Cement leakage in percutaneous vertebroplasty: effect of preinjection gelfoam embolization

Prospective case series.

To determine the safety and feasibility of routine preinjection of gelfoam embolization during percutaneous vertebroplasty.

Percutaneous vertebroplasty has been used effectively in pain relief for vertebral fractures resulting from malignancy and osteoporosis. However, cement extrusion is a common problem and can lead to complications. Gelfoam embolization of venous channels before cement injection has not been widely used as a technique to prevent leakage.

Thirty-one patients who met the inclusion-exclusion criteria for the study underwent percutaneous vertebroplasty. Venography was first performed to determine the flow pattern in the vertebrae and confirm needle placement. Next, routine gelfoam embolization of venous channels was performed. This was followed by low-pressure, minimal-volume cement injection. The outcome measure of cement leakage was assessed after surgery using radiographs and CT scans.

There were no complications. In the 31 patients, 61 levels of vertebroplasty were performed. Overall, there were 16 leaks out of 61 levels in 12 patients (26.2%). In osteoporotic fractures, there were 11 leaks in 49 levels, giving a leakage rate of 22.5%. There was only 1 epidural leak in this group (2%), and this was asymptomatic. Seven leakages were into the adjacent disc, 2 into the body, and 1 into the paravertebral tissues. In malignant fractures, there were 5 leakages out of 12 levels (41.7%). Of these, 2 were epidural leaks (16.7%), which were asymptomatic.

Complications resulting from leakage are the most feared side effect of the procedure. This has resulted in only limited application of vertebroplasty in the United Kingdom. Routine gelfoam embolization together with careful technique has been shown to be a safe and feasible method during vertebroplasty.