Coronary Artery Stenosis. Insight from MAIN-COMPARE Study

Transcription

1 PCI for Unprotected Left Main Coronary Artery Stenosis Insight from MAIN-COMPARE Study Young-Hak Kim, MD, PhD Cardiac Center, University of Ulsan College of Medicine, Asan Medical Center

2 Current Practice for LM Stenosis What is a real-world?

3 What is your plan? 76 Yr Male, Normal EF, LM with 3 Vessel ds

4 What is your plan? 58 Yr Male, Normal EF, Isolated LM

5 What are your considerations? Ejection fraction Age Right coronary involvement Vascular size of LM Lesion location at LM Extra-LM involvement e Long-term data compared with CABG Economy Stentable or not?

6 LM-PCI has already been popular in Korea! Motivation from pioneer Excellent educational course of training Skilled hands Patients preference Cumulative evidence!

7 MAIN-COMPARE Registry N Engl J Med 2008;358.

8 MAIN-COMPARE Study Stenting (BMS or DES) vs. CABG January, 2000 Wave I LMCA disease BMS (N=318) CABG (N=448) Second quarter (May), 2003 Wave II LMCA disease June, 2006 DES (N=784) CABG (N=690) Total (N=2240) PCI (N=1102) CABG(N=1138)

9 Prevalence of PCI or CABG in Korea Unpublished data from MAIN-COMPARE No. of patients PCI CABG

10 Current Practice for LM Stenosis PCI is now a preferable alternative to bypass surgery for unprotected LM stenosis in Korea. With the introduction of DES, PCI for LM stenosis has been a more popular strategy as compared with CABG. In many institutions, PCI seems to be the first line therapy for LM stenosis with STENTABLE angiographic morphology.

11 Safety of Unprotected LM Stenosis

12 Adjustment for Confounder in Non-randomized Study Propensity Score Methodology Replace the collection of confounding covariates with one scalar function of these covariates: the propensity score. Age Gender Location. 1 composite covariate: Propensity Score

13 Creation of Propensity Score Dependent variable : Stent / CABG Independent variable : Demographic characteristics, Cardiac or coexisting conditions and Angiographic characteristics logistic regression

14 Creation of Propensity Score Dependent variable : Stent / CABG Independent variable : Demographic characteristics, Cardiac or coexisting conditions and Angiographic characteristics logistic regression

15 Comparison of Risk Profiles No Overlap : Completely l Different Group Propensity 1 score 0 PCI CABG

16 Comparison of Risk Profiles Propensity 1 score Overlap in Clinical i l Risks 0 PCI CABG

17 Propensity-Match in MAIN-COMPARE Propensity 1 score 596 Pts (unmatched) 542 Pairs (matched) 560 Pts (unmatched) 0 A half of PCI and CABG patients was at a comparable risk in clinical and angiographic g PCI presentations. CABG Our patients treated with PCI are not so simple.

18 Comparable Incidence of Death Propensity-Matched t Populations Overall (542 pairs) BMS (207 pairs) P=0.91 P=0.45 DES (396 pairs) P=0.26

19 Comparable Incidence of Death/QMI/Stroke Propensity-Matched t Populations Overall (542 pairs) BMS (207 pairs) P=0.59 P=0.61 DES (396 pairs) P=0.16

20 Hazard Ratios for Safety Outcomes Outcome Overall BMS DES (542 pairs) (207 pairs) (396 pairs) HR HR HR P P (95% CI) (95% CI) (95% CI) P Death 1.18 ( ) ( ) ( ) 0.26 Death, Q-MI, or stroke ( ) ( ) ( ) 0.15 *HR are for the stenting group, as compared with CABG group

21 Incidence of Death Matched and Unmatched Populations Matched group CABG 92% Un-matched group 94% PCI 92% 90%

22 Incidence of Death Matched and Unmatched Populations Propensity 1 score 560 Pts 94% (unmatched) 90% 92% 542 Pairs 92% (matched) 596 Pts (unmatched) 0 PCI CABG

23 Safety of PCI for Unprotected LM Stenosis PCI for unprotected LM stenosis was comparably safe to CABG for patients at a low or moderate clinical risk. The risk of mortality was more dependent on the baseline clinical risk of patients than the type of treatment.

24 Efficacy of Unprotected LM Stenosis

25 Lower Incidence of TVR By CABG Propensity-Matched t Populations Overall (542 pairs) BMS (207 pairs) P<0.001 P<0.001 DES (396 pairs) P<0.001

26 PCI had a lower efficacy? Outcome Overall BMS DES (542 pairs) (207 pairs) (396 pairs) HR HR HR P P (95% CI) (95% CI) (95% CI) P TVR 4.76 ( ) < ( ) < ( ) < *HR are for the stenting group, as compared with CABG group

27 Outcomes for Representing Efficacy For Approval of Drug or Devices by FDA Class End points Safety data requirement Effect size expected Minimum strength of evidence needed A Mortality or stroke Minimali Small P < B ESRD, MI, CHF Intermediate Small P< C1 Hospitalization, procedure Intermediate to considerable Moderate p < C2 X Symptoms, QOL, Exercise capacity Surrogate markers such as blood pressure, lipidid Considerable Considerable Substantial Substantial P < 0.01 or P < 0.05 in 2 distinct studies P < 0.05 in 2 distinct studies Jolicoeur EM, Am Heart J 2008;155:435

28 Type of Repeat Revascularization % CABG Pts (N=48) PCI Pts (N=136) PCI 4 CABG 22

29 PCI Compared with CABG In Diabetic Patients Death Q MI TVR CVA Death/Q-MI/TVR Death/Q-MI/TVR/CVA Crude Adjusted HR (95% CI) 0.70 ( ) 0.38 ( ) ( ) ( ) 1.53 ( ) 1.41 ( ) P Multivariable adjusted Adjusted HR (95% CI) 0.78 ( ) 0.30 ( ) ( ) ( ) 1.64 ( ) 1.95 ( ) < P <

30 PCI Compared with CABG In Non-bifurcation Outcome Crude Multivariable adjusted HR P HR (95% CI) (95% CI) Death ( ) ( ) CVA 2.0 ( ) ( ) P TVR ( ) ( ) Death, Q-MI, or CVA ( ) ( ) Death, Q-MI, CVA, or TVR ( ) ( ) 0.111

31 PCI Compared with CABG Isolated LM Death Q wave MI TVR MACE PCI better CABG better Adjusted Hazard Ratio (95% CI)

32 DES Compared with BMS Outcome Crude Hazard Ratio (95% CI) P Adjusted for propensity Hazard Ratio (95% CI) P Death 0.93 ( ) ( ) 0.87 MI 122( ( ) ( ( ) 56) TLR 0.39 ( ) < ( ) <0.001 TVR 0.55 ( ) ( ) <0.001 Death/MI 1.04 ( ) ( ) 0.47 Death/MI/TLR 0.84 ( ) ( ) 0.03 Death/MI/TVR 084( ( ) 09) ( ( ) 89)

33 Efficacy of PCI for Unprotected LM Stenosis The risk of repeat revascularization is lower with use of CABG than PCI. However, repeat revascularization is one of outcomes assessing the efficacy of a certain strategy. The majority of restenosis at the LM was treated with PCI. The safety and efficacy of PCI was consistently approved in diverse subgroups of patients.

34 Is it enough to change the guideline? Limitations of MAIN-COMPARE It is evidenced that PCI is a good alternative to surgery. But, it is not sufficient enough to change guideline because

35 Baseline Characteristics of Propensity-Matched Patients (542 pairs) Stents CABG Variable (n=542) (n=542) Previous coronary angioplasty Previous myocardial infarction Previous congestive heart failure Chronic obstructive pulmonary disease Cerebrovascular disease Peripheral vascular disease Renal failure Ejection fraction (%) Median Interquartile range

36 Baseline Characteristics of Propensity-Matched Patients (542 pairs) Variable Electrocardiographic findings Stents (n=542) CABG (n=542) Sinus rhythm Atrial fibrillation Other Clinical indication (%) Silent ischemia Chronic stable angina Unstable angina NSTEMI

37 Feasibility of PCI for High-Risk Patients t? Propensity score 1 0 PCI CABG

38 Angiographic Predictors by Detailed Independent Analysis? Stents CABG Variable (n=542) (n=542) Angiographic characteristics (%) Extent of diseased vessel Left etmain only Left main plus single-vessel disease Left main plus double-vessel disease Left main plus triple-vessel disease Right coronary artery disease Restenotic lesion

39 Incidence of Death Longer F/U in DES patients is required

40 12 Sites in Korea for MAIN-COMPARE by experienced experts in tertiary ti hospitals. - Asan Medical Center, Seoul, Korea - Kangnam St Mary s Hospital, Seoul, Korea - Yoido St Mary s Hospital, Seoul, Korea - Kyungpook National University Hospital, Daegu, Korea - Gachon University Gil Medical Center, Incheon, Korea - Seoul National University Hospital, Seoul, Korea - Seoul National University Bundang Hospital, Seongnam, Korea - Samsung Medical Center, Seoul, Korea - Ajou University Hospital, Suwon, Korea - Yonsei University Medical Center, Seoul, Korea - Chonnam National Univeristy Hospital, Gwangju, Korea - Chung-Nam University Hospital, Daejon, Korea

41 Definitions for LM Study? Where is LM? What is a significant LM stenosis? What is isolated LM disease? What is TLR of LM? What is TVR of LM? How can we measure quantitative angiography at LM?

42 Guideline Level of Evidence from ACC/AHA Level A - Sufficient evidence from multiple RCT and meta-analysis Level B - Limited evidence from single RCT or non- randomized trials Level C - Only expert opinion, case studies

43 Future Studies to Change the Guideline We need randomized studies performed - Across the countries - With a standardized protocol - With an adequate calculation of statistical power - For very long-term F/U - Under the control of experienced organization

44

45 What we need to know further Who can be more benefited by PCI? Who can be more benefited by CABG? Who can be comparably treated with either PCI or CABG?