椎旁功能性肌肉横截面积预测腰椎间盘突出症术后复发的Nomogram建立与验证  

Nomogram development and validation for predicting postoperative recurrence of lumbar disc herniation based on paraspinal functional muscle cross-sectional area

作　　者：唐明 曾凡袆 常新东 方晴晴 贺鸣飞 殷世武[1,2] TANG Ming;ZENG Fanyi;CHANG Xindong;FANG Qingqing;HE Mingfei;YIN Shiwu(Department of Interventional Radiology and Vascular Surgery,Hefei Hospital Affiliated to Anhui Medical University(the Second People s Hospital of Hefei),Hefei 230011,Anhui,China;The Fifth Clinical College of Medicine,Anhui Medical University,Hefei 230032,Anhui,China)

机构地区：[1]安徽医科大学附属合肥医院∥合肥市第二人民医院介入血管疼痛科,安徽合肥230011 [2]安徽医科大学第五临床医学院,安徽合肥230032

出　　处：《暨南大学学报（自然科学与医学版）》2023年第6期602-611,共10页Journal of Jinan University(Natural Science & Medicine Edition)

基　　金：合肥市卫生健康委员会应用医学研究项目(Hwk2021zc001);安徽医科大学研究生科研与实践创新项目(YJS20230207)。

摘　　要：目的:建立腰椎间盘突出症患者行经皮椎间孔镜手术术后复发(PRLDH)的个体化预测模型。方法:回顾性收集2019年6月至2023年1月期间接受椎间孔镜(PETD)手术治疗的227例患者资料及28个其他临床特征。采用Lasson回归和多因素Logistic回归筛选出影响PRLDH发生的独立危险因素,建立Nomogram模型并进行内部验证,评估本模型的准确性、校准性和临床效用。结果:患者的椎间盘夹角、患侧多裂肌功能表面积、Modic改变和BMI是影响PRLDH发生的独立预测因素(P<0.05)。基于筛选出的预测因素建立Nomogram模型。内部验证结果显示,Nomogram模型预测发生PRLDH的曲线下面积为0.867,95%CI:0.787~0.946;Hosmer-Lemeshow拟合优度检验χ^(2)=10.154,P=0.254,校准曲线显示观测值与预测值之间保持较好一致性。决策曲线分析(DCA)预测PRLDH发生的阈值为0.05~0.91,本预测模型提供了临床净收益。临床影响曲线(CIC)表明在阈值概率范围内,预期的复发患者总是多于实际的复发患者。此外,Nomogram模型临床净收益均高于单个危险因素。结论:基于椎旁肌肉构建的Nomogram预测模型具有良好的预测能力,有助于个性化评估PRLDH的发生风险。Objective:The study aimed to create a personalized predictive model for postoperative recurrent lumbar disc herniation(PRLDH)following percutaneous endoscopic transforaminal discectomy(PETD)in patients.Methods:Data from 227 patients who underwent PETD surgery between June 2019 and January 2023,including 28 additional clinical features were collected.Independent risk factors for PRLDH were identified using Lasso and multiple-factor logistic regression,leading to the development of a Nomogram model.Internal validation was performed to assess model accuracy,calibration,and clinical utility.Results:Intervertebral disc angles,functional cross-sectional area of the ipsilesional multifidus muscle,Modic changes,and BMI were identified as independent predictive factors significantly associated with the occurrence of PRLDH(P<0.05).A Nomogram model was developed based on selected predictors.Internal validation resulted in an AUC of 0.867(95%CI:0.787~0.946)for predicting PRLDH.The Hosmer-Lemeshow test yielded non-significance(χ^(2)=10.154,P=0.254),and the calibration curve showed good consistency between observed and predicted values.DCA identified a threshold range of 0.05~0.91 for PRLDH prediction,indicating clinical net benefits.CIC indicated that within this threshold probability range,the expected recurrent cases exceeded the actual ones.Furthermore,the Nomogram model consistently outperformed individual risk factors in clinical benefits.Conclusion:The Nomogram predictive model,based on the assessment of paraspinal muscles,demonstrates excellent predictive capability and contributes to the personalized assessment of the risk of PRLDH.

关 键 词：腰椎间盘突出症 经皮椎间孔镜手术 复发 预测模型 

分 类 号：R274.9[医药卫生—中西医结合]