Long-term clinical outcome and predictors of major adverse cardiac events after percutaneous interventions on saphenous vein grafts

OBJECTIVES The purpose of this study was to examine the long-term clinical outcome after percutaneous intervention of saphenous vein grafts (SVG) and to identify the predictors of major adverse cardiac events (MACE). BACKGROUND Percutaneous interventions of SVGs have been associated with more procedural complications and higher restenosis rates compared with interventions on native vessels. METHODS From 1993 to 1997, 1,062 patients underwent percutaneous intervention on 1,142 SVG lesions. Procedural, in-hospital and long-term clinical outcomes were recorded in a database and analyzed. RESULTS In-hospital MACE occurred in 137 patients (13%) including death (8%), Q-wave myocardial infarction (MI) (2%) and coronary artery bypass surgery (3%). Late MACE occurred in 565 patients (54%) including death (9%), Q-wave MI (9%) and target vessel revascularization (36%). Any MACE occurred in 457 (43%) patients. Follow-up was available in 1,056 (99%) patients at 3 ± 1 year. Univariate predictors were restenotic lesion (odds ratio [OR]: 2.47, confidence interval [CI]: 1.13 to 3.85, P = 0.0003), unstable angina (OR: 1.99, CI: 1.27 to 2.91, P = 0.04) and congestive heart failure (CHF) (OR: 1.97, CI: 1.14 to 3.24, P = 0.02) for in-hospital MACE, and peripheral vascular disease (PVD) (OR: 2.18, CI: 1.34 to 3.44, P = 0.002), intra-aortic balloon pump placement (OR: 2.08, CI: 1.13 to 3.85, P = 0.02) and previous MI (OR: 1.97, CI: 1.14 to 3.25, P = 0.007) for late MACE. Independent multivariate predictors for late MACE were restenotic lesion (relative risk [RR] 1.33, P = 0.02), PVD (RR: 1.31, P = 0.01), CHF (RR: 1.42, P = 0.01) and multiple stents (RR: 1.47, P = 0.004). Angiographic follow-up was available for 422 patients. Angiographic restenosis occurred in 122 (29%) of stented SVGs and 181 (43%) of nonstented SVGs (p = 0.04). Stent implantation did not confer a survival benefit. CONCLUSIONS Despite the use of new interventional devices, SVG interventions are associated with significant morbidity and mortality; SVG stenting is not associated with better three-year event-free survival. This may be due to progressive disease at nonstented sites.