• Erika R Ward, Laurence W Hedlund, William C Kurylo, Charles T Wheeler, Gary P Cofer, Mark W Dewhirst, Lawrence B Marks, and Zeljko Vujaskovic.
• Department of Radiation Oncology, Duke University Medical Center, Durham, NC 27710, USA.
• Int. J. Radiat. Oncol. Biol. Phys. 2004 Apr 1; 58 (5): 1562-9.

PurposeTo assess the usefulness of hyperpolarized helium (3He) MRI, including apparent diffusion coefficient measurements, in the detection and evaluation of radiation-induced lung injury in rats.Methods And MaterialsFemale Fischer-344 rats were treated to the right lung with fractionated dose of 40 Gy (5 x 8 Gy) using 4-MV photons. Conventional proton (1H) and hyperpolarized (3He) MRI were used to image the lungs 3-6 months after radiation treatment. Apparent diffusion coefficient (ADC) maps of hyperpolarized 3He in the lungs were calculated using a nonlinear, least-squares fitting routine on a pixel-by-pixel basis. After imaging, lungs were processed for histologic assessment of damage.ResultsThe effect of radiation was time dependent with progressive right lung damage ranging from mild to moderate at 3 months to severe fibrosis with structural deformation at 6 months after radiation. There was a significant decrease in the apparent diffusion coefficient of hyperpolarized 3He gas in radiation-treated lungs. Areas of decreased ADC in the lungs correlated with fibrosis shown by histology.ConclusionThis is the first study to show that hyperpolarized 3He MRI can detect radiation-induced lung injury noninvasively. Reduced hyperpolarized 3He ADC values postradiation likely reflect reduced alveolar volumes associated with fibrosis of the interstitium. Future studies at earlier time points may determine whether this noninvasive imaging technique can detect lung damage before clinical symptoms. Development of this new approach of magnetic resonance lung imaging in the rat model of radiation-induced lung injury will increase the ability to develop appropriate algorithms and more accurate models of the normal tissue complication probability.

16 keywords...

hide…