基于SEER数据库绘制列线图分析软骨肉瘤患者预后相关因素

目的:分析软骨肉瘤流行病学特征及影响预后的相关因素，并且绘制列线图来个体化预测患者远期生存率。方法:收集SEER数据库2004至2015年诊断的1 453例软骨肉瘤患者的临床数据，回顾性分析软骨肉瘤患者流行病学特征及影响患者预后的相关危险因素，使用随机数字表法将所有纳入对象以7∶3分为建模组（1 017例）和验证组（436例）并构建列线图进行内部验证，预测软骨肉瘤患者3年、5年的生存率，使用多因素COX风险比例模型来确定独立因素，采用一致性指数(C-index)及校准曲线评估该预测模型的准确性。结果:软骨肉瘤患者男女性别比为:1.23∶1，年龄≥50岁患者占比为55.7%，最常见的发病部位是下肢骨，多因素分析提示，影响软骨肉瘤患者预后的因素包括年龄、性别、肿瘤原发部位、肿瘤大小、病理分级、AJCC TNM分期、手术方式、是否化疗、肿瘤大小；建模组与验证组1、3、5年ROC曲线AUC值分别为:0.87、0.838、0.807，0.864、0.754、0.755；列线图C-index指数为:0.805。结论:列线图可以准确预测软骨肉瘤患者生存率，具有较好的预测精度，有助于对患者进行个性化的预后评估和指导临床决策。

Nomograms for predicting overall survival and prognostic factors in patients with chondrosarcoma based on the US SEER database

Objective:The epidemiological characteristics and prognostic factors of chondrosarcoma were analyzed, and the long-term survival rate of patients was individually predicted by drawing a rosette.Methods:Clinical data of 1 453 patients with chondrosarcoma diagnosed in SEER database from 2004 to 2015 were collected. Epidemiological characteristics and prognostic risk factors of chondrosarcoma patients were analyzed retrospectively. Random number table method was used to include all the patients as 7: 3. The patients were divided into a modeling group (1 017 cases) and a validation group (436 cases), and the 3-year and 5-year survival rates of chondrosarcoma patients were predicted by using a multi-factor COX risk ratio model to determine the independent factors. The accuracy of the prediction model was evaluated by C-index and calibration curve.Results:The gender ratio of chondrosarcoma patients was 1.23∶1 The proportion of patients aged ≥50 years was 55.7%, and the most common site of disease was the lower limb bone. Multivariate analysis showed that the factors influencing the prognosis of chondrosarcoma patients included age, gender, primary site of tumor, tumor size, Grade Grade, AJCC TNM stage, surgical method, chemotherapy or not, and tumor size. The AUC values of ROC curves in the modeling group and the verification group at 1, 3 and 5 years were 0.87, 0.838, 0.807, 0.864, 0.754, 0.755, respectively. The C-index of the column chart is 0.805.Conclusion:The rosette map can accurately predict the survival rate of chondrosarcoma patients with good prediction accuracy, which is helpful for personalized prognosis assessment of patients and guide clinical decision making.