Number of evaluated lymph nodes and positive lymph nodes, lymph node ratio, and log odds evaluation in early-stage pancreatic ductal adenocarcinoma: numerology or valid indicators of patient outcome?
BACKGROUND According to some studies, the number of lymph nodes with metastases in relation to the total number of removed lymph nodes, the lymph node ratio (LNR), is one of the most powerful predictors of survival after resection in patients with pancreatic cancer. However, contradictory results have been reported, and small sample sizes of the cohorts and different definitions of a microscopic positive resection margin (R1) hamper the interpretation of data. METHODS The predictive value of LNR for 3-year survival was assessed using a Cox proportional hazards model. From 1992 to 2012, all patients with pancreatic and periampullary cancer operated on with pancreatoduodenectomy were selected from a database. Clinicopathological characteristics were analysed. Microscopic positive resection margin was defined as the microscopic presence of tumour cells within 1 mm of the margins. A nomogram was created. RESULTS Some 760 patients were included. Predictive factors for death in 350 patients with pancreatic ductal adenocarcinoma included in the nomogram were: R1 resection (hazard ratio (HR) 1·55, 95 per cent c.i. 1·07 to 2·25), poor tumour differentiation (HR 2·78, 1·40 to 5·52), LNR above 0·18 (HR 1·75, 1·13 to 2·70) and no adjuvant therapy (HR 1·54, 1·01 to 2·34). The C statistic was 0·658 (0·632 to 0·698), and calibration was good (Hosmer-Lemeshow χ(2) = 5·67, P =0·773). LNR and poor tumour differentiation (HR 4·51 and 3·30 respectively) were also predictive in patients with distal common bile duct (CBD) cancer. LNR, R1 resection and jaundice were predictors of death in patients with ampullary cancer (HR 7·82, 2·68 and 1·93 respectively). CONCLUSION LNR is a common predictor of poor survival in pancreatic, distal CBD and ampullary cancer.