Most cancers show some degree of family clustering. We can use this to estimate the contribution of genetic susceptibility to cancer risk. A small amount of this is due to rare, strong mutations in genes such as BRCA1 or 2. We assume that the rest is ‘polygenic': The combined effect of probably thousands of individual variants or rare mutations, each of small effect. Genome-wide association studies (GWAS) have identified common variants that account for 5-30% of the estimated polygenic component of susceptibility to different cancers. However, the greater part of the susceptibility remains unexplained: The so-called ‘missing heritability'.

In this session, a scene-setting introduction will be followed by three talks, each addressing a different unanswered question:

1. How can existing and potential future knowledge of genetic variants identified from GWAS and from resequencing be used to estimate individual risks, or to stratify the population into levels of risk for screening or prevention studies?

2. What is the ‘missing heritability' and how far are we likely to succeed in finding the genetic variants or interactions that are responsible?

3. What are the prospects for understanding mechanisms in a way that could lead to mechanism-based approaches to prevention by reduction of the increased risk? The polygenic model of susceptibility implies a distribution of risk across the population. The analogy is to think of the genetic variants as a hand of ‘good' or ‘bad' cards dealt out at conception. A woman at high risk of breast cancer, for example, has a ‘bad hand of cards' but another woman at similar risk may have an equally bad hand of different cards. What is the likely extent of heterogeneity in ‘bad hands' of genetic variants, and thus in mechanism? Could this complexity be reduced to a small number of networks or pathways, whose nodes might be therapeutic targets?