Pancreatic ductal adenocarcinoma (PDAC) is the most intractable of human malignancies. Survival rate at 5 years is very low (less than 5%). Patients are most of the time diagnosed while the disease has already spread out and benefits from surgical resection is often cut off due to local recurrence and lack of efficient chemotherapy. Indeed, it is urgent to develop new tools to be used by clinicians in order to propose better management of the disease. This project is based on two specific characteristics of PDAC. First, the existence of a prominent tumor stroma compartment (desmoplasia) consisting of non-neoplastic myofibroblastic pancreatic stellate cells, vascular, nerve and immune cells surrounded by immense quantities of ECM, from far exceeding that found in most other tumor types. Second, the very specific and almost unique PDAC associated neural compartment remodeling (PANR) correlated with the intense neuropathic pain observed in this disease. PANR consists in a modification of nerve fiber structure/density and presence of tumoral cell within nerve fibers, a phenomenon called peri-neural invasion which is highly correlated to local recurrence of primary PDAC tumor. We and others hypothesized that, by dialoguing with cancer cells, non-tumoral stromal cells impact on PDAC tumor biology by modeling the own tumoral structure and fostering the tumor development. Regarding this concept we suppose that the intra-tumoral micro-environment has an active and efficient role on the neural compartment remodeling, its associated pain and local recurrence. By integrating preliminary and ongoing results from transcriptomic and proteomic analysis of human PDAC and endogenous mice model developing PDAC, as well as multiples cell lines co-culture studies, we aim to determine this “stromal PDAC signature” and its specific components involved in the PANR. Such improvement could permit to unravel novel diagnostic, prognostic, and therapeutic options for this deadly malignancy.