The objective of this study was to compare the biomechanical differences of different rod configurations following anterior column realignment (ACR) and pedicle subtraction osteotomy (PSO) for an optimal correction technique and rod configuration that would minimize the risk of rod failure.
A validated spinopelvic (L1-pelvis) finite element model was used to simulate ACR at the L3–4 level. The ACR procedure was followed by dual-rod fixation, and for 4-rod constructs, either medial/lateral accessory rods (connected to primary rods) or satellite rods (directly connected to ACR level screws). The range of motion (ROM), maximum von Mises stress on the rods, and factor of safety (FOS) were calculated for the ACR models and compared to the existing literature of different PSO rod configurations.
All of the 4-rod ACR constructs showed a reduction in ROM and maximum von Mises stress compared to the dual-rod ACR construct. Additionally, all of the 4-rod ACR constructs showed greater percentage reduction in ROM and maximum von Mises stress compared to the PSO 4-rod configurations. The ACR satellite rod construct had the maximum stress reduction i.e., 47.3% compared to dual-rod construct and showed the highest FOS (4.76). These findings are consistent with existing literature that supports the use of satellite rods to reduce the occurrence of rod fracture.
Our findings suggest that the ACR satellite rod construct may be the most beneficial in reducing the risk of rod failure compared to all other PSO and ACR constructs.
Adult spinal deformity (ASD) is prevalent in up to 68% of the population and is associated with poor health-related quality of life [
To address this concern, anterior column realignment (ACR) has become a trusted, minimally invasive alternative to PSO. In an ACR procedure, an interbody cage is used to replace an intervertebral disc and the anterior longitudinal ligament (ALL) is released. ACR has shown similar radiographic correction as PSO, with lesser complications [
Thus, the objective of this study was to use finite element analysis (FEA) to suggest the most optimal procedure and rod configuration for ASD correction. For this purpose, 4 different rod configurations were investigated following ACR: dual-rod configuration, additional medial accessory rods (ACR-MED), additional lateral accessory rods (ACR-LAT), and additional short satellite rods (ACR-SAT). These rod combinations were chosen being among the common configurations seen in clinical practice. Also, these rods configurations have been investigated following PSO in our previous study [
A previously validated osteo-ligamentous spinopelvic model (L1-pelvis) was used to simulate ACR at the L3–4 level in this study [
To simulate the ACR surgery at the L3–4 level, the intervertebral disc at the index segment was completely removed along with the ALL. Additionally, resection of the posterior elements (facets, lamina) was performed at L3–4 level [
Overall, the ACR surgery was simulated with 4 different rod configurations (
• A 30° hyperlordotic cage at L3–4 level and bilateral rod fixation from L1–S1. (
• A 30° hyperlordotic cage at L3–4 level and bilateral rod fixation from L1–S1 + short satellite rods at L3–4. In this configuration, primary rods are not connected to the L3–4 pedicle screws. (
• A 30° hyperlordotic cage at L3–4 level and bilateral rod fixation from L1–S1 + medially affixed accessory rods connected to the primary rods via connectors at L2–3 and L4–5 regions. (
• A 30° hyperlordotic cage at L3–4 level and bilateral rod fixation from L1–S1 + laterally affixed accessory rods connected to the primary rods via connectors at L2–3 and L4–5 regions. (
The pelvis was constrained in all of the models. All of the models were subjected to a 400 N physiological compression load followed by a 7.5 Nm pure moment applied to the L1-superior endplate to simulate flexion/extension, lateral bending, and axial rotation [
The global ROM (L1–S1) for all of the models was analyzed. Also, the maximum von Mises stress in the rod for all of the rod configurations was used to calculate the factor of safety (FOS). FOS was determined by dividing the yield stress of CoCr by the maximum von Mises recorded for a given rod configuration. The higher the FOS lesser are the chances of rod fracture. The data for ACR rod configurations were also compared to the same rod configurations following PSO. The percentage change for ACR 4-rod constructs was calculated with respect to the dualrod ACR construct while percentage change for PSO 4-rod constructs was calculated with respect to the dual-rod PSO construct.
where, CoCr yield stress= 989 MPa.
The predicted global ROM showed that the use of more rods correlated to the lower ROM as higher ROM was observed in the dual-rod ACR model when compared to the 4-rod constructs (
The maximum von Mises stress in rods was computed under all loading conditions (
The location of the maximum rod stress and percentage differences of rod stress when compared to the ACR model are summarized in
The FOS was 3.35 in the ACR model. All 4-rod constructs saw this number increase with the ACR-SAT model showing the greatest FOS at 4.76. The ACR-LAT and ACR-MED showed a FOS of 4.17 and 4.20, respectively.
The previous study done by Vosoughi et al. [
The PSO models done by Vosoughi et al. [
Since PSO and ACR are among the most used procedures in correcting ASD, the dual and 4-rod configurations in either PSO or ACR were compared. Data from a previous study for dual and 4-rod constructs following PSO was used for comparison. Vosoughi et al. [
Moreover, the study by Vosoughi et al. [
A possible explanation for why the use of additional rods in ACR leads to higher reduction of rod stresses is the use of an interbody cage at the index segment, which has been shown to reduce rod stresses in PSO models as well [
The results of our numerical analyses demonstrate a decreased global L1–S1 ROM in all 4-rod constructs compared to the 2-rod construct, suggesting that the use of additional rods leads to a reduction in motion. Our results also indicated that the maximum von Mises stresses on the primary rods were considerably lower in all of the 4-rod constructs compared to the 2-rod construct. Our findings are consistent with the study of Januszewski et al. [
Godzik et al. [
However, with the addition of accessory rods, the high-stress regions on the primary rods migrated away from the ACR site to adjacent to the domino. Shen et al. [
In summary, our FEA assessed the global (L1–S1) ROM and maximum von Mises stress for a dual-rod construct and 3 different 4-rod constructs. The addition of accessory/satellite rods models resulted in higher reduction of maximum von Mises stress on the primary rods when used in the ACR models as compared to the PSO models. The ACR-SAT rod construct showed the highest FOS compared to all other ACR and PSO constructs. The findings of this study are consistent with the literature. Studies done by La Barbera et al. [
As with any computational study, our FEA also had certain limitations. This study did not consider the adverse effects of notches that are created on rods during manual rod contouring in a surgical environment. This study also did not simulate cyclic loading and thus how soon the rod construct will fail cannot be predicted. Another limitation of our study was that the model did not include paraspinal muscles. However, this limitation was addressed by the addition of follower loads that mitigate the muscle contractions and the bodyweight as proposed by Patwardhan et al. [
In conclusion, the 4-rod constructs consistently demonstrated lower maximum rod stresses when compared to the dualrod model following both ACR and PSO. The results of our study suggest that the 4-rod constructs proved to be more effective in ACR models rather than PSO models. These findings were consistent with existing literature that suggested the occurrence of rod fracture decreased when 4-rod constructs were implemented. Additionally, the ACR-SAT model is associated with the highest FOS and lowest maximum rod stress when compared to all other models. ACR reduces the risk for complication compared to PSO, but the use of 4-rod constructs may also be more beneficial in ACR compared to PSO. However, additional
The authors have nothing to disclose.
The work was supported in part by NSF Industry/University Cooperative Research Center at The University of California at San Francisco, San Francisco, CA, The University of Toledo, Toledo, Ohio, The Ohio State University, Columbus, Ohio and Northeastern University, Boston, Massachusetts (
Posterior view of the different anterior column realignment (ACR) rod configurations. (A) Bilaterally fixated ACR model (ACR). (B) Four-rod instrumented ACR model with short satellite rods at L3–4 (ACR-SAT). (C) Four-rod instrumented ACR model with medially affixed accessory rods (ACR-MED). (D) Four-rod instrumented ACR model with laterally affixed accessory rods (ACR-LAT).
Two lateral views of the anterior column realignment (ACR) model. The magnified views show the 30° hyper lordotic cage at L–4.
Comparison of the instrumented L1–S1 global range of motion (ROM) for different loading directions and configurations. ACR, anterior column realignment; ACR-SAT, ACR-short satellite rods; ACR-LAT, ACR-lateral accessory rods; ACR-MED, ACR-medial accessory rods.
Comparison of the maximum von Mises stress (MPa) found for the primary rods in the ACR, ACR-SAT, ACR-LAT, and ACR-MED models under all loading conditions. ACR, anterior column realignment; ACR-SAT, ACR-short satellite rods; ACR-LAT, ACR-lateral accessory rods; ACR-MED, ACR-medial accessory rods.
Material properties assigned to the finite element model [
Component | Element type | Young modulus (MPa) | Poisson ratio |
---|---|---|---|
Bone | |||
Cortical bone | C3D8 | 12,000 | 0.3 |
Cancellous bone | C3D8 | 100 | 0.2 |
Pelvic cortical bone | C3D8 | 17,000 | 0.3 |
Pelvic cancellous bone | C3D8 | 10 | 0.2 |
Intervertebral disc | |||
Nucleus | C3D8H | C1 = 0.12, C2 = 0.003, D1 = 0.0005 | 0.49 |
Annulus ground substance | C3D8 | Hyperelastic (C10, 0.348; D1, 0.3) | |
Annulus fibers | Rebar | 357–550 | |
Ligaments | T3D2 | Non-Linear | |
Apophyseal joints | Nonlinear, soft contact, GAPUNI elements | ||
Sacroiliac joints | Nonlinear, soft contact | ||
Implants | |||
ACR cage (PEEK) | C3D8 | 3,600 | 0.25 |
Screw head (CoCr) | C3D4 | 241,000 | 0.3 |
Primary/supplementary rods (CoCr) | C3D8 | 241,000 | 0.3 |
Ti6Al4V pedicle screw shaft | C3D4 | 11,500 | 0.3 |
ACR, anterior column realignment; PEEK, polyether ether ketone; CoCr, cobalt-chromium.
Values and locations of the maximum von Mises stress recorded on the rods for the 4-rod configurations tested
Motion | ACR | ACR-SAT | ACR-LAT | ACR-MED |
---|---|---|---|---|
Flexion | 277.3 | -47.35% | -19.73% | -20.34% |
ACR Index | L5–S1 | Adjacent to domino | Adjacent to domino | |
Extension | 200 | -41.50% | -47.80% | -49.10% |
ACR Index | L1–2 | Adjacent to domino | Adjacent to domino | |
Right bending | 228 | -35.53% | -25.44% | -28.07% |
ACR Index | L5–S1 | Adjacent to domino | Adjacent to domino | |
Left bending | 253 | -41.50% | -32.41% | -34.51% |
ACR Index | L5–S1 | Adjacent to domino | Adjacent to domino | |
Right rotation | 267 | -36.33% | -20.30% | -24.72% |
ACR Index | L1–2 | Adjacent to domino | Adjacent to domino | |
Left rotation | 260 | -25.00% | -16.54% | -20.85% |
ACR Index | L1–2 | Adjacent to domino | Adjacent to domino | |
Factor of safety | 3.35 | 4.76 | 4.17 | 4.20 |
Maximum values are reported for the ACR model, but percent difference with respect to the ACR model is reported for the ACR-SAT, ACR-LAT, and ACR-MED models. The factor of safety for each rod is also recorded.
ACR, anterior column realignment; ACR-SAT, ACR-short satellite rods; ACR-LAT, ACR-lateral accessory rods; ACR-MED, ACR-medial accessory rods.
Values for the percentage change of global ROM of the 4-rod construct when compared to its respective ACR or PSO dual-rod model under all loading types
Variable | Satellite rods |
Lateral accessory rods |
Medial accessory rods |
|||
---|---|---|---|---|---|---|
ACR model | PSO model | ACR model | PSO model | ACR model | PSO model | |
Extension | -62% | -4% | -62% | -12% | -62% | -16% |
Flexion | -51% | -11% | -48% | -11% | -41% | -15% |
Left bending | -34% | -54% | -40% | -1% | -51% | -8% |
Right bending | -25% | -61% | -29% | -1% | -47% | -8% |
Left rotation | -44% | 31% | -66% | -7% | -69% | -8% |
Right rotation | -47% | 31% | -70% | -7% | -69% | -8% |
Average | -44% | -11% | -52% | -6% | -56% | -11% |
The average percent difference is also reported.
ROM, range of motion; ACR, anterior column realignment; PSO, pedicle subtraction osteotomy.
Values for the percentage change of maximum rod stress of the 4-rod construct when compared to its respective ACR or PSO dual-rod model under all loading types
Variable | Satellite rods |
Lateral accessory rods |
Medial accessory rods |
|||
---|---|---|---|---|---|---|
ACR model | PSO model | ACR model | PSO model | ACR model | PSO model | |
Extension | -42% | -10% | -48% | 5% | -49% | -2% |
Flexion | -47% | -34% | -20% | -4% | -20% | -8% |
Left bending | -42% | -12% | -32% | 0% | -35% | -3% |
Right bending | -36% | -14% | -25% | +3% | -28% | -3% |
Left rotation | -25% | -11% | -17% | +11% | -21% | -2% |
Right rotation | -36% | -12% | -20% | +8% | -25% | 0% |
Average | -38% | -16% | -27% | +4% | -30% | -2% |
The average percent difference is also reported.
ACR, anterior column realignment; PSO, pedicle subtraction osteotomy.