Mechanical response of cervical vertebral endplates to axial loading.

Objective: After anterior cervical discectomy the implantation of a spacer is common practice. The majority of these spacers are trapezoid titanium cages. During the development of a height-adjustable cervical implant we needed to establish the testing limits for this device. A known phenomenon is subsidence of the cage into the vertebral endplates, which leads to a decrease in height and/or angulation of the cervical spinal segment. In contrast to the thoracic and lumbar spines, there are only limited data concerning the load-bearing ability of cervical endplates. The aim of our investigation was to obtain these data.

Methods: Bone density of 16 cervical vertebrae was estimated by quantitative computed tomography. After embedding of the vertebrae into PMMA, each endplate was slowly compressed until failure using a metal indenter resembling the form of a newly developed cervical implant. A fixed protocol with increasing loading cycles was followed. Endpoint was breakage of the endplate as established by failure to resist the increasing loading forces produced by the testing machine.

Results: The mean bone density of the 16 cervical vertebrae was 204 with a standard deviation of 52 mg Ca-HA/mL (range 130-281). The endplates failed with a mean loading of 1084 N +/- 314 (range 340-1550 N). The maximum load correlates with the bone density (R2 = 0.7347). With the 97.79 mm2 load bearing surface of the cage we calculate a mean cervical endplate break strength of 10.47 MPa and a 95 % confidence interval of 12.66-9.51 MPa. An initial settling produced by resting of the anchoring teeth in the cervical endplates was observed in 8 vertebrae at a load of 113 N (range 50-250 N).

Conclusions: In contrast to the thoracic and lumbar spines, cervical endplates show a lower resistance against axial forces. The data are important to understand postoperative cage subsidence and to establish testing limits for the development of new implant designs.