Study Design. A retrospective review and analysis of consecutive patients who underwent single-level vertebroplasty at our institute between March 2002 and March 2006.
Objective. To analyze the risk factors for subsequent fractures after vertebroplasty and to predict the postoperative fracture-free time and rate. The effect of bone cement volume injected was also evaluated.
Summary of Background Data. Previous studies of subsequent fractures after vertebroplasty showed conflicting conclusions about risk factors. The frequency of refracture also varied, ranging from 12% to 52%. Most new fractures occurred at adjacent levels, with different risk factors identified. No data were available on the effect of injected bone cement volume, and no consensus had been reached as to the optimal cement volume.
Methods. All enrolled patients were treated with single-level vertebroplasty and followed a standardized postoperative care protocol. Data from medical records and radiographs were collected and analyzed. Variables included patient constitutional factors, radiographic parameters, and volume of injected bone cement.
Results. A total of 166 patients (76 men, 90 women) with a mean age of 73.4 years were enrolled in this study. The mean follow-up time was 15.3 months. The overall refracture rate was 38%, with a mean fracture-free interval of 32 months. Both a greater volume of bone cement injected and a greater degree of vertebral height restored contributed significantly to the risk of subsequent adjacent fracture. No risk factor for subsequent remote fracture was identified. A greater volume of bone cement injected was positively correlated with deformity correction after vertebroplasty.
Conclusion. Most subsequent fractures occurred at the adjacent level within the first 3 months. Patient preoperative condition did not help predict refracture. Although a greater volume of bone cement injected when performing vertebroplasty contributed to the risk of subsequent adjacent fracture, it resulted in a greater improvement of kyphosis.