Outcome after pancreaticoduodenectomy for T3 adenocarcinoma: A multivariable analysis from the UK Vascular Resection for Pancreatic Cancer Study Group.

BACKGROUND: Most resectable pancreatic cancers are classified as T3, including those involving the porto-mesenteric vein. Survival and perioperative morbidity for venous resection have been found to be comparable to standard resection. We investigate factors associated with short and long term outcomes in pancreaticoduodenectomy with (PDVR) and without (PD) venous resection exclusively for T3 adenocarcinoma of the head of the pancreas. METHODS: This is a UK multicenter retrospective cohort study assessing outcomes in patients undergoing PD and PDVR. All consecutive patients with T3 only adenocarcinoma of the head of the pancreas undergoing surgery between December 1998 and June 2011 were included. Multivariable logistic and proportional hazards regression analyses were performed to determine the association between the surgical groups and in-hospital mortality (IHM) and overall survival (OS). RESULTS: 1070 patients were included of whom 840 (78.5%) had PD and 230 (21.5%) had PDVR. Factors independently associated with IHM were a high creatinine (aHR 1.14, p = 0.02), post-operative bleeding (aHR 2.86, p = 0.04) and a re-laparotomy (aHR 8.42, p = 0.0001). For OS, multivariable analyses identified R1 resection margin status (aHR 1.22, p = 0.01), N1 nodal status (aHR 1.92, p = 0.0001), perineural invasion (aHR 1.37, p = 0.002), tumour size >20mm (aHR 0.63, p = 0.0001) and a relaparotomy (aHR 1.84, p = 0.0001) to be independently associated with overall mortality. CONCLUSION: This study on T3 adenocarcinoma of the head of the pancreas suggests that IHM is strongly associated with perioperative complications whilst OS is affected by histological parameters. Detailed pre-operative disease evaluation and advances in oncological treatment have the potential to improve OS.