Prognostic Nomograms for Patients with Primary Sarcomatoid Carcinoma of The Urinary Bladder: Based on The SEER Database.

PURPOSE: The present study aimed to develop nomograms based on the SEER database to predict the prognosis for patients with primary sarcomatoid carcinoma of the urinary bladder (SCUB). MATERIALS AND METHODS: Patients with primary SCUB were identified in the Surveillance, Epidemiology, and End Results (SEER) database, between 1975 and 2017. Univariate and multivariable Cox analysis were conducted to identify the independent prognostic factors for developing the overall survival (OS) and cancer-specific survival (CSS) nomograms. Then, concordance index (C-index), receiver operating characteristic (ROC) curve and calibration curve were used to evaluate the accuracy of the nomogram model. In addition, the model was further compared with TNM staging system. RESULTS: A total of 238 eligible patients with primary SCUB were selected from the SEER database. As suggested by Cox-analysis, age, sex, T stage, M stage, tumor size, and surgery type of primary site were identified as the independent factors for predicting both OS and CSS. We developed OS and CSS nomograms with a favorable C-index by using these prognostic factors. The C-indexes of the OS and CSS nomogram in the present study were 0.738 (0.701-0.775) and 0.763 (0.724-0.802), which were superior to those of the AJCC TNM staging with 0.621 (0.576-0.666) and 0.637 (0.588-0.686) respectively, showing better discriminatory ability. Subsequently, the ROC curves showed that the 1-, 3- and 5-year AUCs (area under the curve) of OS nomogram (i.e., 0.793, 0.807 and 0.793) were higher than those of the TNM stage((i.e., 0.659, 0.676, 0.659). Similarly, as for CSS model, them ((i.e., 0.823, 0.804 and 0.804) were aslo exceed those of TNM stage((i.e., 0.683, 0.682, 0.682). Furthermore, the calibration curves indicated a good consistency between the predictive survival and the actual survival. Finally, patients were stratified by risk, and Kaplan-Meier survival curve suggested that the prognosis of the low-risk group was significantly better than that of the high-risk group. CONCLUSION: We developed nomograms with the SEER database, which could help predict the prognosis of SCUB individuals more accurately.