Ischemic Heart Disease 265. T1 Weighted Myocardial Perfusion MR Imaging Using Blood Pool Contrast Medium SHU555C and Multi- Slice Hybrid Echo-Planar Sequence

Hajime Sakuma,' Kakuya Kitagawa,' Yasutaka Ichikawa,' Nanaka Ishida,' Kan Takeda,' Naoki Kato,' Satoshi Yoshise.* Mie University, 2-1 74 Edobashi, Tsu, Mie, Japan; 2Nihon Schering, 2-6-64 Nishimiyahara, Osaka, Japan Introduction: With recent advances in fast imaging techniques, contrast enhanced cardiac MR study can provide comprehensive assessments of myocardial perfusion, cardiac function and coronary anatomy in clinical patients. However, conventional gadolinium MR contrast agents demonstrate extracellular distribution, which makes it complicated to quantify myocardial perfusion because the contrast medium leaks rapidly from the blood pool to the interstitial tissue during first-pass. Ultrasmall superparamagnetic iron oxides (USPIO) agents have high T1 relaxivity and remain in the blood pool with a longer vascular half-life. These characteristics make USPIO agents potentially useful for assessments of myocardial perfusion, as well as for improving contrast on coronary MR angiography. However, the USPIO agent exhibits susceptibility effects as well, resulting in transverse dephasing and loss of signal which depend on a particle size of the USPIO agent and echo time of MR imaging sequence. SHU555C is a small-molecular-size subfraction of frucarbotran (Resovist). The smaller size of the iron oxide particles in SHU-555C substantialty reduces susceptibility effects and provides a larger T1/T2 relaxivity ratio. However, a feasibility of first-pass myocardial perfusion MR imaging using a bolus injection of the USPIO blood pool agent and a fast multi-slice perfusion MR sequence has not been demonstrated. Purpose: In the current study, T1-weighted first pass myocardial perfusion MR images were acquired after bolus injection of SHU555C using a cardiac MR imager. The signal to noise ratio (SNR) and contrast to noise (CNR) were also assessed on 3D coronary MRA with navigator echo and fast cine MR images of the left ventricle before and after SHU555C administration. I Methods: MR studies were performed in six beagle dogs. The study protocol was approved by the institutional animal care and use committee. The animals were anesthetized and mechanically ventilated with a respirator. Each dog was placed supine on a Plexiglas holder and inserted into a head coil in the magnet. MR images were acquired with a 1.5T cardiac MR imager (Signa CV/i, GE Medical Systems). After acquiring scout images, MR images were acquired as follows: (1) precontrast 3D coronary MR angiography, (2) pre-contrast breathhold cine MR imaging, (3) first myocardial perfusion MR imaging, (4) second myocardial perfusion MR imaging after 15minutes delay time, (5) post-contrast 3D coronary MR angiography, (6) post-contrast breath-hold cine MR imaging. Total MR imaging time excluding animal preparation was approximately 1 hour. Myocardial perfusion MR images were obtained with a hybrid echo-planar sequence (TR/TE = 6.9/1.4ms, flip angle = 20 deg, ETL = 4, FOV = 32cm, Matrix = 128 X 128, slice thickness = 10mm). Interleaved notched saturation was used to optimize both T1 contrast (TI = 180ms) and number of slice locations. Seven short-axis images of the left ventricle were acquired every other heartbeats. Two sets of contrast enhanced perfusion MR images were acquired in order to assess feasibility of rest and stress perfusion MR assessments. In four animals, twenty micro-mol Fekg of SHU-555C was injected into antecubital vein as a compact bolus (<lml) for each perfusion MR acquisition, followed by lOml of saline flush at a rate of 3mYsec with a power injector. In two animals, O.lmmol/kg of Gd-DTPA was injected for comparison. Coronary MR angiography of the left main coronary artery and left anterior descending artery was acquired with a respiratory gated 3D MRA sequence using navigator echo (TR/TE = 5.6/1.5ms, flip angle = 30deg, FOV = 30 X 22.5cm, matrix = 256 X 160 X 14, slice thickness = 2mm). Breath-hold cine MR images were obtained on short axis and vertical long axis planes of the left ventricle, using a prospectively triggered fast gradient echo cine sequence with k-space segmentation (TR/TE = 6.7/2.3ms, flip