Gastrointestinal Cancer

Background and Aim: Colorectal cancer (CRC) is the third most common cancer worldwide and affects about 15 000 Australians every year. When the cancer is detected early, it is often resectable; however, about 50% of the patients will experience a relapse and succumb to metastatic disease. Despite many advances in cytotoxic and targeted therapies in recent years, the development of resistance remains a challenge. Unfortunately, standard cell line xenograft models do not accurately recapitulate the heterogeneous nature of the human disease. In addition, the use of standard cell lines to determine the efficacy of new drug targets in vitro has revealed inconsistent results when compared with patient treatment response. It is becoming increasingly clear that the lack of appropriate model systems to accurately predict the response of anticancer drug targets limits most research opportunities. Our aim was to establish and characterize a series of CRC patientderived xenografts (PDXs) as a renewable resource to accurately represent the human disease for use in anticancer drug studies. Methods and Results: PDXs are physiologically relevant preclinical models where the patient tumor heterogeneity, genetic profile, and gene expression patterns are retained. We generated our CRC PDXs by engrafting fresh human tumor tissue into immunocompromised mice directly after patient surgery. HE staining, immunohistochemical staining, western blotting, and genomic profiling have been used to compare the expression signatures of the implanted tumors (including after serial transplantation) with the original patient tumor. To date, we have successfully generated 16 PDXs representing different stages of primary colon tumors from different patients and have archived 28 patient samples for future use. Our tumor engraftment success rate following subcutaneous implantation is 66%, with preliminary histopathological analysis highlighting that the features of the original patient tumor are retained in the PDXs after serial transplantation. We are currently working with four PDX lines that show variable tumor development rates, ranging from 1 to 5 months. From these, secondgeneration PDX tumor lines are being generated with acquired 5fluorouracil resistance, a standard-of-care chemotherapeutic drug for patients with CRC, to begin to understand the mechanisms underlying chemotherapeutic resistance. Conclusion: Our established PDXs will serve as a platform that will allow us to do in vivo and in vitromodeling of CRC and study the effects of novel targeted therapies. They are also an excellent resource for studying mechanisms of resistance in CRC. The outcomes of patients with two hepatocellular carcinoma lesions diagnosed simultaneously or sequentially RV APOSTOLOV, ZS ARDALAN, NM KAM, K PATWALA, N KUTAIBA, AG TESTRO, PJ GOW *Liver Transplant Unit, Radiology Department, Austin Health, Melbourne, Victoria, Australia