Introduction The prognosis in oesophago-gastric cancer is poor with less than 15% patients surviving beyond 5 years after diagnosis. The addition of neoadjuvant therapy has been shown to increase survival in patients suitable for curative surgery. However, the additional gains are modest and the majority of patients do not respond sufficiently from therapy to gain any benefit. There is an urgent need to identify markers that can predict response to neoadjuvant therapy in order provide safer, more effective, individualised treatment regimes. Methods A prospective, multi-centre, collaborative study was undertaken in patients with oesophago-gastric cancer undergoing neoadjuvant therapy and potentially curative surgery. Levels of circulating biomarkers M2-Pyruvate kinase, alkaline phosphatase, CA19-9, CEA and CA 72-4 were measured in patients before and after administering the first cycle of chemotherapy. Binary logistic regression analysis was performed to assess the ability of biomarkers to predict histological response to therapy. Results 165 patients were recruited to the main study. 105 patients had complete histopathological data for analysis. There were 27 responders and 78 non-responders to neoadjuvant therapy. There were no differences in pre-therapy demographic, pathological or treatment factors between the two groups. Responders had less post-operative lymphovascular invasion (P= 0.004) and higher R0 resection rates (P=0.03). Pre-therapy M2-Pyruvate kinase levels were lower in responders compared to non-responders (P=0.037) and levels were able to predict response with each unit increase in the biomarker level being associated with a 4.1% decrease in the likelihood of response (P=0.027). M2-PK levels were not associated with any pre-operative demographic, clinical or pathological factors. Conclusions Pre-therapy dimeric M2-PK levels can predict response to neoadjuvant therapy in patients with oesophago-gastric cancer. The test could be of clinical value for 1 in every 8 patients undergoing the test.

The high incidence of gastrointestinal cancers in the general population and the presence of premalignant dysplastic precursor lesions in the gastrointestinal tract make the gastrointestinal tract an ideal environment to study cancer clonality and clonal expansion. Background: Intestinal metaplastic (IM) glands in the human stomach are clonal, contain multiple stem cells and spread by fission. This mechanism of gland fission causes field cancerisation. We hypothesised that gastric adenocarcinoma (GA) progresses through a series of genetic events arising from a founder mutation. A process analogous to niche succession may also take place in the normal oesophagus. We hypothesise that oesophageal squamous cell cancer occurs by a process of field cancerisation of the oesophagus. RHBDF2 has been identified as the gene responsible for tylosis with oesophageal carcinoma (TOC). We hypothesise that RHBDF2 germline gain of function mutations might be lost during tumour progression in TOC and this might affect iRhom2 localisation in the cell. Methods and results: A cohort of 23 patients with dysplasia and a cohort of 51 GA patients were screened for genes accounting for 75% of all somatic mutations previously reported in GA. Only 13% of dysplastic patients and 31.4% of GA patients had mutations. Three dysplastic patients and six GA patients were analysed by microdissection. Small gastric cancer foci in a cohort of hereditary diffuse gastric cancer (HDGC) patients (n=5) were also screened by laser-capture microdissection (LCM) for mutations in TP53. A cohort of 30 patients was screened for common mutations in OSCC and for RHBDF2 mutations. 36.36% of the patients presented mutations. Three patients with mutations were randomly selected and areas of oesophageal squamous cell dysplasia and OSCC were analysed by LCM. Three TOC patients were also analysed by LCM and immunohistochemistry was performed for iRhom2 and ADAM17. Conclusions: The usual mutational events established for GA development during the metaplasiadysplasia- carcinoma sequence (MCS) do not fit the results from either of our two LCM mutation studies in the human stomach. Dysplasia was shown to be clonal and GA demonstrates genetic heterogeneity through clonal evolution. Field cancerisation could not be detected in HDGC using TP53 as a clonal marker. The low incidence of OSCC patients with mutations implies that other genes may be involved in the premalignant pathway leading to OSCC. Oesophageal squamous cell dysplasia and OSCC demonstrate clonal expansion through tumour progression. RHBDF2 mutations do not occur in sporadic OSCC but germline RHBDF2 mutations can be lost during tumour progression in TOC patients with LOH in 17q. Overall, the somatic mutation theory of carcinogenesis seems to hold true for both the progression to GA and OSCC, as both carcinomas seem to evolve from a single mutated stem cell and acquire genetic heterogeneity as the tumours evolve.

Emerging evidence suggests the immune system has a role in preventing cancer, and in advanced cancer evidence of immune dysfunction is widespread. This project focused on cytotoxic T lymphocyte antigen 4 (CTLA4), a key negative regulator of T cell activation found on dedicated regulatory T cells (Treg) and activated T lymphocytes, and asked whether modulation of immune control with anti-CTLA4 blockade led to significant anti-tumour activity. Clinical and laboratory investigation of anti-CTLA4 blockade using tremelimumab in a phase II trial of second-line therapy in advanced oesophageal and gastric adenocarcinomas was combined with an attempt to establish a suitable pre-clinical model based on therapeutic vaccination against the tumour associated antigen (TAA) 5T4.Eighteen patients received tremelimumab. Most drug-related toxicity was mild but there was a single death due to bowel perforation. Four patients had stable disease with clinical benefit; one achieved a partial response after eight cycles (25.4 months) and remains well on study after four years. Markers of regulatory phenotype, forkhead box protein 3 (FoxP3) and CTLA4, doubled transiently in CD4+CD25high lymphocytes in the first month after tremelimumab before returning to baseline. In contrast, CTLA4 increased in CD4+CD25low/negative lymphocytes throughout the cycle of treatment. Post-treatment expanded Treg expressed FoxP3 without interleukin-2 and their defining suppressive function was not abolished despite prolonged anti-CTLA4 blockade. De novo proliferative responses to TAA 5T4 (8 of 18 patients) and carcinoembryonic antigen (CEA; 5 of 15) were detected. Patients with a post-treatment CEA proliferative response had median survival of 17.1 months compared to 4.6 months for non-responders (p=0.002). Baseline interleukin-2 release after T lymphocyte activation was higher in patients with clinical benefit and toxicity. Heterologous mouse 5T4 (m5T4) vaccination showed some evidence of weak therapeutic benefit, but all tumour models investigated had rapidly lethal kinetics. Specific m5T4 immune responses could be detected by serum antibody ELISA and IFN-gamma ELISPOT assays in naive animals but were lower frequency than published responses to h5T4, and were further attenuated in tumour-bearing animals. The addition of anti-CTLA4 blockade did not result in significant augmentation of m5T4 specific immunity after vaccination in non tumour-bearing animals and combination treatment was ineffective as therapy in this autologous model. Results are discussed in the context of emerging immunotherapeutics in melanoma and prostate cancer. In the absence of supportive data from the model system it would not be appropriate to pursue combination heterologous 5T4 vaccine with anti-CTLA4 blockade, but in view of the unusual durability of the best response to tremelimumab, and in vitro evidence of enhanced proliferative responses to relevant TAA, further investigation of drug activity may be warranted in metastatic gastric and oesophageal second-line treatment.

Gastric and esophagogastric junction cancers are heterogeneous and aggressive tumors with an unpredictable response to cytotoxic treatment. New methods allowing for the analysis of drug resistance are needed. Here, we describe a novel technique by which human tumor specimens can be cultured ex vivo, preserving parts of the natural cancer microenvironment. Using a tissue chop- per, fresh surgical tissue samples were cut in 400 μm slices and cultivated in 6-well plates for up to 6 days. The slices were processed for routine histopa- thology and immunohistochemistry. Cytokeratin stains (CK8, AE1/3) were ap- plied for determining tumor cellularity, Ki-67 for proliferation, and cleaved caspase-3 staining for apoptosis. The slices were analyzed under naive conditions and following 2–4 days in vitro exposure to 5-FU and cisplatin. The slice culture technology allowed for a good preservation of tissue morphology and tumor cell integrity during the culture period. After chemotherapy exposure, a loss of tumor cellularity and an increase in apoptosis were observed. Drug sensitivity of the tumors could be assessed. Organotypic slice cultures of gastric and es- ophagogastric junction cancers were successfully established. Cytotoxic drug effects could be monitored. They may be used to examine mechanisms of drug resistance in human tissue and may provide a unique and powerful ex vivo platform for the prediction of treatment response.