Establishment and validation of a nomogram to predict overall survival of patients with gastric neuroendocrine neoplasms / 中华胃肠外科杂志

ABSTRACT

Objective: To establish a novel nomogram to predict overall survival of patients with gastric neuroendocrine neoplasms (g-NEN). Methods: A case control study was conducted. Clinicopathological and follow-up data of patients with g-NEN who were treated in two academic medical centers in Southern China between July 2008 and June 2018 were retrospectively collected, including 174 patients from Sun Yat-sen University Cancer Center and 102 patients from the First Affiliated Hospital of Sun Yat-sen University. Univariate survival analysis using Kaplan-Meier method and multivariate analysis using Cox regression were performed to identify prognostic factors. A nomogram was subsequently established based on prognostic factors. Harrell's concordance index (C-index), receiver operating characteristic (ROC) curve, calibration curve and decision curve analysis (DCA) were used to verify the performance of the model according to differentiation, calibration and clinical utility. Results: A total of 276 patients were enrolled in the study, of whom 189 patients were male and 87 were female. The age at diagnosis was below 60 years old in 150 patients and 60 years or older in 126 patients. There were patients diagnosed with gastric neuroendocrine carcinoma (g-NEC) and 101 patients with gastric neuroendocrine tumor (g-NET). The number of patients with primary tumor locating at upper, middle and lower parts of stomach was 131, 98 and 47, respectively. As for TNM stage, 72 patients were categorized as stage I, 26 patients stage II, 93 patients stage III, and 85 patients stage IV. Univariate analysis indicated that age, pathological type, primary site, Ki-67 index, T stage, N stage, and M stage were associated with overall survival of g-NEN patients (all P<0.05). Multivariate regression analysis testified that high Ki-67 index, advanced T stage and advanced M stage were independent prognostic factors (all P<0.05). The C-index of the nomogram was 0.806 (95%CI 0.769-0.863). The calibration curve of the nomogram showed that the predicted survival rate was consistent with the actual survival rate in g-NEN patients. The ROC curves and DCA showed that the nomogram had better differentiation and clinical utility than the American Joint Committee on Cancer (AJCC) 8th TNM staging system (the area under the ROC curve was 0.862 vs. 0.792). Conclusion: The first nomogram to predict overall survival of patients with g-NEN is established and verified in this study, which provides individual prediction of 3-year overall survival rate and is applicable to both g-NET and g-NEC patients.