Project Abstract Discovering predictive biomarkers that better identify which patients will respond to cancer immunotherapies is a major unmet clinical need for oncology. Immunohistochemistry of antigen status is fraught with false positives and negatives for drug targets like PD-1 and PD-L1, and on this basis, measuring antigen levels with quantitative imaging may provide a more global assessment of drug target expression in all cancer lesions within a patient. These data may in turn empower more sophisticated and robust algorithms for identifying potential responders. With these considerations in mind, this project will develop a high sensitivity imaging tool targeting PD-L1 that is responsive to the special demands of human translation. For instance, we will develop a radiotracer based on a human recombinant Fab against PD-L1, which will both preclude the need for a costly humanization process and minimize the absorbed dose to normal tissues in patients. Moreover, we will radiofluorinate the Fab using a new chemoenzymatic technology that we recently developed and published. The radiolabeling technology may facilitate more rapid translation as it is site specific and it results in higher specific activity and radiochemical yield compared to the current gold standard in the field, N-succinimidyl-[18F]- 4-fluorobenzoate. In three specific aims, the Fab will be radiolabeled and characterized in vitro, proof of concept imaging studies will be conducted to show specific binding in models of cancer known to respond to anti-PD-1/PD-L1 therapies, and longitudinal imaging studies will be conducted to determine if the Fab can detect PD-L1 expression changes due to chemo or radiation therapy that can enhance the impact of anti-PD- 1/PD-L1 immunotherapy. In summary, the data from this project could significantly contribute to the community wide effort to develop better translational predictive biomarkers for important cancer immunotherapies.