Serial digital autoradiography with a silicon strip detector as a high resolution imaging modality for TRT Dosimetry

This study aims to investigate the possibility of implementing serial autoradiography using a silicon strip detector as an imaging modality in pre-clinical radionuclide therapy research, in order to study the effect of non-uniform uptake on absorbed dose distribution and biological response. Tumor tissues expressing CD20 (B-cell lymphoma) or carcinoembryonic antigen (CEA; colorectal cancer) were excised from animals injected with I-131-labelled anti-CD20 or anti-CEA antibodies and antibody fragments. The tumors were cryosectioned at 100 mu m and imaged using a real-time silicon- strip imager with a pixel-size of 50 mu m. Software was developed to correct for image artifacts and to realign the image sections into a volume by a two-step process with least square error and mutual information registration measures. The realigned volumes were convolved with beta dose point kernels to provide the dose rate distribution for I-131 and Y-90 at the time of sacrifice. Using these volumes, comparisons can be made between uptake and penetration of different antibodies and the dose rate uniformity of different radionuclides. Simulations performed using measured I-131 and I-125 energy spectra showed that energy separation with less than 5% error could be performed with 100 counts per pixel. Imaging and subsequent separation of a sample containing both I-131 and I-125 proved the possibility of simultaneous imaging of two targeting agents in the same tissue. Thinner tissue sections were also set aside and successfully used for H&E staining and immunohistochemistry to enable future comparison of uptake and dose rate in different cell-type populations in the tissue. This method successfully provides high-resolution activity and dose rate volumes and has potential for multi-labeling imaging and co-registration with histology. As a complimentary imaging modality it can aid in investigating the effect of non-uniform uptake. Optimization is still needed in both the sectioning protocol and realignment software.