Dynamic nomogram for predicting the overall survival and cancer-specific survival of patients with gastrointestinal neuroendocrine tumor: a SEER-based retrospective cohort study and external validation.

BACKGROUND

Gastrointestinal neuroendocrine tumor (GI-net) is a rare heterogeneous tumor, and there is a lack of models to predict its prognosis. Our study aims to develop and validate two new nomograms to predict the overall survival (OS) and cancer-specific survival (CSS) of GI-net patients and investigate their application value.

METHODS

SEER*Stat 8.4.4 software was used to download clinicopathological information of GI-net patients between 2010 and 2015 from the Surveillance, Epidemiology, and End Results (SEER) database. These patients were randomly divided into a training group (n=3007) and an internal-validation group (n=1289) at a 7:3 ratio. Patients from the Fourth Hospital of Hebei Medical University were enrolled in this study to form the external-validation group (n=86). Univariate and multivariate Cox analyses were performed to explore the independent prognostic factors and establish two nomograms. The concordance index (C-index), area under the time-dependent receiver operating characteristic curve (AUC), calibration curve, and decision curve analysis (DCA) were used to evaluate the nomograms. X-tile was used to divide GI-net patients into high-, medium-, and low-risk groups. Kaplan-Meier (KM) curves and log-rank tests were used to compare survival differences among the three groups.

RESULTS

Seven variables (age, site, size, grade, M stage, surgery, and chemotherapy) were selected to establish the nomogram for OS, and 6 variables (age, size, grade, M stage, surgery, and chemotherapy) were selected for CSS. The C indices (0.785, 0.813, and 0.936 in the training, internal-validation, and external-validation groups for OS; 0.888, 0.893, and 0.930 for CSS, respectively) and AUCs (≥0.7) indicated that the nomograms had satisfactory discriminative ability. Calibration curve analysis and DCA revealed that the nomogram had a satisfactory ability to predict OS and CSS. KM curves indicated that each of the two nomograms clearly differentiated the high-, medium-, and low-risk groups. In addition, two online risk calculators were developed to predict the OS and CSS of these patients visually.

CONCLUSIONS

Our nomograms may play an important role in predicting 3- and 5-year OS and CSS for GI-net patients. Risk stratification systems and online risk calculators can be utilized in clinical practice to help doctors create personalized treatment plans.