Finite element analysis is widely used in spinal biomechanics. At present, it is of great significance in the evaluation and analysis of spinal fusion and ASD[17–20]. In the past few years, XLIF was recognized and adopted by spinal surgeons worldwide. Although this operation has many advantages, the complications caused by it should not be ignored. Therefore, improving the stability of the fusion segment, reducing the incidence of cage subsidence and ASD is still the focus of future research.
After lumbar interbody fusion, the biomechanics of the entire spine will change. Its stress load transmission will also change accordingly. In order to stabilize the fusion segment, internal fixation is often supplemented. Clinical studies had shown that, although stand-alone cage can increase the stability of the segment, compared with the additional plate and pedicle screw fixation, the mobility of the fusion segment was significantly increased [21]. Through imaging analysis, Marchi et al. [22] found that the rate of cage subsidence of stand-alone during XLIF was as high as 30%. Thus, supplementary fixation, such as pedicle screws, is recommended. Studies had shown that lumbar fixation and fusion can significantly reduce the ROM at the fusion level and provide strong stability[23]. Through research in this article, after internal fixation, the ROM of M1-M6 in the L4-5 segment was significantly reduced under all motion patterns. In terms of flexion and extension, the posterior pedicle fixation model (M5、M6) had a significant decrease, followed by M2, while M1, M3, and M4 had relatively low decreases. It showed that all fixation models could reduce the mobility of the fusion segment. The posterior pedicle fixation model provided high stability, followed by M2, while M1, M3, and M4 were relatively low.
Although much literature reported that there was no significant difference in stability and fusion rate between unilateral and bilateral pedicle fixation [24–25], many scholars still believed that the strength of unilateral fixation was not as stable as bilateral fixation [26–27]. Flexion and extension are the most frequent movements in daily human life. In this study, compared with other models, the pedicle screw fixation groups could significantly reduce the mobility of the fusion segment in terms of flexion and extension, indicating that the posterior pedicle fixation group can ensure the stability of the fusion segment, followed by the M2.
Cage subsidence is one of the common complications of lateral lumbar interbody fusion. Macki et al.[28] retrospectively analyzed 21 articles and included 1362 patients undergoing lateral lumbar interbody fusion, and identified that a subsidence incidence of 10.3% and a reoperation rate for subsidence of 2.7%. The maximum stress could be used to predict the sinking risk of the cage. The greater the stress, the higher the sinking risk [29–30]. Insufficient internal fixation strength was one of the important factors for cage subsidence. The use of supplemental internal fixation for lateral lumbar interbody fusion, such as bilateral pedicle screws, served to mitigate subsidence, protect the indirect decompression, and promote arthrodesis[31]. After fusion, the frictional contact between the cage and the endplate was likely to cause stress concentration. Compared with the unilateral fixation of pedicle screws, the bilateral pedicle screws can effectively control the stability of the index segment and reduce the load applied to the interbody fusion cage. In addition, bilateral fixation had relatively little stress damage to the fusion segment, and the protection of the cage itself was also relatively beneficial [32–33].
As shown in Fig. 4, the maximal von mises stress of the cage was obviously higher in M1 than that of other models (except for left bending), and the stresses of M3 and M5 were also relatively high.. In the left bending, the stress of M2, M1, and M4 were lower than in other models. This shows that M1 has a higher risk of cage subsidence, and the lateral internal fixation device can effectively share the stress of the cage during lateral bending.
After lumbar spine fixation and fusion, the mobility of the fixed segment decreases, causing the center of rotation to shift, thereby changing the motion of adjacent segments. Therefore, only by increasing the mobility of adjacent segments to compensate for the loss of mobility of the entire spine caused by the reduction of the mobility of the fixed segment.
In our study, after internal fixation, the ROM and disc stress of each model in the L3-4 segment in all motion cases were greater than that of the normal model, whereas the M5、M6 model had the larger range of activity in flexion and extension,, M6 had a high ROM in left and right bending, and M1 and M2 had the largest ROM in left and right rotation. By analyzing from the stress of the L3/4 disc, it could be concluded that the M6 disc had the highest stress in flexion; M1 had the highest stress in l extension; M1 and M2 had the highest stress in left and right rotation. Increased ROM and intervertebral disc stress in adjacent segments after spinal internal fixation were considered to be the main risk factors for ASD[ 34–35],and ASD was considered to be the result of spinal fusion [36]. With the increase in the stress of the intervertebral disc in the adjacent segment, the intervertebral disc deforms, resulting in an increase in the mobility of the adjacent segment. Therefore, the increase in the mobility of the adjacent segment may be caused by the increase in the intervertebral disc stress. Moreover, the flexion and extension activities were more closely related to the course of ASD [37].
Overall, all internal fixation models increase the mobility of adjacent segments and the stress of the intervertebral disc. Although the increase in mobility and intervertebral disc stress was not proportional to the strength of internal fixation, ASD should be considered when performing XLIF surgery. In clinical practice, the incidence of ASD can be reduced by reducing the ROM and intervertebral disc stress in adjacent segments.
Our study was based on finite element analysis and has several limitations. First of all, the division of elements and the determination of boundary conditions in the modeling process were all artificial settings, which need to be compared with cadaver specimens and in vivo experiments. Secondly, the adult healthy volunteer was selected, degenerative factors were not considered, and the muscles were not modeled. Therefore, the force of the human body could not be fully simulated. Furthermore, the biomechanical effects of internal fixation and surgery was simplified.