The emerging of non-fusion surgery is aimed to solve the long-term complication of fusion surgery that may bring the adjacent disc degeneration. Among several kinds of artificial discs developed in these years, the majority in the market is Prodisc-L (Synthes Inc.) which is designed with the purpose to restore the motions including anteroposterior translation, lateral bending, and axial rotation. These is also one artificial disc called Physio-L (Nexgen Spine) which were hyper-elastic material (Polycarbonate Polyurethanes) and is designed to restore the motions maintioned above plus axial loading. The concept of using hyper-elastic material as disc is to mimic the material properties of intervetebral discs so that this disc both absorb the axial loading and also restore the physiological range of motion. Few studies focused on the biomechanical behavior of hyper-elastic artificial discs have yet been reported. Therefore, the purpose of this study is to compare the biomechanical behavior between Prodisc-L and Physio-L.

A validated three-dimensional finite element model of the L1-L5 lumbar intact spine was used in this study with ANSYS software [Fig.1]. Total disc replacement surgery, partial discectomy, total nuclectomy and removal of the anterior longitudinal ligament were performed at the L3/L4 segment of this intact model, and the Prodisc-L and Physio-L was implanted into L3/L4 segment, respectively. In addition, hyper-elastic materials adopted by Physio-L are usually categorized by their hardness into soft and hard [Fig.2]. Therefore, two kinds of Physio-L were studied. A 400 N follower load and a 10 N-m moment were applied to the intact model to obtain four physiological motions as comparison baseline. The implanted models were subjected to 400 N follower load and specific moments in accordance with the hybrid test method.

For the Prodisc-L model in the surgical segment, the range of motion (ROM) varied by -26%, +17%, -0.01%, and -0.04% in flexion, extension, lateral bending, and axial rotation, respectively, as compared to intact model [Fig.3]. For the Physio-L (soft) model, ROM varied by +10%, +8%, +3%, and +19% in four physiological motions, respectively. For the physio-L (hard) model, ROM varied by +1%, +8%, +1%, and +11% in four physiological motions, respectively. For the Prodisc-L model in the adjacent segments, ROM varied by +4% ∼ +10%, -2% ∼ -5%, -1% ∼ -4%, and +1% ∼ -2% in four physiological motions, respectively. For the Physio-L (soft) model, ROM varied by 0% ∼ -5%, -2% ∼ -5%, -0% ∼ -5%, and -9% ∼ -11% in four physiological motions, respectively. For the physio-L (hard) model, ROM varied by +4% ∼ -2%, +8% ∼ -5%, +1 ∼ -5%, and +11% ∼ -6% in four physiological motions, respectively.

As seemed in the simulation, the behavior of Physio-L (both soft and hard) is similar to that of intact model under flexion and extension, but not in axial rotation. In addition, Physio-L (hard) model is more similar to intact model as compared to Physio-L (soft) model.