ABSTRACT
Objective To investigate the independent risk factors affecting prognosis of patients with retinoblastoma (RB) and construct a nomogram to predict prognosis of patients with RB. Methods Data of 759 RB patients were collected from the SEER database. Patients were randomly assigned to the training group and validation group in a 7:3 ratio. Univariate and multivariate Cox proportional hazard regression analyses were used to determine the independent prognostic factors, based on which a nomogram was constructed. C index, calibration curve, and ROC curve were used to evaluate the predictive efficiency and calibration degree of the nomogram. Results Multivariate analysis identified independent risk factors associated with overall survival, namely, T stage and SEER stage. The C-index of SEER training set was 0.765 (95%CI: 0.744-0.786), the calibration curve was drawn, and the observed and predicted values overlapped well, indicating good consistency. The ROC curve showed that the nomogram could accurately predict three-year (AUC=0.743), five-year (AUC=0.734) and 10-year (AUC=0.720) survival rates of RB patients. Conclusion T stage and SEER stage are independent risk factors related to prognosis of RB patients, and the nomogram can accurately predict the three-year, five-year, and 10-year overall survival rates of patients.