ISAR-LEFT MAIN: A Randomized Clinical Trial on Drug-Eluting Stents for Unprotected Left Main Lesions J. Mehilli, MD Deutsches Herzzentrum Technische Universität Munich Germany

Sponsor Investigator initiated trial with financial supported in part by Deutsches Herzzentrum Munich and in part by unrestricted grant from Cordis, J&J

Background Recent evidence has shown that DES are at least as safe but much more effective in reducing the need for reintervention as BMS in different patient and lesion subsets. Currently available DES might not necessarily be associated with equal performance. Potential differences in safety and efficacy between DES may be important when treating uLMCA lesions. Despite the multitude of comparative DES studies for a variety of lesion characteristics no similar studies have been performed for patients with uLMCA disease.

Objective of ISAR-LEFT MAIN …to assess the relative efficacy of the paclitaxel-eluting stent (Taxus) and sirolimus-eluting stent (Cypher) in patients with unprotected LMCA lesions.

Inclusion Criteria Patients with ischemic symptoms or evidence of myocardial ischemia in the presence of ≥50 % stenosis located in unprotected LM lesions Informed, written consent

Exclusion Criteria Age < 18 years Cardiogenic shock ST-elevation acute myocardial infarction In-stent restenosis Prior coronary artery bypass surgery Left main size >4.5 mm by visual estimation Malignancies with life expectancy <1 year Planned staged PCI within 30 days from index PCI Planned elective surgical procedure necessitating discontinuation of clopidogrel during the first 6 months Pregnancy

Primary End Point Incidence of major adverse cardiac events defined as the composite of death myocardial infarction target lesion revascularization at 1-year follow-up

Secondary End Point left main area analysis Angiographic restenosis at 6-9-month FU angiogram, defined as diameter stenosis ≥50% measured by QCA in the area from left main ostium to 5-mm proximal segments of LAD, LCx as well as of R. intermedius if the latter has a reference diameter >2 mm

Sample size calculation Hypothesis: Taxus is not inferior to Cypher in terms of major adverse cardiac events Assumptions: Incidence of MACE 20% in the Cypher group Margin of non-inferiority 8% Power of 80% -level of 0.05 Needed number of patients for each group: 302

ISAR-LEFT MAIN 607 patients with unprotected left main lesions Intracoronary Stenting and Angiographic Results: Drug-Eluting Stents for Unprotected Coronary Left Main Lesions 607 patients with unprotected left main lesions Clopidogrel 600 mg at least 2h before procedure Aspirin 500mg i.v. Paclitaxel-eluting stent (Taxus) n=302 Sirolimus-eluting stent (Cypher) n=305 Clopidogrel 2x75 mg/day until discharge 75 mg indefinitely Aspirin 200 mg/day

Follow-Up Protocol 30 d 6-9 mo. 12 mo. 600 mg Clopidogrel PCI ASS 500 mg + heparin with/out GPI (10%) or bivalirudin (15%) 30 d 6-9 mo. 12 mo. serial CK + CKMB measurements clinical follow-up (100%) repeat angiography (87%) clinical follow-up (100%)

Baseline clinical characteristics Taxus n=302 Cypher n=305 Age, years 68.8±10.3 69.3±9.3 Women, % 25 20 Art. hypertension, % 70 69 Diabetes, % 30 28 Current smoker, % 10 Hypercholesterolemia, % 78 75

Baseline clinical characteristics (con’t) Taxus n=302 Cypher n=305 Acute coronary syndrome, % 44 40 History of MI, % 25 28 Prior PCI, % 46 50 Parsonnet score 12.8±9.8 12.0±9.1 EuroSCORE 4.7±3.4 4.4±3.2 IABP, % 1

Angiographic characteristics Taxus n=302 Cypher n=305 LV ejection fraction, % 53.4±12.8 54.4±12.5 Coronary artery dominance right left balanced 78 14 8 80 11 9 Trifurcation morphology, % 25 24 Occluded dominant RCA, % 15 LMCA size, mm 3.83±0.56 3.78±0.55

Left main lesion location Ostial % Midshaft Distal Taxus n=302 Cypher n=305

Stenting technique Culotte stenting T-stenting Single stenting Taxus % Culotte stenting T-stenting Single stenting Taxus n=302 Cypher n=305

30-Day Clinical Outcome P=.48 P=.66 P=.56 P=.57 % Taxus Cypher

1-Year Clinical Outcome P=.39 P=.83 P=.47 % Taxus Cypher

1-Year Clinical Outcome (con‘t) P=.71 P=.83 % Taxus Cypher

Pnoninferiority Taxus vs. Cypher<.001 Primary Endpoint: 1-Year MACE RR 0.85; 95% CI 0.56 to 1.28 4 - 2 2 -4 6 8 - 2.2% + 95% CI = 2.7% Pre-specified margin 8% Difference in MACE Pnoninferiority Taxus vs. Cypher<.001 % Taxus Cypher MACE at 1 year

Primary Endpoint in Different Subgroups Taxus Cypher Relative Risk (95% CI) P Value for Interaction No. of Events/Total (%) Diabetes Cypher better Taxus better 2 1 0.41 Yes 12/90 (13.3) 17/86 (19.8) No 29/212 (13.6) 31/219 (14.1) 0.47 Parsonnet score >15 10/106 (9.4) 13/89 (14.6) ≤15 23/193 (11.9) 25/216 (11.6) 0.36 EUROScore ≥6 24/112 (21.4) 24/97 (24.7) <6 17/190 (8.9) 24/208 (11.5)

Secondary Endpoint: Angiographic Restenosis % Taxus Cypher

2-Year Target Lesion Revascularization P=.47 % Taxus Cypher

2-Year Stent Thrombosis % Definite stent thrombosis Probable stent thrombosis Taxus Cypher

Years after randomization Mortality 50 40 RR 1.14 (95% CI, 1.94-0.66), P=.64 30 Cumulative Incidence, % 20 Taxus 10 Cypher 1 2 Years after randomization Patients at Risk Taxus 302 294 276 215 186 Cypher 305 293 280 225 201

Years after randomization B Death, Myocardial Infarction or Reintervention 50 40 RR 0.99 (95% CI, 0.69-1.42), P=.96 30 Cumulative Incidence, % Cypher 20 Taxus 10 1 2 Patients at Risk Years after randomization Taxus 302 278 250 189 160 Cypher 305 271 252 198 176

Summary Treatment of unprotected LMCA lesions in unselected patients with DES is feasible, safe and effective up to two years Both Taxus and Cypher stents provide similar clinical and angiographic outcomes in this form of coronary artery disease