Chen Shulin, Tian Liru, Li Chuan, Zhong Dongmei, Wang Tingting, Chen Yuyu, Zhou Taifeng, Yang Xiaoming, Liao Zhiheng, Xu Caixia
Research Center for Translational Medicine, the First Affiliated Hospital, Sun Yat-sen University, 58 Zhongshan Road 2, Guangzhou, Guangdong 510080, P.R. China.
Department of Clinical Laboratory, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Sun Yat-sen University Cancer Center, Guangzhou 510060, P. R. China.
J Cancer. 2025 Mar 10;16(7):2075-2086. doi: 10.7150/jca.105590. eCollection 2025.
Osteosarcoma (OSC) is a high-morbidity bone cancer with an unsatisfactory prognosis. Timely and accurate assessment the overall survival (OS) and progression-free survival (PFS) in patients with OSC are required to guide and select the best treatment. This study aimed to develop a simple, convenient and low-cost prognostic model based on clinical characteristics and blood biomarkers for predicting OS and PFS in OSC patients. Overall, 158 patients with OSC included from the Sun Yat-sen University Cancer Center in this retrospective study. LASSO-Cox algorithm was used to shrink predictive factor size and established a prognostic risk model for predicting OS and PFS in OSC patients. The predictive ability of the survival model was compared to the Tumor Node Metastasis (TNM) stage and clinical treatment by concordance index (C-index), time-dependent receiver operating characteristic (td-ROC) curve, decision curve analysis (DCA), net reclassification improvement index (NRI), and integrated discrimination improvement index (IDI). Based on results from the LASSO-Cox method, gender, family history of cancer, monocyte (M), red blood cell (RBC), lactic dehydrogenase (LDH), and cystatin C (Cys-C) were identified to construct a novel predictive model for the OSC patients. The C-index of the prognostic model to predict OS and PFS were 0.713 (95% CI = 0.630 - 0.795) and 0.636 (95% CI = 0.577 - 0.696), respectively, which were higher than the OS and PFS of TNM stage and clinical treatment. Td-ROC curve and DCA of the predictive model also demonstrated good predictive accuracy and discriminatory power of OS and PFS compared to TNM stage and treatment. Moreover, the prognostic model performed well across all time frames (1-, 3-, and 5-year) with regards to the IDI and NRI in comparison to the TNM stage, and clinical treatment. The simple, convenient and low-cost prognostic model we developed demonstrated favorable performance for predicting OS and PFS in OSC patients, which may serve as a useful tool for physicians to provide personalized survival prediction for OSC patients.
骨肉瘤（OSC）是一种发病率高且预后不理想的骨癌。需要及时、准确地评估骨肉瘤患者的总生存期（OS）和无进展生存期（PFS），以指导和选择最佳治疗方案。本研究旨在基于临床特征和血液生物标志物开发一种简单、便捷且低成本的预后模型，用于预测骨肉瘤患者的OS和PFS。在这项回顾性研究中，共纳入了158例来自中山大学肿瘤防治中心的骨肉瘤患者。采用LASSO - Cox算法缩小预测因子规模，并建立了用于预测骨肉瘤患者OS和PFS的预后风险模型。通过一致性指数（C指数）、时间依赖性受试者工作特征（td - ROC）曲线、决策曲线分析（DCA）、净重新分类改善指数（NRI）和综合判别改善指数（IDI），将生存模型的预测能力与肿瘤 - 淋巴结 - 转移（TNM）分期及临床治疗进行比较。基于LASSO - Cox方法的结果，确定了性别、癌症家族史、单核细胞（M）、红细胞（RBC）、乳酸脱氢酶（LDH）和胱抑素C（Cys - C），以构建骨肉瘤患者的新型预测模型。该预后模型预测OS和PFS的C指数分别为0.713（95%CI = 0.630 - 0.795）和0.636（95%CI = 0.577 - 0.696），高于TNM分期和临床治疗的OS及PFS。与TNM分期和治疗相比，预测模型的td - ROC曲线和DCA也显示出对OS和PFS具有良好的预测准确性和鉴别力。此外，与TNM分期和临床治疗相比，该预后模型在所有时间框架（1年、3年和5年）的IDI和NRI方面表现良好。我们开发的简单、便捷且低成本的预后模型在预测骨肉瘤患者的OS和PFS方面表现良好，这可能为医生为骨肉瘤患者提供个性化生存预测提供有用工具。
