PROVE - PRostate and OVarian Epigenetics: investigating therapeutic resistance via cfDNA

Chemoresistance remains a significant challenge to alleviating suffering from cancer and improving overall survival, and it is considered a major cause of mortality across several cancer types, including prostate (PCa) and ovarian (OC). Despite improvement in PCa survival rates, this malignancy remains the 3rd leading cause of cancer-related deaths in men in Ireland, whereas OC is the most lethal gynaecological cancer, with 5-year survival rates remaining static at approximately 40%. The advent of liquid biopsies, coupled with a longitudinal study design, allows for evaluation of tumour evolution, while also overcoming spatial and/or temporal heterogeneity associated with the malignancy itself. We hypothesised that molecular changes driving therapeutic resistance in these tumours could be studied using liquid biopsies. The aim of this PhD thesis was to explore how tumours in PCa and OC patients evolve during acquired therapeutic resistance, with a specific focus on the methylome, as this has already been shown to be paramount to therapeutic resistance in many cancer types. First, we evaluated the feasibility of tumour-specific methylation detection in the PCa setting, using two types of liquid biopsies (blood and urine) obtained from four treatment naive metastatic PCa (mPCa) patients. Analysis of differentially methylated CpG sites unveiled similar methylation patterns between both liquid biopsies, with significant contributions from non-tumour sources, consistent with previous reports. Serial methylomic profiling was then carried out in a total of 48 plasma samples from an additional mPCa cohort (n=9), at baseline and multiple therapeutic intervals. While the majority of the methylome remained stable over time, a proportion of dynamic CpG sites was found to be associated with clinical events in these patients, including different therapeutic approaches. Moreover, these sites were predominantly linked with genes involved in immune response pathways, an observation that supports our foregoing findings surrounding the origins of cell-free DNA. We investigated therapeutic resistance patterns in OC by using a different approach, focused on firstly decoding tumour patterns in tissue samples. While OC is composed of several different subtypes, the analyses in this thesis pertain to serous OC only. A personalised omics pilot study, using low-grade serous ovarian cancer (LGSOC) samples for a single patient, resulted in the identification of a BRAF D549G mutation, uncovering new potential therapeutic routes for this patient. While the therapeutic targeting of this alteration resulted in minimal clinical response, we were able to sensitively detect this mutation in a matched plasma sample using droplet digital PCR. Refocusing on the methylomic landscape, and highlighting the high-grade serous subtype (HGSOC), we were able to identify a total of 8 genes with concurrent methylomic and transcriptomic changes in acquired resistant samples. Methylation detection of two of those genes, APOBEC3A and NKAPL, in plasma samples of a longitudinal HGSOC cohort, revealed a degree of association with patient relapse status. Finally, targeted demethylation of the promoter region of NKAPL in an acquired resistance OC cell line model resulted in partial reversal of cisplatin resistance, emphasising the potential role of a methylation-induced drug resistance mechanism in HGSOC patients. The work conducted in this thesis showcases the potential of a personalised medicine approach, in conjunction with the use of liquid biopsies, to deliver insight into novel mechanisms of drug resistance. Methylation signatures were upheld as part of the evolving landscape of both PCa and OC, with a likely involvement in therapeutic response mediation in patients presenting with these malignancies. Future study of these signatures is warranted to determine their performance as minimally-invasive tools for accurate monitoring of PCa and OC patients