Development of a porcine model for pancreatic cancer

N.A. Remmers, J.L. Cox, P.M. Grandgenett, J.A. Grunkemeyer, L.A. Rund, L.B. Schook, M.A. Hollingsworth, M.A. Carlson
American Association for Cancer Research Annual Meeting, April 1-5 2017, Washington D.C.

Pancreatic ductal adenocarcinoma (PDAC) remains a lethal disease with a 5-year survival rate of 4%. Without signifıcant advances in early detection and treatment, PDAC is estimated to become the second leading cause of cancer deaths in the United States by 2020. Many advancements in PDAC have come from murine models and retrograde studies of patient samples. A supplemental pig model would allow such studies as noninvasive image-guided technologies, radiation oncology, drug metabolism, surgical experimentation/training, and early detection screening to be more readily performed. Here, we sought to develop an orthotopic model of porcine pancreatic cancer using transformed pancreatic tumor cells to help address some of these issues. We isolated primary porcine epithelial cells from pancreatic ducts harvested from a normal pig. We isolated epithelial cells from fıbroblasts by serial dilutions and stained our fınal population with cytokeratin 19 (specifıc for pancreatic ductal epithelium), E cadherin, alpha-SMA, and collagen I to show that our population of primary cells were indeed of epithelial origin and not mesenchymal. We used these cells to generate 3 different tumorigenic cell lines: PGKP (contains mutated Kras and p53), PGKPS (PGKP SMAD4 shRNA), and PGKPSC (PGKPS p16 shRNA). We tested these cell lines using in vitro assays to test their tumorigenic properties. The PGKP cells outperformed the primary cells in proliferation, population doubling time, and soft agar growth with p-values 0.03. The additional hits in the PGKPS and PGKPSC cells generated more pronounced invasion through matrigel, migration rate, anchorage-independent growth, proliferation capacity, and population doubling times with p0.01 when compared both to primary cells and the PGKP cells. We used all three cell lines in a subcutaneous injection model in nude mice to determine if they could produce tumors in vivo. We injected 5x106 cells into the hind flanks of mice with groups of 10 mice/cell line. Of the 30 mice injected, 28 grew tumors that were 1.8 cm2 in size within 5 weeks. All of the tumors had large necrotic centers, were hypervascular, and mucous producing. Upon H&E analysis, all tumors displayed an undifferentiated morphology with mild desmoplasia. Representative tumor sections were stained for the same markers as used above in the primary cell isolates with the addition of vimentin. These stains indicate that our tumorigenic cells formed undifferentiated carcinomas with mild desmoplasia. Here, we demonstrate for the fırst time development of porcine pancreatic tumor cells that can induce tumors in vivo. The expression of vimentin in our tumor slides indicates that our cells may have the potential to recapitulate metastatic disease when orthotopically injected into the pancreas of syngeneic pigs. We feel that this model could greatly supplement the existing murine models of pancreatic cancer to help improve imaging, surgical, and future diagnostic techniques.