. (mm)3.00 [2.75, 3.50]3.00 [2.90, 3.50] Max pressure (atm.)14.0 [12.0, 16.0] *P=0.002; **P< * ** Stone GW. TCT 2008.

Quantitative Coronary Angiography TAXUS (L=2642, V=2353) EXPRESS (L=857, V=771) Pre RVD (mm)2.89 ± ± 0.50 Pre MLD (mm)0.35 ± 0.45 Pre %DS87.6 ± ± 15.4 Pre lesion length (mm)17.5 ± ± 8.8 Pre TIMI 0/1, 2, 360.6%, 13.6%, 25.7%57.4%, 15.2%, 27.4% Post RVD (mm)2.93 ± ± 0.50 Post MLD (mm)*2.36 ± ± 0.52 Post %DS*19.9 ± ± 11.1 Acute gain (mm)**2.04 ± ± 0.62 Post TIMI 0/1, 2, 31.7%, 10.7%, 87.6%0.9%, 9.3%, 89.8% *Analysis segment, all lesions, whether stented or not; **stented lesions only; †P=0.006 † Stone GW. TCT 2008.

Aspirin and Thienopyridine Use Regular* aspirin use (%) Regular* thieno. use (%) *Taken >50% of days since last visit Antiplatelet agent use (%) P <0.001 Stone GW. TCT 2008.

Time in Months Primary Efficacy Endpoint: Ischemic TLR Ischemic TLR (%) Diff [95%CI] = -3.0% [-5.1, -0.9] HR [95%CI] = 0.59 [0.43, 0.83] P=.002 TAXUS DES (n=2257) EXPRESS BMS (n=749) % 4.5% Number at risk TAXUS DES EXPRESS BMS Stone GW. TCT 2008.

Time in Months Secondary Efficacy Endpoint: Ischemic TLR Ischemic TLR (%) Diff [95%CI] = -3.0% [-5.2, -0.7] HR [95%CI] = 0.65 [0.48, 0.89] P=.006 TAXUS DES (n=2257) EXPRESS BMS (n=749) % 5.8% Number at risk TAXUS DES EXPRESS BMS Stone GW. TCT 2008.

Time in Months Primary Safety Endpoint: Safety MACE* Safety MACE (%) Diff [95%CI] = 0.1% [-2.1, 2.4] HR [95%CI] = 1.02 [0.76, 1.36] P NI =0.01, P Sup =0.92 TAXUS DES (n=2257) EXPRESS BMS (n=749) % 8.0% Number at risk TAXUS DES EXPRESS BMS * Safety MACE = death, reinfarction, stroke, or stent thrombosis Stone GW. TCT 2008.

Time in Months One-Year All-Cause Mortality Mortality (%) HR [95%CI] = 0.99 [0.64,1.55] P=0.98 TAXUS DES (n=2257) EXPRESS BMS (n=749) % Number at risk TAXUS DES EXPRESS BMS Stone GW. TCT 2008.

Time in Months One-Year Death or Reinfarction Death or MI (%) HR [95%CI] = 0.97 [0.70,1.32] P=0.83 TAXUS DES (n=2257) EXPRESS BMS (n=749) 4 7.0% 6.8% Number at risk TAXUS DES EXPRESS BMS Stone GW. TCT 2008.

Time in Months Stent Thrombosis (ARC Definite or Probable) Stent Thrmobosis (%) HR [95%CI] = 0.92 [0.58,1.45] P=0.72 TAXUS DES (n=2238) EXPRESS BMS (n=744) 2 3.4% 3.1% Number at risk TAXUS DES EXPRESS BMS Stone GW. TCT 2008.

Stent Thrombosis Rates* TAXUS (N=2238) EXPRESS (N=744) Hazard ratio [95%CI] P Value Stent thrombosis, ≤30 days2.3%2.7%0.87 [0.52,1.46] ARC** definite1.9%2.3%0.83 [0.47,1.45] ARC probable0.5%0.4%1.11 [0.31,4.05]0.87 Stent thrombosis, >30d – 1y1.0%0.7%1.39 [0.52,3.68] ARC definite0.9%0.7%1.25 [0.47,3.35] ARC probable0.1%0%–0.42 Stent thrombosis, ≤1 year3.1%3.4%0.92 [0.58,1.45] ARC definite2.6%3.0%0.86 [0.53,1.41] ARC probable0.5%0.4%1.33 [0.38,4.73]0.65 *Kaplan-Meier estimates, **ARC= Academic Research Consortium Stone GW. TCT 2008.

One Year Composite Safety Endpoints* TAXUS (N=2257) EXPRESS (N=749) HR [95%CI]P Value Safety MACE**8.1%8.0%1.02 [0.76,1.36]0.92 Death, all-cause3.5% 0.99 [0.64,1.55] Cardiac2.4%2.7%0.90 [0.54,1.50] Non cardiac1.1%0.8%1.32 [0.54,3.22]0.55 Reinfarction3.7%4.5%0.81 [0.54,1.21] Q-wave2.0%1.9%1.07 [0.59,1.94] Non Q-wave1.8%2.7%0.68 [0.39,1.17]0.16 Stent thrombosis † 3.1%3.4%0.92 [0.58,1.45] ARC definite2.6%3.0%0.86 [0.53,1.41] ARC probable0.5%0.4%1.33 [0.38,4.73]0.65 Stroke1.0%0.7%1.52 [0.58,4.00]0.39 *Kaplan-Meier estimates; **Primary safety endpoint; †ARC (Academic Research Consortium) definite or probable Stone GW. TCT 2008.

Angiographic Follow-up TAXUS DES N=1348 EXPRESS BMS N=452 Randomized Eligible N=1308N= consecutive eligible pts assigned to 13 month angiographic FU* * Randomized in stent arm; stent procedure successful (DS <10%, TIMI-3 flow, ≤NHLBI type A peri-stent dissection); no stent thrombosis or CABG w/i 30 days 4011 Died before angio FU Died before angio FU N=942 (72.0%) N=307 (69.6%) Completed Angio FU Angio FU not performed Angio FU not performed Not received/analyzable Not received/analyzable Out of window Out of window N=911N=293 Analyzed Lesions Lesions Stone GW. TCT 2008.

Follow-up QCA TAXUS (L=1081, V=964) EXPRESS (L=332, V=302) P value TIMI flow - 0/12.8%3.6% %5.0% %91.4%0.55 FU RVD (mm)2.91 ± ± FU MLD in-stent (mm)2.36 ± ± 0.82< FU MLD in-segment (mm)2.08 ± ± 0.76< FU %DS in-stent18.8 ± ± 24.9< FU %DS in-segment28.8 ± ± 22.0< Aneurysm0.5%0.9%0.40 Ulcerated0.5%0.6%0.67 Ectasia0.7%0.9%0.73 Stone GW. TCT 2008.

Binary Analaysis Segment Restenosis at 13 Months ● Patient with Lesion Level Analysis* Binary Restenosis (%) 32 * ITT: Includes all stent randomized lesions, whether or not a stent was implanted, and whether or not non study stents were placed ** Any lesion with restenosis  per pt restenosis RR [95%CI] = 0.44 [0.33, 0.57] P < RR [95%CI] = 0.44 [0.33, 0.57] P < Major 2  endpoint Stone GW. TCT 2008.

Angiographic Late Loss at 13 Month ● Lesions with Stents Implanted P < P=0.07 Late loss (mean, mm) † 33 P=0.18 P < ± 0.42 ± 0.54 ± 0.64 ± 0.70 ± 0.56 ± 0.64 ± 0.47 ± 0.50 Stone GW. TCT 2008.

Binary Angiographic Restenosis at 13 Months ● Lesions with Stents Implanted † 34 Binary Restenosis (%) RR [95%CI] = 0.42 [0.32, 0.54] P < RR [95%CI] = 0.39 [0.29, 0.52] P < Stone GW. TCT P =0.13 P =0.42

Limitations ● Open label design –Potential bias was mitigated by high protocol procedure compliance and use of blinded clinical event adjudication committees and core laboratories ● Underpowered for stent thrombosis and death –The virtually identical rates of MACE in the TAXUS and EXPRESS groups makes it unlikely that major safety differences exist favoring either stent type at 1-year Stone GW. TCT 2008.

Conclusions ● In this large-scale, prospective, randomized trial of pts with STEMI undergoing primary stenting, the implantation of paclitaxel-eluting TAXUS stents compared to bare metal EXPRESS stents resulted in: –A significant 41% reduction in the 1-year primary efficacy endpoint of ischemia-driven TLR, and a significant 56% reduction in the 13 month major secondary efficacy endpoint of binary restenosis –Non inferior rates of the primary composite safety endpoint of all cause death, reinfarction, stent thrombosis or stroke at 1- year Stone GW. TCT 2008.

Conclusions ● The long-term safety and efficacy profile of paclitaxel-eluting TAXUS stents compared to bare metal EXPRESS stents in STEMI will be determined by the ongoing 5 year follow-up of patients randomized in the HORIZONS-AMI trial Stone GW. TCT 2008.