骨质疏松性椎体压缩性骨折椎体强化术后椎体再骨折的相关因素

Related factors of new vertebral compression fracture after vertebral augmentation for osteoporotic vertebral compression fractures

目的分析椎体强化术(percutaneous vertebroplasty,PVP/percutaneous kyphoplasty,PKP)治疗骨质疏松性椎体压缩性骨折(osteoporotic vertebral compression fractures,OVCFs)患者术后继发椎体压缩性骨折的影响因素,同时建立其风险预测模型以评估其术后继发椎体骨折的风险。方法选取2015年8月至2017年12月因OVCFs于宁夏回族自治区人民医院行PVP/PKP患者385例,根据术后2年间是否出现新鲜椎体骨折,分为再骨折组(recompression fracture,RF)和无再骨折组(non-recompression fracture,NRF),收集所有患者的临床资料,包括性别、年龄、骨折椎体节段、手术方式、骨水泥注入量、骨水泥渗漏、弥散情况以及是否规律抗骨质疏松治疗等,分别进行单因素分析和多因素Logistic回归分析,同时建立风险预测模型并采用Bootstrap法对数据进行内部验证。结果385例患者中41例出现术后继发椎体压缩性骨折,再骨折率10.6%。未规律抗骨质疏松治疗(OR=11.802,95%CI:4.331~32.166,P<0.001)、骨水泥弥散情况差(OR=2.797,95%CI:1.121~6.978,P=0.027)、骨水泥与终板未接触(OR=9.198,95%CI:3.320~25.484,P<0.001)是术后继发椎体骨折的独立危险因素。骨水泥未发生渗漏(OR=0.308,95%CI:0.127~0.748,P=0.009)、骨水泥注入量为4.0~4.5 mL/椎体(OR=0.08,95%CI:0.008~0.762,P=0.028)是术后继发椎体骨折的保护因素。模型受试者工作特征曲线(receiver operating characteristic,ROC)曲线下面积(area under the curve,AUC)为0.881,C-index为0.890(95%CI:0.832~0.948),矫正拟合偏倚后的C-index为0.859,校准曲线表现为模型预测值分布曲线与重复抽样矫正拟合偏倚后的分布曲线良好贴合;模型决策曲线远离两条极端的直线。结论未规律抗骨质疏松治疗、骨水泥渗漏、骨水泥弥散情况差和骨水泥与终板未接触是OVCFs椎体强化术后继发椎体压缩性骨折的独立危险因素;以此所构建的风险预测模型具有良好的区分度、校准度和临床有效性,可为OVCFs椎体强化术后继发椎体压缩性骨折的风险预测提供一定的参考。

Objective To analyze the risk factors of adjacent vertebral fractures after percutaneous vertebroplasty(PVP)or percutaneous kyphoplasty(PKP)for osteoporotic vertebral compression fractures(OVCFs),and establish a risk prediction model to evaluate the risk of postoperative new vertebral fracture.Methods According to the occurrence of fresh vertebral fracture in 2 years,385 patients were divided into recompression fracture and non-recompression fracture group.The clinical data of all patients were collected,including gender,age,vertebral segment of fracture,mode of operation,amount of bone cement injected,cement leakage,dispersion and regular anti-osteoporotic treatment.Risk prediction model was established and Bootstrap method was used to verify the data internally.Results Among the 385 patients,41 cases had new vertebral compression fracture after operation,and the re-fracture rate was 10.6%.Irregular anti-osteoporotic therapy,poor diffusion of bone cement,poor bone cement contact with endplate were independent risk factors.No leakage of bone cement and injection of 4.0-4.5 mL of cement were protective factors for new vertebral fracture after operation.The area of the model was 0.881(95%CI:0.832-0.948),and the C-index after correction of fitting bias was 0.859.The calibration curve showed that the distribution curve of the predicted value of the model fit well with the distribution curve of repeated sampling after correction of fitting bias.The decision curve of the model was far from two extreme straight lines.Conclusions Non-regular anti-osteoporotic treatment,bone cement leakage,poor bone cement dispersion,and non-contact of bone cement with the endplate are independent risk factors for new vertebral compression fractures after vertebral augmentation for OVCFs.The risk prediction model has a medium discrimination,good calibration,and clinical validity,which can provide a certain reference for the risk prediction of new vertebral compression fractures after vertebral augmentation for OVCFs.