Individualized Prediction of Overall Survival for Metastatic Ewing sarcoma of Bone

Abstract Background: Few models have been used to estimate the survival rate of patients metastatic Ewing sarcoma of bone are scarce. We aimed to develop nomograms for predicting 3-, 5-year survival for these patients.Methods: We extracted 686 cases of metastatic Ewing's sarcoma diagnosed between 1973 and 2016 from the Surveillance, Epidemiological and End Results (SEER) database. Univariate and multivariate Cox analysis were used to determine independent prognostic factors. The nomograms are based on the results of multivariate Cox analysis. We also evaluate the performance of these prediction models through the analysis of time-dependent receiver operating characteristic curve, concordance index, calibration curve and decision curve.Results: Age, surgery, tumor size, treatment method and chemotherapy were considered to be important predictors of overall survival of bone metastatic Ewing's sarcoma. Based on these factors, the nomogram models were established and verified internally. These models have good identification and calibration characteristics. A risk classification system based on nomogram has also been constructed to promote risk stratification of metastatic Ewing's sarcoma and to optimize clinical management.Conclusions: We developed the first nomograms and corresponding risk classification system to predict the survival of patients with bone metastatic Ewing's sarcoma. These easy-to-use tools can help oncologists and surgeons make accurate survival assessments.