Emerging strategies for paediatric cancer predisposition: detection, surveillance and prevention.

A substantial proportion of childhood cancers occur in children who are genetically predisposed to develop that cancer. Close surveillance of children with cancer predisposition syndromes enables early cancer detection and results in better survival outcomes and less therapy-related toxicity, although many more predisposition syndromes are likely to exist and remain undetected. Building on conventional clinical surveillance protocols, new approaches are being explored, including liquid biopsy, molecular or metabolic biomarkers, enhanced imaging and microbiome profiling for the purposes of both improved surveillance and the discovery of additional predisposing factors. Knowing the specific genes, expressed proteins and biological pathways associated with a specific cancer predisposition syndrome might permit the use of targeted agents or vaccines, as well as the introduction of lifestyle changes to delay, reduce or even eliminate the development of cancer. Preclinical studies investigating the risks and benefits of preventive interventions would be greatly aided by the identification or development of animal models that faithfully recapitulate human syndromes. The incorporation of artificial intelligence to complement the clinical and experimental assessment of cancer risk based on the syndrome, genotype, modifier genes, family history, environmental exposures and other factors will undoubtedly improve the management and, ultimately, the outcomes of children with an increased genetic risk of cancer. In this Perspective, we discuss strategies to identify new cancer predisposition genes, developing approaches to cancer surveillance, emerging efforts for cancer prevention in high-risk patients, and the role of artificial intelligence in improving data analysis and optimizing medical decision-making.