Biomechanical effects of posterior lumbar interbody fusion with vertical placement of pedicle screws compared to traditional placement

Abstract

BACKGROUND

The pedicle screw technique is widely employed for vertebral body fixation in the treatment of spinal disorders. However, traditional screw placement methods require the dissection of paraspinal muscles and the insertion of pedicle screws at specific transverse section angles (TSA). Larger TSA angles require more force to pull the muscle tissue, which can increase the risk of surgical trauma and ischemic injury to the lumbar muscles.

AIM

To study the feasibility of zero-degree TSA vertical pedicle screw technique in the lumbosacral segment.

METHODS

Finite element models of vertebral bodies and pedicle screw-rod systems were established for the L4-S1 spinal segments. A standard axial load of 500 N and a rotational torque of 10 N/m were applied. Simulated screw pull-out experiment was conducted to observe pedicle screw resistance to pull-out, maximum stress, load-displacement ratio, maximum stress in vertebral bodies, load-displacement ratio in vertebral bodies, and the stress distribution in pedicle screws and vertebral bodies. Differences between the 0-degree and 17-degree TSA were compared.

RESULTS

At 0-degree TSA, the screw pull-out force decreased by 11.35% compared to that at 17-degree TSA (P < 0.05). At 0-degree and 17-degree TSA, the stress range in the screw-rod system was 335.1-657.5 MPa and 242.8-648.5 MPa, separately, which were below the fracture threshold for the screw-rod system (924 MPa). At 0-degree and 17-degree TSA, the stress range in the vertebral bodies was 68.45-78.91 MPa and 39.08-72.73 MPa, separately, which were below the typical bone yield stress range for vertebral bodies (110-125 MPa). At 0-degree TSA, the load-displacement ratio for the vertebral bodies and pedicle screws was slightly lower compared to that at 17-degree TSA, indicating slightly lower stability (P < 0.05).

CONCLUSION

The safety and stability of 0-degree TSA are slightly lower, but the risks of screw-rod system fracture, vertebral body fracture, and rupture are within acceptable limits.

Key Words: Vertical pedicle screw, Pedicle screw technique, Transverse section angle, Lumbosacral segment, Finite element analysis

Core Tip: This study aimed to explore the feasibility of employing zero-degree transverse section angles (TSA) vertical pedicle screw technique in the lumbosacral segment to reduce surgical trauma and ischemic injury to the lumbar muscles during pedicle screw insertion. The safety and stability of the zero-degree TSA vertical pedicle screw technique in the lumbosacral segment were slightly lower than those of the 17-degree TSA screw placement technique. However, the risks of screw-rod system fractures, vertebral body fractures, and ruptures were within acceptable limits. The screws exhibited high resistance to pull-out. The use of this zero-degree TSA vertical pedicle screw technique offers good safety and reliability. It can reduce surgical dissection of the multifidus muscles and minimize surgical trauma to the lumbar muscles.

INTRODUCTION

The pedicle screw technique utilizes the gripping force at the pedicle screw-bone interface to provide fixation of the anterior, middle, and posterior spinal columns, offering excellent stability and multidimensional corrective capability. Its safety and effectiveness have been demonstrated through extensive clinical practice and numerous studies, making it one of the most classic posterior spinal fixation techniques widely employed for vertebral body fixation in treating spinal disorders[1]. Presently, various mature screw placement methods have been developed for the pedicle screw technique, such as the Weinstein, Magerl, and Louis methods[2]. However, traditional screw placement methods require paraspinal muscle dissection and pedicle screw insertion at specific transverse section angles (TSA)[3]. Yet, in the lumbosacral segment, the paraspinal muscles, such as the lumbodorsal fascia and multifidus, exhibit richness and robustness. Larger TSA angles require greater force to pull the muscle tissue, which can increase the risk of surgical trauma and ischemic injury to the lumbar muscles. This can lead to chronic fibrosis and loss of function in the multifidus muscle, as well as postoperative complications such as chronic intractable lower back pain[4,5]. Some researchers have proposed changing the surgical approach and using minimally invasive tools to reduce damage to the lumbar muscles caused by the pedicle screw technique. Methods such as the Wiltse approach and minimally invasive MIS-TLIF can somewhat alleviate damage to the paraspinal muscles; however, outcomes have not been entirely satisfactory[6-8]. Damage to the lumbar muscles caused by the pedicle screw technique is primarily owing to the larger TSA, which consequently requires a larger screw placement field of view[9]. This study used thin-slice computed tomography (CT) data from 200 patients in our hospital, focusing on improving the pedicle screw placement technique in the lumbosacral segment by modifying the direction of screw placement. Biomechanical analysis was conducted using finite element modeling to investigate the feasibility of the zero-degree TSA vertical pedicle screw technique.

MATERIALS AND METHODS

Materials and equipment

Sample data: A sample of 200 patients with lumbar degenerative changes, aged 58–76 years with an average age of 65 years, was selected from our hospital. The study sample consisted of 97 males and 103 females. High-resolution CT scans were performed on the L4, L5, and S1 vertebral segments to obtain baseline data. Additionally, a healthy young volunteer was selected. Any abnormalities, such as lumbar spondylolisthesis, lumbar spondylolysis, lumbar degenerative changes, deformities, injuries, or a history of previous lumbar spine surgeries, were excluded. High-resolution CT scans were conducted on the L4–S1 vertebral segments to obtain baseline data.

Hardware equipment

A Siemens Dual-Source CT Scanner (SOMATOM Definition) was used in conjunction with a DELL T7820 advanced computing workstation with an Intel C621 chipset and an NVIDIA Quadro 1T professional graphics card.

Software environment

Mimics 21.0 edition (Materialize, Belgium), Geomagic 2022 edition (Geomagic, United States), Hyper Mesh 2022 edition (Altair, United States), Unigraphics NX10.0 edition (Siemens PLM Software, United States), Abaqus 2022 edition (Dassault Systemes Simulia, France), and ANSYS 2020 edition (ANSYS, United States) were used in this study.

Model creation

Vertebral body model: Healthy young volunteers were selected for this study. High-resolution CT scans of the L4 to S1 vertebral segments were obtained using a Siemens dual-source CT scanner (SOMATOM Definition) with a slice thickness of 1 mm and scan speed of 0.4 s per rotation. The acquired CT data were imported into Mimics software, and threshold-based segmentation was performed based on the grayscale values of various tissues to extract geometric models of different parts of the L4 to S1 vertebral bodies. These models were then exported in the STL format. The extracted vertebral geometric models in STL format were imported into the Geomagic software. Surface deformities, distortions, roughness, and other adverse structural issues in the extracted vertebral geometric models were addressed using surface fitting, mesh division, mesh quality improvement, smoothing, and noise reduction techniques. This process resulted in the creation of solid three-dimensional models of the lumbar and sacral vertebrae. The solid three-dimensional models were subsequently imported into the HyperMesh software, where C3D10M tetrahedral elements were selected to mesh the vertebral models. This allowed the creation of network models of structures such as intervertebral discs and vertebrae that met the requirements for finite element analysis (FEA). Subsequently, the InP network model was exported (Figure 1).

Figure 1 Finite element models of vertebral bodies and pedicle screws.

Pathway and pedicle screw models: Using DICOM-format data from thin-slice CT scans of 200 patients with lumbar degenerative changes from our hospital, we imported the data into the Mimics software. Subsequently, we individually reconstructed the morphology of the L4, L5, and S1 pedicle roots in each patient. Measurements were taken for the widest and narrowest transverse and sagittal diameters of the pedicle roots at L4, L5, and S1. Considering the structural characteristics of the lumbosacral region, we designed straight cylindrical channels with a diameter of 6.5 mm that were tangential to the inner walls of the pedicle roots. We measured the maximum screw insertion length, marked the entry points, and designed straight vertical pedicle screw channels at a zero-degree TSA.

Material properties and model conditions: This study involved lumbar vertebral models that did not include ligaments. The material properties for the lumbar vertebral and pedicle screw models were defined in Mimics, and the properties were set based on the previous literature and actual CT values (Table 1)[10-12].

Table 1 Material property settings.

Material name	Elastic modulus (MPa)	Poisson's ratio
Trabecular bone	200	0.2
Cortical bone	12000	0.3
Pedicle screw	3500	0.25
Intervertebral disc	3.4	0.39
Screw-rod system	120000	0.33

The simulation replicated the actual loading conditions of the lumbar vertebrae during forward bending, extension, and rotation. Forward bending and extension were aligned with the global X, Y, and Z coordinates, whereas rotation was aligned with the tangential direction of the lumbar curve. The vertebral bodies were constrained to the intervertebral discs, and pedicle screws were bound to the vertebral bodies. The lower part of the S1 vertebra was completely fixed. A reference point was established on the upper surface of the L4 vertebra, and the upper surface movements were coupled. A standard lumbar load of 500 N was applied (Figure 2)[13]. The simulation considered force conditions during pedicle screw extraction. The vertebral body was constrained, and a horizontal outward pull-out load was applied to the tail of the pedicle screw, incrementally increasing by 17.5 N with each step.

Figure 2 Diagram depicting application of loads during human forward bending, backward extension, rotation, and lateral bending. A: Extension; B: Forward bending; C: Right lateral bending; D: Left lateral bending; E: Right rotation; F: Left rotation.

Four scenarios were established with TSA of 0°, 7°, 10°, and 17° to simulate pedicle screw pull-out and resistance to the pull-out force was observed under different TSA. Additionally, two scenarios with TSA of 0° and 17° were set to observe the safety and stability of the pedicle screws and vertebral bodies under various conditions.

Main observational parameters

The primary observations in the experiment focused on five aspects: Pedicle screw pull-out force, pedicle screw safety, pedicle screw stability, vertebral body safety, and vertebral body stability.

(1) The pedicle screw pull-out force was evaluated by restraining the vertebral body and applying a horizontal outward pull-out load to the tail of the pedicle screw. The load was increased incrementally from 0 to 17.5 N each time. The maximum pull-out force of the pedicle screw was determined by analyzing the load vs pedicle screw head displacement curve. A sudden increase in screw head displacement indicates screw loosening, with the corresponding load value representing the maximum resistance to the pull-out force of the pedicle screw[14];

(2) Pedicle screw safety was assessed by applying a lumbar load of 500 N and an additional 10 N/m of rotational torque. The maximum stress and stress distributions in the pedicle screws were calculated. Based on previous studies, the threshold for rod system fracture was considered to be 924 MPa[15,16]. If the calculated stress reached or exceeded this value, it was considered a risk factor for rod system fracture, with higher stress values indicating a greater risk;

(3) Pedicle screw stability was evaluated under six different conditions: Simulated flexion, extension, left rotation, right rotation, left lateral bending, and right lateral bending. A 100 N load was applied in each direction, and the displacement of the pedicle screws was measured. These displacements were converted into load-displacement ratios to analyze the stability of the pedicle screws in each direction. A higher load-displacement ratio indicates better stability of the pedicle screws[17].

(4) The safety of the vertebral bodies was assessed by applying a 500 N axial load and a 10 N/m rotational torque. The maximum stress and stress distribution in the vertebral bodies were calculated. Based on previous studies, a permissible bone stress of 110-125 MPa was considered. If the calculated stress values reached or exceeded this range, it indicated a risk of bone fracture, with a higher proximity to this range indicating a greater risk[18];

and (5) Vertebral body stability was assessed by simulating six different conditions: Flexion, extension, left rotation, right rotation, left lateral bending, and right lateral bending. A 100 N load was applied in each condition, and the displacement of the vertebral bodies was calculated. This displacement was then converted into a load-displacement ratio to analyze the stability of the vertebral bodies in various directions. A higher load-displacement ratio indicated better stability of the vertebral bodies.

Statistical analysis

The experimental data were analyzed and processed using SPSS software (version 21.0). For normally distributed continuous data, the mean ± SD is used to represent the results. Analysis of variance was conducted to assess the differences in the vertebral pedicle screw pull-out force, pedicle screw safety, pedicle screw stability, vertebral body safety, and vertebral body stability among the four TSA. Statistical significance was set at P < 0.05.

RESULTS

Resistance to pedicle screw pull-out force

Table 2 presents the statistical results of the pedicle screw pull-out force under 0°, 7°, 10°, and 17° TSA, with the lowest pull-out force observed at 0° (P < 0.05). Stress in the vertebral bodies at different angles did not exhibit statistically significant differences among the groups.

Table 2 Pull-out strength of pedicle screws (unit: N).

TSA	Pull-out resistance	Maximum vertebral stress
0 degrees	132.8 ± 4.99	2.0058 ± 0.0031
7 degrees	139.6 ± 5.24	2.001 ± 0.0029
10 degrees	145.4 ± 6.09	2.0015 ± 0.0058
17 degrees	149.8 ± 6.72	2.0066 ± 0.0075
F	314.683	1.928
P value	0.001	0.182

TSA: Transverse section angle.

Safety of pedicle screws

Table 3 shows the test results for the maximum stress of pedicle screws at 0-degree and 17-degree TSA. In cases of extension and right bending, the maximum stress in the 0-degree TSA group was lower than that of the 17-degree group (P < 0.05). In cases of flexion, left bending, right rotation, and left rotation, the maximum stress in the 0-degree pedicle screw group was greater than that of the 17-degree group (P < 0.05).

Table 3 Maximum stress calculations for pedicle screws (unit: MPa).

TSA	Extension	Flexion	Right lateral bending	Left lateral bending	Right rotation	Left rotation
0 degrees	437.85 ± 11.72	335.1 ± 9.85	433.64 ± 10.15	612.31 ± 22.52	657.5 ± 44.84	582.2 ± 47.57
17 degrees	626.18 ± 44.45	242.8 ± 21.52	648.5 ± 83.31	488.62 ± 36.67	551.12 ± 24.01	548.4 ± 46.03
F	1313.129	196.715	1875.31	546.879	3557.963	1261.657
P value	0.000	0.000	0.000	0.000	0.000	0.000

TSA: Transverse section angle.

Stability of pedicle screws

Table 4 shows the load-displacement ratio values for pedicle screws at 0-degree and 17-degree TSA. In flexion (forward bending) and extension (backward bending), the load-displacement ratio for 0-degree TSA pedicle screws was greater than that for 17-degree TSA pedicle screws (P < 0.05). In right lateral bending, left lateral bending, right rotation, and left rotation, the load-displacement ratio for 0-degree TSA pedicle screws was smaller than that for 17-degree TSA pedicle screws (P < 0.05).

Table 4 Load-displacement ratio for pedicle screws (unit: N/mm).

TSA	Extension	Flexion	Right lateral bending	Left lateral bending	Right rotation	Left rotation
0 degrees	2740.93 ± 165.32	392.80 ± 22.71	483.01 ± 29.98	394.84 ± 20.03	351.67 ± 20.84	396.13 ± 23.69
17 degrees	2165.44 ± 131.12	353.29 ± 20.36	527.13 ± 33.77	425.08 ± 24.35	364.43 ± 19.14	428.47 ± 27.76
F	124.963	33.72	72.107	6.14	4.986	7.093
P value	0.000	0.000	0.000	0.011	0.028	0.001

TSA: Transverse section angle.

Safety of the vertebral body

Table 5 presents the simulation test results of maximum stress in the vertebral body at 0-degree and 17-degree TSA. During simulations of human bending forward, extending backward, bending sideways, and rotating, the maximum stress in the vertebral body at 0-degree TSA was consistently higher than that at 17 degrees (P < 0.05).

Table 5 Maximum stress in the vertebral body (unit: MPa).

TSA	Extension	Flexion	Right lateral bending	Left lateral bending	Right rotation	Left rotation
0 degrees	68.585 ± 4.41	77.245 ± 5.21	69.775 ± 4.19	78.91 ± 6.05	72.192 ± 4.22	68.45 ± 4.09
17 degrees	39.079 ± 2.40	55.681 ± 3.89	65.348 ± 4.09	72.734 ± 5.02	66.227 ± 4.31	45.756 ± 3.19
F	628.454	197.455	78.021	48.819	14.763	368.721
P value	0.000	0.000	0.000	0.000	0.000	0.000

TSA: Transverse section angle.

Stability of the vertebral body

Table 6 shows the simulation results of vertebral body load displacement ratios at 0-degrees and 17-degree TSA. In extension, the vertebral body load displacement ratio at 0 degrees was greater than that at 17 degrees (P < 0.05). In flexion, right lateral bending, left lateral bending, right rotation, and left rotation, the vertebral body load displacement ratio at 0-degree TSA was smaller than that at 17 degrees (P < 0.05).

Table 6 Vertebral load displacement ratio (unit: N/mm).

TSA	Extension	Flexion	Right lateral bending	Left lateral bending	Right rotation	Left rotation
0 degrees	1827.29 ± 114.78	261.87 ± 16.01	318.39 ± 20.71	307.10 ± 21.09	273.52 ± 18.13	264.10 ± 17.33
17 degrees	1443.65 ± 83.75	274.79 ± 18.53	344.18 ± 27.18	327.74 ± 23.35	283.45 ± 17.06	281.93 ± 21.99
F	134.605	6.144	229.409	97.584	3.775	8.134
P value	0.000	0.012	0.000	0.000	0.027	0.001

TSA: Transverse section angle.

Vertebral stress distribution

Figure 3 depicts the stress distribution in the vertebrae at 0-degree and 17-degree TSA. Under simulated conditions of human extension, flexion, rotation, and lateral bending, the stress in the vertebrae was primarily concentrated in the upper portions of L4, L5, and S1, with the highest stress occurring at the location of the vertebral arch, which represents a region of stress concentration. The stress distribution was somewhat similar under both TSA, generally showing that the stress on L5 was higher than on L4 and S1, except for left rotation at 0 degrees TSA angle, where the stress on L4 was higher than that on L5 and S1. However, there were significant differences in stress distribution between the two TSA. Under 0-degree TSA, the stress on L4 was consistently higher than that on S1, while under 17-degree TSA, the stress on S1 was consistently higher than that on L4. The stress differences between L4 and S1 were exactly the opposite under the two angles. Under both TSA, the maximum stress occurred in the L5 vertebral arch during left bending, with a maximum stress of 78.91 ± 6.05 for 0-degree TSA and 72.734 ± 5.02 for 17-degree TSA.

Figure 3 Vertebral stress distribution map. A: Vertebral stress distribution at 0-degree transverse section angle (TSA); B: Vertebral stress distribution at 17-degree TSA.

Screw-rod system stress distribution

Figure 4 shows the stress distribution of the rod system under 0-degree and 17-degree TSA during simulated human motions of extension, flexion, rotation, and lateral bending. There was a significant difference in stress distribution in the rod system between 0-degree and 17-degree TSA under these loading conditions. At 0-degree TSA, the maximum stress was primarily distributed at the head of the L4 pedicle screw and the rod connecting L5 to S1. Conversely, at 17-degree TSA, the maximum stress was mainly distributed at the head of the L5 pedicle screw and the rod connecting L4 to L5.

Figure 4 Rod system stress distribution. A: Stress distribution of the pedicle screw system at 0-degree transverse section angle (TSA); B: Stress distribution at 17-degree TSA.

DISCUSSION

Although pedicle screw techniques have significant advantages in terms of stability and corrective capability, surgical complications are common. To address postoperative chronic refractory lower back pain caused by surgical damage and ischemic injuries, some researchers have proposed techniques such as the Wiltse approach and minimally invasive MIS-TLIF with screw placement. However, in the lumbar and sacral regions, the lumbar back muscles are robust and layered at oblique angles without clear muscle gaps. When using the Wiltse approach, a significant force is required to stretch the lower back muscles. In contrast, when using the MIS-TLIF screw placement technique, the rigid channel is forcefully expanded through the lower back muscles. Consequently, both approaches involve muscle tissue detachment and prolonged intraoperative exposure time and do not significantly reduce damage to the lower back muscles[19]. To address this, this study aimed to reduce the TSA and explore the feasibility of vertically placing screws at a zero-degree TSA. This approach was intended to avoid significant muscle expansion, reduce extensive exposure of the facet joints, and minimize damage to the lower back muscles and other anatomical structures.

Some pedicle screw techniques for vertebral biomechanical experiments use cadaveric dissections or animal vertebrae as specimens. However, owing to variations in vertebral bone density, screw placement points, and screw angles, it is challenging to maintain consistency in the initial experimental conditions. This variability has led to certain controversies in the experimental results and conflicting findings[20-24]. Some researchers have used FEA for biomechanical studies. The stress within the vertebral and screw-rod systems can be calculated more objectively using finite element models that simulate the vertebral fixation system, its motion, and materials. This approach also allows for a visual representation of stress changes and distributions, which aids in identifying areas of stress concentration and potential fracture[25-28]. Therefore, drawing on previous research, this study used a substantial dataset of L4-S1 vertebral CT scans as a foundation for constructing finite element models of the vertebrae, accurately reproducing the morphology of the L4-S1 vertebrae. The experimental results closely matched data from studies conducted by scholars both domestically and internationally, confirming the validity and accuracy of the model.

Kılıçaslan et al[29], in their biomechanical study using pig lumbar vertebrae, found that the screw trajectory is a major factor influencing the biomechanical characteristics of pedicle screw techniques. Fatima et al[30] pointed out in their research that different screw trajectories lead to biomechanical differences primarily because of variations in bone density around the screw trajectory. This study found a significant difference in the pull-out resistance of the screws at different TSA. Specifically, at 0 degrees, 7 degrees, 10 degrees, and 17 degrees, the pull-out resistance values were 132.8 ± 4.99, 139.6 ± 5.24, 145.4 ± 6.09, and 149.8 ± 6.72, respectively. Compared with the mainstream 17-degree screws, the 0-degree TSA screws exhibited an 11.35% reduction in pull-out resistance, and compared with the 7-degree screws, there was a 4.87% reduction. This difference suggests that screws placed vertically at a zero-degree TSA may have a slightly higher risk of loosening or pulling out; however, they still exhibited a strong pull-out strength with an average value of 132.8 N.

Some studies have indicated that loosening or fracture of pedicle screws can result from excessively concentrated stress. Research has shown that stress transfer within the screw-rod system resembles a butterfly shape, in which stress is transmitted from the upper screws through the connecting rod to the lower screws. This pattern may be a major contributor to screw-rod system fractures[31-33]. Research indicates that the fracture threshold for a screw-rod system is 924 MPa, and reaching this threshold represents a very high risk of fracture[34]. This study found that when subjected to a conventional lumbar vertebral weight of 500 N and rotational torque of 10 N/m, the stress in the screw-rod system at a zero-degree TSA ranged from 335.1 to 657.5 MPa, while at a 17-degree TSA, it ranged from 242.8 to 648.5 MPa. Although there was some difference in the maximum stress between the two TSA angles, both were within the fracture threshold, indicating a relatively low risk of screw-rod system fracture.

CONCLUSION

The FEA results indicated that when using vertical screw placement at a zero-degree TSA, the overall maximum stress in both the vertebral body and screw-rod system was higher than that of screw placement at a 17-degree TSA. However, the maximum stress in the vertebral body remained lower than the typical allowable stress for the vertebral bone, ranging from 28.26%-36.87%. This indicates a low risk of vertebral body fractures. Although the pull-out resistance of pedicle screws at a zero-degree TSA decreased by 11.35% compared to that at 17 degrees, reaching 132.8 N indicates a relatively strong pull-out strength. Overall, the biomechanical performance of the zero-degree TSA vertical screw placement technique was inferior to that of the 17-degree TSA approach. However, it still exhibited a high pull-out strength, stability, and safety, indicating that it has a certain feasibility level.