1-s2.0-S1521690X08001711-main

Best Practice & Research Clinical Endocrinology & Metabolism 23 (2009) 317–334

Contents lists available at ScienceDirect

Best Practice & Research ClinicalEndocrinology & Metabolism

journal homepage: www.elsevier .com/locate/beem

3

Surgical versus percutaneous revascularization of coronaryartery disease in diabetic patients

Stephane Cook, MD, Consultant, Stephan Windecker, MD *

Invasive Cardiology, Department of Cardiology, University Hospital Bern, 3010 Bern, Switzerland

Keywords:coronary artery diseasediabetes mellituscoronary artery bypass grafting (CABG)percutaneous coronary intervention (PCI)

* Corresponding author. Tel.: þ41 31 632 44 97;E-mail address: [email protected] (S

1521-690X/$ – see front matter � 2008 Elsevier Ldoi:10.1016/j.beem.2008.12.002

Morbidity and mortality related to coronary artery disease (CAD)remain a great challenge in patients with diabetes mellitus.Revascularization of CAD is an important therapeutic interventionowing to its impact on both symptoms and prognosis. The optimalrevascularization strategy continues to evolve due to the advent ofnew technologies and improved peri-procedural outcome withboth percutaneous coronary interventions and coronary arterybypass grafting. Although clinical outcome following coronaryartery bypass is worse in diabetic as opposed to non-diabeticpatients, surgical revascularization tends to be associated withbetter outcome in stable patients with multivessel disease andreduced left ventricular function. The advent of drug-eluting stentshas challenged the supremacy of coronary artery bypass graftingand has become a valuable alternative to surgery. The safety andefficacy of drug-eluting stents in the treatment of patients withdiabetes and multivessel disease is currently under investigation inseveral ongoing randomized controlled trials. Percutaneous coro-nary intervention is the therapy of choice in patients with acutecoronary syndromes, particularly ST-elevation myocardial infarc-tion. The focus of this review is to present the current evidence,define the role of percutaneous and surgical revascularization inthe treatment of diabetic patients with CAD, and propose a tailoredapproach for clinical decision-making.

� 2008 Elsevier Ltd. All rights reserved.

Fax: þ41 31 632 47 70.. Windecker).

td. All rights reserved.

mailto:[email protected]

www.sciencedirect.com/science/journal/1521690X

S. Cook, S. Windecker / Best Practice & Research Clinical Endocrinology & Metabolism 23 (2009) 317–334318

Diabetes and heart disease

Prevalence of diabetes mellitus in the general population

The prevention of coronary artery disease (CAD) and reduction of associated morbidity and mortalityconstitute great public health challenges worldwide. The prevalence of diabetes mellitus has increasedby 42% in industrialized countries (1995: 51 million, 2005: 72 million)1–5, whereas it has almost tripledin developing countries (1995: 84 million, 2005: 228 million)6 during the last decade. Based onprojections of studies performed between 1991 and 2001, Mokdad et al7 estimate that 22 millionAmericans (11% of the population) will suffer from diabetes mellitus in the year 2025, and one third ofAmericans born in the year 2000 will be at risk of developing diabetes mellitus during their lifetime.

Coronary artery disease in the diabetic population

Decreased nitric oxide bioavailability, the hallmark of insulin resistance, is multifactorial (impairedNO synthesis, trapping of NO by reactive oxygen species) and has been associated with the dissemi-nated nature of coronary artery disease, procoagulant state, reduced protection afforded by ischaemicpreconditioning, and impairment of myocardial function observed in diabetic patients. These factorsnot only increase the predisposition to CAD but are also responsible for the more severe sequelae ofthrombotic events.8

A consistent observation in diabetic compared with non-diabetic patients is a four- to six-foldincreased risk of cardiovascular events. A diagnosis of diabetes mellitus is present in 15–25% of patientsreferred for coronary revascularization, in 20–30% of patients admitted with acute coronary syndromes(ACS), and in >30% of patients presenting with cardiogenic shock. Conversely, 75% of hospitalizationsand 80% of all deaths in diabetic patients are attributed to a cardiovascular cause. An observationalstudy of 1059 diabetic and 1373 non-diabetic individuals during a 7-year period in Finland showeda similar risk of subsequent myocardial infarction among diabetic patients without previousmyocardial infarction as in non-diabetic patients with a history of prior myocardial infarction (cardiacmortality in diabetes without myocardial infarction: 2.5% per year, versus no diabetes with myocardialinfarction: 2.6% per year, P¼ 0.5).9 Therefore, diabetes mellitus is considered as a risk equivalent toCAD. Moreover, diabetic women loose their gender protection against CAD and have the samecardiovascular risk as men.

Patients with insulin-dependent diabetes mellitus (IDDM) compared with those with non-insulin-dependent diabetes (NIDDM) are younger, have carried a diagnosis of diabetes for longer, and haveworse renal function. While some nuances in outcome are observed between patients with andwithout IDDM, the pathophysiological mechanisms are similar, and hyperglycaemia per se is delete-rious.10 Along this line, patients with asymptomatic hyperglycaemia are also at increased risk forcardiovascular events. In the DECODE study of 15,388 men and 7126 women11, the risk of cardiovas-cular mortality was significantly increased in subjects with asymptomatic diabetes defined as eithera fasting plasma glucose >7.0 mmol/L or a 2-hour post-load plasma glucose �11.1 mmol/L.

Practice points

� diabetes magnifies the risk of cardiovascular morbidity and mortality� there is a three-fold increased risk of CAD disease in diabetic compared with non-diabetic

patients� there is a four-fold increase in mortality for patients with type-I diabetes compared with non-

diabetic patients� women with diabetes lose their gender protection against CAD� 15–25% of patients undergoing percutaneous coronary intervention suffer from diabetes� 20–30% of patients presenting with acute coronary syndromes have diabetes� diabetes is considered a risk equivalent to a history of myocardial infarction

S. Cook, S. Windecker / Best Practice & Research Clinical Endocrinology & Metabolism 23 (2009) 317–334 319

Revascularization strategies in diabetic patients

The revascularization strategy in patients with diabetes mellitus is influenced by several factors,including the clinical setting (acute ST-elevation myocardial infarction and cardiogenic shock, non-ST-segment-elevation myocardial infarction or stable angina pectoris), coronary anatomy (extent ofcoronary disease, suitability for coronary artery bypass grafting (CABG) anastomoses, previous CABG),and left ventricular function. Apart from these disease-specific factors, other clinical characteristicsinfluence the choice of revascularization procedures, such as concomitant valvular disease, chestdeformities, prior radiation exposure, peripheral artery disease, and extracardiac conditions such aschronic obstructive pulmonary disease, coagulation disorders, or malignancies.

Percutaneous revascularization of diabetic patients

Diabetic patients undergoing percutaneous coronary revascularization have a higher rate ofprevious myocardial infarction, congestive heart failure, and arterial hypertension than non-diabeticpatients.12 CAD is also more extensive in diabetic than non-diabetic patients. For example, angio-graphic data of 281 diabetic and 1833 non-diabetic patients included into the NHLBI registry between1985 and 1986 showed a higher prevalence of three-vessel disease (28% in diabetics versus 18% in non-diabetics, P< 0.01) and conversely a lower rate of one- and two-vessel disease in diabetic patients.13

Clinical outcome of diabetic patients undergoing percutaneous coronary intervention (PCI) isinferior to that of non-diabetic patients. Several aspects of diabetic CAD appear responsible for thisobservation. Vessel size has been reported to be smaller in diabetic compared with non-diabeticpatients with stable CAD as well as in patients with myocardial infarction. Diabetes has also consis-tently been shown to be associated with higher rates of restenosis in the balloon-angioplasty, bare-metal-stent, and drug-eluting-stent era. Angiographic and ultrasonic studies suggest a higher degree oflate luminal loss and neointimal hyperplasia in diabetic compared to non-diabetic patients asa potential mechanism underlying this phenomenon. An increased rate of restenosis may have animpact on prognosis in diabetic patients. The occlusive form of restenosis has been associated withpoor survival and was an independent predictor of mortality in a study of 603 diabetic patientsundergoing balloon angioplasty between 1987 and 1995. Finally, diabetes constitutes an independentpredictor of early stent thrombosis, both in the bare-metal- and drug-eluting-stent era.14 Morerecently, the early outcome of diabetic patients has improved with the advent of adjunctive phar-macological therapy, including glycoprotein IIb/IIIa antagonist, pre-loading with high-dose clopidogrel,and long-term thienopyridine and statin therapy. Notwithstanding, the long-term risk of death,myocardial infarction, and repeat revascularization remains nearly twice as high in diabetic patientsthan in non-diabetic patients undergoing PCI.15

Bare-metal stents versus drug-eluting stents in diabetic patientsDrug-eluting stents (DES) have successfully addressed the problem of restenosis inherent to bare-

metal stents (BMS), and significantly decreased the need for target-lesion revascularization.16 Tworandomized studies compared clinical outcomes between BMS and DES in diabetic patients (DIABETES,SCORPIUS). Rates of repeat revascularization were markedly reduced in favour of DES in bothstudies.17,18 In 3751 pairs of patients treated with either DES or BMS in Ontario, rates of target vesselrevascularization were also significantly reduced in favour of DES in nearly all subsets of diabeticpatients.19 A collaborative network meta-analysis compared the risk of revascularization between thetwo first-generation DES – sirolimus-eluting stent (SES) and paclitaxel-eluting stent (PES) – and BMS indiabetic patients.15 The risk of target-lesion revascularization was reduced by 71% and 62%, respec-tively, in favour of PES and SES compared with BMS in diabetic patients, similar to observations withnon-diabetic patients. Due to the higher baseline risk of restenosis, the absolute reduction in repeatrevascularization is more pronounced in diabetic than non-diabetic patients, and the use of DES shouldbe strongly considered in this patient subgroup.

Although a recent meta-analysis of four trials suggested an increased risk of death with SEScompared to BMS in diabetic patients20, this finding was not confirmed in a larger analysis of 14 trials.21

The safety and efficacy of the two first-generation DES and BMS in patients with and without diabetes


was also addressed in the above-mentioned network meta-analysis.15,22 There were no significantdifferences in the risk of overall mortality, cardiac mortality, and myocardial infarction among SES, PES,and BMS in diabetic as well as non-diabetic patients. Of note, in trials with less than 6 months’ durationof clopidogrel therapy, the risk of death associated with SES was more than twice the risk associatedwith BMS, whereas trials with clopidogrel of 6 months or longer showed no increase in risk with SEScompared with BMS. This observation suggests that the above-mentioned increase in the risk of deathassociated with SES compared with BMS in diabetic patients was most likely due to the restrictedduration of clopidogrel therapy below 6 months in early trials.20 In summary, DES demonstratea similar safety profile, with comparable rates of death, cardiac death, and myocardial infarction, asBMS in diabetic patients.

Surgical revascularization of diabetic patients

The clinical outcome following CABG is more complicated in diabetic than in non-diabetic patients.Carson and colleagues assessed short-term morbidity and mortality in a retrospective cohort study of146,786 patients undergoing CABG in 434 hospital in the United States in 1997.23 The investigatorsobserved an increased risk of mortality (3.7% versus 2.7%, adjusted OR¼ 1.23, 95% CI 1.15–1.32),morbidity (13.9% versus 9.1%, adjusted OR¼ 1.38, 95% CI 1.33–1.44), and infections (7.9% versus 5.2%,adjusted OR¼ 1.36, 95% CI 1.30–1.43) in diabetic compared with non-diabetic patients.

The long-term outcome following CABG (mean 7 years) of 2544 diabetic and 6581 non-dia-betic patients was investigated by Mohammadi and colleagues.24 Cardiac survival at 5, 7, and 10years amounted to 96.4%, 94.6%, and 90.4% among non-diabetic patients, 95.9%, 93.0%, and 87.0%among patients with NIDDM, and to 92.8%, 86.8%, and 75.7% among IDDM patients (P< 0.001),respectively. Similar results have been reported in a study of 36,641 consecutive patients (31%diabetic patients) followed for a mean duration of 4 years.25 In this cohort study, the annualincidence rate of death was 3.1 deaths per 100 person-years in non-diabetic and 4.4 deaths per100 person-years in diabetic patients, respectively. The difference was markedly increased indiabetic patients with concomitant renal insufficiency or peripheral vascular disease (9.4 deathsper 100 person-years).

Patients undergoing surgical revascularization preferably undergo grafting with the use of the leftinternal mammary artery (LIMA). The latter affords protection from atherosclerosis and is associatedwith low graft failure during long-term follow-up. Tatoulis et al26 reported a patency rate of 98% at 5years, 95% at 10 years, and 88% at 15 years. Furthermore, the survival advantage afforded by CABG indiabetic patients as observed in the BARI (Bypass Angioplasty Revascularization Investigation) trial waslimited to patients who had received a LIMA graft. Arterial revascularization with use of bilateralinternal mammary arteries (BITA) has been shown to decrease the risk of death (hazard ratio¼ 0.72,95% CI 0.57–0.91) and need for reoperation (HR¼ 0.38, 95% CI 0.19–0.77) in both diabetic (n¼ 633) andnon-diabetic (n¼ 3673) patients.27

There have been other improvements in surgical technique, including minimal invasive proceduresand ‘off-pump’ surgery, in attempts to improve on surgical results. In the case of conventional bypasssurgery with cardiopulmonary bypass, the heart and ascending aorta are cannulated, cardiac arrest isinduced, and the ascending aorta is manipulated in order to construct proximal anastomoses. Inaddition to the risk of atherosclerotic emboli, cardiopulmonary bypass induces an inflammatoryresponse with oxidative stress, and activation of the complement system, which may potentially leadto multiple organ dysfunction and/or damage. Given these findings, it is conceivable that miniaturi-zation of the cardiopulmonary bypass pump or its avoidance will substantially decrease the risk ofaortic emboli and end-organ dysfunction, particularly in diabetic patients with more extensiveatherosclerosis. Finally, minimizing the surgical incision as well as trauma to skin, soft tissue, and boneby using mini-thoracotomy may reduce the risk of perioperative infection in diabetic patients.28

Balloon angioplasty and BMS versus CABG in diabetic patients with stable angina pectoris

To date, only one prospective randomized controlled trial (RCT) comparing CABG with PCI specif-ically in diabetic patients has been completed. However, outcomes for subgroups of diabetic patients


enrolled into RCTs comparing the two revascularization strategies have been summarized in reviews29

and a meta-analysis30 (Table 1).The BARI trial randomized 1829 patients with multivessel disease to CABG (LIMA use: 82%) or PCI

(stents: none). The study showed similar mortality in the overall population (PCI: 13.7% versus CABG:10.7%, P¼ 0.19), but decreased survival among diabetic patients (n¼ 353) treated with PCI (PCI: 35.5%versus CABG: 19.4%, P¼ 0.003)31 at 5 years of follow-up. The survival benefit in favour of CABG waslimited to patients revascularized by means of a LIMA graft (5-year mortality of diabetic patientstreated with LIMA: 2.9%), whereas mortality was similar among patients treated with a saphenous veingraft (18.2%) and balloon angioplasty.32 The original observation of the absence of a difference insurvival in the overall population and improved survival in diabetic patients allocated to CABG wasmaintained at 7 years (PCI: 44.3% versus CABG: 25.6%, P¼ 0.001)33 and 10 years (PCI: 54.5% versusCABG: 42.2%, P¼ 0.03)34, respectively. Although widely publicized, it is worth mentioning that theanalysis of diabetic patients was not pre-specified in the protocol, but rather it was a post-hocinvestigation, and the study was underpowered for detecting differences in mortality amongsubgroups. Moreover, the findings were not confirmed in the BARI registry of 2010 patients, who wereeligible for the randomized study but did not provide informed consent and were treated according tophysician preference; of these patients, 65% underwent PCI. At 7 years no difference in mortality wasobserved in both the overall population (PCI: 13.9% versus CABG: 14.2%, P¼ 0.66) as well as diabeticpatients (PCI: 26% versus CABG: 26%, P¼ 0.96). Rates of repeat revascularization were significantlyhigher among diabetic patients treated with PCI (69.9%) than with CABG (11.1%) at 7 years.

The RITA (Randomized Intervention Treatment of Angina) I trial35 randomly assigned 1011 patientswith multivessel disease to PCI (stents: none) or CABG (LIMA graft: 74%) and followed patients fora median of 6.5 years. The study showed no difference in death or myocardial infarction in the overallpopulation. Somewhat opposite to the BARI findings, mortality tended to be lower in diabetic patientstreated with PCI, as did the combined endpoint of death or myocardial infarction. The EAST (EmoryAngioplasty versus Surgery) trial36 was a single-centre randomized study comparing PCI (stents: none)with CABG (LIMA graft: 86%) in 392 patients with multivessel disease, of whom 59 were diabetic. Therewas no difference in survival among diabetic patients at 3 years. During longer-term follow-up, survivalcurves crossed after 5 years and began to diverge in favour of CABG, although the difference was notsignificant at 8 years of follow-up. The Coronary Angioplasty versus Bypass Revascularization Inves-tigation (CABRI)37 randomly assigned 1054 patients with multivessel disease to PCI (stents: none) orCABG (LIMA graft: 81%). Diabetic patients had significantly higher mortality than non-diabetic patients,but the difference among diabetic patients treated with PCI (22.6%) or CABG (12.5%, RR¼ 1.81, 95% CI0.8–4.1) was not significant. The randomized AWESOME (Angina With Extremely Serious OperativeMortality Evaluation) trial compared PCI (stents: 54%) with CABG (LIMA graft: 76%) in 454 patients ofwhom 32% (n¼ 144) had diabetes.38 A total of 1650 patients were eligible for the study but did notprovide consent and were followed in the physician-guided registry. At 3 years of follow-up, mortalitywas similar among PCI- and CABG-treated diabetic patients in the randomized cohort (PCI: 19% versusCABG: 28%, P not significant) as well as in the registry arm (PCI: 29% versus CABG: 27%, P not signifi-cant). The above-mentioned trials were performed prior to the widespread use of stents. Fourrandomized clinical trials compared PCI with the use of bare-metal stents – ARTS I (Arterial Revas-cularization Therapies Study I), SoS (Stenting of Saphenous Vein Grafts), ERACI II (ArgentineRandomized Study: Coronary Angioplasty with Stenting Versus Coronary Bypass Surgery in Multi-Vessel Disease), and MASS II (the Massachusetts stent trial II) – against CABG in patients with multi-vessel disease (Table 1).39–42 All these studies lacked adjunctive pharmacological treatment, includingthienopyridine pre-loading and glycoprotein IIb/IIIa antagonists, and therefore their applicability totoday’s clinical practice is limited.

In ARTS I, a total of 1205 patients were randomly assigned treatment with PCI (BMS: 100%) or CABG(LIMA graft: 93%).39 At 5 years, differences in rates of death were not significant between PCI and CABG(PCI: 8.0% versus CABG: 7.6%, RR¼ 1.05, 95% CI 0.71–1.55, P¼ 0.83) in the overall population. Whilemortality was higher in diabetic (13.4%) as opposed to non-diabetic (6.8%, P¼ 0.03) patients treatedwith PCI, this difference was not observed among patients treated with CABG (diabetes: 8.3% versus nodiabetes: 7.5%, P¼ 0.80). Along this line, mortality tended to be higher with PCI (13.4%) than CABG(8.3%, RR¼ 1.61, 95% CI 0.71–3.63, P¼ 0.39) among the 208 diabetic patients, although differences were

Table 1Randomized controlled trial: bare-metal stents (BMS) versus coronary artery bypass grafting (CAB for multivessel disease (MVD).

Study general characteristics Overall population Diabetic population

Study Enrolmentperiod

Follow-up (y)

Patientsrandomized

Diabetics PCI CABG Mortality Repeatrevascul ization

MACCE Mortality RepeatRevascularization

MACCE

PCI, n CABG,n

PCI,n

CABG,n

% Stent/%DES

% Arterialgraft

PCI(%)

CABG(%)

P-value PCI(%)

CA(%

P-value PCI(%)

CABG(%)

P-value PCI(%)

CABG(%)

P-value PCI(%)

CABG(%)

P-value PCI(%)

CABG(%)

P-value

RITA 1988–1991 6.5 510 501 29 33 98%/0% 74% 8% 9% 0.51 44% 11 0.01 14% 11% ns 7% 24% 0.09 na na na 17% 36% 0.06EAST 1987–1990 3 198 194 29 30 0%/0% 86% 7% 5% ns 57% 14 <0.001 na na na 7% 10% ns na na na na na na

� 5 � � � � � � 13% 6% ns 61% 16 <0.001 na na na 10% 10% ns na na na na na na� 8 � � � � � � 21% 17% 0.4 65% 27 <0.001 na na na 40% 25% 0.23 na na na na na na

CABRI 1988–1992 4 542 493 62 63 0%/0% 81% 11% 7% ns na na na na na na 23% 13% ns na na na na na naBARI 1988–1991 5 915 914 170 173 0%/0% 81% 14% 11% 0.19 54% 8% <0.001 na na na 35% 19% 0.002 na na na na na na

� 10 � � � � � � 29% 27% 0.18 77% 20 <0.001 na na na 55% 42% 0.025 80% 19% <0.001 na na naAWESOME 1995–2000 5 222 232 65 79 54%/ 0% 76% 20% 21% ns na na na 52% 39% 0.01 19% 28% ns na na na na na naARTS 1997–1998 1 600 605 112 96 100%/0% 93% 3% 3% 0.75 17% 4% <0.001 27% 12% <0.001 6% 3% 0.29 8% 0% <0.001 37% 26% <0.001

� 3 � � � � � � 4% 5% 0.47 27% 7% <0.001 34% 17% <0.001 7% 4% ns 15% 8% 0.02 41% 19% <0.001� 5 � � � � � � 8% 8% 0.83 30% 9% <0.001 42% 22% <0.001 13% 8% 0.27 43% 10% <0.001 55% 25% <0.001

SoS 1996–1999 2 488 500 68 74 78%/0% 81% 5% 2% 0.01 21% 6% <0.001 23% 12% <0.001 4% 1% 0.55 25% 5% 0.001 na na na� 5 � � � � � � 8% 4% 0.01 na na na na na na 10% 1% 0.001 na na na na na na� 6 � � � � � � 11% 7% 0.02 na na na na na na 18% 5% 0.01 na na na na na na

ERACI II 1996–1998 1 225 225 39 39 77%/0% 89% 3% 8% 0.02 17% 5% <0.001 na na ns 10% 10% 0.66 na na na na na na5 � � � � � � 7% 12% 0.18 28% 7% <0.001 35% 25% 0.019 10% 10% 0.66 na na na na na na

MASS II 1995–2000 1 205 203 56 59 68%/0% 92% 4% 4% 0,86 12% 0% <0.002 na na na 5% 7% 0.59 na na na na na na5 � � � � � � 16% 13% 0.4 32% 4% <0.001 na na na 16% 15% 0.4 na na na na na na

RITA, Randomized Intervention Treatment of Angina; EAST, Emory Angioplasty versus Surgery Tr l; CABRI, Coronary Angioplasty versus Bypass Revascularization Investigation; BARI,Bypass Angioplasty Revascularization Investigation; AWESOME, Angina With Extremely Serious O erative Mortality Evaluation; ARTS, Arterial Revascularization Therapies Study; SoS,Stenting of Saphenous Vein Grafts trial; ERACI II, Argentine Randomized Study: Coronary Angiop sty with Stenting Versus Coronary Bypass Surgery in Multi-Vessel Disease II; MASS,Massachusetts stent trial; PCI, percutaneous coronary intervention; DES, drug-eluting stents; ns, n significant; na, not applicable.

G)

ar

BG)

%%%%

%

iap

laot


not significant and the study was not powered to address this question. Repeat revascularizationprocedures were more frequent among PCI-treated patients in the overall population (PCI: 30.3%versus CABG: 8.8%, RR¼ 3.46, 95% CI 2.61–4.60, P< 0.001), a difference which was even morepronounced in diabetic patients (PCI: 42.9% versus CABG: 10.4%, RR¼ 4.11, 95% CI 2.20–7.68, P< 0.001).

The SoS trial randomized 988 patients with multivessel disease to treatment with PCI (BMS: 100%)or CABG (LIMA graft: 93%).40 It is the only trial that showed an increased mortality among patientstreated with PCI compared to CABG in the overall population (PCI: 10.9% versus CABG: 6.8%, HR¼ 1.66,95% CI 1.08-2.55, P¼ 0.02) at 6 years. Comparison of mortality at 5 years showed similar outcomeamong PCI patients enrolled into SoS (8.1%), ARTS I (8.0%) and ERACI II (7.1%), whereas mortality amongCABG patients was much lower in SoS (4.3%) than in ARTS I (7.6%) and ERACI II (11.5%) in the overallpopulation. Among diabetic patients, mortality was 17.6% for PCI- and 5.4% for CABG-treated patients,with no evidence that the treatment effect differed among diabetic and non-diabetic patients (inter-action test P¼ 0.15). Of note, mortality of diabetic patients treated with CABG was unexpectedly lowerin diabetic (5.4%) than in non-diabetic patients (7.0%).

ERACI II randomly assigned 450 patients with multivessel disease to undergo PCI (BMS: 100%) orCABG (LIMA graft: 89%).41 At 5 years of follow-up, there were no significant differences in mortalityamong PCI-treated (7.1%) and CABG-treated (11.5%, P¼ 0.18) patients in the overall population. Simi-larly, there was no difference in mortality among diabetic patients (PCI: 10.3% versus CABG: 10.3%). Therandomized MASS II trial compared PCI (n¼ 205; BMS: 68%) with CABG (n¼ 203; LIMA graft: 92%) andmedical treatment (n¼ 203) in patients with multivessel disease.42 At 5 years of follow-up, the studyshowed similar mortality among diabetic patients treated with PCI (16.1%) and CABG (15.3%), whereasmortality was higher among medically treated patients (25.3%), a difference which became significantafter the first year (P¼ 0.04).

An individual patient data meta-analysis43 of four trials (ARTS I, SoS, ERACI II, MASS II), comprising3051 patients, compared the clinical outcome at 12 months between patients undergoing PCI with theutilization of BMS against CABG. The primary endpoint of the study – a composite of death, myocardialinfarction, and stroke – occurred in 8.7% of PCI-allocated and 9.1% of CABG-allocated patients(HR¼ 0.95, 95% CI 0.74–1.2). Repeat revascularization procedures were more common among PCI-treated (18%) than CABG-treated (4.4%, HR¼ 4.4, 95% CI 3.3–5.9) patients. A subgroup analysis indiabetic patients revealed no significant difference in mortality among patients treated with PCI (5.6%)or CABG (3.5%, HR¼ 1.6, 95% CI 0.72–3.6, P¼ 0.30).

More recently, Bravata and colleagues presented the largest systematic review comprising 23 RTCsin which 5019 patients were randomly assigned to PCI (balloon angioplasty or BMS) and 4944 patientsrandomly allocated to CABG.30 Completeness of follow-up was high among all trials, and the follow-upduration extended up to 10 years. Procedural mortality was low, and similar for PCI (1.1%) and CABG(1.8%, P not significant), whereas procedural stroke was more common with CABG (1.2% versus 0.6%,95% CI 0.2–1.0%, P¼ 0.002). Mortality assessed at 1 year (PCI: 3.5% versus CABG: 3.6%) and 5 years (PCI:10.3% versus CABG: 9.3%) was similar among PCI and CABG in the overall population. However, anginarelief was better and repeat revascularization procedures less frequent among CABG- than PCI-treatedpatients. A subgroup analysis among diabetic patients showed no difference in mortality between PCI(20.8%) and CABG (17.8%, P not significant) (Fig. 1).

Drug-eluting stents versus coronary artery bypass grafting in diabetic patients with stable angina pectorisTo date, several registries have compared clinical outcome of patients with multivessel disease

following treatment with DES or CABG44–51, and these are summarized in Table 2 and Fig. 2. ARTS II wasa non-randomized registry to determine the safety and efficacy of sirolimus-eluting stents in 607patients with multivessel disease. In order to make the results of this registry as comparable as possibleto the randomized ARTS I trial comparing BMS with CABG, the same inclusion and exclusion criteria,the same protocol definitions, and the same primary endpoint were chosen. Nevertheless, as comparedwith ARTS I, patients included into ARTS II had more three-vessel disease, a higher incidence of dia-betes, and were treated with more and longer stents. At 1 year of follow-up, the incidence of repeatrevascularization was 8.5% in ARTS II and therefore significantly lower than in the historical BMS arm ofARTS I (21.3%, RR¼ 0.44, 95% CI 0.31–0.61), but still higher than in the historical CABG arm of ARTS I(4.2%, RR¼ 2.03, 95% CI 1.23–3.34). Conversely, the combined endpoint of death, myocardial infarction,

PCI betterCABG better

AWESOME, 2002 (33)BARI, 1997 (37), 1996 (64)EAST, 2000 (80)ERACI II, 2005 (86)MASS II, 2006 (103)RITA, 1998 (110)Overall

8/1145/6926/2935/3947/5627/29

188/233

8/1275/9327/3035/3950/5925/33

220/266

CABGPCI

Surviving Patients/All Patients

Five-year survival in patients with diabetes

Fig. 1. Survival according to revascularization type in diabetic patients. AWESOME, Angina With Extremely Serious OperativeMortality Evaluation; BARI, Bypass Angioplasty Revascularization Investigation; EAST, Emory Angioplasty versus Surgery Trial; ERACIII, Argentine Randomized Study: Coronary Angioplasty with Stenting Versus Coronary Bypass Surgery in Multi-Vessel Disease II;RITA, Randomized Intervention Treatment of Angina; MASS II, Massachusetts stent trial II; CABG, coronary artery bypass grafting;PCI, percutaneous coronary intervention. From Bravata et al (2007, Annals of Internal Medicine 147: 703–716) with permission.


or stroke was lower in ARTS II (3.0%) than the CABG-ARTS I group (8.0%, RR¼ 0.37, 95% CI 0.30–0.51). Ina stratified analysis, diabetic patients had a higher rate of major adverse cardiac and cerebrovascularevents (MACCE) (15.7% versus 8.5%, RR¼ 1.85, 95% CI 1.16–2.97, P¼ 0.09) and repeat interventions(13.4% versus 6.8%, RR¼ 1.97, 95% CI 1.16–3.34, P< 0.01) compared with non-diabetic patients. Therewere no significant differences between ARTS II and the CABG arm of ARTS I in MACCE at 1 year. Morerecently, the 3-year follow-up data of ARTS II have been reported and compared to the historical groupsof ARTS I. Similarly to the 1-year data, MACCE were similar in ARTS II (19.4%) and the CABG arm of ARTSI (16.2%, P¼ 0.21) but more frequent in the BMS arm of ARTS I (34%, P< 0.001). Freedom from revas-cularization remained lowest within the CABG arm of ARTS I (93.4%) followed by the DES arm ARTS II(85.5%, P< 0.001) and topped by the BMS arm of ARTS I (73.7%, P< 0.001). The results of ARTS II have tobe cautiously interpreted in the light of the non-randomized nature of the study. Accordingly, selectionbias and confounding by indication cannot be excluded, and the comparison with CABG requiresconfirmation in prospective randomized clinical trials.

ERACI III enrolled 225 patients with multivessel disease who were treated with DES and respectedthe same inclusion and exclusion criteria as the randomized ERACI II trial comparing BMS with CABG.46

The rate of major adverse cardiac events at 3 years in ERACI III was significantly higher in diabetic thannon-diabetic patients (36.2% versus 19.1%, RR¼ 0.81, 95% CI 0.66–0.99, P¼ 0.02) with no differencesamong DES-treated patients in ERACI III (36.2%) and patients undergoing CABG (30.8%) or BMSimplantation (43.6%, P¼ 0.49) in ERACI II. The need for repeat revascularization was lowest amongpatients undergoing CABG in ERACI II (15.4%), followed by DES-treated patients of ERACI III (21.3%) andtopped by BMS-treated patients in ERACI II (38.5%, P¼ 0.05). Of note, the rate of stent thrombosis inERACI III amounted to 2.6% at 1 year, 3.5% after 2 years, and 4.4% by the end of 3 years. The results ofERACI III have to be interpreted within the same framework as those of ARTS II, namely the fact that thecomparison to CABG and PCI with DES is not randomized but rather a historical one.

A large registry included 1680 patients undergoing revascularization for multivessel disease at theWashington Hospital Center, of whom 1080 patients were treated for two-vessel disease (884 PCI, 196CABG) and 600 patients for three-vessel disease (96 PCI, 505 CABG).52 After multivariable adjustmentfor differences in age, gender, dyslipidaemia, diabetes, left anterior descending artery involvement,prior CABG, peripheral vascular disease, and chronic renal insufficiency, major adverse cardiac andcerebrovascular events were higher for two-vessel (HR¼ 2.29, 95% CI 1.39–3.76, P¼ 0.01) and three-vessel disease (HR¼ 2.90, 95% CI 1.76–4.78, P< 0.001) patients undergoing PCI with DES than CABG.The unfavourable outcome of DES-treated patients compared with CABG was particularly pronouncedin the diabetic population, with a significantly increased risk of major adverse cardiac and

Table 2Studies comparing coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI) with drug-eluting stents (DES).

Study general characteristics Overall population Diabetic population

Study Randomized? Follow-up (y)

Patients Diabetics Mortality Repeatrevascularization

MACCE Mortality Repeatrevascularization

MACCE

PCI-DES,n

CABG,n

PCI-DES,n

CABG,n

PCI-DES,n

CABG,%

P-value PCI-DES,n

CABG,%

P-value PCI-DES,n

CABG,%

P-value PCI-DES,n

CABG,%

P-value PCI-DES,n

CABG,%

P-value PCI-DES,n

CABG,%

P-value

ARTS II Single-arm trialcomparedwithARTS I

3 607 605 158 96 3% 4% ns 14% 7% <0.001 19% 16% 0.21 na na ns na na na 28% 18% 0.09

ERACI III Single-arm trialcomparedwith ERACI II

3 225 225 47 39 6% 10% 0.16 14% 6% 0.002 23% 23% ns na na ns 21% 15% 0.58 36.2% 30.8% 0.49

SamsungMedicalcenter

No 1.8 441 390 163 189 2% 3% 0.17 9% 1% <0.001 13% 4% <0.001 4% 4% 0.49 12% 1% <0.001 18% 5% <0.001

WashingtonHospitalcenter

No 1 979 701 344 257 8% 3% 0.006 na na na 25% 10% <0.001 14% 2% 0.003 na na na 31% 9% <0.001

RuiJinHospital

No 2.1 235 231 75 64 3% 2% 0.77 10% 3% 0.001 15% 8% 0.03 na na na na na na na na na

Briguori et al No 1 � � 69 149 � � � � � � � � � 6% 5% 0.74 19% 5% 0.001 29% 21% 0.17UCLA-Cedars-

SinaiMedicalcenter

No 1 � � 102 103 � � � � � � � � � 10% 8% 0.6 20% 3% <0.001 27% 12% 0.006

MACCE, major adverse cardiac and cerebrovascular events; ARTS II, Arterial Revascularization Therapies Study II; ERACI III, Argentine Randomized Study: Coronary Angioplasty with StentingVersus Coronary Bypass Surgery in Multi-Vessel Disease III; UCLA, University of California, Los Angeles.

MACCE Mortality

PCI better CABG better PCI better CABG better

DM2vCAD

DM3vCAD

Non-DM2vCAD

Non-DM3vCAD

<0.001

<0.001 <0.001

0.03

0.34

0.430.19

0.07

DM2vCAD

DM3vCAD

Non-DM2vCAD

Non-DM3vCAD

Fig. 2. Impact of diabetes and coronary extent on morbidity and mortality. Hazard ratios for major adverse cardiac and cerebro-vascular events (MACCE; left) and mortality (right) in subgroup with (DM) or without (Non-DM) diabetes mellitus and two-vessel(2vCAD) or three-vessel (3vCAD) coronary artery disease (CAD) after multivariate adjustment for age, gender, dyslipidaemia, leftanterior descending artery (LAD) involvement, prior coronary artery bypass grafting (CABG), peripheral vascular disease, and chronicrenal insufficiency. Redrawn from Javaid et al (2007, Circulation 116: I-200–206) with permission.


cerebrovascular events and mortality, whereas there outcome in non-diabetic patients was similar forboth revascularization strategies.

A registry from Samsung Medical Center compared clinical outcomes of 831 patients with multi-vessel disease treated with either DES (n¼ 441) or CABG (n¼ 390) between March 2003 and March2005.47 At 1 year follow-up, results in diabetic patients were similar to those of the overall patientpopulation, with no difference between DES- and CABG-treated patients in terms of overall mortality(3.8% versus 3.8%, P¼ 0.49) and stroke (1.3% versus 05%, P¼ 0.25), but a significantly higher rate ofrepeat revascularization procedures (12.4% versus 0.5%, P< 0.001) and therefore a higher rate of majoradverse cardiac and cerebrovascular events (18.3% versus 4.9%, P< 0.001).

Similar result were reported in a registry from RuiJin Hospital of 466 consecutive patients withmultivessel disease undergoing revascularization with either DES or CABG. 49 Although short-termoutcome (30 days) was better among DES- than CABG-treated patients, there were no significantdifferences between either revascularization strategies at 25� 8 months of follow-up regarding overalldeath, myocardial infarction or stroke (DES: 6.3% versus CABG: 5.6%, P¼ 0.84). However, the need forrepeat revascularization procedures was higher among DES- than CABG-treated patients (DES: 10.4%versus CABG: 2.8%, P¼ 0.001).

Briguori and colleagues reported on 218 consecutive diabetic patients undergoing treatment withCABG (n¼ 149) or DES (n¼ 69).50 At 1 year follow-up, propensity score-adjusted analysis showed nodifference in the rate of the composite endpoint of death, myocardial infarction, or stroke (DES: 13%versus CABG: 12%, HR¼ 0.80, 95% CI 0.80–1.35, P¼ 0.40) but a higher rate of repeat revascularizationamong DES-treated patients (19% versus 5%, HR¼ 2.05, 95% CI 1.12–3.75, P¼ 0.001).

A registry of 205 consecutive diabetic patients with multivessel disease undergoing treatment withDES (n¼ 102 patients) or CABG (n¼ 103 patients) observed similar mortality (8% versus 10%, P¼ 0.6),more strokes among CABG patients (0% versus 4%, P¼ 0.04), and conversely more myocardial infarc-tions potentially related to stent thrombosis and a higher rate of repeat revascularization proceduresamong DES-treated patients at 1 year of follow-up.

In an epidemiological study, Hannan and colleagues assessed clinical outcome of patients withmultivessel disease who underwent revascularization with CABG (n¼ 7437) or DES (n¼ 9963)between 2003 and 2004 in the New York State.52 Patients undergoing CABG were on average older,more likely to be male, non-Hispanic and white, and to have a lower ejection fraction, previousmyocardial infarction, other coexisting conditions, and three-vessel disease. Rates of revascularizationat 18 months were higher among DES- than CABG-treated patients, although the latter observation wasrelated at least in part to reimbursement-driven staged procedures using PCI. Unadjusted rates of deathwere similar for DES and CABG in patients with two-vessel and three-vessel disease. CABGwas associated with lower adjusted rates of death and death or myocardial infarction for patients with

Practice points

� compared with non-diabetic patients, diabetic patients undergoing CABG are at increasedrisk of wound infections and peri-procedural morbidity and mortality� compared with non-diabetic patients, diabetic patients undergoing PCI are at increased risk

of restenosis� PCI with BMS in diabetic patients is associated with higher rates of repeat revascularization

than CABG� compared with BMS, DES have significantly reduced the risk of restenosis following PCI in

diabetic patients; however, repeat revascularization procedures remain more common afterDES than CABG in diabetic patients� completed randomized clinical trials and meta-analyses show similar long-term mortality

among patients with multivessel disease irrespective of revascularization strategy� a recent dedicated randomized trial in diabetic patients confirmed a similar mortality for PCI

and CABG


two-vessel and three-vessel disease. Of note, adjusted rates of death (HR¼ 0.97, 95% CI 0.77–1.20,P¼ 0.75) and death or myocardial infarction (HR¼ 0.84, 95% CI 0.69–1.01, P¼ 0.07) were similar fordiabetic patients irrespective of revascularization strategy. This was in contradiction to a previousreport from the same investigators comparing clinical outcome following revascularization with BMSor CABG in New York State, where the latter group was associated with improved survival amongdiabetic patients with three-vessel disease.53

The UK-based CARDIA (Coronary Artery Revascularization in Diabetes) trial directly compared CABGwith PCI (predominant DES use: 71%) in diabetic patients with multivessel disease in a randomized,non-inferiority study. Due to recruitment difficulties, only 510 of 600 planned patients (85%) wererandomized, and preliminary results were presented at the European Society of Cardiology convention2008 in Munich.54 The primary endpoint, a composite of death, non-fatal MI, and stroke, assessed at1 year showed similar outcome for PCI (11.6%) and CABG (10.2%, P¼ 0.63) with no significant differ-ences in rates of death (PCI: 3.2%, CABG: 3.3%, P¼ 0.83) and myocardial infarction (PCI: 8.4%,CABG: 5.7%, P¼ 0.25), although non-fatal strokes tended to be less common with PCI (0.4%) than CABG(2.5%, P¼ 0.09). Repeat revascularization procedures were more frequent with PCI (9.9%) than CABG(2.0%, P< 0.001) at 1-year follow-up. These findings are in line with the subgroup of diabetic patients(n¼ 512 patients) included in the SYNTAX trial, a large-scale randomized study (n¼ 1800 patients)comparing CABG with DES in the treatment of patients with multivessel disease. The compositeendpoint of death, MI, and stroke at 1 year was similar for CABG (10.3%) and PCI (10.1%, P¼ 0.96),whereas repeat revascularization was more common with the latter (PCI-MACCE: 26.0%, CABG-MACCE:14.2%, P¼ 0.03).55

Revascularization in diabetic patients presenting with acute ST-elevation myocardial infarction

Diabetic patients experiencing an acute ST-elevation myocardial infarction present later to thehospital, are more likely to have congestive heart failure, and have longer door-to-balloon timesthan non-diabetic patients.56 CABG in the setting of acute ST-elevation myocardial infarctionremains reserved for patients with mechanical complications (papillary muscle dysfunction,rupture of the interventricular septum) and the rare case of failed PCI. Primary PCI refers to thecatheter-based emergency reperfusion of the infarct-related artery and has been shown to improvesurvival, reduce the risk of recurrent myocardial infarction, stroke, and intracranial haemorrhage ascompared with systemic thrombolysis in patients with acute ST-elevation myocardial infarction.57

The therapeutic advantage of primary PCI over thrombolysis appears particularly pronounced indiabetic patients. A pooled subset analysis of 11 randomized clinical trials comparing primary PCIwith thrombolysis in diabetic patients showed a significant reduction in death or myocardial

Practice points

� diabetic patients with acute ST-elevation myocardial infarction present later to the hospitalthan non-diabetic patients (pain to door time is increased by 15 minutes)� pre-hospital mortality is increased in diabetic versus non-diabetic patients with acute ST-

elevation myocardial infarction� door-to-balloon time is longer in diabetic compared with non-diabetic patients with acute

ST-elevation myocardial infarction� primary PCI should be considered first-line therapy when initiation is timely and the

procedure is performed in experienced centres� patients with acute ST-elevation myocardial infarction complicated by cardiogenic shock

should undergo emergency revascularization either by percutaneous coronary interventionor coronary artery bypass grafting


infarction at 30 days (9.2% versus 19.3%, P< 0.05).58 Moreover, primary PCI as compared withthrombolysis in diabetic patients results in a higher patency rate of the infarct-related artery (88%versus 31%, P< 0.001), improved left ventricular function (LVEF 49% versus 36%, P< 0.05) as wellas long-term survival.59

Fig. 3. Angiographic examples of good (top) and poor (bottom) candidate for coronary artery bypass grafting (CABG). The goodcandidate has extensive coronary disease of the proximal part of all three major epicardial coronary vessels without significantdisease in the distal part (anastomosis). In the current illustration, the patient presented with two long high-graded stenoses of theleft anterior descending artery (LAD), a chronic occlusion of the ostium of the circumflex artery (LCx), and a long and complexstenosis of the proximal/middle part of the dominant right coronary artery (DA). On the opposite, the coronary artery diseaseaffecting a poor CABG candidate extends predominantly in the distal part of the vessels, precluding the confection of any surgicalanastomosis. In the present case, the patient suffered from a diffuse form of coronary artery disease with a functional occlusion ofthe very distal part of all three major epicardial coronaries, as illustrated here by the left anterior descending artery. This patient alsopresented with two serial 50–70% stenoses of the left subclavian artery (LSCA). LIMA, left internal mammary artery.

Fig. 4. Angiographic examples of good (top) and poor (bottom) candidate for percutaneous coronary intervention (PCI). A good PCIcandidate has a limited disease with focal stenosis. In the current illustration, the patient presented with one high-grade stenosis ofthe distal part of the circumflex artery (LCx, inaccessible to coronary artery bypass grafting, CABG), as well as a high-grade stenosis ofthe proximal part of the dominant right coronary artery (RCA). The left anterior descending artery (LAD) is not affected. A poor PCIcandidate has a diffuse disease with the presence of long and calcified lesions. The patient depicted in this illustration has a chronictotal occlusion of the distal part of LAD and proximal RCA, as well as serial high-grade stenosis of LCx.


The value of glycoprotein IIb/IIIa antagonists in addition to standard antiplatelet therapy includingaspirin and clopidogrel has been evaluated in diabetic patients with acute ST-elevation myocardialinfarction undergoing primary PCI. In the ADMIRAL (Abciximab Before Direct Angioplasty and Stentingin Myocardial Infarction) study, administration of the glycoprotein IIb/IIIa antagonist abciximab indiabetics was associated with a significant reduction in death (0 versus 16.7%, P¼ 0.02), and thecombined endpoint of death, reinfarction, or revascularization (20.7 versus 50.0%, P¼ 0.02) at6 months.60 In an individual-patient data meta-analysis of diabetic patients with ST-elevationmyocardial infarction, Montalescot and colleagues reported improved survival at 3 years in patientstreated with abciximab: 22% versus 40%, P¼ 0.02; number needed to treat (NNT): 6 (6–64).61 Nobenefit in terms of survival or recurrent myocardial infarction was observed in another trial (CADILLAC)assessing the benefit of adjunctive therapy with abciximab in diabetic patients, although the risk ofrepeat revascularization procedures was significantly improved.62

Revascularization in diabetic patients with cardiogenic shock

Patients with acute myocardial infarction complicated by cardiogenic shock within 36 hours ofsymptom onset benefit from emergency revascularization in terms of survival.63 While the majority ofpatients will undergo percutaneous coronary intervention, patients with extensive CAD and those notamenable to PCI should be considered for emergency CABG. In the SHOCK trial, 23 diabetic patientswith cardiogenic shock were referred for emergency surgery (49% of enrolled CABG patients) and 22patients underwent PCI (27% of enrolled PCI patients). Rates of survival in diabetic patients were similarin both groups at 30 days (PCI versus CABG: 62% versus 61%, P¼ 1.00) and at 1 year (PCI versus CABG:


52% versus 48%, P¼ 1.00).64 Accordingly, the choice of revascularization in diabetic patients withcardiogenic shock depends on the extent of CAD and the suitability for either revascularization method.

Revascularization in diabetic patients with acute coronary syndromes (unstable angina/non-ST-elevationmyocardial infarction)

Diabetes is an independent predictor of mortality in patients with acute coronary syndromes(ACS). In the OASIS (Organization to Assess Strategies for Ischemic Syndromes) registry, the mostimportant independent predictors of mortality among patients with ACS were older age (RR¼ 1.58;95% CI: 1.48–1.69, per 10 years), history of heart failure (RR¼ 2.34; 95% CI: 2.01–2.73), and diabetes(RR¼ 1.56; 95% CI: 1.35–1.79). The risk associated with diabetes corresponded to the risk ofnon-diabetic patients, who were on average 10 years older.65 In a pooled analysis of six ACS trials66,30-day mortality among 5,409 diabetic patients was almost two-fold higher than in 19,319 non-diabetic patients (5.5% versus 3.0%, HR 1.7, P< 0.001). The guidelines of the European Society ofCardiology for patients presenting with ACS have designated the status diabetes a high-risk label,suggesting that diabetic patients with ACS should be treated according to the high-risk pathway, i.e.,with an early invasive strategy.

An early invasive strategy including revascularization by either PCI or CABG has been comparedwith an initial conservative management in patients presenting with ACS in several trials. While anearly invasive strategy has been shown to significantly reduce death or myocardial infarction in theoverall population, high-risk groups including patients with elevated levels of troponin and diabeticpatients appear to benefit most. In the FRISC II (Fast Revascularisation during Instability in Coronaryartery disease II) study67,68 patients were randomly assigned to an early invasive strategy whichincluded routine catheterization and revascularization as appropriate within 7 days, or to a conser-vative strategy in which catheterization was deferred unless the patient had objective evidence ofrefractory ischaemia. The composite endpoint of death or myocardial infarction occurred in 10.4% of

Drug-eluting stents

Consider iv Gp IIb/IIIA inhibitorsAvoid contrast nephropathy

Discontinue metformin the day before

Medical therapyBeta-blockers, ACE-inhibitors

statins, aspirinclopidogrel

Symptoms ofangina or

angina equivalent

Objective evidenceof ischemia

How to Treat Diabetes Mellitus and Coronary Artery Disease

Risk Factor Intervention-hyperglycemia (HbA1C<7%)

-hyperlipidemia (LDL<100 mg/dl)

-Body mass index <25 kg/m2-ACE inhibitors

-smoking cessation

-hypertension (<130/85)

Anatomy suitable for PCIOne-, two-, or selected three

vessel disease

CABG

Left main diseaseThree-vessel disease

and reduced LVEF

Anatomy unsuitablefor PCI/CABG

Fig. 5. Treatment algorithm for the treatment of coronary artery disease in diabetic patient. LDL, low-density lipoprotein; ACE,Angiotensin-converting enzyme; PCI, percutaneous coronary intervention; CABG, coronary artery bypass graft; LVEF, left ventricularejection fraction.


patients undergoing an early invasive strategy compared with 14.1% (P¼ 0.005) of patients treatedconservatively. The difference in favour of an early invasive strategy was more pronounced in diabeticpatients (29.9% to 20.6%, OR 0.61; CI 0.36–1.04; P¼ 0.066) than in non-diabetic patients (12.0% to 8.9%,OR 0.72; CI 0.54–0.95; P¼ 0.02). Of 2220 patients enrolled into TACTICS-TIMI 1869, 28% were diabetic.Patients were randomly assigned to an early invasive strategy which included routine catheterizationwithin 4–48 hours and revascularization as appropriate, or to an initial conservative strategy. At6 months, the composite of death, non-fatal myocardial infarction, and re-hospitalization for ACS wassignificantly reduced in patients assigned to an early invasive strategy (OR¼ 0.78; 95% CI: 0.62–0.97,P¼ 0.03). Again, the difference in favour of an early invasive strategy was more pronounced in diabetic(risk reduction: �27%) than in non-diabetic patients (risk reduction: �13%).

Diabetic patients with ACS benefit from adjunctive therapy with glycoprotein IIb/IIIa antagonistswhen undergoing revascularization. A meta-analysis of six studies in diabetic patients with ACSshowed improved survival with the use of glycoprotein IIb/IIIa antagonists.66

Conclusions

Compared with non-diabetic patients, diabetic patients have more extensive atherosclerosis anda worse clinical outcome following revascularization procedures. Notwithstanding, the therapeuticbenefit is particularly pronounced in this high-risk subgroup, justifying aggressive revascularizationand medical treatment for secondary prevention. The best coronary revascularization of diabeticpatients depends on the clinical setting and anatomical factors (Figs. 3–5). In the acute settingincluding ST-elevation myocardial infarction and non-ST-elevation myocardial infarction, PCI appearspreferable. In patients with stable CAD, the extent of disease and non-cardiac morbidity require morecareful evaluation. CABG appears more effective in terms of repeat revascularization procedures inpatients with advanced multivessel disease, while PCI with the use of DES and adjunctive pharma-cological treatment, including thienopyridines and glycoprotein IIb/IIIa antagonist, is a valuablealternative in patients with less extensive disease. Of note, data from RCTs and meta-analyses indicateno difference with respect to long-term ischaemic endpoints between the two revascularizationstrategies. Several ongoing trials will shed more light on the relative efficacy of CABG and PCI amongdiabetic patients as well as the benefit of intensive medical treatment as compared with eitherrevascularization strategy.

Research agenda

� November 2008, BARI 2D: randomizing patients in a 2� factorial design to insulin-providingor insulin-sensitizing medical therapy, and to contemporary PCI or coronary artery bypasssurgery (based on physician judgment) versus no revascularization (http://www.clinicaltrials.gov, NCT00006305)� December 2011, FREEDOM: randomized, multicentre trial comparing PCI with DES to CABG in

diabetics with multivessel disease (http://www.clinicaltrials.gov, NCT00086450)� August 2012, VA CARDS: randomized, multicentre trial comparing PCI with DES to CABG in

diabetics (http://www.clinicaltrials.gov, NCT00326196)

Practice points

� mortality in patients with ACS is twice as high in diabetic than in non-diabetic patients� diabetic patients show a particularly high benefit from an early invasive strategy as compared

to an initial conservative strategy� administration of glycoprotein IIb/IIIa antagonists is associated with improved survival in

diabetic patients with ACS

http://www.clinicaltrials.gov





References

1. Seidell JC. Obesity, insulin resistance and diabetes - a worldwide epidemic. The British Journal of Nutrition 2000; 2000:S5–S8.

2. Kopelman PG & Hitman GA. Diabetes. Exploding type II. Lancet 1998; 352(Suppl. 4): SIV5.3. Cook S & Scherrer U. Insulin resistance, a new target for nitric oxide-delivery drugs. Fundamental & Clinical Pharmacology

2002; 16: 441–453.4. Amos AF, McCarty DJ & Zimmet P. The rising global burden of diabetes and its complications: estimates and projections to

the year 2010. Diabetic Medicine 1997; 14(Suppl. 5): S1–85.5. Girman CJ, Rhodes T, Mercuri M et al. The metabolic syndrome and risk of major coronary events in the Scandinavian

Simvastatin Survival Study (4S) and the Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS). TheAmerican Journal of Cardiology 2004; 93: 136–141.

6. King H, Aubert RE & Herman WH. Global burden of diabetes, 1995–2025: prevalence, numerical estimates, and projections.Diabetes Care 1998; 21: 1414–1431.

7. Mokdad AH, Ford ES, Bowman BA et al. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. TheJournal of the American Medical Association 2003; 289: 76–79.

8. Cook S. Coronary artery disease, nitric oxide and oxidative stress: the ‘Yin-Yang’ effect–a Chinese concept for a worldwidepandemic. Swiss Medical Weekly 2006; 136: 103–113.

9. Haffner SM, Lehto S, Ronnemaa T et al. Mortality from coronary heart disease in subjects with type 2 diabetes and innondiabetic subjects with and without prior myocardial infarction. The New England Journal of Medicine 1998; 339:229–234.

10. Stratton IM, Adler AI, Neil HAW et al. Association of glycaemia with macrovascular and microvascular complications of type2 diabetes (UKPDS 35): prospective observational study. British Medical Journal 2000; 321: 405–412.

11. Group DS & on behalf of the European Diabetes Epidemiology G. Glucose tolerance and cardiovascular mortality:comparison of fasting and 2-hour diagnostic criteria. Archives of Internal Medicine 2001; 161: 397–405.

12. Stein B, Weintraub WS, Gebhart SSP et al. Influence of diabetes mellitus on early and late outcome after percutaneoustransluminal coronary angioplasty. Circulation 1995; 91: 979–989.

13. Kip KE, Faxon DP, Detre KM et al. Coronary angioplasty in diabetic patients: the National Heart, Lung, and Blood InstitutePercutaneous Transluminal Coronary Angioplasty Registry. Circulation 1996; 94: 1818–1825.

14. Wenaweser P, Daemen J, Zwahlen M et al. Incidence and correlates of drug-eluting stent thrombosis in routine clinicalpractice: 4-year results from a large 2-institutional cohort study. Journal of the American College of Cardiology 2008; 52:1134–1140.

15. Stettler C, Allemann S, Wandel S et al. Drug eluting and bare metal stents in people with and without diabetes: collaborativenetwork meta-analysis. British Medical Journal 2008; 337: a1331.

16. Cook S, Meier B & Windecker S. Sirolimus-eluting versus paclitaxel-eluting stents: are there differences in reality? TheInternet Journal of Cardiology 2007: 19–25. Spring.

17. Sabate M, Jimenez-Quevedo P, Angiolillo DJ et al. Randomized comparison of sirolimus-eluting stent versus standard stentfor percutaneous coronary revascularization in diabetic patients: the diabetes and sirolimus-eluting stent (DIABETES) trial.Circulation 2005; 112: 2175–2183.

18. Baumgart D, Klauss V, Baer F et al. One-year results of the SCORPIUS study: a German multicenter investigation on theeffectiveness of sirolimus-eluting stents in diabetic patients. Journal of the American College of Cardiology 2007; 50: 1627–1634.

19. Tu JV, Bowen J, Chiu M et al. Effectiveness and safety of drug-eluting stents in Ontario. The New England Journal of Medicine2007; 357: 1393–1402.

20. Spaulding C, Daemen J, Boersma E et al. A pooled analysis of data comparing sirolimus-eluting stents with bare-metalstents. The New England Journal of Medicine 2007; 356: 989–997.

21. Kastrati A, Mehilli J, Pache J et al. Analysis of 14 trials comparing sirolimus-eluting stents with bare-metal stents. The NewEngland Journal of Medicine 2007; 356: 1030–1039.

22. Stettler C, Wandel S, Allemann S et al. Outcomes associated with drug-eluting and bare-metal stents: a collaborativenetwork meta-analysis. Lancet 2007; 370: 937–948.

23. Carson JL, Scholz PM, Chen AY et al. Diabetes mellitus increases short-term mortality and morbidity in patients undergoingcoronary artery bypass graft surgery. Journal of the American College of Cardiology 2002; 40: 418–423.

24. Mohammadi S, Dagenais F, Mathieu P et al. Long-term impact of diabetes and its comorbidities in patients undergoingisolated primary coronary artery bypass graft surgery. Circulation 2007; 116: I-220–I-225.

25. Leavitt BJ, Sheppard L, Maloney C et al. Effect of diabetes and associated conditions on long-term survival after coronaryartery bypass graft surgery. Circulation 2004; 110: II-41–II-44.

26. Tatoulis J, Buxton BF & Fuller JA. Patencies of 2,127 arterial to coronary conduits over 15 years. The Annals of Thoracic Surgery2004; 77: 93–101.

27. Stevens LM, Carrier M, Perrault LP et al. Influence of diabetes and bilateral internal thoracic artery grafts on long-termoutcome for multivessel coronary artery bypass grafting. European Journal of Cardio-Thoracic Surgery 2005; 27: 281–288.

28. de Jaegere PPT & Suyker WJL. Off-pump coronary artery bypass surgery. Heart 2002; 88: 313–318.29. Flaherty JD & Davidson CJ. Diabetes and coronary revascularization. The Journal of the American Medical Association 2005;

293: 1501–1508.30. Bravata DM, Gienger AL, McDonald KM et al. Systematic review: the comparative effectiveness of percutaneous coronary

interventions and coronary artery bypass graft surgery. Annals of Internal Medicine 2007; 147: 703–716.31. The Bypass Angioplasty Revascularization Investigation (BARI) Investigators. Comparison of coronary bypass surgery with

angioplasty in patients with multivessel disease. The New England Journal of Medicine 1996; 335: 217–225.32. The Bypass Angioplasty Revascularization Investigation (BARI) Investigators. Influence of diabetes on 5-year mortality and

morbidity in a randomized trial comparing CABG and PTCA in patients with multivessel disease. Circulation 1997; 96:1761–1769.


33. Feit F, Brooks MM, Sopko G et al. Long-term clinical outcome in the bypass angioplasty revascularization investigationregistry: comparison with the randomized trial. Circulation 2000; 101: 2795–2802.

34. The Bypass Angioplasty Revascularization Investigation (BARI) Investigators. The final 10-year follow-up results from theBARI randomized trial. Journal of the American College of Cardiology 2007; 49: 1600–1606.

35. Henderson RA, Pocock SJ, Sharp SJ et al. Long-term results of RITA-1 trial: clinical and cost comparisons of coronaryangioplasty and coronary-artery bypass grafting. Randomised intervention treatment of angina [see comments]. Lancet1998; 352: 1419–1425.

36. King SB, Lembo NJ, Weintraub WS et al. A randomized trial comparing coronary angioplasty with coronary bypass surgery.The New England Journal of Medicine 1994; 331: 1044–1050.

37. CABRI Trial Participants. First-year results of CABRI (coronary angioplasty versus bypass revascularisation investigation).Lancet 1995; 346: 1179–1184.

38. Sedlis SP, Morrison DA, Lorin JD et al. Percutaneous coronary intervention versus coronary bypass graft surgery for diabeticpatients with unstable angina and risk factors for adverse outcomes with bypass: outcome of diabetic patients in theAWESOME randomized trial and registry. Journal of the American College of Cardiology 2002; 40: 1555–1566.

39. Serruys PW, Ong ATL, van Herwerden LA et al. Five-year outcomes after coronary stenting versus bypass surgery for thetreatment of multivessel disease: the final analysis of the Arterial Revascularization Therapies Study (ARTS) randomizedtrial. Journal of the American College of Cardiology 2005; 46: 575–581.

40. Booth J, Clayton T, Pepper J et al. Randomized, controlled trial of coronary artery bypass surgery versus percutaneouscoronary intervention in patients with multivessel coronary artery disease: six-year follow-up from the Stent or SurgeryTrial (SoS). Circulation 2008; 118: 381–388.

41. Rodriguez AE, Baldi J, Pereira CF et al. Five-year follow-up of the argentine randomized trial of coronary angioplasty withstenting versus coronary bypass surgery in patients with multiple vessel disease (ERACI II). Journal of the American College ofCardiology 2005; 46: 582–588.

42. Hueb W, Lopes NH, Gersh BJ et al. Five-year follow-up of the Medicine, Angioplasty, or Surgery Study (MASS II):a randomized controlled clinical trial of 3 therapeutic strategies for multivessel coronary artery disease. Circulation 2007;115: 1082–1089.

43. Mercado N, Wijns W, Serruys PW et al. One-year outcomes of coronary artery bypass graft surgery versus percutaneouscoronary intervention with multiple stenting for multisystem disease: a meta-analysis of individual patient data fromrandomized clinical trials. The Journal of Thoracic and Cardiovascular Surgery 2005; 130: 512–519.

44. Serruys, P., What have we learned from ARTS I and ARTS II? AHA, special session, November 4th. 2007, AHA/ESC jointsession: multivessel coronary artery disease d stents vs CABG:room W415a.

45. Serruys PW, Colombo A, Morice MC et al. Arterial revascularisation therapies study part II - sirolimus-eluting stents for thetreatment of patients with multivessel de novocoronary artery lesions. EuroIntervention 2005; 1: 147–156.

46. Rodriguez AE, Maree AO, Mieres J et al. Late loss of early benefit from drug-eluting stents when compared with bare-metalstents and coronary artery bypass surgery: 3 years follow-up of the ERACI III registry. European Heart Journal 2007; 28:2118–2125.

47. Yang JH, Gwon H-C, Cho SJ et al. Comparison of coronary artery bypass grafting with drug-eluting stent implantation for thetreatment of multivessel coronary artery disease. The Annals of Thoracic Surgery 2008; 85: 65–70.

48. Javaid A, Steinberg DH, Buch AN et al. Outcomes of coronary artery bypass grafting versus percutaneous coronary inter-vention with drug-eluting stents for patients with multivessel coronary artery disease. Circulation 2007; 116: I-200–I-206.

49. Yang ZK, Shen WF, Zhang RY et al. Coronary artery bypass surgery versus percutaneous coronary intervention with drug-eluting stent implantation in patients with multivessel coronary disease. Journal of Interventional Cardiology 2007; 20: 10–16.

50. Briguori C, Condorelli G, Airoldi F et al. Comparison of coronary drug-eluting stents versus coronary artery bypass graftingin patients with diabetes mellitus. The American Journal of Cardiology 2007; 99: 779–784.

51. Lee MS, Jamal F, Kedia G et al. Comparison of bypass surgery with drug[hyphen (true graphic)]eluting stents for diabeticpatients with multivessel disease. International Journal of Cardiology 2007; 123: 34–42.

52. Hannan EL, Wu C, Walford G et al. Drug-eluting stents versus coronary-artery bypass grafting in multivessel coronarydisease. The New England Journal of Medicine 2008; 358: 331–341.

53. Hannan EL, Racz MJ, Walford G et al. Long-term outcomes of coronary-artery bypass grafting versus stent implantation. TheNew England Journal of Medicine 2005; 352: 2174–2183.

54. Kapur A. Coronary artery revascularisation in diabetes. The CARDia trial. European Heart Journal 2008. Hotlines II. Session1694–1695: ESC, September 1.

55. Serruys, P.W.; Mohr, F.W.; investigators obotS., The Synergy between percutaneous coronary intervention with TAXUS andcardiac surgery: the SYNTAX study. New Engl J Med 359: in press.

56. Harjai KJ, Stone GW, Boura J et al. Comparison of outcomes of diabetic and nondiabetic patients undergoing primaryangioplasty for acute myocardial infarction. The American Journal of Cardiology 2003; 91: 1041–1045.

57. Keeley EC, Boura JA & Grines CL. Primary angioplasty versus intravenous thrombolytic therapy for acute myocardialinfarction: a quantitative review of 23 randomisned trials. Lancet 2003; 361: 13–20.

58. Grines C, Patel A, Zijlstra F et al. Primary coronary angioplasty compared with intravenous thrombolytic therapy for acutemyocardial infarction: six-month follow up and analysis of individual patient data from randomized trials. American HeartJournal 2003; 145: 47–57.

59. Timmer JR, Ottervanger JP, Thomas K et al. Long-term, cause-specific mortality after myocardial infarction in diabetes.European Heart Journal 2004; 25: 926–931.

60. Montalescot G, Barragan P, Wittenberg O et al. Platelet glycoprotein IIb/IIIa inhibition with coronary stenting for acutemyocardial infarction. The New England Journal of Medicine 2001; 344: 1895–1903.

61. Montalescot G, Antoniucci D, Kastrati A et al. Abciximab in primary coronary stenting of ST-elevation myocardial infarction:a European meta-analysis on individual patients’ data with long-term follow-up. European Heart Journal 2007; 28:443–449.

62. Stuckey TD, Stone GW, Cox DA et al. Impact of stenting and abciximab in patients with diabetes mellitus undergoingprimary angioplasty in acute myocardial infarction (the CADILLAC trial). The American Journal of Cardiology 2005; 95: 1–7.


63. Hochman JS, Sleeper LA, Webb JG et al. Early revascularization in acute myocardial infarction complicated by cardiogenicshock. SHOCK investigators. Should we emergently revascularize occluded coronaries for cardiogenic shock [seecomments]. The New England Journal of Medicine 1999; 341: 625–634.

64. White HD, Assmann SF, Sanborn TA et al. Comparison of percutaneous coronary intervention and coronary artery bypassgrafting after acute myocardial infarction complicated by cardiogenic shock: results from the should we emergentlyrevascularize occluded coronaries for cardiogenic shock (SHOCK) trial. Circulation 2005; 112: 1992–2001.

65. Malmberg K, Yusuf S, Gerstein HC et al. Impact of diabetes on long-term prognosis in patients with unstable angina andnon-q-wave myocardial infarction: results of the OASIS (Organization to Assess Strategies for Ischemic Syndromes) registry.Circulation 2000; 102: 1014–1019.

66. Roffi M, Chew DP, Mukherjee D et al. Platelet glycoprotein IIb/IIIa inhibitors reduce mortality in diabetic patients with non-ST-segment-elevation acute coronary syndromes. Circulation 2001; 104: 2767–2771.

67. Wallentin L, Lagerqvist B, Husted S et al. Outcome at 1 year after an invasive compared with a non-invasive strategy inunstable coronary-artery disease: the FRISC II invasive randomised trial. FRISC II investigators. Fast revascularisation duringinstability in coronary artery disease. Lancet 2000; 356: 9–16.

68. Norhammar A, Malmberg K, Diderholm E et al. Diabetes mellitus: the major risk factor in unstable coronary artery diseaseeven after consideration of the extent of coronary artery disease and benefits of revascularization. Journal of the AmericanCollege of Cardiology 2004; 43: 585–591.

69. Cannon CP, Weintraub WS, Demopoulos LA et al. Comparison of early invasive and conservative strategies in patients withunstable coronary syndromes treated with the glycoprotein IIB/IIIA inhibitor tirofiban. The New England Journal of Medicine2001; 344: 1879–1887.

Documents

1-s2.0-S1521690X08001711-main