EACR 2016
Polygenic risk and the targeted prevention of cancer
Prof Paul Pharoah - University of Cambridge, Cambridge, UK
I was talking today about the potential for polygenic risk in personalising cancer screening programmes, with particular focus on personalising breast cancer screening by mammography.
What do you mean by polygenic risks?
In the last ten years we have identified nearly a hundred variants across the human genome that are associated with differences in risk between individuals. These are variants that you can inherit that affect your risk of cancer, in this case breast cancer, and because these variants can be common in the population some people have multiple variants and so polygenic risk is the risk associated with carrying lots of different variants.
Why did you choose to make breast cancer your focus?
Three main reasons, the first is that it’s my area of research interest, so I know the most about it, but breast cancer is particularly relevant for this. One is because it’s the cancer where we’ve found the most genetic variants associated with risk, so the potential for polygenic risk stratification is perhaps the greatest. Secondly, there’s an established breast cancer screening programme in the NHS and also in many other health systems and so the potential for stratified risk screening is perhaps more relevant, especially as there has been a big debate in the last twenty years about the value of breast cancer screening by mammography.
What clinical benefits might your research bring?
The clinical benefits are that we know that breast cancer screening and other cancer screenings are associated with benefits and harms. The major benefit is a reduction in cancer specific mortality but the harms can be quite serious in that there can be over-diagnosis, which is the diagnosis of cancers that would otherwise never have occurred in somebody’s lifetime, and over-treatment, treatment of those cancers that is effectively unnecessary. The idea is that if we could identify groups of the population that were at higher or lower risk than average the benefit to harm ratio of screening might be improved. We haven’t established scientifically that that’s true, it’s just something that has potential and needs a lot more research in order to show that it really works. My research in the last ten years has focussed on genome wide association studies to identify the risk variants for cancer. Genome wide associations have been particularly successful with the advent of chip genotyping and coupled together with our better understanding of the architecture of the human genome. Then using statistical modelling to put those data together to develop polygenic risk models.
