The magnitude of the survival benefit of internal thoracic artery grafting: Absolute risk reduction

Central MessageThe magnitude of the survival benefit of CABG with internal thoracic artery graft increases with time over decades.See Commentaries on pages 112 and 114. The magnitude of the survival benefit of CABG with internal thoracic artery graft increases with time over decades. See Commentaries on pages 112 and 114. Treatments for coronary artery disease (angina and myocardial infarction) have emerged in the following order: nitroglycerin in the late 1870s, coronary artery bypass grafting (CABG) in the early 1960s, percutaneous coronary intervention (PCI) in the late 1970s, and statins in the late 1980s. CABG was expected to reduce mortality and prevent myocardial infarction from the beginning, and several randomized trials and observational studies that investigated the outcomes of CABG versus drug treatment were already planned and conducted in the first decade. However, at that time, the operative mortality was higher in the CABG group, and the incidence of myocardial infarction was reported to be similar or even higher in the CABG group than in the drug group. The effectiveness of CABG surgery in preventing all-cause mortality was first reported in a meta-analysis by Yusuf and colleagues1Yusuf S. Zucker D. Peduzzi P. Fisher L.D. Takaro T. Kennedy J.W. et al.Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomized trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration.Lancet. 1994; 344: 563-570Abstract PubMed Scopus (1743) Google Scholar in 1994, approximately 30 years after the CABG was first performed. If a treatment is judged to be effective on the basis of the results of multiple randomized trials or meta-analyses, this level A evidence leads to a class I recommendation for the treatment in the guidelines. The treatment then becomes the first choice in clinical practice. However, as medical management and PCI for coronary disease have evolved over the years, the decision to actually perform an invasive procedure such as a CABG, should be not only on the basis of the “presence or absence of treatment effect,” but also on the “magnitude of treatment effect” being worthy of surgery. For example, a 20% reduction in the relative risk of death might sound quite impressive, but its effect on the patients and our surgeons' practice might nevertheless be minimal. This notion is illustrated using a concept called “number needed to treat” (NNT); the number of patients who must undergo CABG during a specific time to prevent 1 adverse outcome. It is the reverse of the absolute risk reduction (ARR). The presence or absence of a treatment effect or superiority can be assessed to some extent in observational studies, but the magnitude of the treatment effect (expressed as NNT) for a given group of patients can only be assessed in randomized trials. Therefore, for cardiac surgeons who are in charge of life and death, assessing the magnitude of the reduction in all-cause mortality (ie, NNT of all-cause mortality), is extremely important when deciding whether or not to perform surgery. In 1977, Gruentzig was the first to successfully perform PCI. In the 1980s, randomized trials and meta-analysis showed that PCI for acute myocardial infarction reduced short-term mortality by 9% to 7%, an ARR of 2% (9% − 7% = 2%).2Michels K.B. Yusuf S. Does PTCA in acute myocardial infarction affect mortality and reinfarction rates? A quantitative overview (Meta-analysis) of the randomized clinical trials.Circulation. 1995; 91: 476-485Crossref PubMed Google Scholar, 3Weaver W.D. Simes R.J. Betriu A. Grines C.L. Zijlstra F. Garcia E. et al.Comparison of primary coronary angioplasty and intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review.JAMA. 1997; 278: 2093-2098Crossref PubMed Google Scholar, 4Keeley E.C. Boura J.A. Grines C.L. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardial infarction: a quantitative review of 23 randomised trials.Lancet. 2003; 361: 13-20Abstract Full Text Full Text PDF PubMed Scopus (3481) Google Scholar The inverse of this ARR (that is, 1 divided by the ARR) is equal to the number of such patients cardiologists would have to treat to prevent 1 death—the NNT. In this case, the cardiologist would have to treat 50 such patients to save a single life (1/0.02 = 50). Statins were approved by the US Food and Drug Administration in 1987, and as early as 1994, a randomized trial (4S) reported that statins were effective in preventing myocardial infarction and improving life expectancy.5Scandinavian Simvastatin Survival Study GroupRandomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S).Lancet. 1994; 344: 1383-1389Abstract PubMed Scopus (11157) Google Scholar In 2005, a meta-analysis that combined 14 randomized trials since the 4S trial (totaling 90,056 patients) reported that statin therapy for secondary prevention improved the 4.7-year survival rate by 1.2% corresponding to a 4.7-year NTT of 83.6Cholesterol Treatment Trialists' (CTT) CollaboratorsEfficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90 056 participants in 14 randomised trials of statins.Lancet. 2005; 366: 1267-1278Abstract Full Text Full Text PDF PubMed Scopus (5580) Google Scholar On the basis of the results of these randomized trials and a meta-analysis, global guidelines classify PCI for acute myocardial infarction and statins for hypercholesterolemia recommended as class I with level of evidence A. The first meta-analysis was by Yusuf and colleagues1Yusuf S. Zucker D. Peduzzi P. Fisher L.D. Takaro T. Kennedy J.W. et al.Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomized trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration.Lancet. 1994; 344: 563-570Abstract PubMed Scopus (1743) Google Scholar in 1994, who reported that the strategy of initial CABG surgery is effective in preventing all-cause mortality (Table 1 and Figure 1). In the report by Yusuf and colleagues the 30-day mortality rate in the CABG group was relatively high at 3.2%. Nonetheless, CABG reduced all-cause mortality by 5.6% at 5 years, 5.9% at 7 years, and 4.1% at 10 years (Figure 2). The magnitude of this treatment effect was greatest at approximately 7 years postoperatively, then decreased, and by 12 years, the effect had disappeared. The level A evidence from the meta-analysis by Yusuf and colleagues has long been the fundamental reason for cardiac surgeons to perform CABG. To date, the European and Japanese guidelines have given class I recommendations for CABG in all stable coronary artery disease patients with left main trunk (LMT) lesions or/and proximal left anterior descending lesions. However, there are now critics and detractors of the all-cause mortality benefit of CABG as reported by Yusuf and colleagues. The criticism comes mainly from the cardiologist community. They argue that this report is a synthesis of randomized trials conducted at a time when statins and PCI were not yet actively performed, and that CABG might no longer be effective in preventing all-cause mortality, or even if it is, it might be less effective. The counter argument to this criticism comes from the cardiac surgeon community: Yusuf and colleagues reported that 90% of CABGs used only vein grafts, so the effect was small after 7 years and disappeared after 12 years. However, modern CABG uses at least 1 internal thoracic artery (ITA) graft in principle, so the effect might last much longer.Table 1ARR and NNT of CABG to prevent 1 death across landmark trials in patients with coronary artery diseaseStudyCoronary artery diseasePatient nAll-cause mortalityARR, %NNTP valueNon-initial CABG strategy, %Initial CABG strategy, %Years after CABGSVG and pre-statin eraYusuf and colleagues1Yusuf S. Zucker D. Peduzzi P. Fisher L.D. Takaro T. Kennedy J.W. et al.Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomized trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration.Lancet. 1994; 344: 563-570Abstract PubMed Scopus (1743) Google Scholar meta-analysis3VD 51%, 2VD 32%, 1VD 10%, LMD 7%, DM 10%, low EF 7%264915.810.255.618<.0001 21.715.875.917<.001 30.526.4104.124.03ITA and statin eraFREEDOM7The FREEDOM Trial InvestigatorsStrategies for multivessel revascularization in patients with diabetes.N Engl J Med. 2012; 367: 2375-2384Crossref PubMed Scopus (1298) Google ScholarDiabetic multivessel disease190016.310.955.419.049 STICH8STICH InvestigatorsCoronary-artery bypass surgery in patients with left ventricular dysfunction.N Engl J Med. 2011; 364: 1607-1616Crossref PubMed Scopus (808) Google Scholar,9STICH InvestigatorsCoronary-artery bypass surgery in patients with ischemic cardiomyopathy.N Engl J Med. 2016; 374: 1511-1520Crossref PubMed Scopus (472) Google ScholarLow EF (≤35%)1212413655.020.12 66.158.9107.214.02 SYNTAX10Head S.J. Davierwala P.M. Serruys P.W. Redwood S.R. Colombo A. Mack M.J. et al.Coronary artery bypass grafting vs. percutaneous coronary intervention for patients with three-vessel disease: final five-year follow-up of the SYNTAX trial.Eur Heart J. 2014; 35: 2821-2830Crossref PubMed Scopus (253) Google Scholar,11The SYNTAX Extended Survival InvestigatorsPercutaneous coronary intervention versus coronary artery bypass grafting in patients with three-vessel or left main coronary artery disease: 10-year follow-up of the multicentre randomised controlled SYNTAX trial.Lancet. 2019; 394: 1325-1334Abstract Full Text Full Text PDF PubMed Scopus (238) Google Scholar3VD109514.69.255.419.006 28.021.0107.014<.05 EXCEL12The EXCEL Trial InvestigatorsEverolimus-eluting stents or bypass surgery for left main coronary artery disease.N Engl J Med. 2016; 375: 2223-2235Crossref PubMed Scopus (674) Google Scholar,13The EXCEL Trial InvestigatorsFive-year outcomes after PCI or CABG for left main coronary disease.N Engl J Med. 2019; 381: 1820-1830Crossref PubMed Scopus (307) Google ScholarLMD (SYNTAX Score <33)19058.25.932.343.11 13.09.953.132<.05ARR, Absolute risk reduction; NNT, number needed to treat; CABG, coronary artery bypass grafting; SVG, saphenous vein graft; VD, vessel disease; LMD, left main disease; DM, diabetes mellitus; EF, ejection fraction; ITA, Intenal Thoracic Artery; FREEDOM, Future Revascularization Evaluation in Patients with Diabetes Mellitus: Optimal Management of Multivessel Disease; STICH, Surgical Treatment for Ischemic Heart Failure; SYNTAX, SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery; EXCEL, Evaluation of XIENCE versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization trial. Open table in a new tab Figure 2Absolute risk reduction (ARR) of all-cause death for saphenous vein coronary artery bypass and consequent number needed to treat (NNT) in the meta-analysis from Yusuf and colleagues.1Yusuf S. Zucker D. Peduzzi P. Fisher L.D. Takaro T. Kennedy J.W. et al.Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomized trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration.Lancet. 1994; 344: 563-570Abstract PubMed Scopus (1743) Google Scholar In the era of SVG, coronary artery bypass grafting (CABG) reduced all-cause mortality by 5.6% at 5 years, 5.9% at 7 years, and 4.1% at 10 years. The magnitude of the survival benefit was greatest at approximately 7 years postoperatively, then decreased, and by 12 years, the effect had disappeared. SVG, Saphenous vein graft.View Large Image Figure ViewerDownload Hi-res image Download (PPT) ARR, Absolute risk reduction; NNT, number needed to treat; CABG, coronary artery bypass grafting; SVG, saphenous vein graft; VD, vessel disease; LMD, left main disease; DM, diabetes mellitus; EF, ejection fraction; ITA, Intenal Thoracic Artery; FREEDOM, Future Revascularization Evaluation in Patients with Diabetes Mellitus: Optimal Management of Multivessel Disease; STICH, Surgical Treatment for Ischemic Heart Failure; SYNTAX, SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery; EXCEL, Evaluation of XIENCE versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization trial. Yusuf and colleagues also noted that the groups with the greatest reduction in all-cause mortality with CABG were those with LMT disease, low cardiac function, and 3-vessel disease. The Bypass Angioplasty Revascularization Investigation (BARI) trial, published in 1996, reported that those with diabetic multivessel disease was a group that showed a significant benefit in preventing all-cause mortality with CABG.14The BARI InvestigatorsComparison of coronary bypass surgery with angioplasty in patients with multivessel disease.N Engl J Med. 1996; 335: 217-225Crossref PubMed Scopus (1541) Google Scholar Since the report from Yusuf and colleagues, the development of aggressive medical therapy and PCI with stents has resulted in the following groups of coronary lesions being prioritized for CABG in clinical practice: LMT disease, 3-vessel disease, and diabetic multivessel disease. Furthermore, with the advent of PCI using drug-eluting stents, randomized trials for these lesion groups were planned in the 2000s, and the results of these trials have been reported since the 2010s. The Future Revascularization Evaluation in Patients with Diabetes Mellitus: Optimal Management of Multivessel Disease (FREEDOM) trial was reported in 2021, which randomized patients who had diabetes and multivessel coronary disease to PCI and CABG.7The FREEDOM Trial InvestigatorsStrategies for multivessel revascularization in patients with diabetes.N Engl J Med. 2012; 367: 2375-2384Crossref PubMed Scopus (1298) Google Scholar In the CABG group 82.6% of the patients were receiving statins and ITA was used in 94.4% of the patients. The 5-year mortality rate was 10.9% in the CABG group and 16.3% in the PCI group (P = .049); CABG reduced all-cause mortality by 5.4% over 5 years, corresponding to a 5-year NTT of 18.5 (= 100/5.4). The Surgical Treatment for Ischemic Heart Failure (STICH) trial was a randomized trial of patients with low cardiac function with an ejection fraction of <35%.8STICH InvestigatorsCoronary-artery bypass surgery in patients with left ventricular dysfunction.N Engl J Med. 2011; 364: 1607-1616Crossref PubMed Scopus (808) Google Scholar,9STICH InvestigatorsCoronary-artery bypass surgery in patients with ischemic cardiomyopathy.N Engl J Med. 2016; 374: 1511-1520Crossref PubMed Scopus (472) Google Scholar It compared a medical therapy group (602 patients) with a CABG group (610 patients). Five-year follow-up was reported in 2011, which showed no significant difference in all-cause mortality. However, the 10-year results reported in 2016 showed that the all-cause mortality was 66.1% in the drug group versus 58.9% in the CABG group (P = .02). CABG reduced all-cause mortality by 7.2% over 10 years, corresponding to a 10-year NNT of 13.9 (= 100/7.2). The SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery (SYNTAX) trial reported in 2014 showed that in patients with 3-vessel disease, 5-year all-cause mortality was 14.2% for PCI with drug-eluting stents and 9.2% for CABG (P < .05); CABG reduced all-cause mortality by 5.0% over 5 years, corresponding to a 5-year NNT of 20 (= 100/5.0).10Head S.J. Davierwala P.M. Serruys P.W. Redwood S.R. Colombo A. Mack M.J. et al.Coronary artery bypass grafting vs. percutaneous coronary intervention for patients with three-vessel disease: final five-year follow-up of the SYNTAX trial.Eur Heart J. 2014; 35: 2821-2830Crossref PubMed Scopus (253) Google Scholar ITA grafting was done in 97% of patients in the CABG group. In 2019, the 10-year results of the SYNTAX trial were reported: in patients with 3-vessel disease, 10-year all-cause mortality was 28% for PCI and 21% for CABG; CABG resulted in a 7% reduction in all-cause mortality over 10 years, corresponding to a 10-year NNT of 14.2 (= 100/7).11The SYNTAX Extended Survival InvestigatorsPercutaneous coronary intervention versus coronary artery bypass grafting in patients with three-vessel or left main coronary artery disease: 10-year follow-up of the multicentre randomised controlled SYNTAX trial.Lancet. 2019; 394: 1325-1334Abstract Full Text Full Text PDF PubMed Scopus (238) Google Scholar In the Evaluation of XIENCE versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization (EXCEL) trial of patients with LMT disease with a SYNTAX score of 32 or less, the reduction in all-cause mortality with CABG was 2.3% at years PCI 5.9% and at 5 years PCI CABG; < EXCEL Trial InvestigatorsEverolimus-eluting stents or bypass surgery for left main coronary artery disease.N Engl J Med. 2016; 375: 2223-2235Crossref PubMed Scopus (674) Google Scholar,13The EXCEL Trial InvestigatorsFive-year outcomes after PCI or CABG for left main coronary disease.N Engl J Med. 2019; 381: 1820-1830Crossref PubMed Scopus (307) Google Scholar On the basis of the results of the randomized trials CABG reduced all-cause mortality in patients with LMT disease, low cardiac function, 3-vessel disease, and diabetic multivessel disease. Therefore, can that CABG has a effect on all-cause mortality of evidence even in the era of aggressive drug therapy and PCI with drug-eluting the magnitude of the reduction in all-cause mortality with CABG using saphenous vein graft was at 10 years than at 5 years, but the magnitude of the effect with CABG using the ITA graft was at 10 years than at 5 years. PCI was in 1977, the or LMT lesions was as early as the when the of CABG in preventing myocardial infarction had not yet been by randomized trials. is as the presence of a bypass graft to the left coronary In and J. E. J. A. et safety and of percutaneous coronary intervention with and coronary artery bypass surgery for multivessel coronary artery disease a meta-analysis with 5-year data from the and PubMed Scopus Google Scholar reported a meta-analysis of randomized trials that compared PCI with stents and CABG, and reported that the incidence of myocardial infarction was similar in The first randomized trial to the effect of CABG was the BARI trial, which was reported in D. S. S. et Bypass Angioplasty Revascularization Investigation Diabetes randomized trial of treatment in diabetes with stable ischemic heart disease: of treatment strategy on cardiac mortality and myocardial PubMed Scopus Google Scholar The that the lesion was than 20 years the BARI trial reported the effect of CABG, and that it is used that the effect of bypass is to the cardiologist community. In the BARI trial aggressive medical therapy was compared with aggressive medical therapy CABG in diabetic patients who were for CABG. The of this the statin rate was and cholesterol was reduced to in and of patients in the aggressive medical therapy group coronary revascularization during the follow-up with statin at and cholesterol reduction to the strategy of initial CABG surgery reduced the 5-year incidence of myocardial infarction by an = corresponding to a 5-year NNT of (= The Coronary Artery Revascularization in Diabetes A. R.J. J. et of percutaneous coronary intervention with coronary artery bypass grafting in diabetic patients results of the Artery Revascularization in PubMed Scopus Google Scholar FREEDOM Trial InvestigatorsStrategies for multivessel revascularization in patients with diabetes.N Engl J Med. 2012; 367: 2375-2384Crossref PubMed Scopus (1298) Google Scholar and E. Colombo A. et al.Coronary artery bypass graft surgery versus percutaneous coronary intervention in patients with three-vessel disease and left main coronary disease: 5-year follow-up of the clinical SYNTAX trial.Lancet. 381: Full Text Full Text PDF PubMed Scopus Google Scholar trials also reported of myocardial infarction in the CABG group. The left main InvestigatorsPercutaneous coronary angioplasty versus coronary artery bypass grafting in the treatment of left main 5-year outcomes from the trial.Lancet. Full Text Full Text PDF PubMed Scopus Google Scholar and EXCEL Trial InvestigatorsFive-year outcomes after PCI or CABG for left main coronary disease.N Engl J Med. 2019; 381: 1820-1830Crossref PubMed Scopus (307) Google Scholar trials of LMT disease also reported low of myocardial infarction during for myocardial infarction in the CABG group. Therefore, can that CABG has a effect on myocardial infarction of evidence even in the era of aggressive medical therapy and PCI with drug-eluting In patients with stable coronary artery disease, the results of randomized trials and meta-analysis Coronary Angioplasty and PCI with stents, as as stents and drug-eluting stents, reported that PCI has improved with in but not all-cause mortality and myocardial P. for of coronary A Med. 2003; PubMed Scopus Google A. coronary for coronary artery disease: a quantitative and a Full Text Full Text PDF PubMed Scopus Google Scholar The results of the BARI D. S. S. et Bypass Angioplasty Revascularization Investigation Diabetes randomized trial of treatment in diabetes with stable ischemic heart disease: of treatment strategy on cardiac mortality and myocardial PubMed Scopus Google Scholar and Revascularization and Drug Evaluation P.M. J.A. et al.Effect of PCI on survival in patients with stable ischemic heart disease.N Engl J Med. PubMed Scopus Google Scholar trials also showed that PCI not all-cause mortality or myocardial infarction compared with medical randomized trials were and patients to the medical also PCI during the follow-up if Therefore, that PCI is but can that the strategy of PCI first was not to the strategy of first then PCI if needed during In to the results of randomized trials on CABG, such as the meta-analysis by Yusuf and Treatment Trialists' (CTT) CollaboratorsEfficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90 056 participants in 14 randomised trials of statins.Lancet. 2005; 366: 1267-1278Abstract Full Text Full Text PDF PubMed Scopus (5580) Google Scholar showed that of patients in the medical group CABG during follow-up (ie, yet the group to CABG had In the BARI trial, which was the first to evidence of the effect of CABG on myocardial of the patients in the drug treatment group coronary revascularization during D. S. S. et Bypass Angioplasty Revascularization Investigation Diabetes randomized trial of treatment in diabetes with stable ischemic heart disease: of treatment strategy on cardiac mortality and myocardial PubMed Scopus Google Scholar The trial also showed that of patients in the treatment group CABG 5 years and 20% 10 InvestigatorsCoronary-artery bypass surgery in patients with left ventricular dysfunction.N Engl J Med. 2011; 364: 1607-1616Crossref PubMed Scopus (808) Google Scholar,9STICH InvestigatorsCoronary-artery bypass surgery in patients with ischemic cardiomyopathy.N Engl J Med. 2016; 374: 1511-1520Crossref PubMed Scopus (472) Google Scholar The treatment effect of CABG is so of if surgery is it must be done as as or the effect be In 1996, and A. K.B. Coronary bypass surgery with on survival over a Engl J Med. 1996; PubMed Scopus Google Scholar compared the survival of the CABG group with ITA patients) and the group patients) in an observational and reported that the group with ITA had survival for years after surgery. The of the life in the groups after the which is to be the life of the SVG, and to the In and S. J. et of diabetes on coronary artery bypass graft PubMed Scopus Google Scholar reported graft for patients who CABG at the over time at 5 years, at 10 years, and at 20 ITA graft of were and for 20 years In and S.J. J. E. E. et after coronary artery bypass grafting versus percutaneous coronary intervention with for coronary artery disease: a of Full Text Full Text PDF PubMed Scopus Google Scholar reported the results of a of data from randomized trials patients). They reported a reduction in 5-year all-cause mortality with CABG grafting across all coronary artery disease disease 3-vessel disease LMT disease corresponding to a 5-year NNT of 50 (= This is level A The cardiologist this 5-year NNT of 50 to be small and CABG over PCI only for coronary lesions in which CABG 5-year all-cause mortality by or 3-vessel disease, diabetic multivessel and the cardiac surgeon also not to the of this 2% However, the era reported by Yusuf and colleagues when CABG was using S. Zucker D. Peduzzi P. Fisher L.D. Takaro T. Kennedy J.W. et al.Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomized trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration.Lancet. 1994; 344: 563-570Abstract PubMed Scopus (1743) Google Scholar ITA is used in modern CABG, so the reduction in all-cause mortality at 5 S.J. J. E. E. et after coronary artery bypass grafting versus percutaneous coronary intervention with for coronary artery disease: a of Full Text Full Text PDF PubMed Scopus Google Scholar is The magnitude of the expected reduction of all-cause mortality with ITA grafting is in the Central and 2% at 5 years is it becomes 7% at years and at 20 years. that the magnitude of the survival benefit of CABG with ITA increases over as in Figure is that the strategy of initial CABG surgery might be for with longer life as as 3-vessel disease, and low cardiac In the of the I would to that be an important in the for CABG. The reported no of The and to of and to or for which have a of The and of this have no of