Bladder cancer is the second commonest malignancy of the urinary tract accounting for 145,000 directly related deaths world-wide per annum. TCC accounts for the majority of bladder cancer presenting in Europe. Approximately 70-80% of all TCC is superficial at presentation. Superficial TCC presents a major healthcare issue as it is a highly recurrent and progressive disease with recurrence rates of up to 70% and pathological progression rates up to 30% depending on grade and stage of disease. Intravesical chemotherapy and immunotherapy are established adjuvant treatments in superficial TCC with the objective of reducing recurrence with chemotherapy and reducing both recurrence and progression with immunotherapy. Muscle invasive bladder cancer requires radical treatment with radiotherapy or cystectomy. Despite significant advances in surgical techniques and perioperative management, cystectomy remains a morbid procedure with the rate of cure having reached a plateau. Radical radiotherapy has undergone significant improvements in targeting and effective dose delivery over the past 3 decades. Experiments in this thesis investigate whether further improvements in radiation induced cell kill can be achieved by manipulating the epidermal growth factor receptor and its down-stream signalling cascades using the tyrosine kinase inhibitor ZD1839. In this thesis the effects of combination treatment with ZD1839 and radiotherapy in the established bladder cancer cell lines MGHU-1 and its mutant radiosensitive clone, S40b are studied. The optimisation schedule for treatment with ZD1839 was established using colony forming assays. The mechanism of action of ZD1839 was investigated using flow cytometry for analysis of cell cycle perturbations and effects on apoptosis. Further experiments were conducted to ascertain whether there was an additive or synergistic effect between ZD1839 and the established radiosensitiser Gemcitabine at the optimum treatment schedules established for both drugs using colony forming assays.

No description available.

No description available.

No description available.

No description available.

The outcome of radical treatment for prostate cancer is appreciably influenced by the presence of hypoxia. Oxygenation status may therefore be another underlying biological parameter, beyond the classic prognostic factors (age, clinical stage, Gleason score and prostate specific antigen), that predicts for treatment failure in this malignancy. Angiogenesis plays a pivotal role in the growth, invasion, metastasis and survival of prostate tumours. Measurements of angiogenesis have been linked with clinical and pathological stage, histological grade and the potential for metastasis formation. They also provide prognostic information and have been correlated with disease-specific survival and progression after treatment. Magnetic resonance imaging techniques are capable of detecting the molecular, biochemical, physiological and metabolic changes that occur due to pathological processes within tissues. Experiments presented in this thesis have sought to evaluate the ability of Dynamic Contrast Enhanced MRI (DCE-MRI), Dynamic Susceptibility Contrast MRI (DSC-MRI), Intrinsic Susceptibility Weighted MRI (also known as Blood Oxygen Level Dependent (BOLD) MRI) and Diffusion Weighted Imaging (DWI) to characterise the oxygenation and vascular status of prostate tumours in animal models and in patients with prostate cancer. This research has demonstrated the feasibility of hypoxia imaging in prostate cancer. Although MRI can not precisely map tissue p02, the combination of BOLD-MRI and dynamic susceptibility contrast MRI provides a valuable surrogate and predicts the pattern of hypoxia, as determined by pimonidazole immunohistochemistry, with reasonable accuracy. The research has also shown that prostate cancer responds to carbogen gas breathing and that androgen deprivation causes profound vascular collapse within one month of starting therapy. These findings should help in the rational design of future studies that aim to target tumour vasculature and combat tumour hypoxia in prostate cancer.

Terahertz (THz) radiation lies between the microwave and infrared regions of the electromagnetic spectrum. THz radiation excites intermolecular interactions and is non-ionising making it a viable tool for medical imaging. This thesis describes the development and validation of spectroscopic methods for diagnosis of tissue pathologies at THz wavelengths. Theoretical techniques were developed to determine the origin of the contrast seen in THz images of biological tissue. Specific biological tissues investigated in this thesis were colonic tissues with the aim of determining the origin of contrast between healthy and diseased tissue in THz images. This thesis investigates the interaction of THz radiation with matter using simple tissue phantoms made from five biologically relevant materials: water, methanol, lipid, sucrose and gelatin. Phantoms are designed to imitate the spectroscopic properties of tissue at specific wavelengths where physical properties of the phantom, such as concentration and homogeneity, can be accurately controlled. The frequency-dependent absorption coefficients, refractive indices and Debye relaxation times of the pure compounds were measured and used as prior knowledge in the different theoretical methods for the determination of concentration. Three concentration analysis methods were investigated, a) linear spectral decomposition, b) spectrally averaged dielectric coefficient method and c) the Debye relaxation coefficient method. These methods were validated on phantoms by determining the concentrations of the phantom chromophores and comparing to the known composition. Two-component phantoms were made comprising water with methanol, lipid, sucrose or gelatin. Two different three-component phantoms were created; one with water, methanol and sucrose and a second with water, gelatin and lipid. The accuracy and resolution of each method was determined to assess the potential of each method as a tool for medical diagnosis at THz wavelengths. Finally, the spectroscopic methods were applied to measurements of ex-vivo colon tissues containing cancerous and dysplastic regions. Statistical analysis of the reflected time-domain waveforms demonstrated good distinction between healthy and diseased tissues with an estimated sensitivity of 89.2% and specificity of 78.3%.

Electrical impedance Tomography (EIT) is a novel medical imaging method which has the potential to provide the revolutionary advance of a method to image fast neural activity non-invasively. by imaging electrical impedance changes over milliseconds which occur when neuronal ion channels open during activity. These changes have been estimated to be c.1% locally in cerebral cortex, if measured with applied current below 100Hz. The purpose of this work was to determine if such changes could be reproducibly recorded in humans non invasive First, a novel recessed electrode was designed and tested to determine to enable a maximal current of 1mA to be applied to the scalp without causing painful skin sensation. Modelling indicated that this produced a peak current density of 0.3A/m2 in underlying cortex, which was below the threshold for stimulation. Next, the signal-to-noise ratio of impedance changes during evoked visual activity was investigated in healthy volunteers with current injected with scalp electrodes and recording of potential by scalp electrodes (Low Frequency EIT) or magnetic field by magnetoencephalography (Magnetic Detection EIT). Numerical FEM simulations predicted that resistivity changes of 1% in the primary7 visual cortex translate into scalp voltage changes of IjiV (0.004%) and external magnetic field changes of 30fT (0.2%) and were independently validated in saline filled tanks. In vivo, similar changes with a signal-to-noise ratio of 3 after averaging for 10 minutes were recorded for both methods the main noise sources were background brain activity and the current source. These studies with non-invasive scalp recording have, for the first time, demonstrated the existence of such changes when measured non-invasively. These are unfortunately too low to enable reliable imaging within a realistic recording time but support the view that such imaging could be possible in animal or human epileptic studies with electrodes placed on the brain or non-invasively following technological improvements this further work is currently in progress.

No description available.

Respiratory motion is a major factor contributing to errors and uncertainties in Radiotherapy (RT) treatment of lung tumours. Knowledge of this motion may improve the planning and delivery of RT treatment for lung cancer patients. This thesis develops and evaluates methods of building patient specific respiratory motion models. These relate the internal motion to respiratory parameters derived from an external surrogate signal that can be measured during data acquisition and treatment delivery. The models offer a number of advantages over current methods of imaging and analysing respiratory motion, in particular their ability to account for variations in the respiratory motion. Computer Tomography (CT) data is acquired over several respiratory cycles to sample some of the variation in the respiratory motion. B-spline registrations are used to recover the motion and deformation from the CT data. The models are then constructed by fitting functions that relate the registration results to the respiratory parameters. This thesis describes the CT data and respiratory parameters that have been used to construct the motion models. It details the registrations protocols used and evaluates their results. The initial models presented in the thesis relate the registration results to a single parameter, the phase of the respiratory cycle, and average out any variation in the respiratory motion. The later models relate the registration results to two respiratory parameters, with the intention of modelling some of the variation. A number of different functions are assessed for both the single and two parameter models. The results show that the models can predict the respiratory motion in the CT data very accurately (mean error < 1.4 mm). This thesis also discusses some of the uses of the motion models in RT and, in particular, explores the use of the motion models for 'tracking' respiratory motion while delivering intensity modulated RT.