ABSTRACT In order to improve cancer patient outcomes, the next generation of cancer therapeutics should combine targeted treatments with the ability to predict the likelihood of an individual’s tumor to benefit from a given treatment. The focus of this project is on advancing a personalized platform based on the individual’s tumor behavior which is determined not only by the cancerous epithelium but the entire tumor microenvironment (TME). The role of TME in disease progression or even therapeutic drug response is well established in variety of cancers with poor outcome including pancreatic ductal adenocarcinoma (PDA) with a five year survival rate of less than 10%. Here, combining 3D patient-derived organoids (PDO) and microfluidic technologies, we will develop a novel tumor-chip device with the potential to generate individualized treatment response predictions ex vivo. In Aim 1, using multiple tissue and functional read outs we will show that our multicellular platform recapitulates the tumor and its microenvironment. In Aim 2, we will show the utility of our system as an individualized ex vivo platform to functionally characterize the tumor response to conventional and novel therapies. Our long-term goal is to optimize patient-specific ex vivo drug response assay technology such that it may be used to rapidly, accurately, and reliably inform individualized treatment decisions. As a proof of concept, the project involves developing a personalized pancreatic cancer model; however this technology can be extended to other solid tumor types with rich TME (i.e., breast cancer) to improve molecular and/or cellular characterizations of cancer biology and therapeutics.