Direct detection of myocardial fibrosis without the use of an exogenous contrast agent would overcome safety limitations and could potentially result in much shorter imaging protocols. Over the past years, there has been increasing interest in MR parameters that directly use changes in tissue composition in cardiomyopathies to generate contrast between healthy and diseased myocardium¹. Two promising techniques are Magnetization Transfer (MT) and T_1ρ MRI. It has been shown in an ex vivo rat model that magnetization transfer rate (MTR) is decreased in acute and chronically infarcted myocardium compared to healthy tissue². Furthermore, it has been demonstrated that the T_1ρ relaxation parameter can be used to detect chronic myocardial infarcts in patients, without the use of a contrast agent³. Goal of the present study was to compare the ability to detect and quantify myocardial scar tissue in a chronic infarct model using MT and T_1ρ mapping.

In vivo MRI was performed on a clinical 1.5 MR scanner (Philips Ingenia) in 3 anesthetized pigs, 4 weeks after 90 minutes occlusion of the left anterior descending artery (LAD). MT Imaging was performed by acquisition of two images, one with and one without an off-resonance saturation pre-pulse (8 hyperbolic shaped RF pulses, flip angle 800 degrees, duration 20 ms, total duration 160 ms, offset 1000 Hz). T_1ρ-mapping was performed using a 2D, single shot T_1ρ-prepared balanced fast field echo sequence. Five images were obtained with different spin-lock (SL) preparation times (SL = 0, 10, 20, 30, 40 ms) in diastole (amplitude 500 Hz). Readout parameters for both methods were: bandwidth/pixel = 723 Hz, TE/TR = 1.74/3.5 ms, resolution = 2 x 2 mm², slice thickness=8 mm, FOV =288x288 mm², flip angle= 35 degrees, SENSE=2.

After MT imaging and T_1ρ-mapping 0.2 ml/kg gadobutrol (Gadovist) was injected and 15 minutes after injection late gadolinium enhancement (LGE) MRI was performed. LGE imaging parameters were: TI = 200-250 ms, TE/TR= 1.18/3.6 ms, resolution = 1.25 x 1.25 mm², slice thickness = 5 mm, FOV = 320x296 mm², flip angle = 25 degrees, 30 TFE shots).

Analysis: T_1ρ-maps were calculated by pixelwise fitting of a mono-exponential decay function in Matlab (Mathworks). Pixelwise MTR-maps were calculated using the following formula: $$$MTR = \frac{M_{0} - M_{t}}{M_{0}}$$$ . The myocardium of the left ventricle was delineated by manually contouring the endocardial and epicardial border on the MTR and T_1ρ-map. MT and T_1ρ-values for infarct and remote myocardium in the animals were measured by manual segmentation of the infarct and remote area, guided by the corresponding LGE images.

The MTR- and T_1ρ-maps were successfully obtained from all animals (Figure 2). The MTR was significantly lower in the infarcted region (0.27 ± 0.01 ms), compared to remote myocardium (0.38 ± 0.01 ms), (p=0.008), (Figure 3). The T_1ρ relaxation time was significantly higher in the infarcted region (87.0 ± 1 ms), compared to healthy remote myocardium (56.4 ± 1 ms), (p=0.001) (Figure 4).We found a significantly lower MTR and a significantly higher T_1ρ relaxation time in the chronic infarct area, compared to remote myocardium. The decreased MTR and the increased T_1ρ relaxation time in the infarct area, suggest that there is a decrease in interaction between water and macromolecules in the infarct region, which can be detected with both techniques. The decrease in MTR in the infarct region is in accordance with other studies in literature^2,4,5. Likewise, the increased T_1ρ relaxation time suggest less interaction between the water protons and the surrounding tissue, resulting in a longer relaxation time. The formation of myocardial scar is associated with an increase of collagen macromolecules in the infarct region. Our hypothesis is, that other changes in tissue composition dominate the observed T_1ρ and MT contrast in cardiomyopathies. After an infarct there is a decrease in the number of viable cardiomyocytes with cellular macromolecules, and an increase of extracellular volume in the cardiac tissue. We hypothesize that the increase of extracellular volume is measured, instead of directly measuring the relaxation effect of the interactions with collagen fibers. Main drawback of MT Imaging is that the amount of Magnetization Transfer observed depends on the number and strength of the off-resonance saturation pre-pulses. However, the T_1ρ relaxation times are also sensitive to the experimental parameters, such as the spin-lock amplitude.In this study we have shown in an animal model of chronic myocardial infarction that both the magnetization transfer as well as the T_1ρ MRI can be used to detect chronic myocardial infarcts without the use of a contrast agent. Both contrast mechanisms indicate a decrease in water-macromolecular interactions.