PLIF with posterior instrumented supplements has been used in the management of lumbar conditions for many years [3, 4]. With decades of development, neural injury, broad dissection of bone and muscle, and blood loss have been improved [3, 4].
The addition of autologous bone graft made from the iliac crest has been considered as the “gold standard”. The high fusion rate is achieved because of the provision of cellular factors that are favorable for bone healing, and the inherent properties of osteoconduction and osteoinduction [9, 16]. Nevertheless, the complications and morbidities such as infection, bleeding, pain at the donor site, and iliac fracture cannot be ignored [9, 17, 18]. Commonly, the spinous process and laminae are dissected during PLIF, so the use of autologous bone graft from the spinous process and laminae can avoid the complications and morbidities while using a bone graft from the iliac crest [6, 8, 9].
Cages made of titanium have been used in interbody fusion for many years. Although the fusion rate is high when using a titanium cage, a high risk of subsidence has been reported due to its high stiffness [1,2,3,4,5, 19]. PEEK has a stiffness close to that of cortical bone and has been widely used to produce cages [10, 20]. Furthermore, using PEEK cages in interbody fusion contributes to a nominal immediate anterior load sharing and restoration of height in the collapsed intervertebral space caused by degenerated discs [21]. The subsidence rate in the interbody fusion using PEEK cages was lower than that when using the titanium cages [2, 7, 22]. Nevertheless, the fusion rate was lower when using the PEEK cages because of its disadvantages of osteoconduction and osteoinduction [2, 7, 23].
To prevent complications when using autologous bone graft harvested from the iliac crest and to overcome the disadvantages of titanium or PEEK cages, a cage made from autologous bone was designed, and its biomechanical performance was evaluated using FEA. The biomechanical behaviors in various surgical models with three kinds of cages on ROM, the maximum stress in the cage-endplate interface, FJF, and IDP in four motion modes were evaluated. As shown in Fig. 9, Table 3, and Table 4, compared to the intact model, the ROM decreased by 97.30–98.21% at the surgical levels, while increased by 5.56–16.25% at adjacent levels in all surgical models, which was consistent with the results of the previous studies [11,12,13,14]. No significant differences were found in the ROM of the three surgical models, suggesting that the use of autologous bone-cage in PLIF could provide stability close to that of traditional solid cages. The reduced ROM at the surgical levels could be good for fusion and healing. As shown in Figs. 10 and 11, and Tables 5 and 6, the IDP at the adjacent levels increased in the surgical models during various motion modes, except for lateral rotation. The FJF at the L4/5 level disappeared in the surgical models because of the cages and rigid fixation with screws and rods. At the adjacent levels, the FJF increased in all motion modes in all surgical models. Nevertheless, no significant differences were found in FJF at the adjacent levels between the models. As shown in Figs. 13 and 14, and Table 6, the maximum stress in the cage-endplate interface was significantly higher in the surgical models of titanium cage and PEEK cage than that of the autologous bone-cage in all motion modes, except in extension. In the inferior cage-endplate interface in axial loading and flexion, the maximum stress in the surgical models with a titanium cage and PEEK cage were dozens of times higher than that in the surgical model with autologous bone-cage. Grant et al. measured the stiffness of different regions on the endplate and found a trend of decrease from the outside to the center of the endplate [24]. If the local stress was higher than the limit of the related regions, microfracture would occur [3, 4, 16,17,18, 24], leading to osteolysis and cage subsidence [3, 4, 16,17,18]. The stress in the cage-endplate interface was only less than 10 MPa in the autologous bone-cage model, which suggested that the lower the stress in the cage-endplate interface, the lower possibility of occurrence of microfractures, osteolysis or cage subsidence.
As the material property of the autologous bone-cage is unclear, the properties were set at a Young’s modulus of 5000 MPa, with a Poisson’s ratio of 0.29. The basis for setting the parameters was that the mixture of the spinous process and laminar bone obtained by compression is clinically roughly similar to that of cortical bone and cancellous bone in a 1:1 ratio. In the present study, the bone-cage after autogenous bone compression was considered as a whole cage. Therefore, no interaction was considered among the compressed parts. Nevertheless, it might influence the results, but further studies will be necessary to characterize the parts. This could provide some theoretical basis on what is happening, but clinically, it might have little value since the bone-cage made from compressed resected bones will be different from one patient to another. Especially, autograft tissue from spinous processes/lamina often contain soft tissue fragments that may change the biomechanical properties of the bone-cage, and this is a limitation of the model. The compressed cage may interfere with or even destroy the microstructure of bone cells. Due to the possible loss of growth factors in bone after extrusion, the time required for intervertebral fusion during our clinical case follow-up was found to be slightly longer than for the other commercial cages, but there was no difference in the overall fusion rates [8]. Some bone resorption could occur, but it did not affect the fusion rate in the clinical study [8]. In the near future, we will conduct further experiments to check whether the osteogenic performance of the compressed cage formed after extrusion has changed.
The limitations of our study should not be ignored. We compared parameters in our study with that of the obtained parameters from the model in the literature. The validation should be conducted by comparing the results in the FE study with that of biomechanical experiments and clinical studies. The FE model of the lumbar spine was simplified to improve the efficiency of convergence in the FE study. It still cannot represent the actual conditions in a real human body. Furthermore, the material property of the autologous bone-cage that was set in this study was still not the precise value, so more biomechanical studies should be conducted in the future.