Finite element analysis of the screw in percutaneous axial lumbosacral interbody fusion

Abstract

Objective:  To establish a three‐dimensional finite element model of the L₅/S₁ motion segment with percutaneous axial lumbosacral screw and analyze the biomechanical stress on the screw.

Methods:  With the help of related software and the Mechanical Virtual Human of China, a three‐dimensional finite element model was established. Three different loading conditions on the screw were analyzed with this model.

Results:  Peak stresses on the screw under three loading conditions were 175.334 MPa, 183.765 MPa and 146.237 MPa, respectively. Generally, stress values were relatively low. The stress values were relatively high at the point of interbody fusion and middle part of the screw, all the highest values being localized to the upper and lower threads closest to the middle part. Comparison among the three conditions showed that torque load was the greatest, followed by vertical load, with lateral bending being the least.

Conclusion:  Percutaneous axial lumbosacral screw easily meets normal loading conditions and may be an effective method for lumbosacral fusion.

Introduction

The finite element method is an important computational mechanics approach. It first appeared in the 1940s and was the product of comprehensive utilization of applied mathematics, mechanics and computer science. In 1973, Belytschko et al. were the first to apply the finite element method to spinal biomechanics research ¹ . Since then, the finite element method has resulted in great developments in spinal biomechanics research and has become an important research method in related spinal biomechanics. Percutaneous axial lumbosacral interbody fusion (AxiaLIF) is a new type of lumbosacral fusion which was first introduced by Cragg et al. ² Although previous studies have reported related biomechanical research on this new technique, very few researchers have focused on biological mechanical analysis of the screw itself. The purpose of this study was to make a preliminary biomechanical analysis of the AxiaLIF screw and provide a theoretical basis for further clinical applications.

Materials and methods

AxiaLIF screw

The screw is made of titanium and has a total length of 70 mm (Rui Zhi Medical Production Equipment, Shanghai, China). It is composed of three parts: the middle for maintaining fusion (length 10 mm), the upper (25 mm) and lower (35mm) parts being inserted into the L₅ and S₁ vertebral bodies.

Modeling process

First, a solid model was established by computer‐aided design (CAD) software (Unigraphics NX 4.0) according to the actual dimensions of the screw. Then, computer graphics were obtained from the IGES format and imported into the finite element analysis software (ANSYS 10.0) to establish the finite element model. A complete human lumbosacral segment (L₅‐S₁) model was directly adopted from the “Chinese Mechanical Virtual Human” model library (Fig. 1a). Then the lumbosacral inter‐vertebral disc was removed, and replaced by the AxiaLIF screw (Fig. 1b) according to the manufacturer's instructions. Thus the integrated model was formed (Fig. 1c). The material properties of each part of the integrated model were ascertained by referring to the relevant literature (Table 1). With the help of ANSYS (ANSYS 10.0), the model was divided by appropriate cell type meshes. When the finite element model had been completed, it was comprised of a total of 815 921 units (Fig. 2a–c).

Figure 1.

Solid model of the lumbosacral spine, screw and integrated model. (a) An intact lumbosacral model from L₅ to the coccygeal vertebrae was chosen from the model library; (b) An AxiaLIF screw was made using CAD software; (c) An integrated model was formed for research purposes.

Table 1.

Material property and cell type

Part	Young's modulus (Mpa)	Poisson's ratio	Cell type
Cortical bone	12 000	0.25	Shell63
Cancellous bone	100	0.20	Solid92
Posterior	3500	0.25	Solid92
End‐plate	12 000	0.25	Shell63
Screw	100 000	0.30	Solid92

Thickness of cortical bone and cancellous bone were 0.5 mm and 0.2 mm, respectively.

Figure 2.

ANSYS was used to divide the solid model with cell type meshes. Finite models of (a) the lumbosacral spine, (b) screw and (c) integrated model were made.

Stress analysis

All the degrees of freedom of the sacroiliac joint and S₁ vertebrae lower surface were constrained according to the activity characteristics of the human lumbosacral joint. Three different conditions: 400 N vertical compression, 10 Nm torque moment and 10 Nm bending moment, were applied on the upper surface of the L₅ vertebra to simulate load stress on the screw (Fig. 3a–c).

Figure 3.

Stress distribution of screw under different load conditions. (a) Stress distribution under vertical compression is shown, different colors stand for different amounts of stress. (b) Torque moment and (c) bending moment are shown similarly.

Results

The stress distribution in the screw was analyzed under three different load conditions. The results showed that the peak load on the screw with vertical stress, torque moment and lateral bending were 175.334 MPa, 183.765 MPa and 146.237 MPa, respectively. The overall stress on the screw was relatively small under each loading condition. Local stress on the upper and lower 3–5 threads of the middle part was relatively high, with the highest value on the threads closest to the center. Additionally, stress gradually reduced towards both upper and lower parts of the screw.

Discussion

Fusion is a very common form of management of spinal degenerative disease at the L₅‐S₁ level. It has been reported that lumbosacral fusion has been applied more and more frequently in recent decades ³ . Although surgical technologies have improved greatly, the rates of pseudoarthrosis, internal fixation failure and complications such as sagittal imbalance are still high after L₅‐S₁ fusion. Local lumbosacral anatomical complexity, relatively poor bone quality and the unique biomechanical characteristics are important factors affecting the postoperative results of lumbosacral fusion ⁴ . Traditional classical surgical treatments often damage the normal structure and physiological function of the spine, which may aggravate lumbosacral instability.

Technological advances in spine surgery have allowed spine surgeons to develop less invasive approaches which put the important anatomic structures at less risk and thus reduce morbidity. In 2004, Cragg et al. first introduced percutaneous anterior lumbosacral axial fusion (AxiaLIF), which was a new operation style for L₅S₁ degenerative disc disease ² . In this operation, L₅S₁ discectomy, bone grafting and axial fixation are performed through a work tunnel which lies close to the middle sacral surface. The technique has the advantage of avoiding destruction of the structures which stabilize the spine, consistent with minimally invasive spine surgery (MISS). Thus, it theoretically has biomechanical advantages. It not only retains the posterior spinal structures, but also decreases anterior disc damage. Ledet et al. have reported that standalone AxiaLIF provides significant construct rigidity ⁵ . They attributed this result to the preservation of the annulus.

Akesen et al. reported the biomechanical results in adult lumbosacral specimens ⁶ . Because the annulus fibrosus remains intact, AxiaLIF can effectively reduce the extent of lumbosacral articulation. The range of movement (ROM) of L₅‐S₁ after AxiaLIF augmented with pedicle screws is similar to that of anterior lumbar interbody fusion cages augmented with pedicle screws. The authors suggest that, in order to ensure successful fusion, supplementary posterior fixation is still necessary. One year later, similar conclusions on AxiaLIFII were reached by Erkan et al. ⁷ AxiaLIF II, AxiaLIF II with articular screw and AxiaLIF II with pedicle screws were compared to intact specimens in this report. All three methods greatly restricted the ROM of L₅‐S₁; there were no statistically differences among the three methods. Although AxiaLIF II alone did provide sufficient biomechanical stability, posterior internal fixation was also suggested to ensure successful fusion. Preliminary promising results have also been reported for AxiaLIF in animal experiments and clinical trials ⁸ , ⁹ , ¹⁰ , as well as in relevant biomechanical studies. However, there is still lack of information on the biomechanical status of the screw itself.

In our study, screw stress was small under all three loading conditions, and the stress peak was much smaller than the yield strength of the titanium screw of 894–1034 MPa ¹¹ (the screw strength in our study was greater than the actual yield strength under peak stress of 895 MPa). This indicates that the screw is capable of bearing more stress than that occurring in normal activities of the human body. In addition we found that, regardless of load conditions, the relative concentration of stress on the screws was always located in the central section and the few upper and lower threads on either side of it. What is more interesting was that all the stress peaks were located in the first thread below the central section. This demonstrated that the stress is relatively large at the lower edge of L₅ vertebra and the upper edge of S₁ vertebra. In order to effectively guarantee stability of fusion, good and full connection between vertebral body and screw must be ensured during surgery. In torque and lateral bending, stress gradually became less from the central section to the extremities of the screw. We speculate that this can be explained by concluding that the fused vertebrae are the main weight‐bearing unit. In the normal human body, the anterior column of lumbar vertebrae supports most of the axial load. It also plays a very important role in stabilizing the lumbar motion segment. This was validated in this study. The result of stress peak comparisons between the three loading conditions was: torque > vertical compression > lateral bending, which suggests that shear stress caused by torque would result in the greatest damage to the screw. This also implies that the facet joints of the lumbar sacral vertebrae, which can effectively prevent shear stress, play an important role in intervertebral stability. The stability of the facet joints for the promotion of fusion has a positive significance. Dong et al. have reported that lumbar motion segment torsional rigidity decreases by 32%–39% after unilateral partial or total removal of the facet joints ¹² .

In brief, the preliminary results of our study indicate that the strength of AxiaLIF can meet daily requirements, such as vertical compression, torsion, lateral bending, in humans. The maximum stress values are much smaller than the yield strength of titanium alloys. Screw stress peaks were all located in the terminal threads below the fusion section, which suggests that it is very important to make sure that good and full contact between the terminal threads of the screw and S₁ is achieved during surgery. The preliminary study discusses the results of finite element analysis of AxiaLIF, further study and verification is still necessary. However, there are some shortcomings: (i) In order to simplify the procedure, the L₅S₁ disc was totally removed when constructing the integrated model, which is not consistent with clinical actuality. Considering that the annular fibrosus is an important stable structure, its presence may help share the local stress on the screws. It is not possible that the procedure we adopted would have had a negative impact on the results. (ii) Because the study was a finite element model simulation study, and the model used a lumbosacral spine from normal adults, the load simulation environment and the actual stress in patients was very difficult to match accurately. Further study will be helpful.