Author + information
- Received July 18, 2016
- Revision received May 30, 2017
- Accepted May 31, 2017
- Published online July 24, 2017.
- Sajjad Raza, MDa,
- Eugene H. Blackstone, MDa,b,
- Penny L. Houghtaling, MSb,
- Jeevanantham Rajeswaran, PhDb,
- Haris Riaz, MDc,
- Faisal G. Bakaeen, MDa,
- A. Michael Lincoff, MDd and
- Joseph F. Sabik III, MDa,∗ ()
- aDepartment of Thoracic and Cardiovascular Surgery, Cleveland Clinic, Cleveland, Ohio
- bDepartment of Quantitative Health Sciences, Cleveland Clinic, Cleveland, Ohio
- cDepartment of Internal Medicine, Cleveland Clinic, Cleveland, Ohio
- dDepartment of Cardiovascular Medicine, Cleveland Clinic, Cleveland, Ohio
- ↵∗Address for correspondence:
Dr. Joseph F. Sabik III, Department of Surgery, University Hospitals Cleveland Medical Center, 11100 Euclid Avenue, Lakeside 7, Cleveland, Ohio 44106-7060.
Background Nearly 50% of patients undergoing coronary artery bypass grafting have diabetes. However, little is known about the influence of diabetes on long-term patency of bypass grafts. Because patients with diabetes have more severe coronary artery stenosis, we hypothesized that graft patency is worse in patients with than without diabetes.
Objectives This study sought to examine the influence of diabetes on long-term patency of bypass grafts.
Methods From 1972 to 2011, 57,961 patients underwent primary isolated coronary artery bypass grafting. Of these, 1,372 pharmacologically treated patients with diabetes and 10,147 patients without diabetes had 15,887 postoperative angiograms; stenosis was quantified for 7,903 internal thoracic artery (ITA) grafts and 20,066 saphenous vein grafts. Status of graft patency across time was analyzed by longitudinal nonlinear mixed-effects modeling.
Results ITA graft patency was stable over time and similar in patients with and without diabetes: at 1, 5, 10, and 20 years, 97%, 97%, 96%, and 96% in patients with diabetes, and 96%, 96%, 95%, and 93% in patients without diabetes, respectively (early p = 0.20; late p = 0.30). In contrast, saphenous vein graft patency declined over time and similarly in patients with and without diabetes: at 1, 5, 10, and 20 years, 78%, 70%, 57%, and 42% in patients with diabetes, and 82%, 72%, 58%, and 41% in patients without diabetes, respectively (early p < 0.002; late p = 0.60). After adjusting for patient characteristics, diabetes was associated with higher early patency of ITA grafts (odds ratio: 0.63; 95% confidence limits: 0.43 to 0.91; p = 0.013), but late patency of ITA grafts was similar in patients with and without diabetes (p = 0.80). Early and late patency of saphenous vein grafts were similar in patients with and without diabetes (early p = 0.90; late p = 0.80).
Conclusions Contrary to our hypothesis, diabetes did not influence long-term patency of bypass grafts. Use of ITA grafts should be maximized in patients undergoing coronary artery bypass grafting because they have excellent patency in patients with and without diabetes even after 20 years.
- coronary artery bypass grafting
- graft occlusion
- graft patency
- internal thoracic artery grafts
- saphenous vein grafts
The FREEDOM (Future Revascularization Evaluation in Patients with Diabetes Mellitus: Optimal Management of Multivessel Disease) trial demonstrated fewer deaths and myocardial infarctions in patients with diabetes and multivessel coronary artery disease after coronary artery bypass grafting (CABG) than after percutaneous coronary intervention (1). Today, nearly 50% of patients undergoing CABG have diabetes (2,3). However, little is known about bypass graft patency in patients with versus without diabetes (4–9). We hypothesized that because patients with diabetes have more severe coronary artery stenosis (10,11), stenosis in their bypass grafts would also be more severe than in patients without diabetes, resulting in lower graft patency. To test this hypothesis, we compared patency of the 2 most commonly used bypass grafts, internal thoracic arteries (ITA) and saphenous veins (SV), in patients with versus without diabetes.
From January 1, 1972, to January 1, 2011, a total of 57,961 patients underwent primary isolated CABG at Cleveland Clinic. Patients were included in this study if the following information was available: 1) knowledge of whether pharmacologically treated diabetes mellitus was present (treated with insulin or oral hypoglycemic agents); 2) detailed pre-operative angiographic data describing location and severity of native coronary system stenosis; 3) results of at least 1 post-operative angiogram before any repeat coronary intervention; and 4) quantitative information on degree of angiographic stenosis of bypass grafts. These criteria were met for 11,519 patients, 1,372 with pharmacologically treated diabetes and 10,147 without (Figure 1, Online Table 1). Pre-operative, operative, and post-operative variables (Online Appendix 1) were retrieved from the Cardiovascular Information Registry. Use of these data for research was approved by the Institutional Review Board, with patient consent waived. Compared with excluded patients, studied patients were younger, operated on in earlier years, had fewer comorbidities, fewer received bilateral ITA grafts, and more had incomplete revascularization (Online Table 2).
Coronary artery bypass grafts
In the study group, 28,876 coronary artery bypasses were performed, 3,881 in patients with diabetes and 24,995 in patients without diabetes (Online Table 3). These included 8,124 ITA grafts and 20,376 SV grafts. Use of ITA grafting substantially increased in the 1980s (Figure 2). Most grafts were used as single conduits: 95.5% of ITA grafts in patients with diabetes and 96.5% in patients without diabetes, and 82% of SV grafts in patients with diabetes and 86% in patients without diabetes. Four general sites for graft anastomoses were defined: 1) left anterior descending (LAD); 2) diagonal; 3) left circumflex; and 4) right coronary artery.
Patients underwent post-operative coronary angiography for a variety of reasons. Early in the series, they underwent a single planned angiogram at 1 year after CABG; subsequently, we presume angiography was performed for suspected ischemic symptoms.
From 1972 to 2011, 15,887 post-operative angiograms were performed in the study group, with 38,753 individual graft observations (Online Figures 1 and 2, Online Table 4); 8,387 patients (73%) had 1 postoperative angiogram, 2,267 (20%) had 2 post-operative angiograms, 623 (5.4%) had 3 post-operative angiograms, and 242 (2.1%) had 4 or more. These angiograms recorded stenosis for 27,969 unique grafts: 7,903 ITA grafts (1,132 in patients with diabetes and 6,771 in patients without diabetes) and 20,066 SV grafts (2,512 in patients with diabetes and 17,554 in patients without diabetes). These 27,969 grafts were the unit of analysis, not the patient.
To investigate possible underestimation of graft patency caused by patients’ reluctance to undergo angiography unless symptomatic, we studied a subgroup of 985 patients who underwent CABG from 1972 through 1975 and had a single planned angiogram at 1 year (11 to 13 months). These angiograms recorded stenosis for 1,883 unique grafts; 433 ITA grafts (27 in patients with diabetes and 406 in patients without diabetes), and 1,450 SV grafts (88 in patients with diabetes and 1,362 in patients without diabetes).
Grafts, particularly SV grafts, tended to be either completely patent or occluded (Figure 3). Therefore, for analysis of this bimodal distribution, a graft was defined as patent if not occluded on follow-up angiography.
Because date of graft occlusion is rarely known, rendering time-to-event (Kaplan-Meier) analysis nonapplicable, we used longitudinal data analysis to study bypass graft patency. The objective of this method was to estimate ensemble average patency across time after CABG from multiple angiographic “snapshots” of the status of patients’ grafts, much as one would do for multiple blood pressure readings across time. The patient was a random effect to account for within-patient variability, because a patient generally had multiple grafts. Patency was compared for grafts in patients with and without diabetes and for type of graft, ITA or SV. Because the analysis identified both an early and late phase of graft occlusion, with the early phase lasting for about 1 year for ITA grafts and about 4 years for SV grafts, statistics for both phases are presented and arbitrarily designated as “early” and “late.” Details of the analysis are given in Online Appendix 2 (“Patency Analysis”).
Risk-adjusted graft patency
Because we hypothesized that graft patency is adversely affected by secondary effects of diabetes, we developed a semisaturated propensity score for having diabetes. Logistic regression was used, incorporating 30 clinical variables (C = 0.84). Effectiveness of the propensity score to match patients is displayed in Online Figure 3. This propensity score was used to adjust all models presented in this study (“Propensity Score Development” in Online Appendix 2).
For the 985 patients with a single planned angiogram available at 1 year, simple logistic regression for occlusion at 1 year (not time to event) was performed with adjustment done for sex, age, body mass index, degree of proximal stenosis, grafts to LAD, and propensity score.
Risk factors for graft occlusion
Prevalence of graft occlusion across time was estimated by averaging patient-specific profiles. Therefore, we sought to identify variables besides diabetes and type of bypass graft that modulated this longitudinal prevalence. For this, we identified variables simultaneously in each phase of graft occlusion using variables in Online Appendix 1 (“Risk Factor Analysis” in Online Appendix 2).
Because patients with diabetes tend to be clinically followed for their diabetes more closely than patients without diabetes, the possibility of work-up bias affecting estimates of graft patency cannot be ignored. Therefore, 3 complementary methods were used to assess time to first post-CABG angiogram and frequency of angiographic assessment: 1) crude comparison between patients with and without diabetes of time to first angiogram and number of angiograms performed per patient; 2) repeat of “1” for 1,328 propensity-matched patients (98% of all possible matches); and 3) New York Heart Association functional class at first angiogram (“Evaluation of Possible Work-up Bias” in Online Appendix 2).
The unanticipated results of this study made it imperative to re-examine this particular cohort of patients with diabetes who, by nature of the study, had to have survived to undergo at least 1 coronary angiogram (“Long-term Survival Analysis” in Online Appendix 2).
Possible influence of cardiac death on patency
Although in longitudinal data analysis death merely terminates observation of graft stenosis, it is possible that increased risk of death in patients with diabetes may lead to overestimation of graft patency because it cannot account for cardiac deaths caused by graft occlusion. Formal statistical methods to quantify the effect of cardiac death on graft patency estimates do not yet exist, but methods do exist to discover if the pattern of graft patency in patients dying of cardiac causes diverges from that of patients still alive. The method is detailed in Online Appendix 2 (“Pattern Mixture Analysis”).
SAS version 9.4 (SAS Institute, Cary, North Carolina) was used for all analyses.
Diabetes and bypass graft patency
In patients with diabetes, 85% of ITA grafts were free of any stenosis on follow-up angiograms, 10% had 1% to 99% stenosis, and 4.6% were occluded (Figure 3A). In patients without diabetes, 86% of ITA grafts were free of any stenosis, 7.8% had 1% to 99% stenosis, and 6.7% were occluded. Unadjusted ITA graft patency at 1, 5, 10, 15, and 20 years after surgery was 97%, 97%, 96%, 96%, and 96% in patients with diabetes and 96%, 96%, 95%, 94%, and 93% in patients without diabetes, respectively (early p = 0.20; late p = 0.30; Figure 4). Early and late patency of ITA grafts was similar in patients operated on in the 1970s, 1980s, and recent years (1990 to 2011; Online Figure 4A).
In the subgroup of patients with single planned angiograms, 1-year ITA graft patency was 93% in patients with diabetes and 96% in patients without diabetes (p = 0.30).
In patients with diabetes, 47% of SV grafts were free of any stenosis, 20% had 1% to 99% stenosis, and 32% were occluded (Figure 3B). In patients without diabetes, 48% of SV grafts were free of any stenosis, 19% had 1% to 99% stenosis, and 34% were occluded. Unadjusted SV graft patency at 1, 5, 10, 15, and 20 years after surgery was 78%, 70%, 57%, 49%, and 42% in patients with diabetes and 82%, 72%, 58%, 48%, and 41% in patients without diabetes (early p < 0.002; late p = 0.60) (Figure 4). Early patency of SV grafts was better in patients operated on in the 1970s (Online Figure 4B), but long-term patency was better in patients operated on in recent years (1990 to 2011).
In the subgroup of patients with single planned angiograms, 1-year SV graft patency was 80% in patients with diabetes and 86% in patients without diabetes (p = 0.07).
After adjusting for patient characteristics, patients with diabetes had higher, not lower, early patency of ITA grafts than those without diabetes (odds ratio: 0.63; 95% confidence limits: 0.43 to 0.91; p = 0.013), but late patency was similar (p = 0.80) (Central Illustration, Figure 5A, Online Figure 5, Online Table 5). Early and late patency of SV grafts was similar in patients with and without diabetes (early p = 0.9; late p = 0.8) (Figure 5A, Online Figure 5, Online Table 5).
In the subgroup of patients with single planned angiograms, adjusted 1-year patency was similar for both ITA grafts (p = 0.60) and SV grafts (p = 0.40) (Figure 5B) in patients with versus without diabetes. As a form of validation, predicted occlusion at 1 year from the multivariable graft occlusion model (Online Table 5) was similar to observed occlusion at 1 year for both ITA and SV grafts in patients with versus without diabetes (Online Table 6).
Other risk factors for graft occlusion
Other risk factors for early graft occlusion included: 1) female sex; 2) ITA grafts to right coronary artery; and 3) lesser degree of proximal coronary artery stenosis for ITA grafts (Online Table 5). Risk factors for late occlusion included: 1) younger age at time of CABG; 2) asymptomatic patients; 3) higher triglycerides; 4) SV grafts to left circumflex coronary artery; and 5) grafts to non-LAD coronary arteries (Online Table 5).
Work-up bias evaluation
Median time to first coronary angiography was shorter by about 7 months for patients with versus without diabetes (p = 0.0009), and 10 months among propensity-matched patients (p = 0.0005) (Online Table 7). However, median number of angiograms performed per patient was similar, 3 each for patients with and without diabetes, and for propensity-matched patients (Online Table 7). Time to first angiography was similar for patients with and without diabetes in the first year (overall p = 0.09; matched p = 0.70), but thereafter, shorter in patients with than without diabetes (overall p < 0.0001; matched p < 0.0001) (Online Figure 6).
More patients with diabetes (64%) were symptomatic than patients without diabetes (55%; p < 0.0001) at first angiogram overall, but results were similar for matched patients (p = 0.14) (Online Table 8).
Among patients with diabetes, overall survival at 1, 5, 10, 15, and 20 years after CABG was 99.4%, 93%, 73%, 51%, and 35%, respectively; for patients without diabetes it was 99.8%, 98%, 88%, 74%, and 58%, respectively (p < 0.0001) (Online Figures 7A and 7B). Among patients with diabetes, mode of death was cardiac in 173 (45%), noncardiac in 102 (27%), and not known in 107 (28%). Among patients without diabetes, mode of death was cardiac in 1,266 (47%), noncardiac in 720 (26%), and not known in 737 (27%).
Among propensity-matched patients, survival for patients with diabetes at these same time points was 99.4%, 93%, 73%, 51%, and 35%, respectively; for patients without diabetes it was 99.6%, 96%, 83%, 69%, and 58%, respectively (p < 0.0001) (Online Figures 7C and 7D). Among propensity-matched patients with diabetes, mode of death was cardiac in 168 (45%), noncardiac in 101 (27%), and not known in 104 (28%). Among propensity-matched patients without diabetes, mode of death was cardiac in 103 (41%), noncardiac in 63 (25%), and not known in 85 (34%).
Possible influence of cardiac death
For ITA grafts, there was no apparent difference in graft patency trend for patients with diabetes experiencing a cardiac death versus those who were alive. However, there was a slight difference in patients without diabetes: those experiencing a cardiac death had slightly higher patency (about 1%) than those who were alive (Online Figure 8A). In patients with diabetes receiving SV grafts, in every instance the longitudinal pattern of those experiencing a cardiac death was slightly lower within the first few years (by a maximum of 3%) and progressively higher/better beyond about 4 years (Online Figure 8B). In patients without diabetes receiving SV grafts, the pattern for those experiencing cardiac death was everywhere somewhat higher.
We studied 20,066 SV grafts and 7,903 ITA grafts and found no influence of diabetes on ITA or SV graft patency over more than 20 years, contrary to our hypothesis. ITA graft patency remained stable over time, whereas SV graft patency declined progressively for patients with and without diabetes. Patient characteristics associated with worse graft patency included women versus men, younger age, asymptomatic patients, and higher triglyceride levels. Grafting strategies associated with worse graft patency included using an ITA to graft coronaries with a lesser degree of proximal stenosis, an ITA to graft the right coronary artery, and an SV to graft the left circumflex coronary artery. Despite similar long-term graft patency in patients with and without diabetes, long-term survival was worse in those with diabetes.
Findings in context
Studies differ regarding the effect of diabetes on bypass graft patency. Supporting our observations, Schwartz et al. (4) found similar graft patency in patients with and without diabetes using angiographic data from the original BARI (Bypass Angioplasty Revascularization Investigation) trial. ITA graft patency was 89% in patients with diabetes versus 85% in patients without diabetes (p = 0.20), and SV graft patency was 71% versus 75% (p = 0.40), respectively, at a mean follow-up of 3.9 years. Hwang et al. (5) found 5-year arterial graft patency of 95% in patients with diabetes and 91% in patients without diabetes. In their study, early, 1-, and 5-year follow-up angiograms were performed independently of patients’ ischemic symptoms. Goldman et al. (9) studied long-term (10-year) patency of 457 ITA grafts and 1,074 SV grafts and identified risk factors for graft occlusion. Similar to our findings, they did not find diabetes to be a risk factor.
Contrary to our findings, Deb et al. (6) found greater SV graft occlusion by angiography at least 5 years after CABG, 25% in patients with diabetes and 16% in patients without diabetes (p = 0.06). Yilmaz et al. (8) also found diabetes to be associated with worse short-term (≤5 years) SV graft patency. Ayan et al. (7) found similar arterial graft patency in matched patients with and without diabetes, but worse SV graft patency in patients with diabetes. We believe that our study, by virtue of its large sample size, long follow-up, and distinctive statistical methodology, despite the unanticipated result, may reflect the truth.
Effectiveness of coronary artery bypass surgery is related to long-term graft patency. Because bypass grafts are equally durable in patients with and without diabetes, coronary surgery should be as effective in those with diabetes. Nevertheless, long-term survival after CABG in patients with diabetes is worse than for patients without diabetes. Why? The reason could be native-vessel disease progression and heart failure, or noncardiac modes of death. Patients with diabetes have more comorbid conditions than those without diabetes, including hypertension, chronic renal insufficiency, peripheral arterial disease, and higher body mass index (2,12). The BARI investigators (12) compared cause of death in patients with and without diabetes at 5 years after CABG and found similar cardiac mortality (5.8% in patients with diabetes and 4.7% in patients without diabetes), but 5-year noncardiac mortality was much higher in patients with diabetes (13% vs. 5.6%).
Similar to other reports, we found that bypass grafts to non-LAD coronary arteries (13–17) were associated with decreased graft patency. That non-LAD bypass grafts have lower patency than grafts to the LAD might be because it is easier technically to graft anterior coronary arteries, and the amount of myocardium supplied by the LAD is greater than that supplied by other coronary arteries, resulting in a larger blood flow demand being placed on bypass grafts to the LAD.
Grafting a noncritically stenosed coronary artery with an ITA was another risk factor for graft occlusion. Unlike SV grafts, arterial grafts can autoregulate their size depending on flow requirements. When flow requirements are low, such as when ITA grafts are used to bypass coronary arteries with noncritical stenoses, grafts may close as a result of competitive blood flow in the native vessel (18–24). In our study, early ITA graft patency was better in patients with than without diabetes; this could be because patients with diabetes have more diffuse coronary artery disease than patients without diabetes, which may lead to decreased competitive flow in the native vessel. Maximum preoperative proximal coronary stenosis between the bypass graft anastomosis and aorta, the surrogate for competitive flow, could overestimate competitive flow in native vessels in patients with diabetes because the maximum proximal stenosis lesion could be longer due to the diffuse nature of coronary artery disease in these patients.
Progression of vein graft atherosclerosis causes bypass graft patency to decrease with time (25–27). Because ITA grafts are resistant to atherosclerosis, they are much more likely than SV grafts to remain patent in patients with and without diabetes (15,28–30).
Unadjusted early patency of SV grafts was better in patients who underwent surgery in the 1970s. This could be caused by the single planned angiogram at 1 year after CABG in the early 1970s. Unadjusted long-term patency of SV grafts was best in the recent cohort of patients, which could be due to statin use. The Post CABG trial showed that aggressive lowering of low-density lipoprotein cholesterol by statins reduces progression of atherosclerosis in SV grafts (31).
We have previously shown that bilateral ITA grafting maximizes long-term survival in patients with diabetes undergoing CABG (32). However, it is also associated with higher occurrence of deep sternal wound infections. Therefore, we recommend that bilateral ITA grafting be used in patients with diabetes (Figure 2) for whom risk of deep sternal wound infection is low; it might be best to avoid bilateral ITA grafting in obese women with diabetes and diffuse atherosclerotic burden, patients at the greatest risk of developing these infections (32).
Methodological considerations and study limitations
We do not know on a case-by-case basis the indication for postoperative angiography. We presume it most likely was for recurrence of ischemic symptoms. Therefore, it could be argued that the results (that graft patency is similar among patients with and without diabetes) are applicable only to patients with ischemic symptoms who undergo angiography and may not be generalizable to the entire CABG population. To account for this, we studied the influence of diabetes on patency of bypass grafts in a subset of patients who underwent single planned angiography 1 year after surgery. The results of this analysis were similar to the overall results of our study and showed that diabetes was not associated with lower bypass graft patency. Moreover, using the multivariable model for graft occlusion for the overall study population, predicted occlusion at 1 year was calculated for patients undergoing single planned angiography and compared with the actual occlusion for this population. For patients with and without diabetes, and ITA and SV grafts, there was no significant difference between observed and predicted occlusion.
A prospective study with accurate, periodic noninvasive surveillance would be the best method for determining graft patency. Although in some prospective studies high proportions of patients have undergone early angiography, by 5 years, patient dropout because of death, reoperation, and refusal to participate altered the characteristics of the remaining population in nonrandom ways (13,33–35). Accurate, noninvasive advanced imaging with minimal radiation exposure is needed.
Patients with diabetes are less likely than those without diabetes to have symptoms from myocardial ischemia (36,37). Therefore, a patient with diabetes might have more coronary or bypass graft atherosclerotic disease before developing symptoms than would a patient without diabetes and be less likely to undergo angiography. Graft patency among patients with diabetes might be lower than our findings. In fact, contrary to this speculation, median time to first angiography was shorter for patients with diabetes. This could make their graft patency seem lower (earlier angiography, earlier detection of graft closure).
To explore the possible influence of cardiac death on longitudinal estimates of graft patency, we performed a pattern-mixture sensitivity analysis to estimate patency trends separately for patients who experienced a cardiac death and patients alive at the time of follow-up closing date. This analysis demonstrated that cardiac death did not substantially alter the results presented. In particular, it does not suggest that cardiac death caused us to overestimate graft patency.
We did not have long-term medical management data for patients in the study and therefore could not assess its influence on patency of grafts over time, except to observe better late graft patency in the most recent cohort.
Long-term patency of bypass grafts is similar in patients with and without diabetes. Therefore, worse long-term outcomes after CABG in patients with diabetes are likely not related to lower graft patency. Use of ITA grafts should be maximized in all patients undergoing surgical revascularization, because they have excellent patency in patients with and without diabetes even 20 years post-operatively.
COMPETENCY IN MEDICAL KNOWLEDGE: Diabetes does not adversely influence the long-term patency of coronary artery bypass grafts. Worse survival after coronary surgery in patients with diabetes is not related to lower graft patency so much as to comorbidities, progressive atherosclerosis, and complications of diabetes.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Internal thoracic artery grafts should be used as much as possible in patients with and without diabetes undergoing coronary artery surgery because of their excellent patency even 20 years post-operatively.
TRANSLATIONAL OUTLOOK: Further studies are necessary to address the factors that compromise long-term survival of patients with diabetes after coronary bypass surgery and to assess the impact of diabetes on long-term patency of radial, gastroepiploic, and inferior epigastric artery grafts.
For supplemental tables and figures, please see the online version of this article.
Funded in part by the Gus P. Karos Registry Fund; the Sheikh Hamdan bin Rashid Al Maktoum Distinguished Chair in Thoracic and Cardiovascular Surgery, held by Dr. Sabik; and the Kenneth Gee and Paula Show, PhD, Chair in Heart Research, held by Dr. Blackstone. Dr. Raza is a Clinical Research Scholar as part of the Cardiothoracic Surgical Trials Network, and his Master of Science in Management–Healthcare degree is being funded by National Institutes of Health/National Heart, Lung, and Blood Institute grant UM1HL088955. These individuals and funding organizations played no role in the collection of data or analysis and interpretation of the data and had no right to approve or disapprove publication of the finished manuscript. Dr. Sabik is the North American principal investigator for the Abbott Laboratories–sponsored left main coronary disease randomized trial (EXCEL); is on the Society for Thoracic Surgeons board of directors; is on the scientific advisory board of Medtronic; and has received consulting income from Medtronic. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Presented at the American College of Cardiology Scientific Sessions 2016, Chicago, Illinois, April 2–4, 2016.
- Abbreviations and Acronyms
- coronary artery bypass grafting
- internal thoracic artery
- left anterior descending coronary artery
- saphenous vein
- Received July 18, 2016.
- Revision received May 30, 2017.
- Accepted May 31, 2017.
- 2017 American College of Cardiology Foundation
- ↵Society of Thoracic Surgeons National Adult Cardiac Surgery Database Harvest Report, October 2016. Available at: http://www.sts.org/sts-national-database/database-managers/adult-cardiac-surgery-database/harvest-schedule. Accessed March 20, 2017.
- Schwartz L.,
- Kip K.E.,
- Frye R.L.,
- Alderman E.L.,
- Schaff H.V.,
- Detre K.M.
- Ayan M.,
- Saurav A.,
- Kabach A.,
- et al.
- Goldman S.,
- Zadina K.,
- Moritz T.,
- et al.
- ↵(1997) Influence of diabetes on 5-year mortality and morbidity in a randomized trial comparing CABG and PTCA in patients with multivessel disease: the Bypass Angioplasty Revascularization Investigation (BARI). Circulation 96:1761–1769.
- Fitzgibbon G.M.,
- Kafka H.P.,
- Leach A.J.,
- Keon W.J.,
- Hooper G.D.,
- Burton J.R.
- Campeau L.,
- Lesperance J.,
- Hermann J.,
- Corbara F.,
- Grondin C.M.,
- Bourassa M.G.
- Raza S.,
- Sabik J.F. 3rd.,
- Masabni K.,
- Ainkaran P.,
- Lytle B.W.,
- Blackstone E.H.