Predictive factors of restenosis after coronary stent placement.

OBJECTIVES The objective of this study was to identify clinical, lesional and procedural factors that can predict restenosis after coronary stent placement. BACKGROUND Coronary stent placement reduces the restenosis rate compared with that after percutaneous transluminal coronary angioplasty (PTCA). However, restenosis remains an unresolved issue, and identification of its predictive factors may allow further insight into the underlying process. METHODS All patients with successful coronary stent placement were eligible for this study unless they had had a major adverse cardiac event during the 1st 30 days after the procedure. Of the 1,349 eligible patients (1,753 lesions), follow-up angiography at 6 months was performed in 80.4% (1,084 patients, 1,399 lesions). Demographic, clinical, lesional and procedural data were prospectively recorded and analyzed for any predictive power for the occurrence of late restenosis after stenting. Restenosis was evaluated by using three outcomes at follow-up: binary restenosis as a diameter stenosis > or =50%, late lumen loss as lumen diameter reduction and target lesion revascularization (TLR) as any repeat PTCA or coronary artery bypass surgery involving the stented lesion. RESULTS Multivariate analysis demonstrated that diabetes mellitus, placement of multiple stents and minimal lumen diameter (MLD) immediately after stenting were the strongest predictors of restenosis. Diabetes increased the risk of binary restenosis with an odds ratio (OR) [95% confidence interval] of 1.86 [1.56 to 2.16] and the risk of TLR with an OR of 1.45 [1.11 to 1.80]. Multiple stents increased the risk of binary restenosis with an OR of 1.81 [1.55 to 2.06] and that of TLR with an OR of 1.94 [1.66 to 2.22]. An MLD <3 mm at the end of the procedure augmented the risk of binary restenosis with an OR of 1.81 [1.55 to 2.06] and that of TLR with an OR of 2.05 [1.77 to 2.34]. Classification and regression tree analysis demonstrated that the incidence of restenosis may be as low as 16% for a lesion without any of these risk factors and as high as 59% for a lesion with a combination of these risk factors. CONCLUSIONS Diabetes, multiple stents and smaller final MLD are strong predictors of restenosis after coronary stent placement. Achieving an optimal result with a minimal number of stents during the procedure may significantly reduce this risk even in patients with adverse clinical characteristics such as diabetes.