Biomechanical Impact of Cement Augmentation on Pedicle Screw Fixation and Adjacent Segment Disease in Multilevel Lumbar Fusion: A Finite Element Analysis
Article information
Abstract
Objective
Cement augmentation is widely used to enhance pedicle screw fixation, particularly in osteoporotic patients. However, its effects on adjacent segment disease (ASD) and implant failure in multilevel lumbar interbody fusion remain unclear. This study aimed to assess the effectiveness of cement augmentation in preventing implant failure and its impact on ASD risk using finite element analysis (FEA).
Methods
A FEA of L2–S1 multilevel lumbar interbody fusion was performed to evaluate the biomechanical effects of cement augmentation. Three models were analyzed under normal and osteoporotic conditions: type 1 (no augmentation), type 2 (upper instrumented vertebra [UIV] augmentation), and type 3 (UIV and UIV+1 augmentation). Range of motion (ROM), intradiscal pressure (IDP), screw pull-out risk, and implant failure were assessed.
Results
Cement augmentation significantly reduced screw pull-out risk, particularly in osteoporotic conditions, where type 1 exhibited a failure rate of 91.5%, while type 2 and type 3 remained below 39%. Cement augmentation did not demonstrate a substantial impact on ASD development, as ROM and IDP changes remained within a minimal range in this FEA model. However, osteoporosis was associated with a substantial increase in IDP, with a result as high as 809%. Despite its benefits, augmentation at UIV+1 increased the risk of pedicle screw breakage and vertebral body fracture, with L1 (UIV+1) lower endplate fracture rate of 82.7% in type 3, compared to 56.6% in type 2 and 52.8% in type 1.
Conclusion
Cement augmentation effectively improves screw fixation and does not appear to significantly increase ASD risk based on this FEA study. Limiting cement augmentation to the UIV level in lumbar multilevel fusion may help reduce the risk of implant failure, though further clinical validation is required to confirm these biomechanical findings.
INTRODUCTION
Implant failure, including screw pull-out, screw breakage, and fractures of instrumented or adjacent vertebral bodies, remains a significant challenge in spine surgery, adversely affecting patient outcomes. These failures often necessitate revision surgeries, which tend to be more complex and less successful than primary procedures [1,2]. Cement augmentation has emerged as a crucial technique for enhancing the fixation strength of spinal implants and preventing implant failure, particularly in osteoporotic patients. By reinforcing the vertebral body with bone cement, such as polymethylmethacrylate, implant fixation is improved, reducing the risk of screw loosening and pull-out [3,4].
Despite these benefits, cement augmentation raises concerns regarding adjacent segment disease (ASD). Altered load distribution and increased stiffness in the augmented segment may accelerate degeneration in adjacent levels, potentially leading to new symptoms and the need for additional surgical intervention [5,6]. However, clinical studies on this issue have reported conflicting results; some studies have observed an increased incidence of ASD following augmentation, while others have suggested that appropriate surgical planning and technique can mitigate these risks. Nevertheless, these studies are often limited by short follow-up periods and small sample sizes, hindering definitive conclusions [6,7].
To overcome these limitations, finite element analysis (FEA) has emerged as a powerful tool for evaluating spine biomechanics under various surgical conditions. FEA enables precise simulations of stress distribution and implant stability, offering valuable insights into potential failure mechanisms [8,9]. In this study, we utilized FEA to simulate multilevel lumbar interbody fusion, specifically oblique lumbar interbody fusion (OLIF) and posterior lumbar interbody fusion (PLIF). These are commonly performed in clinical practice and represent critical cases where implant failure can significantly affect surgical outcomes [10,11].
This study aimed to assess the effectiveness of cement augmentation in preventing implant failure and evaluate its impact on ASD risk. By simulating various spine surgery scenarios with different levels of cement augmentation and osteoporosis, this study provides a detailed biomechanical analysis and predictive insights into surgical outcomes [12]. Through FEA modeling, we aim to overcome the practical limitations of clinical studies and contribute to a better understanding of the biomechanical implications of cement augmentation in lumbar spine surgery.
MATERIALS AND METHODS
This study employed an FEA approach to investigate the effects of cement augmentation in multilevel lumbar interbody fusion. The workflow consisted of (1) constructing a validated 3-dimensional (3D) FEA model; (2) simulating different surgical conditions with OLIF and PLIF; (3) applying physiological loading and boundary conditions; and (4) analyzing biomechanical parameters such as range of motion (ROM), peak von Mises stress (PVMS), and intradiscal pressure (IDP) to assess the risks of implant failure and ASD.
1. 3D Lumbar Spine Finite Element Modeling
This study utilized a previously validated 3D FEA model of the lumbar spine (L1–S1) based on computed tomography (CT) scans with 1-mm slice thickness, originally obtained from a 57-year-old Korean male without degenerative pathology (Supplementary Fig. 1) [13,14]. The 3D lumbar FEA model consisted of the vertebral body, intervertebral disc, posterior elements, and endplates. The mesh size of all models was set to 0.3 mm, with tetrahedral elements used for meshing. The 7 major vertebral ligaments were precisely modeled as truss elements using Abaqus software (ver. 6.14, Dassault Systems, France), ensuring the accuracy and reliability of the simulation. The mechanical properties of the spine, including those for osteoporotic conditions, were assigned based on previous studies and FEA of spinal degeneration, with Young modulus in osteoporotic bone reduced by approximately 60% compared to normal bone to reflect the mechanical characteristics of moderate to severe osteoporosis (T-score≤-2.5) (Table 1) [15-21].
Material properties of the implant and spine used in finite element models
2. Multilevel Lumbar Interbody Fusion Modeling
For implementation of the multilevel lumbar interbody fusion model, ANYPLUS OLIF and PLIF cages (GS Medical Co. Ltd., Korea) and Scoli poly-axial pedicle screws (GS Medical Co. Ltd., Korea) were designed using 3D CAD (computer-aided design) software, and the material properties of the implant were applied accordingly (Table 1).
Lumbar interbody fusion was performed at the L2–S1 levels based on the 3D lumbar spine FEA model. OLIF cages were inserted at L2–5 levels, while PLIF cages were placed at the L5–S1 levels as OLIF at L5–S1 is often challenging due to vascular structures [22]. The annulus fibrosus and ligaments were assumed to be partially removed, and the nucleus pulposus was assumed to be completely removed to allow cage insertion. Each cage was positioned at the center of the upper endplate (Supplementary Fig. 2A–C). For posterior fixation, pedicle screws were inserted, with L2 designated as the upper instrumented vertebra (UIV). The screw entry point was set at the midpoint of the transverse process with a 30° convergence angle, and the rod was fixed (Supplementary Fig. 2D).
Bone cement was applied to the vertebral body following standard surgical techniques. A total of 5 mL of cement (2.5 mL per pedicle) was injected into the L2 vertebral body (UIV), while 6 mL of cement (3 mL per pedicle) was injected into the L1 vertebral body (UIV+1) (Fig. 1A) [5]. The volume of cement was determined based on previous studies and the authors’ clinical experience [23,24]. The samples were classified into 3 types based on cement augmentation at the L1 or L2 levels. Additionally, each type was further analyzed under osteoporotic conditions, resulting in a total of 6 models (Fig. 2B). Each model was simulated once under the same loading and boundary conditions, as repeated runs in deterministic FEA yield identical results.
(A) Cement-augmented pedicle screw insertion at the upper instrumented vertebra (L2) and cement augmentation (vertebroplasty) at the L1 body. (B) Six types of samples. Type 1, without cement augmentation; type 2, L2 cement-augmented screws; type 3, L2 cement-augmented screws with L1 vertebroplasty.
Biomechanical analyses. (A) Screw pull-out at L2. (B) Range of motion of the L1–2 segment. (C) Intradiscal pressure at the L1–2 level. (D) Screw breakage at L2. (E) Fracture risk at L1. PVMS, peak von Mises stress.
3. Loading and Boundary Conditions
To evaluate the effect of each factor on the ROM, this study utilized the hybrid protocol, a method that adjusts the displacement of the implanted spine model to achieve the same angular rotation as that of the normal spine (Supplementary Fig. 3) [25]. A pure moment of 10 N ·m was applied to simulate 6 spinal movements: flexion, extension, left lateral bending, right lateral bending, left axial rotation, and right axial rotation. The S1 lower endplate was constrained in all directions to ensure controlled movement. Additionally, a 400-N follower load was applied to account for the physiological pulling force of the spinal muscles [26]. “Tied contact” was applied under the assumption of complete union between all implants and bones and complete fixation between the implants, ensuring no movement or separation between the components [27].
4. Biomechanical Analysis Using the Established Model
FEA was performed using Abaqus software (ver. 6.14, Dassault Systems, France) to evaluate the potential risks of ASD and implant failure following multilevel interbody fusion. To assess screw loosening at the UIV level, the PVMS at the interface between the UIV (L2) and the pedicle screw was measured and compared with the yield stresses of the cancellous bone (16.3 MPa) and bone cement (92.2 MPa) (Fig. 2A). A higher PVMS at this interface indicated an increased probability of screw pull-out, suggesting a higher likelihood of cracks or breakage at the insertion site [28,29]. The risk of ASD was evaluated by measuring the ROM of the adjacent segment (L1–2) and the IDP of the nucleus pulposus at the L1–2 levels (Fig. 2B and C).
To assess the possibility of screw failure, the PVMS of the pedicle screws was measured and compared with the yield stress of the material (860 MPa) (Fig. 2D). Additionally, to evaluate the risk of adjacent L1 vertebral body fracture, the PVMS at the lower endplate was analyzed (Fig. 2E).
RESULTS
1. Effect of Cement Augmentation on Pull-Out of an L2 Screw
Under normal bone conditions, cement augmentation in types 2 and 3 significantly improved stabilization, reducing pullout percentages to less than 49% for all measured motions compared to type 1. The reduction was particularly notable in left and right lateral bending, with pull-out percentages of 40.18% and 39.32%, respectively. The effect was even more prominent under osteoporotic conditions. In type 1, osteoporosis was accompanied by a dramatic increase in screw pull-out percentage, reaching up to 91.53% in left lateral bending. In contrast, types 2 and 3 maintained pull-out percentages less than 39% for all motions, demonstrating the stabilizing effect of cement augmentation. Cement augmentation at the L2 levels in types 2 and 3 reduced screws pull-out percentages by 61% to 84% compared to type 1 (Supplementary Table 1). These trends are also clearly illustrated in Fig. 3, which shows the screw pull-out rates across all groups and movement directions.
Pull-out rates of L2 screws. Osteoporosis significantly increased the risk of screw pull-out across all stabilization types, with type 1 being the most highly affected. Type 1, without cement augmentation; type 2, L2 cement-augmented screws; type 3, L2 cement-augmented screws with L1 vertebroplasty; type 1-OP, type 1 in osteoporotic vertebra; type 2-OP, type 2 in osteoporotic vertebra; type 3-OP, type 3 in osteoporotic vertebra.
Osteoporosis significantly increased the risk of screw pull-out across all stabilization types, with type 1 being the most highly affected. In type 1, osteoporosis increased pull-out forces by up to 188% compared to normal bone conditions, particularly during axial rotation. While osteoporosis also increased PVMS for pull-out in types 2 and 3, the values remained significantly lower than those observed in type 1.
2. Effect of Cement Augmentation on ROM of the Adjacent L1–2 Segment
Supplementary Table 2 presents the ROM (°) for each type across the 6 movement directions. Under normal bone conditions, types 2 and 3, which included cement augmentation, exhibited a slightly greater ROM than type 1 across all tested movements. However, the differences among the 3 types remained within 5% across all movements. A similar trend was observed under osteoporotic conditions, where types 2 and 3 also demonstrated a slightly increased ROM compared to type 1, remaining within a 5% difference.
Despite these minimal differences between types, osteoporotic conditions led to a significant overall increase in ROM compared to normal bone conditions across all models. Notably, axial rotational capability in type 1 decreased by approximately 83% under osteoporotic conditions compared to normal bone conditions (Fig. 4A).
Risk of adjacent segment disease. (A) Range of motion at L1–2. (B) Intradiscal pressure at L1–2. Type 1, without cement augmentation; type 2, L2 cement-augmented screws; type 3, L2 cement-augmented screws with L1 vertebroplasty; type 1-OP, type 1 in osteoporotic vertebra; type 2-OP, type 2 in osteoporotic vertebra; type 3-OP, type 3 in osteoporotic vertebra.
3. Effect of Cement Augmentation on IDP at the L1–2 Level
The IDP at the L1–2 level was analyzed across stabilization types (Supplementary Table 2). Under normal bone conditions, the FEA results indicated that the IDP in types 1–3 remained similar, with variations within a 5% threshold. A similar trend was observed under osteoporotic conditions, where the relative differences between types also remained within a 5% range.
However, a comparison between normal and osteoporotic conditions revealed that osteoporosis substantially increased IDP across all movements, with the most notable increase occurring during lateral bending (Fig. 4B). During right lateral bending, type 2 under osteoporotic conditions exhibited an IDP of 9.53 MPa, an increase greater than 809% compared to the 1.05 MPa in normal bone conditions. Similarly, the increase in IDP due to osteoporosis was 797% in type 3 and 760% in type 1.
4. Evaluation of L2 Pedicle Screw Failure Under Various Movements
Under normal bone conditions, cement augmentation in types 2 and 3 resulted in slight increase in PVMS and failure rate for L2 screw breakage compared to type 1, with changes up to 10% during flexion. Under osteoporotic conditions, type 3 exhibited the highest PVMS and failure rates for L2 screws, followed by types 2 and 1. During lateral bending, the screw failure rate in type 3 and 2 were 20% and 11% higher than that in type 1, respectively. Additionally, during right axial rotation, type 3 showed the highest PVMS value of 699.9 MPa, with failure rates reaching up to 81.4%. Fig. 5A and Table 2 show the distribution of screw failure rates across all types and movements.
Risk of implant failure. (A) Failure of L2 pedicle screws. (B) Failure of the L1 vertebral body (lower endplate). Type 1, without cement augmentation; type 2, L2 cement-augmented screws; type 3, L2 cement-augmented screws with L1 vertebroplasty; type 1-OP, type 1 in osteoporotic vertebra; type 2-OP, type 2 in osteoporotic vertebra; type 3-OP, type 3 in osteoporotic vertebra.
Failure analysis of L2 pedicle screw and L1 vertebral body
5. Fracture Risk of the L1 Vertebral Body
Under normal bone conditions, the fracture rates of the L1 vertebral body were comparable across groups, with variations within 5%. However, under osteoporotic conditions, the results differed significantly from those observed in normal bone conditions. Type 3 exhibited the highest fracture risk of the L1 body, followed by types 2 and 1. Notably, lateral bending posed the highest fracture risk in type 3, reaching 82.7%, while the fracture risks in types 2 and 1 were 56.6% and 52.8%, respectively. These fracture risk patterns are illustrated in Fig. 5B and are summarized with detailed values in Table 2.
DISCUSSION
The primary objective of this study was to evaluate the biomechanical effects of cement augmentation in multilevel lumbar interbody fusion using FEA. Specifically, we aimed to assess its impacts on screw pull-out resistance, ASD risk, and implant failure under both normal and osteoporotic bone conditions. While cement augmentation is widely used to enhance implant stability, concerns remain regarding its potential effects on adjacent segments and long-term mechanical integrity. By simulating various surgical conditions, this study provides a comprehensive analysis of the stabilization effects of cement-augmented screws and explores their clinical implications.
Previous clinical studies and FEA research have consistently demonstrated that cement augmentation effectively prevents screw pull-out. A meta-analysis reported that the screw loosening rate was significantly lower in the cement-augmented screw group than in the conventional screw group (odds ratio [OR], 0.13; 95% confidence interval [CI], 0.07–0.22), while the postoperative fusion rate was higher in the cement-augmented screw group (OR, 2.80; 95% CI, 1.49–5.25) [30]. Chevalier et al. [31] conducted a study utilizing experimental testing, micro-CT, and microfinite element models on 6 cadaver specimens. Their findings demonstrated strong correlations between the measured experimental fixation strength and the modeled prediction of pull-out, validating the modeling approach. Their study concluded that cement augmentation around pedicle screws increased fixation stiffness in both pull-out and bending, particularly in osteoporosis.
Our findings were consistent with these results. Notably, under osteoporotic conditions, type 1 exhibited a failure rate of 91.53%, whereas types 2 and 3, which incorporated cement augmentation, achieved pull-out percentages less than 39%. These results indicate that cement augmentation effectively prevents screw loosening and pull-out at the UIV level. Together with previous studies, our findings confirm that cement augmentation is highly effective in preventing screw pull-out. However, based on the results of types 2 and 3, as well as type 2-OP and type 3-OP, vertebroplasty at the adjacent segment did not result in significant differences in screw loosening or pull-out.
Regarding the development of ASD, this FEA study demonstrated that cement augmentation in types 2 and 3 slightly increased ROM under both normal and osteoporotic bone conditions. However, the differences in ROM among the types remained within a 5% range. Similarly, IDP variations across types remained within a 5% range under both conditions. These findings indicate that cement augmentation was not significantly associated with ASD development. Zhou et al. [5] analyzed the PVMS of the intervertebral disc and the ROM in the adjacent segments (L3–4 and L5–S1) after L4–5 fusion using an L3–S1 3D FEA model. They concluded that cement augmentation was more likely to increase the potential risk of ASD. However, the differences in ROM and IDP between the 2 groups were minimal, with variations within 0.4° and 0.19 MPa across all movements. Additionally, the study simulated a single level and did not compare normal bone with osteoporosis. While cement augmentation itself was not a significant factor, the increased stiffness of interbody fusion constructs may still contribute to adjacent segment stress. In this context, posterolateral fusion without interbody cages has been associated with a lower inci-dence of ASD in clinical studies [32]. Further FEA comparisons could help evaluate the biomechanical differences between these strategies.
In addition to surgical technique, patient-related factors such as bone quality may also influence ASD development. This study demonstrated that osteoporosis was significantly associated with ASD development. IDP increased by up to 807% when comparing normal to osteoporotic conditions, regardless of type. These findings suggest that changes in material properties due to osteoporosis may have a greater effect on adjacent segment biomechanics than the localized stiffness changes from cement augmentation. This is particularly important because clinical studies primarily focus on osteoporotic patients when evaluating cement augmentation, hindering assessment of the direct impact of osteoporosis itself on ASD. Consistent with our results, Li et al. [33] reported that osteoporosis, compared to normal bone mineral density (BMD), led to a deterioration of biomechanical characteristics in the adjacent disc, potentially increasing the incidence of ASD after percutaneous transforaminal endoscopic discectomy. Similarly, a study on OLIF found that poor BMD exacerbates motility compensation of adjacent segments following OLIF, which may lead to a higher risk of ASD [34]. However, further comparative clinical studies are needed to validate these findings by examining how osteoporosis and cement augmentation independently affect ASD development.
The effects of cement augmentation on pedicle screw breakage and UIV+1 level vertebral body fracture also are noteworthy. Our study demonstrated that cement-augmented screws slightly increased the risk of breakage compared to conventional screws under normal bone conditions. However, in osteoporotic conditions, the risk of screw failure increased dramatically with cement augmentation. The greatest effect was observed in type 3, where cement augmentation extended to UIV+1, leading to the highest risk of screw breakage. Similarly, for the L1 (UIV+1) vertebral body, the fracture risk did not vary significantly among types under normal bone conditions. However, under osteoporotic conditions, type 3, which included cement augmentation at UIV+1, exhibited the highest fracture risk at the L1 lower endplate (82.7%) compared to type 2 (56.6%) and type 1 (52.8%).
Although cement augmentation is typically considered for osteoporotic patients, this study revealed that augmentation at the UIV+1 level increases the risk of screw breakage and UIV+1 vertebral body fracture in such patients. However, the screw pull-out rates at the UIV level were similar between types 2 and 3. Therefore, in lumbar multilevel fusion, cement augmentation at the UIV level appears to provide benefits in screw fixation while minimizing the risk of implant failure [35]. This conclusion contradicts the current consensus in adult spinal deformity surgery, where cement augmentation is often extended to UIV+1 in fusions extending above the thoracolumbar junction [36,37]. This discrepancy may be attributed to differences between lumbar fusion levels and those extending beyond the thoracolumbar junction, where a long lever arm and thoracic kyphosis influence spinal biomechanics [38]. Another possible explanation relates to the way cement augmentation was modeled in our simulation. In the FEA model, cement at the UIV+1 level was confined to a central spherical mass, without extension toward the endplates or anterior cortex. This conservative distribution reflects clinical practice aimed at minimizing cement leakage, but may leave cancellous bone near the endplates susceptible to stress concentration and fracture, particularly in osteoporotic conditions. Alternative cement distribution patterns may alter the biomechanical response of the vertebra. Therefore, further studies are necessary to explore this distinction and evaluate whether different augmentation strategies at the UIV+1 level influence adjacent vertebral fracture and the progression of ASD over time.
This study has several limitations. First, as an FEA-based study, it relies on computational modeling rather than clinical outcomes, and the material properties used to simulate osteoporosis may not fully capture the actual changes in bone quality observed in osteoporotic patients. Since individual variations in BMD and stage of degeneration were not considered, future studies should incorporate these factors to improve the accuracy of the results [39]. Second, there are limitations in the selection of boundary conditions. In this study, all contact conditions between the bones and implants were set as “tied contact,” assuming a completely fused state after surgery. However, in clinical practice, the postoperative period is a critical phase during which complications such as screw loosening or vertebral body fractures can occur [40]. To enhance clinical relevance, future research should consider frictional coefficients and simulate the immediate postoperative condition to better reflect real-world surgical outcomes. Third, this study relied solely on FEA without complimentary experimental or clinical validation. While FEA provides valuable insights into biomechanical behavior, its findings must be validated through experimental studies and clinical trials [41]. Additionally, this study focused exclusively on L1– S1 multilevel fusion, and the findings may not be directly applicable to other spinal fusion procedures, particularly those involving the thoracolumbar junction [36,37]. Further research is required to evaluate the biomechanical characteristics across fusion levels and surgical conditions.
Despite these limitations, this study provides valuable insights into the biomechanical effects of cement augmentation in multilevel lumbar fusion, one of the most commonly performed procedures for degenerative lumbar diseases.
CONCLUSION
This FEA study demonstrated that cement-augmented screws effectively prevented screw pull-out in both normal and osteoporotic bones, with a particularly dramatic reduction in pullout risk under osteoporotic conditions. More importantly, cement augmentation did not increase the possibility of ASD, as assessed by the ROM and IDP of the adjacent segment. Instead, ASD appeared to be more closely associated with osteoporosis rather than cement augmentation. However, vertebroplasty at the level above the UIV should be carefully considered. Contrary to expectations, and unlike in adult spinal deformity surgery involving the thoracolumbar junction, pedicle screw breakage at the UIV and vertebral body fractures above the UIV were significantly increased, particularly in osteoporotic conditions. This study provides biomechanical insights into the effects of cement augmentation in multilevel lumbar interbody fusion, reflecting clinically common scenarios. However, further clinical validation is necessary to confirm the findings of this biomechanical FEA study.
Supplementary Materials
Supplementary Tables 1–2 and Supplementary Figs. 1–3 are available at https://doi.org/10.14245/ns.2550294.147.
Establishment of a 3-dimensional (3D) lumbar spine finite element model using geometric data extracted from computed tomography (CT) scans and model segmentation.
(A) Three-dimensional modeling and dimensions of cages and pedicle screws used in this study, (B) Positions of OLIF cages at the L2–5 levels, (C) Positions of PLIF cages at the L5–S1 levels, (D) Insertion points and convergence angles of pedicle screws. OLIF, oblique lumbar interbody fusion; PLIF, posterior lumbar interbody fusion.
Loading and boundary conditions. Six movements (flexion, extension, left lateral bending, right lateral bending, left axial rotation, right axial rotation) simulated using a pure moment of 10 N · m and a follower load of 400 N.
Loosening and pull-out of L2 screw
Range of motion and intradiscal pressure of the adjacent L1–2 level
Notes
Conflict of Interest
The authors have nothing to disclose.
Funding/Support
This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Education (2021R1I1A1A01059501).
Author Contribution
Conceptualization: SJL, YHK, JSL, HYP; Data curation: MYJ, JHA; Formal analysis: MYJ, SJL, JHA, YHK, JSL, HYP; Funding acquisition: HYP; Methodology: MYJ, SJL, JHA, HYP; Project administration: SJL, YHK, HYP; Writing – original draft: MYJ, HYP; Writing – review & editing: MYJ, SJL, JSL, HYP.