Although big data is useful for generating drug candidates, and cell culture models give us a starting point on biological relevance, there currently is no substitute for the whole organism when testing the safety and efficacy of new drugs. Together, these laboratory approaches help us to understand the human disease, and are critical for testing new therapies before going into clinical trial. The whole system is required to measure tumor progression and the response to new therapeutic interventions. 

​

Many pre-clinical mouse models of cancer are problematic:

• One technique involves treating animals with carcinogens, but this classical approach causes liver damage, and leads to the formation of tumors throughout the body, instead of a defined tumor type.

• Another common technique involves implantation of human cancer cells into the flank of the animal, but these so-called ‘xenograft’ models do not replicate the unique characteristics of cancers that naturally arise in each tissue.

• Very often, human cancer cells are injected into animals with no immune system, so they cannot reject the graft, but this leads to the formation of tumors that do not reproduce the key characteristics of the human disease.

 

We use a more realistic pre-clinical mouse model to study therapeutic efficacy:

• Tissue-specific stem cells are isolated from adult mice and cultured in the absence of serum, to maintain their original genetic and biochemical characteristics.

• These cells are then modified to transform them into cancer cells, using mutations like those found in human beings with that cancer.

• The newly malignant cells are then implanted into the brains of mice from a wild-type background, with this ‘syngeneic’ model eliminating the possibility of host rejection. [Since we know that host factors, including immune responses, play a role in the growth of cancer cells, it is important to keep these factors in an animal model, as we test the growth rate of the tumor with various therapeutic interventions.]

• We implant the cancer-initiating cells directly into the tissue, so the tumor microenvironment in the animal model is similar to the human condition.

• The resulting tumors have been classified by trained clinical pathologists, in accordance with WHO criteria, and verified to replicate the histopathological features of the human disease, including high rates of proliferation and invasion into surrounding tissue.

 

With these excellent preclinical models of cancer, the time-course of tumor growth can be compared between treatment groups. To make careful assessments of drug efficacy, we use a blinded, placebo-controlled, endpoint-defined study design.  These preclinical studies are used to evaluate either  small molecule therapies or gene therapies, providing an excellent path to clinical trial after initial lead development.