Edinburgh Research Archive

Genomic catastrophes: complex structural variants in high grade serous ovarian cancer

Item Status

Embargo End Date

Authors

Brown, Stuart L.

Abstract

High-grade serous ovarian cancer (HGSOC) is the most common and lethal type of ovarian cancer. HGSOC accounts for 70-80% of deaths from all forms of ovarian cancer, roughly 98000-110000 deaths worldwide each year according to WHO estimates. It is characterized by ubiquitous mutations in the tumour suppressor gene TP53. Roughly half of HGSOC tumours also have mutations in genes involved in the homologous recombination repair pathway, primarily BRCA1 and BRCA2, leading to homologous recombination deficiency (HRD). Cancers with HRD must rely on alternative DNA repair pathways to repair breaks in their DNA. The PARP protein is key to the base excision repair/single-strand break repair pathway, and recent HGSOC drugs inhibit the PARP protein to make HRD tumours unable to efficiently repair genomic breaks and leading to apoptosis. However, these cancers often gain resistance to PARP inhibitors by regaining the ability to use the homologous recombination DNA repair pathways, which renders PARP inhibitors ineffective. On average only 45% of patients diagnosed with HGSOC survive for more than 5 years after diagnosis. HGSOC is highly genomically rearranged and contains hundreds to thousands of structural variants, though their impacts on tumour function and evolution are poorly understood. Structural variants usually accumulate over many cell cycles, but it has recently become clear that large numbers can be acquired in a single cell cycle in complex patterns called complex structural variants (cSV). It is known that cSVs are catastrophic genomic events that can generate hundreds of structural changes across the genome, and at least some cSV types have been reported in HGSOC. The changes induced by cSVs can increase or decrease the copy number of genes, change their genomic location and orientation, or disrupt the gene entirely. The increased use of whole genome sequencing in large cancer cohorts has allowed an increasing number of cSV types to be identified. However, investigations into cSVs often focus on a single cSV type, frequently using differing computational criteria for the identification of cSVs. This means that the relationships between cSVs remains understudied, and interpretability of the impacts of cSVs between studies is challenging. In this work, eight cSV types were identified across the 324 whole-genome-sequenced HGSOC samples using previously published criteria: chromothripsis, chromoplexy, breakage fusion bridges, ecDNA, pyrgo, rigma, tymphonas and seismic amplification. The prevalence, distribution, and impact on survival of each cSV type was assessed, and their relationships with other genomic features, such as HRD and whole genome duplication (WGD) were investigated. By studying eight cSV types together across a uniformly processed well annotated cohort, many novel insights into HGSOC structural evolution were gained. I have shown that there are two main routes to genomic diversity in the HGSOC cohort. One of these routes involves HRD while the other route involves WGD and cSVs, and it appears that WGD can buffer the deleterious effects of cSV. I also found that the presence of cSVs was not associated with significant variation in survival rates, but there was a trend that patients with tumours containing more severe cSV events had better survival rates compared to those with less severe events of the same type. This was observed for chromothripsis, a particularly disruptive cSV often described as ‘chromosome shattering’. My results also revealed a novel hotspot for multiple cSV types on chromosome 19 which covers the known HGSOC oncogene CCNE1, which confers worse survival on patients when amplified. I have shown that the amplification of CCNE1 by breakage fusion bridges results in higher copy numbers than by simple amplification. Overall, this thesis investigates the prevalence, co-occurrence, distribution, and impact of multiple cSV types on survival in HGSOC and provides novel insights into the genomic diversity of HGSOC.

This item appears in the following Collection(s)