SAHA PCI or CABG for Left Main and Multi-Vessel Disease: when I would definitely/ maybe/never refer my patient for PCI or CABG

Transcription

1 SAHA 2017 PCI or CABG for Left Main and Multi-Vessel Disease: when I would definitely/ maybe/never refer my patient for PCI or CABG David P Taggart MD PhD FRCS FESC Professor of Cardiovascular Surgery, University of Oxford Conflicts of Interest: (i) Clinical: Cardiac Surgeon (ii) Commercial: Consultant to Medistim, Medtronic, VGS, Somahlution, Stryker) (iii) One of 25 ESC/EACTS Guidelines Writers on Myocardial Revascularization (iv) Chairman of Surgical Committee of the EXCEL trial

2 50 Years Ago: First report of SYSTEMATIC use of SV grafts for CABG ATS [Dec ] 40 Years Ago: Gruentzig reports first PTCA (AHA 1977)

3 PCI or CABG for Left Main and Multi-Vessel Disease: when I would definitely/ maybe/never refer my patient for PCI or CABG (stable or urgent patients but NOT STEMI who should get PCI) UK Most interventions are RECOMMENDED by a HEART TEAM based on 1 ESC/EACTS Guidelines (Evidence Based) 2 Potential contra-indications to the recommended intervention? 3 Patient Preferences ( what is the best treatment for me? )

4 Complex CAD should be discussed by Heart Team IC 79% 66%

5 Mutlivessel Disease (No Left main)

6 6054 patients: HR CABG vs PCI 0.73 ( ); p < JAMA 2013

7 6054 patients RR CABG vs DES for MI,RR,CVA [Siphai et al JAMA 2013] MI: RR 0.58 ( ) p<0.001 Revasc: RR 0.29 ( ) p<0.001 Stroke: RR 1.36 ( ) p=0.06 CONCLUSIONS AND RELEVANCE: In patients with multivessel coronary disease, compared with PCI, CABG leads to an unequivocal reduction in long-term mortality and myocardial infarctions and to reductions in repeat revascularizations, regardless of whether patients are diabetic or not. These findings have implications for management of such patients.

8 [IJC 2016] 5 RCTS 4563 patients Survival with CABG continues to increase past 5 years with diverging survival curves CABG results could be even better with more arterial grafts and OMT 0.9%

9 SYNTAX RCT (5 Years): 3 Vessel Disease [EHJ 2013] PCI CABG nos Death (-5.4%).006 Low <23 nos DEATH (0.9%) CVA MI Cardiac Death (-5.2%).001 MI (-7.3%) < % D+C+M Revasc CVA (+0.6%).66 nos D+C+M (-8%) <.001 Revasc (-12.8%) <.001 (i) Consistent with PM registry data > 10 years (ii) Similar rate of stroke in PCI/CABG Int % DEATH (7%) CVA MI D+C+M Revasc nos Survival: Accelerating Divergence at 5 years High >32 DEATH (9%) CVA MI D+C+M Revasc

10 Accelerating Divergence of Survival benefit for CABG [JACC 2016]

11 FREEDOM 1900 patients MVD in DM NEJM % 5.4%

12 Comparison of coronary artery bypass surgery and percutaneous coronary intervention in patients with diabetes: a meta-analysis of randomised controlled trials [Lancet Diabetes Endocrinol 2013] Verma S et al: [LANCET DIABETES and ENDOCRINOLOGY 2013] 8 trials with 3612 patients

13 Patency of RIMA to 20 years [Tatoulis et al Curr Op Cardiol 2011]

14 Left Main

15 o<90% of LMS are distal/bifurcation (very high risk of restenosis) o<90% have multivessel CAD (CABG already offers survival benefit)

16 Similar outcomes at 3 years for Death, and Composite Death/MI/Stroke but Much Greater Need for Target Vessel Revascularization with Stents MAIN-COMPARE Registry of UPLM disease in 2240 Patients: 1102 stents and 1138 CABG followed for 3 years BMS DES

17 o Competitive flow in CABG if low SYNTAX scores ie less proximal CAD o Accelerating Divergence of Survival Curves in Favour of CABG in >32 o Used to define patients in the EXCEL trial (Syntax Scores <33) SYNTAX LEFT MAIN Left SYNTAX Maintrial 705 RCT patients CIRC 5 years 2014 CIRC 2014

18 CAUTION: ONLY 3 YEARS FOLLOW-UP!!!! LM: SYNTAX < RCT patients 1000 Registry Patients NEJM 2016 At 5 years? No Difference in Stroke

19 EXCEL: The Money Shot From randomization to 30 days From 30 days to 3 years PCI (n=948) CABG (n=957) HR [95%CI] P value PCI (n=939) CABG (n=947) HR [95%CI] P value Death, stroke or MI 4.9% 7.9% 0.61 [0.42, 0.88] % 7.9% 1.44 [1.06, 1.96] Death 1.0% 1.1% 0.90 [0.37, 2.22] % 4.9% 1.44 [0.98, 2.13] Stroke 0.6% 1.3% 0.50 [0.19, 1.33] % 1.8% 1.00 [0.49, 2.05] MI 3.9% 6.2% 0.63 [0.42, 0.95] % 2.5% 1.71 [1.00, 2.93] 0.05 By 3 years CABG mortality 2.3% lower (p=0.06) BUT WITH DIVERGING SURVIVAL CURVES and NO increased risk of stroke

20 LM: 1201 RCT patients No Registry Patients Lancet 2016 Mortality 12% 9% MI 7% 2% REVASC 16% 10% STROK E 5% 2%

21 1 3 REASONS WHY CABG HAS A SURVIVAL BENEFIT OVER PCI Anatomically, atheroma is mainly located in the proximal coronary arteries Placing bypass grafts to the MID CORONARY VESSEL has TWO effects (i) Complexity of proximal CULPRIT lesion is irrelevant (ii) Over the long term offers prophylaxis against FUTURE proximal culprit lesions In contrast, PCI only treats SUITABLE localised proximal culprit lesions but has NO PROPHYLACTIC BENEFIT against new proximal disease 2 IMA elutes NO into coronary circulation reducing risk of further disease [CIRC 2007] impairs re-endothelialization, downstream endothelial function and creates pro-thrombotic milieu 3 PCI means incomplete revascularization (Hannan Circ 2006) Of 22,000 PCI 69% had incomplete revascularization >2 vessels (+/- CTO) HR for mortality 1.4 (95% CI = ) Residual SYNTAX score >8 increases mortality and MACCE (Farooq, Serruys CIRC 2013) PCI will never match the results of CABG for LM/MVD (POBA;BMS;DES)

22 Ratio Elective PCI:CABG per 100,000 population in 24 OECD countries? USA 4 x difference in European states with similar populations and economies

23 [CMAJ 2012] 17 cardiac centres in Ontario,CA 5 x difference in PCI:CABG ratio 4% of patients discussed at MDT (96% NOT) INTERPRETATION The physician performing the diagnostic catheterization and the treating hospital were strong independent predictors of the mode of revascularization. Opportunities exist to improve transparency and consistency around the decision-making process for coronary revascularization, most notably among patients with non-emergent multivessel disease.

24 WHY RECOMMENDATIONS for INTERVENTION SHOULD BE BY HEART TEAM [BMJ 2014] Only 1% said for symptoms only Only 1% correctly identified that ELECTIVE PCI was for symptoms only

25 Multi-Vessel Disease (No Left Main): ESC Guidelines 2013

26 Left Main: ESC Guidelines 2013

27 AR Gruentzig (NEJM 1979) We estimate that only about 10 to 15 per cent of candidates for bypass surgery have lesions suitable for this procedure (PCI). A prospective randomized trial will be necessary to evaluate its usefulness in comparison with surgical and medical management. Opie LH, Commerford PJ, Gersh BJ Lancet 2006; 367:69-78

28 Complex CAD should be discussed by Heart Team IC 66% 79% CABG would be even better with more arterial grafts and greater use of OMT

29 200 patient with stable angina and significant stenoses >80% and FFR <0.7 RCT of PCI vs sham invasive procedure At 6 weeks no difference in exercise test nor frequency or severity of angina

30 Summary and Conclusions: PCI vs CABG % of patients with 3 vessel CAD (SYNTAX >22) and 66% with LM (SYNTAX >32) have strong survival advantage with CABG (reduced MI and repeat revasc) by 3 years and continuing to increase past 5 years CABG is superior to PCI despite inferior OMT and 80% of all grafts being vein grafts (would be even better with widespread use of arterial grafts) In 21% of patients with 3VD (SYNTAX scores <23) and 34% with LM (SYNTAX scores <33), similar 5 year survival between CABG and PCI but less repeat revasc with CABG In contemporary trials CABG causes a non significant increase in stroke with 3VD and LM Consistent unwarranted variation in ratios of PCI:CABG between countries and within countries ABSENCE of Heart Team (using approved guidelines) results both in most elective PCI patients failing to understand its rationale and also a large number of inappropriate or wrong PCI interventions Guidelines are transparent and protect the patients (from receiving wrong interventions) and doctors (from administering wrong interventions) and should be mandatory Professional bodies should persuade statutory bodies/payers that they only reimburse interventions which are approved by the Heart Team based on official guidelines (or clear documentation why guidelines were not followed).

31 Interpreting Efficacy of PCI vs CABG: 3Key Questions 1. Are most trial patients typical of routine practice?: NO (except SYNTAX) 2. What is the duration of follow-up? USUALLY < 5 YEARS (ie Interim Analysis): few exceptions 3. Use of OPTIMAL (Guideline Based) medical therapy? CABG substantially inferior to PCI

32 Ann Thorac Surg 2006;82: [ALL REPORTED PCI EQUAL TO CABG for SURVIVAL]

33 Survival benefit of CABG increases with time (< 5 yr follow-up is interim analyses) SYNTAX 1,095 pts: EHJ 2013 FREEDOM 1,900 pts: NEJM % 5.4% ASCERT 189,793 pts: NEJM % 4.4% NY Registry 16,242 pts: ATS 2013 ACCELERATING DIVERGENCE OF SURVIVAL CURVES BEYOND 5 YEARS!!!

34 [NEJM 2016] N= % of OMT X to CABG 7.2% 8.8%

35 50 Years Ago: First report of SYSTEMATIC use of SV grafts for CABG ATS [Dec ] 1986: Loop et al SYSTEMATIC Use of ITA graft 1999: Lytle et al SYSTEMATIC Use of two ITA grafts 40 Years Ago: Gruentzig reports first PTCA (AHA 1977)

36

37 1. Are RCT patients typical of routine patients? CABG first systematically performed in 1967 and PCI in 1977 Trials of coronary revascularization by CABG or stents (PCI) >30 years CABG remains one of the most commonly performed major operations (in 2016 worldwide > 1 million CABG) No other surgical operation has ever been subjected to the same scientific scrutiny as CABG Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomised trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration* [Lancet 1994] 1 2 Summary CABG: BETTER if 2 or 3 vessel CAD involving proximal LAD CABG: NO benefit for 1 or 2 VD NOT involving proximal LAD We carried out a systematic overview using individual patient data from the seven randomised trials that have compared a strategy of initial coronary artery bypass graft (CABG) surgery mortality in patients with stable coronary heart disease (stable angina not severe enough to necessitate surgery on grounds of alone, or infarction). prognostically important clinical and angiographic risk factors were integrated to stratify patients by risk levels and the extension of survival at 10 years was examined (change in survival [SE 3 1] months in low-risk group, 5 0 [4 2] months in moderate-risk group, and 8 8 [5 4] months in high-risk group; p for trend <0 003). A strategy of initial CABG surgery is associated with lower mortality than one of medical management with delayed with Could one of this initial medical information therapy to assess be the used effects onto design a trial to show that PCI is as effective as CABG?

38 CIRC 2015 Substantially inferior OMT in CABG group mortality and MACCE

39 Q: Could industry (who want to sell stents) design an RCT to prove that stents are equal to CABG Original population:100% Angiographic proven multivessel CAD Original population:24 % Angiographic Criteria Original population:6% Cardiologist/Surgeon Agree riginal population:4% Randomized patients (1 or 2 VD + good LV) EXCLUDE 76% (outcome known to be better with CABG: severe 3VD, LM, occluded vessels, poor LV) EXCLUDE 18% (CABG can treat all lesions but stents cannot) EXCLUDE 2% (Patients Refuse Participation) (i) Can now Confidently Predict: No difference in survival (Yusuf 1994) (ii) Generalize results in publications: All patients with coronary disease (iii) Organize Sympathetic Editorials Ignore major flaws/limitations of RCT (iv) Use Trials to Underpin Guidelines

40 Surely this could not REALLY happen: ESPECIALLY in such a prestigious field of medicine dominated by EBM and RCTs (led by distinguished and famous investigators)?

41 [ATS 2006] Society ACC/AHA Circulation 2006 ESC Eur Heart J 2005 BCS Heart 2005 Summary of Guidelines Recommendations for stents vs CABG based on 15 RCTs Patients with 2 or 3 vessel disease who are otherwise eligible for CABG including diabetes NO SURGICAL OPINION RECOMMENDED all patients except diabetics with multivessel disease, unprotected left main, CTO NO SURGICAL OPINION RECOMMENDED patients to be fully informed in decisions, treatment options (GMC Good Medical Practice) NO SURGICAL OPINION RECOMMENDED almost all patients can be treated by PCI NONE RECOMMEND SURGICAL OPINION Written by 23 cardiologists 1 surgeon 46 cardiologists 0 surgeon 8 cardiologists 1 surgeon 77 cardiologists 2 surgeons surgical societies should no longer provide a token surgeon on cardiology guidelines as they are hopelessly outgunned and ineffectual against what are, in effect, exclusive cardiology dictates. If surgical opinion is genuinely to be heard, there must be comparable numbers of surgeons on writing committees

42 Taggart DP. Lancet 2009; 373: Most significantly, the randomized trials only enrolled around 5%-10% of the eligible population, the majority of whom had single or double vessel disease and normal left ventricular function [2], a group in whom it was already well established that there was no prognostic benefit of CABG [3]. By largely excluding patients with a known survival benefit from CABG (left main+/- triple vessel coronary artery disease and especially with impaired ventricular function [3]), the trials ignored the prognostic benefit of surgery in more complex coronary artery disease. Nevertheless, the inappropriate generalization of the trial results from their highly select populations to most patients with multivessel disease has been ubiquitous in the literature and has, at least in part, justified the explosive growth in PCI in developed countries. [2] Taggart DP. Thomas B. Ferguson Lecture. Coronary artery bypass grafting is still the best treatment for multivessel and left main disease, but patients need to know. Ann Thorac Surg 2006;82: [3] Yusuf S, et al. Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomised trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration. Lancet 1994 ;344:

43 [2016] Primary outcome is 10-year survival (in 2018)

44 15,583 patients followed for a mean of >9 years [CIRC 2014]

45 Has the difference in mortality between percutaneous coronary intervention and coronary artery bypass grafting in people with heart disease and diabetes changed over the years? A systematic review and meta-regression Peter Herbison, Cheuk-Kit Wong [BMJ 2015] In DM even with 3 rd generation stents CABG still has strong survival

46 Left Main (+/- Mutlivessel Disease)

47 [JACC Cardiovasc Intervention 2013] 24 studies (3 RCT) with 14,203 patients followed to 5 years DEATH (5 yr): No Difference MI: ( 1-3yrs) TVR: ( 1-5 yr) CVA: ( 1-5yr) 1 yr: 0.8% vs 2.8% 5 yr:1.7% vs 4.7% (Δ 0.9% vs 1.9%):? OMT Different from 3VD where CABG death, MI, RR and NS for CVA

48 LM: CABG BEST ONLY FOR HIGH TERCILES (>32) DEATH MI MACCE TVR

49 LEFT MAIN SYNTAX trial 705 RCT patients 5 years CIRC 2014 Different from 3VD!!

50 Politics of PCI vs CABG BMJ 2005 The current tendency of some cardiologists to exclusively investigate and treat patients with severe multivessel disease without a surgical opinion not only belittles the traditional multidisciplinary approach but ensures that the best and most balanced advice is unlikely to be consistently offered. Most importantly, by effectively denying patients the opportunity of making a fully informed choice, it falls far short of best practice.

51 Current evidence for PCI and CABG in multi-vessel and left main Documented unwarranted variations in ratio of PCI and CABG Documented inappropriate use of investigations and interventions Documented poor patient understanding of the rationale(s) for treatment

52 [EJCTS 2014] 13 x difference for lowest (0.36) and highest (4.74) although only 40 miles apart!!

53 JAMA Int Med ,225,562 angiograms: 25% ASYMPTOMATIC (range 1%-76%)

54 Adherence of Catheterization Laboratory Cardiologists to ACC/AHA Guidelines for PCI and CABG: What happens in Actual Practice? [Hannan et al Circ 2010] o16142 catheter lab patients in New York otreatment decision made by catheter lab cardiologist alone in 64% ACC/AHA Recommendation Numbers % CABG % PCI % Medical None CABG PCI <1 CABG or PCI <1 Neither Total <1 o92% of PCI procedures ad hoc (ie no time for real choice/ genuine consent) ochance of PCI increased in hospitals with PCI facilities

55 JAMA Int Med 2014 In taped interviews benefits of PCI were accurate in 5% BUT overstated in 48% [explicitly (13%) or implicitly (35%)]

56 PCI vs CABG in 2017 THREE KEY AREAS TO ADDRESS 1. EFFICACY of PCI vs CABG: Four important issues (i) Are RCT patients typical of routine practice? (ii) Duration of follow-up? (iii) SYNTAX scores? ie severity of CAD (iv) Use of OPTIMAL (guideline based) medical therapy? 2. EVIDENCE BASIS for PCI and CABG in Multivessel and Left MainDisease 3. POLITICS of PCI vs CABG (Rationale for the Heart Team)

57

58 Relative Efficacy of PCI and CABG: 4 Questions 1.Are patients enrolled in RCTs typical of routine practice? USUALLY NO. With the single exception of the all-comer SYNTAX trial, 19 other RCTs of PCI vs CABG enrolled <10% of the eligible population ie those with low severity CAD (but then generalized the results to the whole population) 2. What is the duration of follow-up? SHOULD BE AT LEAST 5 YEARS. The benefits of CABG (improved survival, reduced MI and repeat revascularization) continue to increase with time (< 5yr follow up is only an interim analyses) 3. What are the SYNTAX scores of the study population? NO SYNTAX SCORE = NO IDEA OF SEVERITY OF CAD Cannot recommend best treatment 4. Did CABG patients in RCTs receive Optimal Medical Therapy? USUALLY NO. In most trials CABG patients received substantially inferior guideline based medical therapy (OMT) leading to increased mortality and MACCE

59

60

61 ojoint Cardiology (ESC) and Cardiac Surgery (EACTS) o25 members from 13 European countries 9 non interventional cardiologists, 8 interventional cardiologists, 8 cardiac surgeons Reflects the Heart Team!!! oextensively reviewed by external referees before publication

62 14 chapters 270 references

63 Increased EDRFs (especially NO) produced by IMA results (i) in superior graft patency (ii) protects native coronary artery circulation % progression of native CAD IMA SVG Kitamura (1987) Loop (1996) Manninen (1998) Hamada (2001) Borges (2010) AVERAGE CIRC 2007 impairs re-endothelialization, creates pro-thrombotic environment impairs distal endothelial function

64 James McGill Glasgow University 1756 Sir William Osler Regius Professor of Medicine Oxford

65 [ATS 2009] Scottish forefathers!!

66 MAIN-COMPARE Registry of UPLM disease in 1102 stents and 1138 CABG BMS DES there was a trend toward higher rates of death and the composite end point in the group that received DES

67 SYNTAX 5 years 1095 RCT patients EHJ %

68 JAHA 2013 Death DEATH CARDIAC Cardiac DEATH Death REVASC