We are evaluating genomic risk predictions for prostate, breast, colorectal cancers and melanoma. This includes well-established existing risk predictions as well as new risk predictions developed within the Study. Our particular focus is on so-called genomic risk scores (often called polygenic risk scores or PRS), which combine information across hundreds or thousands of genetic variants. We are assessing how well such scores can predict cancer risk, on their own or combined with other risk factors for cancer (such as age). Our work considers different ways of assessing risk (e.g. risk relative to population average, or absolute risk of cancer diagnosis within 5 years) and a range of important performance measures for the risk predictions.
Our analyses use information from the latest global studies that examine genetic variants for disease risk (so-called genome-wide association studies), as well as data from large cohorts. This includes the new genomic data we have generated for the 45 and Up Study, combined with existing data on 45 and Up Study participants. In collaboration with teams at QIMR Berghofer and Cancer Council Victoria, we are also analysing data from two other major Australian cohorts: QSkin and the Melbourne Collaborative Cohort Study. For a key international comparison, we are also evaluating risk predictions in the UK Biobank cohort.
The risk prediction analyses within the Australian Cancer Risk Study are primarily led by Dr Hamzeh Mesrian Tanha and Mr Philip Ly, supervised by A/Prof. Julia Steinberg (all three based at the Daffodil Centre), and with key collaborators including Prof. Anne Cust (Daffodil Centre), Prof. David Whiteman and A/Prof. Matthew Law (QIMR Berghofer), A/Prof. Robert MacInnis and Prof. Roger Milne (Cancer Council Victoria), and their teams.