BackgroundAngiographic and clinical parameters are poor predictors of in-stent restenosis. Bone marrow–derived CD34+ cells that coexpress a receptor for vascular endothelial growth factor (kinase insert domain receptor [KDR]) are committed to endothelial lineage. Mobilization and infusion of CD34+/KDR+ cells accelerates re-endothelialization and reduces neointimal thickness in vascular injury models. Bioengineered stents capturing CD34+ cells also show expedited re-endothelialization. We examined whether circulating CD34+/KDR+ cell counts can be used to predict restenosis in a bare-metal stent (BMS). MethodsCD34+/KDR+ cells were counted by flow cytometry in 124 nondiabetic patients before BMS implantation and the relation to in-stent late luminal loss (LLL) was examined by angiography at 6 months (primary end point). Neointima was also quantified as the maximum percentage area stenosis (M%AS) and percentage volume intima hyperplasia (%VIH) on intravascular ultrasonography (secondary end points). ResultsMultiple linear regression analysis, taking into account implanted stent length and diameter, revealed no relation between CD34+/KDR+ cell counts and LLL (partial regression coefficient b = 0.11; 95% confidence interval [CI], −0.19-0.42; P = 0.46). Similarly, no relation between CD34+/KDR+ cell counts and M%AS or %VIH could be demonstrated. Moreover, the increase in CD34+/KDR+ cell counts over 6 months was unrelated to LLL (b = −0.15; 95% CI, −0.42-0.12; P = 0.28), M%AS, and %VIH. ConclusionsAlthough our study does not exclude a pathophysiologic role for CD34+/KDR+ cells in the formation of neointima, cell counts before percutaneous coronary intervention proved to be unrelated to LLL or intravascular ultrasonographically derived restenosis parameters in coronary BMSs at 6 months.