Establishment and Validation of A Prognostic Nomogram for Pediatric Ependymoma Based on SEER Database

Objective To establish and validate a Nomogram for predicting the survival of patients with pediatric ependymoma based on SEER database.

Methods We collected the clinicopathological data from 1975 to 2016 in the SEER database. Univariate and multivariate Cox proportional hazard regression models were used to identify potential predictors. A Nomogram was constructed to predict 5- and 10-year overall survival of patients with pediatric ependymoma. The consistency index (C-index), receiver operating characteristic curve and calibration curve were used to verify the discrimination and accuracy of the Nomogram. The decision curve analysis was performed to verify the clinical applicability of the Nomogram.

Results A Nomogram model was established based on multivariate Cox proportional hazards model of training set. C-index values of the Nomogram were 0.713 (95%CI: 0.680-0.747) and 0.734 (95%CI: 0.681-0.787) in the training and validation sets, respectively. ROC curves also showed good discrimination in the training set. The calibration curves showed satisfactory consistency between Nomogram and ideal models. The decision curve analysis demonstrated the considerable clinical usefulness of the Nomogram.

Conclusion The Nomogram model is constructed based on age at diagnosis, gender, race, primary tumor sites, tumor grade, surgery treatment and SEER registry to predict the survival of patients with pediatric ependymoma. It has good discrimination and accuracy degree, providing useful guidance to make more accurate and personalized survival prediction for patients in clinic.