摘要
目的分析肺癌脊柱转移瘤开放手术患者术后的生存预后,并评估现有生存期预测模型的准确度。方法回顾性收集2019年1月至2021年11月于广东省人民医院骨肿瘤科收治的76例肺癌脊柱转移行开放手术的患者资料,男49例、女27例,年龄(59.3±10.3)岁(范围26~80岁)。对骨转移数目、病理类型、内脏转移情况、表皮生长因子受体(epidermal growth factor receptor,EGFR)编码基因突变是否为敏感的靶向突变位点、血清碱性磷酸酶(alkaline phosphatase,ALP)水平、血红蛋白水平、Frankel分级等与术后生存期的关系进行Cox逻辑回归分析,绘制Kaplan-Meier曲线,确定潜在的预后因素。通过绘制受试者工作特征曲线(receiver operating characteristic curve,ROC),验证Tomita评分、Tokuhashi修订评分(2005年)、Katagiri New评分、新英格兰脊柱转移评分系统(New England spinal metastasis score,NESMS)、骨骼肿瘤学研究组(Skeletal Oncology Research Group,SORG)机器学习算法在预测术后生存期的准确度。结果患者术后随访时间18.0(2.3,36.0)个月,术后中位生存期12.6个月[95%CI(10.8,14.4)],6个月生存率71.6%,12个月生存率为52.0%。Cox回归分析示ALP[HR=0.23,95%CI(0.11,0.48),P<0.001]、血红蛋白[HR=4.48,95%CI(2.07,9.70),P<0.001]、EGFR突变情况[HR=2.22,95%CI(1.04,4.76),P=0.040]是影响患者生存期的危险因素。Tomita评分、Tokuhashi修订评分(2005年)、Katagiri New评分、NESMS对术后1年死亡率预测准确度分别为58.7%、65.7%、70.5%、65.0%,对6个月死亡率预测准确度分别为63.7%、62.2%、61.2%、56.8%。SORG机器学习算法术后1年和90d死亡率预测准确度分别为81.1%和67.5%。结论无EGFR突变、ALP>164 U/L和血红蛋白≤125 g/L是脊柱肺癌脊柱转移患者术后生存期的危险因素;SORG机器学习算法对肺癌脊柱转移患者术后生存率的预测具有良好的准确度。
Objective To analyze the prognostic factors and evaluate the accuracy of existing survival prediction models in patients with lung cancer-derived spinal metastases who have undergone open surgery.Methods According to the inclusion criteria,the data of 76 patients with spinal metastasis of lung cancer who underwent open surgery in the department of Orthopedics in Guangdong Provincial People's Hospital were collected from January 2019 to November 2021.The relationship between the number of bone metastasis,pathological type,visceral metastasis,epidermal growth factor receptor mutation,serum alkaline phosphatase(ALP),hemoglobin(Hb),Frankel grade and postoperative survival time in 76 cases was analyzed by Cox logical regression analysis and Kaplan-Meier method to determine the potential prognostic factors.The accuracy of Tomita score,Tokuhashi revised score,Katagiri New score,New England Spinal Metastasis Score score(NESMS)and Skeletal Oncology Research Group(SORG)machine learning algorithm in predicting postoperative survival time was verified by drawing receiver operating characteristic(ROC)curve.Results The median follow-up time of the patients was 18.0 months(2.3-36.0 months).The median survival time was 12.6 months[95%CI(10.8,14.4)].The survival rates at 6 and 12 months after operation were 71.6%and 52.0%,respectively.Multivariate regression analysis showed that ALP[HR=0.23,95%CI(0.11,0.48),P<0.001],Hb[HR=4.48,95%CI(2.07,9.70),P<0.001]and EGFR mutation[HR=2.22,95%CI(1.04,4.76),P=0.040]were independent predictors of prognosis.The accuracy of Tomita score,Tokuhashi revised score(2005),Katagiri New score and NESMS score in predicting 1-year mortality was 58.7%,65.7%,70.5%and 65%respectively,and the accuracy in predicting 6-month mortality was 63.7%,62.2%,61.2%and 56.8%respectively.The accuracy of SORG machine learning algorithm in predicting 1-year and 90 d mortality was 81.1%,67.5%,respectively.Conclusion No EGFR mutation,ALP>164 U/L and Hb≤125 g/L were risk factors affecting the survival of patients with spinal
作者
钟国庆
王晓岚
周洁龙
何玥
曾龙辉
解居宁
赖华昊
严渊
姚孟宇
程实
张余
Zhong Guoqing;Wang Xiaolan;Zhou Jielong;He Yue;Zeng Longhui;Xie Juning;Lai Huahao;Yan Yuan;Yao Mengyu;Cheng Shi;Zhang Yu(Department of Bone Oncology,Guangdong Provincial People's Hospital(Guangdong Academy of Medical Sciences),Guangzhou 510080,China)
出处
《中华骨科杂志》
CAS
CSCD
北大核心
2022年第24期1605-1614,共10页
Chinese Journal of Orthopaedics
基金
国家自然科学基金(U21A2084)。
关键词
脊柱
肺肿瘤
肿瘤转移
存活率
预测
Spine
Lung neoplasms
Neoplasm metastasis
Survival rate
Forecasting
