Development of orthotopic mouse models for mid-low rectal cancer

Mid-low rectal cancer is one of the most common types of rectal cancer and has a poor prognosis. Surgery and chemoradiotherapy are the main treatments for early and advanced rectal cancer with an overall 5-year relative survival rate of only 56.9%. Development of novel antitumor agents is needed. Animal models of disease are indispensable for drug development. The most commonly used animal models of rectal cancer are established by inducing tumors by the subcutaneous transplantation, cecum or peritoneal injection, but not injection in the rectum. Their tumor microenvironment differs from that of rectal tumors in situ, which is hard to precisely simulate the occurrence and development process and drug response of human rectal cancer. In this study, we established orthotopic mouse models of mid-low rectal cancer with primary tumors originating from the rectum, including two models that could simulate the early and advanced stages of the disease, respectively. In the first model, the local primary tumor was restricted to the rectal area of the anal verge by rectal submucosal injection, its growth could be monitored with IVIS live imaging and magnetic resonance imaging. Histological analysis confirmed that the tumor originated from the submucosal layer and then invaded the muscular layer without metastatic tumors. This model may be useful for evaluating drugs for early mid-low rectal cancer in the future. The second model featuring a rectal primary tumor accompanied with abdominal metastases was established via rectal serosal injection. In this model, a large tumor formed at the rectal injection site and then metastasized to the abdominal cavity, reproducing the process from occurrence to metastasis of mid-low rectal cancer, and may be a good tool for the evaluation of drugs for advanced-stage disease. The injection methods used in these models do not require the aid of special colonoscopes, are simple and easy to operate, and have high tumor tumorigenicity and reproducibility. These results suggest that our staged modeling can provide targeted choices for preclinical drug research of mid-low rectal cancer at different stages.