腰椎间盘突出症患者行经皮椎间孔镜髓核摘除术术后复发的列线图模型构建

Construction of a nomogram model for predicting recurrence after percutaneous transforaminal endoscopic discectomy in patients with lumbar disc herniation

目的分析腰椎间盘突出症(LDH)患者行经皮椎间孔镜髓核摘除术(PTED)术后复发的影响因素,构建列线图模型并验证其预测价值。方法2022年1—12月南阳市中心医院诊治LDH患者268例,均行PTED,术后随访1年,根据复发情况分为复发组40例和未复发组228例,比较2组体质量指数、病程、Pfirrmann分级及纤维破坏、髓核摘除情况等临床资料,采用多因素logistic回归分析LDH患者行PTED术后复发的影响因素;根据影响因素构建预测LDH患者行PTED术后复发的列线图模型;绘制ROC曲线,评估列线图模型预测LDH患者行PTED术后复发的效能;采用Bootstrap法进行内部验证,采用校准曲线评估列线图模型的校准度,采用决策曲线评估列线图模型的净收益。结果复发组年龄[(62.55±8.38)岁]大于未复发组[(53.99±5.01)岁](t=6.266,P<0.001),体质量指数[(25.65±2.13)kg/m^(2)]及病程≥6个月、Pfirrmann分级Ⅲ~Ⅳ级、有纤维破坏、髓核摘除不彻底比率(70.00%、60.00%、52.50%、50.00%)均高于未复发组[(23.44±1.25)kg/m^(2)、41.67%、32.02%、25.44%、24.12%)(t=6.364,χ^(2)=11.001、11.538、11.988、11.307,P均<0.05)。年龄(OR=1.275,95%CI:1.158~1.402,P<0.001)、体质量指数(OR=2.209,95%CI:1.540~3.170,P<0.001)、病程(OR=4.954,95%CI:1.624~15.110,P=0.005)、Pfirrmann分级(OR=4.064,95%CI:1.338~12.339,P=0.013)、纤维破坏(OR=3.851,95%CI:1.346~11.016,P=0.012)、髓核摘除(OR=3.090,95%CI:1.126~8.478,P=0.028)是LDH患者行PTED术后复发的影响因素。列线图模型预测LDH患者行PTED术后复发的AUC为0.939(95%CI:0.903~0.965,P<0.001),灵敏度为90.00%,特异度为82.46%;列线图模型的C指数为0.839(95%CI:0.796~0.882,P<0.001),校准曲线与理想曲线基本接近,具有较好一致性;决策曲线远离横纵坐标,预测LDH患者行PTED术后复发风险阈值>0.16,临床净收益较好。结论年龄大、体质量指数高、病程≥6个月、Pfirrmann分级Ⅲ~Ⅳ级、有纤维破坏及髓核摘除不彻底的LDH患者行PTED术后复发的风险较大,构建的列线图模型预测LDH患者行PTED术后复发有较高价值。

Objective To analyze the factors influencing recurrence after percutaneous transforaminal endoscopic discectomy(PTED)in patients with lumbar disc herniation(LDH),to construct a nomogram model and to verify the predictive value of model.Methods Totally 268 LDH patients underwent PTED in Nanyang Central Hospital from January to December 2022,and were divided into recurrence group(n=40)and non-recurrence group(n=228)according to one-year follow-up results.The clinical data as body mass index,duration of LDH,Pfirrmann grade,fibrous destruction and nucleus removal were compared between two groups.Multivariate logistic regression analysis was used to evaluate the influencing factors of recurrence of LDH after PTED.A nomogram model was constructed based on the influencing factors to predict recurrence of LDH after PTED,and its predictive efficiency was assessed by ROC curve.Bootstrap method was used for internal verification.The calibration and net benefit of the nomogram model were assessed by calibration and decision curves.Results The age was older in recurrence group[(62.55±8.38)years]than that in non-recurrence group[(53.99±5.01)years](t=6.266,P<0.001).The body mass index,and proportions of duration of LDH≥6 months,Pfirrmann gradeⅢ-Ⅳ,presence of fibrous destruction and nucleus incomplete removal were higher in recurrence group[(25.65±2.13)kg/m^(2),70.00%,60.00%,52.50%,50.00%]than those in non-recurrence group[(23.44±1.25)kg/m^(2),41.67%,32.02%,25.44%,24.12%](t=6.364,χ^(2)=11.001,χ^(2)=11.538,χ^(2)=11.988,χ^(2)=11.307,all P values<0.05).Age(OR=1.275,95%CI:1.158-1.402,P<0.001),body mass index(OR=2.209,95%CI:1.540-3,170,P<0.001),duration of LDH(OR=4.954,95%CI:1.624-15,110,P=0.005),Pfirrmann grade(OR=4.064,95%CI:1.338-12.339,P=0.013),fibrous destruction(OR=3.851,95%CI:1.346-11.016,P=0.012),and nucleus removal(OR=3.090,95%CI:1.126-8.478,P=0.028)were the influencing factors of LDH recurrence after PTED.The AUC of nomogram model for predicting LDH recurrence after PTED was0.939(95%CI:0.903-0.965,P<0.001),with a sensitivity of 90.00%and a specificity of 82.46%.The C-index of nomogram model was 0.839(95%CI:0.796-0.882,P<0.001),and the calibration curve was basically close to the ideal curve,with a good consistency.The decision curve was far away from the horizontal and vertical coordinates,predicting a recurrence risk threshold of>0.16 after PTED in LDH patients,with a good net clinical benefit.Conclusions The old age,large body mass index,duration of LDH≥6 months,Pfirrmann gradeⅢ-Ⅳ,fibrous destruction,and nucleus incomplete removal indicate a high risk of recurrence after PTED in LDH patients.To construct a nomogram model achieves a high predictive value for LDH recurrence after PTED.