The study subject was a 32-year-old healthy male volunteer. This study was approved by the Ethics Committee of Guangxi Medical University, and informed consent was obtained from the volunteer. A 3D finite element model of C3–C7 was reconstructed with computed tomography (CT) data using Mimics 20.0 (Materialise, Leuven, Belgium). Next, the 3D model was smoothed and polished using Geomagic 12.0 (Geomagic, United States). In the model, the cortical and cancellous bone, endplate, annulus fibrosus, nucleus pulposus, posterior elements, anterior longitudinal ligament, posterior longitudinal ligament, capsular ligament, transverse ligament, ligamentum flavum, interspinous ligament, supraspinous ligament and capsular ligament were reconstructed. Table 1 lists all the material properties and element types of these tissues according to Polikeit et al. (2003).

To obtain accurate data, the mesh of the model was validated. The mesh convergence test was performed with the 3D finite element model of C3–C7, and the mesh was divided by five sizes (0.5, 1, 1.5, 2, and 3 mm) (Figure 1). The five mesh models were subjected to testing, with the maximal von Mises stress on the vertebral body being the parameter of interest. Figure 2 shows the relationship between the stress and the mesh size. Based on the calculation time and results, the requirement of a change rate of <5% for the maximal von Mises stress was met (Dai et al., 2022). Hence, the unit size of 1 mm was used as the final mesh size for this study (Table 2).

Setup of the C3–C7 model (Figure 3): All facet joints were set as contact, with a friction co-efficient of 0.1 (Liu et al., 2011). All degrees of freedom of the endplate beneath C7 were restricted, and a pre-load of 50 N was applied to the endplate above C3 to simulate the weight of the head, and the additional bending moment of motion was 1 Nm (Lee et al., 2011). To validate the C3–C7 finite element model, a bending moment of 1.5 Nm was applied to the C3 plane, and the model was loaded in flexion, extension, lateral bending, and axial rotation (Lee et al., 2011), followed by calculation, data extraction, and determination of the range of motion (ROM).

3D physical modeling was performed for a novel ATPS-plate system of the cervical spine (Patent No: ZL 2018 2 0814089.9) (Figure 4) and for a titanium mesh cage (TMC) by using Solidworks 2015 (Dassault Systemes, France). C6 corpectomy and C5/6 and C6/7 discectomy were simulated, and a TMC packed with cancellous bone was implanted into the decompression groove. The components of the screw-plate system were then assembled. For contact setup: the intervertebral disc, nucleus pulposus, and endplate were bound to each other; the screws and vertebrae contacted inseparably, the screws and plate and the plate and vertebrae contacted in a face-to-face manner. There was separable rough contact between the TMC and endplate, which did not allow for sliding. The finite element model of four types of internal fixation in ACCF following reconstruction for single-level corpectomy and decompression was simulated (Figure 5): UATPS, including ipsilateral transpedicular screw-contralateral vertebral body screw (ITPS-CVBS) and cross transpedicular screw-vertebral body screw (CTPS-VBS); BATPS; and VBS.

The 4-type finite element model was imported into Ansys Workbench 18.0 (ANSYS, United States), and working conditions were established and calculated with reference to the original model of the full set. Subsequently, the ROM, the maximal von Mises stress on TMC, the stress distribution and maximal von Mises stress on screws, and the maximum sliding displacement of screws were analyzed for the four models.

Under pure moments and motion loads, the predicted results of ROM were compared with the results of the validated model to assess the validity of the new model. The results were congruent with the literature (Lee et al., 2011), thus, the model was validated (Figure 6).

Compared with the original model, all the internal fixation models exhibited reduced ROM in the six orientations, namely, reduction of ROM by 42.1%–48.1% in flexion, by 44.3%–47.5% in extension, by 20.2%–33.5% in left lateral bending, by 15.7%–22.1% in right lateral bending, by 20.7%–24.6% in left rotation, and by 20.0%–22.2% in right rotation. The ROM differed insignificantly between the four models and was in the ascending order: BATPS<CTPS-VBS<ITPS-CVBS<VBS (Figure 7).

Under the loading conditions in flexion, extension, lateral bending, and lateral axial rotation, the maximal von Mises stress on TMC was lowest in the BATPS group (12.18, 43.58, 29.58, 30.05, 51.59, and 40.28 MPa, respectively) and highest in the VBS group (14.92, 51.10, 24.93, 36.92, 47.36, and 49.17 MPa, respectively), while that in the UATPS group was between the BATPS group and the VBS group (CTPS-VBS: 14.20, 48.41, 31.72, 33.00, 41.99, and 41.48 MPa, respectively; ITPS-CVBS: 14.64, 50.26, 32.41, 33.11, 42.03, and 42.09 MPa, respectively) (Figure 8). The stress on BATPS in flexion, extension, and lateral bending was significantly lower than that on UATPS, but the differences were insignificant in rotation. However, regardless of bilateral or unilateral ATPS, the stress on TMC was lower compared with that on VBS.

Comparison of UATPS (ITPS-CVBS vs. CTPS-VBS) showed that the stress reduced by 37.4% on ITPS-CVBS when compared to CTPS-VBS in left bending, but the group differences were insignificant in flexion, extension, right lateral bending, and lateral rotation (Figure 9). BATPS endured the maximum stress of 82.07 MPa in extension. VBS endured stress of 23.38, 58.92, 78.80, 36.09, 44.56, and 40.66 MPa, respectively, under the six loads, which were significantly lower than those on transpedicular screws (ITPS-CVBS, CTPS-VBS, and BATPS) (Figure 9). Figure 10 shows the stress distribution on the screws under the six loads in each model group.

The maximum displacement of screws was largest in the VBS group under loads in flexion, extension, left and right lateral bending, and left and right lateral axial rotation (Figure 10). When compared with VBS, the maximum displacement of screws reduced by 8.1%, 2.3%, 4.1%, 7.7%, 22.1%, and 6.4%, respectively, with ITPS-CVBS; by 29.4%, 11.9%, 32.7%, 16.6%, 22.1%, and 6.6%, respectively, with CTPS-VBS; and by 39.6%, 37.9%, 64.1%, 64.6%, 53.1%, and 50.7%, respectively, with BATPS (Figure 11).

Following cervical spinal fusion, a smaller ROM is associated with higher stability and reduced likelihood of loosening of internal fixation devices. Wu et al. (2018) simulated ATPS and VBS internal fixation with six cervical spine specimens and found that ATPS had a smaller ROM than VBS, and that ATPS internal fixation can offer adequate stability for three-column injury to the lower cervical spine. Our results demonstrated that compared with the original model, the four screw insertion methods, namely, BATPS, ITPS-CVBS, CTPS-VBS, and VBS, exhibited significantly reduced overall ROM of the cervical spine under all six loads, with the ROM in the ascending order of ATPS<ITPS-CVBS<CTPS-VBS<VBS (Figure 7). When compared with VBS, ATPS can reduce the ROM of the cervical spine, which is consistent with the findings of Wu et al. (2018). It can be inferred that the use of ATPS can reduce the ROM of the cervical spine and thus increase the stability of the cervical spine, and that such advantages may be more pronounced in the internal fixation for multilevel cervical corpectomy and spinal fusion. Notably, the number of screws used and the method of screw insertion influence the ROM of the cervical spine differently. An increased number of ATPS used is associated with a decreased ROM; given a same number of ATPS or VBS used, cross insertion of ATPS resulted in a smaller ROM and a better overall stability compared with ipsilateral insertion.

Implant displacement and settlement is associated with the stress on the implant-endplate interface. An excess load of the endplate may lead to implant displacement and endplate damage, ultimately resulting in failure of the internal fixation. In this study, the maximal von Mises stress on TMC was lowest in flexion and highest in extension in each group, possibly because the screws and plate in front of the cervical spine can offset some stress in flexion. However, the magnitude of the stress varied among the different methods of fixation. Under various working conditions in motion, the maximal von Mises stress on TMC was lowest with BATPS, was lower with CTPS-VBS than with ITPS-CVBS, and was highest with VBS. Hence, internal fixation with VBS yields greater stress on TMC and is likely to cause damage to the endplate bone.

Owing to bending, deformation, and loosening, the anchoring components and rods between the screw and bone are liable to displacement (Oda et al., 2022). Fogel et al. (2003) also reported that fracture is likely to occur in the presence of failure between the screwhead and screw. In the clinical setting, the screw is usually fractured at the junction of the nut and the plate, and the stress on this junction is key to facture of the screw. In this study, the stress cloud map showed that the stress was predominantly concentrated at a point two-thirds of the length from the tail of the screw after ACCF, the stress of the ATPS from the cranial part was higher than that of the caudal part. The maximal von Mises stress was greater on unilateral or bilateral ATPS compared with that on VBS. VBS inserted into the anterior and middle columns of the vertebral body and into the cancellous bone of the vertebral body can bear a small stress during movements of the cervical spine. In contrast, ATPS penetrates the anterior, middle, and posterior columns of the vertebral body; hence, it bears a great stress during movements of the cervical spine, thereby avoiding loosening. The screw stress cloud maps did not show any red areas reflective of a concentration of the stress on VBS in the six orientations of movement, possibly because of transfer of the stress onto the TMC. Hence, the likelihood of TMC displacement is high with VBS. With respect to UATPS fixation, the stress was reduced by 37.4% with ITPS-CVBS when compared with CTPS-VBS in left bending, possibly because no ATPS shared the stress in ITPS-CVBS at the left side of the cervical spine. Hence, in light of the maximal von Mises stress on screws and the stress cloud map, ATPS can tolerate greater stress than VBS during cervical spine movements. BATPS can balance the stress during cervical spine movements better than UATPS. In terms of UATPS, CTPS-VBS can tolerate lateral bending better than ITPS-CVBS.

The maximum displacement of the screw is proposed as a stability parameter (Li et al., 2013), which can reflect the overall stability of the screw-plate system better than the ROM of the cervical spine and the stress on internal fixation devices. Screw loosening results from insufficient stress on the screw-bone interface, and the bone density of the vertebral body, the length of screw, the thread type, the screw diameter, and single or double cortical fixation all influence screw stability (Zhang et al., 2006; Matsukawa et al., 2016). Pedicle is the most abundant area of cortical bone in the cervical spine. Koller et al. (2008a) demonstrated that the fixation strength of ATPS was 2.5 times that of conventional anterior VBS. In this study, under loads in flexion, extension, lateral bending, and lateral axial rotation, the maximum displacement was largest with VBS, smallest with BATPS, and was moderate with ITPS-CVBS and CTPS-VBS. UATPS showed even smaller displacement than VBS under the six loads, suggesting a good stability, which is consistent with the literature (Koller et al., 2010; Zhao et al., 2018). With respect to UATPS, CTPS-VBS exhibited smaller screw displacement and better stability compared with ITPS-CVBS in flexion, extension, and lateral bending, but not in rotation. Hence, in terms of the ATPS technique, stability is highest with BATPS, followed by CTPS-VBS, and is lowest with ITPS-CVBS.

Taken together, based on the ROM, the maximal von Mises stress on TMC, the stress distribution and maximal von Mises stress on screws, and the maximum sliding displacement of screws, we compared different methods of anterior screw insertion into the cervical spine in ACCF and demonstrated that the stability was highest with BATPS, followed by CTPS-VBS, ITPS-CVBS, and then VBS. Hence, the ATPS technique can reduce the incidence of screw loosening and TMC displacement.

This study does have some limitations. The study investigated the initial stability of different screw fixations following single-level ACCF, and further studies need to be conducted on the ultimate mechanical properties and fatigue resistance of these screw fixations. In addition, no finite element model involves muscles, and the data may change with the addition of muscles. Furthermore, the model in this study did not simulate multilevel ACCF or osteoporosis-related working conditions. Our data are expected to change in multilevel spinal fusions and osteoporosis, but the overall trend of stability may not change. In addition, in vitro biomechanical testing and clinical studies need to be performed to appraise the results of this study.

The stability of various methods of anterior screw insertion in the cervical spine for ACCF differs. The findings of this study hold the potential to aid in the development of an optimal fixation method for lower cervical spinal fusions, with the goal of reducing internal fixation failures following anterior cervical spinal fusions. In the context of finite element analysis (FEA), BATPS is recommended whenever feasible, and in cases where only UATPS insertion is possible, CTPS-VBS is recommended. Prior to conducting clinical trials, it remains essential to perform in vitro experiments to further analyze and validate the results.