RESUMO
PURPOSE: The extreme microscopic heterogeneity of tumors makes it difficult to characterize tumor hypoxia. We evaluated how changes in the spatial resolution of oxygen imaging could alter measures of tumor hypoxia and their correlation to radiation therapy response. METHODS AND MATERIALS: Cherenkov-Excited Luminescence Imaging in combination with an oxygen probe, Oxyphor PtG4 was used to directly image tumor pO2 distributions with 0.2 mm spatial resolution at the time of radiation delivery. These pO2 images were analyzed with variations of reduced spatial resolution from 0.2 mm to 5 mm, to investigate the influence of how reduced imaging spatial resolution would affect the observed tumor hypoxia. As an in vivo validation test, mice bearing tumor xenografts were imaged for hypoxic fraction and median pO2 to examine the predictive link with tumor response to radiation therapy, while accounting for spatial resolution. RESULTS: In transitioning from voxel sizes of 200 µm to 3 mm, the median pO2 values increased by a few mm Hg, and the hypoxic fraction decreased by more than 50%. When looking at radiation-responsive tumors, the median pO2 values changed just a few mm Hg as a result of treatment, and the hypoxic fractions changed by as much as 50%. This latter change, however, could only be seen when sampling was performed with high spatial resolution. Median pO2 or similar quantities obtained from low resolution measurements are commonly used in clinical practice, however these parameters are much less sensitive to changes in the tumor microenvironment than the tumor hypoxic fraction obtained from high-resolution oxygen images. CONCLUSIONS: This study supports the hypothesis that for adequate measurements of the tumor response to radiation therapy, oxygen imaging with high spatial resolution is required to accurately characterize the hypoxic fraction.
