Impact of attenuation and scatter correction in estimating tumor hypoxia-related kinetic parameters for FMISO dynamic animal-PET imaging

Tumor hypoxia promotes tumor progression and reduces the efficacy of radiation and chemotherapy. A potentially important non-invasive hypoxia imaging technique is positron-emission-tomography (PET) with the radiotracer 18F-fluoromisonidazole (FMISO). FMISO binds to macromolecules and is trapped within cells in the absence of oxygen. The time course of this process may be imaged with dynamic PET, and a pharmacokinetic compartmental model has been used to evaluate the rate constants between different compartments in a tumor. In our Small-Animal-Imaging Facility, most animal-PET data are reconstructed without attenuation-correction (AC) or scatter-correction (SC). We have investigated how these correction factors affect the rate constant’s estimation in hypoxic tumor. A 4D dynamic digital PET phantom is created to simulate a 300-gram rat with a 5-gram xenografted subcutaneous human colorectal adenocarcinoma tumor, injected with a 2-mCi FMISO bolus via the tail vein, and imaged on our Focus-120 micro-PET for 90 minutes. The 4D dynamic image data is then forward projected to simulate the 4D dynamic sinogram. FORE+2DOSEM and FORE+FBP reconstruction protocols are used with and without AC&SC. The reconstructed dynamic images are then analyzed with the pharmacokinetic compartmental model and the estimated kinetic parameters are compared with the original one. It is found that the image-based plasma input function plays an important role in kinetic analysis. Present PET technology could not fully recover the sharp peak of the plasma input function, and makes estimated kinetic parameters differ from the true ones by 5.7∼8.8% with ACSC, and 10.4∼12.6% without ACSC.