Presented at NACOB 98: North American Congress on Biomechanics Canadian Society for Biomechanics - American Society of Biomechanics University of Waterloo Waterloo, Ontario, Canada August 14-18, 1998

---

BIOMECHANICAL STUDY ON STABILITY OF THE CERVICAL GRAFT FUSION

R.N.Natarajan, B.H.Chen, H.S. An, T.H.Lim and G.B.J. Andersson
Department of Orthopedic Surgery, Rush-Presbyterian-St.Luke's Medical Center
Chicago, Illinois

INTRODUCTION

The Smith-Robinson tricortical iliac crest graft that is placed between the endplates is most commonly used for interbody fusion. Biomechanical studies on this construct are lacking, particularly with respect to flexibility changes with multiplaner loadings and stress generated in the graft itself. Knowing these biomechanical properties of the surgical construct is important for selection of grafts with appropriate strength and for guiding patients in terms of postoperative bracing and rehabilitation. Using finite element modeling, the present study attempts to quantify percentage changes in flexibility in C5-C6 motion segment, contact surface changes of the graft, and stresses in the graft under moment loadings in varying directions when (a) the load is transferred to graft through a sliding contact surface between the endplate and the graft surfaces simulating a loose interbody graft, and (b) the graft is rigidly connected to the vertebral surfaces, simulating a well fitting tight graft. Clinically, interbody graft fixation will be somewhere between the two constructs.

MATERIALS & METHOD

A three dimensional non-linear finite element model of an intact C5-C6 cervical spine segment was developed using CT scans of 38-year-old female normal subject. The model included fiber reinforced annular fibers represented by cable elements whose cross-sectional area decreased along radial direction from outer portion of annulus. Facet joints with cartilage layers attached were assumed inclined at an angle of 45 degrees to transverse plane and modeled by three-dimensional moving contact surface elements. Material properties for the current model were taken from the literature. A graft of mean cross-sectional area equal to 65% of end plate area was placed with its anterior surface at a distance of 2.25 mm from the anterior most location of the vertebra. The elastic property of the graft was taken as 2500 MPa to simulate iliac crest bone. The biomechanical function of the intact model under external load was compared with in-vitro model results available in literature (Moroney et al., 1988). In the first anterior fusion model the graft was attached to the end plates through 3D contact surface elements (with coefficient of friction =0.15) representing a loose interbody graft. The second anterior fusion model represented the case of a well fitting tight graft between the superior and inferior vertebra. The predicted responses due to various moment loads (1.8Nm) along with a pre-load as a result of the two model variations were studied to provide all the three displacements and rotations of C5 vertebra with respect to C6 vertebra.

RESULTS

The primary and coupled motions of the intact model under different moment loadings were in excellent agreement with the available experimental data

The model in which the contact between the graft and end plate was through sliding contact, predicted decreased motion in flexion (45%) and extension moments (50%) while increased motion under torsion (466%) and lateral bending moments (22%) (Table 1) as compared to intact model. A reduction in motion was observed under all moment loads when the graft was assumed fixed with the vertebrae. Maximum reduction in motion of 66% was calculated under flexion moment while minimum reduction in motion (33%) was observed under lateral bending moment load. It was observed that under all loading conditions the entire graft surfaces were not in contact with the end plates. Maximum percentage of contact area between the graft and end plate was observed under torsion moment load while the least percentage of contact area was calculated under lateral bending moment load. Maximum compressive stress (Table 2) in the graft was observed under flexion and extension moment loads in both models. Torsion and lateral bending moment loads produced higher compressive stress in the graft when the graft was connected to the end plates using sliding contact surfaces rather than rigid connection. The maximum compressive stress in the graft reached 21 MPa under extension moment loading which may cause failure of the graft (An et al., 1994).

CONCLUSIONS

The finite element model study showed that graft effectively reduced the motion under all moment loadings only when rigid connection construct is achieved. Contact conditions between the graft and the vertebrae were found to affect motion particularly under torsion and lateral bending moment conditions. In the immediate post-operative period, torsion may be most harmful followed by lateral bending, and flexion/extension moment if the graft is not rigidly fixed to the end plates. In developing interbody graft devices or cages, one should consider the torsional stability of the construct as an important element. Flexion/extension moment loadings increased the compressive stress in the graft and therefore graft fracture or collapse may be most susceptible in flexion/extension.

REFERENCES

(1) Moroney et al., (1988),J.Bio.Mech,21:769-779

(2) An et al., (1994),Spine,19:2358-2363