Purpose: To report the systematic comparison of magnetic resonance imaging (MRI) with contrast-enhanced computed tomography (CT) for evaluating 8 different aortic stent-graft models.
Methods: MR angiography (MRA) was performed using a 1.5-T whole body system within 2 days of a CT examination (4 detector row scanner) on 8 patients with one of these stent-graft models: AneuRx, Endofit, PowerLink, Excluder, LifePath, Talent, Vanguard, or Zenith. Using a 4-point scale (maximum score 112 points), 4 independent readers (1 vascular surgeon and 3 radiologists) rated the impact of stent-related artifacts on the diagnostic quality of each imaging method for 28 parameters: length, diameter, collateral aortic side branches, stent-graft prostheses, and contrast. Each examiner also scored his personal diagnostic confidence with each stent-graft model.
Results: The scores for diagnostic confidence in the CT imaging were 4 points for each stent-graft, with the exception of the LifePath (3 points). The diagnostic confidence in the MR images was mainly poor, with a median score of only 1; however, 3 stent-grafts (AneuRx, Excluder, and Vanguard) received > or =3 points. The total scores for comparative assessment were significantly different (p<0.05) between CT imaging (111.5) and MR (58.5). CT studies of all stent-grafts received >101 points, while only 3 devices acquired >80 points (AneuRx, Excluder, and Vanguard). Bland-Altman analysis showed that the reliability of the 4 readers was higher using the CT method. The total assessment scores of the stent-graft systems were related only on the different imaging methods (p<0.0001) and not to the different readers (p=0.983).
Conclusions: CT and MRI are fast, reliable means of providing all relevant information for stent-graft surveillance. Of 8 different stent-graft models, only 3 could be adequately assessed by MRA. Therefore, the potential advantages of the MR technique (e.g., use of minimally nephrotoxic contrast media, lack of ionizing radiation) are available only to a small proportion of patients.