Diabetes and the associated metabolic syndrome have reached epidemic proportions in many areas of the world.1,2 Atherosclerotic cardiovascular disease has been correlated with the presence, duration and severity of diabetes.3,4 Patients with diabetes have worse clinical outcomes following presentation for acute coronary syndrome (ACS)5,6 as well as following both surgical (coronary artery bypass graft [CABG])7–11 and percutaneous revascularisation (percutaneous coronary intervention [PCI])12–16 compared with their non-diabetic counterparts. Although coronary stenting has improved outcomes (compared with balloon angioplasty) following PCI, both angiographic restenosis and the requirement for repeat revascularisation are increased in diabetics compared with non-diabetics, and also limit the durability of PCI compared with CABG.17–20 Stent-based elution of paclitaxel or sirolimus from biostable polymers has been demonstrated to reduce coronary late lumen loss as well as angiographic and clinical restenosis compared with bare-metal stent (BMS) deployment in patients both with and without diabetes.21–24 Although the advent of drug-eluting stents (DES) promises a paradigm shift away from surgical revascularisation in diabetic patients (especially those with multivessel disease) and towards PCI,25 the presence of diabetes remains a significant predictor of adverse clinical and angiographic outcomes following PCI.26,27 Continued evolution of both DES and adjunctive pharmacotherapies is required to achieve optimal clinical outcomes following PCI with DES in patients with diabetes.
The Diabetic ‘Problem’
The propensity for diabetic patients to experience adverse outcomes following revascularisation has been ascribed to smaller-calibre target vessels, a greater degree of underlying vascular inflammation, a prothrombotic milieu and more frequent associated cardiovascular risk factors.28,29 Immunity, inflammation and heredity appear central to the pathogenesis of insulin resistance and hyperglycemia,30,31 which in turn contribute to the development of atherothrombotic disease by stimulating pro-coagulant factors (fibrinogen, von Willebrand factor, factor VII), endothelial dysfunction, oxidative stress and inflammatory cytokines (interleukin-6, C-reactive protein) while reducing levels of antithrombotic factors (antithrombin III, tissue-plasminogen activator).28,32 The prothrombin milieu, endothelial dysfunction and excessive inflammatory response to vessel injury, which is in part mediated by advanced glycation end-products, likely contribute to the exaggerated smooth-muscle cell and neointimal proliferation following PCI observed in diabetic patients.28,33,34
Furthermore, the decreased numbers of circulating endothelial progenitor cells present in diabetics may contribute to delayed endothelial regeneration and healing.35 Finally, maladaptive arterial remodelling contributes to restenosis following balloon angioplasty in diabetic patients.36 Although coronary stent (BMS) deployment improved both early and late clinical and angiographic outcomes in diabetic cohorts compared with balloon angioplasty, late restenosis and repeat revascularisation remained significantly more common in diabetics than in non-diabetics. Randomised comparative studies of PCI (balloon angioplasty or BMS) and CABG demonstrated an increased morbidity and mortality among diabetics compared with their non-diabetic counterparts and improved late survival trends following multivessel surgical revascularisation compared with PCI.7–11
The potential mechanisms for diminished benefit following PCI in patients with diabetes include both the increased risk of target lesion/vessel restenosis and more diffuse atherosclerosis with resultant greater disease progression in non-target vessels. Even following the introduction of DES for multivessel PCI, the available comparative studies of multivessel PCI versus CABG have suggested higher rates of major adverse clinical events (MACE) in late follow-up of PCI-treated patients, driven in large part by the increased requirement for repeat revascularisation.37,38 The appreciable rate of disease progression observed in DES-treated patients underscores the importance of concomitant optimisation of cardiovascular risk factor treatment and glycaemic control. Nevertheless, the ‘gap’ between revascularisation rates with CABG versus BMS implantation has been narrowed by the introduction of DES, and rates of death and/or myocardial infarction appear to be similar at mid-term follow-up (one to two years) for both DES- and CABG-treated patients.
More definitive comparisons of DES versus CABG for multivessel revascularisation in both diabetics and non-diabetics will be provided by ongoing randomised controlled clinical trials such as Synergy between PCI with Taxus and Cardiac (SYNTAX),39 Coronary Artery Revascularization in Diabetes (CARDIA),40 Comparison of Two Treatments for Multivessel Coronary Artery Disease in Individuals with Diabetes (FREEDOM)41 and Coronary Artery Revascularization in Diabetes (VA CARDS).42
The Benefits of Drug-eluting Stents
Diabetes remains a significant predictor of adverse outcomes, including death, myocardial infarction and angiographic and clinical restenosis, as well as stent thrombosis following the deployment of DES.43–47 Nevertheless, multiple randomised controlled trials, meta-analyses and registries have demonstrated the superiority of DES versus BMS in patients with and without diabetes who undergo PCI21–25,48–51 (see Figure 1). Stent-based elution of either paclitaxel or sirolimus from biostable polymers has been demonstrated to reduce angiographic coronary late lumen loss as well as the occurrence of both angiographic and clinical (target lesion or vessel) restenosis compared with BMS. Despite demonstrated differences in coronary late lumen loss by quantitative coronary angiography between the available sirolimus- and paclitaxel-eluting stent platforms,52 pooled analyses of both randomised trials and registries suggest similar rates of clinical events – including death, myocardial infarction, target vessel revascularisation, stent thrombosis and cumulative MACE – following deployment of these ‘first-generation’ DES devices53,54 (see Figure 2). Preliminary data suggest that clinical outcomes may be improved following deployment of a second-generation thin-strut everolimus-eluting DES compared with the first-generation paclitaxel-eluting device.55
Despite the use of DES, survival free from MACE remains diminished following multivessel (versus single-vessel) PCI in both diabetic and non-diabetic patients.56,57 Recent data suggest that the occurrence of death and non-fatal myocardial infarction in late follow-up of diabetic patients may be reduced by DES (versus BMS) in proportion to the duration of dual antiplatelet (aspirin plus thienopyridine) therapy.58 The maximum relative benefit of DES was observed in those patients who received clopidogrel therapy for longer than nine months following stent deployment. A recent randomised controlled trial that compared stent-based paclitaxel elution from a biostable polymer versus paclitaxel elution from a reservoir-based bioresorbable polymer demonstrated improved angiographic as well as clinical outcomes in those patients randomly assigned to receive the biostable polymer platform.59,60 The superiority of the biostable polymer platform in reducing adverse clinical and angiographic events in both diabetic and non-diabetic patients may be explained by the higher dose of paclitaxel loaded onto the stent (100 versus 10μg) and/or differences in polymer drug release kinetics, which resulted in more effective suppression of restenosis compared with the bioresorbable polymer device.60
The Importance of Diabetic Control
Both the presence and severity of diabetes, as reflected by the requirement for insulin treatment, adversely affect early (in-hospital) and late survival as well as the requirement for repeat revascularisation.61,62 Multiple studies from the era of BMS deployment demonstrate an increase in coronary late lumen loss in-stent, increased neointimal hyperplasia and an increase in MACE among diabetic patients (versus non-diabetics), especially those who require insulin treatment.34,63 Previous studies have demonstrated a relationship between fasting and/or hospital admission blood glucose and subsequent mortality following either elective PCI or primary PCI for ST-segment elevation myocardial infarction.61,64,65 Similarly, the level of glyated haemoglobin (HbA1c) as a measure of pre-procedural glycaemic control has been correlated with target vessel revascularisation following BMS deployment, as well as with the occurrence of major adverse cardiovascular and cerebrovascular events following multivessel DES or surgical (CABG) revascularisation.38,62 In these reports, an HbA1c level >7.0% was associated with adverse outcomes. In the largest available evaluation of baseline (pre-procedural) HbA1c and subsequent clinical and angiographic outcomes following DES deployment, no specific ‘threshold’ level (i.e. >7.0 versus <7.0%) of HbA1c was defined.60 However, when analysed as a continuous variable across a total of 1,675 patients (469 of whom were medically treated diabetics), a statistically significant correlation between HbA1c and cumulative MACE or clinically driven target vessel revascularisation was observed (see Figure 3). Of note, 77 patients in this study had an elevated HbA1c >6.5% in the absence of previously diagnosed diabetes. The relationship between HbA1c and clinical outcomes was less evident for those patients with diagnosed diabetes when analysed separately.
Recent data also suggest a correlation between the specific type of oral hypoglycemic therapy employed and the presence and/or extent of vascular disease, as well as the frequency of vascular events in diabetic patients.66,67 Specifically, insulin-sensitising medications (versus insulin-providing) have been associated with improved outcomes in diabetic patients following ACS and PCI. Indeed, recent data from randomised controlled trials suggest that thiazolidinedione (TZD) therapy reduces the risk of target vessel revascularisation and angiographic late lumen loss following BMS deployment in diabetic patients.68,69
Suppression of inflammatory cytokine production (sCD40L, C-reactive protein, E-selectin, von Willenbrand factor) has been demonstrated following TZD administration.70–72 Furthermore, the TZD rosiglitazone has demonstrated antiplatelet effects that may contribute to the clinical benefit observed following treatment with this agent, and which appear to be independent of the known effects of rosiglitazone on insulin resistance.73
Peri-procedural Adjunctive Pharmacotherapies
The use of adjunctive peri-procedural glycoprotein (GP) IIb/IIIa inhibitor therapy may provide additional clinical benefit to diabetic patients undergoing PCI with either DES or BMS. Platelet GPIIb/IIIa inhibition reduces the incidence of peri-procedural myocardial infarction, stent thrombosis and urgent repeat revascularisation. Randomised controlled clinical trials and meta-analyses have demonstrated a survival advantage following PCI with adjunctive GPIIb/IIIa inhibitor therapy that is particularly evident in higher-risk patient subsets (patients with elevated cardiac biomarkers, diabetics, etc.).74–77 In a pooled analysis of multiple trials, the benefit of adjunctive peri-procedural GPIIb/IIIa inhibition for reduction of MACE was particularly evident for insulin-dependent diabetics following sirolimus-eluting stent deployment.78 In addition, a randomised placebo-controlled trial demonstrated that peri-procedural abciximab administration was associated with a reduction in angiographic and clinical restenosis following BMS deployment in diabetic patients.79 Adjunctive peri-procedural GPIIb/IIIa inhibition may also provide benefit to patients treated with bivalirudin as the procedural anticoagulant. In a large single-centre registry of patients treated with DES, all-cause mortality over two years of follow-up was significantly lower in those individuals who received peri-procedural heparin plus GPIIb/IIIa inhibitor therapy compared with those treated with bivalirudin.80
The potential inadequacy of bivalirudin monotherapy in patient cohorts with high levels of baseline platelet activation (acute myocardial infarction, diabetes, etc.) was observed in the Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial infarction (HORIZONS-AMI) trial, which compared bivalirudin with unfractionated heparin plus GPIIb/IIIa blockade in patients undergoing primary PCI for ST-segment elevation myocardial infarction.81 Acute (<24 hours), definite or probable stent thrombosis was increased in bivalirudin (1.3%) compared with heparin plus GPIIb/IIIa (0.3%; p<0.001) treated patients and was associated with a relatively higher incidence of MACE during the first 15 days following trial enrolment.
Similarly, despite the potential for bivalirudin to inhibit thrombin-mediated platelet activation and aggregation, those patients at highest risk following presentation for ACS (creatine kinase Mb [CKMB]/troponin elevation, Thrombolysis In Myocardial Infarction [TIMI] risk score >3, no thienopyridine pre-treatment prior to PCI) appear to derive the greatest relative benefit from adjunctive platelet GPIIb/IIIa inhibition.82 The administration of cilostazol, which suppresses intimal hyperplasia in both animal and human studies, as well as platelet P-selectin expression and leukocyte MAC-1 upregulation following coronary stent deployment, has been demonstrated to reduce both in-stent coronary late lumen loss by quantitative coronary angiography and binary (>50%) angiographic restenosis in diabetic patients following BMS or DES deployment.83,84 In a randomised controlled clinical trial, cilostazol added to therapy with aspirin plus thienopyridine (triple-drug therapy) resulted in a reduction in both in-stent and in-segment late lumen loss, as well as binary angiographic restenosis following DES deployment, in diabetic patients compared with aspirin plus thienopyridine (dual-drug therapy) alone.84
Both the duration and intensity of thienopyridine-mediated platelet P2Y12 receptor inhibition may be particularly important in diabetic patients, who often manifest abnormalities in platelet size and function. A differential response to standard doses of both aspirin and clopidogrel has been described in diabetics and may at least in part be overcome by increasing aspirin/clopidogrel dose.85–87 Similarly, an extended duration of clopidogrel therapy following DES deployment may be particularly beneficial to diabetic patients.58
Finally, concomitant therapy with a 3-hydroxy-3-methylglutaryl coenzyme- A reductase inhibitor (statin) and an angiotensin-converting enzyme inhibitor or angiotensin receptor blocker should also be mandated in diabetic patients if clinically tolerated. Indeed, the prescription of multiple American College of Cardiology/American Heart Association clinical practice guideline-adherent88 medications appears to provide incremental survival advantage to patients following presentation for ACS.89,90
Diabetes presents an increasingly prevalent clinical challenge that significantly influences the presence and severity of atherothrombotic cardiovascular disease as well as clinical outcomes following coronary revascularisation procedures. Although DES have significantly reduced the requirement for repeat revascularisation following PCI in diabetic patients, the optimal strategy for multivessel revascularisation in this complex patient cohort remains to be defined by ongoing randomised, controlled clinical trials. Adjunctive peri-procedural and post-PCI pharmacotherapy may provide particular clinical benefit to patients with diabetes, and should at least include clinical practice guideline-adherent medications. Serious consideration should be given to the prescription of medications (cilostazol, thiazolidinediones) with demonstrated specific benefit in diabetics following PCI. Finally, stringent glycaemic control and aggressive management of associated cardiovascular risk factors is required to optimise clinical outcomes following either percutaneous or surgical coronary revascularisation in diabetic patients.