Genetic mechanisms and heterogeneity in familial myeloid leukemia

A53

Initial and downstream steps in leukemogenesis may be elucidated through studying families inheriting a predisposition to a purely haematological malignancy. These families allow the unique opportunity to study the “life” of a cancer stem cell throughout disease formation and shed light on the mechanisms involved in disease progression. To date, only two genes have been shown to be involved in familial AML, transcription factors RUNX1 and CEBPA. Only limited knowledge of the downstream events following these germline mutations has been attained. The study presented here is a comprehensive screen of not only germline RUNX1 and CEBPA mutations in familial AML, but also somatic events occurring downstream of these mutations throughout disease progression.
 We have identified novel mutations in RUNX1 and a previously identified mutation in CEBPA in five pedigrees (four previously unpublished). We have also identified five pedigrees in which neither RUNX1 nor CEBPA are responsible. This finding confirms that genetic heterogeneity exists and additional leukaemia driver genes, that likely play important roles in sporadic leukaemia, await discovery.
 In six pedigrees with germline RUNX1 or CEBPA mutation we further investigated the downstream events occurring during progression of disease, utilising affected tissue from different stages of leukemogenesis. We identified several novel secondary hits, including the first described cases of secondary hits to RUNX1 following germline mutation, and an individual with triallelic RUNX1 aberration. We also observed a marked reduction in expression from two separate RUNX1 alleles with distinct mutations; one nonsense and one missense.
 Finally, using a combination of karyotype and SNP chip analyses, we observed both known and novel genomic copy number variations (CNVs) at several loci in 3 individuals. These loci involved genes such as RUNX1, ATM, FLT3 and the homeobox gene CDX2. However, in the remaining 4 individuals studied, we did not identify any potentially pathogenic CNVs. This implicates the involvement of other unidentified hits, such as epigenetic and/or point mutations, or loss of heterozygosity (LOH) without copy number variation. LOH will soon be addressed utilising SNP chip homozygosity call differences between affected and unaffected DNA samples.
 Familial studies of leukemogenesis, such as presented here, will shed light on the initial steps in malignant transformation, and allow further characterisation of the pathways and mechanisms involved in leukemogenesis. This will aid in the development of novel therapeutic and treatment initiatives.