Please use this identifier to cite or link to this item: http://hdl.handle.net/2440/121304
Full metadata record
DC FieldValueLanguage
dc.contributor.advisorPenfold, Scott-
dc.contributor.advisorDouglass, Michael-
dc.contributor.advisorNguyen, Giang-
dc.contributor.authorAustin, Annabelle Mary-
dc.date.issued2019-
dc.identifier.urihttp://hdl.handle.net/2440/121304-
dc.description.abstractCancer is a highly prevalent disease that places a significant economic burden upon society. Radiotherapy is commonly utilised as a treatment for benign and malignant tumours. A fundamental challenge in radiotherapy is delivering a sufficient dose of radiation to eradicate a tumour while minimizing the dose deposited in surrounding healthy tissue. Excessive radiation damage to these tissues can result in treatment toxicities that may have adverse effects on patient quality of life. Proton therapy offers the potential for increased sparing of normal tissue compared with X-ray therapy, which is more commonly used in radiotherapy. However, the degree of this sparing can be highly variable between patients. Furthermore, data from Phase III clinical trials can quickly become outdated due to the long follow-up times that are required to observe late effects, together with the rapid evolution of technology. The process of deciding whether to refer a patient for proton therapy can be complex as a result. In addition, proton therapy is significantly more expensive as a treatment compared with Xray therapy. This suggests that patients who are expected to receive the greatest benefit should be prioritised. Computer models can offer a possible solution to this dilemma, by predicting the clinical outcome that may be expected as a result of a given treatment. In this work, a Markov simulation tool was developed which is capable of producing such predictions and comparing proton and X-ray radiotherapy treatment plans on an individual patient basis. The radiobiological effect of a given treatment plan is estimated in terms of the probabilities of tumour control, radiation-induced injuries and radiation-induced second cancers. These are combined in the Markov model to efficiently estimate the clinical outcome resulting from a given treatment plan. This outcome is quantified in terms of the quality adjusted life expectancy (QALE), or number of quality adjusted life years (QALYs), which is an adjustment of the raw life expectancy to account for the effect of time spent with injury or disease. The result is a model that uses several input parameters to produce a single quantitative output, indicative of the relative quality of a treatment plan. The predictions of the model can be affected by uncertainties in the radiobiological model parameters and uncertainties in dose delivery. The latter can arise as a result of changes in the target volume relative to the radiation field over the course of treatment. A consideration of these effects was incorporated into the model, as they have the potential to influence whether a patient is selected to receive proton therapy. The cost-effectiveness of a treatment is of particular importance in the current resource limited healthcare environment. The Markov model was developed to include treatment costs, including treatment of radiation therapy side effects. An application of the model to a cohort of base of skull chordoma patients revealed that all patients could be treated with proton therapy cost-effectively due to the potential for sparing of critical structures. Base of skull chordoma is typically regarded as a standard indication for proton therapy. In contrast, in a study of a cohort of left-sided breast cancer patients, it was found that the majority of patients could not be treated cost-effectively with proton therapy. This was likely due to the cardiac toxicity rate being particularly low with the deep inspiration breath hold X-ray treatment technique used for the patients in this cohort, resulting in no significant advantage from proton therapy. The developed model has the potential to form the basis of a clinically viable patient selection tool. However, the model requires external validation before being suitable for clinical implementation. Due to the limited availability of proton therapy, such a model may prove useful as Australia prepares to begin treating cancer patients with proton therapy.en
dc.language.isoenen
dc.subjectMarkov modelen
dc.subjectdecision aiden
dc.subjectproton therapyen
dc.subjectradiobiological modelsen
dc.title[EMBARGOED] A radiobiological Markov model for aiding decision making in proton therapy referralen
dc.typeThesisen
dc.contributor.schoolSchool of Physical Sciences : Physicsen
dc.provenanceThis thesis is currently under Embargo and not available.en
dc.description.dissertationThesis (Ph.D.) -- University of Adelaide, School of Physical Sciences, 2019en
Appears in Collections:Research Theses

Files in This Item:
File Description SizeFormat 
Austin2019_PhD.pdfLibrary staff access only11.73 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.