A71. Identification of scarred myocardium using first pass MRI and delayed enhancement MRI

J. Wang,B. Xiang,M. Grewul,H. Liu,G. Li,M. Jackson,B. Tomanek,R. Deslauriers,G. Tian
DOI: https://doi.org/10.1016/j.yjmcc.2006.03.411
IF: 5.763
2006-01-01
Journal of Molecular and Cellular Cardiology
Abstract:Pathological changes of scarred myocardium greatly affect signal intensity of its contrast-enhanced magnetic resonance imaging (MRI). This is due to scar-associated changes in microcirculation and distribution volume for contrast reagent. Therefore, the present study was to determine whether scarred myocardium could be identified using contrast-enhanced first-pass MRI and delayed enhancement MRI. Chronic myocardial infarction was induced in pig hearts by occlusion of two diagonal branches of the left anterior descending (LAD) coronary artery. Four weeks after the occlusion, pig hearts were excised and perfused in a Langendorff apparatus. An MRI contrast reagent, Gadolinium-diethylenetriamine pentaacetic acid (Gd-DTPA), was injected into as a bolus into aortic perfusion line. Its first pass was followed by T2* weighted imaging. T1 relaxation times were then measured continuously for 30 minutes using nine inversion recovery times. At the end of experiment, hearts cut into 5 mm thick slices and stained with TTC for measurement of infarct size. T2* imaging showed that wash-in slope of Gd-DTPA was similar in both scarred and normal myocardium. However, wash-out slope of Gd-DTPA was significantly greater in infarct region than in normal regions, suggesting that less contrast reagent diffused out of the capillaries in infarct region during its first pass. This may be an indication of decreased capillary permeability. It was also found that T1 relaxation times measured during 30-minute contrast perfusion were significantly shorter in infarct than in normal myocardium. This suggests that distribution volume for the contrast reagent was still significantly greater in chronic infarct region than in normal region. Our results demonstrate that both T*- and T-weighted imaging techniques can be used to delineate chronic myocardial infarction.
What problem does this paper attempt to address?