Project Summary In this project, we propose to correct [18]F-Misonidazole (FMISO) Positron Emission Tomography (PET) images to make PET be more accurate in measuring hypoxia. Hypoxia - a deficiency in oxygen - is a strong promotor of tumor radioresistance, yet there are no established clinical methods to accurately image oxygenation in vivo to improve treatment. Preclinical studies that acquired pO2 measurements using electron paramagnetic resonance (EPR) imaging have shown that targeting hypoxic tumor sub-regions with a boost of TCD95 after irradiating the whole tumor with TCD15, significantly increased tumor control when compared to the boosting normoxic tumor sub-regions with TCD95. So far this has been demonstrated in fibrosarcoma (p = 0.04) and mammary carcinoma (p = 0.013) tumor types. This shows that if hypoxic tumor sub-regions can be accurately imaged and targeted with a boosted dose – better known as dose painting – the overall dose deposition to surrounding healthy tissues can be decreased while improving tumor control. In recent years, clinical trials have used FMISO PET for radiation therapy dose plans with the hypothesis that targeting hypoxic tumor regions with a boost of radiation, as defined by FMISO uptake, would increase survival. These trials did not, however, show improvement when compared to treatment plans with and without a boost of radiation. These outcomes are in direct contrast to the EPR hypoxia studies previously mentioned. We hypothesize this is due to the imperfect binding mechanisms of FMISO to hypoxic cells, as well as general limitations of PET tracers being unable to reach hypoxic tumor regions due to their chaotic vasculature or acidosis. In response to the failure of using only FMISO PET to target hypoxia, we propose to include anatomical and physiological information of tumor vasculature and pH imaged with Dynamic Contrast Enhanced (DCE) MRI and chemical exchange saturation transfer (CEST) MRI, respectively. DCE-MRI and CEST-MRI will be used to correct the PET image to show hypoxic tumor regions, using EPR as the ground truth of hypoxia. The specific aims of the proposal are to develop and execute a study design to (1) evaluate discrepancies between tumor hypoxia regions as defined by FMISO PET and EPR, (2) identify tumor features in DCE-MRI and CEST-MRI that result in poor overlap between hypoxia as defined by PET and EPR, and (3) develop a correction PET/MRI algorithm to improve PET measurement of hypoxia, using EPR hypoxia measurements as the gold standard. Upon completion, aim 1 will demonstrate the need for correcting FMISO PET images. Aim 2 will identify correlations between vascular permeability and pH features and low or high PET uptake, where it might not match with hypoxia as defined by EPR. The third aim will take those features to create a decision tree algorithm to correct FMISO PET definition of hypoxia. We intend for this PET/MRI correction algorithm to eventually be implemented in clinical ...