Introduction: The incidence of gastro-oesophageal cancers has increased considerably over the last decade. As the disease is associated with a poor prognosis, there is a need to identify markers of treatment response which could be used in the future to improve the management of gastro-oesophageal tumours. Aims: 1) To compare the ability of published histological grading criteria (Becker, Mandard and Ninomiya) to assess response to neo-adjuvant chemotherapy (NCT) in gastro-oesophageal cancers. 2) To evaluate the expression of thymidylate synthase (TS) in pre-treatment diagnostic biopsy samples as a predictive marker of response to NCT. 3) To investigate whether measurements of hypoxia obtained using pimonidazole are prognostic for treatment outcome in patients with gastro oesophageal adenocarcinoma (ACC). 4) To study the prognostic significance and clinicopathological associations of epithelial mesenchymal transition (EMT) related proteins S100A4, Vimentin and Snail1 in gastro-oesophageal junction (GOJ) tumours. 5) To evaluate the ability of dynamic contrast enhanced (DCE) MRI to assess chemoradiotherapy (CRT) induced changes in oesophageal cancer.Methods: 1) Formalin fixed paraffin embedded (FFPE) tumour blocks and haematoxylin and eosin stained slides of samples from resected tumours (n=66) were obtained from patients who received NCT for gastric and GOJ tumours. The slides were scored independently by two observers and kappa scores calculated. 2) Pre-treatment diagnostic tissue biopsy samples were obtained from 45 patients with gastric and GOJ cancer who received NCT. TS expression was assessed using immunohistochemistry (IHC) and scored by two observers. The clinical and pathological data were analysed. 3) 57 patients were prospectively administered intravenous pimonidazole and the tumour specimens were collected both at staging laparoscopy and resection. IHC was performed to assess pimonidazole expression and determine its association with clinico-pathological factors. 4) Tissue microarrays were prepared from resection specimens from GOJ ACC. IHC was performed to investigate EMT related protein expression. 5) DCE-MRI was performed on five patients diagnosed with oesophageal cancer treated with CRT. Multiple pharmacokinetic parameters were evaluated. Findings: 1) Becker's histological grading criteria was the most reproducible and prognostic of outcome. The incidence of complete histological response (5%) was low in patients receiving NCT. 2) No prognostic benefit of TS expression was identified. 3) Results from only 34 patients were available for analysis. 77% pimonidazole positivity was observed. Preoperative anaemia was associated with significant tumour pimonidazole expression (p=0.04). Pimonidazole was not prognostic for outcome. 4) The overall positive expression was S100A4 (85%), Vimentin (14%) and Snail1 (89%). The increased expression of S100A4 at the tumour body (p=0.02) and luminal surface (p=0.01) was associated with a poor outcome. 5) Significant changes were measured in DCE-MRI pharmacokinetic parameters after CRT. Conclusion: 1) Becker's histological response grading criteria should be further studied in routine clinical practice for response assessment to NCT. 2) TS should be explored further as a marker of NCT response in gastro-oesophageal cancer. 3) Hypoxia is a characteristic feature of upper gastrointestinal ACC and is associated with anaemia. 4) S100A4 is the most useful marker of EMT in GOJ adenocarcinoma. 5) DCE-MRI tracer kinetic parameters should be explored in a larger study to assess their ability to monitor the efficacy of and predict response to neo-adjuvant treatment.