LAUSR

Background Our study aimed to construct and validate prognostic nomograms for predicting survival for patients with Nonfunctional Pancreatic neuroendocrine tumor (NF-pNET). Methods This retrospective study included 1824 patients diagnosed with NF-pNET in the Surveillance, Epidemiology and End Results database between 2004 and 2016. Randomization divided the patients into training (n = 1278) and validation (n = 546) cohorts. Prognostic factors were determined using Cox regression analyses, nomograms based on AJCC 7th and 8th staging system were constructed separately. The prediction models were validated using internal validation and external validation. Results Age, year of diagnosis, primary tumor site, grade, 7th or 8th TNM stage, surgery, tumor size were determined as prognostic indicator to construct two nomograms. Harrell’s concordance index (C-index) of two nomograms exhibited a clinical predictive ability of 0.828 (95%CI, 0.808~0.849) vs 0.828 (95% CI, 0.808~0.849) in the internal verification. The c-index in the external validation was 0.812 (95%CI, 0.778~0.864) vs 0.814 (95% CI, 0.779~0.848). The predictive power of the two nomograms is comparable. Conclusions Our nomogram may be a effective tool for predicting overall survival in patients with NF-pNET. The AJCC 8th-edition system provides discrimination similar to that of the 7th-edition system.