1

Introduction

Percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) are established strategies for coronary revascularization in the setting of ischemic heart disease. Although CABG was the standard of care for patients with multivessel disease, the improvement of interventional techniques, the introduction of bare metal stents (BMS) and later drug eluting stents (DES) led to increased use of PCI in managing patients with multivessel disease. Several randomized controlled trials (RCTs) compared the two strategies in the plain old balloon angioplasty (POBA) era , the BMS era , and the contemporary DES era .

While data in the DES era coming from RCT comparing DES vs. CABG are consistent in reporting increased incidence of Major cardiac and cerebral Events (MACCE) especially in diabetic patients among all trials, that is not the case for long-term data on the incidence of death, MI and stroke at 5 years. While SYNTAX showed a non-significant difference in the composite of death/MI/stroke between the two strategies at 5 years in both diabetic and non-diabetic groups, FREEDOM showed a higher incidence of this composite in diabetic patients treated with PCI.

Using meta-analysis to pool data from multiple RCTs provides a more precise assessment of the effects of treatment, and also increases the number of patients within clinical subgroups of interest, often providing adequate statistical power to evaluate outcomes in these subgroups . The only meta-analysis of data exclusively derived from RCTs comparing PCI and CABG in stable ischemic heart disease included only patients from either the pre DES, or those from the left main subgroup or combined diabetes subgroup from the BMS with those in the DES era . Data from the pre DES meta-analysis contributed important comparisons of mortality rates resulting from the two strategies; but did not provide information on other important outcomes including myocardial infarction (MI), target vessel revascularization (TVR) and stroke. These data also did not include patients managed with DES, and thus are not representative of contemporary percutaneous management of multivessel disease.

This study reports results from a meta-analysis of six RCTs evaluating outcomes in patients receiving PCI with DES versus CABG in the contemporary era.

2

Methods

Relevant studies were identified through electronic searches of MEDLINE and the Cochrane Central Register of Controlled Trials databases from 01/01/2003 to 05/31/2013. The start date was defined as 01/01/2003 as the FDA approved DES use in 2003 . The search strategy used the terms “percutaneous coronary intervention,” “stent(s),” “drug-eluting stent,” “sirolimus-eluting stent,” or “paclitaxel-eluting stent,” paired with “coronary artery bypass graft.” In addition, we searched bibliographies of relevant studies, reviews, editorials, letters, and meeting abstracts. The analysis was restricted to include only prospective RCTs or pre-specified sub-analyses from RCTs that randomized patients to PCI with DES versus CABG; and reported both safety and efficacy outcomes. The quality of the identified studies was assessed with respect to control for confounders, measurement of exposure, completeness of follow-up, and blinding. We followed a scoring system based on a check list derived from recommended criteria recommended by the QUOROM (The Quality of Reporting of Meta-analyses) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines to assess the quality of the trials used in this meta-analysis .

Two reviewers (ZF and WZ) independently extracted data from the list of included studies. Extracted data included authorship, study period, publication year, study design, study region, baseline characteristics of patients, sample size, clinical events, and duration of follow-up. The following outcomes were extracted: all-cause death, myocardial infarction [MI], stroke [cerebrovascular accident (CVA)], and target vessel revascularization [TVR]. Major cardiac and cerebral Events (MACCE) data were excluded as they were reported differently by each trial.

A random effects model meta-analysis following the DerSimonian–Laird method was used to determine relative risks for the PCI and CABG groups for each endpoint . This method treats study affiliation as a random effect and considers both the within-study variance and the between-study variance. Heterogeneity (inconsistency of results among studies) was assessed using the DerSimonian–Laird I2 statistic, which calculates the proportion of between-study variability that cannot be attributed to sampling variation . All meta-analyses were conducted with R statistical software for Windows with the package meta (R Foundation for Statistical Computing, Vienna, Austria).

2

Methods

Relevant studies were identified through electronic searches of MEDLINE and the Cochrane Central Register of Controlled Trials databases from 01/01/2003 to 05/31/2013. The start date was defined as 01/01/2003 as the FDA approved DES use in 2003 . The search strategy used the terms “percutaneous coronary intervention,” “stent(s),” “drug-eluting stent,” “sirolimus-eluting stent,” or “paclitaxel-eluting stent,” paired with “coronary artery bypass graft.” In addition, we searched bibliographies of relevant studies, reviews, editorials, letters, and meeting abstracts. The analysis was restricted to include only prospective RCTs or pre-specified sub-analyses from RCTs that randomized patients to PCI with DES versus CABG; and reported both safety and efficacy outcomes. The quality of the identified studies was assessed with respect to control for confounders, measurement of exposure, completeness of follow-up, and blinding. We followed a scoring system based on a check list derived from recommended criteria recommended by the QUOROM (The Quality of Reporting of Meta-analyses) and PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines to assess the quality of the trials used in this meta-analysis .

Two reviewers (ZF and WZ) independently extracted data from the list of included studies. Extracted data included authorship, study period, publication year, study design, study region, baseline characteristics of patients, sample size, clinical events, and duration of follow-up. The following outcomes were extracted: all-cause death, myocardial infarction [MI], stroke [cerebrovascular accident (CVA)], and target vessel revascularization [TVR]. Major cardiac and cerebral Events (MACCE) data were excluded as they were reported differently by each trial.

A random effects model meta-analysis following the DerSimonian–Laird method was used to determine relative risks for the PCI and CABG groups for each endpoint . This method treats study affiliation as a random effect and considers both the within-study variance and the between-study variance. Heterogeneity (inconsistency of results among studies) was assessed using the DerSimonian–Laird I2 statistic, which calculates the proportion of between-study variability that cannot be attributed to sampling variation . All meta-analyses were conducted with R statistical software for Windows with the package meta (R Foundation for Statistical Computing, Vienna, Austria).

3

Results

The initial search identified 123 potential articles. Of these, 6 were RCTs that included patient treated with DES . Five of the 6 trials included only DES patients in the PCI arm . One trial included both BMS and DES but reported separate comparisons of DES to CABG for the two groups . Outcomes at one year were reported by all 6 trials . Four of the six trials reported 2-year outcomes , and 3 trials reported 5-year outcomes . Table 1 shows the patients characteristics of each of these RCTs. The six trials included 1-year outcomes for 5123 patients. Longer-term follow-up reports displayed some discrepancy in reporting between different trials, so a cohort of 4498 patients from four trials was used to compare outcomes at 1 and 2 years . Another cohort of 4202 patients from three trials was used to compare outcomes at 1 and 5 years . Table 2 shows the patients characteristics of these 3 cohorts.

Table 1

RCT of PCI with DES vs. CABG.

	SYNTAX	FREEDOM	PRE-COMBAT	VA CARD	CARDIA	Boudriot et al.
Population	General	Diabetics	Left Main	Diabetics	Diabetics	Left Main
Study Size	1800	1900	600	198	502	201
Diabetics	25%	100%	32%	100%	100%	36%
DES stent use	100%	100%	100%	100%	71%	100%
Primary outcome	Composite of death, MI, stroke and TVR at 12 months	Composite of death, MI, stroke at 2 years.	Composite of death, MI, stroke and TVR at 12 months	Composite of death and MI at 2 years	Composite of death, MI, stroke and TVR at 12 months	Freedom of death, MI and TVR at 12 months
Secondary outcomes	Death, MI, stroke and TVR	Composite of death, MI, stroke and TVR,	Death, MI, stroke and TVR	Death, MI, stroke and TVR	Death, MI, stroke and TVR	Death, MI and TVR
Primary outcome Results	MACCE at 12 months: Higher in the PCI group (17.8%, vs. 12.4%; P = 0.002)	The primary outcome at 5 years: Higher in the PCI group (26.6% vs. 18.7%; P = 0.005.	MACCE at 12 months: Higher in the PCI group (8.7% vs. 6.7%; P = 0.01)	Primary outcome at 2 years: Under-powered (HR: 0.89; 95% CI: 0.47–1.71)	Primary outcome at 1 year: No Difference (HR: 1.25, 95% CI: 0.75–2.09; P = 0.39)	Primary outcome at 1 year: Higher in the PCI group (19% vs. 13.9%; P = 0.19 for non-inferiority)
Revascularization rate	Higher in PCI group (13.5% vs. 5.9%, P < 0.001)	Higher in PCI group (12.6% vs. 4.8%; P < 0.001)	Higher in PCI group (9.0% vs. 4.2%; P = 0.02)	Under-powered (HR: 0.93; 95% CI: 0.42–2.07)	Higher in PCI group (HR: 1.77, 95% CI: 1.11–2.82; P = 0.02)	Higher in PCI group (14% vs. 5.9%; non-inferiority P = 0.35)
Follow-up period	1, 2, 3, 4 and 5 years	1, 2 and 5 years	1 and 2 years	1 and 2 years	1 and 5 years	1 year
Inclusion Period	2006–2007	2005–2010	2004–2009	2006–2010	2002–2007	2003–2009
EuroSCORE	3.8 ± 2.6	2.8 ± 2.4	2.6 ± 1.8	N/A	N/A	N/A
SYNTAX score	28.7 ± 11.5	26.1 ± 8.6	N/A	22.1 ± 8.6	N/A	23.50 (14.8–29.0)

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Primary endpoint results of the OMEGA Study: One-year clinical outcomes after implantation of a novel platinum chromium bare metal stent
2. Rational use of rotational atherectomy in calcified lesions in the drug-eluting stent era: Review of the evidence and current practice
3. Transcollateral retrograde diagnostic coronary angiogram — important therapeutic implications for an occluded arterial coronary artery bypass graft
4. Large coronary intramural hematomas: a case series and focused literature review

Stay updated, free articles. Join our Telegram channel

Join