Detection of stent fractures: a comparison of 64-slice CT, conventional cine-angiography, and intravascular ultrasonography

We evaluated the ability of 64-slice computed tomography (CT), conventional cine-angiography, and intravascular ultrasound (IVUS) to detect stent fractures under ideal conditions. Coronary stent fracture has been implicated as one of the mechanisms of stent thrombosis and, perhaps, in-stent restenosis. However, the preferred imaging modality in detecting fractures in coronary stents has not been well established. Four different types of commonly used coronary stents (Cypher, Taxus, Vision, Hepacoat) each with three strut fractures (Cypher, 5; Taxus, 5; Vision, 4; Hepacoat, 5) were nominally deployed in polyurethane tubes and imaged with 64-slice CT, conventional cine-angiography, and IVUS. For each stent type, an unfractured control stent was also imaged. Overall accuracy (84.1% vs. 73.9%), sensitivity (80.7 vs. 77.2%), and specificity (100% vs. 58.3%) for stent fracture detection was higher with 64 multislice CT compared to conventional cine-angiography. Stent fractures were not accurately detected by IVUS. Fracture detection by multislice CT was best when the stents were imaged at 45 degrees to the z-axis. Under ideal in vitro conditions, CT has a high accuracy when used to evaluate coronary stent fractures. The overall accuracy, sensitivity, and specificity of detecting stent fractures are lower by conventional cine-angiography. Stent fractures were not detected using IVUS.