Impact of Diabetes on Five-Year Outcomes after Vein Graft Interventions Performed Prior to the Drug-Eluting Stent Era

While an increasing proportion of the grafts used during coronary artery bypass (CABG) procedures are arterial, the majority of grafts implanted at most centers have been and continue to be saphenous vein grafts. On long-term follow up, up to 50% of such vein grafts may be occluded.^1,2 Consequently, a significant number of patients with previous CABG eventually require percutaneous coronary intervention (PCI) of diseased vein grafts. Compared to native coronary arteries, saphenous vein grafts (SVG) have a higher plaque burden and are more likely to have lesions that are diffuse, concentric and friable.³ These factors contribute to the significantly worse early and long-term outcomes seen after SVG PCI.^4–7Diabetes mellitus is also an important clinical factor associated with worse outcomes after PCI.^5,8,9 Compared to nondiabetic patients, diabetic patients are more likely to have endothelial dysfunction, dyslipidemia, increased platelet aggregability and increased circulating procoagulants.⁸ These factors contribute to the higher risk of atherosclerotic heart disease and increased cardiovascular mortality among diabetic patients.^10,11 Similar factors also play a role in the higher early and long-term mortality rates and higher restenosis rates seen after PCI of native coronary arteries in diabetic patients.^12–14 Recent reviews suggest that an increasing number of patients undergoing SVG PCI have diabetes.¹⁵ While SVG PCI is known to be associated with worse short- and intermediate-term outcomes, the impact of diabetes on long-term outcomes after SVG PCI has never been clearly established. In recent years, the use of distal protection devices during SVG PCI has been shown to improve short-term outcomes,^16,17 but the use of these devices is still not widespread and their impact on long-term survival has not been demonstrated. Similarly, while drug-eluting stents have had a major impact on outcomes in patients undergoing native coronary artery PCI,^18,19 their use in vein grafts has not been studied extensively,²⁰ and data on their effect on long-term outcomes will not be available for a few more years. This study aimed to assess the impact of diabetes on 5-year outcomes in patients undergoing SVG PCI in the era prior to the availability of drug-eluting stents and distal protection devices. Methods Data on subjects undergoing PCI at Emory Hospital from 1981 to 2001 were collected prospectively and entered into a computerized database. All patients undergoing elective PCI of a SVG during this time period were included in the analysis. Patients with an acute myocardial infarction were excluded from the analysis. When a patient had more than one SVG intervention during the study period, we included only the first SVG intervention as the index procedure. Diabetic patients were defined as those receiving treatment with either insulin or an oral hypoglycemic agent at the time of the intervention. Diet-controlled diabetics were included if they had a documented fasting blood glucose level > 140 mg/dl, or a random blood glucose of > 200 mg/dl during the hospitalization. Variables evaluated were age, gender, hypertension, smoking history, family history of coronary artery disease (CAD), congestive heart failure (CHF), left ventricular ejection fraction (LVEF), severity of angina, and a prior history of myocardial infarction (MI). CHF severity was defined and classified according to the New York Heart Association (NYHA) criteria, and severity of angina was assessed using the Canadian Cardiovascular Society classification. PCIs were performed using standard techniques.²¹ Baseline angiographic characteristics, lesion severity and angiographic complications were entered into a computerized database. Lesions were classified based on the American College of Cardiology/American Heart Association consensus guidelines for PCI. Angiographic success, defined as residual stenosis Patient follow up. Follow-up information after PCI is routinely obtained by the review of medical records and through letters or telephone interviews with patients or their referring physicians. The information is then entered into the databank. For this study, we focused on data obtained regarding mortality, recurrent MI and the need for subsequent revascularization. Long-term follow up data were available on 2,299 of the 2,556 (90%) patients who met the initial criteria. The protocol for maintaining data in the Emory databank was approved by the Institutional Review Board at Emory University. Statistical analysis. Continuous data are expressed as mean ± standard deviation (SD), and categorical data as proportions. Continuous data were compared by unpaired t-test analysis and categorical data by the Chi-square test. A p-value of 22 were used to assess correlates of in-hospital and long-term outcomes. Multivariate correlates of in-hospital and long-term mortality were expressed as odds and hazard ratios, respectively. The discrimination of the models was assessed by the C index, and the models were validated and calibrated according to the method of Harrell.²³ Overall survival was determined and expressed by the Kaplan-Meier method, with differences between groups assessed by the log-rank method.²⁴ Results There were 2,556 SVG PCIs that met our inclusion and exclusion criteria (1,780 in the nondiabetic group and 776 in the diabetic group). The baseline clinical characteristics of the group are shown in Table 1. At baseline, there were no differences in the rates of angina or prior history of MI in the diabetic and nondiabetic groups. Diabetic patients were more likely to be female (27% vs. 17%; p p p Angiographic, early and intermediate-term outcomes. In-hospital and angiographic outcomes are shown in Table 3. The rates of angiographic success were similar in the two groups (95.1% in the diabetes group vs. 94.1% in the patients without diabetes; p-value 0.34). There was a trend towards more postprocedure in-hospital Q-wave MIs in the diabetic group, but this did not reach statistical significance (1.5% vs. 0.9%; p-value 0.15). There were more in-hospital deaths in the diabetes group (2.1% vs. 0.7%; p = 0.004). Multivariate predictors of in-hospital mortality in the study population were diabetes mellitus (odds ratio 3.6; 95% CI =1.7–7.5; p-value = 0.0007), NYHA Class II–IV CHF (odds ratio 3.8; 95% CI = 1.8–8.0; p-value = 0.0005), and older age (odds ratio per 10-year increase 1.04; 95% CI = 1.0–1.9; p-value = 0.02). Thirty-day mortality was significantly higher in the diabetic group compared to the nondiabetic group (4.4% vs. 1.9%; p-value p-value Five-year outcomes. The 5-year survival rate was also worse for the diabetic group (62.9% vs. 78.5%; p-value p-value p = 0.12). The combined endpoint of death, MI or revascularization was also significantly more common in the diabetic patients at 5 years. The survival rate free of the combined endpoint was 23% in the diabetic group and 31% in the nondiabetic group (p-value = 0.0001). Kaplan-Meier curves demonstrating a survival rate free of the combined endpoint of death, MI or revascularization during the 5-year period after the SVG PCI are shown in Figure 2. Outcomes in bare metal stent patients. The number of patients receiving bare metal stents during SVG PCI during the entire study period was higher in the diabetic group than in the nondiabetic group (48.0% [n = 372/776] vs. 37.8% [n = 673/1780] p-value p-value = 0.92). There was a nonsignificant trend towards increased in-hospital mortality in the diabetes group (1.6% vs. 0.5%; p-value 0.13), and higher 30-day mortality (4.8% vs. 2.0%; p-value 0.28). On the other hand, when intermediate and long-term outcomes were evaluated in the stented patients, 1-year survival (89.4% vs. 95.5%; p-value = 0.008), and 5-year survival (78.2% vs. 87.1%; p-value = 0.009), were significantly worse in the diabetic group. However, there were no significant differences at 5 years in the rates of repeat revascularization in both diabetic and nondiabetic patients who underwent SVG stenting (44.1% vs. 46.6%; p = 0.52). Kaplan-Meier curves demonstrating survival free of death up to 5 years in the bare metal stent population are shown in Figure 3. Impact of improved PCI techniques and adjuvant therapy during the study period. As the study period extended over two decades, there were changes in medical therapy and improvements in PCI techniques over this duration. In order to evaluate the effects of such changes on the differences in the long-term outcomes between the diabetic and nondiabetic cohorts, we also assessed data from more recent time periods separately. One- and 5-year outcomes were assessed separately for the time periods from 1991–1995 and 1996–2001. The results are shown in Table 5. For both of these time periods, 1- and 5-year survival remained significantly worse in the diabetic subjects. Five-year survival rates were significantly better, in both diabetic and nondiabetic patients, in the time period from 1996–2001, compared to the 1991–1995 time period. Discussion This is the largest study to date looking at outcomes after SVG PCI in diabetic patients. The study was an attempt to establish long-term outcomes after vein graft PCI over the two decades of interventions prior to the availability of distal protection devices and drug-eluting stents. It is important to establish long-term outcomes in this population so that historical references would be available for comparison as we begin to gather data from the use of newer devices that are now available. The results of the study show that, as expected, diabetic patients undergoing SVG PCI had an increased in-hospital and 30-day mortality and decreased 1-year survival compared to nondiabetic patients. When 5-year outcomes were assessed, both diabetic and nondiabetic patients had a poor long-term survival free of death (Figure 1). Survival free of the combined endpoint of death, MI or revascularization at 5 years after the initial SVG intervention was also remarkably poor in both groups (Figure 2). After adjusting for other significant clinical predictors, the presence of diabetes was an independent predictor of worse 5-year survival. The study also demonstrates the fact that the overall improvement in long-term outcomes that has occurred in the more contemporary era of SVG PCI have not translated into an improvement in the difference in survival between diabetic and nondiabetic patients. Comparison to previous studies. Only limited data are available on outcomes after PCI of SVGs in diabetic patients. Ahmed et al.²⁵ evaluated outcomes in patients undergoing vein graft stenting and showed significantly worse outcomes in diabetic patients up to 1 year. Our study showed similar findings, with higher rates of in-hospital adverse outcomes as well as impaired 1-year survival seen in the diabetic group. Another previous study showed that diabetes was an independent risk factor for restenosis after SVG PCI.²⁶ In the current study, this was assessed indirectly, as overall rates of repeat PCI at 1 year were evaluated and not shown to be significantly different in the diabetic and nondiabetic groups. The current study is the first to also show that worse clinical outcomes after SVG PCI in diabetic patients continued to increase up to 5 years. The 5-year survival rates in both the diabetic and nondiabetic groups of 63% and 79%, respectively, in this study were clearly worse than 5-year survival after PCI of native coronary arteries noted in previous studies. In patients undergoing native coronary artery PCI, 5-year survival rates of 89% and 93% were reported in a study by Stein et al.,¹³ and 66% and 86% in the BARI study²⁷ in the diabetic and nondiabetic groups, respectively. In addition, the current study shows that the rates of major adverse cardiovascular events (death, MI and revascularization) during the 5-year follow-up period were remarkably high in both groups. An encouraging finding from the study was the fact that long-term survival appears to be significantly better in the contemporary era of stents and adjuvant medical therapy. However, even with such improvements in survival, adverse outcomes remained significantly higher among diabetic patients. Reasons for the increased adverse outcomes in diabetic patients. There were higher rates of CHF and hypertension, as well as higher rates of angiographic and in-hospital complications seen in the patients with diabetes. However, in the current study, diabetic patients had impaired survival at 5 years, even after multivariate analysis to adjust for these and other significant clinical predictors. This may have been due to higher initial plaque burden and more progressive vein graft disease subsequently in diabetic patients. The increase in late adverse outcomes may also have been related to more native vessel disease progression in patients with diabetes.²⁸ The latter may also have resulted in the surprising finding of a trend towards lower rates of repeat revascularization in diabetic patients at 5 years. Poor distal targets due to diffuse distal disease may prevent redo CABG in many diabetic patients. This factor in turn could have contributed to the poor 5-year survival rates in these subjects. Differences in medical treatment between the groups were not compared and may have been partially responsible for the worse outcomes in the diabetic patients. However, we would expect patients with diabetes to receive at least equivalent amounts of lipid-lowering therapy. In addition, agents such as angiotensin converting enzyme inhibitors that are beneficial in the secondary prevention of atherosclerosis^29,30 have long been used for prevention of renal disease³¹ in a large proportion of diabetic patients. The potential benefits and higher use of these agents in diabetic patients could actually have reduced the gap in survival between diabetic and nondiabetic patients. Effect of bare metal stenting. The use of bare metal stents varied considerably between the two groups during the period of this study, mainly due to low rates of stent use prior to 1996. The study by Ahmed et al.²⁵ looking at outcomes after SVG PCI only included bare metal stent patients, and showed that all outcomes up to 1 year were significantly worse in the diabetic group. By contrast, in the current study, when only data from the bare metal stent patients were evaluated, periprocedural MI, in-hospital mortality and 30-day mortality rates were not significantly different in the diabetic and nondiabetic groups. On the other hand, 1-year and 5-year survival rates remained significantly worse in the diabetic patients. Most previous studies have shown better outcomes during SVG PCI with the use of stents compared to balloon angioplasty only,^32–34 although the incremental benefit is relatively small. In this study, survival appeared to be better in the stented patients compared to the overall study population. However, during the period that was evaluated in the study, there were also significant advances with PCI techniques and with medical therapy. Despite these improvements, as shown in Table 5, long-term outcomes were still worse in the diabetic compared to the nondiabetic patients in the more recent time periods of the study. Thus, despite the use of bare metal stents, improvements in PCI techniques and major advances in medical therapy, diabetic patients continued to have a poor long-term prognosis compared with nondiabetic patients in this study. An obvious question that was not addressed due to the design of the study is whether the use of drug-eluting stents will have a major impact on these poor long-term outcomes. While drug-eluting stents have not been used long enough for long-term data to be obtained, a recent study that evaluated the use of such stents suggests that diabetes continues to remain the strongest predictor of major adverse cardiovascular events at 6 months.²⁰ This suggests that drug-eluting stents may not necessarily resolve the issue of worse long-term outcomes in diabetic patients undergoing SVG PCI. Study limitations. Our study has the inherent limitations of a retrospective analysis from databases, and also used the older definitions of diabetes that were in place for the majority of the time period evaluated in the study. This definition was revised in the latter period of the study.³⁵ We also did not assess for differences in the use of glycoprotein IIb-IIIa receptor antagonists during the SVG intervention in our study population. The use of these agents, however, may not be as important in SVG PCI as they are in native coronary artery PCI procedures. A recent review of the data suggests that these agents do not confer any benefit when administered during SVG PCI.³⁶ More recent advances, such as the use of distal protection devices and drug-eluting stents, were specifically excluded by the study design and duration. These new devices were not available during the study period. While such devices have been shown to improve early outcomes after SVG PCI,^16,17,20 it remains unclear whether these recent advancements will improve the outcomes in the diabetic population undergoing SVG PCI. Conclusions The present study confirms the fact that in the era prior to the availability of drug-eluting stents, diabetic patients had significantly worse 5-year clinical outcomes compared to nondiabetic patients after vein graft PCI. The use of bare metal stents during SVG PCI appeared to narrow the differences in early outcomes, but not 1 and 5-year outcomes between the two groups. Improvements in adjuvant medical therapy and PCI techniques over the duration of the study also did not appear to narrow the differences in long-term outcomes between the diabetic and nondiabetic groups. Given the increasing number of patients with diabetes undergoing SVG PCI, an emphasis must be placed on developing and using strategies, such as distal protection and drug-eluting stenting, that will narrow differences in outcomes between the diabetic and nondiabetic population. As both diabetic and nondiabetic patients had remarkably poor long-term outcomes after vein graft PCI, patients undergoing such procedures should be targeted for very aggressive secondary prevention.

1. Fitzgibbon GM, Kafka HP, Leach AJ, et al. Coronary bypass graft fate and patient outcome: Angiographic follow-up of 5,065 grafts related to survival and re-operation in 1,388 patients during 25 years. J Am Coll Cardiol 1996;28:616–626. 2. Bourassa MG, Enjalbert M, Campeau L, et al. Progression of atherosclerosis in coronary arteries and bypass grafts: Ten years later. Am J Cardiol 1984;53:102C–107C. 3. Motwani JG, Topol EJ. Aortocoronary saphenous vein graft disease: Pathogenesis, predisposition and prevention. Circulation 1998;97:916–931. 4. Kereiakes D. Platelet Glycoprotein IIb/IIIa Inhibition and atheroembolism during bypass graft angioplasty: A cup half full (editorial). Circulation 2002;106:2994–2996. 5. Hong MK, Mehran R, Dangas G, et al. Are we making progress with percutaneous saphenous vein graft treatment? A comparison of 1990 to 1994 and 1995 to 1998 results. J Am Coll Cardiol 2001;38:150–154. 6. Smith SC, Dove JT, Jacobs AK, et al. ACC/AHA guidelines for percutaneous coronary intervention (revision of the 1993 PTCA guidelines). J Am Coll Cardiol 2001;37:2215–2238. 7. de Feyter PJ, van Suylen RJ, de Jaegere PP, et al. Balloon angioplasty for the treatment of lesions in saphenous vein bypass grafts. J Am Coll Cardiol 1993;21:1539–1549. 8. Ellis SG, Narins CR. Problem of angioplasty in diabetics. Circulation 1997;96:1707–1710. 9. Shaw RE, Anderson VE, Weintraub WS, et al. Development of a risk adjustment mortality model from the ACC-NCDR experience: 1998–2000. J Am Coll Cardiol 2002;39:1104–1112. 10. Kannel WB, McGee DL. Diabetes and cardiovascular disease. The Framingham study. JAMA 1979;241:2035–2038. 11. Stamler J, Vaccaro O, Neaton JD, et al. Diabetes, other risk factors, and 12-year cardiovascular mortality for men screened in the Multiple Risk Factor Intervention Trial. Diabetes Care 1980;3:650–654. 12. Abizaid A, Kornowski R, Mintz GS, et al. The influence of diabetes mellitus on acute and late the clinical outcomes following coronary stent placement. J Am Coll Cardiol 1998;32:584–589. 13. Stein B, Weintraub WS, Gebhart SS, et al. Influence of diabetes mellitus on early and late clinical outcomes after percutaneous transluminal coronary angioplasty. Circulation 1995;91:979–989. 14. Ghazzal Z, Ashfaq S, Douglas JS, et al. Prognostic implication of creatine kinase release following elective percutaneous coronary intervention in the pre-IIb-IIIa antagonist era. Am Heart J 2003;145:1006–1012. 15. Mathew V, Clavell AL, Lennon RJ, et al. Percutaneous coronary interventions in patients with prior coronary artery bypass surgery: Changes in patient characteristics and outcome during two decades. Am J Med 2000;108:127–235. 16. Baim DS, Wahr D, George B, et al. Randomized trial of distal embolic protection device during percutaneous intervention of saphenous vein aorto-coronary bypass grafts. Circulation 2002;105:1285–1290. 17. Stone GW, Rogers C, Hermiller J, et al. Randomized comparison of distal protection with a filter-based catheter and a balloon occlusion and aspiration system during percutaneous intervention of diseased saphenous vein aorto-coronary bypass grafts. Circulation 2003;108:548–553. 18. Stone GW, Ellis SG, Cox DA, et al. One-year clinical results with the slow-release, polymer-based, paclitaxel-eluting TAXUS stent: The TAXUS-IV trial. Circulation 2004;109:1942–1947. 19. Morice MC, Serruys PW, Sousa E, et al. Randomized study with the sirolimus-coated Bx Velocity balloon-expandable stent in the treatment of patients with de novo native coronary artery lesions. N Engl J Med 2002;346:1773–1780. 20. Ge L, Iakovou I, Sangiorgi GM, et al. Treatment of saphenous vein graft lesions with drug-eluting stents: Immediate and midterm outcome. J Am Coll Cardiol 2005;43:989–994. 21. Douglas JS, King SB III. Percutaneous coronary interventions. In: Fuster V, Alexander RW, O’Rourke RA, et al. (Eds.) Hurst’s The Heart, 10th Edition. 2001, pp. 1437–1470. 22. Cox DR. Regression models and life tables. J R Stat Soc Bzvol 1972;34:187–202. 23. Harrell F. Design: S Functions for biostatistical/epidemiological modeling, testing, estimation, validation, graphics, prediction, and typesetting. Programs available from statlib@lib.stat.cmu.edu. 24. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–481. 25. Ahmed JM, Hong MK, Mehran R, et al. Influence of diabetes mellitus on early and late clinical outcomes in saphenous vein graft stenting. J Am Coll Cardiol 2000;36:1186–1193. 26. Heidland UE, Heintzen MP, Michel CJ, et al. Risk factors for the development of restenosis following stent implantation of venous bypass grafts. Heart 2001;85:312–317. 27. The Bypass Angioplasty Revascularization Investigation (BARI) Investigators. Comparison of coronary bypass surgery with angioplasty inpatients with multivessel disease. N Engl J Med 1996;335:217–225. 28. Alderman EL, Corley SD, Fisher LD, et al. 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. 29. Yusuf S, Sleight P, Pogue J, et al. Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. N Engl J Med 2000;342:145–153. 30. Heart Outcomes Prevention Evaluation (HOPE) Study Investigators. Effects of ramipril on cardiovascular and microvascular outcomes in people with diabetes mellitus: Results of the HOPE study and MICRO-HOPE study. Lancet 2000;355:253–259. 31. Lewis EJ, Hunsicker LG, Bain RP, et al. The effect of angiotensin-converting-enzyme inhibition on diabetic nephropathy. N Engl J Med 1993;329:1456–1462. 32. Wong SC, Baim DS, Schatz RA, et al. Immediate results and late outcomes after stent implantation in saphenous vein graft lesions: The multicenter U.S. Palmaz-Schatz stent experience. J Am Coll Cardiol 1995;26:704–712. 33. Savage MP, Douglas JS, Fischman DL, et al. for the Saphenous Vein de novo (SAVED) investigators. Stent placement compared with balloon angioplasty for obstructed coronary bypass grafts. N Engl J Med 1997;337:740–747. 34. Brener SJ, Ellis GS Apperson-Hansonn C, et al. Comparison of stenting and balloon angioplasty for marrowings in aorto-coronary saphenous vein conduits in place for more than 5 years. Am J Cardiol 1997;79:13–18. 35. Gabir MM, Hanson RL, Dabelea D, et al. The 1997 American Diabetes Association and 1999 World Health Organization criteria for hyperglycemia in the diagnosis and prediction of diabetes. Diabetes Care 2000;23:1108–1112. 36. Roffi M, Mukherjee D, Chew DP, et al. Lack of benefit from intravenous platelet glycoprotein IIb/IIIa receptor inhibition as adjunctive treatment for percutaneous interventions of aortocoronary bypass grafts: A pooled analysis of five randomized clinical trials. Circulation 2002;106:3063–3067.