Pancreatic cancer patient-derived xenograft tumors for translational studies in precision cancer medicine

Abstract public domain (NIH RePORTER)

Two major transcriptional subtypes of pancreatic ductal adenocarcinoma (PDAC), Classical and Basal-like, are presently the most powerful predictors of patient outcomes. While three-quarters of patients harbor Classical tumors, this subtype is grossly underrepresented in cell lines. Thus, there is a critical need to develop experimental tools that represent the clinical spectrum to test emerging therapies. Our preliminary experiments demonstrate that patient-derived xenograft (PDX) tumor models recapitulate the distribution of Classical and Basal-like PDAC observed in patients. Examining the effects of FOLFIRINOX in these PDX models revealed a preferential response of Classical tumors to FOLFIRINOX than Basal-like tumors. Moreover, we have identified four subtypes of mouse-derived tumor stroma that differ in the levels of myofibroblastic cancer associated fibroblasts (myCAFs) and sensitivity to inhibitors of TGFβ signaling. Based on these exciting preliminary results, our central hypothesis is that PDX tumors recapitulate the heterogeneity of PDAC that contributes to the variable treatment response of patients. Experiments in this proposal will employ a unique collection of molecularly annotated low-passage PDX tumors, NanoString GeoMx spatial transcriptomic profiling, and high-resolution analysis of extracellular vesicles to define molecular subtypes of tumors that respond to standard of care and investigational therapies. Three specific aims are proposed to test the central hypothesis: 1) Define the molecular subtypes of PDAC that are responsive to standard of care chemotherapy, 2) Elucidate the contribution of stromal subtypes to chemoresistance, and 3) Identify biomarkers on extracellular vesicles for the molecular subtyping of PDAC. In the first aim, orthotopic PDX tumors will be evaluated for response to FOLFIRINOX and gemcitabine + Abraxane using doses and regimens that mimic the clinical setting. Transcriptional profiling will be performed to define the mechanisms that contribute to acquired resistance to chemotherapy. In the second aim, spatial transcriptomic studies will ascertain the extent the stromal subtypes identified in our PDX tumors reflect stromal composition of the matched patient tumors. Employing PDX models with stroma subsets that diverge in their myCAF content, orthotopic tumor studies will be performed to determine the contribution of TGFβ signaling in mediating resistance to FOLFIRINOX therapy. For the third aim, single-extracellular vesicle analysis of plasma from PDX tumor-bearing mice will define biomarkers of PDAC and stroma subtypes, and the clinical application of these biomarkers will be examined using archival patient plasma samples. The research proposed in this application is innovative because it considers the impact of both cancer and stromal heterogeneity in the response of pre-clinical models to therapy. The proposed research is significant because it will provide strong scientific evidence for the rational use of the PDX platform in the development of new therapeutic strategies for PDAC.

Public health relevance public domain (NIH RePORTER)

NARRATIVE Despite advances in chemotherapy and surgical techniques, pancreatic cancer remains the third leading cause of cancer-related deaths in the United States. This proposal characterizes the use of patient-derived xenograft tumors to model the heterogeneity of pancreatic cancer. The insight gained from these studies will reveal the mechanisms that contributes to the variable treatment response of patients and provide for a rational use of tumor models to investigate novel therapeutics.