Preview

PDF, English Download (6MB) | Terms of use

Abstract

Genetic changes play crucial roles in shaping the development and progression of tumors, affecting their growth, response to treatment and overall prognosis. This thesis aims at reconstructing the genetic history of oligodendroglioma, a slowly growing yet incurable brain tumor. The dynamics of its ongoing genetic evolution and the resulting progression remain poorly understood. Therefore, in this work I develop a mathematical approach to estimate malignant growth and infer evolutionary trajectories of oligodendrogliomas from whole-genome sequencing data.

I analyse deep whole-genome sequencing data from six primary-relapse oligodendroglioma pairs. This type of data enables quantitative analysis of subclonal evolution during the early stages of the disease and its subsequent progression. By applying a population genetics model to these samples, I identify a common early tumorigenesis pathway characterized by mutations in the IDH1/2 genes and the loss of chromosomes 1p and 19q, which are defining properties of these tumor entities. Dating the founding cell (most recent common ancestor) of individual patients suggests that oligodendrogliomas originate in childhood and evolve over decades with an increasing growth rate. These dynamics are consistent with rapid oligodendrocyte production in childhood and provides an explanation for the incidence peak of oligodendrogliomas in mid-life. Accelerated growth is linked to the selection of subclones, with or without known driver mutations. Specifically, I find that TERT promoter mutations are associated with strong selection, whereas weak subclonal selection is linked to mutations in CIC, NOTCH1, ZBTB20 or ATM, or could not be explained by a known driver. For individual tumors, I predict the time to recurrence based on the inferred growth rate of the primary tumor. Overall, I show how genome sequencing can provide valuable insights into the dynamics of tumor evolution, allowing for personalized predictions of recurrence based on the genomic analysis of the primary tumor.

In the second part of the thesis, I explore a potential non-genetic factor influencing subclonal selection of oligodendrogliomas: DNA methylation. Specifically, differentially methylated sites between recurrent and primary tumor samples are identified to better understand oligodendrogliomas progression. However, I could not detect evidence for epigenetic subclonal driver events.