WHY DO VERTEBRAL FRACTURES USUALLY AFFECT THE CRANIAL ENDPLATE?

Abstract

Introduction: Endplate fractures are clinically important. They are very common, are associated with an increased risk of back pain, and can probably lead on to intervertebral disc degeneration. However, such fractures tend to damage the cranial endplate much more often than the caudal. In this study, we test the hypothesis that the vulnerability of cranial endplates arises from an underlying structural asymmetry in cortical and cancellous bone.

Methods: Sixty-two “motion segments” (two vertebrae and the intervening disc and ligaments) were obtained post-mortem from human spines aged 48–92 yrs. All levels were represented, from T8–9 to L4–L5. Specimens were compressed to failure while positioned in 2–6o of flexion, and the resulting damage characterised from radiographs and at dissection. 2mm-thick slices of 94 vertebral bodies (at least one from each motion segment) were cut in the mid-sagittal plane, and in a para-sagittal plane through the pedicles. Microradiographs of the slices were subjected to image analysis to determine the thickness of each endplate at 10 locations, and to measure the optical density of the endplates and adjacent trabecular bone. Comparisons between measurements obtained in cranial and caudal regions, and in mid-sagittal and pedicle slices, were made using repeated measures ANOVA, with age, level and gender as between-subject factors. Linear regression was used to determine significant predictors of compressive strength (yield stress).

Results: Fracture affected the cranial endplate in 55 specimens and caudal endplate in 2 specimens. Endplate thickness was low centrally and higher towards the periphery. Cranial endplates were thinner than caudal, by 14% and 11% in mid-sagittal and pedicle slices respectively (p=0.003). Differences were greater in central and posterior regions. Cranial endplates were supported by trabecular bone with 6% less optical density (p=0.004) with this difference also being greatest posteriorly. Caudal but not cranial endplates were thicker at lower spinal levels (p=0.01). Vertebral yield stress (mean 2.21 MPa, SD 0.78 MPa) was best predicted by the density of trabecular bone underlying the cranial endplate in the mid-sagittal slices of the fractured vertebral bodies (r2 = 0.67, p=0.0006).

Conclusions: When vertebrae are compressed by adjacent discs, cranial endplates usually fail before caudal endplates because they are thinner and supported by less dense trabecular bone. These asymmetries in vertebral structure may be explained by the location of back muscle attachments to vertebrae, and by the nutritional requirements of adjacent intervertebral discs.