Magnetic resonance imaging versus multislice computed tomography of thoracic aortic endografts

To compare the potential of magnetic resonance imaging (MRI) to multislice computed tomography (CT) for evaluating stent-graft placement in the thoracic aorta.

Susceptibility artifacts in 2 different stent-graft systems (Talent and Excluder) were evaluated in vitro in 2 angulations (straight and 33 degrees curved) using 3 different MRI gradient echo sequences (True FISP, 2-dimensional FLASH, and 3-dimensional Turbo FLASH). The size of the stent-related artifact was measured, and the relative stent lumen was calculated. In vivo stent demarcation, stent patency, and additional findings were determined in 13 patients (3 Talent, 9 Excluder, and 1 combined) and compared to CT findings.

In vitro, both endograft systems proved to be MR compatible, with the relative stent lumen value ranging from 82% to 100% in the straight configuration; in a curved model, the relative stent lumen value ranged from 56% to 92% with the 3D Turbo FLASH sequence, which provided the smallest susceptibility artifacts. The Excluder endoprosthesis caused significant signal inhomogeneity within the stent in a curved configuration. In vivo, MRI and multislice CT showed similar results, with CT imaging slightly superior in stent demarcation and MRI better in demonstrating thrombus. CT beam hardening artifacts were pronounced in the Talent system, while the Excluder device caused significant signal inhomogeneity within the stent on magnetic resonance angiography.

Multislice CT and contrast-enhanced MRI are fast, reliable means of providing all relevant information for surveillance of fully MR-compatible stent-grafts in the thoracic aorta.