Abstract
BACKGROUND: Malignant bone tumors are rare and highly heterogeneous tumors with poor clinical prognosis and numerous challenges in treatment. Traditional prognostic models may lead to biased assessment of tumor-specific mortality risk due to failure to account for competing risk events such as non-tumor causes of death. The objective of this study was to develop and validate a competing risk model for cancer-specific survival (CSS) in patients diagnosed with malignant bone tumors, and to improve the accuracy of prognostic prediction. METHODS: A total of 3,508 patients with osteosarcoma, chondrosarcoma, and Ewing sarcoma from the Surveillance, Epidemiology, and End Results (SEER) database between 2000 and 2022 were included, and divided into a training set (2,455 cases) and a validation set (1,053 cases) at a ratio of 7:3. Univariate and multivariate Cox regression analyses were used to screen independent risk factors for cancer-specific mortality (CSM), construct a competing risk model, and draw a nomogram. The model performance was evaluated using the consistency index (C-index), area under the receiver operating characteristic curve (AUC), calibration curve, and decision curve analysis (DCA), and compared with the TNM (tumor, node, metastasis) staging system. RESULTS: Age, gender, primary tumor site, tumor size, clinical stage, surgery, TNMstage, pathological type (Ewing sarcoma), radiotherapy, and chemotherapy were identified as independent risk factors for CSM. The C-index of the model was 0.77 [95% confidence interval (CI): 0.75-0.79] in the training set and 0.79 (95% CI: 0.77-0.82) in the validation set, with AUC >0.8 in both. The calibration curve showed a high degree of agreement between predicted and actual survival rates. DCA results indicated that the clinical net benefit of this model was significantly better than the TNM staging system. Risk stratification showed that the 5-year CSM rate in the high-risk group (65%) was significantly higher than that in the low-risk group (22%, P<0.001). CONCLUSIONS: The competing risk model constructed in this study can accurately predict the CSS probability of patients with malignant bone tumors, with better performance than traditional staging systems, providing a new tool for the development of individualized treatment plans and the identification of high-risk patients.