Related Experiment Video
Updated: May 23, 2026
Nine-Grid Area Division Method: A New Ideal Bone Puncture Region for Percutaneous Vertebroplasty in Lumbar Spine
Published on: August 9, 2024
[Cement leakages in percutaneous vertebroplasty:analysis of postoperative computed tomography]
Bin Tie1, Shi-cheng He, Gao-jun Teng
1Department of Radiology, Affiliated Zhongda Hospital, Southeast University, Nanjing 210009, China.
Objective:
To evaluate the influencing factors of cement leakage in vertebroplasty for the treatment of osteoporosis vertebral compression fracture (OVCF) and vertebral metastases (VM).
Methods:
Retrospective analysis was conducted for 653 vertebrae in 356 patients undergoing vertebroplasty at our hospital from May 2007 to January 2011. 251 cases had 438 vertebrae with painful OVCF while 105 cases had 215 vertebrae with VM. Pre-operative computed tomography (CT) was performed to determine the presence of cortical defects or osteolysis and within 3 days after PVP to observe the distribution of polymethylmethacrylate (PMMA) in vertebrae and whether leakage occurred. Volume of PMMA injected into each vertebral body and types of cement leakage were compared between the OVCF and VM groups by Z test or χ². The correlation between cortical defects and cement leakages around vertebrae was assessed with Pearson correlation coefficient.
Results:
The successful rate of PVP was 100%. The mean volume of PMMA injected into each vertebra was (5.0 ± 2.0) ml and (4.0 ± 1.7) ml in the OVCF and VM groups respectively (P < 0.05). Asymptomatic PMMA leakage was demonstrated by CT in 93 vertebrae (21.2%) in the OVCF group and in 53 vertebrae (28.8%) in the VM group respectively (P < 0.05). Cement leakages into disk were found in 58 vertebrae in the OVCF group and 16 vertebrae in the VM group respectively (P = 0.025). Cement leakages into paravertebral vein were found in 12 vertebrae in the OVCF group and 26 vertebrae in the VM group respectively (P < 0.0001). Correlation was found between cortical defects and cement leakage into paravertebral soft tissues in the OVCF group (r = 0.14) or in the VM group (r = 0.27), between end-plate defects and cement leakage into disk in the OVCF group (r = 0.29) or in the VM group (r = 0.31).
Conclusion:
As a common occurrence in vertebroplasty, cement extravasation is well-tolerated in most patients. It occurs more frequently in the patients with VM than those with OVCF, especially in cases of leakage into paravertebral vein. Cement leakage into disc or paravertebral soft tissue is predisposed in vertebrae with end plate, cortical defects or osteolysis.
