Preclinical Hyperpolarized 129Xe MRI: Development, Applications, and Dissemination
Hyperpolarized (HP) gas MRI is emerging as a powerful, non-invasive method for imaging lung function. MRI of HP 129Xe and 3He was first introduced in small animals and was soon followed by its clinical implementation. 3He was preferred for imaging since it was more straight-forward to hyperpolarize in large volumes and had favorable magnetic resonance properties for high-resolution. However, the scarcity and high cost of this isotope has driven a transition to abundantly available 129Xe. This transition has stimulated a lot of clinical 129Xe MRI research. 129Xe ventilation, barrier-tissue uptake and red-blood-cell (RBC) transfer can now be depicted separately and three-dimensionally by exploiting xenon solubility and large chemical shifts in different pulmonary micro-environments. With this powerful capability, this technique has found clinical application across a broad range of lung diseases.
As clinical implementation progresses, it has become increasingly important to test these methods in well-controlled animal models. Such preclinical studies enable the testing of experimental drugs, tracking of disease progression by longitudinal imaging, validation against histology, and provide a platform to rapidly develop and validate novel methods of image acquisition and analysis. However, among the ~20 centers worldwide that have HP gas MRI capability, only 5 have demonstrated the capability to conduct preclinical studies. Preclinical 129Xe MRI is challenging owing to extensive requirements of animal handling, reliable delivery of polarized gas, and the challenges of high-resolution multi-breath imaging. While some applications for HP gas MRI in small animals have emerged, these have mostly been with 3He and the handful of work on 129Xe has been limited to 2D imaging. As was the case in the clinic, there is now an equally urgent need to drive the transition from 3He to 129Xe in the preclinical setting, demonstrate sufficient image quality, and rapidly discover new applications.
The objective of this work is to establish a robust and comprehensive 129Xe MRI infrastructure to investigate rodent models of lung disease, and to lay the foundation for the reverse-translation and dissemination of this capability. To this, the work in this thesis describes several milestones toward establishing routine, high-resolution 3D 129Xe MRI of gas-exchange on a modern preclinical imaging platform.
First, we established routine 3D 129Xe MRI in mice on a GE 2 Tesla magnet. Through rigorous optimization of multi-breath image acquisition strategies with constant-volume ventilation, we demonstrated high-resolution imaging of 129Xe gas- and dissolved-phases in mice with 156-µm and 312-µm isotropic resolution. In addition to imaging, we also comprehensively characterized 129Xe spectroscopic lineshapes in mice. The in vivo resonances of 129Xe are sensitive to micron-scale changes in lung physiology, but have been analyzed and reported inconsistently and inaccurately in the literature. Using innovative spectroscopic acquisition methods and robust fitting techniques, we introduced methodology to identify an accurate 129Xe reference frequency in vivo, and characterized the many dissolved-phase resonances that are arise as 129Xe is transported to distal tissues in the thoracic cavity.
Until this point, animal studies using 129Xe MRI required sacrificing the animal upon completion of imaging, owing to the requirement of tracheostomy to ventilate the rodent with HP gas. Also, our experiments could only be conducted on a single 2 Tesla magnet, because the ventilator and physiological monitoring system was hard-wired to this scanner. In order to address these limitations, we built a new ventilator with integrated physiological monitoring with a focus on portability, minimizing cost, and compatibility for longitudinal imaging. The portable and integrated ventilator made possible our first dissemination of preclinical 129Xe MRI—to the University of Oxford, UK.
Our robust 129Xe MRI and spectroscopy protocol was deployed to investigate two key mouse models of lung disease at 2 Tesla: lung cancer and invasive pulmonary aspergillosis (IPA). In lung cancer, longitudinal 129Xe MRI revealed tumors on 1H MRI and histology, and severe ventilation and gas-exchange defects. 129Xe spectroscopy additionally revealed a robust signature of cancer-associated cachexia. 129Xe MRI in IPA also revealed significant and complex ventilation and gas-exchange defects, which was bolstered by spectroscopic features.
Having a portable ventilator enabled experiments to be carried out at other magnets at our center. Since modern preclinical magnets now operate at high field strengths, we established preclinical 129Xe MRI on a Bruker 7 Tesla magnet at our center, to facilitate its broader dissemination. This is the most widely available preclinical imaging platform with an installed base estimated to exceed 500 units. This transition involved a comprehensive characterization and optimization of the noise floor of the system to maximize SNR, and developing several new image acquisition strategies to rapidly image short-lived 129Xe signal at 7 Tesla (dissolved-phase T2* ~0.5 ms). On this platform, we developed a robust 129Xe MRI protocol for quantitative gas-exchange mapping in rats, identical to that used by our clinical program to facilitate translation/reverse translation.
Finally, we used the new 7 Tesla platform to investigate the monocrotaline (MCT) rat model of pulmonary arterial hypertension (PAH). This model was chosen for 2 reasons: first, it provided a unique opportunity to deploy 129Xe gas-exchange MRI in a model that is translationally relevant to current clinical investigations; second, there is also a dire need for non-invasive assays to elucidate the pathogenesis of this disease in the lung, and to enable early detection. In this study, we comprehensively characterized the imaging and spectroscopic markers of this disease and validated results with histology. 129Xe MRI revealed significantly reduced signal in RBCs, as well as interesting abnormalities in the barrier-uptake and gas-phase signal that were consistent with the pathobiology of this disease model. This is the first study to have demonstrated the potential of 129Xe to be a valuable tool for assessing rodent models of pulmonary vascular disease.
This body of work has thus established a robust preclinical 129Xe MRI framework that can be routinely used for imaging across field strengths, vender platforms, rodent species, be translated/reverse-translated to/from our clinical program, and be disseminated to other centers. We have also demonstrated the potential of this imaging platform to identify different disease signatures in several clinically-relevant rodent models. We anticipate that this work will provide a fundamentally new capability to accelerate progress in lung imaging research.
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.
Rights for Collection: Duke Dissertations