Patients with type 2 diabetes mellitus represent 25% of those requiring myocardial revascularisation.1 Although current guidelines favour coronary bypass surgery (CABG) over percutaneous coronary intervention (PCI) in most diabetics who require revascularisation, 2–3 substantial variability exists in practice patterns between individual hospitals, suggesting a lack of clinical consensus.4 The major advantage of CABG over bare-metal stent (BMS) implantation is the lower risk of repeat revascularisation procedures through the follow-up.5–7 This advantage may be eliminated by drug-eluting stents (DES).8–12 It is unknown, however, whether this translates into better long-term clinical outcome. Indeed, concerns have been raised from a recent meta-analysis of 1,748 patients enrolled in four randomised trials evaluating the safety of sirolimus-eluting stents (SES) compared with BMS.13 In the 428 patients with diabetes, a significant difference in the four-year survival rate was observed in favour of the BMS over the SES group (95.6% versus 87.8%; hazard ratio (HR) for death in the SES group, 2.9; 95% confidence interval (CI), 1.38–6.10; p=0.008). The lower survival rate among patients with diabetes who were treated with SES was due to increased numbers of deaths from both cardiovascular and non-cardiovascular causes. No difference in survival rate was detected among the patients without diabetes. Among the patients with diabetes, there was a small excess of very late stent thrombosis (occurring more than one year after the procedure). Although the number of fatal events in patients with diabetes was small, and a lower than expected mortality was noted among the patients with diabetes in the BMS group, these data warrant additional studies.
It is common experience in clinical practice that patients with type 2 diabetes mellitus encompass a heterogeneous population with different severity of cardiovascular disease and therefore different outcome. Insulin treatment has largely been used as an index of higher risk. However, this approach seems to be quite simplistic. It is time for more accurate criteria for risk-stratification in diabetic patients. In this context, microvascular complications may represent a useful index to differentiate high- versus low-risk patients.
Microvascular complications have been identified as risk markers for cardiovascular death in diabetic subjects.14–16 Several mechanisms (such as chronically elevated plasma glucose levels, vascular endothelium abnormalities and insulin resistance) have been proposed as explanations for the development and progression of microvascular complications.17–19 The causes of generalised endothelial dysfunction in type 2 diabetes mellitus are unknown, although there are many possible candidates, such as increased protein kinase C activity, non-enzymatic glycation and oxidative stress, alterations in redox potential and increased expression of transforming growth factor (and/or tumour necrosis factor).17–19
Furthermore, variability among individuals in terms of endothelial vulnerability to injury (potentially due to genetic factors) might account for the observation that only a subset of patients appears susceptible to the development of microangiopathy. The coincidence of albuminuria, retinopathy and cardiovascular disease has been explained by the Steno hypothesis,19 according to which albuminuria, retinopathy and cardiovascular disease reflect generalised vascular leakage caused by genetically determined alterations in the basement membrane metabolism associated with hyperglycaemia. Diabetic retinopathy (DR) is a frequent and early sign of microvascular complication of diabetes mellitus, and remains a leading cause of blindness in Japan, Europe and North America.20 The balance of evidence suggests that there is a graded relation between severity of DR and risk of all-cause mortality in both type 1 and type 2 diabetes. Whether this is simply because DR is a reflection of poor diabetic status or whether its presence constitutes a separate and additive adverse influence on mortality and the incidence of heart disease remains uncertain. Moreover, after adjustment for potential confounders, proliferative DR and progression to proliferative DR remain associated with increased all-cause mortality and incidence of coronary heart disease.21–22 An increased spontaneous and induced platelet aggregation, with enhanced thromboxane A2 production and expression of more glycoprotien IIb/IIIa receptors, has been demonstrated in diabetic patients. Moreover, greater coagulation activity co-exists with increased procoagulant factors (fibrinogen, factor VII and von Willebrand factor) and fibrinolytic system impairment, with reduced synthesis of prostacyclin and increased plasminogen activator inhibitor type 1 (PAI-1).23 Whether these abnormalities are more pronounced in diabetics with microvascular complications has not been investigated.
Diabetic nephropathy24–25 and retinopathy25–29 have been identified as independent predictors of major adverse cardiac events (MACE) after both PCI and bypass surgery in diabetic patients. Briguori et al.25 reported that MACE at 12 months occurred in 21% of patients in the group without microvascular complications, in 40% in the group with nephropathy, in 51% in the group with retinopathy and in 73.5% in the group with both microvascular complications (p<0.001; see Figure 1). The impact of microvascular complications on long-term outcome following DES implantation has been assessed in 371 type 2 diabetic patients who had elective DES implantation on native vessels at our institutions from April 2002 to June 2005.30 The complexity of coronary artery disease was quantified using the SYNergy between PCI with TAXUS and Cardiac Surgery (SYNTAX) score.31
At 14±3 months, MACE occurred in 134 out of 371 patients (36%; MACE group), whereas 237 patients were event-free. Patients in the MACE group were older, with a higher rate of microvascular complications (68% versus 44%; p<0.001), with higher SYNTAX score (21±6 versus 15±8; p<0.001) and with higher rate of small (<2.75mm) vessel diameter. Independent predictors of MACE were microvascular complications (HR, 1.88; 95% CI, 1.20–2.95; p=0.006), SYNTAX score >19 (HR, 1.72; 95% CI, 1.04–2.45; p=0.027) and small vessel reference diameter (HR, 1.61; 95% CI, 1.08–2.85; p=0.035). The independent predictors of the composite of death and myocardial infarction were age ≥65 years (HR, 3.06; 95% CI, 1.28–7.32; p=0.012), presence of microvascular complications (HR, 1.95; 95% CI, 1.45–13.08; p=0.008), and previous myocardial infarction (HR, 3.03; 95% CI, 1.16–8.33; p=0.024).
According to these findings, the presence of microvascular complications may help physicians to identify high-risk diabetic patients undergoing DES implantation. Although the goals of DES are to lower restenosis at the stented site, the therapy is local and obviously will not prevent non-culprit lesion progression. Screening diabetic patients for both nephropathy and retinopathy prior to myocardial revascularisation appears to be an effective way to risk-stratify this group of patients and thus to focus preventative measures. Data from the FREEDOM trial (multivessel sirolimus stenting versus surgery in diabetics), the CARDia trial (enrolling patients with diabetes mellitus and multivessel disease D to either SES plus abciximab or CABG surgery) and the BARI 2D trial (comparing tight diabetic control with insulin-providing versus insulin-sensitising therapy with and without a revascularization procedure of choice) will contribute to our understanding of the appropriate revascularisation strategy for diabetic patients.