The Biomechanical Effect on the Adjacent L4/L5 Segment Under S1 Superior Facet Arthroplasty: A Finite Element Analysis

Zewen Shi Ningbo University Lin Shi Ningbo No 2 Hospital: Ningbo Huamei Hospital University of Chinese Academy of Sciences Xianjun Chen Ningbo No 2 Hospital: Ningbo Huamei Hospital University of Chinese Academy of Sciences Jiangtao Liu Ningbo No 2 Hospital: Ningbo Huamei Hospital University of Chinese Academy of Sciences Haihao Wu Ningbo No 2 Hospital: Ningbo Huamei Hospital University of Chinese Academy of Sciences Chenghao Wang Ningbo No 2 Hospital: Ningbo Huamei Hospital University of Chinese Academy of Sciences Zeming Chen Ningbo University Fang Yang Ningbo University Sheng Yu Ningbo University Qingjiang Pang (  pangqingjiang@ucas.ac.cn ) Ningbo No 2 Hospital: Ningbo Huamei Hospital University of Chinese Academy of Sciences


Introduction
In recent years, percutaneous transforaminal endoscopic discectomy (PTED) has been accepted as an alternative treatment for disc herniation due to its advantages over traditional open surgery [1][2][3]. However, the working channel is often di cult to establish in L5/S1 segment as high iliac crest, hyperplastic articular process and narrow foramen intervertebrale. In these cases, the articular process arthroplasty is needed [4]. In the long term, the degeneration of responsible segment and adjacent segments are clinically common after facet arthroplasty. Though the effect of S1 superior articular process arthroplasty on responsible segment had been reported [5], the effect on adjacent segment (L4/L5) has not been unreported.
In this study, nite element model was used to simulate S1 superior articular process arthroplasty. The S1 superior articular process was formed parallel to the S1 upper endplate from the apex to the base and perpendicular to the S1 upper endplate from the ventral to the dorsal. The effect of adjacent segment (L4/L5) biomechanical change was explored.

Research object and data collection
Eight healthy volunteers were selected, whose age range from 22 to 29 years old. X-ray of the lumbosacral vertebra was taken to exclude pathological conditions. Computed tomography scan of the lumbar spine was obtained with 1.0 mm thickness. The work has been approved by the Hospital Ethical Committee (Hwa Mei Hospital, University of Chinese Academy of Science) and that subjects gave informed consent to the work.

Normal nite element model
The computed tomography images were post-processed for boundary detection with the Mimics 17.0 (Materialise, Belgium), then geometric models were established. Geomagic Studio 10.0 (Geomagic,USA) was used to import the spine modes for repair, noise reduction. Hypermesh 13.0 (Altair USA) was used to generate the FE mesh for analysis. Nastran 2012 (MSC, USA) was used to construct nite element models (FEMs). The material properties of the model were listed in Table 1 using the results of previously published studies [1,6,7]. The surgical models were constructed based on the validated intact model (M1). Geomagic was used to simulate unilateral S1 superior superior articular process arthroplasty. The S1 superior superior articular process was graded formed (1/5, 2/5, 3/5, 4/5, 5/5) parallelly to the S1 upper endplate from the apex to the base (transverse plasty) and perpendicularly to the S1 upper endplate from the ventral to the dorsal (longitudinal plasty), respectively. The models established were de ned as M2-M10 (M2-M5 represents transverse plasty 1/5 − 4/5, M6-M9 represents longitudinal plasty 1/5 − 4/5, and M10 represents plasty 5/5) ( Fig. 1).

Boundary and loading conditions
The inferior surface of the S1 vertebra was constrained completely (Fig. 2). A vertical load of 500 N and a torque of 10 N·M were applied to the L4 to simulate the weight of the body and various loading conditions of the lumbar spine. The torque along the axis generate exion, extension, forming contralateral exion, forming side exion, forming contralateral rotation, forming side rotation. The ROM and intervertebral disc von Mises stress of adjacent segment (L4/L5) were quanti ed.

Statistical analysis
SPSS 19.0 software was adopted for statistical analysis in this study. Data was represented by x ± s. ANOVA was used between groups, and Dunnett analysis was used for pairwise comparison. Herein, P < 0.05 was considered to be statistically signi cant.

Discussion
Due to the anatomical characteristics of L5/S1 segment, posterior approach is generally the preferred method of spinal endoscopy. However, the posterior approach is di cult for patients with stenosis of vertebral canal, small inter-laminar space and extreme lateral lumbar disc herniation. In addition, research has shown that the lumbar instability would occur after the medial of inferior articular process is removed more than 1/2 through the posterior approach [8]. At this time, the posterolateral approach is needed. However, in L5/S1 segment, the superior articular process arthroplasty is usually needed to enlarge the foramen intervertebrale under posterolateral approach so as to reduce the interference in nerves and expand the surgical indications [9].
Unilateral S1 superior articular process reconstruction has a great impact on the biomechanics of the responsible segment, and the results has been published in previous studies by our research group [5]. Adjacent disc degeneration often occurs after PTED, and the annual risk rate of clinically related adjacent segmental diseases is reported to be 0.6-3.9% [10,11]. However, there has been no consensus about the effects of S1 superior articular process forming on the biomechanics of adjacent L4/L5 segment. Hence, it is of great signi cance to explore the effect of S1 superior facet arthroplasty on the biomechanics of L4/L5 through nite element analysis, and indirectly re ect the effect of S1 superior articular process arthroplasty on the risk of adjacent segment degeneration.
The biomechanical study of facet joint and spinal degeneration has been deepened gradually, with the popularity of the technology of PTED. Matsuo et al [12] showed that the degenerative lumbar spondylolisthesis is signi cantly related to the sagittal and axial angles of the facet joints. It can be seen that the facet joint plays an important role in spinal degeneration. As we know, facet arthroplasty can reduce the stability and increase the risk of degeneration of the responsible segment [5,13,14]. However, the effect of S1 superior articular process arthroplasty on the ROM of adjacent L4/L5 segment has not been reported. The results of this study showed that the ROM of L4/L5 segment increased signi cantly in exion, lateral exion and lateral rotation, when the longitudinal forming was more than or equal to 3/5. Compared with the normal model, the difference was statistically signi cant. In lateral bending and lateral rotation, the ROM of L4/L5 segment increased signi cantly, when the transverse forming was more than or equal to 2/5. Compared with the normal model, the difference was statistically signi cant. These results suggest that the ROM of adjacent L4/L5 segment would be affected after the longitudinal shape of S1 superior articular process is more than or equal to 3/5 or the transverse shape is more than or equal to 2/5.
As we know, 25% of the axial compressive stress and 40%−65% of the rotational and shear stress of the lumbar spine are borne by the facet joint [15]. The increased stress of the intervertebral disc caused by the asymmetry of the articular process contributes to the increased risk of lumbar degeneration [16]. Qian et al [14] reported that 1/4 of L5 unilateral superior articular process forming could increase the stress of the same segment of intervertebral disc. However, few study about the effect of S1 superior facet arthroplasty on the adjacent disc stress has been reported. The results of this study showed that the disc stress of L4/L5 segment increased signi cantly in exion, contralateral exion and contralateral rotation after the longitudinal forming was more than or equal to 3/5. Compared with the normal model, the difference was statistically signi cant. In exion, extension, lateral exion and contralateral rotation, the disc stress of L4/L5 segment increased signi cantly after the transverse forming was more than or equal to 1/5. Compared with the normal model, the difference was statistically signi cant. It was worth noting that the stress of the L4/L5 disc increased most obviously when it rotated to the contralateral side of the forming. This is consistent with the statement that the facet joint plays a major role in the torsional stiffness of the intervertebral disc [17]. In other words, the disc stress of adjacent L4/L5 segment would be affected after the longitudinal form of S1 upper joint is more than or equal to 3/5 or the transverse form is more than or equal to 1/5, resulting in an increased risk of degeneration of adjacent segment.
In terms of experimental methods, three-dimensional nite element method has been widely used in orthopedic eld, especially as a high simulation method in spine [18]. However, in another way, the accuracy of the research is decreased by the nite element method owing to it is a method of simplifying the complexity. In addition, this study infers the relationship between facet arthroplasty and adjacent segment degeneration from the immediate in uence, failing to monitor the whole degeneration process dynamically. In the experimental design, the apex and basal part are common locations for S1 superior articular process arthroplasty. In addition, the original shape of foramen intervertebrale should be maintained and the destruction of anatomical structure should be reduced as far as possible. Two experimental methods were designed, parallel to S1 upper endplate from the top to the base and perpendicular to S1 upper endplate from ventral to dorsal under the premise of taking the upper edge of S1 pedicle as the lowest level.

Conclusions
In conclusion, the S1 superior articular process of the lumbar spine is not only of great signi cance to the biomechanics of the same segment, but also to the adjacent segment. It is possible that the stability of adjacent L4/L5 segment would be decreased and the stress of intervertebral disc would be increased after unilateral S1 superior facet arthroplasty. It is suggested to form ventral to dorsal of unilateral S1 Superior articular process arthroplasty should be controlled less than or equal to 3/5 and it is not recommended to form from apex to base, combining with the previous study of the effect of S1 superior articular process arthroplasty on the same segment [5]. Otherwise, the long term risk of adjacent segment degeneration would be increased.  Lumbar motion of L4/L5 segment after S1 superior facet arthroplasty Lumbar disc stress of L4/L5 segment after S1 superior facet arthroplasty