Predictors of postoperative and oncological outcome after surgical resection of pancreatic cancer

Pancreatic cancer resection is a major undertaking for most patients, associated with high rates of morbidity and mortality, leading to a delay in the commencement of much needed systemic adjuvant therapy. This is compounded with the long-term effects of exocrine pancreatic insufficiency, post-pancreatectomy diabetes mellitus, malnutrition, and disorders of gut motility. Therefore, a significant proportion of patients endure the morbid effects of pancreatic resection whilst deriving minimal oncological benefit. In order to optimise care of the pancreatic cancer patient, there is a need to be able to identify predictors of both postoperative and survival outcomes after resection. With regard to postoperative outcome, novel predictors of its key determinant - post-operative pancreatic fistula (POPF) - are evaluated in this thesis. For the first time, intra-operative assessment of peri-pancreatic fluid for amylase concentration is shown to be a predictor of POPF. In subsequent chapters, the concept of the acinar score is introduced. The acinar score and post-operative pancreatitis are shown to be inherent to the development of POPF. A method is also described by which pre-operative computed tomography may be used to accurately predict the acinar score and POPF. Regarding oncological outcome, a panel of prognostic biomarkers is rigorously evaluated for its ability to predict survival after surgery. Patterns of expression in a triple biomarker panel are shown to accurately stratify patient survival postoperatively and to represent various clinical phenotypes of pancreatic cancer. Lastly, this thesis demonstrates that this triple biomarker panel, which also correlates with known genomic subtypes of pancreatic cancer, may be used to determine whether a patient with resectable or borderline resectable pancreatic cancer will benefit from either a neoadjuvant treatment approach or upfront surgical resection