Survival Nomograms for Patients with Pancreatic Neuroendocrine Tumors: A Population-based Study

Background: Although Pancreatic neuroendocrine tumors(PNETs) considered as indolent tumors, most patients are diagnosed at an advanced stage. Herein, we aimed to establish a nomogram to predict the survival of PNETs patients for clinical use via Surveillance, Epidemiology, and End Results (SEER) database. Methods: Based on the SEER program, the data of 1103 patients with PNETs were enrolled and randomly divided into training set and validation set. We performed Kaplan-Meier analysis, Cox proportional hazard regression analysis in training set to evaluate the value of prognostic factors. A nomogram was constructed obtained these independent prognostic factors for predicting overall survival(OS) and specific-cancer survival(CSS). C-index, calibration curve, decision curve analysis were used to evaluate the predictive accuracy of the nomogram. Results: Age, primary site, TNM stage, grade, and surgery were associated with OS and CSS in the multivariate models. Nomograms were established depend on these risk factors and had a better discrimination power than TMN stage. The validation technologies showed that the nomogram was able to predict 3- and 5-year OS and CSS accurately, and also proved the superiority. Age, primary site, TNM stage, grade, and surgery were associated with OS and CSS in the multivariate models. Conclusions: Nomograms were established depend on these risk factors and had a better discrimination power than TMN stage. The validation technologies showed that the nomograms were able to predict 3- and 5-year OS and CSS accurately, and also proved the superiority.