Skip to 0 minutes and 14 secondsAs we have seen, cancer is a disease of the genome caused by the accumulation of many mutations. But importantly, not all tumours share the same genetics. So one breast cancer, for example, may be very different genetically to another. Mutations in a few key genes are very common, but even if a gene is mutated in only a small proportion of patients, it might very important for the growth of their tumours. This is leading to what's known as genomics-guided personalised medicine. The International Cancer Genome Consortium was set up by researchers wanting to catalogue the different genetic changes that occur in tumors of all different types. This knowledge is now the basis for lots of subsequent work towards the goal of personalised therapy.

Skip to 1 minute and 0 secondsIt allows us to understand which faults drive cancer and to understand how genetics can affect the success of treatment. It can also provide information on new targets for treatment. And all this information can be used in the fight against cancer, and particularly, hard-to-treat cancers. One of those hard-to-treat cancers is pancreatic cancer. Over the past 40 years, we've seen great improvements in survival in many different types of cancer, but pancreatic cancer is lagging behind, and we have a great deal of work to do. Pancreatic cancer is the fourth leading cause of cancer-related death in the Western world. Every year in the UK there are 7,000 cases diagnosed, and the survival rate is less than 5%.

Skip to 1 minute and 47 secondsIn most cases, the tumour is too advanced for surgery. This is partly because of anatomical location, and partly because patients present when the disease is already very advanced. The tumours themselves are highly resistant to chemotherapy and current treatments don't work. Even patients who do have tumours that can be removed surgically almost always recur with metastasis because the tumor is so aggressive. Perhaps the most powerful tools available now to study cancer are Genetically-Engineered Mice Models, or GEMMs. Using new smarter models-- which I'll talk about briefly in a moment-- researchers can determine whether mutations found in human tumours by efforts like the ICGC are driver or passenger mutations, ie, do they really matter for tumour initiation or survival?

Skip to 2 minutes and 38 secondsOr do they accumulate as tumors become more genomically unstable, but play no part in tumour progression. These models also provide clinically relevant models for testing personalised therapies because it's easier to kill cancer cells in a plastic dish, but, unfortunately, it's much harder to do in a mouse or human. These models can also be a source of cell lines for high throughput in vitro work to find therapies for tumours when it's not clear from the cancer genome what therapy to use. Thanks to new techniques, GEMMs have become incredibly advanced. It's now possible to engineer mice to express an enzyme called Cre within a specific set of cells or within a specific organ, in this example, the pancreas.

Skip to 3 minutes and 27 secondsCre acts as molecular scissors wherever it finds its target sequence. We can also engineer mice that will express a mutant oncogene, but only once this enzyme, Cre, has cut out a stop sequence. When these mice are crossed together, the result is a mouse in which the mutant oncogene is expressed, but only in the target organ. And importantly, the gene is expressed normally and not overexpressed, resulting in models much closer to the human disease.

Skip to 3 minutes and 59 secondsThese models can be used in two ways. First, to validate the mutation of interest. Is this mutation important in driving cancer? And second, to test new therapies in preclinical trials. For example, high tumour levels of a protein called mTOR are found in about 15% of human pancreatic cancers and when you look at the survival graph of patients, high levels of this protein are associated with much poorer survival. Mouse models can answer the first question of whether or not mTOR mutation is important for driving cancer. When mice are genetically engineered to have high levels of mTOR in the pancreas, we see that those mice with high mTOR develop tumours very rapidly.

Skip to 4 minutes and 46 secondsSo that tells us that mutations in mTOR are important for driving cancer. And therefore, this gene and its protein might be a very good drug target in the small subset of patients who overexpress it. Now, this same model can then be used to answer the second question. Will inhibiting mTOR work specifically in these tumours? And again, we can see the answer to this question here. Those tumours with high mTOR respond much better to mTOR inhibitor and that suggests to us that we have a real possibility of improving survival for this subset of patients. Researchers and clinicians are also figuring out smarter ways to see quickly if a treatment is working.

Skip to 5 minutes and 32 secondsImaging using various techniques, including ultrasound, PET, and MRI scanning offers a window into the body. It can be used to visualise cancer growth and biology, provide screening and diagnostic tests, and monitor responses to personalised treatments. Preclinical imaging using GEMMs is very useful in that it is translatable to patients and allows testing of imaging protocols that may be used to assess responses to personalised therapies. So for example, if a drug is known to be working when proliferation is reduced, it's possible to use an imaging probe that lights up only in proliferative areas. If that signal goes away, the drug is working and imaging this allows clinicians to make decisions on treatment very quickly and without the need for invasive tests.

Skip to 6 minutes and 27 secondsSo where are we going with genomics-guided personalised medicine? We have a wide variety of tools at our disposal. These include identifying and advising people at high risk genetically, and using a wide range of screening and diagnostic tests, including genetic screening. Advances in genetic screening mean that now we are not that far away from achieving this routinely. With the information that brings, clinicians would be able to choose from a wide range of effective personalised treatments already available, and they would be able to make use of techniques like imaging to tell quickly if a treatment is working. Using this linked-up approach, the ultimate goal is to make cancer a manageable disease that can be detected and treated.

Mouse models in pancreatic cancer research

Dr Jen Morton gives us insight into how mouse models are being used in the fight against pancreatic cancer.

Share this video:

This video is from the free online course:

Cancer in the 21st Century: the Genomic Revolution

University of Glasgow