Objective: The purpose of this study was to evaluate and compare the perceptibility of 75% and 95% in-stent stenoses with CT angiography and MR angiography using six stent types in a phantom model.
Materials and methods: Six different stent types were placed into tubes filled with contrast agent (ioversol or gadoteric acid), and nylon cylinders (8 mm diameter) bored in the central axis (2 and 4 mm) to mimic 75% and 95% stenoses were inserted into the stents inside the tubes. CT angiography (16- and 64-MDCT scanners using three different kernels at 120 and 140 kV) and MR angiography (1.5 T) were performed. On 2-mm coronal sections, signal intensities in the stenosed stents were compared with unstenosed segments. In addition, perceptibility of the residual lumen was assessed using a subjective score. Image analysis was performed by two experienced and blinded radiologists.
Results: Sixteen-slice CT angiography showed relative in-stent signal intensities of 72-87%, whereas 64-MDCT angiography showed relative in-stent signal intensities of 63-99%. Sixty-four-slice CT angiography showed nearly no difference between 75% and 95% stenoses in the subjective scores. The high-contrast kernel was superior to intermediate- and low-contrast kernels. MR angiography showed relative in-stent signal intensities of 57-98%. The presence of localized artifacts and resulting inhomogeneous luminal signal caused lower subjective perceptibility ratings than the objective score would suggest.
Conclusion: CT angiography was superior in the differentiation between 95% stenoses and occlusions. 64-MDCT angiography was superior to 16-MDCT (mean +/- SD, 83.0 +/- 2.9 vs 78.3 +/- 3.3; p = 0.006), especially with high-contrast kernels (89.7 +/- 2.1 vs 78.3 +/- 3.3; p = 0.001). For detection of 75% stenoses, MR angiography seems to be suitable subjectively, even though no statistical significance was found.