Open Collections

Warner, Evan

University of British Columbia

Prostate cancer is the most common cancer in Canadian men, and despite decades of intensive clinical research, metastatic prostate cancer (mPCa) remains invariably lethal. Some patients relapse with metastases after surgery or radiation therapy, while a subset present with metastatic disease at diagnosis. Optimal management of mPCa demands further investigation to solidify prognostic and predictive biomarkers to guide treatment selection. Both tumour tissue and peripheral blood are a viable source of DNA for identification and study of genomic biomarkers, but neither are optimal for profiling all disease states. The research presented in this thesis describes next-generation sequencing of 2,275 tissue and cell-free DNA (cfDNA) samples from 922 men with mPCa. In advanced, heavily-treated mPCa, ctDNA analysis distinguished monoallelic from biallelic defects in DNA damage repair (DDR) genes BRCA2, ATM, and CDK12. BRCA2-deficient samples exhibited frequent disruption of tumour suppressors via chromosomal rearrangements, while CDK12 mutations were associated with oncogene amplification. ATM-mutation appeared mutually exclusive with TP53 mutation. Loss of BRCA2 and CDK12—but not ATM—was associated with inferior clinical outcomes compared to DDR-intact disease. In patients with untreated de novo mPCa, comprehensive sampling of all primary tumour foci, pelvic lymph node metastases, and cfDNA revealed an aggressive landscape of genomic alterations. Intra-prostate heterogeneity was pervasive and impacted clinically-relevant genes, resulting in frequent discordance between select primary-restricted foci and synchronous metastases. CfDNA profiling was frequently confounded by pervasive low tumour fraction across different foci. Additional complexity was driven in several instances by polyclonal metastatic seeding from the reservoir of phylogenetically-related primary populations. When simulating standard clinical practice relying on a single tissue focus, genomic heterogeneity plus highly variable per-sample tumour fraction caused false genotyping of dominant disease. However, in silico modeling demonstrated that analysis of multiple biopsy cores can rescue misassigned somatic genotypes. Overall, this body of work demonstrates the effective use of genomic profiling in both tissue and blood-derived tumour DNA. Heterogeneity was a defining characteristic of mPCa genomes, from mechanisms of gene biallelic inactivation to alteration profiles of intratumoural foci. This knowledge is essential to future biomarker-guided trials seeking to improve treatment selection and personalized management of mPCa.

Text

2022-10-12

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/82853

Genome Science and Technology

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/