Medicine (St Vincent's) - Theses
Permanent URI for this collection
Now showing 1 - 2 of 2
ItemBreast cancer risk assessment and risk management:: Development of a personalised, web- based, decision support tool (iPrevent®)Collins, Ian ( 2016)Breast cancer risk is a complex interaction of environmental factors, inheritance and genetics. Awareness of her personal breast cancer risk can allow a woman to make informed decisions regarding the management of her risk, through screening or risk reduction measures such as surgery or medication. The effect of risk factors, both inherited and modifiable, on breast cancer risk is complex but mathematical models exist to estimate an individual’s risk. These models use epidemiological data to quantify the risk for an individual, based on a range of their risk factors. As these models were developed primarily for research purposes, they are not designed for ease-of-use by a range of clinicians, nor designed to be used by those unfamiliar with estimating breast cancer risk. Once estimated, a woman’s breast cancer risk must then be explained in a way that is comprehensible to the woman, together with ways to manage that risk in a similar format. If this were achieved for all women, it may allow them to make informed choices and potentially even prevent breast cancer or reduce its impact greatly. The ultimate aim of this research was to develop a user-friendly computerised, web-based breast cancer risk assessment and risk management support tool. This tool, called iPrevent©, uses the existing mathematical models to estimate individualised breast cancer risk, but using a user friendly interface. It then goes further, and provides Cancer Australia guideline-based recommendations, based on that risk, for each individual woman. It presents the risks and befits of each evidence-based intervention in a similar manner to the risk so that women can make an informed choice regarding their breast cancer prevention strategy. Before developing iPrevent©, I first examined the other possible effects of being a carrier of a mutation in breast cancer predisposition genes, as it was hypothesised that other factors such as fertility effects, could have a large bearing on any future decisions women may make, including risk reducing surgery. I then explored current behaviours to reduce risk among women at highest risk, in an attempt to understand the magnitude of the possible benefits of iPrevent© in this highest risk group. Through focus group studies, I examined the information needs of clinicians to facilitate breast cancer risk discussions. Understanding the needs of end-user clinicians of iPrevent© ensures it could meet their needs. This usability may increase uptake and use, of both the tool and breast cancer risk management strategies where appropriate. This tool, iPrevent©, is currently undergoing clinical validation studies, outside the scope of this thesis, but will shortly become freely available with the aim of increasing individual awareness of each woman’s own breast cancer risk, enabling her to manage that risk according to the evidence, Cancer Australia guidelines and her own preferences.
ItemThe contribution of genetic variations in the region of the parathyroid hormone-like hormone, PTHLH, gene to breast cancer susceptibilityFreeman, Adam Noel ( 2016)Aims: • Analyse the evidence for PTHLH and PTHrP, its protein product, playing a role in breast cancer and update the empirical definition of the gene. • Describe the 3-Dimensional structure of the PTHLH region and determine its system of regulatory interactions, including remote regulatory elements affecting PTHLH. • Integrate existing tools in addition to the novel perspectives generated above to enable a comprehensive annotation and analysis of genetic variants identified through molecular epidemiological techniques including Genome-Wide Association Studies (GWAS) and somatic DNA sequencing of tumour tissue to derive putative molecular mechanisms for the region’s involvement in breast cancer susceptibility. Methodology: • Review published studies and databases, integrating findings from diverse sources. • Acquire and analyse DNA and RNA sequencing, regulatory, expression, algorithm-inferred, and proximity-ligation data from multiple public data sources including ENCODE, ROADMAP, dbGAP, COSMIC and other to update the definition of PTHLH, and advance concepts of structure and regulatory function in the region. • Acquire and analyse primary GWAS data, performing imputation with multiple algorithms and references, and annotating associated variants with a suite of tools. • Use genome browsers including UCSC, WUSTL, and Golden Helix SVS, and their associated databases and tools, to analyse and visually integrate findings. Results: • PTHrP has multiple discretely functional segments active throughout the cell. It likely plays a bivalent and context-dependent role in cancer biology. Analysis of somatic variation in cancer suggests PTHrP may have a tumourigenic role within the nucleus. • PTHLH sits within a 1.3Mb TAD featuring multiple sub-structures that are integrated with the region’s regulatory function. There are activated chromatin hubs (ACHs) at protein-coding genes with evidence of extensive interaction between them. This is facilitated by the TAD’s structure, collocating them at the neck of the TAD. • The ACHs at MRPS35, KLHL42, and CCDC91 each monopolise a subordinate regulatory sub-net with a hierarchical structure. They each appear to act as important remote regulatory elements that integrate regulatory signals generated within their respective sub-nets, transferring them to PTHLH, and other genes, via ACH-ACH interactions. • There appear to be multiple discrete GWAS breast cancer association signals in the PTHLH region. Annotation of the associated variants suggests three particular regulatory elements may be its key drivers. In the context of the regulatory concepts developed in this thesis, the variants may affect a particular regulatory signal at multiple points in its assembly. PTHLH is the likely downstream target of this signal. • There are multiple poorly-describe coding, and non-coding, genes in the region that are also potential actors in breast cancer and should be investigated. Conclusion: • The PTHLH region is likely involved in the pathogenesis of breast cancer through the modification of PTHLH expression. There are likely to be other mechanisms in parallel that are yet to be fully described.