In 2011, cardiovascular disease (CVD) is the number one killer in the United States accounting for 31.3% or about 1 in 3 deaths (786,000 of 2.51 million). Heart failure represented 36.1% of the total number of CVD deaths or 1 in 9 total US deaths. A key prognostic indicator of heart failure is the presence and extent of myocardial fibrosis, which is found in all three main CVD etiologies including systolic dysfunction, non-ischemic cardiomyopathy, and valvular disease. Non-invasive myocardial tissue characterization to detect and characterize MF is highly sought after clinical tool because of its prognostic value. Current technologies include both contrast-based and contrast-free cardiovascular magnetic resonance (CMR) imaging approaches. Although contrast-based CMR methods are clinically in routine use, a significant portion of CVD patients (about 1 in 3) also have renal insufficiency requiring a contrast-free approach.
One promising contrast-free CMR approach is diffusion CMR, which is capable of unique tissue characterization by being sensitivity to the microscopic motion of water molecules in tissue. Specific to myocardial tissue characterization of fibrosis, it potentially offers the highest contrast (>50% signal change) between myocardial fibrosis compared to normal myocardium. More importantly, it has the potential to map myofiber architecture, which holds the promise of evaluation of novel myocardial therapy such as regenerative-based treatments. Despite the strong biological connection between diffusion CMR and myocardial fibrosis, in vivo diffusion CMR currently is not widely used in a clinical setting because of the major technical limitations of the technique which includes (i) bulk motion artifacts, (ii) distortion/susceptibility artifacts, (iii) limited spatial resolution (58mm3), (iv) limited spatial coverage, (v) long scan times (>20min), and (vi) low signal-to-noise ratio. Among the major technical challenges listed above for diffusion CMR, the greatest challenge is overcoming bulk motion because it is fundamentally tied to diffusion encoding. This dissertation aims to address some of these technical challenges to ultimately yield a clinically translatable in vivo diffusion CMR technique.
To address challenges (i), (ii), (iii), and (v), a novel diffusion encoding strategy (M2 gradient moment nulling) was developed in conjunction with a unique diffusion acquisition approach (diffusion preparation) allowing for 3D multi-shot readouts. The two developed technologies were initially demonstrated in healthy subjects to reproducibly yield diffusion-weighted images (DWI) free of motion artifacts and comparable estimates of apparent diffusion coefficients (ADC) when compared to conventional diffusion sequences.
The novel diffusion CMR sequence was then applied in a pre-clinical setting testing its ability to detect and characterize myocardial fibrosis in a chronic MI porcine animal model. The chronic MI porcine model acted as an excellent test scenario since myocardial fibrotic tissue in this model also has reduced bulk motion, which goes counter to the expected increase in estimated ADC. The novel diffusion CMR sequence able to detect myocardial fibrosis and yield an expected increased in ADC of fibrotic tissue. This lends credence to the new sequence’s robustness against bulk motion in the detection of myocardial fibrosis.
Lastly, the novel diffusion CMR sequence was applied in a pilot clinical setting to detect myocardial fibrosis in hypertrophic cardiomyopathy patients. Compared to the gold-standard sequence, it was able to yield high sensitivity, specificity, and accuracy in detecting myocardial fibrosis. Additionally, a simple threshold was only necessary to yield comparable characterization of the diffuse presentations of myocardial fibrosis.