ABSTRACT Pancreatic cancer is a devastating disease with a five-year survival rate below 10%. One of the main factors underscoring this low survival rate is the lack of effective clinical treatments. The chemotherapy gemcitabine is the most widely used agent for pancreatic cancer due to its well tolerated profile, even though treatment only marginally extends survival. In other cancers, gemcitabine can be very effective. The limited utility of gemcitabine in pancreatic cancer is thought to result from non-cancerous cells in the tumor creating a physical barrier limiting drug delivery. According to this model, chemotherapeutic agents are unable to penetrate the tumor and reach the cancer cells. We found that tumor associated macrophages (TAMs), a non-cancerous immune cell type, abundantly secrete the nucleoside deoxycytidine (dC), and this directly inhibits the cytotoxic activity of gemcitabine. In this research proposal, we will define how dC is made and released by TAMs and how dC is obtained and utilized by pancreatic cancer cells to promote gemcitabine resistance. We will also test the hypothesis that dC release is a TAM property that can be reversed by reprogramming the TAM phenotype. These studies will be accomplished using metabolomics techniques in combination with inhibitors of metabolism and signal transduction. In parallel, we will disrupt TAM-pancreatic cancer dC crosstalk in human patient-derived microtumor models and in syngeneic mouse models to determine the translation value. The clinical application of insights from these studies could have an immediate impact on patients. A means to predict gemcitabine response based on TAM properties and/or to enhance gemcitabine efficacy by targeting the TAM phenotype would increase the utility of this well-tolerated, mainstay treatment option for patients with pancreatic cancer.