The role of self-gated MRI in murine models of heart failure

Abstract

Heart failure is a major and escalating global health care problem. Basic research is aimed at discovering the complex mechanisms behind heart disease and the progression into heart failure, both for understanding the natural development of particular diseases, but also with the aim of finding new targets and strategies for therapy. As a tool in this setting, mouse models with specific phenotypes are common, displaying reproducible cases of disease and failure, allowing for detailed studies of global, local and cellular features of disease progression.

As a part of this process, non-invasive imaging techniques play an important role. Especially magnetic resonance imaging (MRI) has enabled an array of different methods to assess pathological differences in heart failure. Since an MR image with sufficient image quality will have to be acquired over several heart beats, special attention has also been given to the techniques used to detect the beating heart and acquire an image in the desired spot of the cardiac cycle, a procedure known as gating. A new technique for gating called self-gated (SG) MRI has been developed. In contrast to conventional cardiac MRI, which uses the electrical signals of the heart, the electrocardiogram (ECG), to gate the image acquisition, the SG technique obtains motion information from the MR signal itself, interleaved with or as a part of the main imaging sequence, and uses this information to retrospectively gate the image acquisition.

The aims of this thesis were to implement and evaluate SG MRI sequence in murine models of heart disease, and to use SG MRI in a longitudinal fashion to assess development of heart disease in mice.

The performance of ECG-gated sequences is affected by heart failure, in a similar manner, it has been proposed that the performance of SG MRI can also be deteriorated in hearts with impaired cardiac function. However, in this thesis, it was shown that SG MRI was not subject to a physiological bias, and could be used in different disease models without the performance being affected by the diversity in cardiac dysfunction. In a longitudinal fashion, the sequence, like MRI in general, revealed information on global cardiac structure and function that could be used to differentiate diseased and healthy animals. Additionally, it was shown that this particular SG sequence can be combined with both gadolinium and manganese contrast agents to assess infarction size.

In addition to testing the performance and usability of SG MRI, the sequence was a part of a multimodal regime in a study aimed at explaining the increased mortality seen in diabetic patients after infarction. In a diabetes infarction model in mice, we observed an increased rate of mortality, accompanied by a detrimental effect on cardiac Ca2+ handling in this group, leading to an increased risk of ventricular arrhythmias and death. High-intensity interval training showed to normalize the abnormal Ca2+ homeostasis observed after infarction in the setting of diabetes, providing evidence of the important role of exercise training in preventing death after infarction in diabetic patients.