Should the Guidelines for Unprotected LM PCI Change

Should the Guidelines for Unprotected LM PCI Change? Yes… No… Maybe… Gregg W. Stone MD Columbia University Medical Center The Cardiovascular Research Foundation

DISCLOSURES Gregg W. Stone, MD Honoraria – Boston Scientific Corporation, Abbott Vascular

ACC/AHA Guidelines Post SYNTAX IIa 2007: LMCA PCI is reasonable in pts with class III angina and >50% LM stenosis who are not eligible for CABG IIb 2009 (post SYNTAX): Stenting the LMCA as an alternative to CABG may be considered in pts with anatomic conditions associated with a low risk of PCI procedural complications and clinical conditions that predict an increased risk of adverse surgical outcomes ACC/AHA 2009 Focused Updates for STEMI and PCI. Circulation 2009; 120: 2271– 2306

SYNTAX Eligible Patients De novo disease (n=1800) Limited Exclusion Criteria Previous interventions Acute MI with CPK>2 x Concomitant cardiac surgery Left Main Disease (isolated, +1, +2 or +3 vessels) 3 Vessel Disease (revasc all 3 vascular territories) N=705 N=1095 Serruys PW et al. NEJM 2009; 360: 961 -72

MACCE to 1 Year (primary endpoint) (All-cause death, stroke, MI, any repeat revasc) Cumulative Event Rate (%) CABG 20 (N=897) TAXUS (N=903) P=0. 0015* 12. 1% 10 0 17. 8% 0 6 Months Since Allocation Serruys PW et al. NEJM 2009; 360: 961 -72 12 ITT population

MACCE to 1 Year Left Main Subset Cumulative Event Rate (%) CABG 40 (N=348) TAXUS P=0. 44* 20 0 (N=357) 15. 8% 13. 6% 0 Event rate ± 1. 5 SE, *Fisher exact test 6 Months Since Allocation Serruys PW et al. NEJM 2009 12 ITT population

Interaction Test for Left Main Vs. No Left Main; 1 Year MACCE Hazard Ratio ± 95% CI Left main (n=705) 3 vessel ds only (n=1095) 0 0. 5 Reduced 1 1. 5 Increased P value for interaction= 0. 11 Serruys PW. TCT 2008 2 2. 5

Death (All-cause) to 1 Year Left Main Subset CABG (N=348) Cumulative Event Rate (%) 40 TAXUS (N=357) P=0. 88* 20 4. 4% 4. 2% 0 0 6 Months Since Allocation Event rate ± 1. 5 SE, *Fisher exact test 12 ITT population Serruys PW. TCT 2008

Myocardial Infarction to 1 Year Left Main Subset CABG (N=348) Cumulative Event Rate (%) 40 TAXUS (N=357) P=0. 97* 20 4. 3% 4. 1% 0 0 6 Months Since Allocation Event rate ± 1. 5 SE, *Fisher exact test 12 ITT population Serruys PW. TCT 2008

CVA (Stroke) to 1 Year Left Main Subset CABG (N=348) Cumulative Event Rate (%) 40 TAXUS (N=357) P=0. 009* 20 0 2. 7% 0. 3% 0 6 Months Since Allocation Event rate ± 1. 5 SE, *Fisher exact test 12 ITT population Serruys PW. TCT 2008

Death/CVA/MI to 1 Year Left Main Subset CABG (N=348) Cumulative Event Rate (%) 40 TAXUS (N=357) P=0. 29* 20 9. 1% 7. 0% 0 0 6 Months Since Allocation Event rate ± 1. 5 SE, *Fisher exact test 12 ITT population Serruys PW. TCT 2008

Symptomatic Graft Occlusion & Stent Thrombosis to 1 Year Left Main Subset CABG (n=348) TAXUS (n=357) Patients (%) P=0. 49 3. 7 2. 7 n=11 n=9 CABG TAXUS Serruys PW. TCT 2008 ITT population

Revascularization* to 1 Year Left Main Subset CABG (N=348) Cumulative Event Rate (%) 40 TAXUS (N=357) P=0. 02* 20 12. 0% 6. 7% 0 0 Event rate ± 1. 5 SE, *Fisher exact test 6 Months Since Allocation *Any Serruys PW. TCT 2008 12 revascularization (PCI or CABG); ITT population

MACCE to 2 Years by SYNTAX Score Tercile Left Main SYNTAX Score 33 CABG (N=149) TAXUS (N=135) Left Main Cumulative Event Rate (%) 40 CABG PCI P-value Death 4. 1% 10. 4% 0. 04 CVA 4. 2% 0. 8% 0. 08 MI 6. 1% 8. 4% 0. 48 Death, CVA or MI 11. 5% 15. 6% 0. 32 Revasc. 9. 2% 21. 8% 0. 003 P=0. 02 30 29. 7% 20 17. 8% 10 0 0 12 Months Since Allocation Cumulative KM Event Rate ± 1. 5 SE; log-rank P value 24 Site-reported data; ITT population

MACCE to 2 Years by SYNTAX Score Tercile Left Main SYNTAX Scores 0 -32 CABG (N=196) TAXUS (N=221) Left Main Cumulative Event Rate (%) 40 30 P=0. 48 20. 5% 20 CABG PCI P-value Death 7. 9% 2. 7% 0. 02 CVA 3. 3% 0. 9% 0. 09 MI 2. 6% 3. 8% 0. 59 Death, CVA or MI 12. 1% 6. 9% 0. 06 Revasc. 11. 4% 14. 3% 0. 44 18. 3% 10 0 0 12 Months Since Allocation Cumulative KM Event Rate ± 1. 5 SE; log-rank P value 24 Site-reported Data; ITT population

Conclusions Are the current guidelines for elective LM PCI (class IIb) appropriate? YES But, is LM PCI for selected lesions (e. g. SYNTAX score ≤ 32) preferred? I believe so, but…. . A large-scale, randomized trial of LM PCI vs. DES is required before the guidelines should change!

Do we really need another randomized trial of PCI vs. CABG for LM disease? • YES: SYNTAX leaves many questions unanswered 1) SYNTAX suggests (but doesn’t prove) that: - PCI and CABG for LM ds. have similar rates of death/MI/stroke - PCI may be acceptable or superior for certain LM subsets 2) Could the results be further improved with a better DES? 3) What is the optimal approach to the distal bifurcation? 4) Could IVUS and/or FFR improve outcomes?

EXCEL: Study Design Draft design 4000 pts with left main disease SYNTAX score ≤ 32 Consensus agreement by heart team Yes (N=2500) PCI and CABG registries R PCI (Xience Prime) (N=1250) No (N=1500) (limited in-hosp data) CABG (N=1250) Clinical follow-up: 30 days, 6 months, yearly through 5 years This trial design has not yet been reviewed by the US FDA and is subject to change