T Imamura1, Y Yamamoto1, T Sugiura1, Y Okamura1, T Ito1, R Ashida1, K Ohgi1, A Todaka2, A Fukutomi2, T Aramaki3, K Uesaka1. 1. Division of Hepato-Biliary-Pancreatic Surgery, Shizuoka Cancer Centre, Shizuoka, Japan. 2. Division of Gastrointestinal Oncology, Shizuoka Cancer Centre, Shizuoka, Japan. 3. Division of Interventional Radiology, Shizuoka Cancer Centre, Shizuoka, Japan.
Abstract
BACKGROUND: The length of tumour-vein contact between the portal-superior mesenteric vein (PV/SMV) and pancreatic head cancer, and its relationship to prognosis in patients undergoing pancreatic surgery, remains controversial. METHODS: Patients diagnosed with pancreatic head cancer who were eligible for pancreatoduodenectomy between October 2002 and December 2016 were analysed. The PV/SMV contact was assessed retrospectively on CT. Using the minimum P value approach based on overall survival after surgery, the optimal cut-off value for tumour-vein contact length was identified. RESULTS: Among 491 patients included, 462 underwent pancreatoduodenectomy for pancreatic head cancer. PV/SMV contact with the tumour was detected on preoperative CT in 248 patients (53·7 per cent). Overall survival of patients with PV/SMV contact exceeding 20 mm was significantly worse than that of patients with a contact length of 20 mm or less (median survival time (MST) 23·3 versus 39·3 months; P = 0·012). Multivariable analysis identified PV/SMV contact longer than 20 mm as an independent predictor of poor survival, whereas PV/SMV contact greater than 180° was not a predictive factor. Among patients with a PV/SMV contact length exceeding 20 mm on pretreatment CT, those receiving neoadjuvant therapy had significantly better overall survival than patients who had upfront surgery (MST not reached versus 21·6 months; P = 0·002). CONCLUSION: The length of PV/SMV contact predicts survival, and may be used to suggest a role for neoadjuvant therapy to improve prognosis.
BACKGROUND: The length of tumour-vein contact between the portal-superior mesenteric vein (PV/SMV) and pancreatic head cancer, and its relationship to prognosis in patients undergoing pancreatic surgery, remains controversial. METHODS:Patients diagnosed with pancreatic head cancer who were eligible for pancreatoduodenectomy between October 2002 and December 2016 were analysed. The PV/SMV contact was assessed retrospectively on CT. Using the minimum P value approach based on overall survival after surgery, the optimal cut-off value for tumour-vein contact length was identified. RESULTS: Among 491 patients included, 462 underwent pancreatoduodenectomy for pancreatic head cancer. PV/SMV contact with the tumour was detected on preoperative CT in 248 patients (53·7 per cent). Overall survival of patients with PV/SMV contact exceeding 20 mm was significantly worse than that of patients with a contact length of 20 mm or less (median survival time (MST) 23·3 versus 39·3 months; P = 0·012). Multivariable analysis identified PV/SMV contact longer than 20 mm as an independent predictor of poor survival, whereas PV/SMV contact greater than 180° was not a predictive factor. Among patients with a PV/SMV contact length exceeding 20 mm on pretreatment CT, those receiving neoadjuvant therapy had significantly better overall survival than patients who had upfront surgery (MST not reached versus 21·6 months; P = 0·002). CONCLUSION: The length of PV/SMV contact predicts survival, and may be used to suggest a role for neoadjuvant therapy to improve prognosis.