Improving Noninvasive Therapy Evaluation of Stem-Cell-Based Treatment for Heart Failure with Diffusion Tensor MRI
Heart disease is the leading cause of death in the world. Advance noninvasive imaging techniques are sought out to guide and assess the efficacy of novel therapies to repair the heart such as exosomes from cardiosphere-derived cells (CDC-XO), developed by Eduardo Marbán, MD, PhD. One key advanced imaging tool developed by our group is Diffusion Tensor MRI (DTI) that can reveal the heart’s underlying fiber architecture. Applying DTI in the heart is technically challenging because of its inherent sensitivity to bulk motion, which in a moving organ like the heart is substantial. We aim to address these technical challenges, thereby yielding new, unique insight into evaluation of CDC-XO therapy.
To address the inherent motion sensitivity, we developed a second-order motion compensation gradient moment nulling (M2) that rephases the deleterious bulk motion-induced phase shifts. Without M2, these phase shifts will destroy the desired diffusion MRI contrast, resulting in signal voids. Further development was also made in the MRI readout employing a novel diffusion preparation scheme that allows for 3-D high-resolution.
This newly designed sequence was applied in 12 porcine with myocardial infarction in a randomized placebo controlled study to evaluate how the myocardial fiber orientations (HAT: helix angle transmurality) derived from DTI were affected by the CDC-XO treatment. Conventional MRI used to measure scar size (SS) and cardiac function (EF) was also acquired.
In the CDC-XO group (Figure 1), conventional MRI (LGE) revealed decreased SS (red arrow), and in the same region of reduced scar, DTI showed regrowth of the tissue realigning (blue to red) to surrounding tissue. Overall, the CDC-XO group demonstrated no significant change in HAT and a significant reduction in SS, representing a halt in disease progression. In the placebo group, both SS and HAT significantly changed in an adverse manner, reflecting disease progression. Receiver operator characteristic (ROC) curves (Figure 2) demonstrated that combined SS and HAT (0.93) had significantly improved prediction of the therapeutic endpoint EF in comparison to either SS (0.76) or HAT (0.82) alone.
Our preliminary results in porcine show that CDC-XO treatment is not only reducing scar but also re-growing viable tissue that has partially realigned with the normal tissue. This effectively results in an overall halt to the disease progression of myocardial infarction. Furthermore, we have demonstrated that DTI has added value in predicting the therapeutic end point of cardiac function when combining it with conventional MRI. With the development of a robust DTI technique for the heart, cardiologists at Cedars-Sinai can potentially characterize the progression of heart disease in their patients and monitor the use of the novel therapies developed in their laboratories.
We developed a robust in-vivo cardiac DTI sequence capable of imaging in a clinical setting by overcoming major technical challenges. We also demonstrated DTI has added value in improving the prediction of the therapeutic efficacy of stem cell-based treatments.