Introduction
Disc replacements are promised to restore the kinematics of a spine segment with degenerated disc as well as preventing degeneration at the adjacent level.  More recently, to resolve some of the deficiencies of anterior lumbar arthroplasty like the procedure itself, post-surgery pain in facets and difficulty in revision, posterior disc designs are being pursued.  They have the advantage of being easy to revise, and dealing with all the pain generators in a motion segment including the nerves, facets and disc.  This finite element based study compares the biomechanical effects of a novel posterior disc replacement across the implanted and adjacent levels of the spine versus intact.

Figure 1 - 3D view of L3-S1 implanted with posterior disc.

Materials and Methods
A three dimensional, ligamentous, experimentally validated L3-S1 finite element (FE) model described in our previous studies (Goel et al, 2005) was used to simulate total disc replacement at L4-L5 level (Figure 1).  A surgical procedure of complete removal of nucleus, posterior longitude ligament and ligamentum flavum, complete laminectomy and partial unilateral medial facetectomy (40% of right facet), was simulated to replace the SMART DISC at L4-L5 level.  The disc components had the material properties of titanium endplates coupled with cobalt chrome for the instantaneous centerline nucleus.  The top and bottom endplates of the implant were tied to respective endplates of vertebra.  Mechanical frictionless interactions were defined between articulating surfaces of the implant.
A compressive follower load of 400 N plus a bending moment of 10 Nm were applied to implanted and intact models to simulate physiological extension, flexion, left & right lateral bending and left & right axial rotation.  Motion, facet loads and intradiscal pressure (IDP) across the segments were computed for implanted and intact models.

Results
The motion at the implanted level increased significantly in flexion and lateral bending versus intact (Figure 2).  At the adjacent segments the motions were close in all loading cases.  Both intact and implanted models had similar IDP values at L3-L4 and L5-S1 except in left and right bending.  Facet loads increased significantly at L4-L5 in extension and left bending for left facet and right bending for right facet respectively.

Figure 2 - Comparison of angular displacement, facet loads and IDP at different levels between implanted and intact spine.

Conclusion
SMART DISC seems to be able to restore the kinematics of the spine at the implanted and the adjacent levels to normal, except in extension and lateral bending.  The intradiscal pressures also did not change following disc replacement as compared to the intact case and facet loads were closed in most of loading cases.  Additional biomechanical evaluations are essential for a final design.