Identifying predictive and prognostic markers in localised oesophagogastric cancer

Abstract

The survival of patients with localised oesophagogastric (OG) cancer is poor even with multimodality therapy and modern surgical techniques. The need to identify robust but easily-reproducible prognostic and predictive biomarkers is paramount for better patient selection, as is the investigation of rational therapeutic combinations for patients with poor prognostic tumours. Pooled analysis of data from five large randomised controlled trials shows that females suffer more toxicity from cytotoxic chemotherapy but their survival is better than males following neoadjuvant chemotherapy and surgery. Older patients can achieve the same outcome as younger patients when treated with the same perioperative paradigm. Central pathological assessment of samples taken from over 1500 clinical trial patients confirms that the lymph node status in the resection specimen following neoadjuvant chemotherapy is the single most important prognostic factor in resected OG cancer. Determination of regression grading within either the primary tumour or metastatic lymph nodes is not required to define prognosis any further. The impact of chemotherapy on the immune tumour microenvironment of patients treated with current standard-of-care perioperative FLOT chemotherapy was evaluated via multiplex immunofluorescence. Patients with higher PD-1+ immune cell infiltrates at baseline have better survival than patients with low levels, with a similar trend also observed for high activated CD8+ T-cells and high FOXP3+ T-cells. In addition, PD-1+ immune cells predict better tumour responses to chemotherapy. Due to the immunogenic activity of chemotherapy agents within the FLOT regimen, synergy with checkpoint inhibition is likely to improve responses. Interim analysis results from the phase II ICONIC (Perioperative Immuno-Chemotherapy in Operable oesophageal aNd gastric Cancer) study show this novel combination to be safe with no unexpected toxicity or surgical detriment. The results presented in this thesis outline an important evaluation of clinical and pathological markers which have been defined from large trial datasets. In addition, this thesis has sought to investigate the interplay between chemotherapy and tumour infiltrating immune cells as we head towards chemo-immunotherapy possibly playing a role in the management of these tumours with the hope of improving patient outcomes.