In vitro testing of lumbar disc arthroplasty devices

Abstract

Background Context: Interest in lumbar disc arthroplasty as an alternative to fusion surgery continues to grow. The goal of disc arthroplasty is to replace the diseased disc while preserving and/or restoring motion at the operated spinal level. Different paradigms exist in the design of total disc arthroplasty devices. Purpose: The purpose of this study was to compare the in vitro biomechanics of a more constrained ball-and-socket design (Prodisc-L, Synthes Spine and Maverick, Medtronic) and a less constrained mobile-bearing design (Charité, DePuy). The biomechanical performance of the disc prostheses was compared to harvested and fused spine conditions. The fused model was simulated using single-level posterior pedicle screw fixation instrumentation. Study Design/Setting: In vitro test to compare the biomechanical properties of three different lumbar disc replacement devices in a human cadaveric model. Methods: Twenty human cadaveric lumbar spines (L1-sacrum) were tested in flexion, extension, lateral bending, and axial rotation under displacement control to a target bending moment of 8Nm. The spine conditions tested were: harvested spine (n=19); L5-S lumbar disc replacement using Prodisc-L (n=13); Maverick (n=7); Charité (n=6); and L5-S pedicle screw fixation (PSF) (n=19). The first 12 spines were split into 2 groups: 6 were instrumented with the Charité and 6 with the ProDisc-L. The next 7 spines tested were split into 2 groups: 4 instrumented with ProDisc-L and 3 with the Maverick. After completing all tests on the second group of 7, the Maverick and ProDisc-L discs were swapped between spines and retested. The Click'X pedicle screw system (Synthes Spine) was used to simulate the fusion model in all spines tested. For axial rotation tests, a 100N compressive load was applied. Measurements included vertebral motions, total spine rotation, and applied loads. The percent contribution of rotation at the instrumented (L5-S) level relative to total rotation (L1-S), as well as at the remaining adjacent levels relative to total rotation, was determined a common load limit (8Nm) and compared using a one-way ANOVA and SNK test (P<0.05). Results: A significant reduction in motion occurred at the operated level of PSF condition compared to the three disc arthroplasty conditions for all loading modes. No differences occurred between the 3 disc conditions for all modes tested, except at the instrumented level of the ProDiscL (93% of H) and Maverick (128% of H) spine conditions during combined flexion+extension. The reduced motion at the operative level of the PSF condition was transferred to the adjacent levels and caused a significant increase in motion during combined flexion+extension at all adjacent levels for the 3 disc arthroplasty conditions, during combined right+left lateral bending at L1-L2 for all disc conditions and at L3-L4 for the Charité, and during combined right+left axial rotation at L3-L4 for all three disc conditions. Conclusions: Issues pertaining to adjacent segment disease (ASD) with pedicle screw fixation were supported by increased motion contributions at multiple sub-adjacent segments. However, disc arthroplasty eliminated any significant increase and may prevent ASD. Compared to pedicle screw fixation, the three differently designed disc prostheses (Prodisc-L, Maverick, and Charité) remained stable and provided improved lumbar mobility. The only notable difference between the disc designs was the increased combined flexion+extension motion at the operative level of the Maverick disc compared to the ProDisc-L device. © DiAngelo et al.