Biomechanical comparison of multilevel lateral interbody fusion with and without supplementary instrumentation: a three-dimensional finite element study

BACKGROUND: Lateral lumbar interbody fusion (LLIF) is a popular, minimally invasive technique that is used to address challenging multilevel degenerative spinal diseases. It remains controversial whether supplemental instrumentation should be added for multilevel LLIF. In this study, we compared the kinematic stability afforded by stand-alone lateral cages with those supplemented by bilateral pedicle screws and rods (PSR), unilateral PSR, or lateral plate (LP) fixation using a finite-element (FE) model of a multi-level LLIF construct with simulated osteoporosis. Additionally, to evaluate the prospect of cage subsidence, the stress change characteristics were surveyed at cage-endplate interfaces. METHODS: A nonlinear 3-dimensional FE model of the lumbar spine (L2 to sacrum) was used. After validation, four patterns of instrumented 3-level LLIF (L2-L5) were constructed for this analysis: (a) 3 stand-alone lateral cages (SLC), (b) 3 lateral cages with lateral plate and two screws (parallel to endplate) fixated separately (LPC), (c) 3 lateral cages with bilateral pedicle screw and rod fixation (LC + BPSR), and (d) 3 lateral cages with unilateral pedicle and rod fixation (LC + UPSR). The segmental and overall range of motion (ROM) of each implanted condition were investigated and compared with the intact model. The peak von Mises stresses upon each (superior) endplate and the stress distribution were used for analysis. RESULTS: BPSR provided the maximum reduction of ROM among the configurations at every plane of motion (66.7–90.9% of intact spine). UPSR also provided significant segmental ROM reduction (45.0–88.3%). SLC provided a minimal restriction of ROM (10.0–75.1%), and LPC was found to be less stable than both posterior fixation (23.9–86.2%) constructs. The construct with stand-alone lateral cages generated greater endplate stresses than did any of the other multilevel LLIF models. For the L3, L4 and L5 endplates, peak endplate stresses caused by the SLC construct exceeded the BPSR group by 52.7, 63.8, and 54.2% in flexion, 22.3, 40.1, and 31.4% in extension, 170.2, 175.1, and 134.0% in lateral bending, and 90.7, 45.5, and 30.0% in axial rotation, respectively. The stresses tended to be more concentrated at the periphery of the endplates. CONCLUSIONS: SLC and LPC provided inadequate ROM restriction for the multilevel LLIF constructs, whereas lateral cages with BPSR or UPSR fixation provided favorable biomechanical stability. Moreover, SLC generated significantly higher endplate stress compared with supplemental instrumentation, which may have increased the risk of cage subsidence. Further biomechanical and clinical studies are required to validate our FEA findings. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12891-017-1387-6) contains supplementary material, which is available to authorized users.