骨质疏松性胸腰椎骨折PVP/PKP术后二次骨折的危险因素分析及预测模型建立

Risk factors and prediction model of refracture after percutaneous vertebroplasty and percutaneous kyphoplasty for osteoporotic thoracolumbar fracture

中文关键词:  胸腰椎骨折  骨质疏松症  经皮椎体成形术  经皮椎体后凸成形术  椎体二次骨折

英文关键词:Thoracolumbar fracture  Osteoporosis  Percutaneous vertebroplasty  Percutaneous kyphoplasty  Vertebral refracture

中文摘要:

  【摘要】 目的：探讨分析骨质疏松性胸腰椎骨折经皮椎体成形术（percutaneous vertebroplasty，PVP）或经皮椎体后凸成形术（percutaneous kyphoplasty，PKP）术后发生椎体二次骨折的危险因素，建立并验证术后二次骨折风险的预测模型。方法：回顾性分析2015年1月~2019年12月在我院确诊为骨质疏松性胸腰椎骨折行PVP或PKP术的患者130例，其中男33例，女97例，年龄70.20±8.46岁（55~92岁），随访时间14.75±2.17个月（12~22个月）。根据术后有无椎体二次骨折分为二次骨折组（n=26）和无二次骨折组（n=104），比较两组的年龄、性别、体质指数（body mass index，BMI）、骨密度（bone mineral density，BMD）、既往骨折史、手术椎体数、PVP/PKP、术前术后椎体前缘高度（anterior vertebral height，AVH）差、术前术后局部后凸（segmental kyphosis，SK）差、单个椎体骨水泥注射量和有无骨水泥渗漏等因素。将单因素分析有意义的指标纳入多因素Logistic回归分析，根据结果建立骨质疏松性胸腰椎骨折PVP/PKP术后二次骨折风险的预测模型，用受试者工作特征（receiver operating characteristic，ROC）曲线、决策曲线及校正曲线进行内部验证。结果：单因素分析及多因素Logistic回归分析发现高龄[odds ratio（OR）=44.33，95% confidence interval（CI）=2.98~659.51，P=0.0059]、低BMD（OR=10.70，95%CI=2.50~45.75，P=0.0014）、既往骨折史（OR=14.76，95%=2.40~90.57，P=0.0036）和多手术节段（OR=6.36，95%CI=1.51~26.72，P=0.0115）是骨质疏松性胸腰椎骨折PVP/PKP术后椎体二次骨折的独立危险因素（P＜0.05）。建立基于Logistic回归分析的骨质疏松性胸腰椎骨折PVP/PKP术后二次骨折风险预测模型，并绘制了基于上述预测模型的列线图，经ROC、决策曲线及校正曲线验证发现模型拟合效果较好。结论：高龄、低骨密度、多手术节段、既往骨折史是骨质疏松性胸腰椎压缩骨折PVP/PKP术后二次骨折的独立危险因素。骨质疏松性胸腰椎骨折患者PVP/PKP术后二次骨折风险预测模型拟合效果良好，为术后评估二次骨折的风险提供了参考。

英文摘要:

  【Abstract】 Objectives: To explore and analyze the risk factors of vertebral refractures after percutaneous vertebroplasty(PVP) or percutaneous kyphoplasty(PKP) for osteoporotic thoracolumbar fractures and to establish a prediction model. Methods: The data of 130 patients diagnosed with osteoporotic thoracolumbar fractures who underwent PVP or PKP surgery in our hospital between January 2015 and December 2019 were retrospectively analyzed. There were 33 males and 97 females, averaged 70.20±8.46(55-92) years old and were followed up for 14.75±2.17(12-22) months. According to whether occured vertebral refracture after surgery, the patients were divided into the refracture group(n=26) and non-refracture group(n=104). Parameters such as age, gender, body mass index(BMI), bone mineral density(BMD), history of fracture, number of operative vertebrae, PVP/PKP, the difference between preoperative and postoperative anterior vertebral height(AVH), the difference between preoperative and postoperative segmental kyphosis(SK), the average volume of bone cement injection, and with/without cement leakage were compared between the two groups. And the statistically significant factors in univariate factor analysis were included in multivariate logistic regression analysis, based on which, a prediction model of the risk of vertebral refractures after PVP/PKP for osteoporotic thoracolumbar fractures was established, and it was verified according to the receiver operating characteristic curve(ROC), decision-making curve and calibration curve. Results: Univariate analysis and multivariate logistic analysis showed that old age[odds ratio(OR)=44.33, 95% confidence interval(CI)=2.98-659.51, P=0.0059], lower BMD(OR=10.70, 95%CI=2.50-45.75, P=0.0014), history of fractures(OR=14.76, 95%CI=2.40-90.57, P=0.0036) and multiple surgical segments(OR=6.36, 95%CI=1.51-26.72, P=0.0115) were the independent risk factors for vertebral refracture after PVP/PKP for osteoporotic thoracolumbar fracture patients(P＜0.05). A risk prediction model of vertebral refracture after PVP/PKP for osteoporotic thoracolumbar fracture patients was established based on logistic regression, and thus a nomograph was drawn. The prediction model was verified with good fitting effect according to ROC, decision-making curve and calibration curve. Conclusions: Old age, lower BMD, multiple surgical segments, history of fractures are independent risk factors of refracture after PVP/PKP for osteoporotic throacolumbar fracture. The prediction model of the risk of vertebral refracture in patients with osteoporotic thoracolumbar fracture after PVP/PKP is well fitted, which provides a reference for evaluating the risk of secondary fracture after surgery.