Author + information
- Received October 26, 2016
- Revision received January 15, 2017
- Accepted February 6, 2017
- Published online April 17, 2017.
- Mouin S. Abdallah, MD, MSca,
- Kaijun Wang, PhDb,
- Elizabeth A. Magnuson, ScDb,c,
- Ruben L. Osnabrugge, MScd,
- A. Pieter Kappetein, MD, PhDd,
- Marie-Claude Morice, MDe,
- Friedrich A. Mohr, MD, PhDf,
- Patrick W. Serruys, MD, PhDd,
- David J. Cohen, MD, MScb,c,∗ (, )
- SYNTAX Trial Investigators
- aRobert and Suzanne Tomsich Department of Cardiovascular Medicine, Cleveland Clinic, Cleveland, Ohio
- bSaint Luke’s Mid America Heart Institute, Kansas City, Missouri
- cUniversity of Missouri-Kansas City School of Medicine, Kansas City, Missouri
- dErasmus University Medical Center, Rotterdam, the Netherlands
- eHôpital Privé Jacques Cartier, Massy, France
- fHerzzentrum Universität Leipzig, Leipzig, Germany
- ↵∗Address for correspondence:
Dr. David J. Cohen, Saint Luke’s Mid America Heart Institute, University of Missouri-Kansas City School of Medicine, 4401 Wornall Road, Kansas City, Missouri 64111.
Background In the SYNTAX (Synergy Between Percutaneous Coronary Intervention With Taxus and Cardiac Surgery) trial, patients with 3-vessel or left main coronary artery disease (CAD) had improved long-term outcomes with coronary artery bypass graft (CABG) surgery compared with percutaneous coronary intervention (PCI) with drug-eluting stents (DES), improvements driven mainly by differences in myocardial infarction and repeat revascularization.
Objectives This study compared the long-term quality-of-life benefits of DES-PCI versus CABG for patients with 3-vessel or left main CAD.
Methods Between 2005 and 2007, the SYNTAX trial randomized 1,800 patients with 3-vessel or left main CAD to either CABG or DES-PCI. Health status was assessed at baseline and at 1, 6, 12, 36, and 60 months by using the Seattle Angina Questionnaire (SAQ) and the 36-Item Short Form Health Survey.
Results At 5-year follow-up, CABG was superior to DES-PCI on several SAQ domains including angina frequency and physical function, as well as the role physical and role emotional scales of the 36-Item Short Form Health Survey. Subgroup analysis demonstrated a significant interaction between angiographic complexity (as assessed by the SYNTAX score) and angina relief (mean difference in the SAQ angina frequency score for CABG vs. PCI of −0.9, 3.3, and 3.9 points for low, intermediate, and high SYNTAX score patients, respectively; p = 0.048 for interaction).
Conclusions Among patients with 3-vessel or left main CAD, both CABG and DES-PCI were associated with substantial and sustained quality-of-life benefits over 5 years of follow-up. In general, CABG resulted in greater angina relief, although the absolute treatment benefit was small. Angina relief at 5 years was enhanced with CABG among patients with high SYNTAX scores, a finding reinforcing the recommendation that CABG should be strongly preferred for such patients. (Synergy Between Percutaneous Coronary Intervention With Taxus and Cardiac Surgery [SYNTAX]; NCT00114972)
The SYNTAX (Synergy Between Percutaneous Coronary Intervention With Taxus and Cardiac Surgery) trial compared the outcomes of percutaneous coronary intervention (PCI) with drug-eluting stents (DES) versus coronary artery bypass graft (CABG) surgery among patients with 3-vessel or left main coronary artery disease (CAD) (1). At 1-year follow-up, CABG provided superior angina relief compared with DES-PCI as assessed by the Seattle Angina Questionnaire (SAQ) and a small advantage on the Medical Outcomes Study 36-Item Short Form Health Survey (SF-36) general health subscale (2). At 5-year follow-up, clinical results from SYNTAX showed that CABG was superior to DES-PCI with respect to the composite endpoint of death, myocardial infarction, stroke, or repeat revascularization, with the benefit driven predominantly by decreased rates of repeat revascularization and myocardial infarction with no significant differences in rates of death or stroke (3).
Given the similar long-term outcomes of CABG and DES-PCI with respect to clinical endpoints, other factors including quality of life (QOL) should be considered when patients are suitable candidates for both revascularization strategies. Accordingly, to provide a complete picture of the outcomes after CABG versus DES-PCI and guide the choice of revascularization strategy for patients with 3-vessel or left main CAD, we performed a prospective health status substudy alongside the SYNTAX trial. The goal of this substudy was to assess both disease-specific and generic health status from the patients’ perspective up to 5 years after initial revascularization for 3-vessel or left main CAD with either DES-PCI or CABG.
The design, methods, and clinical results of the SYNTAX trial have been published previously (1,3,4). Briefly, between 2005 and 2007, 1,800 patients with 3-vessel or left main CAD from 85 centers in 18 countries primarily in Europe and the United States were randomized to revascularization with either CABG or PCI with a first-generation paclitaxel-eluting stent (TAXUS Express, Boston Scientific, Natick, Massachusetts). All patients enrolled in the trial were deemed appropriate candidates for either revascularization procedure by a local heart team that included an interventional cardiologist and a cardiothoracic surgeon. The treatment goal for all patients was complete revascularization with the assigned treatment strategy along with optimal medical therapy chosen on the basis of contemporary guidelines. All patients gave written informed consent, and locally appointed ethics committees and regulatory agencies approved the study protocol.
Measurement of health status
Health status was assessed in all patients at baseline and at 1, 6, 12, 36, and 60 months after randomization. Disease-specific health status was assessed using the SAQ, and generic health status was assessed using the SF-36 (5,6). The SAQ is a 19-item questionnaire that measures 5 domains of health status related to CAD: angina frequency (AF), physical limitations (PL), disease perception/QOL, angina stability (AS), and treatment satisfaction (TS) (5,7–9). Scores range from 0 to 100, with higher scores indicating fewer symptoms and better health status. Previous studies demonstrated that differences of 8 to 10 points are clinically meaningful on an individual patient level (7). The SF-36 is a 36-item questionnaire that measures health status across 8 domains: physical functioning; role limitations resulting from physical problems; bodily pain; vitality; general health perception; social function; role limitations resulting from emotional problems; and mental health. Scores range from 0 to 100, with higher scores reflecting better health status. Summary scores for physical and mental health are standardized to a mean of 50 and an SD of 10 (10). The minimum clinically important difference for the SF-36 summary scales is approximately 2 points (11). Questionnaires were administered in each patient’s native language. Validated translations were available for the SAQ in 15 of 18 participating countries and for all 18 countries for the SF-36. Missing items for each subscale were handled according to published guidelines provided by the developer for each instrument.
The population for our study included all patients with available baseline health status, analyzed by intention-to-treat. Baseline patient characteristics and health status were compared between the 2 groups using Student t tests for continuous variables and chi-square tests for categorical variables. Within-group comparisons were performed using paired Student t tests to evaluate changes in health status between baseline and 1 month, 6 months, 1 year, and 5 years as well as between 1 and 6 months. Differences in health status between CABG and DES-PCI at each time point were estimated using analysis of covariance models, adjusting for baseline health status. To account for missing data, we repeated the analysis using longitudinal random-effect growth curve models that incorporated all available health status scores, including those for patients who subsequently died, withdrew from the study, or were lost to follow-up. Covariates for the growth curve models included the baseline score of the health status domain being assessed, treatment assignment, follow-up time, and interactions between treatment and time. The intercept and linear time effects were estimated using both fixed and random effects, whereas quadratic and cubic effects of time were modeled as fixed effects to avoid overparameterization. These models were optimized using a backward elimination procedure, starting with the highest order of time by treatment interaction.
To evaluate whether between-group differences in health status were clinically significant, we created cumulative response curves that describe the proportion of patients (by treatment group) who achieved health status improvement of a given magnitude (12).
An additional analysis was performed in which the SAQ-AF score was treated as a categorical variable. Consistent with previous studies, patients were categorized as being angina free (i.e., SAQ-AF score of 100) or as having monthly angina (SAQ-AF score, 70 to 90), weekly angina (SAQ-AF score, 40 to 60), or daily angina (SAQ-AF score ≤30). This categorical variable was analyzed using ordinal logistic regression, which estimates the odds of a 1-level improvement in angina-frequency with CABG versus DES-PCI.
Finally, stratified analyses were performed to compare the effect of CABG versus DES-PCI on the SAQ-AF scale within pre-specified subgroups according to age (<75 years vs. ≥75 years), diabetes status, 3-vessel versus left main CAD, SYNTAX score (≤22 vs. 23 to 32 vs. ≥33), and baseline angina severity (none vs. monthly vs. daily or weekly). These analyses were performed by introducing a treatment by subgroup interaction term in the analysis of covariance models at both the 3-year and 5-year time points.
All tests of statistical significance were 2-tailed, and a p value <0.05 was considered to indicate statistical significance without adjustment for multiple comparisons. All analyses were performed using SAS for Windows software, version 9.3 (SAS Institute, Cary, North Carolina).
A total of 1,800 patients with 3-vessel or left main CAD were randomized to revascularization with either CABG or DES-PCI. Of these patients, 1,731 had baseline health status available (n = 848 CABG-treated patients and n = 883 PCI-treated patients) and constituted the analytic cohort for our study (Figure 1). Baseline characteristics were well balanced between treatment groups without major differences in clinical characteristics, disease severity, or health status according to both the SAQ and the SF-36 (Table 1). At baseline, according to the SAQ-AF scale, 11% of the patients had daily angina, 27% had weekly angina, 39% had monthly angina, and 22% had no angina. Completion rates for the SAQ and SF-36 are summarized in Online Table 1. In general, completion rates were high and similar for both treatment groups through 3 years of follow-up. At 5 years, there was a slight decline in compliance with QOL follow-up, with rates of 70% to 75% for the SAQ and 75% to 80% for the SF-36. When we compared patients who did and did not respond to the QOL surveys at 5 years, nonresponders were older, more likely to smoke, and more likely to have had a prior myocardial infarction (Online Table 2). In addition, nonresponders had greater physical limitations at baseline as determined on the basis of both the SAQ and the SF-36.
Disease-specific health status
Changes in disease-specific health status over time and mean differences between treatment groups are summarized in the Central Illustration and Table 2. Compared with baseline, both the DES-PCI and CABG groups demonstrated significant improvement in cardiac-specific health status as assessed by the SAQ by 1-month follow-up, with further improvement at 6 months and 1 year; these changes were largely sustained at 3 and 5 years (Online Table 3). Compared with CABG, patients assigned to DES-PCI had faster health status improvement with higher scores on the SAQ-PL, SAQ-QOL, and SAQ treatment satisfaction scales at 1 month. These differences were not sustained, however, and CABG resulted in improved scores on the SAQ-AF and SAQ-QOL scales at 1-year follow-up. At 3-year follow-up, there were no differences in disease-specific health status between treatment groups; however, at 5-year follow-up, there were small but statistically significant differences favoring CABG on both the SAQ-AF and the SAQ-PL subscales (mean differences of 2.1 and 3.0 points, respectively).
Results were generally similar when these analyses were repeated using longitudinal growth curve models to account for missing data (Online Table 4) and when the SAQ-AF scale was analyzed as a categorical variable (Figure 2). In particular, the proportion of patients who were angina free was significantly greater with CABG than with DES-PCI at both 1-year follow-up (76.8% vs. 71.5%; p = 0.01) and 5-year follow-up (78.8% vs. 74.1%; p = 0.03).
Compared with baseline, generic health status improved substantially by 1-month follow-up in the DES-PCI group and by 6-month follow-up in the CABG group (Online Table 3). Changes in generic health status over time and mean differences between treatment groups are summarized in Figures 3A and 3B and Table 2. There were significant differences in virtually all the SF-36 subscales in favor of DES-PCI at 1 month, but these differences were no longer apparent at 6 months. Although there were no differences in generic health status between CABG-treated patients and DES-PCI–treated patients at 1 and 3 years, by 5 years there were small but statistically significant differences in the role physical and role emotional subscales as well as the mental component summary scale in favor of CABG (mean differences of 3.1, 4.0, and 1.5 points, respectively).
Cumulative response curves demonstrating the proportion of patients who experienced changes in health status of a given magnitude are displayed in Online Figures 1 to 4. For the SAQ-AF scale, the major benefit of CABG over DES-PCI at 5 years was in the proportion of patients who experienced large improvements in angina. Specifically, ∼22% of patients treated with CABG experienced at least a 40-point improvement in the SAQ-AF scale, whereas only 17% of patients treated with DES-PCI experienced this great an improvement (absolute risk difference: 4.4%; number needed to treat [NNT] = 23). Conversely, for the SAQ-PL scale, the major benefit of CABG was in avoidance of modest declines in physical limitations. With PCI, approximately 15% of DES-PCI–treated patients experienced at least a 10-point decline in the SAQ-PL scale compared with only 11% of CABG-treated patients (absolute risk difference: 4.4%; NNT = 23). For a 20-point decline in the SAQ-PL scale at 5-years (a large decline), the absolute risk difference was 2.8% (NNT = 36).
The estimated effect of CABG versus DES-PCI on the SAQ-AF scale within pre-specified subgroups is shown in Figure 4A (3-year treatment effect) and Figure 4B (5-year treatment effect). There were no significant treatment-by-subgroup interactions at 3 years. However, at 5-year follow-up, patients with higher baseline SYNTAX scores had greater angina relief following CABG than after DES-PCI (mean treatment difference for CABG vs. DES-PCI of −0.9, 3.3, and 3.9 points for patients with low, intermediate, and high SYNTAX scores, respectively; p = 0.048 for interaction). Patients with high SYNTAX scores were more frequently angina free at 5-years with CABG than with DES-PCI (82.2% vs. 73.3%); however, there were no significant differences in the proportion of angina-free patients with low (72.3% vs. 72.5%; p = 0.94) or intermediate (81.6% vs. 75.6%; p = 0.21) SYNTAX scores (Online Table 5).
Antianginal medication use rates over time are summarized in Online Table 6. There were no significant differences between the DES-PCI and CABG treatment groups with respect to use of β-blockers and calcium-channel blockers at 3- and 5-year follow-up. Patients assigned to initial DES-PCI were more likely to be taking long-acting nitrates at each follow-up time point, however.
In this study, we used SYNTAX trial data to compare the impact of CABG versus DES-PCI on long-term health status among patients with 3-vessel or left main CAD. The study demonstrated several important findings: 1) both CABG and DES-PCI were associated with angina relief as early as 1 month after revascularization, and this benefit was largely sustained over 5 years; 2) health status recovery was faster after DES-PCI compared with CABG; 3) at both 1- and 5-year follow-up, CABG resulted in statistically significant differences in cardiovascular-specific and generic health status, but the absolute magnitude of these between-group differences was relatively small and of uncertain clinical relevance; and 4) perhaps most importantly, there was significant interaction between the anatomic SYNTAX score and the extent of angina relief at 5 years, thus favoring CABG among patients with high SYNTAX scores.
Several previous studies compared health status after CABG versus PCI using a variety of different techniques including balloon angioplasty (BARI [Bypass Angioplasty Revascularization Investigation], RITA [Randomized Intervention Treatment of Angina], GABI [German Angioplasty Bypass Surgery Investigation], and CABRI [Coronary Angioplasty Versus Bypass Revascularization Investigation]) (13–16), bare-metal stents (ARTS I [Arterial Revascularization Therapies Study] and SoS [Stent or Surgery]) (17–19), and most recently, drug-eluting stents (ARTS II and FREEDOM [Future Revascularization Evaluation in Patients With Diabetes Mellitus: Optimal Management of Multivessel Disease]) (20,21). Although the general pattern of health status changes noted in SYNTAX is similar to that observed in previous trials (earlier recovery with PCI and a long-term advantage with CABG), the magnitude of benefit provided by CABG over PCI differed from the older trials. For example, at 1-year follow-up, the difference in the frequency of any angina between CABG and balloon angioplasty was 15% in the BARI trial (13) and 10.6% between CABG and PCI with bare-metal stents in ARTS (17), whereas in the SYNTAX trial, this difference was just 5.3%. This diminishing difference in cardiovascular-specific health status between CABG and DES-PCI treatment groups at early time points may reflect the benefits of drug-eluting stents on restenosis, which generally occurs within the first 12 months of PCI and is frequently associated with recurrent angina (22).
Although the differences between CABG and DES-PCI groups in disease-specific health status at 5 years (2.1 and 3.0 points on the SAQ-AF and SAQ-PL domains, respectively) are smaller than the minimum clinically important differences for these scales, it is important to recognize the distinction between the minimum clinically important difference for an individual (8 to 10 points for the SAQ subscales) and the difference in group means (which is the principal result of our analysis). Patient-level cumulative response curves indicated that these between-group differences were driven by modest differences (3 to 5 percentage points) in the proportion of patients who experienced clinically important changes in health status at 5 years. To provide further context, it is worth noting that the 5% absolute difference in the frequency of any angina between CABG and DES-PCI treatment groups at 5 years is larger than the differences between PCI and medical therapy groups at 1- and 3-year follow-up in the COURAGE (Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation) trial (23).
The long-term health status differences noted in our trial were similar to those previously reported from FREEDOM, another contemporary trial comparing CABG and DES-PCI to examine long-term patient-reported outcomes (21). Similar to SYNTAX, FREEDOM (which was specifically conducted among patients with diabetes and multivessel CAD) demonstrated no significant differences in disease-specific or generic health status between CABG and DES-PCI treatment groups at 3-year follow-up, but an increasing benefit of CABG after 3 years. The specific mechanism underlying this late benefit of CABG over DES-PCI remains uncertain, but it may relate to differences in completeness of revascularization or greater protection from progression of proximal disease with bypass grafting as compared with the more focused revascularization provided by PCI.
Subgroup analyses demonstrated that the benefits of CABG over DES-PCI with respect to angina relief were consistent across most pre-specified patient-related characteristics including age, diabetes, and the presence of significant left main CAD. However, the benefit of CABG over DES-PCI did differ according to the extent of anatomic complexity (as assessed by the baseline SYNTAX score). Specifically, patients with intermediate (23 to 32) and high (≥33) SYNTAX scores derived significantly greater angina relief with CABG compared with DES-PCI at 5 years, whereas patients in the lowest SYNTAX score tertile (0 to 22) derived no additional angina relief from CABG. These findings paralleled the clinical benefits of CABG over DES-PCI in the SYNTAX trial, which demonstrated significantly better long-term outcomes with CABG, particularly among the highest SYNTAX score tertile (3), and strongly suggested that CABG should be the treatment of choice for such high-risk patients (24,25). Conversely, the lack of meaningful long-term differences in clinical outcomes or health status, coupled with the early QOL benefits seen with DES-PCI, suggests that DES-PCI would be preferred by most patients with anatomically less complex disease. In addition, with long-term follow-up, the interaction between baseline angina frequency and treatment group that had previously been noted at 1 year (with greater benefit of CABG among patients with daily or weekly angina) (2) was no longer apparent at either 3 or 5 years.
This study should be viewed in light of several limitations. The SYNTAX trial enrolled a select group of patients with 3-vessel or left main CAD who were believed to be suitable candidates for either PCI or CABG. As a result, it is unlikely that our results can be applied to patients who are not suitable candidates for equivalent revascularization by these 2 techniques. Another important limitation in SYNTAX is missing QOL data, which were <20% through 3-year follow-up but ∼25% in year 5. Although it is possible that missing data resulted in bias, the rates of missing data were virtually identical between the CABG and DES-PCI groups at each follow-up time point. Moreover, the fact that our main findings were similar to those in other studies of PCI versus CABG in multivessel CAD and were unchanged in analyses that used longitudinal growth curve models as their basis is relatively reassuring in this regard. Finally, the SYNTAX trial used only first-generation paclitaxel-eluting stents for the index revascularization procedures. Because recent meta-analyses have suggested that newer-generation DES (in particular, everolimus-eluting stents) result in lower rates of target vessel revascularization and stent thrombosis compared with paclitaxel-eluting stents (26,27), it is possible that some of the health status benefits of CABG seen in SYNTAX would have been attenuated with use of these improved devices.
The health status results from the SYNTAX trial demonstrated that for patients with 3-vessel or left main CAD, a strategy of initial CABG provided better angina relief than did DES-PCI at 5 years, differences that were driven mainly by patients with higher SYNTAX scores. These findings reinforce the current recommendation that CABG should be the preferred revascularization treatment for patients with anatomically complex 3-vessel or left main CAD.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Patients with 3-vessel or left main CAD derive greater long-term health and QOL benefits from CABG compared with PCI using first-generation DES. These benefits are confined mainly to patients with anatomically complex CAD, a finding suggesting that DES-PCI may be a reasonable alternative to CABG for patients with less complex disease.
TRANSLATIONAL OUTLOOK: Future trials should evaluate whether newer-generation DES can provide benefits comparable to those of CABG for patients with left main and complex 3-vessel CAD.
For supplemental tables and figures, please see the online version of this article.
This work was funded by Boston Scientific and was designed collaboratively by the steering committee and the sponsor. The data were collected by the investigators at each site and were analyzed by the Health Economics and Technology Assessment Group at Saint Luke’s Mid America Heart Institute. Dr. Cohen has received institutional research grants from Boston Scientific during the conduct of the study; has received institutional research grants from Boston Scientific, Medtronic, and Abbott Vascular outside the submitted work; and has received consulting fees from Medtronic and Cardinal Health. Dr. Magnuson has received research grants from Boston Scientific during the conduct of the study; and has received research grants from Abbott, AstraZeneca, Daiichi-Sankyo, Edwards Lifesciences, Eli Lilly, and Medtronic. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- coronary artery bypass graft
- coronary artery disease
- drug-eluting stent(s)
- percutaneous coronary intervention with drug-eluting stent
- percutaneous coronary intervention
- quality of life
- Seattle Angina Questionnaire
- Seattle Angina Questionnaire-angina frequency
- Seattle Angina Questionnaire-physical limitations
- 36-Item Short Form Health Survey
- Received October 26, 2016.
- Revision received January 15, 2017.
- Accepted February 6, 2017.
- 2017 American College of Cardiology Foundation
- Mohr F.W.,
- Morice M.C.,
- Kappetein A.P.,
- et al.
- Spertus J.A.,
- Winder J.A.,
- Dewhurst T.A.,
- et al.
- Spertus J.A.,
- Jones P.,
- McDonell M.,
- Fan V.,
- Fihn S.D.
- Pocock S.J.,
- Henderson R.A.,
- Seed P.,
- Treasure T.,
- Hampton J.R.
- Hamm C.W.,
- Reimers J.,
- Ischinger T.,
- Rupprecht H.J.,
- Berger J.,
- Bleifeld W.
- Wahrborg P.
- Zhang Z.,
- Mahoney E.M.,
- Stables R.H.,
- et al.
- Serruys P.W.,
- Ong A.T.,
- van Herwerden L.A.,
- et al.
- van Domburg R.T.,
- Daemen J.,
- Morice M.C.,
- et al.
- Arnold S.V.,
- Magnuson E.A.,
- Wang K.,
- et al.
- Kolh P.,
- Windecker S.
- Fihn S.D.,
- Blankenship J.C.,
- Alexander K.P.,
- et al.
- Bangalore S.,
- Kumar S.,
- Fusaro M.,
- et al.