Search

Dysplasia arising in Barrett’s oesophagus (BO) confers risk of progression to oesophageal adenocarcinoma (OAC), but it is variable and confounded by sampling error and diagnostic inter-observer variability. There is a need for biomarkers to accurately define risk and guide surveillance and treatment strategies. Oesophagectomy is the preferred treatment of high grade dysplasia (HGD), although this has been challenged by the emergence of endoscopic therapy. This thesis has shown image cytometric DNA ploidy analysis (ICDA) is accurate when compared to flow cytometry. In a multi centre biomarker validation study, ICDA predicted cancer risk in non dysplastic BO. ICDA also predicted relapse following photodynamic therapy (PDT). Nucleotyping (NT) is an imaging technique that evaluates textural changes of nuclei, which correlate with chromatin content and organisation. In this thesis NT was demonstrated to yield additional diagnostic information over ICDA alone, and the combination was highly accurate for dysplasia classification. In an experiment evaluating replication licensing factors in OAC, those representing the S-G2-M phases of the cell cycle correlated with aneuploidy. Polo like kinase 1 upregulation had the strongest correlation, the first time this has been shown in BO. Elastic scattering spectroscopy detected DNA ploidy abnormalities with equal accuracy using a standard or near infra-red enhanced spectrometer. In a prospective in vivo surveillance study, fewer biopsies were taken without a change in the diagnostic yield for dysplasia. A field carcinogenesis effect was also demonstrated, independent of both dysplasia and DNA ploidy abnormalities. Finally, in a randomised control trial of PDT for HGD, patients with BO less than 7cm had significantly higher efficacy and better tolerability with ALA than photofrin. Radiofrequency ablation was shown to be a safe and effective rescue therapy for patients that failed PDT. In summary these studies advance our understanding of biomedical optics for the risk stratification and treatment of Barrett’s oesophagus.

This work utilised two common methods of pruning individual urban trees: crown-reduction (partial shoot removal), which is commonly used to restrict the water use of trees through to be at risk within the context of subsidence, and crown-thinning (removal of whole side shoots or branches), which is more commonly used to reduce the shading effects of trees near buildings. These treatments were applied to individual, mature Prunus avium trees in an orchard in the South East of England. Despite rapid recovery of canopy LA within the crown-reduced trees, both treatments significantly reduced the water use of individual wild cherry trees during the first season following application, with crown-reduction having the greatest effect. Crown-thinning increased daily water use per unit LA, while crown-reduction decreased water use per unit when compared to that of the control trees. These differences in water use per unit LA and whole tree water use related well to the contrasting influence of the treatments on canopy architecture and rates of canopy regrowth. Boundary layer conductance (gb) and leaf irradiance were constantly lower within the crown-reduced trees and higher in the crown-thinned trees, when compared to control values. Differences in leaf level physiology were also observed as a result of canopy manipulation, with crown-reduced trees having the highest rates of assimilation (A) and stomatal conductance (gs) in the first year after manipulation. This difference was possibly the result of a smaller shoot:root ratio (S:R). These physiological differences resulted in the crown-reduced trees having a higher discrimination against the heavier 13C (Δ13C), traditionally implying low water use efficiency (WUE). However, measurements of the δ18O of leaf water and whole tree sap flux indicated that this treatment had a lower water use than either the crown-thinned or control trees.