A Nomogram to Predict Individual Survival of Patients with Liver-Limited Metastases from Gastroenteropancreatic Neuroendocrine Neoplasms: A US Population-Based Cohort Analysis and Chinese Multicenter Cohort Validation Study

Introduction: Although gastroenteropancreatic neuroendocrine neoplasms (GEP-NENs) with liver metastasis encompass a wide variety of clinical conditions with various prognosis, no statistical model for predicting the prognosis of these patients has been established. We sought to establish a more elaborative and individualized nomogram to predict survival of patients with liver-limited metastatic GEP-NENs. In addition, this nomogram was validated by both the Surveillance, Epidemiology, and End Results (SEER) database and a Chinese multicenter cohort. Methods: Patients diagnosed with GEP-NENs with liver-limited metastasis between 2010 and 2016 were identified from the SEER database. Kaplan-Meier survival analysis was performed to analyze survival outcomes. A nomogram was established based on the independent prognostic variables identified from univariate and multivariate Cox regression analyses. The nomogram was evaluated in both an internal validation SEER dataset and an external validation dataset composed of patients from the Chinese multicenter cohort. Results: A total of 1,474 patients from the SEER database and 192 patients from the multicenter cohort were included. Age, tumor size, differentiation, primary tumor resection, and liver metastasis resection were identified as independent prognostic factors by univariate and multivariate Cox analyses and were verified by Kaplan-Meier survival analysis (all p < 0.0001). A nomogram was developed and validated by calibration curves and areas under the curve of the external validation cohort, which showed good consistency and veracity in predicting overall survival. Conclusion: A nomogram was developed for the first time to predict the survival of patients with liver-limited metastases from GEP-NENs. Both internal and external validation demonstrated excellent discrimination and calibration of our nomogram. Based on this prognostic model, clinicians could develop more personalized treatment strategies and surveillance protocols.