Comparison of anterior column reconstruction techniques after en bloc spondylectomy: a finite element study

Abstract

Total en bloc spondylectomy (TES) effectively treats spinal tumors. The surgery requires a vertebral body replacement (VBR), for which several solutions were developed, whereas the biomechanical differences between these devices still need to be completely understood. This study aimed to compare a femur graft, a polyetheretherketone implant (PEEK-IMP-C), a titan mesh cage (MESH-C), and a polymethylmethacrylate replacement (PMMA-C) using a finite element model of the lumbar spine after a TES of L3. Several biomechanical parameters (rotational stiffness, segmental range of motion (ROM), and von Mises stress) were assessed to compare the VBRs. All models provided adequate initial stability by increasing the rotational stiffness and decreasing the ROM between L2 and L4. The PMMA-C had the highest stiffness for flexion-extension, lateral bending, and axial rotation (215%, 216%, and 170% of intact model), and it had the lowest segmental ROM in the instrumented segment (0.2°, 0.5°, and 0.7°, respectively). Maximum endplate stress was similar for PMMA-C and PEEK-IMP-C but lower for both compared to MESH-C across all loading directions. These results suggest that PMMA-C had similar or better primary spinal stability than other VBRs, which may be related to the larger contact surface and the potential to adapt to the patient's anatomy.

Figure 1

Construction of the finite element model of the intact lumbar spine. The 3D geometries of the vertebrae were created by segmentation of the QCT images (a–b). The volumetric regions of the vertebrae (c) and the intervertebral discs (d) were defined in accordance with the data published in the literature. The finite element model of the lumbar spine was created by assembling the meshes of the vertebrae and intervertebral discs and adding the ligaments (e).

Figure 2

The surgical finite element models of the lumbar spine with the four investigated vertebral body replacements (a), and the contact surface of the VBRs (yellow) and the bone grafts (red) at the two adjacent endplates in percentage of the endplate area (b).

Figure 3

Kinematic parameters of the surgical model of the lumbar spine. The rotational stiffness of the models in % of the intact model (a). Segmental ROMs of the fixed L2–4 segment in flexion–extension, lateral bending, and axial rotation (b). ROM: range of motion.

Figure 4

Maximum von Mises stress in different load cases in the inferior bony endplate of L2 (a), in the superior bony endplate of L4 (b), and in the posterior rods (c). L: left rod, R: right rod.

Figure 5

Distribution of von Mises stress in different load cases in the inferior bony endplate of L2 from a bottom view (a) and in the superior bony endplate of L4 from a top view (b). FLE flexion, EXT extension, LLB left lateral bending, RLB right lateral bending, LAR left axial rotation, RAR right axial rotation.