In-hospital outcomes and usage of embolic protection devices in percutaneous coronary intervention for coronary artery bypass grafts: Insights from a Japanese nationwide registry.

OBJECTIVES We evaluated the in-hospital outcomes of percutaneous coronary intervention (PCI) for bypass graft vessels (GV-PCI) compared with those of PCI for native vessels (NV-PCI) using data from the Japanese nationwide coronary intervention registry. METHODS We included PCI patients (N = 748,229) registered between January 2016 and December 2018 from 1,123 centers. We divided patients into three groups: GV-PCI (n = 2,745); NV-PCI with a prior coronary artery bypass graft (pCABG) (n = 23,932); and NV-PCI without pCABG (n = 721,552). RESULTS GV-PCI implementation was low, and most cases of PCI in pCABG patients were performed in native vessels (89.7%) in contemporary Japanese practice. The risk profile of patients with pCABG was higher than that of those without pCABG. Consequently, GV-PCI patients had a significantly higher in-hospital mortality than NV-PCI patients without pCABG after adjusting for covariates (odds ratio [OR] 2.36, 95% confidence interval [CI] 1.66-3.36, p < .001). Of note, embolic protection devices (EPDs) were used in 18% (n = 383) of PCIs for saphenous vein grafts (SVG-PCI) with a significant variation in its use among institutions (number of PCI: hospitals that had never used an EPD vs. EPD used one or more times = 240 vs. 345, p < .001). The EPDs used in the SVG-PCI group had a significantly lower prevalence of the slow-flow phenomenon after adjusting for covariates (OR 0.45, 95% CI 0.21-0.91, p = .04). CONCLUSION GV-PCI is associated with an increased risk of in-hospital mortality. EDP use in SVG-PCI was associated with a low rate of the slow-flow phenomenon. The usage of EPDs during SVG-PCI is low, with a significant variation among institutions.