Author + information
- Received October 22, 2018
- Revision received December 7, 2018
- Accepted December 31, 2018
- Published online April 15, 2019.
- Rodrigo Modolo, MDa,b@RodrigoModolo5,
- Ply Chichareon, MDa,c,
- Norihiro Kogame, MDa,
- Ovidiu Dressler, MDd,
- Aaron Crowley, MAd,
- Ori Ben-Yehuda, MDd,
- John Puskas, MDe,
- Adrian Banning, MDf,
- David P. Taggart, MDf,
- A. Pieter Kappetein, MD, PhDg,
- Joseph A. Sabik, MDh,
- Yoshinobu Onuma, MD, PhDi,j,
- Gregg W. Stone, MDk and
- Patrick W. Serruys, MD, PhDl,∗ ()
- aDepartment of Cardiology, Academic Medical Center, University of Amsterdam, Amsterdam, the Netherlands
- bDepartment of Internal Medicine, Cardiology Division, University of Campinas (UNICAMP), Campinas, Brazil
- cDivision of Cardiology, Department of Internal Medicine, Faculty of Medicine, Prince of Songkla University, Songkhla, Thailand
- dCardiovascular Research Foundation, New York, New York
- eMount Sinai Heart at Mount Sinai St. Luke’s, New York, New York
- fOxford Heart Centre, Oxford University Hospitals NHS Foundation Trust, Oxford, United Kingdom
- gMedtronic, Dublin, Ireland
- hDepartment of Surgery, UH Cleveland Medical Center, Cleveland, Ohio
- iErasmus Medical Center, Rotterdam, the Netherlands
- jCardialysis Clinical Trials Management and Core Laboratories, Rotterdam, the Netherlands
- kNewYork-Presbyterian Hospital, Columbia University Medical Center, and the Cardiovascular Research Foundation, New York, New York
- lDepartment of Cardiology, Imperial College of London, London, United Kingdom
- ↵∗Address for correspondence:
Prof. Patrick W. Serruys, Erasmus University Medical Center, P.O. Box 2125, 3000 CC Rotterdam, the Netherlands.
Background Although results of percutaneous coronary intervention (PCI) have been steadily improving, whether surgical outcomes have improved over time is not fully elucidated.
Objectives This study sought to compare the current outcomes of patients undergoing coronary artery bypass grafting (CABG) with prior surgical results, in the context of randomized trials including the left main (LM) coronary artery stem.
Methods The authors performed a propensity-matched analysis of patients randomized to CABG in the SYNTAX (Synergy Between PCI With Taxus and Cardiac Surgery) (enrollment period 2005 to 2007) and EXCEL (Evaluation of XIENCE Versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization) (enrollment period 2010 to 2014) trials. All patients had left main (LM) disease with or without multivessel disease. Adjustment was based on 15 clinical and angiographic variables, including anatomic SYNTAX score, with a 2:1 ratio for the EXCEL and SYNTAX trials, collectively analyzing 909 subjects (n = 580 and n = 329, respectively). The primary endpoint was the composite of all-cause death, myocardial infarction (MI), stroke, or ischemia-driven revascularization at 3 years.
Results Baseline characteristics, anatomic SYNTAX score, number and types of grafts, and duration of hospitalization for the procedures were similar in both groups. CABG procedures in the EXCEL compared with the SYNTAX trial were more often off-pump (29.6% vs. 15.4%; p < 0.001), and guideline-directed medical therapies were used more frequently in the EXCEL surgical cohort. The primary endpoint occurred in 14.0% and 20.9% (p = 0.008) of patients in the EXCEL and SYNTAX trials, respectively. With the exception of MI (4.1% vs. 3.7%), all nonhierarchical events tended to contribute to the improved outcomes in the more recent trial: all-cause death (5.5% vs. 8.5%), stroke (3.1% vs. 5.1%), and ischemia-driven revascularization (7.1% vs. 9.4%) in the EXCEL and SYNTAX trials, respectively.
Conclusions Over a 5- to 7-year period, significant improvement in event-free survival after surgical revascularization for LM disease at 3 years was noted between the SYNTAX and EXCEL trials, consistent with improving results with cardiac surgery over time. (Synergy Between PCI With Taxus and Cardiac Surgery [SYNTAX]; NCT00114972; Evaluation of XIENCE Versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization [EXCEL]; NCT01205776)
Coronary artery bypass graft (CABG) surgery was introduced >50 years ago. In 1970, Favaloro et al. (1) reported the largest series of venous grafting, demonstrating its efficacy with low morbidity and mortality. Since then, CABG has been established as an effective treatment for coronary artery disease (CAD) not responding to pharmacological treatment, and is currently the preferred modality to treat complex CAD, with lower rates of death, myocardial infarction (MI), and repeat revascularization compared with percutaneous coronary intervention (PCI) (2–4).
In this half century of evolution, new procedural techniques have been implemented to improve CABG outcomes. Debates over the best techniques, including on-pump versus off-pump surgery, single versus bilateral mammary artery grafts, radial artery versus right internal mammary versus carefully harvested saphenous vein grafts, continue (5–8). However, despite the evolution of CABG over the last decade, little is known about whether clinical outcomes after surgical revascularization have improved over time. By contrast, advancements in PCI devices and techniques have clearly been shown to result in enhanced event-free survival (9–12).
We therefore sought to determine whether clinical outcomes in patients undergoing CABG for the treatment of left main (LM) stem coronary artery disease (LMCAD) have improved in the last decade by analyzing data from 2 major randomized clinical trials: The SYNTAX (Synergy Between PCI With Taxus and Cardiac Surgery) trial and the EXCEL (Evaluation of XIENCE Versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization) trial.
Study population and propensity matching
The present study is a post hoc, propensity-matched analysis of patients randomized to CABG treatment of LMCAD in the SYNTAX and EXCEL trials. The EXCEL study methodology and main results have been previously published (13). Briefly, the EXCEL trial was an international, open-label, multicenter randomized trial that compared PCI with everolimus-eluting stents with CABG in patients with LMCAD. Recruitment occurred between 2010 and 2014, and the trial included 957 patients assigned to CABG. Patients were enrolled with a site-reported SYNTAX score (SS) ≤32 (low to intermediate anatomic complexity). Complete anatomic revascularization of all vessels 1.5 mm or larger in diameter in which the angiographic diameter stenosis was ≥50% was the treatment strategy; the use of arterial grafts was strongly recommended.
These outcomes were propensity matched to patients with LMCAD that were randomized to CABG in the SYNTAX trial. Details of the SYNTAX trial have been published elsewhere (2). In brief, the SYNTAX trial was a randomized, prospective trial comparing outcomes of patients with 3-vessel disease or LMCAD undergoing PCI with paclitaxel-eluting stents versus CABG. The trial randomized 1,800 total patients between 2005 and 2007 in Europe and the United States. Of those, 897 patients were assigned for CABG, including 348 with LMCAD.
Endpoints and definitions
The primary endpoint of interest was the composite of all-cause death, stroke, MI, or ischemia-driven revascularization (IDR) at 3 years, that is, major adverse cardiovascular or cerebrovascular events (MACCE). Secondary endpoints included the composite of all-cause death, stroke, or MI at 30 days and at 3 years. Each individual, nonhierarchical component of the primary outcome was also evaluated as secondary outcomes at 30 days and at 3 years. Definitions of the endpoints are described elsewhere (2,13). The definitions of death, stroke, IDR, and spontaneous MI were otherwise similar between the SYNTAX and EXCEL trials. However, the MI definitions were discordant between the 2 trials due to the requirement in the SYNTAX trial of the development of new Q waves to define a periprocedural MI, which was not applicable to the EXCEL trial. Thus, to properly match MI definitions, this definition was also applied for the EXCEL trial. All primary and secondary endpoints were adjudicated by an independent clinical events committee.
The anatomic SS was evaluated by specialized core laboratories (Cardialysis, Rotterdam, the Netherlands, for the SYNTAX trial, and the Cardiovascular Research Foundation, New York, New York, for the EXCEL trial). The SSII was calculated by the core laboratories using the patients’ clinical characteristics added to the anatomic SS (14).
Trials protocols were approved by the ethics committee at each participating center, and all patients signed informed consent. Five-year follow-up has finished for the SYNTAX trial (15) and is ongoing for the EXCEL trial (for 5 years), with all patients having completed 3-year follow-up at the time of this report.
All analyses were performed in the intention-to-treat population, which included all patients according to the arm to which they were randomly assigned, regardless of the received treatment. A propensity score model was developed to control for selection bias in the comparison between CABG patients in the SYNTAX and EXCEL trials. The propensity score for an individual is defined as the probability of being included in the SYNTAX (vs. EXCEL) trial conditional on the individual’s baseline characteristics. The propensity score was estimated using a logistic regression model in which the study variable (SYNTAX vs. EXCEL trial) is the outcome and the baseline characteristics are the covariates. The following variables known to affect outcomes were included in the propensity score model as covariates: age, sex, body mass index, smoking status, presence of diabetes, heart failure, hyperlipidemia, hypertension, previous MI, prior cerebrovascular accident, chronic obstructive pulmonary disease, peripheral vascular disease, pulmonary hypertension, creatinine >1.7 mg/dl, and SS as determined by the core laboratory. A greedy matching algorithm was used to match (1:2) SYNTAX patients with available control patients in the EXCEL trial, due to the higher sample size of LMCAD in the EXCEL trial with its exclusive inclusion of LMCAD. A caliper of 0.15 times the standard deviation of the propensity score was enforced to ensure closer matches (16).
Continuous data are presented as mean ± SD or as median (interquartile range) according to data distribution. Comparisons were performed using Student’s t-test or Mann-Whitney U test as appropriate. Categorical data are presented as absolute number and percentages, and were compared with the chi-square test or Fisher exact test. Event rates were based on Kaplan-Meier estimates in time-to-first-event analyses and compared with the log-rank test. A 2-sided p value of 0.05 was considered indicative of statistical significance. All statistical analyses were performed with SAS software, version 9.4 (SAS Institute, Cary, North Carolina).
The propensity-matched study population comprised 909 patients from the SYNTAX (n = 329) and the EXCEL trials (n = 580). Except for an increased frequency of family history of CAD in the EXCEL group and a higher systolic blood pressure in the SYNTAX group, baseline characteristics and comorbidities were comparable between the 2 groups (Table 1). Patients in both groups also had similar SS and SSII (Table 2). As per inclusion in the analyses all patients had LMCAD, but SYNTAX subjects were more likely to have concomitant 3-vessel-disease with LMCAD (34.3% vs. 17.4%), whereas the EXCEL patients presented more 1-vessel disease with LMCAD (31.3% vs. 20.1%) when compared with their counterparts. The time from randomization to surgery was greater in the SYNTAX trial than the EXCEL trial (15 vs. 6 days). CABG procedures were more often elective in the SYNTAX trial compared with the EXCEL trial, whereas urgent and emergent CABG procedures were more commonly performed in EXCEL patients.
The duration of bypass, cross clamping, and of the entire surgical procedure were similar between groups. Nor did the total number of arterial and venous conduits differ. The CABG procedures in the EXCEL trial compared with the SYNTAX trial were more often off-pump (29.6% vs. 15.4%), with more blood cardioplegia (64.7% vs. 52%). Use of an in situ left internal mammary artery (98.7% vs. 94.4%; p = 0.002) and radial artery as a graft (10.3% vs. 5.7%; p = 0.01) were more common in the SYNTAX compared with the EXCEL trial, although there were no statistically significant differences between the groups in the total number of arterial or venous grafts, or the use of bilateral IMAs (Table 3).
Thirty-day and 3-year follow-up medication use is reported in Table 4. Guideline-directed medical therapy (GDMT) use was much more frequent in CABG patients at both time periods in the EXCEL surgical cohort compared with the SYNTAX surgical cohort, including more frequent use of antiplatelet agents, lipid-lowering drugs, beta-blockers, and agents affecting the renin-angiotensin axis.
The composite endpoint of MACCE at 3 years was higher in the SYNTAX group than the matched population from the EXCEL trial (20.9% vs. 14.0%; p = 0.008) (Central Illustration). The composite 3-year endpoint of death, stroke, or MI was also higher in the SYNTAX trial (14.0% vs. 9.6%; p = 0.05) (Table 5). At 30 days, there was no significant difference in MACCE between the groups (Table 5). With the exception of MI, all nonhierarchical components of the primary endpoint contributed to the better outcomes in the EXCEL trial compared with the SYNTAX trial: all-cause death (5.5% vs. 8.5%), any stroke (3.1% vs. 5.1%), and IDR (7.1% vs. 9.4%), respectively (Figure 1).
The present study evaluated whether clinical outcomes of patients undergoing CABG for LMCAD have improved over the last decade. The major finding of the study is that 3-year rates of MACCE after surgical revascularization for LMCAD improved between the performance of the SYNTAX and EXCEL trials. Some differences in technique were also noted between the studies, most notably an increase in off-pump CABG and more extensive use of blood cardioplegia, although for the most part, measurable differences in surgical technique were not observed, including the use of single or multiple arterial grafts and procedural durations. Notably, the use of antiplatelet agents and other GDMT were substantially greater throughout the follow-up period in the EXCEL surgical group compared with the SYNTAX trial surgical group.
Major evolution of coronary revascularization technique with PCI has been easy to document, with shifts from balloon angioplasty to predominant use of bare-metal stents to first- and then second-generation drug-eluting stents. In addition, imaging and physiological guidance techniques have become increasingly used, and adjunct pharmacotherapy has improved (9,17,18). Along with these advances, PCI outcomes have steadily improved over the last 3 decades (11,12). However, whether clinical outcomes of patients undergoing CABG have improved over time has been more difficult to demonstrate (8), as exemplified by the 10-year follow-up of the ART (Arterial Revascularization Trial) demonstrating no significant long-term difference with use of a single versus bilateral IMA grafts (19). The present study demonstrates that surgical outcomes have been improving over time, at least for LMCAD revascularization. The improvement in the 3-year rates of MACCE after CABG in the EXCEL trial compared with the SYNTAX trial underlie why the relative outcomes of CABG versus PCI for treatment of LMCAD have remained consistent over time despite use of clearly superior PCI devices and technique in the EXCEL trial compared with the SYNTAX trial (e.g., everolimus-eluting compared with paclitaxel-eluting stents, more frequent use of intravascular ultrasound guidance).
When comparing the 2 trials, the observed change in surgical approach is the increase in the use of off-pump surgery. Potential advantages of off-pump surgery include reduction of the risk of death, stroke, and MI in higher-risk patients (20). The outcomes of off-pump surgery are highly dependent on the expertise of the operator; incomplete revascularization and reduced graft patency may be more frequent when off-pump surgery is performed by less experienced surgeons (21–23). Many surgeons avoided off-pump surgery in the past in patients with LMCAD due to the hemodynamic disturbances that could occur with the displacement of the heart during the grafting process (24). Nonetheless, progression in equipment and techniques such as stabilizers and shunts have improved the reliability and outcomes of off-pump surgery (25,26). A recent large observational study demonstrated a reduction of mortality with off-pump compared with on-pump surgery, regardless of the number of grafts (20)—if performed by experienced surgeons. These findings could be perceived as if the higher use of off-pump CABG in the EXCEL trial compared with the SYNTAX trial played a role in the improvement in clinical outcomes after LMCAD CABG, in particular regarding the reduction of stroke, but also possibly mortality. However, debate over the benefits of the off-pump surgery persists, especially given the long-term results from the CORONARY (Coronary Artery Bypass Surgery [CABG] Off or On Pump Revascularization Study) (similar 5-year rates of death, stroke, MI, renal failure, or repeat revascularization) (6), and the findings from a recent meta-analysis comprising 8,145 patients in which mortality at 4 years or longer was slightly higher (13.9% vs. 12.3%) with off-pump compared with on-pump surgery (27).
A major difference between the EXCEL and SYNTAX trials that may have contributed the most to the long-term improvement in surgical outcomes over time was the greater use of GDMT in the EXCEL trial, especially antiplatelet agents, statins, beta-blockers, and agents inhibiting the renin-angiotensin axis. The most recent guidelines of the American College of Cardiology/American Heart Association clearly state the benefits of the use of aspirin after CABG, or if the patient is intolerant, the use of an alternative antiplatelet (clopidogrel) (28). Both agents were used at 30 days and 3 years more frequently alone and together in the EXCEL group than the SYNTAX group. Of note, a recent study reported higher graft patency with the use of DAPT compared with monotherapy (29). The survival benefit of and reduction in MI in patients with CAD treated with statins is well documented (30,31). In addition, although systolic blood pressure was found to be lower at recruitment in the EXCEL than in the SYNTAX trial, the use of beta-blockers, angiotensin-converting enzyme inhibitors, and angiotensin receptor blockers were greater in the EXCEL trial. A 20 mm Hg decrease in systolic blood pressure has been associated with up to 50% reduction in the risk of death due to ischemic heart disease and up to 62% reduction in death due to any stroke (32). These agents have also been associated with improved survival in patients with left ventricular dysfunction and chronic kidney disease (33–35). The possible key to the improvement in outcomes from the SYNTAX trial to the EXCEL trial may lie on the greater use of GDMT. This comes as an important message to cardiologists treating post-CABG patients, to never underestimate the power of optimized GDMT (36).
Another aspect of the present report supporting the improvement of CABG is that even with a significantly larger amount of patients undergoing either emergent or urgent procedures (thus with possibly worse prognosis) in the EXCEL trial compared with the SYNTAX trial, the primary outcome was still lower in the EXCEL trial.
The present report is based on a post hoc analysis of propensity-matched populations from 2 major randomized trials, and thus has inherent limitations. The differences in technique and GDMT were not randomized, and propensity matching may not have accounted for unmeasured confounders. In addition, the multifactorial nature of all the changes that may have occurred, such as surgical technique, post-procedural and nursing care, and use of cardiac rehabilitation, cannot be accounted for. With the time difference between the trials and obvious difference in variables such as medication use (36), sites and operators, and differences in population characteristics, conclusion as to which particular variable might be responsible for the improvement observed cannot be made. The results of the present analysis should thus be considered hypothesis-generating, and describe associations, but not causality.
These limitations notwithstanding, the present comparison of surgical outcomes from the older SYNTAX trial and the contemporary EXCEL trial in patients with LMCAD over a 5- to 7-year period demonstrated improved event-free survival after CABG during 3-year follow-up, consistent with improving results over time with cardiac surgery.
COMPETENCY IN MEDICAL KNOWLEDGE: The incidence of MACCE following coronary artery bypass surgery have decreased over a period of 5 to 7 years due to improvements in technology and technique.
TRANSLATIONAL OUTLOOK: Further studies are needed to determine whether improvement in clinical outcomes will be sustained over a longer period of time.
The SYNTAX trial was supported by Boston Scientific, and the EXCEL trial was funded by Abbott Vascular. Dr. Modolo has received financial support from Sao Paulo Research Foundation (FAPESP) grant number 2017/22013-8. Dr. Banning has been partially funded by the NHS Oxford NIHR Biomedical Research Centre; has received institutional sponsorship for a fellowship from Boston Scientific; and has received lecture fees from Boston Scientific, Medtronic, and Abbott Vascular. Dr. Puskas has been a consultant to Medtronic. Dr. Kappetein is an employee of Medtronic. Dr. Sabik has been a consultant to Medtronic, Edwards Lifesciences, and Sorin; and has served on an advisory board for Medtronic Cardiac Surgery. Dr. Onuma is an employee of Cardialysis. Dr. Stone has been a consultant to Claret, Backbeat, Sirtex, Matrizyme, Miracor, Neovasc, V-wave, Shockwave, Valfix, TherOx, Reva, Vascular Dynamics, Robocath, HeartFlow, Gore, Ablative Solutions, and Ancora; has received speaker honoraria from Amaranth and Terumo; holds equity in Ancora, Cagent, Qool Therapeutics, Aria, Caliber, Applied Therapeutics, SpectraWave, the MedFocus family of funds, and the BioStar family of funds; is director of SpectraWave; and his employer, Columbia University, receives royalties from sale of the MitraClip manufactured by Abbott Vascular. Prof. Serruys has been a consultant to Abbott, Biosensors, Medtronic, Micell Technologies, QualiMed, SINOMED, St. Jude Medical, Stentys, Svelte, Philips/Volcano, and Xeltis. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Michael J. Mack, MD, served as Guest Associate Editor for this paper.
Listen to this manuscript's audio summary by Editor-in-Chief Dr. Valentin Fuster on JACC.org.
- Abbreviations and Acronyms
- coronary artery bypass graft
- coronary artery disease
- guideline-directed medical therapy
- ischemia-driven revascularization
- left main
- left main stem coronary artery disease
- major adverse cardiovascular or cerebrovascular events
- myocardial infarction
- percutaneous coronary intervention
- SYNTAX score
- Received October 22, 2018.
- Revision received December 7, 2018.
- Accepted December 31, 2018.
- 2019 American College of Cardiology Foundation
- Benedetto U.,
- Caputo M.,
- Gaudino M.,
- Mariscalco G.,
- Bryan A.,
- Angelini G.D.
- Lamy A.,
- Devereaux P.J.,
- Prabhakaran D.,
- et al.
- Gaudino M.,
- Benedetto U.,
- Fremes S.,
- et al.
- Taggart D.P.,
- Altman D.G.,
- Gray A.M.,
- et al.
- Escaned J.,
- Collet C.,
- Ryan N.,
- et al.
- Modolo R.,
- Collet C.,
- Onuma Y.,
- Serruys P.W.
- Palmerini T.,
- Biondi-Zoccai G.,
- Della Riva D.,
- et al.
- Raber L.,
- Kelbaek H.,
- Taniwaki M.,
- et al.
- Stone G.W.,
- Sabik J.F.,
- Serruys P.W.,
- et al.
- Farooq V.,
- van Klaveren D.,
- Steyerberg E.W.,
- et al.
- ↵Taggart DP. Randomized comparison of single versus bilateral internal thoracic artery grafts in 3102 CABG patients: Major cardiovascular outcomes at ten years of follow up. Paper presented at: ESC 2018; August 26, 2018; Munich, Germany.
- Kowalewski M.,
- Pawliszak W.,
- Malvindi P.G.,
- et al.
- Benedetto U.,
- Lau C.,
- Caputo M.,
- et al.
- Benedetto U.,
- Gaudino M.,
- Di Franco A.,
- et al.
- Shroyer A.L.,
- Hattler B.,
- Wagner T.H.,
- et al.
- Sergeant P.
- Smart N.A.,
- Dieberg G.,
- King N.
- Hillis L.D.,
- Smith P.K.,
- Anderson J.L.,
- et al.
- Zhao Q.,
- Zhu Y.,
- Xu Z.,
- et al.
- Stone N.J.,
- Robinson J.G.,
- Lichtenstein A.H.,
- et al.
- Ponikowski P.,
- Voors A.A.,
- Anker S.D.,
- et al.
- Iqbal J.,
- Zhang Y.J.,
- Holmes D.R.,
- et al.