Diabetes in Coronary Artery Disease: Implications for Revascularization
Abstracts & Commentary
Synopsis: Multiple analyses support the recommendation of the initial NHIBI clinical alert recommending surgery over angioplasty in diabetics with both 2 and 3 vessel disease.
Sources: Ledru F, et al. J Am Coll Cardiol. 2001;37:1543-1550; Detre KM, et al. N Engl J Med. 2000;342:989-997; Niles NW, et al. J Am Coll Cardiol. 2001;37:100-115.
It is well known that diabetics have a greater risk of death and a shortened survival following myocardial infarction (MI); and patients with diabetes who undergo revascularization have a less optimal outcome than nondiabetics. In 1996, the BARI Trial, a comparison of coronary artery bypass surgery (CABG) with angioplasty in subjects with multivessel disease, reported that the diabetic cohort enrolled in the trial had a considerably lower 5-year mortality rate in those randomized to CABG (19.4% vs 34.5% with percutaneous intervention [PCI]; P = .003). An NHLBI clinical alert was issued recommending CABG over PCI in diabetic patients. Subsequently, analyses of other comparative revascularization trials and databases have demonstrated conflicting results as to whether diabetics with multivessel disease have a worse prognosis with PCI vs. CABG. Three recent reports bring information to this controversy; all appear to support a strategy of bypass surgery for the diabetic with 2 and 3 vessel coronary disease.
In an angiographic study from France, Ledru and colleagues confirm that diabetic subjects have more diffuse and extensive coronary atherosclerosis, a greater prevalence of mild, moderate, and severe stenoses, and a 2-fold higher vessel occlusion rate than patients without diabetes, after adjustment for multiple risk factors or clinical symptoms. Diabetics categorized into mild diabetes, with fasting plasma glucose in the lowest range, were more comparable to non- diabetics with respect to coronary anatomy and outcome. Four hundred sixty-six consecutive patients referred for coronary angiography were analyzed. Two-thirds of the patients presented with stable or unstable angina. Two independent angiographers analyzed the coronary anatomy lesion, an extent score, severity score, and atheroscrotic score. There were 93 diabetics who were older that had more hypertension and positive family history but otherwise presented with a similar clinical picture.
Enrolled patients had no history of MI. Forty were newly diagnosed diabetics. Typical diabetic dyslipidemia was noted, with high TG, low HDL, and normal LDL cholesterol in the diabetic group as a whole. Analysis of the angiograms indicated substantially more multivessel disease, extent of CAD, atherosclerosis scores, numbers of diseased segments, and total occlusions, after adjustment for multiple risk factors. There was no difference in average stenosis severity between diabetics and nondiabetics. A step-wise assessment of hyperglycemic cohorts indicated an increasing atherosclerotic burden with higher plasma glucose; however, hyperglycemia was a less potent predictor of disease than LDL cholesterol, age, gender, or hypertension. There were more individuals with only mild (25-49%) coronary obstructions in the diabetic cohort; Ledru et al raise the question whether diabetics have more early or young vulnerable lesions susceptible to plaque rupture. They discuss multiple studies in the literature, and point out a variety of problems with interpretation of this data, including variable sample size, various methodologies of assessing CAD severity, and inclusion of only insulin dependent diabetics. They conclude that sustained hyperglycemia in diabetics results in substantially more severe coronary atherosclerosis. They speculate that in type 2 diabetics with relatively low fasting glucose, the somewhat greater prevalence of mild lesions may represent a surrogate for plaque vulnerability and potential rupture, leading to acute myocardial infarction (AMI).
Two other studies examine the relative advantage of CABG over angioplasty in subjects with multivessel coronary disease who are or are not diabetic. Detre and associates performed an analysis of the BARI cohort, with respect to the outcome of patients who had an AMI following the initial revascularization procedure, and provide evidence that CABG appears to protect the diabetic subjects with postrevascularization AMI compared to PCI, with no such differential in survival in the nondiabetics with an AMI after initial enrollment. The original BARI study, as well as the BARI registry cohort, were included in this analysis of 3603 patients, all of whom underwent revascularization. The primary end point was all-cause death; this analysis focused on spontaneous Q-wave MI (clinical or silent), in follow-up of all BARI patients. Diabetic patients randomized to CABG who had an AMI fared substantially better than those randomized to PCI with an AMI. At 5 years after revascularization, 64% of the diabetics and 58% of the nondiabetics had undergone a CABG. The incidence of spontaneous Q-wave MI was 4.8%; diabetics were 1.9 times as likely as nondiabetics to have an MI. There was no difference in the likelihood of Q-wave infarcts between the CABG and PCI patients.
Overall mortality at 5 years was 8% for 2962 patients without diabetes and 20% for the 641 patients with diabetes, (P = < .001). The mortality in diabetics was 18% in those who had CABG and 25% in those who did not. In diabetics who had a subsequent MI, "the protection provided by CABG was dramatic;" mortality was 17% in those who underwent CABG vs. 80% in those who had a spontaneous Q-wave MI and underwent angioplasty. The mortality in the diabetic CABG patients was comparable to the nondiabetics. Thus, CABG in the diabetic cohort resulted in 81% risk reduction for death (P = .001) in those subjects who had a spontaneous Q-wave MI, and a 35% reduction in mortality (P = .02) in the diabetics who did not have a spontaneous infarct. Detre et al point out that relatively few diabetics subsequently had a spontaneous Q-wave MI, and most importantly, that CABG reduced mortality among the majority of diabetics who did not have an MI when compared to the angioplasty strategy. Nondiabetics who received a CABG did not have a statistically significant reduction in mortality with or without a subsequent Q-wave infarction. Detre et al emphasize the apparent benefit of using internal thoracic artery grafts as a primary CABG strategy. Patients who underwent angioplasty had substantially more jeopardized myocardium than those who had CABG. They suggest that the PCI strategy resulted in a greater burden of residual ischemic myocardium than with CABG.
A recent analysis of a large database from the New England Cardiovascular Disease Study Group supports the view that bypass grafting is a better strategy for multivessel revascularization in diabetics. Niles and colleagues evaluated a large number of diabetic patients who underwent coronary revascularization between 1992 and 1996 in 4 states, including Massachusetts. A complex statistical analysis and assessment of multiple variables, including the priority of need for revascularization, was included in this observational study. The study population consisted of 2756 patients with diabetes and multivessel disease. Seven hundred thirty-six underwent PCI and 2030 underwent CABG, all at the discretion of the physician. Primary outcome was all-cause mortality at 5 years. CABG patients had far more complete revascularization than PCI individual subjects, who tended to undergo a "culprit lesion strategy," with at least 75% of such individuals having incomplete revascularization. On the other hand, the PCI individuals were younger and healthier than the CABG cohort. Stents and IIb/IIIa drugs were not used in these patients. This was an observational study, it is not surprising that 56% of the CABG cohort had 3 vessel disease compared to 16% of PCI patients. The results, using adjusted survival outcomes, indicated a higher long-term mortality in the PCI patients, who were 49% more likely to die than those who received CABG (P = .037). Hospital mortality was similar; curve separation began at approximately 3 months. Triple vessel disease more than doubled the mortality risk for PCI vs. CABG. There was a 33% increase in mortality in the PCI patients with 2 vessel disease who received PCI compared to CABG. Niles et al reviewed 5 other cohorts, including the BARI patients. All studies demonstrated a trend favoring CABG over PCI, with the EAST trial insulin diabetics and the BARI randomized trial clearly favoring CABG. As mentioned, the long-term analysis of BARI, including the registry, demonstrates a major benefit for CABG. Furthermore, late analysis of the EAST trial also resulted in a favorable survival with CABG. Niles et al conclude that their analysis supports the recommendation of the initial NHIBI clinical alert recommending surgery over angioplasty in diabetics with both 2 and 3 vessel disease.
Comment by Jonathan Abrams, MD
The literature is rich with respect to reports on revascularization outcomes in diabetics. The subject is of considerable importance, in that approximately 6% of the US population has overt diabetes, with perhaps 2 times as many patients who do not know that they have diabetes or are prediabetic. Coronary disease accounts for 50-70% of the mortality in type 2 diabetics. The BARI observations have triggered multiple retrospective as well as prospective studies, culminating in these reports. Expert consensus guidelines, clinical trials, and prominent interventional clinical investigators, such as William O’Neil, support a preferential strategy of CABG for diabetics who have multivessel disease. The Northern New England Cardiovascular Disease Study Group report underlines the difficulty in carrying out such a blanket recommendation, by documenting that the large number of angioplasties performed in diabetics provide incomplete revascularization by addressing the lesion responsible for an AMI or acute coronary syndrome, leaving other obstructive lesions unprotected. The French angiographic study confirms multiple other reports that the degree of coronary atherosclerosis in the diabetic is more extensive, diffuse, and severe, making it difficult for multiple PCI interventions to "cure" or adequately treat the entire coronary circulation. Furthermore, there is evidence that multiple factors in the diabetic patient result in less good outcomes following angioplasty even with stenting. These include a prothrombotic-platelet-activated state, oxidative stress, and glycation of proteins. Stenting improves the outcomes in diabetics undergoing PCI, but this approach does not provide the diabetic with a comparable long-term outlook as the nondiabetic following PCI. Furthermore, other data from the Montreal Heart Institutes suggest that late stent stenosis, resulting in decreased LV function, may play a little recognized but important role in the adverse outcome of diabetics undergoing angioplasty with stenting.
Conclusions
The large database, in the aggregate, supports a generic recommendation for bypass grafting in the diabetic with severe 2 and 3 vessel disease warranting revascularization. Whether the widespread use of internal thoracic arteries is responsible for some or most of the protection with surgery is speculative; outcomes in nondiabetics in EAST, BARI, and other databases do not suggest a substantial outcome differential in the nondiabetic who receives a PCI vs. CABG. Physicians caring for diabetics who refer these patients for cardiac catheterization or interventions need to understand the robust data underlying the construct that bypass surgery is the optimal revascularization strategy for the diabetic with multivessel disease.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content