Estimating risk of pulmonary neoplastic embolism during vertebroplasty

Vertebroplasty was simulated on a pig model.

To evaluate the risk of neoplastic tissue migration into lungs during vertebroplasty.

The application of vertebroplasty in spinal metastasis is not well documented. The risk of neoplastic tissue migration into the lungs during vertebroplasty remains unknown.

A cancer model was built in 11 Landrace pigs (50 kg) by injecting 99mTc-labeled albumin macroaggregates into the center of L5 and L6 prior to vertebroplasty. Continuous scintigraphic imaging was performed with 1-minute frames over the lungs and vertebrae before and after injection to ensure steady state and baseline. We surveyed free TcO4- in thyroid. Twenty minutes after the 99mTc injection, 2-level vertebroplasty was performed at L5 and L6 with 3 Jamshidi needles in each vertebra. Into each vertebra, on average, 2.8 ± 1.1 mL of poly(methyl methacrylate) cement (Depuy CMW, Blackpool, UK) was injected. Quantitative scintigrams were obtained within 90 minutes after vertebroplasty. X-rays and quantitative computed tomography scans quantified cement distribution. Means of 99mTc activity before and after vertebroplasty were compared in a paired t test.

In this cancer model, we found an 80% risk of tissue migration to the lungs when performing vertebroplasty. In average, the study showed a significant amount of macroaggregate migration of 1.87% total range from 0% to 8% (CI: 0.05%-0.37%) with P = 0.045. There was no free TcO4- in the thyroid. Despite the standardized procedure, we found a large interindividual variation of pulmonary embolism.

It is demonstrated that there exists a significant risk of exporting neoplastic disease or fatty tissue to the lungs when performing vertebroplasty. A similar adverse effect can be expected with balloon kyphoplasty. In patients with metastatic disease, vertebroplasty should be limited to those with short life expectancy.