The Impact of Diabetic Status on Coronary Artery Bypass Graft Patency
Insights From the Radial Artery Patency Study
Background— Despite worse outcomes in diabetics after coronary artery bypass grafting surgery, studies have not examined graft patency in this high-risk group. This study examined the impact of diabetes on graft patency, 1-year postcoronary artery bypass grafting, using data from a multicenter randomized trial.
Methods and Results— The Radial Artery Patency Study enrolled 561 patients undergoing coronary artery bypass grafting, comparing graft patency of the saphenous vein (SV) versus radial artery 1-year postcoronary artery bypass grafting. Angiographic follow-up was acquired for 440 patients (115 diabetics, 325 nondiabetics), each with a study radial artery and a control SV graft. Preoperative characteristics were similar. The proportion of small-sized target vessels was greater in diabetics (P=0.04). At 1 year, 33 of 230 study grafts (14.4%) were occluded in the diabetics versus 63 of 650 (9.7%) in the nondiabetics (P=0.052). Multivariable regression found diabetes to be a significant independent predictor of 1-year graft occlusion (relative risk, 1.45; 95% CI, 1.03 to 2.05; P=0.03) along with female gender, SV conduit, and small target-vessel size. A significantly higher proportion of SV grafts were occluded in the diabetics (19% versus 12%, P=0.04). Radial artery grafting was protective in the diabetic cohort (radial artery: 11 of 115 occluded [9.5%] versus SV: 22 of 115 occluded [19.1%], McNemar corrected P=0.05; relative risk, 0.42; 95% CI, 0.16 to 1.01) and nondiabetics (radial artery: 25 of 325 occluded [7.7%] versus SV: 38 of 325 occluded [11.7%], McNemar corrected P=0.11; relative risk, 0.63; 95% CI, 0.35 to 1.10).
Conclusions— Coronary artery bypass grafting occlusions were more common among diabetics versus nondiabetics at 1-year angiography, mainly because of more frequent SV graft failure in diabetics. Radial artery, compared with SV grafting, is protective in both diabetic and nondiabetic patients.
The prevalence of diabetes mellitus is increasing to epidemic proportions. Diabetes increases the risk of developing coronary artery disease by 2- to 6-fold.1 Their disease severity, compared with nondiabetics, is worse in the form of more extensive, diffuse, and distal coronary artery disease.1,2 Because percutaneous coronary interventions are optimal for well-defined, discrete lesions, the increasing prevalence of diabetes suggests that revascularization will increasingly and preferably be accomplished for such patients by coronary artery bypass grafting (CABG) surgery.3
There is considerable information on coronary revascularization options and subsequent poorer clinical course for patients with diabetes mellitus; however, little has been published on the angiographic status of CABGs in this high-risk cohort.4 Adverse outcomes have not been distinguished from graft failure, native vessel disease progression, or other morbidity prevalent in the diabetic population.4 The interaction between the choice of conduit, saphenous veins versus mammary or other arterial grafts, and diabetes has also not been well described with regard to angiographic patency.3,5
The multicenter Radial Artery Patency Study (RAPS) was a randomized clinical trial that aimed to determine the relative patency of radial arteries to saphenous vein grafts.6 The primary results revealed that fewer radial artery grafts were totally occluded at 1 year compared with saphenous veins. To date, it is the largest prospective, randomized study with angiographic follow-up comparing patency of any arterial conduit with a saphenous vein when used to bypass nonleft anterior descending artery targets.
The primary aim of this study was to determine the influence of diabetic status on overall graft patency, 1 year after CABG, using angiographic data from the multicenter RAPS randomized clinical trial database. Second, we aimed to compare the patency of the radial artery with that of the saphenous vein in nondiabetics and diabetics separately to elucidate any influence of conduit type and diabetes on patency.
Materials and Methods
Institutional ethics board approval was attained for this study; patients provided informed consent.
Study Population and Randomization
The study population was comprised of those enrolled in the multicenter RAPS randomized clinical trial that underwent 1-year follow-up coronary angiography. Details of the study protocol have been previously published.6 Briefly, patients enrolled were <80 years of age undergoing nonemergent primary isolated CABG for graftable triple-vessel disease with an estimated left ventricular ejection fraction >35%. The left internal mammary artery was used to bypass the anterior circulation. Patients were randomly assigned to one of 2 graft strategies: (1) the radial artery was used to graft the circumflex territory and a saphenous vein graft was used for the right coronary system; or (2) the radial artery was directed to the right coronary territory and a saphenous vein graft used for the circumflex territory. The details of the randomization protocol have previously been published.6
All surgeries were performed using cardiopulmonary bypass and cardioplegic arrest using open harvesting techniques for the radial artery and saphenous vein. The free radial artery pedicle was dilated in situ by a slow intraluminal injection of 4 to 5 mL of a dilute solution of 5 mg verapamil and 65 mg papaverine buffered in 16 mL lactated Ringer’s solution. Patients were given 325 mg aspirin daily within 6 hours postoperatively and continued indefinitely. Study protocol dictated that patients receive 1 to 10 μg/kg/min intravenous nitroglycerin during the first 24 hours postoperatively in the intensive care unit. Vasoconstrictor agents were used only in settings of significant peripheral vasodilatation by protocol. Oral nifedipine was initiated on the first postoperative day and continued for 6 months postoperatively for prophylaxis against radial artery spasm.
Patients underwent follow-up angiography 8 to 12 months after surgery. Nitroglycerin was injected into each graft before filming. Angiograms were adjudicated in an independent blinded manner by 2 cardiologists with a third review in the case of disagreement. Patency of only the 2 study grafts (the study radial artery and the study saphenous vein) was required by the study protocol. Grafts were determined to be occluded if no contrast dye injected into the graft opacified the distal coronary artery (ie, TIMI 0 flow).
The primary objective was to compare overall graft occlusion rates for all study grafts comparing diabetics versus nondiabetics. This is an observational, between-patient, nonrandomized approach. Student t test was used to compare continuous variables and χ2 or Fisher exact test was used to compare categorical variables where appropriate. Multivariable logistic regression analysis was used to ascertain significant independent predictors of graft occlusion.
The secondary objective analyzed occlusion rates for the study radial artery versus the study saphenous vein grafts within the diabetic cohort and within the nondiabetic cohort separately. This maintained the within-patient, randomized comparison. Appropriate statistical tests were used, which included the corrected McNemar test for paired proportional data. We used a generalized estimating equations approach to perform multivariable analysis to account for within-patient clustering incorporating patient demographics, operative, anatomic, and medication data.7
Model predictors, specified a priori, included graft type (radial artery or saphenous vein), age in years (continuous), categorical (yes or no) factors such as myocardial infarction within 30 days before operation, gender (male or female), preoperative smoking status, preoperative diagnosis of diabetes, hypertension, hyperlipidemia, peripheral or cerebrovascular disease (stroke, transient ischemic events, endarterectomy, claudication, peripheral angioplasty, bypass or aneurysm repair), preoperative left ventricular function <35%, proximal stenosis of target native coronary vessel graded by percent stenosis (continuous), size of target native coronary vessel graded by the operating surgeon’s visual assessment of the preoperative angiogram (categorical: small, medium, large), number of bypass grafts (continuous), graft to the right coronary versus circumflex artery (yes or no), use of perioperative vasoconstrictors (yes or no), aspirin prescribed at discharge (yes or no), calcium channel blocker prescribed at discharge (yes or no), and lipid-lowering agent prescribed at discharge (yes or no).
Statement of Responsibility
The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
Of the 561 patients randomized in the multicenter RAPS, 440 patients underwent coronary angiography at 1 year postsurgery (mean time of follow-up angiography was 11±4 months and was not different between the diabetic and nondiabetic cohorts). As previously reported, there were no demographic differences between patients who were randomized and those who underwent study angiography.7
Of the 440 patients who underwent angiography, 325 (74%) were nondiabetic and 115 (26%) were diabetic. Within the latter cohort, 93 (80%) were on oral therapy and 22 (20%) required insulin.
Patient baseline characteristics are listed in Table 1. Both the diabetic and nondiabetic cohorts were comparable with the exception of there being more females in the diabetic group (19%) compared with the nondiabetic group (11%; P=0.02). Preoperative angiographic characteristics, also shown in Table 1, demonstrated that the severity of native vessel occlusion was similar in the various coronary territories in the diabetic group compared with the nondiabetic groups. However, the native vessel size was more often smaller in the patients with diabetes when compared to the patients without diabetes (P=0.04). The number of grafts anastomosed and the operative times were no different between the 2 groups of patients. The use of in-hospital vasoconstrictors and the prescription of aspirin, calcium channel blockers, and lipid-lowering agents on discharge were also no different between groups.
Clinical Outcomes and Primary Angiographic End Point
The clinical outcomes at 1 year and the results of our primary end point, overall graft patency in diabetics versus nondiabetics, are summarized in Table 2. There were no deaths in either group. The rate of repeat revascularization by percutaneous intervention was the same in both the diabetic and nondiabetic patients (there were no coronary reoperations). There were significantly more nonfatal perioperative myocardial infarctions in the nondiabetic cohort (12% versus 5%, P=0.02). Myocardial infarction was broadly defined by electrocardiographic criteria that were centrally read by a blinded committee without any information pertaining to clinical history, which may explain the high event rate.
Overall study graft failure, shown in Table 2, was our primary end point. This was an observational, between-patient, nonrandomized analysis. It indicates a higher rate of study graft failure/occlusion in the patients with diabetes (14.4% versus 9.7%, P=0.05). The rate of graft occlusion was even worse for diabetics for the study saphenous veins (diabetics: 19%, nondiabetics: 12%, P=0.04). However, when comparing the study radial arteries between the patients with diabetes and the nondiabetic patients, there was no statistically significant difference.
Multivariable predictors of study graft failure are listed in Table 3. Diabetes, female gender, and small target vessel size were all predictors of study graft failure. The use of the radial artery and more severe target proximal stenosis were protective. Significant interactions were found for graft type by gender and graft type by peripheral vascular disease history.
Secondary End Point
Our secondary objective was to compare graft failure of the study radial artery versus the study saphenous veins within the nondiabetic patients and within the patients with diabetes. This within-patient comparison maintains randomization.
Within nondiabetic patients, radial artery occlusion rate was 8% versus 12% for study saphenous vein grafts. Using corrected McNemar test for paired proportional data suggested that the trend that the use of the radial artery was protective against study graft failure (relative risk, 0.63; 95% CI, 0.35 to 1.10; P=0.1). However, the degree of protection against graft occlusion was of greater magnitude in the higher-risk patients with diabetes. Study radial artery graft occlusion in the diabetic cohort was 10% versus 19% for the study saphenous veins, resulting in a relative risk of 0.42 (radial being protective against graft failure), which was statistically significant (95% CI, 0.16 to 1.01; P=0.05).
The RAPS was a well-designed, multicenter randomized clinical trial in which 440 of 561 patients received 1-year follow-up angiography.6 It is the largest prospective, randomized study with angiographic follow-up comparing the patency of any arterial conduit with a saphenous vein when used to bypass a nonleft anterior descending coronary target. Our secondary analysis of these patients, according to their diabetic status, showed that overall CABG occlusions were more common among diabetics compared with nondiabetics at 1-year angiography. This was mainly attributable to more frequent saphenous vein graft failure in diabetics. Diabetes was a significant, independent predictor of study graft failure, whereas the use of the radial artery was protective against graft occlusion. Furthermore, within-patient randomized analysis demonstrated that the protective effect of radial arteries against graft occlusion was likely greater in the diabetic than the nondiabetic patients.
Although the literature on clinical outcomes in diabetics undergoing surgery is vast, there is a paucity of angiographic follow-up data related to diabetics and graft patency. The Bypass Angioplasty Revascularization Investigation (BARI) trial compared midterm (4-year) CABG graft patency, through angiography, in diabetics versus nondiabetics4 and found that diabetes did not appear to adversely affect graft patency. However, this secondary study from BARI compared the patency of the internal mammary artery with the saphenous vein. It also is limited in that only 99 of the 292 diabetics (34%) and 469 of the 1234 nondiabetics (38%) returned for follow-up angiography. Subsequently, a 1-year angiographic study by van der Meer et al,8 as well an earlier 5-year angiographic follow-up substudy of the Coronary Artery Surgery Study,9 opposed the results from BARI, showing diabetes to be a predictor of graft patency. Again, the grafts reviewed were the internal mammary and saphenous veins.
The radial artery is increasingly being used in clinical practice and contemporary research suggests it may have equivalent or superior early and midterm results when compared with the saphenous vein.4,10–13 Radial artery grafts are thought to be superior to venous conduits because they have been shown in vitro to have preserved endothelial function when grafted.3,14,15 The influence of diabetes on radial artery grafts is not well described in the literature. Choi et al report a series of 517 patients (214 diabetic) undergoing total arterial revascularization through contemporary off-pump CABG. Diabetic status did not emerge as a predictor of 1-year angiographic patency.
This study is one of the first to compare radial artery graft patency with saphenous vein graft patency in patients with diabetes. The within-patient randomized comparison is a result of the RAPS design, which is one strength of this analysis. The sample size and angiographic follow-up are large relative to the cardiac surgical literature. The rigorous multivariable modeling also controls for the inherent limitations of the observational aspects of our study. Findings pertaining to conduit type and gender and peripheral vascular disease were also hypothesis-generating for further study, because the samples within each group were too small to draw any significant conclusions.
In conclusion, graft occlusion 1 year after CABG is significantly higher in diabetics, mainly due to more frequent saphenous vein graft occlusion. However, the use of the radial artery is protective against graft failure compared with the saphenous vein; the benefits even greater in the higher-risk cohort of diabetics.
Source of Funding
The Radial Artery Patency Study was supported by a grant (MT-13883) from the Canadian Institutes of Health Research.
Presented at the American Heart Association Scientific Sessions, November 4–7, 2007, Orlando, Fla.
Schwartz L, Kip KE, Frye RL, Alderman EL, Schaff HV, Detre KM. Coronary bypass graft patency in patients with diabetes in the Bypass Angioplasty Revascularization Investigation (BARI). Circulation. 2002; 106: 2652–2658.
Desai ND, Naylor CD, Kiss A, Cohen E, Feder-Elituv R, Miwa S, Radhakrishnan S, Dubbin J, Schwartz L, Fremes SE; for the Radial Artery Patency Study Investigators. Impact of patient and target-vessel characteristics on arterial and venous bypass graft patency: insight from a randomized trial. Circulation. 2007; 115: 684–691.
van der Meer J, Hillege HL, van Gilst WH, de la Riviere AB, Dunselman PH, Fidler V, Kootstra GJ, Mulder BJ, Pfisterer M, Lie KI. A comparison of internal mammary artery and saphenous vein grafts after coronary artery bypass surgery: no difference in 1-year occlusion rates and clinical outcomes. Circulation. 1994; 90: 2367–2374.
Alderman EL, Corley SD, Fisher LD, Chaitman BR, Faxon DP, Foster ED, Killip T, Sosa JA, Bourassa MG. Five-year angiographic follow-up of factors associated with progression of coronary artery disease in the Coronary Artery Surgery Study (CASS). J Am Coll Cardiol. 1993; 22: 1141–1154.
Verma S, Szmitko PE, Weisel RD, Bonneau D, Latter DA, Erret L, LeClerc Y, Fremes SE. Clinician update: should radial arteries be used routinely for coronary artery bypass grafting? Circulation. 2004; 110: e40–e46.
Gardner TJ. Searching for the second-best coronary artery bypass graft: is it the radial artery? Circulation. 2007; 115: 678–680.
Johnstone MT, Creager SJ, Scales KM, Custco JA, Lee BK, Creager MA. Impaired endothelium-dependent vasodilation in patients with insulin-dependent diabetes mellitus. Circulation. 2001; 38: 1307–1312.