Development of a Cancer Risk Prediction Tool for Use in the UK Primary Care and Community Settings

Artitaya Lophatananon, Juliet Usher-Smith, Jackie Campbell, Joanne Warcaba, Barbora Silarova, Erika A. Waters, Graham A. Colditz, Kenneth R. Muir

Several multivariable risk prediction models have been developed to asses an individual's risk of developing specific cancers. Such models can be used in a variety of settings for prevention, screening and guiding investigations and treatments. Models aimed at predicting future disease risk that contains lifestyle factors may be of particular use for targeting health promotion activities at an individual level. This type of cancer risk prediction is not yet available in the UK. We have adopted the approach used by the well-established U.S. derived "YourCancerRisk" model for use in the UK population which allow users to quantify their individual risk of developing individual cancers relative to the population average risk. The UK version of YourCancerRisk" computes 10 year cancer risk estimates for 11 cancers utilising UK figures for prevalence of risk factors and cancer incidence. Since the prevalence of risk factors and the incidence rates for cancer are different between the US and the UK population, this UK model provides more accurate estimates of risks for a UK population.  Using an example of breast cancer and data from UK Biobank cohort we demonstrate that the individual risk factor estimates are similar for the US and UK populations. Assessment of the performance and validation of the multivariate model predictions based on a binary score confirm the model's applicability. The model can be used to estimate absolute and relative cancer risk for use in Primary Care and community settings and is being used in the community to guide lifestyle change.