A novel web-based dynamic prognostic nomogram for gastric signet ring cell carcinoma: a multicenter population-based study.

Abstract

Gastric signet ring cell carcinoma (GSRCC) is a rare and highly malignant disease with a poor prognosis. To assess the overall survival (OS) and cancer-specific survival (CSS) of patients with GSRCC, prognostic nomograms were developed and validated using common clinical factors. This retrospective cohort study included patients diagnosed with GSRCC between 2011 and 2018 from the National Cancer Center (n = 1453) and SEER databases (n = 2745). Prognostic nomograms were established by identifying independent prognostic factors using univariate and multivariate Cox regression analyses. The calibration curve and C-index were used to assess the predictions. The clinical usefulness of the survival prediction model was further evaluated using the DCA and ROC curves. The models were internally validated in the training cohort and externally validated in the validation cohort. Two web servers were created to make the nomogram easier to use. Patients with GSRCC were divided into training (n = 2938) and validation (n = 1260) cohorts. The nomograms incorporated six predictors: age, race, tumor site, tumor size, N stage, T stage, and AJCC stage. Excellent agreement was observed between the internal and exterior calibration plots for the GSRCC survival estimates. The C-index and area under the ROC curve were roughly greater than 0.7. Both nomograms had adequate clinical efficacy, as demonstrated by the DCA plots. Furthermore, we developed a dynamic web application utilizing the constructed nomograms available at https://jiangyujuan.shinyapps.io/OS-nomogram/ and https://jiangyujuan.shinyapps.io/DynNomapp-DFS/. We developed web-based dynamic nomograms utilizing six independent prognostic variables that assist physicians in estimating the OS and CSS of patients with GSRCC.