Is off-pump coronary artery bypass grafting superior to drug-eluting stents for the treatment of coronary artery disease? A meta-analysis of randomized and nonrandomized studies.

IJCA-18083; No of Pages 14 International Journal of Cardiology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard

Is off-pump coronary artery bypass grafting superior to drug-eluting stents for the treatment of coronary artery disease? A meta-analysis of randomized and nonrandomized studies☆ Di Lu a,1, Ximing Nie b, Jun Wan a, Shengping He a, Songlin Du a, Zhen Zhang a, Zhenkang Wang a, Wujun Wang a,⁎ a b

Department of Cardiovascular and Thoracic Surgery, Nanfang Hospital, Southern Medical University, Guangzhou 510515, China Department of Neurology, Nanfang Hospital, Southern Medical University, Guangzhou 510515, China

a r t i c l e

i n f o

Article history: Received 12 September 2013 Received in revised form 15 March 2014 Accepted 17 April 2014 Available online xxxx Keywords: Coronary artery disease Off-pump coronary artery bypass grafting Drug-eluting stent Meta-analysis Randomized controlled trial Observational cohort

a b s t r a c t Background: As drug-eluting stent (DES) has almost overcome the disadvantage of frequent restenosis, off-pump coronary artery bypass grafting (OPCAB) has been introduced to avoid complications of cardiopulmonary bypass. However, which approach may promise better outcomes for patients with coronary artery disease remains controversial. Methods: Three databases were searched. The outcomes of interest were major adverse cardiac and cerebrovascular events (MACCE), all-cause death, target vessel revascularization (TVR), repeat revascularization (RRV), myocardial infarction (MI), and cerebrovascular events (CVE). The relative risk (RR) was calculated as the summary statistic. Results: 11,452 patients from 22 studies were included, of which 4949 patients underwent OPCAB and 6503 patients received DES. The cumulative rates of MACCE (RR [95% CI] = 0.43 [0.34, 0.54], P b 0.00001), all-cause death (RR [95% CI] = 0.56 [0.33, 0.96], P = 0.03), TVR (RR [95% CI] = 0.33 [0.21, 0.53], P b 0.00001), RRV (RR [95% CI] = 0.22 [0.11, 0.42], P b 0.00001) and MI (RR [95% CI] = 0.13 [0.05, 0.29], P b 0.00001) at 3 years were all lower in OPCAB group. The incidences of in-hospital death (RR [95% CI] = 1.31 [0.81, 2.13], P = 0.27) and MI (RR [95% CI] = 1.03 [0.60, 1.78], P = 0.92) were not different between groups, but the rate of in-hospital CVE was lower (RR [95% CI] = 2.6355 [1.0033, 6.9228], P = 0.05) in DES group. Conclusions: OPCAB presents better long-term outcomes of MACCE, all-cause mortality, TVR, RRV and MI but uncertain outcome of postoperative CVE without influencing the incidences of in-hospital death and MI. © 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction According to the World Health Organization (WHO), there're more than 7 million deaths each year attributed to coronary artery disease (CAD) worldwide [1]. Each year in America, 635,000 people will have a new coronary attack, 280,000 will have a recurrent attack, and an additional 150,000 silent first myocardial infarctions will occur, which means approximately every 34 s, an American will have a coronary event, and approximately every minute, someone will die of one [2]. On the other hand, it's estimated that 20,000–40,000 individuals of

☆ The authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation. ⁎ Corresponding author at: Department of Cardiovascular and Thoracic Surgery, Nanfang Hospital, Southern Medical University, 1838 N, Guangzhou Avenue, Guangzhou 510515, China. Tel.: +86 20 61641821. E-mail address: [email protected] (W. Wang). 1 Di Lu contributed to the manuscript as the first author.

the population per million suffer from angina in most European countries [3]. Coronary artery bypass grafting (CABG) still remains superior to percutaneous coronary intervention (PCI) in treating stable ischemic heart disease according to guidelines written by the American Heart Association (AHA) [4] and the European Society of Cardiology (ESC) [3] because of their different levels of evidence. However, PCI has become the key treatment for ST-elevation myocardial infarction (STEMI) according to the guidelines [5,6], and urgent CABG rather than PCI is only recommended when patients with STEMI and coronary anatomy not amenable to PCI have ongoing or recurrent ischemia, cardiogenic shock, severe heart failure (HF), other high-risk features, or mechanical defects, mainly due to its need for more preoperative preparation [5,6]. Ever since the technology of percutaneous transluminal coronary angioplasty (PTCA) was introduced in 1977 [7], the high rate of restenosis has always been its biggest problem that kept itself from replacing CABG in treating CAD, which in turn encouraged the development of bare metal

http://dx.doi.org/10.1016/j.ijcard.2014.04.182 0167-5273/© 2014 Elsevier Ireland Ltd. All rights reserved.

Please cite this article as: Lu D, et al, Is off-pump coronary artery bypass grafting superior to drug-eluting stents for the treatment of coronary artery disease? A meta-a..., Int J Cardiol (2014), http://dx.doi.org/10.1016/j.ijcard.2014.04.182

2

D. Lu et al. / International Journal of Cardiology xxx (2014) xxx–xxx

stent (BMS) [8] and drug eluting stent (DES) [9] thereafter. According to the latest outcome of the SYNTAX trial, PCI with the first-generation paclitaxel-eluting stent (PES) seems to be an acceptable alternative for patients with less complex disease (low SYNTAX scores) or left main coronary disease (low or intermediate SYNTAX scores) [10]. As it's minimally invasive and more easily to be accepted by patients at almost the same curative effect level when compared with the conventional CABG, the latter one seems to be challenged. On the other hand, especially for the most benefits of the high-risk people, in order to avoid the complications of cardiopulmonary bypass (CPB) which includes the metabolic acidosis, electrolyte disturbances, coagulation dysfunction, brain damage, cognitive abnormalities, kidney injury and so on, off-pump coronary artery bypass grafting (OPCAB) was first introduced in 1960s [11]. With the development of the minimally invasive technology for surgery, minimally invasive CABG without CPB using an anterolateral minithoracotomy [12–14] and total endoscopic computer enhanced CABG [15,16] were developed one after another. Then reasonably, comparison of the novel OPCAB and the conventional on-pump CABG has become one of the most attractive research topics in the area of therapeutic strategy for CAD. It was demonstrated that OPCAB might reduce the incidence of post-operative stroke compared with the conventional CABG [17], especially in high-risk patients [18], while not significantly reduce the incidence of short-term [17,18] and long-term [18] all-cause mortality and post-operative myocardial infarction [17]. By this taken, there seems to be a trend that OPCAB may take the place of the conventional CABG or at least be another effective alternative. However, there have been few registered clinical trials comparing OPCAB and PCI with DES for treating patients with CAD. To date, several observational studies have been done, which aimed specifically to compare these two different means, but no consensuses on the outcomes of OPCAB and DES have been established to our knowledge. Furthermore, according to the results of the SYNTAX trial mentioned above, the rates of death and myocardial infarction were similar between CABG and PES at 12 months, while stroke was significantly more likely to occur with CABG (85% on-pump) [19]. Considering the comparison results of on-pump and off-pump CABG mentioned above, whether the incidence of post-operative stroke in OPCAB treated patients is similar to that in DES treated patients remains unclear.

2.3. Outcomes and data extraction The outcomes of interest included target vessel revascularization (TVR), repeat revascularization (RRV), major adverse cardiac and cerebrovascular events (MACCE), all-cause death, cardiac death (also called as cardiovascular death), myocardial infarction (MI) and cerebrovascular events (CVE). TVR was defined as any repeat revascularization performed on initially treated vessels including target lesion revascularization (TLR) and new lesion revascularization within the target vessels, while RRV consisted of TVR and revascularization performed on new lesions of non-target vessels (non-TVR). MACCE included death of any cause, nonfatal MI, CVE, and repeat revascularization by percutaneous intervention or surgery. Time points for analysis were in-hospital, 30 days, 12 months, 2 years, 3 years and 5 years, as there were few studies reporting outcomes at 4 years. Subgroup analyses divided by the design of studies were performed. Two investigators (L.D. and N.X.) reviewed all the references achieved through literature search independently for eligible studies at the level of title and/or abstract, and documented disagreements were solved through discussion with a third reviewer (W.W.). In addition, the related meta-analyses and reviews within the search results were further investigated at full-text level to achieve the possible hidden data. The details of outcomes at established time points from these eligible studies were extracted by using a certain kind of spreadsheet that we developed to improve the efficiency of data extraction and avoid possible mistakes. Besides the outcomes, basic information was extracted as follows: first author, affiliation, published date, patient enrollment period, co-morbid conditions, CAD types, mean age, intervention details and follow-up duration. 2.4. Quality assessment Methodological quality of the included studies was assessed independently by the two investigators mentioned above. For non-randomized cohort studies, the Newcastle– Ottawa scale (NOS) [20], a rating system consisted of three domains: selection, comparability and outcome, was used. Total score achieved from each section (selection 0–4, comparability 0–2, and outcome 0–3) ranged from 0 star to 9 stars, which was positively correlated with the study's quality. Studies awarded with more than 5 stars were considered to be of acceptable quality. See the detailed scores of the included studies in the table from Appendix B. Quality evaluation of the included randomized controlled trials (RCTs) was performed according to the Cochrane Collaboration's tool for assessing risk of bias (5.1.0) [21] with the following methodological items: random of sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting and other potential source of bias. Each item was classified as low risk, high risk, or unclear, which determined the general quality when taken together. The risk of bias graph and summary was presented in the figure from Appendix C. Disagreements were also resolved through discussion with the third reviewer. 2.5. Statistical analysis

2. Methods 2.1. Strategy for literature search To achieve eligible evidence, MEDLINE (PubMed interface), Embase and the Cochrane Library were searched without any restrictions on publication status, type, date or language. The following terms were used in our search: “off(−)pump”, “OPCAB”, “off-pump bilateral internal thoracic arterial grafting”, “BITA”, “minimally invasive direct coronary artery bypass”, “MIDCAB”, “beating heart surgery”, “octopus”, “off-pump internal thoracic artery”, “without extra-corporeal circulation”, “drug(−)eluting stent(s)”, “DES(s)”, “sirolimus-eluting stent”, “SES”, “paclitaxel-eluting stent”, “Cypher”, “Taxus”, “everolimus-eluting stent”, “EES”, “limus-eluting stent”, “LES”, “Resolute zotarolimus-eluting stent”, “R-ZES”, “Endeavor zotarolimus-eluting stent”, “E-ZES”, “biolimus A9-eluting stent”, “novolimus-eluting stent”, “pimecrolimus-eluting stent”, “cobalt-chromium everolimus eluting stent”, “CoCr-EES”, “Xience V”. The last search was conducted on August 24th, 2013. Details for search strategy were presented in Appendix A.

2.2. Study selection The major inclusion criteria were as follows: (i) adult patients diagnosed with singleor multi-vessel CAD suitable for revascularization with either OPCAB or DES; (ii) assessing OPCAB versus DES specifically, or CABG versus PCI with outcome details of both OPCAB and DES; (iii) reporting at least one pertinent clinical outcome of short-term, mid-term or long-term follow-up; and (iv) containing original data sufficient for calculating the hazard ratio (HR) or P value. The included studies should be published in English, while with no restrictions on publication types. Studies were excluded according to the following criteria: (i) not reporting the outcomes of both OPCAB and DES simultaneously; (ii) irretrievable or insufficient data for statistical analysis; (iii) duplication; and (iv) unavailable full text of original articles.

Outcomes of interest in this study were all treated as dichotomous variables, and the relative risk (RR) and 95% confidence interval (95% CI) were used as the summary statistics to combine the incidence of these outcomes at each time point, respectively. Heterogeneity was measured by using the chi-squared test and calculating the I 2 statistic, which estimates the percentage of total variation across studies that is due to heterogeneity rather than chance. Significant heterogeneity was considered when Phet ≤ 0.1 and I2 N 50%. In order to identify sources of heterogeneity, sensitivity analyses were performed by deselecting studies one by one to detect the influence of each study on I2 and the pooled RR. The fixed effects model of Mantel–Haenszel method was preferred to calculate the pooled RR with acceptable heterogeneity (I 2 ≤ 50%), while the random effects model was used with substantial heterogeneity. Forrest plots were presented for overall impression of the contribution of each study and the pooled statistic. Funnel plots were generated to evaluate publication bias visually and Egger's test was used for statistical assessment. We were unable to conduct specific analyses considering confounding factors because original data were unavailable. All P values were 2sided, of which except Phet the significance level was set at 0.05. Outcome combining analyses and related plots were managed by Review Manager version 5.2.6 (Cochrane Collaboration) and Egger's test was performed with Stata/SE version 12.0 (Stata Corp LP). This meta-analysis was conducted in compliance with recommendations from PRISMA (preferred reporting items for systematic reviews and meta-analyses) [22].

3. Results 3.1. Eligible studies As shown in Fig. 1, a total of 582 records were yielded by our search strategy. After eliminating 128 duplicate records, 454 records were screened at the level of title and abstract. Then 46 publications including



3

Fig. 1. Flow chart presenting study selection.

31 research records and 15 related reviews and meta-analyses were selected and perused for eligibility and identifying other valuable studies possibly hidden beyond the search results. Finally, 22 eligible studies [23–44] were included for the following quantitative synthesis, including 20 observational cohorts (OC) [23–31,33–36,38–44] and 2 RCTs [32,37], which met the inclusion criteria and had acceptable quality. Among these studies, 5 were published as conference abstracts [25,29, 35,36,38], while the other 17 were journal articles [23,24,26–28, 30–34,37,39–44]. When taken together, 11,452 patients were included, of which 4949 patients underwent OPCAB and 6503 patients underwent PCI with DES. 675 patients from 2 studies [26,42] underwent left main coronary artery (LMCA) revascularization, 2476 patients from 8 studies [23,25,27–29,32,37,38] underwent left anterior descending artery (LADA) revascularization and 2518 patients from 7 studies [24,35,36,39,40,43,44] underwent revascularization for multivessel disease (MVD), while the remaining 5 studies [30,31,33,34, 41] didn't supply detailed data of subgroups divided by type of CAD. Among the included population, more than 2509 patients had co-morbid diabetes mellitus. The first generation DES, including SES and PES, was used in at least 15 studies [23–25,27,28,30–32,34, 36,37,39,40,43,44], while the PCI in 3 studies was performed with the second generation DES [25,38,41]. Among the included studies, as there were a few studies with the same first authors, or performed in the same centers, or even with

the same enrollment periods, the detailed characteristics of each study were summarized in Table 1 to identify duplicates. However, when analyzing a particular outcome, among the similar studies applying available data, only the one with the most patients would be considered eligible to avoid the possible overlapping of data and make the best of what we got. Because Sata et al. [36] reported outcomes of OPCAB versus DES in two different cohorts grouped by age, we divided this conference report into 2 studies, documented as Sata 2009 and Sata 2009a, for the conveniences of the following analysis. 3.2. Clinical outcomes 3.2.1. MACCE As presented in Fig. 2, 14 OC studies [24–29,35,36,39–44] were involved for analyzing MACCE. The cumulative incidences of MACCE in hospital (RR [CI] = 1.55 [0.96, 2.49], P = 0.07; I2 = 40%, Phet = 0.15), at 30 days (RR [CI] = 0.80 [0.22, 2.95], P = 0.74; I2 = 0%, Phet = 0.48) and at 12 months (RR [CI] = 0.53 [0.27, 1.03], P = 0.06; I2 = 81%, P het = 0.001) of OPCAB and DES were not significantly different. However, the cumulative rates of MACCE in OPCAB group were lower at 2 years (RR [CI] = 0.40 [0.30, 0.54], P b 0.00001; I2 = 0%, Phet = 0.71) and 3 years (RR [CI] = 0.43 [0.34, 0.54], P b 0.00001; I2 = 27%, Phet = 0.25). Although there was no significant difference between the two groups for cumulative MACCE at 5 years (RR [CI] =


4

OPCAB

DES

Median follow-up period

Completeness

Reference type

166

MIDCAB

SES

22.5 m

100%

Journal

200

200

SES

12 m

NA

Journal

NA

386

188

Off-pump BITA OPCAB

SES

18 m

NA

Journal

66 ± 9/63 ± 9 59.7 ± 10.2/63.6 ± 9.3 67.5 ± 9.7 (CABG)/ 63.6 ± 10.3 63/62

218 463 163

218 463 163

OPCAB MIDCAB OPCAB

SES, PES NA NA

NA NA NA

NA NA 100%

Journal Conference Journal

308

308

MIDCAB

SES, PES

30 ± 27 m/24 ± 10 m

99.4%/98.1%

Journal

63.4 ± 12/62.4 ± 12

456

456

MIDCAB

SES, PES

70.6 m/67.5 m

99.6%/98.1%

Journal

204 350 2854

204 350 2854

MIDCAB MIDCAB OPCAB

EES, CoCr-EES NA NA

NA 32 ± 17 m/28 ± 10 m NA

NA NA NA

Conference Conference Journal

Study ID

Design

Extent of CAD

Enrollment period (OPCAB/DES)

Age (mean ± SD) (OPCAB/DES)

Ben-Gal 2006

OC

2002.05–2003.11

NA

376

Herz 2004

OC

2002.06–2003.06

NA

Herz 2005

OC

2002.06–2003.06

Briguori 2007 Buszman 2011 Chieffo 2006

OC OC OC

LADA and/or MVD LAD and/or MVD LAD and/or MVD MVD (DM) P-LAD LMCA

2002.04–2004.12 2004–2009 2002.03–2004.07

Etienne 2009

OC

LAD

Etienne 2013

OC

LAD

Ungureanu 2013 Glineur 2009 Mack 2008

OC OC OC

Moshkovitz 2005

OC

Sasaki 2012

OC

Sata 2009

OC

Sata 2009a

OC

Yamagata 2010 Yang 2008

OC OC

P-LAD LAD LAD and/or MVD LAD and/or MVD LMCA and/or MVD MVD (aged 65–75) MVD (age ≥ 75) MVD (DM) MVD

Yi 2008 Yi 2012a Yi 2012 Yi 2013

OC OC OC OC

Hong 2005 Thiele 2009

RCT RCT

MVD MVD LMCA LAD and/or MVD P-LAD P-LAD

1997.04–2006.07/ 2002.06–2005.09 1997.04–2011.02/ 2002.06–2005.10 NA NA 2004.02.01–2004.07.31

Patients enrolled (OPCAB/DES)

NA 63 ± 13/63 ± 10 NA

Patients included (OPCAB/DES)

2002.06–2003.12

NA

522

232

OPCAB

SES

12 m

NA

Journal

2005.01–2010.12

82.6 ± 2.7/82.9 ± 3.2

262

262

OPCAB

NA

35.1 m

NA

Conference

2004–2008

NA

383

383

OPCAB

SES

2.8 y

NA

Conference

2004–2008

NA

189

189

OPCAB

SES

2.8 y

NA

Conference

2004.08–2006.12 2003.03–2005.03

208 750

208 750

OPCAB OPCAB

SES NA

508 508 512 1821

388 290 256 1294

OPCAB OPCAB OPCAB OPCAB

100% 96.2%(CABG)/ 95.2% 98.50% 97.90% 97.70% 98%

Journal Journal Journal Journal

2003.03–2003.11 2003.01–2007.10

61.4 ± 9.9/60.5 ± 9.6 66/66

189 130

189 130

SES, PES NA NA EES, CoCr-EES, E-ZES SES, PES SES

42 ± 8 m 23.1 ± 8.8 m (CABG)/ 21.1 ± 6.7 m 621.1 ± 259.9 d 58.8 ± 11.5 m 38 ± 20 m 23.0 ± 13.0 m

Journal Journal

2003.07–2005.06 2003.07–2005.06 2003.07–2007.06 2008.01–2011.12

67 ± 7/70 ± 9 63 ± 8 (CABG)/ 63 ± 10 61.9 ± 9.2/63.3 ± 62.7 ± 9.3/62.6 ± 65.0 ± 8.6/62.3 ± 65.4 ± 8.6/64.4 ±

NA 43 m

100% 100%

Journal Journal

10.3 10 10.7 10.0

MIDCAB MIDCAB

RCT = randomized controlled trial, OC = observational cohort, OPCAB = off-pump coronary artery bypass grafting, BITA = bilateral internal thoracic arterial grafting, MIDCAB = minimally invasive direct coronary artery bypass, CABG = coronary artery bypass grafting (including both off-pump and on-pump), SES = sirolimus-eluting stent, PES = paclitaxel-eluting stent, EES = everolimus-eluting stent, E-ZES = Endeavor zotarolimus-eluting stent, CoCr-EES = cobaltchromium everolimus eluting stent, LMCA = left main coronary artery, LADA = left anterior descending artery, MVD = multi-vessel disease, DM = diabetes mellitus, d = day, m = month, y = year, NA = not available, SD = standard deviation.



Table 1 Characteristics of studies included in this meta-analysis.


5

Fig. 2. Forrest plots of the relative risk (RR) of cumulative major adverse cardiac and cerebrovascular events (MACCE) after off-pump coronary artery bypass grafting (OPCAB) versus drugeluting stenting (DES) at different time points: RR of in-hospital MACCE (panel A), RR of MACCE at 30 days (panel B), RR of MACCE at 12 months (panel C), RR of MACCE at 2 years (panel D), RR of MACCE at 3 years (panel E), and RR of MACCE at 5 years (panel F).


6


0.60 [0.35, 1.03], P = 0.06; I2 = 88%, Phet b 0.0001), the trend of fewer MACCE in OPCAB couldn't be ignored in long-term follow-up, when taking the aforementioned results together. 3.2.2. Mortality All-cause mortalities of the two groups were quantitatively compared with the data extracted from 17 related studies, including 16 OCs [23,24,26–30,33–36,39,41–44] and a RCT [32] according to Fig. 3. In-hospital mortalities of OPCAB and DES groups were not significantly different in the subgroups of OC (RR [CI] = 1.34 [0.82, 2.19], P = 0.24; I2 = 0%, P het = 0.78) and RCT (RR [CI] = 0.85 [0.08, 9.20], P = 0.89), and remained similar when taken together (RR [CI] = 1.31 [0.81, 2.13], P = 0.27; I2 = 0%, Phet = 0.85). Likewise, the cumulative mortality at 12 months (RR [CI] = 1.43 [0.71, 2.86], P =0.31; I2 = 0%, Phet = 0.53) and 2 years (RR [CI] = 1.20 [0.61, 2.36], P =0.60; I2 = 0%, Phet = 0.41) showed no statistical difference between the two groups. However, the cumulative rate of allcause death at 3 years in OPCAB group was lower than that in DES group (RR [CI] = 0.56 [0.33, 0.96], P = 0.03; I2 = 0%, Phet = 0.54). And there was a trend of lower all-cause mortality (RR [CI] = 0.94 [0.71, 1.25], P = 0.67; I2 = 22%, Phet = 0.28) and cardiac mortality (RR [CI] = 0.61 [0.35, 1.07], P =0.08; I2 = 0%, Phet = 0.59) at 5 years in OPCAB group without statistical difference. 3.2.3. TVR and RRV A total of 13 OCs [23,24,26–29,34,38,39,41–44] and 2 RCTs [32,37] were selected for evaluating TVR, while 9 OCs [23,24,26,28–30,34,36, 44] and 1 RCT [37] for RRV. In Fig. 4, no statistical differences were detected when comparing the in-hospital TVR rate of OPCAB and DES in not only the subgroups of OC (RR [CI] = 2.76 [0.69, 11.14], P = 0.15; I2 = 26%, Phet = 0.26) and RCT (RR [CI] = 0.85 [0.16, 4.52], P = 0.85), but also the total group (RR [CI] = 1.72 [0.62, 4.81], P = 0.30; I2 = 18%, Phet = 0.30). For TVR at 12 months, there was significant difference between OPCAB and DES in the subgroup of OC (RR [CI] = 0.29 [0.15, 0.56], P = 0.0002; I2 = 0%, Phet = 0.72), but no significant difference in the subgroup of RCT (RR [CI] = 0.11 [0.01, 2.11], P = 0.14). However, the overall TVR rate at 12 months was still lower in the OPCAB group (RR [CI] = 0.28 [0.15, 0.52], P b 0.0001; I 2 = 0%, P het = 0.79). Likewise, the cumulative incidences of TVR at 2 years (RR [CI] = 0.17 [0.09, 0.30], P b 0.00001; I 2 = 0%, P het = 0.81), 3 years (RR [CI] = 0.33 [0.21, 0.53], P b 0.00001; I 2 = 0%, P het = 0.99) and 5 years (RR [CI] = 0.26 [0.18, 0.39], P b 0.00001; I 2 = 36%, P het = 0.21) in OPCAB group were lower when compared with DES group. On the other hand, the rates of RRV at 12 months (RR [CI] = 0.26 [0.15, 0.45], P b 0.00001; I2 = 0%, Phet = 0.57), 2 years (RR [CI] = 0.39 [0.28, 0.54], P b 0.00001; I 2 = 0%, P het = 0.55) and 3 years (RR [CI] = 0.22 [0.11, 0.42], P b 0.00001; I 2 = 0%, P het = 0.83) were significantly lower in OPCAB group (data not shown). Similarly, for the cumulative incidence of RRV at 12 months, OPCAB showed better outcomes with statistical significance in the subgroup of OC (RR [CI] = 0.28 [0.16, 0.48], P b 0.00001; I2 = 0%, Phet = 0.47), but not in the subgroup of RCT (RR [CI] = 0.12 [0.01, 2.15], P = 0.15). 3.2.4. MI There were 11 OCs [23,24,26,27,29,34,36,39–41,44] and 2 RCTs [32,37] included for analyzing the incidence of MI, which were shown in Fig. 5. There showed no significant differences in the overall rates of in-hospital MI (RR [CI] = 1.03 [0.60, 1.78], P = 0.92; I2 = 0%, Phet = 0.55), MI at 30 days (RR [CI] = 1.69 [0.49, 5.78], P = 0.40; I2 = 25%, Phet = 0.26; not shown in Fig. 5) and at 12 months (RR [CI] = 1.37 [0.80, 2.34], P = 0.25; I2 = 0%, Phet = 0.41) between OPCAB and DES. When separated into subgroups, OCs also didn't present statistical differences of in-hospital MI (RR [CI] = 1.05

[0.59, 1.88], P = 0.86; I2 = 0%, Phet = 0.42), MI at 30 days (RR [CI] = 0.52 [0.07, 3.63], P = 0.51; I2 = 0%, Phet = 0.55) and at 12 months (RR [CI] = 1.17 [0.66, 2.06], P = 0.59; I2 = 0%, Phet = 0.57) between the two groups, nor did RCTs (in-hospital: RR [CI] = 0.85 [0.16, 4.52], P = 0.85; 30 days: RR [CI] = 5.71 [0.65, 50.39], P = 0.12; 12 months: RR [CI] = 5.33 [0.64, 44.32], P = 0.12). Though without significant difference, there seemed to be a trend of lower incidence of MI in OPCAB group at 2 years (RR [CI] = 0.40 [0.16, 1.02], P = 0.05; I2 = 0%, Phet = 1.00). However, OPCAB showed a better outcome of cumulative MI at 3 years than DES (RR [CI] = 0.13 [0.05, 0.29], P b 0.00001; I2 = 0%, Phet = 0.54). 3.2.5. CVE According to Fig. 6, there were 10 OCs [24,26,27,29,34,36,38,39, 42,44] and 1 RCT [32] used for the quantitative analysis of CVE. When analyzing the incidence of in-hospital CVE, though there showed no significant differences between OPCAB and DES in both the subgroup of OC (RR [CI] = 2.47 [0.89, 6.84], P = 0.08; I2 = 0%, P het = 0.72) and the subgroup of RCT (RR [CI] = 5.07 [0.21, 122.80], P = 0.32), DES presented a better outcome in the total group with significant difference (RR [CI] = 2.6355 [1.0033, 6.9228], P = 0.05; I 2 = 0%, Phet = 0.79). On the other hand, there were no statistical differences between the two groups for the cumulative rates of CVE at 12 months (RR [CI] = 2.29 [0.69, 7.54], P = 0.17; I 2 = 31%, Phet = 0.24), 2 years (RR [CI] = 1.31 [0.39, 4.45], P = 0.66; I2 = 0%, Phet = 0.50) and 3 years (RR [CI] = 1.16 [0.55, 2.43], P = 0.70; I2 = 0%, Phet = 0.41). 3.3. Heterogeneity analysis By using the statistical method we mentioned above, significant heterogeneity across studies was found in the meta-analyses of MACCE incidences at 12 months and 5 years. No single study within each analysis seemed to be responsible for the significant heterogeneity by deselecting studies one by one. Therefore, the random effects model was used and the pooled RRs were affected. With the application of fixed effects model, there were significant differences in both the analyses of MACCE incidences at 12 months (RR [CI] = 0.54 [0.41, 0.71], P b 0.0001; I 2 = 81%, P het = 0.001) and 5 years (RR [CI] = 0.59 [0.50, 0.71], P b 0.00001; I 2 = 88%, P het b 0.0001), but statistical significance no longer existed after converting to random effects model (MACCE at 12 months: RR [CI] = 0.53 [0.27, 1.03], P = 0.06; I 2 = 81%, P het = 0.001; MACCE at 5 years: RR [CI] = 0.60 [0.35, 1.03], P = 0.06; I 2 = 88%, Phet b 0.0001). For the analysis of MACCE at 12 months, 4 studies were included [24,26,40,44]. Chieffo 2006 focused on LMCA revascularization [26], Brigouri 2007 focused on the diabetic people with MVD [24], while Yang 2008 and Yi 2008 studied MVD revascularization [40,44]. On the other hand, when only combining the results of Yang 2008 and Yi 2008, RR [CI] equaled 0.29 [0.17, 0.49] with P b 0.0001 and I 2 equaled 0% with P het = 0.50. Hence, the population baseline of each study might be the main source of heterogeneity. For the combined results of MACCE at 5 years, also 4 studies were involved [27,35,42,43]. Etienne 2013 studied LAD revascularization, Sasaki 2012 focused on patients with MVD and LMCA disease, Yi 2012 studied LMCA revascularization, and Yi 2012a focused on MVD. In that way, it seemed hard to perform subgroup analysis divided by the type of CAD. 3.4. Sensitivity analysis As we described above, sensitivity analysis was performed by deselecting studies one by one alphabetically. The pooled RRs from analyses of in-hospital MACCE, cumulative MACCE at 12 months and



7

Fig. 3. Forrest plots of the relative risk (RR) of cumulative death after off-pump coronary artery bypass grafting (OPCAB) versus drug-eluting stenting (DES) at different time points: RR of in-hospital death (panel A), RR of all-cause death at 12 months (panel B), RR of all-cause death at 2 years (panel C), RR of all-cause death at 3 years (panel D), and RR of all-cause death at 5 years (panel E).


8


Fig. 4. Forrest plots of the relative risk (RR) of cumulative target vessel revascularization (TVR) after off-pump coronary artery bypass grafting (OPCAB) versus drug-eluting stenting (DES) at different time points: RR of in-hospital TVR (panel A), RR of cumulative TVR at 12 months (panel B), RR of cumulative TVR at 2 years (panel C), RR of cumulative TVR at 3 years (panel D), and RR of cumulative TVR at 5 years (panel E).



5 years, death at 3 years and in-hospital CVE were influenced by removing certain study data. In the analysis of cumulative MACCE at 12 months, the study of Chieffo 2006 [26] compared the outcomes of DES versus CABG and the subgroup data of OPCAB were extracted in our study, which may cause different baseline characteristics of the selected population and different outcomes of MACCE at 12 months from the other studies. In addition, the significant heterogeneity was not affected by deselecting it from the analysis. For MACCE at 5 years, the patients enrolled in the study of Sasaki 2012 [35] were all older than 80 years old, while the mean ages of the other three studies were all around 60 [27,42,43], which might cause different outcomes of MACCE from the others. The significant heterogeneity was also not influenced by deselecting it from the analysis. For analyzing all-cause mortality at 3 years, Yi 2013 [41] enrolled much more patients than the sum number of patients from the other two studies, which might have a dominant influence on the overall result. For the meta-analyses of in-hospital MACCE and CVE, they both showed no significant heterogeneity and included the studies of Etienne 2013 [27] and Yi 2012 [42], which presented quite different results. However, these studies affected the overall results with non-dominant weight. On the other hand, all studies included in this research had acceptable quality. In this way, there was no need to remove Etienne 2013 and Yi 2012 from these two analyses. 3.5. Publication bias As we mentioned above, funnel plots were constructed for visual evaluation of publication bias and Egger's test was performed for statistical assessment. None except for the meta-analyses of inhospital death, TVR and CVE, all-cause death at 12 months and MACCE at 2 years presented significant publication bias (data not shown). 4. Discussion By taking all the results together, it seems that incidences of longterm MACCE, all-cause death, TVR, RRV and MI were lower in the OPCAB group, but a higher incidence of post-operative CVE than DES without affecting the rates of in-hospital death and MI was detected. As mentioned above, there were no statistical differences in the meta-analyses of CVE at 12 months, 2 years and 3 years between OPCAB and DES partly due to the small sample size of each analysis. Because the concepts of CVE and stroke are not quite distinguished from each other, only the studies with definite data of CVE and/or stroke were enrolled separately for the best of accuracy, which resulted in the small sample size in analyzing CVE and insufficient data for analyzing stroke. In our results of short-term outcomes, a higher rate of post-operative CVE might occur with OPCAB, but none of the differences were statistically significant in the subgroups of LM, LADA and MVD revascularization (data not shown). In this case, it seems consistent with the fact that the significance wouldn't appear until there's a sample size large enough. However, Edelman et al. demonstrated that the rates of stroke between OPCAB and PCI with BMS or DES were similar in their meta-analysis [45], and another related meta-analysis suggested a beneficial effect of less post-operative stroke on OPCAB over traditional CABG [17]. Although we enrolled more studies of more patients to enlarge our sample size with a more specific focus, the propensity bias couldn't be denied as we included more OCs and fewer RCTs than them. In addition, significant publication bias was found in this analysis of in-hospital CVE, which made this result less convincible. It could be easily understood that aortic cross-clamping technique and going on-

9

pump are well-known risk factors for post-operative strokes [46–48] and the clampless OPCAB strategy seemed to decrease the rate of post-operative stroke when compared with on-pump [49]. However, none of the studies included in the present analysis demonstrated the detailed data about the application of the clampless strategy or the no-touch aorta technique. Therefore, according to our results, the increased incidence of post-operative CVE could be attributed to a number of confounding factors such as different baselines of patients' ages, gender, co-morbidity and operation strategy selection, and whether the clampless OPCAB strategy could decrease the increased incidence of post-operative CVE remains unclear in the present study but promising [49,50]. In that way, comparing the clinical results of clampless OPCAB with that of DES may have profound guiding significance for the strategy of treating CAD. Furthermore, if the off-pump technique, clampless strategy or no-touch aorta technique such as the HEARTSTRING device [51] and the PAS-Port system [50], prophylactic anticoagulation therapy, atrial appendage exclusion and other novel methods could be well developed and applied, surgical revascularization may beat PCI in avoiding post-operative stroke. However, there's still the possibility of conversion to CPB during OPCAB, which may result in increased in-hospital and long-term mortality [52]. And stroke is more common in converted OPCAB patients [53]. Despite the fact that establishing CPB and the operation on the aorta may produce microemboli out of calcified plague, hypotension or hemodynamic instability that mainly causes the conversion may lead to cerebral hypoperfusion or embolization [54,55] and then result in CVE. Therefore, it means a great deal to comprehensively assess the condition of each patient and recognize the risk factors [52] of conversion before determining the operation plan in order to reduce the incidence of conversion as possible as one can. On the other hand, incomplete revascularization of OPCAB has always been denounced when compared with traditional CABG, though the long-term results haven't been much influenced [56], which seems similar to the results of meta-analysis comparing off-pump and on-pump CABG [17]. Nevertheless, with the increasing experience of OPCAB obtained by more and more surgeons and the development of ancillary equipments, it has been proved that OPCAB could provide satisfactory revascularization completeness [57]. According to our results, all the studies included for meta-analyzing the rates of TVR and RRV at 12 months, 2 years, 3 years and 5 years individually presented better outcomes of restenosis occurrence in OPCAB-treated patients, which in turn led to unanimous combining results indicating that OPCAB remained superior to DES in the aspect of restenosis at the level of larger sample size. Moreover, no statistical difference was found in the combining rates of in-hospital TVR between OPCAB and DES and no substantial change presented during sensitivity analysis. There seemed no difference of the incidences of short-term repeat revascularization between groups, but significant publication bias was also found. On the other hand, 3 studies [24,27,37] out of 7 were removed from this analysis because no in-hospital TVR occurred in both groups. In this case, the incidence of in-hospital TVR is too low to be compared between groups, which could also be influenced by many confounding factors. Hence, related data from large-scale RCTs should be collected to solve this disagreement. Similarly, no agreement has been achieved in the analysis of in-hospital MI, in which no study significantly favored OPCAB or DES. Considering the long-term results of MI, TVR and RRV we mentioned above, significant differences tended to occur during longer-term follow-up and this cluster of results combining MI incidences at certain time points seemed to be consistent with the conclusion of long-term rates of TVR and RRV on the other side. In addition, the minimally invasive characteristic of PCI has always been emphasized and used to explain its low post-operative mortality.


10


Fig. 5. Forrest plots of the relative risk (RR) of cumulative myocardial infarction (MI) after off-pump coronary artery bypass grafting (OPCAB) versus drug-eluting stenting (DES) at different time points: RR of in-hospital MI (panel A), RR of cumulative MI at 12 months (panel B), RR of cumulative MI at 2 years (panel C), and RR of cumulative MI at 3 years (panel D).



11

Fig. 6. Forrest plots of the relative risk (RR) of cumulative cerebrovascular events (CVE) after off-pump coronary artery bypass grafting (OPCAB) versus drug-eluting stenting (DES) at different time points: RR of in-hospital CVE (panel A), RR of cumulative CVE at 12 months (panel B), RR of cumulative CVE at 2 years (panel C), and RR of cumulative CVE at 3 years (panel D).

Yan et al. found that DES was associated with lower rates of all-cause mortality at 30 days than CABG (including both off-pump and onpump) [58], but Edelman et al. demonstrated similar rates of all-cause mortality at 30 days between PCI and OPCAB [45], which was consistent with our result. In this way, the disadvantage of too much harm of

conventional CABG might be compensated by OPCAB. However, it was also indicated that OPCAB had no substantial effect on postoperative mortality when compared with on-pump CABG from a meta-analysis [59]. Therefore, the underlying different baselines of patient characteristics and the development of surgical skills and


12


post-operative intensive care could be the explanation for this disagreement. Significant publication bias found in this analysis also made its result less persuasive. For the combining results of all-cause mortality, there was significant difference in 3-year outcome but none in 5-year outcome, which seemed contradictory. And sensitivity analysis also didn't make any significant changes to the pooled RR of all-cause mortality at 5 years. To solve this disagreement, more raw data of cardiac death are needed for detecting the possible beneficial effect of OPCAB over DES because there're a lot more kinds of death included in all-cause death. Although most patients enrolled only suffered from cardiovascular disease, death of other reasons, such as cancer, accident, suicide and so on, couldn't be prevented during the follow-up period. Although the results of cumulative MACCE at 12 months and 5 years turned into no significant differences after the conversion into random effects model due to unavoidable heterogeneity, there's at least a trend that fewer MACCE occurred with patients who underwent OPCAB. As the concept of MACCE includes a bunch of basic outcomes, meta-analyzing this outcome may introduce a lot more confounding factors and hardly reveal the tangible results altogether. However, taking all these short-term results together, it's easy to understand the non-significant difference of in-hospital MACCE between OPCAB and DES as it included death of any cause, nonfatal MI, CVE, and repeat revascularization. Hence, summarizing adverse events by each year or a certain time period seems to be more cautious for presenting different overall results between OPCAB and DES. This meta-analysis has a number of limitations. First of all, most studies included were OCs with small sample size, which may influence the combining results with patients' and/or doctors' preference. Additionally, a few OCs didn't perform propensity match for enrolled patients resulting in different baselines between the two groups. Hence, more trials comparing OPCAB versus DES should be performed to explore the benefits of each intervention. Second, subgroup analysis and risk factor assessment cannot be done as original data were unavailable. Although EES was found to be better than PES/SES in 1 year outcomes [60] and two different generations of DES were both included in this analysis, unavailable raw data stopped us from comparing OPCAB with the first-generation DES and the second-generation DES respectively. Similarly, more subgroup analyses divided by the extent of CAD, comorbidity, gender, age and race could have been done to reveal a full understanding of therapy selection strategy, if more original data were available. Moreover, raw data could be used to analyze more precise outcome index such as TLR, cardiac death, non-fatal MI and stroke. Third, as we mentioned above, in order to combine as much data as possible, a few “similar” studies were involved. Although measures toprevent dataoverlappingweretaken,there'sstillthepossibilitythatthefinalresults might be affected by invisible duplications. Lastly, as OPCAB requires certain cardiac surgery skills and experience, most studies included were performed by different large-scale medical centers for cardiovascular surgery. The preference in patient selection, adjuvant medical treatment and result reporting cannot be ignored, which might possibly end up with non-objective conclusions. Therefore, morebasicanalysesoforiginaldataaboutauthors'preferenceseemnecessary for the quality of the meta-analysis.

5. Conclusion According to our results of meta-analyzing 22 reports, OPCAB remains superior to DES with better long-term outcomes of MACCE, all-cause mortality, TVR, RRV and MI without influencing the incidences of in-hospital death and MI, but whether OPCAB displays equally to DES in the aspect of post-operative CVE still lies nebulous.

Appendix AA.1. Search strategy A.1.1. PubMed #1 off-pump[All fields] OR off pump[All fields] OR OPCAB[All fields] OR off-pump bilateral internal thoracic arterial grafting[All fields] OR BITA[All fields] OR minimally invasive direct coronary artery bypass [All fields] OR MIDCAB[All fields] OR beating heart surgery[All fields] OR octopus[All fields] OR off-pump internal thoracic artery[All fields] OR without extra-corporeal circulation[All fields] #2 drug-eluting stents[All fields] OR drug-eluting stent[All fields] OR DES[All fields] OR DESs[All fields] OR drug eluting stents[All fields] OR drug eluting stent[All fields] OR sirolimus-eluting stent[All fields] OR SES[All fields] OR paclitaxel-eluting stents[All fields] OR Cypher[All fields] OR Taxus[All fields] OR everolimus-eluting stents[All fields] OR EES[All fields] OR limus-eluting stents[All fields] OR LES[All fields] OR Resolute zotarolimus-eluting stents[All fields] OR R-ZES[All fields] OR Endeavor zotarolimus-eluting stents[All fields] OR E-ZES[All fields] OR biolimus A9-eluting stent[All fields] OR novolimuseluting stent[All fields] OR pimecrolimus-eluting stent[All fields] OR cobalt-chromium everolimus eluting stents[All fields] OR CoCrEES[All fields] OR Xience V[All fields] #3 #1 AND #2 A.1.2. EMBASE #1 ‘off-pump’ OR ‘off pump’ OR ‘OPCAB’ OR ‘off-pump bilateral internal thoracic arterial grafting’ OR ‘BITA’ OR ‘minimally invasive direct coronary artery bypass’ OR ‘MIDCAB’ OR ‘beating heart surgery’/syn OR ‘octopus’/syn OR ‘off-pump internal thoracic artery’ OR ‘without extra-corporeal circulation’ #2 ‘drug-eluting stents’/syn OR ‘drug-eluting stent’/syn OR ‘DES’ OR ‘DESs’ OR ‘drug eluting stents’/syn OR ‘drug eluting stent’/syn OR ‘sirolimus-eluting stent’ OR ‘SES’ OR ‘paclitaxel-eluting stents’ OR ‘Cypher’ OR ‘Taxus’/syn OR ‘everolimus-eluting stents’ OR ‘EES’/syn OR ‘limus-eluting stents’ OR ‘LES’ OR ‘Resolute zotarolimus-eluting stents’ OR ‘R-ZES’ OR ‘Endeavor zotarolimus-eluting stents’ OR ‘E-ZES’ OR ‘biolimus A9-eluting stent’ OR ‘novolimus-eluting stent’ OR ‘pimecrolimus-eluting stent’ OR ‘cobalt-chromium everolimus eluting stents’ OR ‘CoCr-EES’ OR ‘Xience V’/syn #3 #1 AND #2 Appendix B

Table Quality assessment of OCs based on NOS. (Each asterisk represents one point.) Study ID

Selection

Comparability

Outcome

Total score

Ben-Gal 2006 Briguori 2007 Buszman 2011 Chieffo 2006 Etienne 2009 Etienne 2013 Glineur 2009 Herz 2004 Herz 2005 Mack 2008 Moshkovitz 2005 Sasaki 2012 Sata 2009 Ungureanu 2013 Yamagata 2010 Yang 2008 Yi 2012 Yi 2012 Yi 2012a Yi 2013

*** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** ***

** * * * ** ** ** * ** ** ** ** * * ** ** ** ** ** **

*** ** ** *** *** *** ** ** ** ** ** ** ** ** *** *** *** *** *** ***

8 6 6 7 8 8 7 6 7 7 7 7 6 6 8 8 8 8 8 8



13

Appendix C

Figure. Risk of bias graph and summary for the included RCTs.

References [1] WHO publishes definitive atlas on global heart disease and stroke epidemic. Indian J Med Sci 2004;58:405–6. [2] Go AS, Mozaffarian D, Roger VL, et al. Heart disease and stroke statistics—2013 update: a report from the American Heart Association. Circulation 2013;127: e6-245. [3] Fox K, Garcia MA, Ardissino D, et al. Guidelines on the management of stable angina pectoris: executive summary: the Task Force on the Management of Stable Angina Pectoris of the European Society of Cardiology. Eur Heart J 2006;27:1341–81. [4] Fihn SD, Gardin JM, Abrams J, et al. 2012 ACCF/AHA/ACP/AATS/PCNA/SCAI/STS Guideline for the diagnosis and management of patients with stable ischemic heart disease: a report of the American College of Cardiology Foundation/ American Heart Association Task Force on Practice Guidelines, and the American College of Physicians, American Association for Thoracic Surgery, Preventive Cardiovascular Nurses Association, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol 2012;60:e44-164. [5] O'Gara PT, Kushner FG, Ascheim DD, et al. 2013 ACCF/AHA guideline for the management of ST-elevation myocardial infarction: executive summary: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2013;61:485–510. [6] Steg PG, James SK, Atar D, et al. ESC Guidelines for the management of acute myocardial infarction in patients presenting with ST-segment elevation. Eur Heart J 2012;33:2569–619.

[7] Gruntzig A, Schneider HJ. The percutaneous dilatation of chronic coronary stenoses— experiments and morphology. Schweiz Med Wochenschr 1977;107:1588. [8] Sigwart U, Puel J, Mirkovitch V, Joffre F, Kappenberger L. Intravascular stents to prevent occlusion and restenosis after transluminal angioplasty. N Engl J Med 1987;316:701–6. [9] Honda Y, Grube E, de La Fuente LM, Yock PG, Stertzer SH, Fitzgerald PJ. Novel drugdelivery stent: intravascular ultrasound observations from the first human experience with the QP2-eluting polymer stent system. Circulation 2001;104:380–3. [10] Mohr FW, Morice MC, Kappetein AP, et al. Coronary artery bypass graft surgery versus percutaneous coronary intervention in patients with three-vessel disease and left main coronary disease: 5-year follow-up of the randomised, clinical SYNTAX trial. Lancet 2013;381:629–38. [11] Kolessov VI. Mammary artery-coronary artery anastomosis as method of treatment for angina pectoris. J Thorac Cardiovasc Surg 1967;54:535–44. [12] Benetti F, Mariani MA, Sani G, et al. Video-assisted minimally invasive coronary operations without cardiopulmonary bypass: a multicenter study. J Thorac Cardiovasc Surg 1996;112:1478–84. [13] Robinson MC, Gross DR, Zeman W, Stedje-Larsen E. Minimally invasive coronary artery bypass grafting: a new method using an anterior mediastinotomy. J Card Surg 1995;10:529–36. [14] Diegeler A, Falk V, Matin M, et al. Minimally invasive coronary artery bypass grafting without cardiopulmonary bypass: early experience and follow-up. Ann Thorac Surg 1998;66:1022–5. [15] Reichenspurner H, Boehm DH, Gulbins H, et al. Robotically assisted endoscopic coronary artery bypass procedures without cardiopulmonary bypass. J Thorac Cardiovasc Surg 1999;118:960–1.


14


[16] Falk V, Diegeler A, Walther T, et al. Total endoscopic computer enhanced coronary artery bypass grafting. Eur J Cardiothorac Surg 2000;17:38–45. [17] Afilalo J, Rasti M, Ohayon SM, Shimony A, Eisenberg MJ. Off-pump vs. on-pump coronary artery bypass surgery: an updated meta-analysis and meta-regression of randomized trials. Eur Heart J 2012;33:1257–67. [18] Marui A, Okabayashi H, Komiya T, et al. Benefits of off-pump coronary artery bypass grafting in high-risk patients. Circulation 2012;126:S151–7. [19] Serruys PW, Morice MC, Kappetein AP, et al. Percutaneous coronary intervention versus coronary-artery bypass grafting for severe coronary artery disease. N Engl J Med 2009;360:961–72. [20] Stang A. Critical evaluation of the Newcastle–Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses. Eur J Epidemiol 2010;25:603–5. [21] Higgins JPT, Green S; Cochrane Collaboration. Cochrane handbook for systematic reviews of interventions Version 5.1.0. Chichester, England; Hoboken, NJ: WileyBlackwell; 2008 [updated March 2011]. [22] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009;339:b2535. [23] Ben-Gal Y, Mohr R, Braunstein R, et al. Revascularization of left anterior descending artery with drug-eluting stents: comparison with minimally invasive direct coronary artery bypass surgery. Ann Thorac Surg 2006;82:2067–71. [24] Briguori C, Condorelli G, Airoldi F, et al. Comparison of coronary drug-eluting stents versus coronary artery bypass grafting in patients with diabetes mellitus. Am J Cardiol 2007;99:779–84. [25] Buszman PP, Krol M, Cisowski M, et al. DES vs MIDCAB for proximal LAD disease: long term registry results. J Am Coll Cardiol 2011;58:B53–. [26] Chieffo A, Morici N, Maisano F, et al. Percutaneous treatment with drug-eluting stent implantation versus bypass surgery for unprotected left main stenosis: a singlecenter experience. Circulation 2006;113:2542–7. [27] Etienne PY, D'Hoore W, Papadatos S, et al. Five-year follow-up of drug-eluting stents implantation vs minimally invasive direct coronary artery bypass for left anterior descending artery disease: a propensity score analysis. Eur J Cardiothorac Surg 2013;44:884–90. [28] Etienne PY, Glineur D, Papadatos S, et al. Comparison of minimally invasive direct coronary artery bypass surgery with implantation of drug-eluting stents in patients with left anterior descending coronary artery disease. Innovations (Phila) 2009;4:340–4. [29] Glineur D, Boodhwani M, Noirhomme P, et al. Short and mid-term clinical outcome following single vessel LAD revascularization with MIDCAB versus DES. Innovations (Phila) 2009;4:159. [30] Herz I, Mohr R, Moshkovitz Y, et al. Revascularization of the left anterior descending artery with drug-eluting stents: comparison with arterial off-pump surgery. Heart Surg Forum 2004;7:E490–2. [31] Herz I, Moshkovitz Y, Hendler A, et al. Revascularization of left anterior descending artery with drug-eluting stents: comparison with off-pump surgery. Ann Thorac Surg 2005;79:88–92. [32] Hong SJ, Lim DS, Seo HS, et al. Percutaneous coronary intervention with drug-eluting stent implantation vs. minimally invasive direct coronary artery bypass (MIDCAB) in patients with left anterior descending coronary artery stenosis. Catheter Cardiovasc Interv 2005;64:75–81. [33] Mack MJ, Prince SL, Herbert M, et al. Current clinical outcomes of percutaneous coronary intervention and coronary artery bypass grafting. Ann Thorac Surg 2008;86:496–503 [discussion]. [34] Moshkovitz Y, Mohr R, Braunstein R, et al. Revascularization of left anterior descending coronary artery in patients with single and multivessel disease: comparison between off-pump internal thoracic artery and drug-eluting stent. Chest 2005;128:804–9. [35] Sasaki K, Fukui T, Tabata M, Takanashi S. Five-year results of off-pump coronary artery bypass grafting and percutaneous coronary intervention with drug eluting stent for multivessel coronary artery disease in elderly patients. Innovations (Phila) 2012;7:140. [36] Sata Y, Kataoka Y, Yagi N, et al. Comparison of three year outcome between sirolimus-eluting stent and off-pump coronary bypass surgery in elderly patients with multi-vessel disease. Am J Cardiol 2009;104:136D. [37] Thiele H, Neumann-Schniedewind P, Jacobs S, et al. Randomized comparison of minimally invasive direct coronary artery bypass surgery versus sirolimus-eluting stenting in isolated proximal left anterior descending coronary artery stenosis. J Am Coll Cardiol 2009;53:2324–31. [38] Ungureanu C, Laruelle C, Khoury E, et al. Retrospective comparison of minimally invasive direct coronary artery bypass surgery versus Xience drug-eluting stenting in isolated proximal left anterior descending coronary artery stenosis. Acta Cardiol 2013;68:104–5.

[39] Yamagata K, Kataoka Y, Kokubu N, et al. A 3-year clinical outcome after percutaneous coronary intervention using sirolimus-eluting stent and off-pump coronary artery bypass grafting for the treatment of diabetic patients with multivessel disease. Circ J 2010;74:671–8. [40] Yang JH, Gwon HC, Cho SJ, et al. Comparison of coronary artery bypass grafting with drug-eluting stent implantation for the treatment of multivessel coronary artery disease. Ann Thorac Surg 2008;85:65–70. [41] Yi G, Joo HC, Youn YN, Hong S, Yoo KJ. Stent versus off-pump coronary bypass grafting in the second-generation drug-eluting stent era. Ann Thorac Surg 2013;96:535–41. [42] Yi G, Youn YN, Hong S, Song SW, Yoo KJ. Midterm outcome of off-pump bypass procedures versus drug-eluting stent for unprotected left main coronary artery disease. Ann Thorac Surg 2012;94:15–22. [43] Yi G, Youn YN, Hong S, Song SW, Yoo KJ. Comparison of long-term outcome of offpump coronary artery bypass grafting versus drug-eluting stents in triple-vessel coronary artery disease. Am J Cardiol 2012;109:819–23. [44] Yi G, Youn YN, Yoo KJ, Hong YS. Comparison of off-pump coronary artery bypass grafting with percutaneous coronary intervention versus drug-eluting stents for three-vessel coronary artery disease. Ann Thorac Surg 2008;86:1438–43. [45] Edelman JJ, Yan TD, Padang R, Bannon PG, Vallely MP. Off-pump coronary artery bypass surgery versus percutaneous coronary intervention: a meta-analysis of randomized and nonrandomized studies. Ann Thorac Surg 2010;90:1384–90. [46] Brizzio ME, Zapolanski A, Shaw RE, Sperling JS, Mindich BP. Stroke-related mortality in coronary surgery is reduced by the off-pump approach. Ann Thorac Surg 2010;89:19–23. [47] Emmert MY, Salzberg SP, Seifert B, et al. Is off-pump superior to conventional coronary artery bypass grafting in diabetic patients with multivessel disease? Eur J Cardiothorac Surg 2011;40:233–9. [48] Sedrakyan A, Wu AW, Parashar A, Bass EB, Treasure T. Off-pump surgery is associated with reduced occurrence of stroke and other morbidity as compared with traditional coronary artery bypass grafting: a meta-analysis of systematically reviewed trials. Stroke 2006;37:2759–69. [49] Emmert MY, Salzberg SP, Seifert B, et al. Clampless off-pump surgery reduces stroke in patients with left main disease. Int J Cardiol 2012;167:2097–101. [50] Borgermann J, Hakim K, Renner A, et al. Clampless off-pump versus conventional coronary artery revascularization: a propensity score analysis of 788 patients. Circulation 2012;126:S176–82. [51] Emmert MY, Grunenfelder J, Scherman J, et al. HEARTSTRING enabled no-touch proximal anastomosis for off-pump coronary artery bypass grafting: current evidence and technique. Interact Cardiovasc Thorac Surg 2013;17:538–41. [52] Chowdhury R, White D, Kilgo P, et al. Risk factors for conversion to cardiopulmonary bypass during off-pump coronary artery bypass surgery. Ann Thorac Surg 2012;93:1936–41 [discussion 42]. [53] Mukherjee D, Rao C, Ibrahim M, et al. Meta-analysis of organ damage after conversion from off-pump coronary artery bypass procedures. Ann Thorac Surg 2011;92:755–61. [54] Caplan LR, Hennerici M. Impaired clearance of emboli (washout) is an important link between hypoperfusion, embolism, and ischemic stroke. Arch Neurol 1998;55:1475–82. [55] Tufo HM, Ostfeld AM, Shekelle R. Central nervous system dysfunction following open-heart surgery. JAMA 1970;212:1333–40. [56] Robertson MW, Buth KJ, Stewart KM, et al. Complete revascularization is compromised in off-pump coronary artery bypass grafting. J Thorac Cardiovasc Surg 2013;145:992–8. [57] Puskas JD, Williams WH, Duke PG, et al. Off-pump coronary artery bypass grafting provides complete revascularization with reduced myocardial injury, transfusion requirements, and length of stay: a prospective randomized comparison of two hundred unselected patients undergoing off-pump versus conventional coronary artery bypass grafting. J Thorac Cardiovasc Surg 2003;125:797–808. [58] Yan TD, Padang R, Poh C, et al. Drug-eluting stents versus coronary artery bypass grafting for the treatment of coronary artery disease: a meta-analysis of randomized and nonrandomized studies. J Thorac Cardiovasc Surg 2011;141:1134–44. [59] Sa MP, Ferraz PE, Escobar RR, et al. Off-pump versus on-pump coronary artery bypass surgery: meta-analysis and meta-regression of 13,524 patients from randomized trials. Rev Bras Cir Cardiovasc 2012;27:631–41. [60] Cassese S, Piccolo R, Galasso G, De Rosa R, Piscione F. Twelve-month clinical outcomes of everolimus-eluting stent as compared to paclitaxel- and sirolimuseluting stent in patients undergoing percutaneous coronary interventions. A metaanalysis of randomized clinical trials. Int J Cardiol 2011;150:84–9.


Coronary-Artery Bypass Grafting.

Comparison of percutaneous transluminal coronary angioplasty versus coronary artery bypass grafting for multivessel coronary artery disease.

Safety of Simultaneous Coronary Artery Bypass Grafting and Carotid Endarterectomy Versus Isolated Coronary Artery Bypass Grafting: A Randomized Clinical Trial.

Optimal revascularization for left main coronary artery disease-coronary artery bypass grafting versus percutaneous coronary intervention.

Revival of the radial artery for coronary artery bypass grafting.

Contemporary coronary artery bypass grafting.

Redo coronary artery bypass grafting.

Comparison of percutaneous coronary intervention with drug eluting stents versus coronary artery bypass grafting in patients with multivessel coronary artery disease: Meta-analysis of six randomized controlled trials.

Intrathecal sufentanil for coronary artery bypass grafting.

Off-pump coronary artery bypass grafting versus conventional coronary artery bypass grafting: What we don't know.

Percutaneous transluminal coronary angioplasty and the changing indications for coronary artery bypass grafting for single-vessel coronary artery disease.

Left main coronary artery stenosis: a meta-analysis of drug-eluting stents versus coronary artery bypass grafting.

Outcomes of coronary artery bypass grafting versus percutaneous coronary intervention with second-generation drug-eluting stents for patients with multivessel and unprotected left main coronary artery disease.

Optimal coronary artery bypass grafting strategy for acute coronary syndrome.

Is microplegia superior to regular blood cardioplegia during coronary artery bypass grafting?

A giant coronary aneurysm formed after coronary artery bypass grafting.

Coronary artery bypass surgery for left main coronary artery disease.

Progression from stenosis to occlusion in the proximal native coronary artery after coronary artery bypass grafting.

Hybrid coronary revascularization versus coronary artery bypass grafting for multivessel coronary artery disease: systematic review and meta-analysis.

Patient's Perception About Coronary Artery Bypass Grafting.

Exercise echocardiography after coronary artery bypass grafting.

Coronary artery bypass grafting saves lives!

Conduits in coronary artery bypass grafting.

Radial artery and right internal thoracic artery: jousting for the throne of coronary artery bypass grafting.