J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 7 , N O . 5 , 2 0 1 4

ª 2 0 1 4 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N I S S N 1 9 3 6 - 8 7 9 8 / $ 3 6 . 0 0

P U B L I S H E D B Y E L S E V I E R I N C . h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j c i n . 2 0 1 3 . 1 2 . 2 0 2

Coronary Artery Bypass Graft Surgery VersusPercutaneous Coronary Intervention WithFirst-Generation Drug-Eluting Stents

A Meta-Analysis of Randomized Controlled Trials

Juwairia Al Ali, MD,* Caroline Franck, MSC,y Kristian B. Filion, PHD,yzxMark J. Eisenberg, MD, MPHyxkMontreal, Quebec, Canada

Objectives This study sought to compare the efficacy of coronary artery bypass graft surgery (CABG) to that ofpercutaneous coronary intervention (PCI) with first-generation drug-eluting stents among patients withmultivessel disease (MVD), unprotected left main (LM) disease, and single-vessel proximal left anteriordescending (LAD) disease.

Background The efficacy and safety of CABG versus PCI with drug-eluting stents in patient subgroups remainscontroversial.

Methods We systematically searched Cardiosource, Circulation, Clinicaltrials.gov, the Cochrane Library, EMBASE,and Medline for articles published through June 2013 for randomized controlled trials comparing CABG with PCI.Primary endpoints included mortality, myocardial infarction, revascularization, and stroke. Data were meta-analyzed with random-effects models.

Results We identified 7 randomized controlled trials (N ¼ 5,835): 2 of MVD (n ¼ 2,410, 100% diabetic), 2 of LMdisease (n ¼ 1,206, 29.0% diabetic), 1 of 3-vessel or LM disease (n ¼ 1,900, 25.5% diabetic), and 2 of single-vesselproximal LAD disease (n ¼ 319, 36.3% diabetic). In MVD patients, CABG reduced the risk of mortality (risk ratio[RR]: 0.70, 95% confidence interval [CI]: 0.57 to 0.87), myocardial infarction (RR: 0.47, 95% CI: 0.36 to 0.61), andrepeat revascularization (RR: 0.36, 95% CI: 0.24 to 0.52), but increased stroke risk (RR: 1.72, 95% CI: 1.02 to 2.90). Inpatients with LM disease, CABG reduced revascularization risk (RR: 0.60, 95% CI: 0.46 to 0.78) and increasedstroke risk (RR: 2.89, 95% CI: 1.15 to 7.27). Data for patients with single-vessel proximal LAD disease wereinconclusive.

Conclusions CABG is more efficacious than is PCI with first-generation drug-eluting stents in patients with LMand MVD, at the cost of increased rates of stroke. No conclusion can be drawn for patients with single-vesselproximal LAD disease. (J Am Coll Cardiol Intv 2014;7:497–506) ª 2014 by the American College of CardiologyFoundation

From the *Division of Cardiology, McGill University Health Center, Montreal, Quebec, Canada; yCenter for Clinical Epidemiology, Lady Davis

Institute for Medical Research, Jewish General Hospital, Montreal, Quebec, Canada; zDivision of Clinical Epidemiology, McGill University, Montreal,

Quebec, Canada; xDepartment of Epidemiology, Biostatistics, and Occupational Health, McGill University, Montreal, Quebec, Canada; and the

kDivision of Cardiology, Jewish General Hospital, McGill University, Montreal, Quebec, Canada. Dr. Filion is a Canadian Institutes of Health Research

(CIHR) New Investigator. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.

Manuscript received November 25, 2013; accepted December 5, 2013.

Al Ali et al. J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 7 , N O . 5 , 2 0 1 4

CABG Versus PCI With DES M A Y 2 0 1 4 : 4 9 7 – 5 0 6

498

The efficacy and safety of coronary artery bypass graft surgery(CABG) versus percutaneous coronary intervention (PCI)with first-generation drug-eluting stents (DES) remainscontroversial for the treatment of various patient subgroups.

See page 507

Although this topic has been researched in many observa-tional studies (1–5), the recommended procedural choicefor patients within subgroups of vessel disease remains un-clear due to a lack of data from randomized controlledtrials (RCTs). Therefore, our objective was to conduct ameta-analysis of RCTs comparing the efficacy and safety of

Abbreviationsand Acronyms

CABG = coronary artery

bypass graft surgery

CI = confidence interval

DES = drug-eluting stent(s)

EES = everolimus-eluting

stent(s)

HR = hazard ratio

LAD = left anterior

descending

LM = left main

MI = myocardial infarction

MIDCAB = minimally invasive

direct coronary artery bypass

graft surgery

MVD = multivessel disease

PES = paclitaxel-eluting

stent(s)

PCI = percutaneous coronary

intervention

RCT = randomized controlled

trial(s)

RR = risk ratio

SES = sirolimus-eluting

stent(s)

CABG to that of PCI with first-generation DES among patientswith multivessel disease (MVD),unprotected left main (LM) dis-ease, and single-vessel proximalleft anterior descending artery(LAD) disease.

Methods

Search strategy. We systemati-cally searched the CochraneLibrary, EMBASE, andMedlinefor articles published throughJune 2013, with the terms “cor-onary angiography” and “coro-nary artery bypass surgery” and“coronary artery bypass graft”and “drug-eluting stents”. Thesearch was restricted to RCTsconducted in humans and pub-lished in English. In addition, wesearched Cardiosource, Circula-tion, and Clinicaltrials.gov, andwe hand-searched the bibliogra-phies of previous studies, relevantreviews, and previous meta-analyses to identify additional

studies not found by our initial search. We conducted andreported our meta-analysis according to guidelines describedin the PRISMA (Preferred Reporting Items for SystematicReviews and Meta-Analyses) (6) statement.Study selection. We restricted inclusion to RCTs thatcompared the efficacy and safety of CABG to that of PCIwith first-generation DES, as well as reported outcomesof mortality, myocardial infarction (MI), revascularization,or stroke. In addition, we only included trials in which themajority of patients who underwent PCI received DES.All studies not published in English were excluded.Data extraction. Data extraction was performed by 2 re-viewers, with disagreements resolved by consensus or by

a third reviewer. Extracted data included study design,enrollment period, duration of follow-up, vessel subtypes(MVD, LM disease, proximal LAD disease), and the num-ber of patients assigned to each group. Baseline participantcharacteristics included age, sex, body mass index, type2 diabetes mellitus, systemic hypertension, dyslipidemia,previous MI, mean ejection fraction, and smoking. Extractedoutcomes of interest were all-cause mortality, MI, revascu-larization, and stroke. Outcome data were extracted ascount data following an intention-to-treat approach at themaximum available follow-up for all trials but FREEDOM(Future Revascularization Evaluation in Patients WithDiabetes Mellitus: Optimal Management of MultivesselDisease), where 5-year results were extracted.Quality assessment. We used the Cochrane Collaborationtool for assessing risk of bias to determine the quality ofincluded RCTs (7). This tool assesses the risk of selectionbias, performance bias, detection bias, attrition bias,reporting bias, and other bias. Each RCT is categorized onthe basis of criteria determining the likelihood of potentialthreats to validity. Quality assessment was independentlyperformed by 2 reviewers.Data analysis. We used DerSimonian-Laird random-effectsmeta-analysis models with inverse variance weighting to cal-culate relative risks and corresponding 95% confidence in-tervals (CIs). In our primary analyses, we restricted inclusionto trials in which all patients in the PCI group received DES.In sensitivity analyses, we included RCTs in which the ma-jority of patients received DES rather than bare-metal stents.The amount of heterogeneity present was estimated usingthe I2 statistic. To examine potential sources of heterogeneity,we stratified our analyses by the following 3 types of coronarydisease: MVD, LM disease, and proximal LAD disease. Wevisually inspected funnel plots and used the Egger test to assesspublication bias. All analyses were conducted using Stata(version 11.2, Stata Corp., College Station, Texas).

Results

Search results. A total of 6,431 potentially relevant studieswere identified in our initial literature search (Fig. 1). Afterscreening the titles and abstracts of these studies, the full-length texts of 68 potentially relevant publications wereretrieved and assessed for eligibility. Of these, 7 studies metour inclusion criteria and were included in our meta-analysis.No additional studies were identified through our manualsearch of references of published studies, relevant reviews,and previous meta-analyses.Study characteristics. The earliest RCT we identifiedcomparing CABG with PCI with first-generation DES waspublished in 2005. Included studies had sample sizes rangingfrom 130 to 1,900 patients and had follow-up durationsranging from 6 to 60 months (Table 1). Three RCTs (8–10)and a subgroup analysis (11) from the SYNTAX (Synergy

Figure 1. Flow Chart

Flow chart describing the systematic literature search for randomized controlled trials comparing coronary artery bypass graft surgery and percutaneous coronaryintervention with drug-eluting stents (DES).

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 7 , N O . 5 , 2 0 1 4 Al Ali et al.

M A Y 2 0 1 4 : 4 9 7 – 5 0 6 CABG Versus PCI With DES

499

Between Percutaneous Coronary Intervention With Taxusand Cardiac Surgery) trial compared CABG with PCI withDES in patients with MVD (n ¼ 4,210). Of note, theMVD subgroup of the SYNTAX trial included only patientswith 3-vessel disease. Two RCTs (12,13) and another sub-group (14) of patients from SYNTAX examined patientswith LM disease (n ¼ 801). Two RCTs (15,16) examinedpatients with single-vessel proximal LAD disease (n ¼ 319).All included RCTs had a low risk of bias according to theCochrane criteria (Online Appendix 1).Procedure characteristics. Included RCTs varied withrespect to the chosen surgical revascularization technique(Table 1). Both trials in patients with LM disease (12,13)and 1 trial of patients with MVD (8) employed standardCABG, whereas trials of patients with proximal LADdisease (15,16) employed minimally invasive direct coro-nary artery bypass graft surgery (MIDCAB). The 2remaining trials (9,10) of patients with MVD selected thesurgical revascularization technique according to surgeons’recommendations.

All stents employed in the included trials were first-gen-eration DES, including sirolimus-eluting stents (SES) andpaclitaxel-eluting stents (PES). Bare-metal stents were usedin a very small minority of patients with MVD (n ¼ 31).Patient characteristics. Baseline patient characteristics werewell balanced across treatment groups. Most patients weremiddle-aged men, one-quarter of whom were currentsmokers (Online Appendix 2). Mean ejection fractionranged from 52.5% to 66.5%. Approximately two-thirds ofpatients had hypertension, and two-thirds were hyper-lipidemic. Among patients in both PCI and CABG groups,about one-fifth had experienced a previous MI.

Trials of patients with MVD were predominantly (76.7%)composed of patients with diabetes: 2 of the 3 largest RCTsincluded in our meta-analysis, CARDia (Coronary ArteryRevascularization in Diabetes) (8) and FREEDOM (10),both included an exclusively diabetic patient population(Online Appendix 2). Slightly over one-quarter of patientsin the SYNTAX trial had diabetes. Diabetes was present in29% of patients with LM disease and 36.3% of patientswith single-vessel proximal LAD disease.Overall meta-analysis results. Overall, when data werepooled across subgroups, CABG was superior relative toPCI for mortality (Fig. 2) and MI (Fig. 3). However,CABG, compared with PCI, resulted in increased rates ofstroke. Heterogeneity may be explained by vessel diseasesubgroup, and these overall results were driven by thestrength of beneficial effect of CABG in patients with MVD(Table 2).Meta-analysis of patient subgroups. Patients with MVD hada significantly reduced risk of all outcomes with CABGcompared with those for PCI with DES, with the exceptionof stroke, which was increased (Fig. 4). Specifically, patientswith MVD who underwent CABG had a lower risk ofmortality (Fig. 2), MI (Fig. 3), and revascularization (Fig. 5).

There was no significant difference between CABG andPCI with DES in patients with LM disease for mortality(Fig. 2) or MI (Fig. 3). However, patients with LM diseasehad a lower risk of repeat revascularization (Fig. 5) and agreater risk of stroke (Fig. 4) when undergoing CABG.

Treatment effect estimates for patients with single-vesselLAD disease were accompanied by very wide CIs. Conse-quently, no significant difference was found between pro-cedures for all outcomes.

Table 1. Characteristics of RCTs Comparing CABG to PCI with DES in Patients With MVD, LM Disease, and Proximal LAD Disease

Study, Year (Ref. #) CountryPCI(n)

CABG(n)

DES(%)

PES(%)

SES(%)

BMS(%) Type of Surgery

IMA(%) Inclusion Exclusion

Follow-Up(Months)

MVD

FREEDOM, 2012 (10) INT 953 947 100 43 51 0 Surgeons’ choice(unspecified % MIDCAB)

0 DM, MVD, eligible for bothPCI and CABG

Severe CHF, previous cardiacsurgery, ACS

60

CARDia, 2010 (8) UK 256 254 69 0 69 31 Standard CABG 94 DM, MVD, complex SVD ACS, LM previous intervention 12

LM Disease

PRECOMBAT, 2011 (13) Korea 300 300 100 0 100 0 Standard CABG 0 LM ACS, previous intervention,EF <30%

24

Boudriot, 2011 (12) Germany 100 101 100 2 98 0 Standard CABG 50 LM ACS, previous surgicalintervention, severe PVD

12

3-Vessel or LM Disease*

SYNTAX, 2009 (9) USA, Europe 903 897 100 100 0 0 Surgeons’ choice(15% MIDCAB)

0 MVD and/or LM Previous intervention, acute MI 48

Proximal LAD Artery Disease

Hong, 2005 (16) Korea 119 70 100 20 80 0 MIDCAB 37 Isolated LAD stenosis ACS, previous intervention 6

Thiele, 2009 (15) Germany 65 65 100 0 100 0 MIDCAB 50 Isolated LAD stenosis ACS, previous intervention 36

*SYNTAX included patients with 3-vessel or LM disease. Secondary analyses were conducted in which patients were stratified by type of underlying coronary disease. Where possible, we present the results of these subgroup analyses in the current meta-analysis,

which were reported at the end of follow-up (48 months).

ACS ¼ acute coronary syndrome; BMS ¼ bare-metal stent; CABG ¼ coronary artery bypass graft surgery; CARDia ¼ Coronary Artery Revascularization in Diabetes; CHF ¼ congestive heart failure; DES ¼ drug-eluting stent; DM ¼ diabetes mellitus; EF ¼ ejection

fraction; FREEDOM ¼ Future Revascularization Evaluation in Patients With Diabetes Mellitus: Optimal Management of Multivessel Disease; IMA ¼ internal mammary artery; INT ¼ international; LAD ¼ left anterior descending; LM ¼ left main; MI ¼ myocardial

infarction; MIDCAB ¼minimally invasive direct coronary artery bypass graft surgery; MVD ¼multivessel disease; PCI ¼ percutaneous coronary intervention; PES ¼ paclitaxel-eluting stent; PRECOMBAT ¼ Premier of Randomized Comparison of Bypass Surgery Versus

Angioplasty Using Sirolimus-Eluting Stent in Patients With Left Main Coronary Artery Disease; PVD¼ peripheral vascular disease; RCT¼ randomized controlled trial; SES ¼ sirolimus-eluting stent; SVD¼ single-vessel disease; SYNTAX ¼ Synergy Between Percutaneous

Coronary Intervention With Taxus and Cardiac Surgery.

AlAliet

al.JACC:CARDIO

VASCULAR

INTERVENTIO

NS,VOL.7,NO.5,2014

CABGVersus

PCIWith

DES

MAY

2014:4

97–506

500

Figure 2. Forest Plot of RCTs Comparing the Effect of CABG on All-Cause Mortality to That of PCI With DES

CABG ¼ coronary artery bypass graft surgery; CI ¼ confidence interval; DES ¼ drug-eluting stents; FREEDOM ¼ Future Revascularization Evaluation in Patients WithDiabetes Mellitus: Optimal Management of Multivessel Disease; LAD ¼ left anterior descending; LM ¼ left main; MVD ¼ multivessel disease; PCI ¼ percutaneouscoronary intervention; PRECOMBAT ¼ Premier of Randomized Comparison of Bypass Surgery Versus Angioplasty Using Sirolimus-Eluting Stent in Patients With LeftMain Coronary Artery Disease; RCT ¼ randomized controlled trial; RR ¼ risk ratio; SYNTAX ¼ Synergy Between Percutaneous Coronary Intervention With Taxus andCardiac Surgery.

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 7 , N O . 5 , 2 0 1 4 Al Ali et al.

M A Y 2 0 1 4 : 4 9 7 – 5 0 6 CABG Versus PCI With DES

501

Visual inspection of the funnel plot (Online Appendix 3)and Egger test (p ¼ 0.025) suggested the presence ofpublication bias.

Discussion

Our study was designed to compare the efficacy and safetyof CABG with that of PCI with first-generation DESacross subgroups of coronary artery vessel disease in termsof mortality, MI, revascularization, and stroke. Our resultssuggest that in patients with MVD, CABG reduces mor-tality, MI, and repeat revascularization, but increases therisk of stroke when compared with first-generation DES.In patients with LM disease, CABG was associated witha reduced risk of revascularization, with an increased riskof stroke, and no significant differences in death or MI.Available data were insufficient to draw meaningful conclu-sions regarding the relative efficacies of CABG and PCI withDES in patients with single-vessel proximal LAD disease.

As an important caveat, both the FREEDOM andCARDia trials were conducted in diabetic patients. Diabetesis a well-known risk factor for cardiovascular disease, andseveral studies have concluded that CABG is preferableto PCI in diabetics (17,18). Two of the 3 largest RCTsexamining patients with MVD included an exclusively dia-betic population; as a result, our findings showing a signif-icant benefit of CABG versus DES in MVD may be drivenby this effect. A recent study stratifying the results of theSYNTAX trial by diabetes status found that diabetic pa-tients with 3-vessel disease had a marked clinical benefitwith CABG versus PCI with DES (18). This benefit wasalso present among nondiabetic patients, albeit less pro-nounced. Among both diabetic and nondiabetic patients,the rates of all-cause death were similar with both CABGand PCI, whereas cardiac death occurred more frequentlywith PCI than with CABG (nondiabetic hazard ratio [HR]:1.62, 95% CI: 1.03 to 2.55; diabetic HR: 2.01, 95% CI:1.04 to 3.88). Both diabetic and nondiabetic patients

Figure 3. Forest Plot of RCTs Comparing the Effect of CABG on MI to That of PCI With DES

Abbreviations as in Figure 2.

Al Ali et al. J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 7 , N O . 5 , 2 0 1 4

CABG Versus PCI With DES M A Y 2 0 1 4 : 4 9 7 – 5 0 6

502

experienced higher rates of repeat revascularization after PCIrelative to CABG (diabetic HR: 2.75, 95% CI: 1.78 to 4.24,p < 0.001; nondiabetic HR: 1.82, 95% CI: 1.39 to 2.38).Nondiabetic patients had a higher rate of MI followingPCI relative to CABG (HR: 2.90, 95% CI: 1.70 to 4.70),whereas no such difference was found among diabetic pa-tients (HR: 1.62, 95% CI: 0.77 to 3.41). Neither diabeticnor nondiabetic patients had significantly different risks ofstroke across CABG or PCI. Together, these results suggestthat a protective effect conferred by CABG to patients with

Table 2. Treatment Effects Obtained From the Meta-Analysis of RCTs Comparingand Proximal LAD Disease*

Outcome Mortality

MVD 0.73 (0.56–0.96) 0.4

Unprotected LM disease 1.08 (0.75–1.57) 0.6

Single-vessel proximal LAD disease 2.34 (0.57–9.64) 1.8

Overall 0.85 (0.64–1.12) 0.5

Values are risk ratio (95% confidence interval). *Data were pooled across trials using random-effects m

Abbreviations as in Table 1.

3-vessel disease exists for both diabetic and nondiabeticindividuals.

CABG may also be beneficial for patients with highdisease complexity and severity. The 3-vessel disease patientsubgroup in the SYNTAX trial also reported that patientswith more severe disease (as determined by the SYNTAXscore [19]) who received PCI had a significantly higher riskof mortality than did patients with high scores who receivedCABG. Similarly, a recent meta-analysis of observationaland randomized studies found that, compared with patients

CABG to PCI With DES Overall and in Patients With MVD, LM Disease,

MI Revascularization Stroke

5 (0.34–0.60) 0.36 (0.26–0.51) 1.72 (1.02–2.90)

9 (0.41–1.16) 0.60 (0.46–0.78) 2.89 (1.15–7.27)

5 (0.46–7.40) 0.55 (0.05–6.24) 5.07 (0.21–122.80)

8 (0.40–0.84) 0.51 (0.39–0.67) 1.79 (1.23–2.62)

eta-analytic models.

Figure 4. Forest Plot of RCTs Comparing the Effect of CABG on Stroke to That of PCI With DES

Abbreviations as in Figure 2.

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 7 , N O . 5 , 2 0 1 4 Al Ali et al.

M A Y 2 0 1 4 : 4 9 7 – 5 0 6 CABG Versus PCI With DES

503

with low SYNTAX scores, those with high scores hadsignificantly higher rates of major adverse cardiovascularand cerebrovascular events with PCI than with CABG at12, 24, and 36 months of follow-up (20). That is, thebenefits of CABG versus PCI differ across disease severity.Additional support is lent to this reasoning based on therecent release of the 5-year follow-up data from theSYNTAX trial: after 5 years, patients with severe disease(as indicated by a high SYNTAX score) had lower mor-tality rates among the CABG treatment group, relative tothe PCI group (21). Among patients with a low or inter-mediate SYNTAX score, mortality rates were similar acrossboth treatment groups. However, patients with an inter-mediate SYNTAX score had significantly higher rates ofMI and repeat revascularization with PCI than withCABG. Overall, the 5-year SYNTAX results show thatapproximately two-thirds of patients with complex coro-nary disease derive greater benefit from CABG than fromPCI with first-generation DES, whereas the remainingpatients may derive comparable benefit from PCI. Thesefindings support the use of CABG in patients with MVD,particularly among those with complex or severe disease.

Evidence for the optimal treatment of LM disease is lessclear. Results from our meta-analysis suggest a possible

benefit favoring CABG, albeit statistically inconclusive.These inconclusive results may be due in part to the rela-tively smaller sample size of patients with LM disease.However, results from a large retrospective multinationalregistry comparing CABG and PCI with first-generationDES in patients with LM disease also showed no differencebetween procedures for death, MI, and cerebrovascular ac-cidents at a median follow-up of 1,295 days (22). Our an-alyses show that in patients with LM disease, CABG resultsin a decreased risk of revascularization (Fig. 5) and anincreased risk of stroke (Fig. 4). Increasingly, however,revascularization is not being considered as clinically sig-nificant as death, MI, or stroke (20,23). Thus, the evidencefavoring CABG over PCI with first-generation DES forpatients with LM disease is not as strong as for patientswith MVD.

Our study had insufficient power to draw clinicallydirective conclusions for patients with single-vessel proximalLAD disease. There are only 2 trials in patients withproximal LAD disease, and these had a substantially shorterduration of follow-up relative to all other included RCTs.Although the evidence comparing CABG with PCI withfirst-generation DES is limited, a previous trial comparingPCI with bare-metal stents found that patients undergoing

Figure 5. Forest Plot of RCTs Comparing the Effect of CABG on Repeat Revascularization to That of PCI With DES

Abbreviations as in Figure 2.

Al Ali et al. J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 7 , N O . 5 , 2 0 1 4

CABG Versus PCI With DES M A Y 2 0 1 4 : 4 9 7 – 5 0 6

504

PCI had a higher risk of major adverse cardiovascular eventsafter 5 years (24). However, with the exception of a higherrisk of repeat revascularization with PCI, the risk of deathand MI was similar in both groups. Although these findingssuggest PCI may be comparable to CABG in patients withproximal LAD disease, statistically inconclusive resultshighlight the need to rely on additional clinical consider-ations to aid in procedural choice.

Results from a recent study validating the SYNTAXscore II (25), an enhanced version of the SYNTAX scorewith prognostically important clinical variables, confirmedthe importance of considering clinical characteristics inweighing the benefits of CABG against those of PCI. Theinvestigators demonstrated that the low, intermediate, andhigh anatomical complexity SYNTAX score categoriesincorrectly categorized lower-risk patients in high SYNTAXscore groups and higher-risk patients in low SYNTAX scoregroups. To correct for this misclassification, the SYNTAXscore II considered additional patient factors includingage, creatinine clearance, left ventricular ejection fraction,sex, chronic obstructive pulmonary disease, and peripheralvascular disease. The presence of these clinical characteristics

modified the threshold of the SYNTAX score at which4-year mortality rates were similar between CABG andPCI. Among patients with LM disease who underwentPCI, higher anatomical SYNTAX scores were required toachieve similar long-term prognosis as for patients whoreceived CABG. By contrast, among patients with 3-vesseldisease, the SYNTAX score was representative of greaterdisease complexity than that for patients with LM diseasewith identical SYNTAX scores. Thus, patients with 3-vesseldisease would have better long-term outcomes with CABG.These findings highlight the importance of consideringall patient factors, particularly in the event where the choiceof procedure is not straightforward.

There is suggestive evidence that first- and second-gen-eration stents may have differential efficacy profiles, whichin turn could affect the optimal procedural choice betweenCABG and PCI with DES. Results from the SPIRIT IV(Clinical Evaluation of the Xience V Everolimus ElutingCoronary Stent System IV) trial (26) revealed that, at 1 yearafter PCI, patients treated with everolimus-eluting stents(EES), compared with patients treated with PES, had alower occurrence of MI (EES: 1.9%, PES: 3.1%, p ¼ 0.02),

J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 7 , N O . 5 , 2 0 1 4 Al Ali et al.

M A Y 2 0 1 4 : 4 9 7 – 5 0 6 CABG Versus PCI With DES

505

definite stent thrombosis (EES: 0.3%, PES: 1.1%,p ¼ 0.004), target lesion revascularization (EES: 2.5%,PES: 4.6%, p ¼ 0.001), target vessel revascularization(EES: 3.9%, PES: 5.9%, p ¼ 0.009), and major adversecardiovascular events (EES: 4.2%, PES: 6.9%, p < 0.001).Similarly, 1-year post-PCI, results from the XAMI (XienceV Stent vs. Cypher Stent in Primary PCI for AcuteMyocardial Infarction) trial (27) showed major adversecardiovascular event rates were significantly lower amongpatients treated with EES relative to SES (RR: 0.52, 95%CI: 0.27- to 1.00, p ¼ 0.048). Although there may beclinical differences attributable to stent type and generation,it remains unclear to what extent these affect the choicebetween second-generation DES and CABG.Study limitations. First, the included RCTs varied in studydesign, patient characteristics, procedural differences, andduration of follow-up. To account for this between-studyheterogeneity, we used random-effects models. We alsostratified our analyses by type of coronary disease to explicitlyexamine the impact of this source of heterogeneity. Second,we found evidence of publication bias, which may haveresulted in an overestimate of the protective effects ofCABG. Although this may have affected our results for LMor proximal LAD patients, it is unlikely to have biased ourMVD analyses as all trials in this patient subpopulation werelarge (i.e., >500 patients) and thus likely to be publishedregardless of statistical significance. Third, diabetic patientswere over-represented in trials conducted in patients withMVD. In addition, CABG and PCI with DES were onlycompared in patients with single-vessel proximal LADdisease in 2 relatively small RCTs to date; analyses con-ducted in this subpopulation, therefore, had only modeststatistical power. Despite the varying duration of follow-upacross studies, the absence of reported event rates necessi-tated the calculation of pooled relative risks. Among patientswith LAD disease, surgical outcomes referred solely toMIDCAB, limiting our ability to comment on the efficacyof standard CABG for LAD disease. Finally, only first-generation DES were employed in all included RCTs,limiting our ability to draw conclusions for second-genera-tion devices.

Conclusions

We found that CABG has superior efficacy when comparedto PCI with first-generation DES in patients with MVDbut results in an increased risk of stroke. Among patientswith LM disease, CABG reduces the risk of revasculariza-tion and increases the risk of stroke. Results were incon-clusive in patients with single-vessel proximal LAD disease.Our observed treatment benefits in patients with MVDbut not others is consistent with the need to consider riskprofiles and disease complexity in selecting the optimalrevascularization procedure.

AcknowledgmentThe authors thank Jennifer Reoch for her help with dataextraction and the reviewers for their thoughtful suggestions.

Reprint requests and correspondence: Dr. Mark J. Eisenberg,Divisions of Cardiology and Clinical Epidemiology, Jewish Gen-eral Hospital/McGill University, 3755 Côte Ste. Catherine Road,Suite H-421.1, Montreal, Quebec H3T 1E2, Canada. E-mail:mark.eisenberg@mcgill.ca.

REFERENCES

1. Luo Y, Yu X, Chen F, et al. Impact of diabetes mellitus on patients withunprotected left main coronary artery lesion disease treated with eitherpercutaneous coronary intervention or coronary-artery bypass grafting.Coron Artery Dis 2012;23:322–9.

2. Lee SW, Kim SH, Kim SO, et al. Comparative long-term efficacyand safety of drug-eluting stent versus coronary artery bypass graftingin ostial left main coronary artery disease: analysis of the MAIN-COMPARE registry. Catheter Cardiovasc Interv 2012;80:206–12.

3. Kawecki D, Morawiec B, Fudal M, Milejski W, Jache�c W, Nowalany-Kozielska E. Comparison of coronary artery bypass grafting withpercutaneous coronary intervention for unprotected left main coronaryartery disease. Yonsei Med J 2012;53:58–67.

4. Head SJ, Bogers A, Kappetein AP. Drug-eluting stent implantation forcoronary artery disease: current stents and a comparison with bypasssurgery. Curr Opin Pharmacol 2012;12:147–54.

5. Ahmed K, Jeong MH, Chakraborty R, et al., for the Korea AcuteMyocardial Infarction Registry Investigators. Percutaneous coronaryintervention with drug-eluting stent implantation vs. coronary arterybypass grafting for multivessel coronary artery disease in metabolicsyndrome patients with acute myocardial infarction. Circ J 2012;76:721–8.

6. Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement forreporting systematic reviews and meta-analyses of studies that evaluatehealthcare interventions: explanation and elaboration. BMJ 2009;339:b2700.

7. Higgins J, Green S, Cochrane Collaboration. Cochrane Handbook forSystematic Reviews of Interventions. Hoboken, NJ: JohnWiley & Sons,2008.

8. Kapur A, Hall RJ, Malik IS, et al. Randomized comparison of percu-taneous coronary intervention with coronary artery bypass grafting indiabetic patients: 1-year results of the CARDia (Coronary ArteryRevascularization in Diabetes) trial. J Am Coll Cardiol 2010;55:432–40.

9. Serruys PW, Morice MC, Kappetein AP, et al., for the SYNTAXInvestigators. Percutaneous coronary intervention versus coronary-arterybypass grafting for severe coronary artery disease. N Engl J Med 2009;360:961–72.

10. Farkouh ME, Domanski M, Sleeper LA, et al., for the FREEDOMTrial Investigators. Strategies for multivessel revascularization in pa-tients with diabetes. N Engl J Med 2012;367:2375–84.

11. Mohr F, Redwood S, Venn G, et al. Final five-year follow-up of theSYNTAX trial: optimal revascularization strategy in patients with three-vessel disease (abstr). J Am Coll Cardiol 2012;60:B13.

12. Boudriot E, Thiele H, Walther T, et al. Randomized comparison ofpercutaneous coronary intervention with sirolimus-eluting stents versuscoronary artery bypass grafting in unprotected left main stem stenosis.J Am Coll Cardiol 2011;57:538–45.

13. Park SJ, Kim YH, Park DW, et al. Randomized trial of stents versusbypass surgery for left main coronary artery disease. N Engl J Med 2011;364:1718–27.

14. Serruys P, Religa G, Ruzyllo W, et al. Long-term safety and efficacy ofpercutaneous coronary intervention with stenting and coronary arterybypass surgery for left main disease: final five-year follow-up of theSYNTAX trial (abstr). J Am Coll Cardiol 2012;60:B13.

15. Thiele H, Neumann-Schniedewind P, Jacobs S, et al. Randomizedcomparison of minimally invasive direct coronary artery bypass surgery

Al Ali et al. J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S , V O L . 7 , N O . 5 , 2 0 1 4

CABG Versus PCI With DES M A Y 2 0 1 4 : 4 9 7 – 5 0 6

506

versus sirolimus-eluting stenting in isolated proximal left anteriordescending coronary artery stenosis. J Am Coll Cardiol 2009;53:2324–31.

16. Hong SJ, Lim DS, Seo HS, et al. Percutaneous coronary interventionwith drug-eluting stent implantation vs. minimally invasive directcoronary artery bypass (MIDCAB) in patients with left anteriordescending coronary artery stenosis. Catheter Cardiovasc Interv 2005;64:75–81.

17. Ellis SG. Coronary revascularization for patients with diabetes: updateddata favor coronary artery bypass grafting. J Am Coll Cardiol 2013;61:817–9.

18. Kappetein AP, Head SJ, Morice MC, et al., for the SYNTAX In-vestigators. Treatment of complex coronary artery disease in patientswith diabetes: 5-year results comparing outcomes of bypass surgery andpercutaneous coronary intervention in the SYNTAX trial. Eur J Car-diothorac Surg 2013;43:1006–13.

19. Serruys PW, Onuma Y, Garg S, et al., for the ARTS II Investigators. 5-year clinical outcomes of the ARTS II (Arterial RevascularizationTherapies Study II) of the sirolimus-eluting stent in the treatment ofpatients with multivessel de novo coronary artery lesions. J Am CollCardiol 2010;55:1093–101.

20. Alam M, Huang HD, Shahzad SA, et al. Percutaneous coronaryintervention vs. coronary artery bypass graft surgery for unprotectedleft main coronary artery disease in the drug-eluting stents eradanaggregate data meta-analysis of 11,148 patients. Circ J 2013;77:372–82.

21. Mohr FW, Morice MC, Kappetein AP, et al. Coronary artery bypassgraft surgery versus percutaneous coronary intervention in patients withthree-vessel disease and left main coronary disease: 5-year follow-up ofthe randomised, clinical SYNTAX trial. Lancet 2013;381:629–38.

22. Chieffo A, Meliga E, Latib A, et al. Drug-eluting stent for left maincoronary artery disease: the DELTA registry: a multicenter registryevaluating percutaneous coronary intervention versus coronary artery

bypass grafting for left main treatment. J Am Coll Cardiol Intv 2012;5:718–27.

23. Jang JS, Choi KN, Jin HY, et al. Meta-analysis of three randomizedtrials and nine observational studies comparing drug-eluting stentsversus coronary artery bypass grafting for unprotected left main coronaryartery disease. Am J Cardiol 2012;110:1411–8.

24. Thiele H, Oettel S, Jacobs S, et al. Comparison of bare-metal stentingwith minimally invasive bypass surgery for stenosis of the left anteriordescending coronary artery: a 5-year follow-up. Circulation 2005;112:3445–50.

25. Farooq V, van Klaveren D, Steyerberg EW, et al. Anatomical andclinical characteristics to guide decision making between coronary arterybypass surgery and percutaneous coronary intervention for individualpatients: development and validation of SYNTAX. score II. Lancet2013;381:639–50.

26. Stone GW, Rizvi A, Newman W, et al. Everolimus-eluting versuspaclitaxel-eluting stents in coronary artery disease. N Engl J Med 2010;362:1663–74.

27. Hofma SH, Brouwer J, Velders MA, et al. Second-generationeverolimus-eluting stents versus first-generation sirolimus-eluting stentsin acute myocardial infarction: 1-year results of the randomized XAMI(XienceV Stent vs. Cypher Stent in Primary PCI for Acute MyocardialInfarction) trial. J Am Coll Cardiol 2012;60:381–7.

Key Words: coronary artery bypass graft - drug-elutingstent(s) - meta-analysis.

APPENDIX

For supplemental information, please see the online version of this paper.