MRI evaluation of a stented abdominal aorta of a rabbit

Abstract

BACKGROUND MRI is an attractive non-invasive angiographic modality not requiring nephrotoxic contrast agents and damaging exposure to radiation such as traditional angiography and multislice CT. It has previously been shown to provide consistently accurate non-invasive information from the stented regions of peripheral vessels even under stenotic conditions [1-3]. The aim of this study is to show that phase contrast magnetic resonance (PC-MR) can be used to provide accurate quantitative information of blood flow in smaller stented vessels such as the coronaries, using an in-vitro model and a rabbit aorta model. The ultimate goal of the project is to evaluate in-stent restenosis in stented coronary vessels, ultimately describing the degree of stenosis based on blood velocity indices such as translesional pressure gradient and peak systolic velocity ratio (PSVR). 2. MATERIALS AND METHODS A vascular phantom was used to evaluate the signal intensity in 3.5 mm generic 54/46 NiTi stents (Boston Scientific, Natick, MA) implanted in a polyvinyl chloride tube. The phantoms were placed in the isocenter of the 3 T Philips Intera (Philips Medical Systems, Bothell, WA) for image acquisition. They were perfused with steady flow 130-160 ml/min using a flow loop with a Little giant pump (Oklahoma, OK,). The MR settings were optimized and were: slice thickness= 5 mm, pixel size= 0.39 mm, V enc= 200 cm/s, TE= 7 msec, TR= 19 msec, and flip angle= 10°. The clinical experiment consisted of 3.5 mm stents deployed in the descending aorta of New Zealand white adult rabbits. IACUC approval was received for the study. The rabbits were anesthetized, transported to the MR facility placed in the isocenter of the 3 T Philips Intera. The MR settings for the rabbit experiment were: slice thickness = 2 mm, TE= 4 msec, TR= 6 msec, pixel size= 0.29, V enc= 150 cm/s and flip angle = 10°. Flow velocities were calculated using the raw PC-MR images and the equation: V = V enc × φ v/180° (1) Where φ v is the phase shift, (180° to -180°), V enc is the encoding velocity and V is the velocity of blood flow calculated by PC-MR. 3. RESULTS The phantom studies provided excellent visibility in the stented region as shown by the modulus (a) and phase contrast (b) images in Fig. 1. Fig. 2 shows the velocity profile in the lumen voxel by voxel from the PC-MR images calculated using equation 1. There were 29 voxels inside the stented region that provided useful signal. The presence of the stent does not impair the signal from voxels in the center of the lumen, but impairs the ability to receive a signal from voxels close to the vessel wall and stent. The comparison indicates good agreement with the theoretical Hagen Poiseuille profile. The average velocity of 35 cm/s is in good agreement with the actual flow (160 ml/min). Similar results were observed in the clinical studies. Fig. 3 shows the modulus images and Fig. 4 the PC-MR images inside and outside the stented region in the rabbit abdominal aorta. The out of stent plane was taken just upstream of the stented area. Excellent lumen visibility is observed in the stented area. Table 1 shows the comparison from 3 experiments when the Region of interest (ROI) is increased from a few voxels at the centerline of the lumen, to the whole useful signal area in the whole lumen. Fig. 5 shows the time velocity history of blood flow throughout the cardiac cycle. This calculation considers all voxels within the lumen and as a result the blood flow in the stented region is underestimated in comparison to the unstented area in the vessel since little signal is received from voxels near the wall. 4. CONCLUSIONS Lumen visibility and measurement of blood flow velocity in stented coronary arteries may be possible by PC-MR. Velocity calculation is more accurate at the center of the vessel since there is signal loss from voxels near the stented wall. This shows promise that MR magnitude images and PC-MR can be used in the non-invasive assessment of stented lumen patency and in-stent restenosis.