Devoted to diagnostic and interventional spine imaging and therapeutics


Vertebral End Plate Fractures: an Indicator of the Abnormal Forces Generated in the Spine Following Vertebroplasty 2005

Interventional Spine

Andrew T Trout, BA, Non ASSR Member
David F Kallmes , MD, ASSR Member
Kent R Thielen , MD, ASSR Member
Kennith F Layton , MD, Non ASSR Member


Biomechanical simulations indicate that, following vertebroplasty, increased stress is localized to the endplates of adjacent vertebral bodies (1, 2). Among spontaneous osteoporotic fractures, isolated superior endplate fractures are substantially more common than isolated inferior endplate fractures. We hypothesized that, if new-onset fractures of vertebral bodies immediately above a cemented vertebrae localized disproportionately to the inferior endplate, one might implicate the cement as causative of this unusual fracture pattern.

Methods & Materials

We performed a retrospective review of 86 patients with documented new vertebral fractures following vertebroplasty for osteoporosis induced fractures. Fractures were diagnosed on the basis of MRI or bone scan. We define “prevalent” fractures as those present prior to vertebroplasty, and “incident” fractures as new-onset fractures following vertebroplasty. We defined fracture location as “endplate” (superior or inferior) if the marrow along the opposite endplate was normal on MRI, if a distinct fracture line could be identified on MRI, or if plain films showed a clear endplate deformity. Fractures not meeting these criteria were considered holo-vertebral.

Chi squared analysis was used to compare the observed proportion of inferior endplate fractures in both prevalent and incident fractures. Our null hypothesis was that incident fractures above a treated level would demonstrate the same proportion of inferior endplate fractures as other groups, including 1) prevalent fractures, 2) incident fractures immediately below a treated level, and 3) incident fractures not adjacent to treated levels.


The 86 patients selected for analysis presented with 313 vertebral fractures and were treated at a total of 137 vertebral levels. Following vertebroplasty, 186 new fractures developed (median number of new fractures=2, range 1-10). MRIs were available for 274 of the 313 prevalent fractures and 128 of the 186 incident fractures. In prevalent (baseline) fractures, superior endplate fractures were significantly more common than either inferior endplate or holo-vertebral fractures (57% superior endplate, 11% inferior endplate, and 32% holo-vertebral, p<.0001) (Figure 1). Of the 186 incident (new-onset) fractures, 77 (41%) were adjacent to treated vertebrae. Non-adjacent, incident fractures showed a fracture distribution that was different from prevalent fractures only in terms of the proportion of holo-vertebral fractures; superior endplate fractures still predominated over inferior endplate fractures (69%, 25%, 6% for superior endplate, inferior endplate, and holo-vertebral fractures, respectively, p<.0001). Incident fractures immediately above treated levels showed a disproportionate number of inferior endplate fractures compared to all other groups (30% superior, 57% inferior, and 13% holo-vertebral fractures, respectively, p<.0001) (Figure 1). Incident fractures immediately below a treated level showed a disproportionate number of superior end plate fractures when compared to prevalent fractures (84% superior, 12% inferior, and 4% holo-vertebral fractures, p=0.002).


The typical location for vertebral compression fractures in osteoporotic patients is along the superior end plate. As predicted by biomechanical simulations, vertebroplasty disrupts this typical fracture distribution through its effect on adjacent vertebral bodies. New-onset fractures that occur immediately above a cemented vertebral body demonstrate a striking localization to the inferior endplate (Figure 2). This pattern was not seen in prevalent fractures, in non-adjacent incident fractures, or in incident fractures immediately below treated levels. In addition, new-onset fractures immediately below a treated level demonstrate a predominance of superior end plate fractures when compared to prevalent fractures. These data add further, circumstantial, evidence that vertebroplasty may play a causative role in new fractures.


1. Baroud G, Heini P, Nemes J, et al., Biomechanical explanation of adjacent fractures following vertebroplasty. Radiology 2003;229(2):606-7; author reply 607-8
2. Baroud G, Nemes J, Heini P, et al., Load shift of the intervertebral disc after a vertebroplasty: a finite-element study. Eur Spine J 2003;12(4):421-6