E3227. Comparison of Healthy Patients and Those With Lung Disease Using Hyperpolarized Xe-129 3D Single-Breath Chemical Shift Imaging
  1. Jack Yang; University of Virgina School of Medicine
  2. Jill Nehrbas; University of Virgina School of Medicine
  3. Steven Guan; University of Virgina School of Medicine
  4. Nick Tustison; University of Virgina School of Medicine
  5. Yun Shim; University of Virgina School of Medicine
  6. John Mugler; University of Virgina School of Medicine
  7. Jaime Mata; University of Virgina School of Medicine
3D single breath chemical shift imaging (3D SB-CSI) utilizes a combination of magnetic resonance imaging (MRI) with hyperpolarized Xe-129 (HP Xe-129) gas to generate spectroscopic maps of the lung. Participants inhale a volume of HP Xe-129 gas that diffuses into the lung parenchyma (tissue) and red blood cells (RBCs). As a result, spectroscopic maps can be generated to locate ventilation defects and physiological changes in different lung compartments. The purpose of this study was to quantify and identify physiological differences in the lung maps generated by healthy patients and those with idiopathic pulmonary fibrosis (IPF), cystic fibrosis (CF), or chronic pulmonary obstructive disease (COPD).

Materials and Methods:
A total of 16 healthy subjects and 29 subjects with lung disease (IPF [N=11], CF [N=13], and COPD [N=5]) were imaged with 1.5T MRI. Subjects laid supine and inhaled a volume of gas equal to 1/3 of their FVC, with a maximum volume capped at 1 L of isotopically enriched (83%) Xe-129 mixed with nitrogen gas. Xe-129 was polarized to ~35% using a commercial polarizer. Participants held their breath for <10 sec, and proton with either 3D SB-CSI or ventilation images were acquired. An RF coil tuned to the Xe-129 frequency was used. The 3D SB-CSI images were then postprocessed in MATLAB, with whole-lung averages computed.

Patients with CF and COPD had significantly more ventilation defects than IPF and healthy patients (p<0.001), which correlated with the FEV1 predicted (R=-0.74). CF and COPD subjects also had more regions of no ventilation than IPF and healthy subjects (p<0.01). IPF had more ventilation defects than healthy subjects (p<0.001). COPD subjects also had more defects than CF subjects (p<0.01). COPD subjects had the lowest RBC/gas ratio, 0.15±0.068AU. The RBC/Gas ratio for CF was 0.35±0.094AU, IPF was 0.28±0.061AU, and healthy was 0.39±0.079AU (p<0.001). The differences between healthy and COPD, and between CF and COPD subjects were most significant (p<0.001). The difference between healthy and IPF was also significant (p<0.01).

The results of this study correlate closely with the known pathophysiology of IPF, CF, and COPD, and demonstrate the utility of 3D SB-CSI to study lung disease. As expected, patients with lung disease have increased ventilation defects when compared to controls. Furthermore, COPD and IPF patients were found to have lower RBC/gas ratios than healthy and CF subjects, indicating impaired gas diffusion into the RBCs. Traditionally, diffusion through alveolar-capillary interfaces can be measured through the diffusion capacity of carbon monoxide (DLCO). When compared with studies that measured DLCO in diseased patients, we expected lower RBC/gas ratios in IPF and COPD since both have impaired gas transfer, due to increased pulmonary fibrosis and vasculature loss respectively. RBC/Gas ratios in CF subjects were similar to healthy subjects, which correlates with a normal DLCO reported in early-stage CF. As imaging techniques improve, 3D SB-CSI may be useful for monitoring disease progression and characterizing disease phenotypes.