Role of vessel size as a predictor for the occurrence of in-stent restenosis in patients with diabetes mellitus

Intracoronary stents have been shown to reduce the rate of restenosis when compared with balloon angioplasty, but in-stent restenosis continues to be an important clinical problem. It was therefore the aim of this registry to identify procedural and angiographic predictors for the occurrence of in-stent restenosis. We analyzed 368 patients with 421 lesions who underwent coronary stent implantation between January 1998 and February 2000. Indications for the placement of a coronary stent were severe dissections (37%), suboptimal angiographic results (38%), restenotic lesions (20%), and graft lesions (4%). Angiographic follow-up was obtained in 270 patients (73%) with 293 lesions after 6 months. Clinical and angiographic variables were analyzed by univariate and multivariate models for the ability to predict the occurrence of in-stent restenosis, defined as a diameter stenosis >50%. In-stent restenosis was angiographically documented in 67 patients and 68 lesions (23%). Under all tested variables the reference luminal diameter before stent implantation (p = 0.006) and diabetes mellitus (p = 0.023) were identified as independent predictors for the occurrence of in-stent restenosis. The comparison of diabetic and nondiabetic patients according to vessel size revealed a 2 times higher rate of in-stent restenosis in small vessels (44% vs 23%, p = 0.002), whereas in vessels >3.0 mm the rate of in-stent restenosis was not significantly different between the 2 groups. In this registry, the clinical variable diabetes and the procedural variable reference vessel size were independent predictors for the occurrence of in-stent restenosis. In these patients, the rate of in-stent restenosis was as high as 45%.