A pictorial classification atlas of cement extravasation with vertebral augmentation.

BACKGROUND CONTEXT: Minimally invasive procedures for the treatment of vertebral compression fractures (VCFs) have been in use since the mid-1980s. A mixture of liquid monomer and powder is introduced through a needle into one or both pedicles, and it polymerizes within the vertebral body in an exothermic chemical reaction. The interaction between cement and the fractured vertebral body determines whether and how the cement stabilizes the fragments, alters morphology, and extravasates. The cement is intended to remain within the vertebral body. However, some studies have reported cement leakage in more than 80% of the procedures. Although cement leakage can have no or minimal clinical consequences, adverse events, such as paraplegia, spinal cord and nerve root compression, cement pulmonary embolisms, or death, can occur. The details of how the cement infiltrates a vertebral body or extravasates out of the body are poorly understood and may help to identify strategies to reduce complications and improve clinical efficacy.

PURPOSE: Apply novel techniques to demonstrate the cement spread inside vertebrae as well as the points and pattern of cement extravastation.

STUDY DESIGN: Ex vivo assessment of vertebral augmentation procedures.

METHODS: Vertebrae from six fresh whole human cadaver spines were used to create 24 specimens of three vertebrae each. The specimens were placed in a pneumatic testing system, designed to create controlled anterior wedge compression fractures. Unipedicular augmentation was performed on the central vertebra of 24 specimens using polymethylmethacrylate/barium sulfate Vertebroplastic cements (DePuy Spine, Raynham, MA, USA). The volume of cement injected into each vertebra was recorded. Fine-cut computed tomography (CT) scans of all segments were obtained (Brilliance 64; Philips Medical Imaging, Amsterdam, The Netherlands). Using multiplanar reconstructions and volume compositing three-dimensional imaging (Osirix, www.osirix-viewer.com), each specimen was carefully assessed for cement extravasation. Specimens were then immersed in a 50% sodium hypochlorite solution until all overlying soft tissues were removed, leaving the bone and cement intact. The specimens were dried and visually examined and photographed to assess cement extravasation and fracture patterns. Specimens were cut in the axial or sagittal plains to assess the gross morphology of cement infiltration and extravasation. Finally, 25-mm block sections were removed from selected specimens and imaged at 14-μm resolution using a GE Locus-SP micro-CT system (GE Healthcare, London, Ontario, Canada).

RESULTS: Infiltration was characterized by an intimate capture of trabecular bone within the cement, forming an irregular border at the perimeter of the cement that is determined by the morphology of the trabeculae and marrow spaces. Extravasation of the cement was assessed as "any" if any small or large amount of extravastation was detected and was also assessed as severe if a large amount of extravasation was found. Out of the 23 levels studied, some extravasation was visibly apparent in all levels. A wide spectrum of filling patterns, leakage points, and interdigitation of the cement was observed and appeared to be determined by the interaction of the cement with the trabecular morphology. The results support the fact that the cement generally advances through the vertebrae with relatively regular and easily identifiable borders.

CONCLUSIONS: Using a cadaver VCF model, this study demonstrated the exact filling and extravastation patterns of bone cement inside and out of fractured vertebrae. These data enhance our understanding of the vertebral augmentation and extravastation mechanics.