Cancer-associated somatic mutations frequently drive clonal expansions in normal ageing tissues. However, the factors governing whether pre-cancerous cells transform into cancer are poorly understood, hindering identification of clones that are clinically significant rather than benign sequelae of ageing. The main aim of this dissertation has been to explore this process in the haematopoietic system, where leukaemia-associated mutations are detectable in >10% of individuals over the age of 50. This phenomenon, termed clonal haematopoiesis (CH), is associated with an increased risk of blood cancers, though only a small minority of individuals progress. Acute myeloid leukaemia (AML) is the commonest acute leukaemia in adults, and usually presents abruptly with complications of bone marrow failure. Using deep targeted sequencing of stored blood DNA samples from individuals who went on to develop AML and controls, we identified features of CH that predict leukaemic progression. The number, type and burden of genetic changes, as well as certain clinical variables, were predictive of AML-free survival. Examining the pre-clinical evolution of lymphoid malignancies using a similar study design and broader sequencing approach also revealed genetic and clinical features predictive of malignant transformation.
The final part of this study investigates the prevalence of clonal haematopoiesis in childhood cancer survivors treated with intensive chemo- or radiotherapy. In contrast to adult cancer patients, the prevalence of CH in children is not dramatically increased by cytotoxic treatment. Collectively, these findings provide proof of principle that benign and pre-malignant clonal expansions in normal blood (and perhaps other tissues) may be distinguishable years prior to overt malignant transformation. This could in future enable earlier detection of those at high risk of blood cancers, and stimulate research into possible interventions to reduce the risk of progression.