| Objectives To investigate the risk factors of bone cement leakage following percutaneous vertebroplasty (PVP) for osteoporotic vertebral compression fracture (OVCF). Methods The clinical and radiological characteristics of 160 patients (167 vertebrae) who underwent PVP in our hospital from October 2020 to June 2022 were retrospectively reviewed. There were 39 males and 121 females, aged (72.8±7.9) years old (58-93 years old). According to the occurrence of bone cement leakage, the patients were divided into leakage group (42 cases, 42 vertebrae) and non-leakage group (118 cases, 125 vertebrae). Univariate analysis was used to analyze the differences between the two groups in gender, age, body mass index (BMI), bone mineral density, fracture level, cortical disruption, intravertebral cleft, vertebrobasilar venous foramen, volume of the injured vertebra, compression rate of the vertebral body, bone cement volume, ratio of bone cement injection volume to volume of the injured vertebra, and ratio of bone cement volume to fractured vertebra loss volume. Factors related to bone cement leakage were included in multivariate Logistic regression analysis to identify the independent risk factors of bone cement leakage after PVP. Receiver operating characteristic (ROC) curve was drawn, and according to the area under curve (AUC) the value of prediction tests was evaluated. Results Univariate analysis showed that there were significant differences in cortical disruption, intravertebral cleft, vertebrobasilar venous foramen, bone cement volume and the ratio of bone cement volume to fractured vertebra loss volume between groups (P＜0.05). Multivariate Logistic regression analysis showed that intravertebral cleft ［OR=5.215, 95% CI (2.006, 13.159), P＜0.001］, vertebrobasilar venous foramen ［OR=3.357, 95% CI (1.205, 9.356), P=0.021］, bone cement volume ［OR=2.519, 95% CI (1.148, 4.477), P=0.002］ and the ratio of bone cement volume to fractured vertebra loss volume ［OR=12.305, 95% CI (1.875, 80.756), P=0.009］ were independent predictors of postoperative bone cement leakage after PVP. The ROC curve showed that the AUC of ratio of bone cement volume to fractured vertebra loss volume in the prediction of bone cement leakage was 0.641, and P＜0.01, which had certain prediction value. The best cut-off point for the prediction of bone cement leakage with ratio of bone cement volume to fractured vertebra loss volume was 61.82%, sensitivity was 69.00% and specificity was 38.40%. Conclusion Intravertebral cleft, vertebrobasilar venous foramen, bone cement volume and the ratio of bone cement volume to fractured vertebra loss volume are independent risk factors for postoperative bone cement leakage. The ratio of bone cement volume to fractured vertebra loss volume should be kept below 61.82% to reduce the risk of bone cement leakage. |
