LAUSR

BACKGROUND Secondary percutaneous coronary intervention (PCI) in patients with prior coronary artery bypass graft surgery is increasingly common. Graft vessel PCI has higher rates of adverse events compared with native coronary vessel PCI. AIM To investigate the clinical outcomes of patients with prior CABG who underwent secondary PCI of either a graft vessel (GV), a native coronary vessel (NV) or both graft and native (NG) vessels. METHODS 220 patients (84% male) who underwent PCI in our institution to either GV (n=89), NV (n=103) or both GV and NV (NG group) (n=28) were studied. The study population underwent 378 procedures (GV group; n=126, NV group; n=164 and NG group; n=88). Median follow up was for 36months [range 2-75months]. RESULTS Target vessel revascularisation (TVR) occurred in 12.5% of the GV group and 3.6% in the NV group [p=0.0004], and was predominantly due to in-stent restenosis. Patients who had PCI due to TVR were more likely to suffer from diabetes and peripheral vascular disease. History of chronic renal failure was associated with higher risk (HR 2.21, p=0.005) whereas preserved left ventricular ejection fraction (LVEF) with lower risk (HR 0.17, p=0.0007) of death. The median survival (interval between CABG and end of follow-up period) was lower in the GV compared with the NV group (315 vs 372months p=0.005). CONCLUSION This registry demonstrates inferior long term outcome for patients undergoing secondary PCI of GV versus NV. Where possible, a strategy of NV rather than GV target PCI should be considered in patients with prior CABG. CONDENSED ABSTRACT Secondary PCI in patients with prior CABG surgery is increasingly common. Graft vessel PCI has inferior outcomes with high rates of restenosis and occlusion compared with native coronary vessel PCI. We studied the clinical outcomes of 220 patients with prior CABG who underwent secondary PCI to either a graft vessel (GV), a native coronary vessel (NV) or both graft and native (NG) vessels. Target vessel revascularisation was 5 times higher in the GV compared with the NV group. History of CRF and impaired left ventricular function were associated with higher risk of death. We also found that the median survival (interval between CABG and end of follow-up period) was better in the NV group compared with GV group. This registry study demonstrates inferior long term outcome for patients undergoing secondary PCI of GV. A strategy of NV rather than GV target PCI should be considered in patients with prior CABG.