Coronary artery revascularization in stable patients with diabetes mellitus

INTRODUCTION — Coronary heart disease (CHD) is a major cause of morbidity and mortality among patients with diabetes mellitus. Compared with nondiabetic patients, individuals with diabetes, and especially those with kidney disease, are more likely to have CHD. Furthermore, patients with diabetes are more likely to have coronary artery disease that is complex, diffuse, and involves multiple vessels. Atherosclerosis tends to be more rapidly progressive and occurs at an earlier age. Diabetic patients with CHD have a lower long-term survival rate than nondiabetic patients with CHD [1]. (See "Prevalence of and risk factors for coronary heart disease in patients with diabetes mellitus".)

Indications for the use of coronary artery revascularization are generally similar in patients with and without diabetes. However, patients with diabetes are at higher risk of future cardiovascular events following revascularization with either percutaneous coronary intervention or coronary artery bypass graft surgery [2]. This increase in risk is likely due to more advanced coronary artery disease, more rapidly progressive disease, and comorbidities associated with diabetes [3]. (See 'Outcomes' below.)

This topic will focus on our approach to coronary artery revascularization in stable patients with diabetes. Issues related to medical versus interventional therapy for stable angina in nondiabetic patients and silent ischemia are discussed separately. (See "Chronic coronary syndrome: Indications for revascularization" and "Silent myocardial ischemia: Epidemiology, diagnosis, treatment, and prognosis".)

Our approach to revascularization in patients with acute coronary syndromes, with or without diabetes, is presented elsewhere. (See "Non-ST-elevation acute coronary syndromes: Selecting an approach to revascularization" and "Acute ST-elevation myocardial infarction: Selecting a reperfusion strategy".)

INDICATIONS FOR REVASCULARIZATION — We believe that the indications for revascularization in stable patients are similar for those with and without diabetes. The decision to proceed with revascularization, as opposed to continuing medical therapy, is made in three groups:

●Patients with activity-limiting symptoms despite maximum medical therapy.

●Active patients who want revascularization for improved quality of life compared with medical therapy, such as those who are not tolerating medical therapy or want to increase their activity level.

●Patients with anatomy for which revascularization has a proven survival benefit, such as significant left main coronary artery disease (CAD; greater than 50 percent luminal narrowing) or multivessel CAD with a reduction of left ventricular ejection fraction (LVEF) and a large area of potentially ischemic myocardium. (See "Chronic coronary syndrome: Indications for revascularization".)

OUR APPROACH TO REVASCULARIZATION — For most patients with diabetes who have multivessel or complex disease and indications for revascularization (see 'Indications for revascularization' above), coronary artery bypass graft surgery (CABG) is preferable to percutaneous coronary intervention (PCI). This preference is based on the results of randomized clinical trials. The FREEDOM trial found that outcomes were improved with CABG. These findings are consistent with the SYNergy Between PCI With TAXus and Cardiac Surgery (SYNTAX) trial and a patient-level meta-analysis [4,5]. For patients with low SYNTAX score (see "Revascularization in patients with stable coronary artery disease: Coronary artery bypass graft surgery versus percutaneous coronary intervention", section on 'Outcomes based on lesion severity'), single-, or two-vessel disease who have lesions that are focal and amenable to PCI, we believe that PCI is a reasonable alternative. In patients with three-vessel disease and a low SYNTAX score, it is not unreasonable to consider PCI, particularly for those with focal stenoses and increased surgical risk.

Regardless of the procedure chosen, all patients with diabetes and CAD should be treated with aggressive risk factor reduction, including optimal management of diabetes, blood pressure control, aspirin, lipid lowering therapy, and lifestyle changes including exercise, dietary modifications, and smoking cessation when indicated. (See "Prevention of cardiovascular disease events in those with established disease (secondary prevention) or at very high risk".)

OUTCOMES — Patients with diabetes comprise as many as 25 to 30 percent of those who undergo revascularization, consistent with the approximately 10 percent prevalence of diabetes in the general population and their two- to threefold increased risk for cardiovascular disease compared with the nondiabetic population. Procedural success is similar in the two groups, but long-term outcomes after revascularization with either percutaneous coronary intervention (PCI) or coronary artery bypass graft surgery (CABG) are worse in diabetic compared with nondiabetic patients which may reflect the burden of comorbidities in patients with diabetes [2].

Revascularization versus medical therapy for stable angina — In the broad population of patients with diabetes, risk of death and future myocardial infarction are similar between revascularization and optimal medical therapy [6-9]. Patients at high angiographic or clinical risk may benefit from revascularization [10].

The best available evidence regarding a comparison of revascularization and optimal medical therapy comes from the Bypass Angioplasty Revascularization Investigation 2 Diabetes (BARI) 2D trial [6,7]. In BARI 2D, 2368 patients with type 2 diabetes mellitus and nonacute ischemic heart disease (defined as either a ≥50 percent stenosis of a major epicardial coronary artery associated with a positive stress test or ≥70 percent stenosis and classic angina) were randomly assigned to either prompt revascularization (PCI or CABG) plus medical therapy or medical therapy only. The most appropriate mode of revascularization (PCI or CABG) was determined by the investigator prior to randomization. At five years, the primary end points of the rates of survival or freedom from major cardiovascular events (eg, death, myocardial infarction [MI], or stroke) did not differ significantly between the prompt revascularization group and the intensive medical therapy alone group (88.3 versus 87.8 percent and 77.2 versus 75.9 percent, respectively). However, in subgroup analysis, the rate of freedom from major cardiovascular events was significantly higher with CABG (77.6 versus 69.5 percent), predominantly attributable to a reduction in nonfatal MI. In the PCI stratum, there were no differences in major adverse cardiovascular events with prompt revascularization, when compared with intensive medical therapy alone (77.0 versus 78.9 percent).

In the COURAGE trial, an initial revascularization strategy with PCI did not reduce the risk of death or MI when compared with an initial trial of medical therapy in a broad group of patients (see "Chronic coronary syndrome: Indications for revascularization", section on 'Indications'). However, patients treated with initial revascularization were more likely to be free from angina at one year (66 versus 58 percent). In the subgroup of patients with diabetes (n=766), patients treated with initial medical therapy had similar cardiovascular outcomes as patients treated with revascularization (HR 0.99; 95% CI 0.73-1.32) [8,9].

In the ISCHEMIA trial, patients with chronic coronary disease were randomly assigned to an invasive (medical therapy plus angiography and revascularization if feasible) or a conservative approach (medical therapy alone with revascularization if medical therapy failed). In this study, 43 percent of participants had diabetes. In a diabetes substudy of ISCHEMIA, participants with diabetes had increased hazards for death or MI over a median 3.1 years of follow-up (HR 1.49; 95% CI 1.31-1.70) [11]. However, there was no difference in death or MI according to initial invasive strategy versus medical therapy. Similarly, in the REVIVED-SCIS2 trial of patients without angina who had left ventricular dysfunction with evidence of myocardial viability, 41 percent had diabetes [12]. In this study, patients who had revascularization with percutaneous coronary intervention compared with medical therapy had similar risks of death or hospitalization for heart failure.

In an attempt to understand further which patients might benefit from CABG, a risk score was developed from baseline angiographic (myocardial jeopardy index, total number of coronary lesions, prior coronary revascularization, and LVEF) and clinical variables (Framingham Risk Score) [10]. Patients were then stratified into high- and low-risk subgroups using either angiographic or clinical variables or both. The composite endpoint of death, MI, or stroke occurred significantly less often in the CABG stratum than in the intensive medical therapy group (24.8 versus 36.8 percent) among those in the highest angiographic risk tertile, and this effect was even more pronounced in patients with both high angiographic and high clinical risk (27.1 versus 47.3 percent; hazard ratio 2.10, p = 0.0051). Although this analysis is limited by small sample size in subgroups and lack of validation of the angiographic risk score used, it suggests that in patients with diabetes, those at highest angiographic and clinical risk benefit the most from revascularization with CABG. In the 381 patients from the CABG stratum that were considered high risk, the rates of death, MI, or stroke at five years were 24.8 percent in the CABG arm and 36.8 percent in the medical therapy arm. In contrast, there were no differences between PCI and medical therapy in the high-risk PCI stratum (29.7 versus 29.9 percent).

After percutaneous coronary artery intervention — Among diabetic patients who undergo elective PCI, the procedural success rates are similar to patients without diabetes [13-17]. However, patients with diabetes have increased rates of long-term restenosis and progression of coronary disease compared with patients without diabetes, as well as lower rates of event-free survival [13-15,17-27]. The following are examples of this point:

●In an analysis of 10,778 patients in the j-Cypher registry who underwent PCI with sirolimus-eluting stents, there were 966 patients with insulin-treated diabetes, 3404 with diabetes not treated with insulin, and 6378 without diabetes [24]. At three years, the rate of target lesion revascularization was significantly higher in the insulin-treated and noninsulin-treated diabetes groups compared with those without diabetes (19, 14, and 10 percent, respectively). Predictors of restenosis in diabetic patients included smaller vessel caliber, greater length of the stented segment, and lower body mass index, similar to the broad population of patients with CAD.

●The SYNTAX trial, which compared PCI with CABG in patients with complex CAD (left main and/or three vessel disease), included 452 patients with diabetes [28]. At five-year follow-up, the rate of repeat revascularization of patients undergoing PCI (n = 231) was higher in patients with diabetes (29 versus 19 percent). Mortality was also increased in the diabetic population (20 versus 12 percent). (See "Revascularization in patients with stable coronary artery disease: Coronary artery bypass graft surgery versus percutaneous coronary intervention", section on 'Outcomes based on lesion severity'.)

After coronary artery bypass graft surgery — Among patients who undergo elective CABG, the procedural short-term success rates are similar in patients with and without diabetes. However, during follow-up, death and adverse nonfatal outcomes after CABG are higher in patients with diabetes.

In addition, the risk of sternal wound infection and renal failure is higher in patients with diabetes [29-31]. It has been suggested that bilateral internal thoracic (mammary) artery grafting may be contraindicated in diabetic patients in whom the incidence of sternal wound infection is as high as 14 percent [31]. Maintenance of strict glycemic control during the perioperative period and skeletonization of the internal mammary artery during harvesting are associated with lower rates of sternal wound infection [32,33]. (See 'Intraoperative and postoperative glycemic control' below and "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Sternal wound infection and mediastinitis' and 'Internal thoracic artery grafts' below.)

Although diabetes does not appear to affect in-hospital mortality after CABG, short- and long-term survival after CABG are significantly reduced in diabetic patients. In different large observational studies, compared with nondiabetic patients, patients with diabetes had higher mortality rates at 30 days (5 versus 2.5 percent) and at 5 and 10 years (22 versus 12 percent and 50 versus 29 percent, respectively) [29,30,34-42]. As with mortality, nonfatal coronary outcomes after CABG also may be somewhat more common in diabetic patients. In a report of over 12,000 patients, diabetes was associated with a significantly lower rate of freedom from PCI at 10 years (83 versus 86 percent) but no difference in either MI or additional CABG at 5 or 10 years [40]. Another report evaluated 1615 patients who underwent CABG at a single center and had one or more repeat angiograms, and they were followed for ≥25 years [41]. Diabetes was associated with a relative risk of 1.39 for progression of atherosclerosis or recurrent symptoms.

PCI versus CABG — For most patients with diabetes and complex multivessel CAD who require revascularization (see 'Revascularization versus medical therapy for stable angina' above), the evidence suggests outcomes are improved with coronary artery bypass graft surgery (CABG). The evidence comes from the FREEDOM trial and subgroup analysis of the BARI 2D trial of patients with diabetes (see 'Revascularization versus medical therapy for stable angina' above) and subgroup analyses of the SYNTAX trial. (See "Revascularization in patients with stable coronary artery disease: Coronary artery bypass graft surgery versus percutaneous coronary intervention", section on 'Outcomes with coronary artery bypass graft surgery and percutaneous coronary intervention'.)

Prior to coronary revascularization, the patient should have the relative benefits and costs of PCI and CABG fully described.

Left main coronary artery disease and SYNTAX score ≤32 — Similar to patients with without diabetes, CABG or PCI may be reasonable revascularization approaches in those with diabetes. Factors such as the complexity of associated disease have been identified that may help in deciding between the two. (See "Left main coronary artery disease", section on 'Percutaneous coronary intervention versus coronary artery bypass graft surgery'.)

The EXCEL trial randomized patients with significant left main coronary artery stenosis and low/intermediate SYNTAX score (≤32). All patients had to be candidates for both PCI or CABG. [2] (see "Left main coronary artery disease", section on 'Randomized trials'). In EXCEL, there were 1905 patients (554 patients with diabetes) assessed to have low or intermediate CAD complexity based on a SYNTAX score ≤32. The following findings were noted in EXCEL:

●The three-year composite primary end point (all-cause death, stroke, or MI) was higher in patients with diabetes compared with those without (20.0 versus 12.9 percent; hazard ratio [HR] 1.60, 95% CI 1.26-2.04).

●In patients with diabetes, the rate of the primary end point was similar between those who received PCI or CABG (20.7 versus 19.3 percent, respectively; HR 1.03, 95% CI 0.71-1.50).

●The rate of all-cause death was higher in those treated with PCI (13.6 versus 9.0 percent, respectively; p = 0.046), and there was no significant interaction regarding diabetes status. However, there were only 59 total deaths in the subgroup with diabetes (n = 38 [PCI] versus n = 21 [CABG]), so the study was underpowered to evaluate death as a single endpoint. (See "Revascularization in patients with stable coronary artery disease: Coronary artery bypass graft surgery versus percutaneous coronary intervention", section on 'Outcomes based on lesion severity'.)

Multivessel or complex disease — In the general population of patients who require coronary revascularization, CABG is preferred to PCI in many patients with multivessel disease. In patients without multivessel coronary artery disease, complex anatomy such as distal left main or proximal left anterior descending coronary artery stenosis, or heavily calcified lesions in tortuous vessels, may influence decision making in favor of CABG. (See "Revascularization in patients with stable coronary artery disease: Coronary artery bypass graft surgery versus percutaneous coronary intervention", section on 'Our approach'.)

The FREEDOM trial is the largest relevant randomized controlled trial evaluating PCI and CABG in patients with multivessel coronary artery disease. In the FREEDOM trial, 1900 patients with diabetes and multivessel CAD (83 percent three-vessel disease) were randomly assigned to either PCI with early generation drug-eluting stents (DES) or CABG. Both groups received optimal medical therapies for the secondary prevention of cardiovascular disease [43]. The median follow-up among survivors was 3.8 years. The following findings were noted:

●The primary composite outcome of death from any cause, nonfatal MI, or nonfatal stroke occurred significantly more often in the PCI group at five years (26.6 versus 18.7 percent; absolute difference of 7.9 percent, 95% CI 3.3-12.5). The benefit from CABG existed in all SYNTAX score subgroups. (See "Revascularization in patients with stable coronary artery disease: Coronary artery bypass graft surgery versus percutaneous coronary intervention", section on 'Outcomes based on lesion severity'.)

●With regard to the individual components of the composite end point, all-cause mortality and MI were significantly higher with PCI (16.3 versus 10.9 and 13.9 versus 6.0 percent, respectively), whereas stroke occurred significantly less often (2.4 versus 5.2 percent).

●At one year, there was a higher rate of repeat revascularization in the PCI group (12.6 versus 4.8 percent; HR 2.74, 95% CI 1.91-3.89).

The SYNTAX trial compared first-generation DES with CABG in patients with multivessel coronary artery disease. In this trial, 1800 patients with three-vessel or left main coronary artery disease were randomized to paclitaxel-eluting stent or CABG. A total of 452 patients (25 percent) had diabetes. The following findings were noted:

●In the overall trial cohort at 12 months, major adverse cardiac or cerebrovascular events were higher in the PCI group (17.8 percent [PCI] versus 12.4 percent [CABG]; p = 0.002). The results were consistent in the subgroup with diabetes.

●At three years, in the cohort with diabetes, major adverse cardiac and cerebrovascular event rates were higher with PCI (37 percent [PCI] versus 22.9 percent [CABG]; p = 0.002). The need for repeat revascularization was significantly higher in patients with diabetes treated with PCI (28 percent [PCI] versus 12.9 percent [CABG]; p<0.001) [44].

●At 10 years, there were no significant differences in death between patients treated with PCI compared with CABG (27 percent [n = 244 with PCI] versus 24 percent [n = 211 with CABG]; p = 0.09). In the cohort with diabetes, in whom 10-year outcomes were available in 452, a total of 150 had died. There were no differences in the risk of death at 10 years in patients treated with PCI compared with CABG (HR 1.10; 95% CI 0.80-1.52) [5].

These data should be interpreted with the understanding that the DES used in this trial (paclitaxel-eluting) has been shown to be less effective and inferior to the current generation of DES.

A patient-level meta-analysis that included data from 11 randomized trials evaluating PCI and CABG provides additional evidence for the superiority of CABG for patients with diabetes and multivessel coronary artery disease [4].

●Mortality at five years was significantly higher with PCI in those patients with diabetes and multivessel coronary artery disease compared with CABG (15.5 versus 10.0 percent; HR 1.48; 95% CI 1.019-1.84; p = 0.0004).

●There was no difference in mortality between PCI and CABG in patients with multivessel without diabetes (HR 1.08; 95% CI 0.86-1.36; p = 0.49). (See "Revascularization in patients with stable coronary artery disease: Coronary artery bypass graft surgery versus percutaneous coronary intervention", section on 'Outcomes with coronary artery bypass graft surgery and percutaneous coronary intervention'.)

One- or two-vessel disease — For patients with single-vessel or noncomplex disease for whom revascularization is indicated (see 'Indications for revascularization' above), PCI is a reasonable choice. While this issue has not been specifically studied, most of our experts prefer PCI to CABG.

Previous coronary artery bypass graft surgery — The optimal revascularization approach in diabetic patients with post-CABG angina is not known. The general approach to the management of post-CABG angina is discussed separately. (See "Late recurrent angina pectoris after coronary artery bypass graft surgery", section on 'Management'.)

PCI is a reasonable first choice for most patients with diabetes and unacceptable angina (with appropriate anatomy) after CABG. Repeat CABG can be considered for patients with multigraft failure or no prior placement of a left internal thoracic artery graft.

PCI CONSIDERATIONS — Among patients with diabetes who undergo percutaneous coronary intervention (PCI), outcomes such as the rates of target lesion revascularization, MI, or survival may be influenced by the stent type as well as diabetes-related factors.

Stent type — Similar to the general population of patients who undergo PCI with stent placement, patients who receive current generation drug-eluting stents (DES) have a lower rate of restenosis and subsequent revascularization than those who receive either bare metal stents or early generation DES [45-58]. The newest generation of ultra-thin-strut stents may reduce target lesion failure and MI [59]. However, a patient-level meta-analysis of ultra-thin stents in patients with diabetes did not demonstrate increased benefit [60]. (See "Intracoronary stents: Stent types", section on 'Ultra-thin-strut bioresorbable drug-eluting stents'.)

Diabetes-related factors — Several diabetes-related clinical factors have been evaluated for their impact on outcome after PCI:

●The rate of restenosis after PCI in diabetic patients may be lower in the setting of lower glycemia. This was suggested in an analysis from 2004 of 239 patients (179 with diabetes) who underwent elective PCI (67 percent with stenting) after measurement of glycated hemoglobin (HbA1c) [61]. Diabetic patients with a preprocedural HbA1c ≤7 percent had a rate of target vessel revascularization at 12 months that was comparable to that in nondiabetic patients (15 versus 18 percent), while diabetic patients with an HbA1c >7 percent had a significantly higher rate of target vessel revascularization (34 percent). Generalization is limited by the study design (non-randomized observational study), its small size, and potential for confounding.

●The experimental observations that the thiazolidinediones inhibit vascular smooth muscle cell proliferation and migration and reduce intimal proliferation after vascular injury provided the rationale for assessing their effect on limiting in-stent restenosis in both nondiabetic and diabetic patients. (See "Thiazolidinediones in the treatment of type 2 diabetes mellitus".)

The potential efficacy of this approach has been evaluated in two small randomized trials in which patients with type 2 diabetes who underwent PCI with stenting were randomly assigned to either a thiazolidinedione (pioglitazone or rosiglitazone) or placebo [62,63]. In both studies, the rate of angiographic restenosis was significantly lower in the thiazolidinedione group. Further studies are required to confirm the efficacy and safety of this approach, particularly with the use of DES since stent technology has evolved since the conduct of these studies.

Lipid-related factors — Lipid-lowering therapy is the cornerstone of secondary prevention measures to reduce future coronary events in patients following PCI. (See "Prevention of cardiovascular disease events in those with established disease (secondary prevention) or at very high risk", section on 'Dyslipidemia'.)

CORONARY ARTERY BYPASS GRAFT SURGERY CONSIDERATIONS

Comorbidities — The rates of long-term death and adverse nonfatal outcomes after coronary artery bypass graft surgery (CABG) are higher in patients with diabetes compared with those without diabetes. (See 'Outcomes' above.)

A variety of factors contribute to the worse prognosis. Diabetic patients who undergo CABG have a worse risk profile, tending to be older with more three-vessel disease and lower LVEF than nondiabetic patients [34,36,39,64].

Comorbidities of diabetes influence post-CABG survival. This was illustrated in a cohort of 36,641 patients undergoing isolated CABG between 1992 and 2001: 70 percent did not have diabetes, 22 percent had diabetes without renal failure (requiring preoperative dialysis) or vascular disease (peripheral or cerebrovascular), and 8 percent had diabetes with renal failure and/or vascular disease [65]. The annual mortality rate for patients without diabetes (3.1 percent) was similar to that for patients with diabetes but no renal failure or vascular disease (4.4 percent). Patients with diabetes who had renal failure and/or vascular disease had a significantly higher annual mortality rate (9.4 percent).

Another observational study of 905 patients found (on multivariate analysis) that an elevated serum creatinine (not renal failure as in the prior study) was an independent predictor of five-year mortality (adjusted hazard ratio 1.5) [66]. Proteinuria was also an independent predictor of mortality.

Preoperative glycemic control is associated with outcomes following CABG. In 3555 consecutive patients undergoing elective CABG at an academic medical center between 2002 and 2006, increasing HbA1c was associated with increased risk of mortality, MI, and deep sternal wound infection [67]. After adjusting for potential confounders, there was an increased risk of mortality (odds ratio 1.40; 95% CI 1.06-1.86) for every 1 percent increase in HbA1c.

Internal thoracic artery grafts — All patients undergoing CABG, including those with diabetes, should receive an internal thoracic (mammary) artery (ITA) graft to improve survival. (See "Coronary artery bypass graft surgery: Graft choices", section on 'Arterial grafts'.)

In addition, we prefer bilateral to single ITA grafting for many patients with diabetes based on two observational studies. This recommendation is made despite the fact that the rate of sternal wound infection may be modestly increased with bilateral compared with single ITA grafting and that diabetes may be a risk factor for the development of postoperative mediastinitis [68,69]. (See 'After coronary artery bypass graft surgery' above and "Postoperative mediastinitis after cardiac surgery", section on 'Risk factors'.)

The issue of whether bilateral ITA improves outcomes compared with single ITA in patients with diabetes was evaluated in a study of 1107 consecutive patients who received either a single or bilateral ITA [70]. After propensity matching, two cohorts of 414 individuals were created. There was no significant difference between the single and bilateral ITA groups in operative mortality (2.4 versus 3.1 percent, respectively), sternal wound infection (1.7 versus 3.1 percent), or total complications (17.1 percent). However, late median survival was significantly lower with single ITA grafting (9.8 versus 13.1 years). Similar improvement in long-term survival after placement of bilateral ITA among patients with diabetes was reported in another 2012 observational study [71].

Radial grafts should be considered in patients undergoing CABG, particularly in those patients with diabetes who are relatively young. In a meta-analysis of six trials evaluating the use of radial artery grafts in patients undergoing CABG (mean age 67 and 35 percent with diabetes), the use of radial grafts reduced the rate of death, MI, or revascularization when compared with saphenous vein grafts (hazard ratio [HR] 0.67, 95% CI 0.49-0.90) [72]. Radial arterial grafts were also at lower risk of graft failure (HR 0.44; 95% CI 0.28-0.70). Outcomes were similar in patients with and without diabetes.

Intraoperative and postoperative glycemic control — For most patients with diabetes, we attempt intensive glycemic control intra- and postoperatively to improve both cardiac and noncardiac outcomes in diabetic patients after CABG [32,73,74]. The following observations illustrate the range of findings:

●In a prospective trial, 141 patients with diabetes who were undergoing CABG were randomly assigned to tight glycemic control (125 to 200 mg/dL [6.9 to 11.1 mmol/L]) with a glucose-insulin-potassium (GIK) infusion or standard therapy with intermittent subcutaneous insulin [74,75]. Patients treated with GIK had lower mean serum glucose concentrations in the perioperative period (138 versus 260 mg/dL [7.7 versus 14.4 mg/dL]). At two years, GIK patients had significantly lower rates of wound infections (1 versus 10 percent), recurrent ischemia (5 versus 19 percent), and mortality (2 versus 10 percent).

●In a randomized trial [76], 352 patients (with or without known diabetes) undergoing CABG were randomized to intensive glucose target (100 to 140 mg/dL) or to a conservative target (141 to 180 mg/dL). In the cohort with diabetes, there were no differences in the rates of complications in patients treated in the intensive or conservative group (49 versus 48 percent, p = 0.87). (See "Early noncardiac complications of coronary artery bypass graft surgery", section on 'Sternal wound infection and mediastinitis' and "Coronary artery bypass surgery: Perioperative medical management", section on 'Glycemic control'.)

RECOMMENDATIONS OF OTHERS — Our recommendations are similar to those from the American College of Cardiology Foundation/American Heart Association/American Association for Thoracic Surgery/Preventive Cardiovascular Nurses Association/Society for Cardiovascular Angiography and Interventions/Society of Thoracic Surgeons in 2014 and those from the European Society of Cardiology/European Association for the Study of Diabetes in 2013 [77-81].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Percutaneous coronary intervention" and "Society guideline links: Coronary artery bypass graft surgery".)

SUMMARY AND RECOMMENDATIONS

●The indications for revascularization in stable patients are similar for those with and without diabetes. (See 'Indications for revascularization' above.)

●Indications – For patients with diabetes and stable angina who do not have a clear indication for revascularization, we suggest intensive medical therapy as the initial strategy rather than immediate revascularization (Grade 2B). (See 'Revascularization versus medical therapy for stable angina' above.)

For patients with diabetes and multivessel disease (particularly if it is complex) who require revascularization, we recommend CABG rather than percutaneous coronary intervention (PCI) (Grade 1B). (See 'PCI versus CABG' above.)

For patients with diabetes undergoing CABG, we suggest placement of bilateral, rather than single, internal thoracic artery (ITA) grafts (Grade 2C). It is reasonable to not perform bilateral ITA grafting in individuals older than 70 to 75 years, as the survival benefit from bilateral ITA may be small in older individuals. (See 'Internal thoracic artery grafts' above.)

For patients with diabetes who have recurrence of unacceptable angina after CABG, we suggest repeat revascularization with PCI for appropriate lesions (Grade 2C). (See 'Previous coronary artery bypass graft surgery' above.)

For patients with diabetes who undergo CABG, we suggest tight glycemic control intra- and postoperatively to improve both cardiac and noncardiac outcomes (Grade 2C). (See 'Intraoperative and postoperative glycemic control' above.)

ACKNOWLEDGMENT — The UpToDate editorial staff thank Dr. Arie P. Kappetein for his past contributions as a section editor to prior versions of this topic review.