Biomechanical Comparison of Anterior Cervical Corpectomy Decompression and Fusion, Anterior Cervical Discectomy and Fusion, and Anterior Controllable Antedisplacement and Fusion in the Surgical Treatment of Multilevel Cervical Spondylotic Myelopathy: A Finite Element Analysis

Abstract

Purpose

Multilevel cervical spondylotic myelopathy poses significant challenges in selecting optimal surgical approaches, warranting a comprehensive understanding of their biomechanical impacts. Given the lack of consensus regarding the most effective technique, this study aims to fill this critical knowledge gap by rigorously assessing and comparing the biomechanical properties of three distinct surgical interventions, including anterior controllable antedisplacement and fusion (ACAF), anterior cervical corpectomy decompression and fusion (ACCF), and anterior cervical discectomy and fusion (ACDF). The study offers pivotal insights to enhance treatment precision and patient outcomes.

Methods

The construction of the cervical spine model involved a detailed process using CT data, specialized software (Mimics, Geomagic Studio, and Hypermesh) and material properties obtained from prior studies. Surgical instruments were modeled (titanium mesh, anterior cervical plate, interbody cage, and self‐tapping screws) to simulate three surgical approaches: ACAF, ACCF, and ACDF, each with specific procedures replicating clinical protocols. A 75‐N follower load with 2 Nm was applied to simulate biomechanical effects.

Results

The range of motion decreased more after surgery for ACAF and ACDF than for ACCF, especially in flexion and lateral bending. ACCF have higher stress peaks in the fixation system than those of ACAF and ACDF, especially in flexion. The maximum von Mises stresses of the bone–screw interfaces at C3 of ACCF were higher than those of ACAF and ACDF. The maximum von Mises stresses of the bone–screw interfaces at C6 of ACDF were much higher than those of ACAF and ACCF. The maximum von Mises stresses of the grafts of ACCF and ACAF were much higher than those of ACDF. The maximum von Mises stresses of the endplate of ACCF were much higher than those of ACAF and ACDF.

Conclusion

The ACAF and ACDF models demonstrated superior cervical reconstruction stability over the ACCF model. ACAF exhibited lower risks of internal fixation failure and cage subsidence compared to ACCF, making it a promising approach. However, while ACAF revealed improved stability over ACCF, higher rates of subsidence and internal fixation failure persisted compared to ACDF, suggesting the need for further exploration of ACAF's long‐term efficacy and potential improvements in clinical outcomes.

Our finite element analysis research revealed that anterior controllable antedisplacement and fusion (ACAF) can preserve a similar spinal stability to anterior cervical discectomy and fusion (ACDF), which is far more satisfying than anterior cervical corpectomy decompression and fusion (ACCF). ACAF also showed a lower risk of internal fixation failure and cage subsidence than ACCF, while higher rates of subsidence, screw loosening and internal fixation failure might still be detected when compared with ACDF, especially in axial rotation.

Introduction

Multilevel cervical spondylotic myelopathy (CSM) is a formidable challenge in clinical management, predominantly affecting individuals aged 50 and above, causing spinal cord compression and consequent neurological impairment. ¹ Surgical decompression has emerged as an effective recourse, particularly through anterior approaches, for addressing persistent CSM. ² However, controversies persist regarding the optimal anterior decompression technique for multilevel CSM, signifying the pressing need to clarify the most efficacious approach among anterior cervical discectomy and fusion (ACDF), anterior cervical corpectomy decompression and fusion (ACCF), and the novel anterior controllable antedisplacement and fusion (ACAF). ² , ³

Although ACDF is known to preserve spinal stability with a lower incidence of fixation system‐related complications, it has drawbacks, such as potential incomplete decompression and increased risk of pseudarthrosis. ² , ³ , ⁴ , ⁵ In contrast, ACCF offers extensive decompression but is associated with a higher incidence of fixation system‐related complications. ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ Firstly introduced in 2017, ACAF presents a promising alternative, uniquely hoisting ventral compressive lesions to attain adequate decompression without resection, and promising favorable short‐term clinical outcomes with minimal complications. ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ However, its long‐term stability and complications remain elusive, necessitating comprehensive biomechanical evaluations against established techniques.

The distinctive biomechanical features of ACAF, stemming from its fusion mode and similarities with ACDF in vertebral body attachment, separate it from both ACCF and ACDF approaches. Notably, the detachment of the affected segment's vertebral body following the creation of bilateral grooves in ACAF mirrors aspects of ACCF. This inherent variability underlines the urgency for a comparative analysis of ACAF, ACCF, and ACDF through finite element analysis (FEA) to discern their biomechanical nuances, predict long‐term complications, assess internal fixation risks, refine ACAF surgical techniques, and inform judicious clinical decision‐making.

This study thus aims to bridge critical knowledge gaps by undertaking a comprehensive biomechanical comparison of these anterior decompression techniques for multilevel CSM, aiming to elucidate their distinct biomechanical attributes and anticipate their clinical implications. The findings are anticipated to enrich our understanding of these surgical modalities, potentially enhancing treatment precision and optimizing patient outcomes.

Methods

Construction of the Cervical Spine Model and Instruments

A three‐dimensional (3D) finite element (FE) model of the C2–C7 cervical vertebrae was constructed based on the computed tomography (CT) data from a 30‐year‐old healthy male volunteer, with a height of 175 cm and weight of 75 kg. The CT scan data with a space interval of 0.625 mm was obtained using a CT scanner (SOMATOM Definition Flash, Siemens, Germany) and saved in (Digital Imaging and Communications in Medicine) Dicom format. After CT scan data was imported into Mimics 10.1 software (Materialize, Belgium), a 3D geometric model of the cervical spine (C2–C7) was established initially by setting a thresholding of bone (226‐2998 Hounsfield Unit [HU]). Next, the rough 3D model was further denoised, smoothed, and filled to transform into a Non‐Uniform Rational B‐Splines (NURBS) 3D solid model in Geomagic Studio 2013 software (Raindrop, USA). Then, the model was imported into Hypermesh 2017 software for FE mesh generation (Altair, USA). Finally, boundary conditions were defined, and the FEA was constructed using Abaqus 2016 software (Simulia, USA).

The FE model included the cortical bone, cancellous bone, end plates, annulus fibrosus, nucleus pulposus, and seven major groups of ligaments (the anterior longitudinal ligament [ALL], posterior longitudinal ligament [PLL], interspinous ligament [IL], intertransverse ligament [ITL], supraspinous ligament [SL], capsular ligament [CL], and ligamentum flavum [LF]). In this FE model, the thickness of the cortical bone and end plates was 0.5 mm on average. ¹⁹ In the intervertebral disc, the volume ratio of the nucleus pulposus to the annulus fibrosus was 4:6. ²⁰ Additionally, annulus fibers accounted for approximately 20% of the annulus fibrosus volume and were set as tension‐only linear contact elements with an orientation of 25° relative to the horizontal plane. ²¹ Seven major ligaments were attached to the corresponding vertebrae as tension‐only linear contact elements. The facet joint gap was set as 0.5 mm, overlaid with an articular cartilage layer. The relationship between the superior and inferior facet joint surfaces was defined as frictionless contact. ²⁰ The finite element model of the intact cervical spine is shown in Figure 1.

FIGURE 1.

Finite element of the intact cervical spine (C2–C7). (A) Anterior view. (B) Posterior view. (C) Lateral view.

According to the actual measured size, the FE models of the surgical instruments, which encompassed the titanium mesh, anterior cervical plate, interbody cage, and self‐tapping screws, were developed using ProE Wildfire 5.0 software (PTC, USA). The screw–plate interface was set as shared nodes with no relative motion. The bone graft–vertebrae and screw–vertebrae interfaces were set as tie constraints. ²⁰ The bone graft–cage and bone graft–mesh interfaces were set as face–face sliding contact with a friction coefficient of 0.07, while the friction coefficient of the cage–vertebrae and mesh–vertebrae interfaces was set as 0.3. ²² The material properties used to reconstruct the FE model are listed in Table 1 and were obtained from previous studies. ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ The number of elements and nodes are shown in Table 2.

TABLE 1.

Material properties of the spinal components

Terms	Yong modulus (MPa)	Poisson ration	Element type	Cross section (mm2)
Cortical bone	12,000	0.29	C3D4H	‐
Cancellous bone	450	0.29	C3D4H	‐
Endplate	500	0.4	C3D4H	‐
Facet joint cartilage	10.4	0.4	C3D4H
Annulus fibrosus	4	0.4	C3D4H	‐
Nucleus pulposus	1	0.475	C3D4H	‐
Anterior cervical plate	110,000	0.3	C3D10M	‐
Titanium mesh	110,000	0.3	C3D8R	‐
Screw	110,000	0.3	C3D10M	‐
Interbody cage	3400	0.3	C3D8R	‐
ALL	7.8	0.3	T3D2 (truss)	6.0
PLL	10	0.3	T3D2 (truss)	5.0
LF	15	0.3	T3D2 (truss)	5.0
SL	8	0.3	T3D2 (truss)	5.0
IL	10	0.3	T3D2 (truss)	10.0
CL	8	0.3	T3D2 (truss)	46.0
ITL	10	0.3	T3D2 (truss)	2.0

Abbreviations: ALL, anterior longitudinal ligament; CL, capsular ligament; IL, interspinous ligament; ITL, intertransverse ligament; LF, ligamentum flavum;; PLL, posterior longitudinal ligament SL, supraspinous ligament.

TABLE 2.

The number elements and nodes of cervical spine model

	Element	Node
C2	119,633	23,860
C3	176,424	33,172
C4	150,712	29,140
C5	143,732	27,766
C6	155,852	30,460
C7	123,172	24,752
C2/3	7285	1704
C3/4	10,488	2189
C4/5	12,465	2586
C5/6	14,153	2926
C6/7	10,366	2226
ALL	10	5
PLL	10	5
LF	38	19
IL	10	5
SL	10	5
CL	20	10
ITL	80	40

Abbreviations: ALL, anterior longitudinal ligament; CL, capsular ligament; IL, interspinous ligament; ITL, intertransverse ligament; LF, ligamentum flavum; PLL, posterior longitudinal ligament; SL, supraspinous ligament.

Intact Model Validation and Surgical Approach Simulation

The intact FE model of the C2–C7 cervical vertebrae was fixed at the inferior endplate of the caudal vertebrae. A compressive follower load of 75 N (imitate muscle force and head weight) and a pure moment of 2.0 Nm (sagittal, transverse, and frontal planes) were applied to the superior surface of the cephalic vertebrae to simulate flexion, extension, lateral bending, and axial rotation. The plane motion of each segment was compared against the data published in previous studies on biomechanics and FEA to validate the model. ²⁶ , ²⁷ , ²⁸

According to the standardized clinical surgical procedures, the following three surgical models were processed and constructed based on the intact FE model. ⁵ , ¹² , ²⁹ , ³⁰ In our FE model, the length of the self‐tapping screw was 16 mm, with the external and inner thread diameters of φ4.5 and φ2.5 mm, respectively. According to the number of the designed vertebrae, the anterior cervical plate was set as 59.2, 16, and 2 mm, respectively, in length, width, and thickness. Additionally, two well‐known instruments, the titanium mesh and the interbody cage, were also used in our study. The dimensions of the interbody cage, including transverse diameter, anteroposterior diameter, height, and diameter of the inner bore, were set at 15, 12.5, 6.5, and 7.5 mm, respectively. The height of the titanium mesh was 22.5 mm, and the inner and external diameters were 10 and 12 mm. Given C4 and C5 vertebrae are the most involved levels in clinical practice, they were selected as surgical levels in our study. ²² The finite element of the three surgical models are shown in Figure 2.

FIGURE 2.

Finite element of the three surgical models. (A) ACAF model: (a) exploded view, (b) anterior view, and (c) oblique view. (B) ACCF model: (a) exploded view, (b) anterior view, and (c) oblique view. (C) ACDF model: (a) exploded view, (b) anterior view, and (c) oblique view. ACAF, anterior controllable antedisplacement and fusion; ACCF, anterior cervical corpectomy decompression and fusion; ACDF, anterior cervical discectomy and fusion.

Two‐Level Anterior Controllable Antedisplacement and Fusion (Two Vertebrae Bodies Were Hoisted and Three Intervertebral Discs Were Replaced with Cages)

First, the intervertebral discs at C3/C4, C4/C5, and C5/C6, ALL from C3 to C6, and PLL at C3/C4 and C5/C6 were removed, while the PLL behind C4 and C5 including C4/5 were reserved. According to the findings in our previous study, an average of 3 mm thickness of the anterior portions of C4 and C5 vertebrae bodies were resected in this study. Second, three separate interbody cages filled with bone graft were inserted into corresponding intervertebral space. The anterior cervical plate was fixed at C3 and C6 vertebrae using four self‐tapping screws. Third, bilateral grooves were created with the thickness of 3 mm in the medial of uncinate process. Finally, the C4 and C5 vertebral bodies were pulled ventrally tightening the middle four self‐trapping screws.

Two‐Level Anterior Cervical Corpectomy Decompression and Fusion (Two Vertebrae Bodies and Three Intervertebral Discs Were Replaced by a Titanium Mesh)

First, the C4 and C5 vertebrae bodies and intervertebral discs at C3/C4, C4/C5, and C5/C6 were removed. Additionally, the ALL from C3 to C6 together with PLL from C3/C4 to C5/C6 were removed. Second, a titanium mesh with bone graft was placed centrally between the C3 inferior and C6 superior end plates. Finally, the anterior cervical plate was fixed at C3 and C6 vertebrae using four self‐tapping screws.

Three‐Level Anterior Cervical Discectomy and Fusion (Three Intervertebral Discs Were Removed and Replaced with Cages)

First, the intervertebral discs at C3/C4, C4/C5, and C5/C6 level were removed. ALL from C3–C6 and PLL corresponding to C3/C4, C4/C5, and C5/C6 were removed as well. Second, three separate interbody cages filled with bone graft were inserted into the corresponding intervertebral space. Finally, the anterior cervical plate was fixed at C3–C6 vertebrae using eight self‐tapping screws.

Outcome Measures

The range of motion (ROM) refers to the degree of movement occurring within segments of the spine. It includes movements such as flexion, extension, lateral bending, and axial rotation. ROMs were measured using FEA within the constructed cervical spine models. These models simulated various movements to calculate the degrees of motion at specific vertebral levels. ROMs were quantified in degrees for different movements (flexion, extension, lateral bending, and axial rotation) at designated levels. The measurements were compared against established data from previous biomechanical studies to validate the accuracy and consistency of the models.

Von Mises stress is a scalar quantity representing the magnitude of stress experienced by a material under complex loading conditions. The maximum von Mises stress signifies the peak stress within a structure or component. These stresses were calculated using FEA, which applies boundary conditions and loads to the finite element models to simulate mechanical behaviors. The maximum von Mises stresses were measured in megaPascals (MPa) or similar units. They were compared across different surgical models (ACAF, ACDF, and ACCF) and specific components (screws, plates, and bone–screw interfaces) to assess stress levels under various loading conditions.

Stress distribution refers to the spatial pattern of stresses within a structure or component, demonstrating how stress is distributed across the analyzed area. Stress distribution was visualized using color‐coded maps generated through FEA. These maps illustrate stress patterns and concentration areas within the cervical spine models. The distribution patterns were assessed to identify regions of stress concentration, stress dispersion, or stress dissipation within the different surgical models. The concentration and dispersion of stress were analyzed in relation to the anatomical structures (bone–screw interfaces, endplates, grafts) and the fixation systems (screws, plates) to understand stress propagation and potential failure points.

Results

Validation of the Intact Cervical Spine Model

The ROMs of the intact cervical spine model for flexion at C3/C4, C4/C5, and C5/C6 were 6.8°, 7.3°, and 7.4°, respectively. The ROMs for extension at C3/C4, C4/C5, and C5/C6 were 5.2°, 5.8°, and 5.3°, respectively. The ROMs for left–right lateral bending at C3/C4, C4/C5, and C5/C6 were 5.1°, 7.6°, and 6.2°, respectively. The ROMs for left–right axial rotation at C3/C4, C4/C5, and C5/C6 were 5.3°, 6.1°, and 5.8°, respectively. As shown in Figure 3, the predicted ROMs of the present study were within the range of the data published by previous literature about biomechanics and FEA.

FIGURE 3.

The comparison of the predicted ranges of motion (ROMs) between the intact model and previously published data. (A) The ROMs of the cervical spine for flexion at C3/C4, C4/C5, and C5/C6. (B) The ROMs of the cervical spine for extension at C3/C4, C4/C5, and C5/C6. (C) The ROMs of the cervical spine for lateral bending at C3/C4, C4/C5, and C5/C6. (D) The ROMs of the cervical spine for axial rotation at C3/C4, C4/C5, and C5/C6.

Range of Motion of the Three Surgical Models

As shown in Figure 4, the ROMs of ACAF were higher than that of ACDF in flexion (25%), extension (100%), and lateral bending (33.3%) but showed no difference in axial rotation. However, compared with ACCF, the ROMs of ACAF were lower in flexion (−50%), extension (−50%), lateral bending (−42.9%), and axial rotation (−66.6%). The detailed ROMs of three surgical models are shown in Supplementary Material 1.

FIGURE 4.

The comparison of the range of motion of the surgical segments among the three surgical models. ACAF, anterior controllable antedisplacement and fusion; ACCF, anterior cervical corpectomy decompression and fusion; ACDF, anterior cervical discectomy and fusion.

Fixation System Stresses

Figure 5 demonstrates the maximum von Mises stresses and stress distribution of the fixation systems. The detailed maximum von Mises stresses of the fixation systems are shown in Supplementary Material 2. In the comparison of the three groups, the maximum von Mises stress of the screws of ACAF was higher than that of ACDF but lower than ACCF in flexion (13.7%, −52.7%), extension (168.0%, −37.0%), lateral bending (14.2%, −46.6%), and axial rotation (100.0%, −42.5%). Similarly, the maximum von Mises stress of the cervical plate of ACAF was also higher compared with ACDF but lower compared with ACCF in flexion (14.1%, −46.2%), extension (24.8%, −53.9%), lateral bending (94.5%, −22.4%), and axial rotation (65.9%, −34.8%). The stresses in the ACCF model were concentrated mainly at the cranial and caudal screw–plate interfaces in all load types. Analogously, the ACAF model also presented a stress concentration at the screw–plate interface. In contrast, the stress distribution of the whole internal fixation in the ACAF model was more homogeneous compared with ACCF. In contrast, the stresses in the ACDF model are relatively concentrated in the shrinkage of the titanium plate instead of the screw–plate interface.

FIGURE 5.

The comparison of the maximum von Mises stresses and stress distribution of the fixation systems among the three surgical models. (A) Maximum von Mises stresses of the plate. (B) Maximum von Mises stresses of the screw. (C) Stress distribution on the fixation systems. ACAF, anterior controllable antedisplacement and fusion; ACCF, anterior cervical corpectomy decompression and fusion; ACDF, anterior cervical discectomy and fusion.

Bone–Screw Interfacial Stresses

Figure 6 demonstrated the maximum von Mises stresses and stress distribution of the bone–screw interfaces. The detailed maximum von Mises stresses of the bone–screw interfaces are shown in Supplementary Material 3. The stress of the bone–screw interface was dependent on the type of surgery but showed a diverse trend on different levels. Stresses of the bone–screw interface of the ACAF model at C3 were lower than the stresses in the ACCF model: flexion (−54.2%), extension (−46.1%), lateral bending (−46.2%), and axial rotation (−34.6%). Compared with ACDF, the C3 bone–screw interface stresses were higher in the ACAF model: flexion (76.8%), extension (73.3%), lateral flexion (140.6%), and axial rotation (99.4%). At C6, the maximum von Mises stresses of the bone–screw interface of ACAF were still lower than that of ACCF during all load types and were still higher than ACDF for all load types except lateral bending: flexion (−27.9%, 113.7%), extension (−38.7%, 70.6%), lateral bending (−47.9%, −16.8%), and axial rotation (−9.9%, 15.1%). Both the cranial and caudal bone–screw interfaces in ACCF sustained higher stresses than those in ACAF, while the bone–screw interfaces in ACDF sustained relatively lower stresses.

FIGURE 6.

The comparison of the maximum von Mises stresses and stress distribution of the bone–screw interfaces among the three surgical models. Maximum von Mises stresses of the bone–screw interfaces at C3 (A) and C6 (B). Stress distribution on the bone–screw interfaces at C3 (C) and C6 (D). ACAF, anterior controllable antedisplacement and fusion; ACCF, anterior cervical corpectomy decompression and fusion; ACDF, anterior cervical discectomy and fusion.

Endplate Stresses

Figure 7 demonstrated the maximum von Mises stresses and the stress distribution of the endplates. The maximum von Mises stress of ACCF on the C3 and C6 endplate was observed for all load types (flexion, extension, lateral bending, and axial rotation) (Supplementary Material 4), and the results shown by the FE models of ACAF were lower than those of ACCF in extension (−18.6%, −28.7%), lateral bending (−89.5%, −41.7%), and axial rotation (−70.0%, −50.8%). However, the endplate stresses of ACAF and ACCF differed slightly in flexion at C3 (3.5 Mpa, 3.4 Mpa), while the endplate stress of ACAF was still lower than ACCF at C6 in flexion (−34.6%). Moreover, opposite results were observed when it came to the comparison between ACAF and ACDF. The endplate stresses of ACAF were higher than those of ACDF in flexion (34.6%, 6.3%) and extension (141.4%, 41.4%) at the C3 and C6 level. Interestingly, the endplate stresses of ACAF were lower at C3 in lateral bending and axial rotation (−67.9%, −36.4%) and C6 in axial rotation (−41.0%) when compared with ACDF, but ACAF showed higher endplate stress in lateral bending (82.6%) at C6. The stress concentration in ACCF was more obvious compared with the other two models, which generally concentrated on the interface between the titanium mesh and the endplate, and the stress concentration in flexion of C6 was the most significant. However, there was no obvious stress concentration on endplates in ACAF and ACDF.

FIGURE 7.

The comparison of the maximum von Mises stresses and stress distribution of the endplates among the three surgical models. Maximum von Mises stresses of the endplates at C3 (A) and C6 (B). Stress distribution on the endplates at C3 (C) and C6 (D). ACAF, anterior controllable antedisplacement and fusion; ACCF, anterior cervical corpectomy decompression and fusion; ACDF, anterior cervical discectomy and fusion.

Graft Stresses

Figure 8 demonstrated the maximum von Mises stresses and the stress distribution of the grafts. The detailed maximum von Mises stresses of the grafts are shown in Supplementary Material 5. The maximum von Mises stresses of the grafts of ACAF were much higher than those of ACDF in flexion (30.9%), extension (50.7%), and lateral bending (117.3%), especially in axial rotation (242.9%). The maximum von Mises stresses of the grafts of ACAF were lower than those of ACCF in flexion (−71.7%), extension (−41.2%), lateral bending (−41.5%), and axial rotation (−41.9%). In all directions, the concentrations of stresses were observed on the interfaces between the grafts and the endplates in both ACAF and ACDF models, but the overall stress distribution was uniform. However, stresses in ACCF in flexion and axial rotation were concentrated at the cranial and caudal side of the titanium mesh, but the stresses were concentrated at one side of the titanium mesh body in extension and lateral bending.

FIGURE 8.

The comparison of the maximum von Mises stresses and stress distribution of the grafts among the three surgical models. (A) Maximum von Mises stresses of the grafts. (B) Stress distribution on the grafts. ACAF, anterior controllable antedisplacement and fusion; ACCF, anterior cervical corpectomy decompression and fusion; ACDF, anterior cervical discectomy and fusion.

Discussion

This study aimed to comprehensively assess and compare the biomechanical properties of three surgical interventions—ACAF, ACCF, and ACDF—for multilevel CSM. The results indicated that ACAF exhibited higher ranges of ROMs in flexion, extension, and lateral bending compared to ACDF while displaying lower ROMs than ACCF across all motion types. Furthermore, stress analysis revealed distinct patterns, showcasing differences in stress distribution among the fixation systems, bone–screw interfaces, endplates, and grafts across the three surgical models. ACAF demonstrated favorable stability compared to ACCF and distinct stress concentration patterns different from ACDF. The clinical significance lies in highlighting ACAF's potential as a promising approach with improved stability and reduced fixation failure risks, yet it necessitates further investigation for long‐term efficacy and outcome enhancements.

Spinal Stability

The postoperative stability of the cervical spine plays a crucial role in determining clinical outcomes and cervical secondary changes, rendering it a vital criterion for evaluating the effectiveness of various cervical surgery. As mentioned above, in the treatment of multilevel CSM, ACDF and ACCF have their own disadvantages. Previous clinical studies and human specimen biomechanical experiments have also proved that compared with ACDF, ACCF had poorer postoperative stability. ⁶ , ¹⁰ , ³¹ , ³² In particular, our previous study demonstrated that the postoperative ROMs at the C3–4 level in the ACCF group exceeded those in ACDF or ACAF groups in all directions. At the C4–5 and C5–6 levels, the postoperative flexion and extension angles in the ACCF group were significantly larger than those in the ACDF and ACAF groups. ³² These results indicated that patients who underwent ACCF experienced lower spinal stability compared to those who underwent ACDF or ACAF.

To prevent the incidence of postoperative complications in multilevel CSM, hybrid decompression and fusion (HDF) was proposed to provide an alternative option. ³¹ , ³³ Moreover, it was verified through biomechanical experiments that the postoperative stability of HDF was better than that of multisegment ACCF. ³¹ , ³⁴ It is worth noting that the design of ACAF is similar to that of HDF to some extent, both of which preserve the cervical vertebral structure as much as possible, thereby improving the inherent mechanical structural stability of the cervical spine. Therefore, ACAF could theoretically obtain better postoperative stability than ACCF. After the comparison of biomechanical stability among two‐level ACAF, two‐level ACCF, and three‐level ACDF, the decrease in surgical segmental ROMs suggested that no apparent difference in ROMs of surgical segments was found between ACAF and ACDF. Similarly, in the study by Kong et al. (2022) on human cadaveric cervical spines, no significant difference was found between the ACDF and the ACAF groups for the ROMs in all directions. ³² Compared with ACCF, ACAF and ACDF decreased more ROMs after surgery in extension, lateral bending, and axial rotation. These results indicated that ACAF can obtain similar postoperative stability to ACDF and was significantly better than ACCF, which was consistent with the conclusions of previous clinical studies and biomechanical experiments and confirmed our biomechanical hypothesis about the postoperative stability of ACAF.

Fixation System‐Related Risks

Due to the elevated elastic modulus of the titanium mesh, ACCF has much higher stiffness and sustains more stress of the load. ²⁰ Elevated stress in the screws and cervical plates can compromise the long‐term stability of ACCF's construct, thereby elevating the risk of internal fixation failure, such as plate migration, dislodgment, or screw loosening. Complications associated with internal fixation failure have been reported in previous studies. ²⁴ However, in the clinical studies on ACAF, no instances of screw loosening have been identified. As our results show, the maximum von Mises stress of cervical plates and screws of ACAF and ACDF in all directions were less than ACCF, which indicated that they had a lower risk of internal fixation failure than ACCF. This similarity might be attributed to the reconstructed structure of ACAF, which mirrors ACDF in terms of internal fixation. Each segment is fixed with screws. With the same fixed length, the more segments of screw fixation can effectively disperse the stress of the internal fixation system. The screws can evenly distribute the stress on each vertebral body to avoid stress concentration. Given that multisegment screw placement in ACAF serves as a means to distribute stresses, patients undergoing ACAF face a diminished risk of subsidence and internal fixation failure than ACCF.

However, the screw–plate interfaces and bone–screw interfaces at the cranial side of ACAF exhibited greater stresses than ACDF, particularly during rotation and lateral bending. This suggests that the likelihood of screw dislodgment and loosening might be higher in ACAF than in ACDF. This might be due to the most structure of the bilateral uncovertebral joints were removed during ACAF surgery, which played an important role in restricting rotation of cervical spine. This study reminds us to pay greater attention to the risk of loosening of the cranial screws after ACAF, especially to limit the postoperative rotation of the cervical spine. Therefore, selecting a more rigid external fixation brace that can restrict the movement of the cervical spine after long‐segment ACAF is recommended.

Cage Subsidence

Cage subsidence, defined as the movement of the titanium mesh or interbody cage into the adjacent vertebral body, has been commonly documented in cases of ACDF and ACCF. ³⁵ , ³⁶ Wu et al. (2014) reported that the occurrence of cage subsidence is closely related to endplate conditions like loaded stress and the degree of osteoporosis. ³⁷ According to DiAngelo et al. (2000), ³⁸ the ultimate cyclic compression load in adults was 200–225 N, and overlong bone graft may surpass this threshold, leading to excessive cyclic compression. This could potentially cause fractures in the endplate, subsidence of the cage, screw loosening, and displacement of the titanium plate. Previous biomechanical studies have reported that ACCF might induce much higher stress on the inferior and superior endplates compared to ACDF. ⁵ , ³⁹ The contact area (point‐to‐point contact) between the titanium mesh cage and the endplate is small, and this characteristic might also be one of the reasons for the concentrated stress, ⁴⁰ which is closely associated with subsidence and other complications related to instruments. In clinical studies by Noordhoek and Zhang, the mean rate of cage subsidence after ACCF and ACDF was 80% and 21.2%, respectively. ⁴¹ The results of this study indicate that the stresses of the inferior and superior endplates of ACCF were greater than those in ACDF and ACAF, especially in the flexion and extension positions, suggesting that the risks of cage subsidence in ACDF and ACAF were lower than that of ACCF. This conclusion is consistent with the above clinical and biomechanical studies. Compared with ACAF and ACDF, the higher endplate stress of ACCF might be caused by the difference in the reconstructed anterior column. The reconstructed anterior column of ACAF and ACDF is consist of three interbody cages and C4 and C5 vertebral bodies, which increased the graft–endplate interface and decreased load sharing, while that of ACCF was only a titanium mesh. In addition, cage subsidence has not been reported in clinical studies with short‐term follow‐up on the ACAF technique. ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ This suggests that ACAF might yield superior outcomes in preventing cage subsidence.

However, the findings of this study also revealed that in extension and lateral bending, ACAF exhibited higher maximum stress at the caudal endplate compared to ACDF, suggesting that the risk of subsidence of the internal body in ACAF still existed. As mentioned above, bilateral grooves were created at the medial uncinate process, making the intermediate vertebral body dissociated, which led to the stress load transferring to the internal fixation system and internal body, resulting in increased stress at the contact interface between the internal body and endplate. However, as shown in this study, the maximum stress values of ACAF and ACDF endplates were significantly lower than the previously mentioned ultimate cyclic compression load values (200–225 N). Additionally, uniform distribution of stresses in both surgical types indicates a low risk of subsidence for both surgical methods. It is worth noting that although the present follow‐up results of ACAF showed no transposition of the interbody cage, the results of this FEA highlight that we should pay close attention to the subsidence of the caudal interbody cage in future follow‐up studies of ACAF.

Bone Fusion

In anterior decompression and fusion surgery, the short‐term spinal stability is obtained via the fixation system, while the long‐term spinal stability relies more on effective bone fusion. Moreover, effective bone fusion serves as a crucial clinical efficacy indicator following fusion surgery. The appropriate graft stress is necessary for effective bone fusion, while inadequate graft stress will lead to bone resorption. ⁴² Previous studies ⁴³ , ⁴⁴ have reported that the fusion rate of ACCF was much higher than that of ACDF, which was consistent with our results. In multilevel CSM, an increased number of grafts and interfaces might cause the inferiority of ACDF in fusion. ⁶ Although the number of grafts and interfaces in ACAF matched that of ACDF, it showed sufficient and rapid bone fusion postoperatively in the present clinical observations. ¹⁴ , ¹⁸ The elevated fusion rate of ACAF can be attributed to the following factors. Compared with ACDF, bone union in ACAF would have occurred not only at multiple contact surfaces between vertebral bodies and the interbody spacers but also at lateral grooves in a vertical direction. More bone marrow exposure under sufficient stress stimulus by making lateral slits would enhance solid union. ⁵ , ⁴⁵ As indicated by our results, graft stresses in ACAF and ACCF surpassed those in ACDF. This might imply that ACAF was superior to ACDF in bone fusion. However, clinical studies with larger sample sizes and longer‐term follow‐up are necessary in the future to confirm this conclusion.

Limitations

Our study had some limitations. First, because all the FE models in this study were constructed using CT scans from a healthy individual with ideal material properties, our findings might not be directly applicable to patients with severe spinal pathologies and deformities, including ossification of the posterior longitudinal ligament (OPLL), ankylosing spondylitis, and acromegaly. Moreover, the presence of ossified PLL significantly affects cervical spine stability, making it challenging to fully simulate the stability of OPLL patients after ACAF with the current model. A cervical spine model with OPLL should be designed for further 3D finite element studies. Further, to reduce the computational load, we did not set the details of the vertebral attachments at the cranial and caudal of the FE model, which ignored the problem of adjacent segment degeneration (ASD). We will explore this problem in a future study. Finally, the bilateral osteotomy will be fully grafted in practical ACAF surgery, which will increase the friction between the dissociated vertebrae body and bilateral groove walls to improve the stability. After fusion, the stability and stress distribution of ACAF will be improved. However, these factors cannot be effectively simulated in a 3D finite element model. Therefore, this study cannot fully reflect the actual biomechanical characteristics of ACAF. Additionally, further cadaveric and follow‐up studies are needed to verify our results in the future due to differences between FEA and real‐life situations.

Conclusion

Our FEA research revealed that ACAF can preserve a similar spinal stability to ACDF, which is far more satisfactory than ACCF. ACAF also showed a lower risk of internal fixation failure and cage subsidence than ACCF, while higher rates of subsidence, screw loosening, and internal fixation failure might still be detected when compared with ACDF, especially in axial rotation. The above results demonstrate the effectiveness of the dispersion of fixation stress by multisegment implantation of screws in ACAF. However, restriction of postoperative movement in axial rotation should be paid additional attention to avoid the failure of internal fixation. This study provides sufficient information for predicting the long‐term complications and improvements in surgical technique.

Author Contributions

Qingjie Kong, Fudong Li, and Jiangang Shi led the conceptualization of the study, conducted investigations, and were instrumental in visualizing the data. Additionally, Qingjie Kong, Fudong Li, Chen Yan, Peidong Sun, Jun Ou‐Yang, and Shizhen Zhong were primarily responsible for constructing and analyzing the finite element models, contributing significantly to the technical aspects and analysis procedures. Jingchuan Sun conducted data curation and formal analysis, playing a crucial role in ensuring the accuracy and integrity of the data used for the study. Yuan Wang provided resources, supervised the project, and contributed to software management, ensuring the necessary tools and resources were available for the study's execution. Jiangang Shi played a pivotal role, contributing to conceptualization, providing resources, and supervising the study's progress. All authors contributed to reviewing and editing the manuscript, ensuring the accuracy, clarity, and coherence of the content presented.

Funding Information

This work was supported by the Shanghai Shenkang Hospital Development Center (SHDC2020CR1024B), the Shanghai Committee of Science and Technology (No. 23Y11903700), and the National Natural Science Foundation of China (No. 81802218).

Conflict of Interest Statement

There are no relevant conflicts of interest for this study.

Ethics Statement

This study was approved by the Ethics Committee of Shanghai Changzheng Hospital.

Supporting information

Supplementary Material.