The total disc replacement (TDR) devices are gaining popularity because of their capability of allowing joint motion at the index level. Studies have shown that motion preservation can reduce the likelihood of further degeneration at the adjacent level with better surgical outcome. Current lumbar TDR devices require an anterior approach for implantation. However, it is known that its clinical outcome may depend on implant insertion and placement during surgery. Only limited number of biomechanical studies regarding the effect of placement orientation on the clinical outcome is currently available. The purpose of this study was to investigate effects of various surgical placement of a lumbar TDR on the kinematics and load-sharing characteristics using finite element method (FEM).

A previously-validated 3-D nonlinear FE model of the intact lumbar motion segment (L3-S1) based on computer tomography (CT) images of a cadaveric specimen (male, age 56, no pathologies) was used as the baseline FE model. Then, implantation of ProDisc-L (Spine Solutions, Inc., Synthes, Paoli, PA, USA) was simulated into the L4–L5 disc space through anterior approach with removal of the nucleus, anterior longitudinal ligament, and the anterior part of the annulus. The location of lumbar TDR was varied in the sagittal and the coronal planes. In the sagittal plane, the implants were placed anteriorly at 3-mm (S-3), 5-mm (S-5), and 7-mm (S-7) offset from the posterior margin of the endplate. In the coronal plane, the devices were shifted from the baseline position laterally to the right by 1-mm (C-1), 2-mm (C-2), and 3-mm (C-3) from the mid-sagittal line along the lower endplate. All of the models were subject to 150N compressive pre-load and flexion/extension moments of 10Nm at the superior endplate L4, while the inferior endplate of L5 was fully constrained. Changes in motion (ROM) and facet loads at the index and adjacent levels were assessed at different implant position.

Results showed that deviation from the central placement (from S-3 to S-7 and from C-1 to C-3) decreases ROM while increasing facet load at the index level. The effect was more pronounced in the sagittal plane than in the coronal plane:10% decrease in ROM and 1% increase in facet load in the sagittal plane vs. no significant change in the coronal plane. As expected, changes were more evident during extension than in flexion. While the kinematics of the spine was restored to the pre-operative stage at the index level (L4-5), the ROM decreased at the adjacent level (L5-S1) in a compensating manner. The overloading of the facet seemed to indicate mal-alignment of the implant can further trigger facet degeneration, which may require unwanted revision or additional surgical treatment.