Acute myeloid leukaemia (AML) is an aggressive type of cancer which claims lives of 70-80% of patients within five years from diagnosis. Although AML is known to develop through a series of acquired mutations, the mechanisms driving its early stages remain poorly understood. Most AML cases as well as its precursor – myelodysplastic syndrome (MDS), originate from clonal haematopoiesis (CH), a common age-associated process. However, only a small proportion of individuals with CH progress to malignancy. Predicting which individuals will advance to cancer remains a major challenge, likely due to our limited understanding of the mechanisms driving clonal expansion and acquisition of secondary mutations.
Our research aims to investigate how cell-extrinsic factors, such as inflammation, influence pre-leukaemic clonal dynamics. The human proteome represents the complex interactions between genetic and epigenetic regulation of biological systems. By leveraging proteomic and genetic data from UK Biobank, we aim to characterise the profiles of people who developed MDS and/or AML. We will then develop a model that will classify individuals based on their future blood cancer development.
Previous studies show that blood cancer-associated mutations can be detected years or even decades before AML diagnosis. We aim to verify whether these early changes are also detectable at the protein level. Since identifying CH depends on detection sensitivity and can be challenging in smaller clones, we will assess if protein-level alterations precede detectable mutations.
The primary objective of our research is to evaluate whether the proteomic profiles of pre-leukaemic individuals can be used to predict future onset of AML or MDS. We aim to investigate whether integrating proteomic data with other data types, such as lifestyle factors and genetic information including mutational signatures, can improve our predictive accuracy.