Electron Paramagnetic Resonance (EPR) spectroscopy is a technique similar to Nuclear Magnetic Resonance (NMR) spectroscopy, with specificity to detect paramagnetic species such as free radicals. Examples of paramagnetic species are molecular oxygen or those generated by exposing tissue to X-rays. Using non-toxic free radical spin probes which are infusible, it is possible to perform EPR imaging (EPRI) similar to the clinically used technique, Magnetic Resonance Imaging (MRI). While MRI non-invasively provides a spatial image of tissue water to represent anatomical image, EPRI provides spatial information (image) of the spin probe distribution in vivo. Since NMR spectra of water protons are relatively invariant in the body and the water content in most tissue is does not vary significantly, anatomical images with high spatial resolution can be obtained. However, since neither the NMR spectral intensity nor the spectral property (line width) vary significantly, it is not straight forward to obtain valuable functional/physiological information using MRI. On the other hand, the EPR spectral properties (line width and intensity) of paramagnetic spin probes are very sensitive to, and are modulated by important physiological properties such as tissue oxygen and redox status. Such EPR spectroscopic information, when spatially encoded in imaging experiments, provides functional/physiological images, which can be co-registered with anatomical images. The feasibility of implementing EPRI in the clinic depends on two factors: 1) biological effects of radiofrequency (RF) radiation, 2) toxicity of the paramagnetic spin probe. Based on the vast information available from the MRI experience on the biological effects of RF radiation, appropriate RF frequency range and power levels can be adapted to carry out EPRI without adverse biological effects. The toxicological properties of the paramagnetic spin probe has been studied at the cellular level as well as in vivo. The tests indicate that the free radical spin probes are well tolerated in vivo and EPRI experiments can be carried out at about 2000 times lower than the maximally tolerated dose. The physiological image information from EPRI will be useful in clinical oncology such as in radiation therapy where treatment of individual tumors can be optimized based on redox status and oxygen status of tumors. Solid human tumors are vascularly compromised and contain zones of hypoxia and become resistant to radiation therapy. Hence, spatial information on the hypoxic zones provided by EPRI could help in radiation treatment planning so that sufficient radiation doses can be delivered to the hypoxic zones to effectively sterilize the tumor. This has been one of the reasons for the Radiation Biology Branch to develop non-invasive functional/physiological imaging methods. The prototype instrument capable of performing small animal studies was available in 1997 and tested with phantom objects. Subsequently, imaging experiments were validated in small objects such as capillaries and the vasculature in the tail of experimental animals. The efforts in the past years (1996 ? 1998) were directed towards the design, construction and implementation of a small animal imaging system capable of providing well resolved images of free radicals in experimental animals such as mice after intravenous administration of an exogenous probe. The sensitivity of detection was optimized such that, at safe and non-toxic doses of the free radical probe, high resolution EPR images were feasible in mice. The efforts in the last year were directed towards developing experimental and computational strategies to obtain physiological information from the spatial EPR images obtained after iv administration. Since EPR images inherently contain spectral information with sufficient discrimination to extract physiological information from the local environment, two separate approaches have been taken to implement this.1) Spectral-spatial imaging: This approach is to obtain mapping oxygen in vivo. Spatial imaging gives intensities of the probe in a voxel without any information regarding the spectral content in the voxel. The EPR spectral properties are sensitively related to local physiology such as tissue oxygen, extraction of this information involves including a spectral dimension to spatial imaging. In EPR, spatial images are obtained by using static gradients of fixed magnitude and various projections are taken at different gradient orientations. Spectral information can also be obtained by varying the gradient strength in each gradient direction and projection data collected. Such experiments were validated in phantom objects containing spin probes under defined pO2. The images represented the local oxygen status with sufficient spatial resolution as well as spectral variation corresponding to the changes in the oxygenation levels. Subsequent experiments in in vivo models provided the feasibility of such experiments in live objects and make it now possible to study important issues related to tumor pO2. Relaxation weighted-oxymetric imaging: This approach will provide a three-dimensional map of oxygen distribution from the spatial images by treating the data computationally. Since the length of the time domain responses of the spin probe vary depending on the oxygen status, the rate of decay can be correlated to local pO2 status. Using a three-dimensional image data set, initially, the object can be visualized for the spin probe distribution. However, if, in each voxel, the signal can be weighted for the relaxation times and the image reprocessed, a relaxation weighted image can be obtained which can be used to indicate regions of normoxia vs hypoxia. This approach has been validated with defined phantoms and the images provided qualitative discrimination between voxels differing in oxygen status. Subsequent experiments were carried out in mice administered with the spin probe. The images were initially processed to get the overall spatial distribution of the spin probe. This same data set was then treated with the relaxation weighted imaging and the image reprocessed. The reprocessed images provide good discrimination in oxygen status between various organs. Such experiments will then provide several snap shots of oxygen maps in animal non-invasively. With such capabilities, the changes in oxygen status as a function of therapies can be investigated repeatedly.
