Background: Randomized clinical trials have demonstrated the superiority of drug-eluting stents (DESs) compared with bare-metal stents in reducing the need for revascularization and major adverse cardiac events (MACEs) in low-risk patients with single-vessel lesions. Rapid DES uptake has necessitated shifting the paradigm to active DES-controlled noninferiority study models with most studies using surrogate angiographic measurements to attain adequate statistical power. No previous prospective trial has specifically compared a new DES with an active-control DES in a high-risk patient population using primary clinical end points.
Objective: COSTAR II is designed to compare use of the investigational Costar stent (Conor MedSystems, Palo Alto, CA) with the Taxus (Boston Scientific, Maple Grove, MN) stent in single- and multivessel percutaneous coronary intervention. The primary end point is the clinical composite of MACE at 8 months supported by consistent results in the evaluation of 8-month MACE rates in the single- and multivessel cohorts and of in-segment late loss in a small angiographic substudy at 9 months.
Methods: A total of 1700 patients, 50% with single-vessel and 50% with multivessel disease, are randomized in a 3:2 ratio to receive either Costar or Taxus stent(s) in this prospective, multicenter, noninferiority study design. Because no prior data were available to determine control multivessel MACE rates, an imputed placebo statistical analysis plan incorporating a variable delta based on actually observed control DES MACE rates will be implemented. The results of COSTAR II will provide information about a novel coronary stent device as well as unique data regarding both control and test DES use in more complex "real-world" patients.