A novel prognostic nomogram for patients with surgically resected perihilar cholangiocarcinoma: a SEER-based study

Abstract

BACKGROUND: This study aimed to compare the predictive efficacy of four different lymph node (LN) staging systems on the overall survival (OS) of patients with surgically resected perihilar cholangiocarcinoma (pCCA), and construct a novel prognostic nomogram to predict OS in pCCA patients. METHODS: Patients with pCCA that underwent surgical resection between 2004 to 2016 were selected from the Surveillance, Epidemiology, and End Results (SEER) database (n=1,173). Patients were randomly divided into a modeling cohort and an internal verification cohort. To compare the prognostic efficacy of four different N staging systems [American Joint Committee on Cancer (AJCC) 7(th) and 8(th) edition N stages, lymph node ratio (LNR), and log odds of positive lymph nodes (LODDS)], we used three different evaluation methods: Harrell's index of concordance (C-index), Akaike information criterion (AIC), and area under the receiver operating characteristic (ROC) curve (AUC). Multivariate analysis was used to identify independent prognostic factors and validate LODDS in the modeling cohort. A nomogram was then constructed to predict 1-, 3-, and 5-year survival. The nomogram was validated using Harrell's C-indexes and calibration curves. RESULTS: Of the four different N staging methods, LODDS was considered to be the most effective LN staging system for evaluating the prognosis of patients with surgically resected pCCA, according to the values calculated for C-index, AUC, and AIC. After validation by C-indexes and calibration curves, the constructed nomogram accurately predicted the OS of pCCA patients. CONCLUSIONS: For patients with surgically resected pCCA, LODDS was found to be the most accurate N staging system. The novel LODDS-based nomogram constructed in this study provides an accurate method for predicting patient survival in pCCA.