The clonal dynamics underlying the genesis and regression of myeloid disorders

All cells of the human body contain DNA, that is copied to the daughter cells during cell divisions. In this process, errors, or 'mutations' can occur. Some of these mutations are oncogenic driver mutations, i.e. cancer-causing mutations. Older individuals have more mutations, hence a higher risk to develop cancer. However, also children develop cancers, and some types of cancer even occur more often in children compared to young adults, such as leukemia. In this thesis, we studied the clonal evolution of different myeloid leukemias.
First, we aimed to better understand pediatric leukemia by studying Down syndrome patients from a somatic evolutionary view. Down syndrome patients are prone to develop leukemia, and in the case of acute myeloid leukemia (AML), this is often preceded by a transient disease at the neonatal age. This transient leukemia can very rarely also occur in non-DS neonates, and we studied this phenomenon by a large international collaboration to collect patient data and reach consensus. Using these data, we developed a guideline for clinicians to diagnose and treat patients with neonatal myeloproliferative disease.
Classification and thus treatment stratification of pediatric AML is mainly based on underlying genetic aberrations. We studied a subset of pediatric AML with a worse prognosis, namely NUP98-rearranged AML. Within this category, we identified a small subset of patients with specific genetic aberrations that have a relatively better prognosis, namelijk NUP98-KDM5A AML with chromosome 13 aberrations. Also, we showed that NUP98-rearranged AML is a heterogeneous group, mainly caused by the different translocation partners. To further delineate the origin of pediatric AML with different fusions. By both bulk and single-cell whole genome sequencing (WGS) of leukemic blasts, hematopoietic stem and progenitor cells (HSPCs) and differentiated cells within the same patient we coud show that normal HSPCs and differentiated cells share the oncogenic aberrations with the AML blasts, and we could trace these aberrations back to years before AML diagnosis, indicating that other (non-genetic) changes may be needed for the AML to fully emerge.
Lastly, we focussed on therapy-related myeloid neoplasms (t-MN), occuring in children after chemotherapy treatment for their first cancer. We could show that both AML blasts and normal HSPCs have an increased mutation load after treatment, and that those mutations are both directly and indirectly caused by specific chemotherapies. Furthermore, we showed different working mechanisms of chemotherapies, with thiopurine treatment only affecting dividing cells, and cisplatin treatment affecting all cells. However, the latter effect is dependent on TP53, a tumor suppressor gene. Cells that are TP53-deficient can divide during cisplatin exposure, and thus these cells have a selective advantage during therapy, enabling them to expand into t-MN. This is specifically important for Li-Fraumeni syndrome patients, that already have a constitutional TP53 mutation, and thus only one wild-type allele for TP53.