Methods
In May 2013, we searched in the published English language literature using online databases (PubMed, EMBASE, and Cochrane Library) and cardiology society websites (cardiosource.com, tctmd.com, crtonline.org) for studies reporting outcomes after implantation of DES in CTO. Query terms included chronic total occlusion, total coronary occlusion, chronic coronary occlusion, DES, BMS, sirolimus-eluting stent (SES), paclitaxel-eluting stent (PES), everolimus-eluting stent (EES), and zotarolimus-eluting stent (ZES). The bibliography of the retrieved articles was searched for additional citations. Studies reporting comparative outcomes between first- and second-generation DESs in CTOs were included. Drug-eluting stent registries that reported outcomes specifically for the subgroup of CTO were also included. Case reports, reviews, editorials, and letters were excluded. All articles were assessed by two reviewers (VL and ESB) before inclusion in the review. Extracted data included study design, sample size, baseline characteristics, stent types, duration of follow-up, routine angiographic follow-up, and outcome data. Outcomes collected included death, myocardial infarction (MI), target lesion revascularization (TLR), target vessel revascularization (TVR), stent thrombosis as defined by the Academic Research Consortium (ARC) criteria, and major adverse cardiac event (MACE) rate, defined as the composite of death, MI, and TVR. Angiographic outcomes included binary in-segment angiographic restenosis (defined as minimum lumen diameter in the recanalized artery of <30% of the reference diameter within the stent and the 5 mm proximal and distal adjacent segments) and reocclusion (defined as 100% stenosis of the target vessel at follow-up angiography).
The odds ratio (OR) for each outcome was calculated from individual studies and pooled with the DerSimonian-Laird random-effects method using Review Manager (RevMan Version 5.2, Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2012). Statistical heterogeneity was assessed with the Chi test, with P-values <.10 suggesting significant heterogeneity. Inconsistency across the trials was assessed using I, where I <25% suggests mild, I between 25% and 50% suggests moderate, and I >50% suggests extensive statistical inconsistency. Publication bias was assessed by visual inspection of funnel plots. Analyses were performed according to the Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group recommendations and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.