Drug discovery efforts around metastasis, which is responsible for 90% of cancer deaths, have been de-prioritized for the lack of both translatable in vitro and in vivo platforms. For the want of clinically relevant in vitro triaging, many molecules end up for screening in pre-clinical animal models.
However, currently available models, from induced to spontaneous metastasis do not reflect the complete biology or are limited by metastasis take rate, surgical complications, and longer experimentation timelines, often ranging from 4 to 6 months. Therefore, it is imperative to engineer better metastasis models that not only reflect holistic biology but are also time-sensitive and cost- effective. We have previously shown that cancer invasiveness in vitro is directly proportional to the plasticity ratio (PR; a ratio of mesenchymal to epithelial nature of a cell) at any given point of time.
We show here that lower PR promotes tumorigenesis, but not metastasis, and the latter increases by increasing PR. In an orthotopic xenograft model of colorectal cancer, we surgically implanted engineered cells that had high PR, in the cecal wall of immunodeficient NOD-SCID mice, thereby decreasing tumorigenesis but increasing metastatic spread. The cells were trackable by bioluminescent imaging to enable live imaging of both primary tumors and metastatic foci/lesions. At the end of 6 weeks, all the animals with engineered cells showed 100% mesenteric lymph node (mLN) invasion and 70% showed liver metastasis, compared with 12.5% animals with mLN invasion and 25% animals with liver metastasis in the wild type cells. A pilot study with repurposed drugs in this animal model showed a significant decrease in liver metastasis, after six weeks of compound treatment compared to the untreated animals. Our initial data suggests that this in vivo modeling paradigm can be developed into robust screening platforms of anti-metastatic therapies for multiple solid tumors.