基于SEER数据库美国脊柱骨肉瘤患者数据:治疗结果及预后预测模型的建立与验证  

Data of spinal osteosarcoma patients in United States based on SEER database:construction and validation of a prediction model for treatment outcomes and prognosis

作　　者：徐志 陈运动 孙玉洁 宫宵男 李豫皖 Xu Zhi;Chen Yundong;Sun Yujie;Gong Xiaonan;Li Yuwan(Department of Orthopedics,Zhangjiagang City Fifth People’s Hospital,Zhangjiagang 215600,Jiangsu Province,China;Department of Orthopedics,First Affiliated Hospital of Xinxiang Medical College,Xinxiang 453000,Henan Province,China;Orthopedic Joint Surgery,Dongying First People’s Hospital,Dongying 257000,Shandong Province,China;Department of Orthopedics,First Affiliated Hospital of Zhejiang University School of Medicine,Hangzhou 310009,Zhejiang Province,China)

机构地区：[1]张家港市第五人民医院骨科,江苏省张家港市215600 [2]新乡医学院第一附属医院骨科,河南省新乡市453000 [3]东营市第一人民医院关节外科,山东省东营市257000 [4]浙江大学医学院附属第一医院骨科,浙江省杭州市310009

出　　处：《中国组织工程研究》2025年第30期6583-6590,共8页Chinese Journal of Tissue Engineering Research

基　　金：国家自然科学基金青年项目(82302853),项目负责人:李豫皖。

摘　　要：背景:脊柱骨肉瘤是一种罕见且侵袭性强的恶性肿瘤,现有的研究大多基于小样本量,且结果不一,难以提供可靠的临床指导。特别是在中国,由于脊柱骨肉瘤的发病率较低,相关研究较为有限,临床医生在治疗过程中缺乏有效的预后工具。目的:构建并验证基于监测、流行病学和最终结果(SEER)数据库的脊柱骨肉瘤患者生存期预测的列线图模型,以期为临床提供科学依据,尤其是对中国患者的治疗方案优化提供借鉴。方法:回顾性分析SEER数据库中2000-2021年被诊断为脊柱骨肉瘤的美国患者数据,首先通过单因素和多因素Cox比例风险模型分析筛选出与脊柱骨肉瘤特异性死亡相关的独立预后因素;随后利用这些独立预后因素,在Rstudio中使用“rms”包构建了脊柱骨肉瘤特异生存率的列线图模型。模型的区分度通过C指数进行评估,预测能力通过受试者工作特征曲线和曲线下面积值验证,校准度通过校准曲线(Calibration plot)评估,临床价值则通过决策曲线分析衡量。此外,进行Kaplan-Meier生存分析以检测列线图分组的合理性。结果与结论:①最终模型包括化疗、肿瘤尺寸、组织学类型、分级、种族和是否手术6个变量;②模型在训练集和验证集中的C指数分别为0.685和0.673,表明模型区分度良好;③校准曲线显示预测生存概率与实际生存概率一致性高;④决策曲线分析表明模型在广泛的死亡风险范围内具有较大的净收益;⑤Kaplan-Meier生存分析显示高危组和低危组患者的预后存在显著差异;⑥此次研究构建的列线图模型能够准确预测脊柱骨肉瘤患者的1年、2年和3年生存期,具有较高的临床应用价值;该模型不仅为美国患者提供了有效的生存预测工具,也为中国脊柱骨肉瘤患者的治疗方案优化提供了重要借鉴;未来研究应进一步验证该模型在不同人群中的适用性,并探索新型治疗手段对脊柱�BACKGROUND:Spinal osteosarcoma is a rare and highly aggressive malignant tumor.Most existing studies are based on small sample sizes and have inconsistent results,making it difficult to provide reliable clinical guidance.Especially in China,due to the low incidence of spinal osteosarcoma and limited related research,clinicians lack effective prognostic tools during treatment.OBJECTIVE:To construct and validate a nomogram model for predicting the survival of spinal osteosarcoma patients based on the Surveillance,Epidemiology,and End Results(SEER)database,providing scientific evidence for clinical decision-making,particularly for optimizing treatment plans for Chinese patients.METHODS:This study conducted a retrospective analysis of patient data diagnosed with spinal osteosarcoma from the SEER database between 2000 and 2021.First,independent prognostic factors associated with specific mortality from spinal osteosarcoma were identified through univariate and multivariate Cox proportional hazards models.Subsequently,these independent prognostic factors were used to construct a nomogram model for predicting survival rates of spinal osteosarcoma patients using the“rms”package in RStudio.The model’s discrimination was assessed using the C-index.Predictive ability was validated through receiver operating characteristic curves and area under the curve values.Calibration was evaluated by calibration plots,and clinical value was measured using decision curve analysis.Additionally,Kaplan-Meier survival analysis was performed to assess the rationality of the nomogram groupings.RESULTS AND CONCLUSION:(1)The final model included six variables:chemotherapy,tumor size,histological type,grade,race,and surgical intervention.(2)The C-indices of the model in the training and validation sets were 0.685 and 0.673,respectively,indicating good discrimination.(3)Calibration curves showed high consistency between predicted survival probabilities and actual survival probabilities.(4)Decision curve analysis indicated that the model pro

关 键 词：脊柱骨肉瘤 列线图模型 生存期 COX回归 预后因素 Kaplan-Meier生存分析 工程化组织构建 

分 类 号：R459.9[医药卫生—治疗学] R318[医药卫生—临床医学] R678.4