removal of fixation construct could mitigate adjacent segment stress after lumbosacral fusion a finite element analysis

Methods: Using a validated lumbosacralfinite-element model, three variations at the L4–L5 segment were ana-lyzed: 1 moderate disc degeneration, 2 instrumented with a stand-alone cage and

Yueh-Ying Hsieha, Chia-Hsien Chena, Fon-Yih Tsuangb,c, Lien-Chen Wua,c,

Shang-Chih Lind, Chang-Jung Chianga,e,⁎

a

Department of Orthopaedics, Shuang Ho Hospital, Taipei Medical University, Taiwan

b

Division of Neurosurgery, Department of Surgery, National Taiwan University Hospital, Taiwan

c

Institute of Biomedical Engineering, National Taiwan University, Taiwan

d

Graduate Institute of Biomedical Engineering, National Taiwan University of Science and Technology, Taiwan

e Department of Orthopaedics, School of Medicine, College of Medicine, Taipei Medical University, Taiwan

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 5 February 2016

Accepted 21 February 2017

Background data: Combined usage of posterior lumbar interbody fusion and transpedicularfixation has been ex-tensively used to treat the various lumbar degenerative disc diseases The transpedicularfixator aims to increase stability and enhance the fusion rate However, how the fused disc and bridged vertebrae respectively affect ad-jacent-segment diseases progression is not yet clear

Methods: Using a validated lumbosacralfinite-element model, three variations at the L4–L5 segment were ana-lyzed: 1) moderate disc degeneration, 2) instrumented with a stand-alone cage and pedicle screwfixators, and 3) with the cage only after fusion The intersegmental angles, disc stresses, and facet loads were examined Four motion tests,flexion, extension, bending, and twisting, were also simulated

Findings: The adjacent-segment disease was more severe at the cephalic segment than the caudal segment After solid fusion andfixation, the increase in intersegmental angles, disc stresses and facet loads of the adjacent seg-ments were about 57.6%, 47.3%, and 59.6%, respectively However, these changes were reduced to 30.1%, 22.7%, and 27.0% after removal of thefixators This was attributed to the differences between the biomechanical char-acteristics of the fusion andfixation mechanisms

Interpretation: Fixation superimposes a stiffer constraint on the mobility of the bridged segment than fusion The current study suggested that the removal of spinalfixators after complete fusion could decrease the stress at ad-jacent segments Through a minimally invasive procedure, we could reduce secondary damage to the paraspinal structures while removing thefixators, which is of utmost concern to surgeons

© 2017 The Authors Published by Elsevier Ltd This is an open access article under the CC BY-NC-ND license (http://

creativecommons.org/licenses/by-nc-nd/4.0/)

Keywords:

Adjacent segment disease

Spinal fixator

Interbody fusion

Finite element

1 Background

Posterior lumbar interbody fusion has gradually been used to

imme-diately restore the dehydrated disc to its original height (Corniola et al.,

2015; Hikata et al., 2014) A transpedicularfixator is instrumented to

stabilize the anterior vertebrae and enhance the bony fusion, thus

avoiding cage subsidence and back-out at the bone-cage interfaces

(Lequin et al., 2014; Oh et al., 2016) However, the rigidity-raising effect,

resulting from interbody fusion and transpedicularfixation, potentially

induces adjacent segment disease (ASD) problems that accelerate the

degeneration of the adjacent discs and facet joints (Kwon et al., 2013;

Lawrence et al., 2012; Lee et al., 2014; Nakashima et al., 2015) Such

an instrumentation-induced problem has been attributed to the fact that the constrained mobility and loads of the instrumented segments

is compensated for by the adjacent segments (Lu et al., 2015; Okuda

et al., 2014)

As an alternative, some dynamicfixators have been designed to pro-vide theflexibility to limit both kinematic and kinetic constraints on the instrumented segments, thus mitigating the post-operative risk of ASD progression (Barrey et al., 2016; Galbusera et al., 2011; Hudson et al., 2011; Kim et al., 2011) There have been a great many attempts to de-signflexibility into the dynamic fixator, such as a rod-rod joint (i.e ISO-BAR), a rod-screw joint (i.e Dynesys), a screw hinge type (i.e COSMIC), and aflexible rod (i.e BioFlex) Some clinical reports showed

satisfacto-ry results for achieving a good bony fusion rate while suppressing ASD progression (Hudson et al., 2011; Kim et al., 2011) However, there are still some studies that showfixator failure (screw loosening and compo-nent wear) and post-operative complications (Barrey et al., 2016; Galbusera et al., 2011) Consequently, static, rather than dynamic

⁎ Corresponding author at: Department of Orthopaedics, Shuang Ho Hospital, Taipei

Medical University, No 291, Zhongzheng Rd., Zhonghe District, New Taipei City 23561,

Taiwan.

E-mail address: cjchiang@s.tmu.edu.tw (C.-J Chiang).

http://dx.doi.org/10.1016/j.clinbiomech.2017.02.011

Contents lists available atScienceDirect

Clinical Biomechanics

j o u r n a l h o m e p a g e :w w w e l s e v i e r c o m / l o c a t e / c l i n b i o m e c h

fixators, are still the principal method of treating such lumbosacral

problems

Recently, minimally invasive spine surgery (MISS) technique for

interbody fusion and transpedicularfixation has been extensively

adopted (Bourgeois et al., 2015; James and William, 2015; Niesche et

al., 2014) Compared with the traditional technique, the screws and

rods can be instrumented and assembled through small hole-like

wounds which could cause less injury to the paraspinal soft tissue

struc-tures Whether traditional or MISS technique is adopted, however, the

metallicfixation inevitably induces kinematic and kinetic compensation

from the instrumented to adjacent segments (Kwon et al., 2013;

Lawrence et al., 2012; Lee et al., 2014; Lu et al., 2015; Nakashima et al.,

2015; Okuda et al., 2014) Using static rather than dynamicfixation,

the current authors have not yet found enough literature report to

re-veal an effective technique to mitigate the ASD progression Intuitively,

it seems that post-operative removal of the staticfixator mitigates the

stress on adjacent segments However, removing the spinalfixator

from the traditional midline approach has been a major concern, due

to massive destruction of the posterior musculature again For the spinal

fixators used in MISS, however, a similar attempt to remove the static

fixator via paramedian approach might be practical (Fig 1) From the

authors' experience, the size of an entry wound to remove the MISS

fixator, through the previous surgical wound, may only be around 20–

30 mm (Fig 1D)

After complete solid fusion has occurred, the current authors have

attempted to remove the screws and rods by MISS technique for

disassembling the highly structural constraint of the staticfixator on

the fused segment (Fig 1) If this could decrease stiffness of fusion

segments and reduce the disc stress of adjacent segments, this attempt potentially provides a trade-off between ASD mitigation and muscula-ture destruction This study used the validated nonlinearly lumbosacral model to evaluate the biomechanical differences between the ‘fusion-fixation’ and ‘fusion-only’ models Special effort was taken to illustrate the difference in the structural constraint between fusion andfixation

If the effects of the ASD mitigation are significant, the removal of the in-ternalfixator by MISS technique can be recommended after posterior lumbar intervertebral fusion

2 Methods 2.1 Lumbosacral models The lumbosacral model from L1 to S1 segments has been developed and validated in the previous studies of the current authors (Chien et al., 2014; Chuang et al., 2012; Chuang et al., 2013) For a paired facet joint, the orientation and separation of the articulating surfaces were cau-tiously established to ensure a consistent unloaded neutral position within a range of around 0.5 mm Other than the L4–L5 segment, the re-maining segments were assumed healthy The geometric size and mate-rial strength of the L4–L5 segment was simulated as ‘moderate degeneration’ The contractions of the five muscle groups were

simulat-ed as distributsimulat-ed loads to stabilize the lumbosacral column (Fig 2) The concentrated loads (M: moment and C: compression) were the result of body weight and the contractions of the abdominal muscles The hybrid use of compression (=150 N) and moment (=10 Nm) was applied at the lumbosacral top to activate lumbosacral motion The lumbosacral

Fig 1 The X-ray images and the operation wounds of the same patient subjected to interbody fusion and transpedicular fixation (A) X-ray of fusion with MISS fixator (B) The operation

fixator (D) The new wounds after removing the MISS fixator.

column was rigidly constrained at the bottom and activated by the

dis-tributed and concentrated loads There were four types of lumbosacral

motion simulated in this study:flexion, extension, bending, and

twist-ing Theflexion and extension are in sagittal plane, and bending and

twisting are in the coronal and transverse planes

2.2 Intervertebral fusion and transpedicularfixation

For the fusion/fixation model, the L4–L5 segment was immobilized

by a transpedicularfixator and fused by a stand-alone cage (Fig 2)

For the fusion model, thefixator in the fusion/fixation model was

re-moved The stand-alone cage and two-sided screws and rods were

as-sumed symmetric in the sagittal plane The longitudinal rods were

consistently 5.0 mm in diameter and the screw diameters (5.5 mm) of

all models were the same across equivalent tests The specification of

the banana cage was 30 mm in length, 10 mm in width, and 10 mm in

height The metallic components of thefixator were consistently made

from titanium-based alloy (Ti-6Al-4V ELI) The stand-alone cage was

made from PEEK (Wiltrom, Taiwan) A spinal surgeon was engaged to

monitor the development of the fusion/fixation and fusion models, to

confirm the proper instrumentation

2.3 Finite-element analyses

Except for the facet joint, no slippage and separation were allowed

be-tween the tissues and implants The interfaces of the facet joints were

modeled as surface-to-surface contact elements in which articulating

fric-tion is ignored and only transmitted normal forces are considered The

cri-terion for controlling the same displacement of the lumbosacral top was

adopted as a reasonable approach to evaluate the effects of the fusion

andfixation on the adjacent segments (Chuang et al., 2013) The materials

of all implants were assumed to have linearly elastic and isotropic

mate-rial properties throughout The calculated von Mises stresses of the cage

and fixator were compared with the yielding strength of the

corresponding material, to validate the assumption of linear elasticity For the fusion/fixation and fusion models, the solid fusion of the L4–L5 segment was simulated as the intimate bonding at the bone-cage inter-faces The remaining nucleus pulposus and annulus fibrosus were modeled as dehydration grade III to simulate the loss of the disc elasticity Experimental and numerical comparisons were used to validate the sim-plifications and assumptions of the finite-element model (Chuang et al.,

2013) Using the cadaveric data, the results of the intact model were val-idated by the ROM of all discs forflexion, extension, rotation, and bending During validation, the initially chosen elastic moduli of the disc and some ligaments were slightly modified within the physiological range to im-prove agreement with the cadaveric results Then, the intact model was transformed into the moderately degenerative model to further compare with the numerical data of the literature counterpart During extension and rotation, the predicted forces of the different facet joints were com-pared for validation

Three types of comparison indices were chosen to evaluate the kine-matic and kinetic responses at the L3–L4 and L5–S1 segments: interseg-mental angles, disc stresses, and facet loads The interseginterseg-mental angle was defined as the change in disc angles before and after exerting loads and denoted as the loss of the intersegmental mobility The disc stress and facet load was defined in terms of von Mises stress and com-pressive force in this study All indices of the fusion/fixation and fusion models were normalized to the corresponding values of the degenera-tive model The differences in the normalized indices provide informa-tion about the biomechanical effects of removing thefixator on the ASD progression

3 Results The kinematic and kinetic responses of two instrumented models at the L3–L4 and L5–S1 segments are shown inFig 3 As compared with the degenerative model, the normalized increases in intersegmental angle of the fusion/fixation model were 61.3% and 53.8% at the cephalic

Fig 2 Fusion/fixation model The concentrated and distributed loads were applied onto the lumbosacral column One stand-alone cage and one MIS transpedicular fixator were instrumented at the L4–L5 segment (A) Coronal view (B) Sagittal view.

and caudal segments, respectively (Fig 3A) For the fusion model, these

increases were reduced to 28.0% and 27.0% The removal of the spinal

fixator can thus suppress the compensation of the adjacent vertebral

motion by 33.3% and 26.8%

Similarly to the kinematic results, fusion/fixation induces the

adja-cent discs to be subjected to the transferred loads from the

instrument-ed segment (Fig 3B) For the fusion/fixation model, the normalized

increases in disc stress at the L3–L4 and L5–S1 segments were 50.1%

and 44.5%, respectively For the fusion model, the disc stresses at the

ce-phalic and caudal segments only increase to 26.0% and 23.2%,

respec-tively After removing thefixator, the stresses at the L3–L4 and L5–S1

discs can thus be further reduced by 24.1% and 21.3%, respectively

The kinetic changes due to instrumentation of the normalized facet

loads are shown inFig 3C Compared with the degenerative model,

the cephalic and caudal facet loads of the fusion/fixation model were

in-creased by 64.5% and 54.7%, respectively Similarly to the

intersegmen-tal angle and disc stress, the loading compensation at the cephalic

segment was higher than at the caudal segment For the fusion model,

the aforementioned increases were reduced by 37.2% and 28.1%,

respec-tively Between the two models, the differences in the compensated

facet loads were 27.3% and 26.6% at the L3–L4 and L5–S1 segments,

re-spectively This indicates that thefixation can greatly constrain the

in-tervertebral mobility and that the disassembly of thefixator is worth

executing in the situation of solid fusion and where assessed as

surgical-ly safe

4 Discussion

Consistent with the available literature reports,(Kwon et al., 2013;

Lawrence et al., 2012; Lu et al., 2015) this study showed more severe

stress and mobility at the cephalic than the caudal segment (Fig 3)

For the fusion/fixation model, the increases of the intersegmental

angles, disc stresses, and facet loads were 61.3%, 50.1%, and 64.5% at the L3–L4 segment and 53.8%, 44.5%, and 54.7% at the L5–S1 segment, respectively The current authors used the moment-arm effect to ex-plain the worse ASD deterioration at the cephalic segment (Fig 4) The distance between the cephalic disc and screws is significantly less

Fig 3 Three comparison indices of the fusion/fixation and fixation models (A) Intersegmental angles (B) Disc stresses (C) Facet loads.

Fig 4 This study used the moment-arm effect to illustrate the severer ASD progression at the cephalic than caudal segment The distance between the disc center and pedicle screw was used as an index to transfer the constrained mobility of the instrumented segment to the adjacent segments.

than that of the caudal segment This results in higher compensation of

the disc mobility and facet loads from the instrumented segment to the

cephalic than caudal segment

After instrumentation, the inserted cage and bridgedfixator have

definitely altered the structural characteristics of the adjacent segments

(Kwon et al., 2013; Lawrence et al., 2012; Lee et al., 2014; Lequin et al.,

2014; Nakashima et al., 2015; Oh et al., 2016) The inserted cage will

stiffen the instrumented disc; thus limit the intersegmental mobility

and shift the motion center from the posterior (i.e point A) to some

re-gion (i.e point B) within the cage (Fig 5) (Kim et al., 2015) The

motion-center shift of the dehydrated and stiffened disc alters the

biomechani-cal behavior of the fused segment The constrained mobility of the fused

segment will be transferred to the adjacent segments to increase the

ki-nematic and kinetic demand around them This is the fusion effect of the

sandwiched cage (Fig 5) For the fusion/fixation model, the linkage of

the screws and rods further deteriorates the biomechanical

compensa-tion at the adjacent segments (Figs 3 and 5) This is thefixation effect

that is attributed to the longitudinal rods and pedicle screws The

fixa-tion effect can make the mofixa-tion center more posterior (i.e point C)

(Highsmith et al., 2007)

The current authors suggest the stiffness-increasing mechanism as

the potential reason of thefixation-induced compensation (Fig 4)

Even for the spinalfixator applied by MISS technique, the bridging

con-struct of the polyaxial screws and rods is still stiffer than the motion

seg-ment At the initial stage, the intimate contact at the screw shank, screw

head, and longitudinal rod raises the construct stiffness higher than that

of the intact segment In the situation of loosening the screw shank/

head, the stiffness of the bridged segment was still higher than the

in-tact segment The stiffness-increasing effect suppresses the deformation

of the bridged segment, transferring the load to the adjacent segments

This can account for the higher kinematic and kinetic demand at the

ad-jacent segments after thefixation Another explanation is that the

Young's modulus of titanium-alloy pedicle screw is 110 GPa, which is

30 times stronger than the PEEK cage (E = 3.5 GPa) The screw system

enhanced the stiffness of the fused motion segment and then caused the

stress concentration at adjacent levels, especially cephalically

Compared with the fusion/fixation model, removing the fixator can

decrease the intersegmental angles, disc stresses, and facet loads by

33.3%, 24.1%, and 27.3% at the L3–L4 segment and 26.8%, 21.3%, and

26.6% at the L5–S1 segment, respectively In the literature, the reported

mitigation of ASD while using the dynamicfixator was still controversial

(Barrey et al., 2016; Galbusera et al., 2011; Hudson et al., 2011; Kim et al., 2011) Even though these are positive results, the kinematic and ki-netic increases of using the dynamicfixator only ranged between 12.3% and 21.2% for these numerical and experimental studies From our re-sults, the stress could be more equally distributed in adjacent segments after removal of the spinalfixators, and possibly have remarkable im-provement of the mitigation of ASD compared with the use of the dy-namicfixator In addition, the pedicle screw fixators are unable to induce bone remodeling, and have the problems such as implant failure due to fatigue (Chen et al., 2005) For better clinical outcomes, removal

of the pedicle screwfixators should be considered to decrease the pos-sibility of screw irritation Removal of the pedicle screwfixators by MISS technique is advantageous due to not only small incisions compared to a traditional large incision, but also significantly less secondary soft tissue injury

Due to the characteristics offinite-element simulation, there were some limitations inherent in this study Relative to the original CT-scan-ning data, some degenerative changes such as lordotic progression, facet hypertrophy, endplate sclerosis, and annular tears were not con-sidered in this study Due to the complexity of a partial resection of facet joints in MISS technique, the instability of the instrumented facet joints was not modeled for the sake of efficiency and avoidance of highly nonlinear simulation Little evidence can be used to validate the effects

of removing the spinalfixator by MISS technique Although this study had a limited number of case, the numerical results of were still repre-sentative Clinical and experimental studies should be conducted to val-idate thefindings of the current study

In conclusion, the hybrid use of fusion andfixation leads to signifi-cant increases in load of the adjacent tissues After successful fusion, the removal of the spinalfixator by MISS technique might be recom-mended as an option to effectively mitigate ASD progression in the ab-sence of spondylolysis and spondylolisthesis

Competing interests The authors declare that they have no competing interests in con-nection to this study

Authors' contributions CJC and SCL conceived of the study, participated in the design of the study and performed the data analyses YYH, FYT, CHC and LCW formu-lated the model and drafted the manuscript with the help of SCL All au-thors carried out thefinite-element analyses and approved the final manuscript

References

Barrey, C., Freitas, E., Perrin, G., 2016 Pedicle screw-based dynamic stabilization devices

in the lumbar spine: biomechanical concepts, technologies, classification, and clinical results Advanced Concepts in Lumbar Degenerative Disk Disease 633–664.

Bourgeois, A.C., Faulkner, A.R., Bradley, Y.C., Pasciak, A.S., Barlow, P.B., Gash, J.R., Reid Jr., W.S., 2015 Improved accuracy of minimally invasive transpedicular screw placement

in the lumbar spine with 3-dimensional stereotactic image guidance: a comparative meta-analysis J Spinal Disord Tech 28, 324–329.

Chen, C.S., Chen, W.J., Cheng, C.K., Jao, S.H., Chueh, S.C., Wang, C.C., 2005 Failure analysis

of broken pedicle screws on spinal instrumentation Med Eng Phys 27, 487–496.

Chien, C.Y., Kuo, Y.J., Lin, S.C., Chuang, W.H., Luh, Y.P., 2014 Kinematic and mechanical comparisons of lumbar hybrid fixation using Dynesys and Cosmic systems Spine

39, E878–E884.

Chuang, W.H., Lin, S.C., Chen, S.H., Wang, C.W., Tsai, W.C., Chen, Y.J., Hwang, J.R., 2012 Bio-mechanical effects of disc degeneration and hybrid fixation on the transition and ad-jacent lumbar segments Spine 37, E1488–E1497.

Chuang, W.H., Kuo, Y.J., Lin, S.C., Wang, C.W., Chen, S.H., Chen, Y.J., Hwang, J.R., 2013 Com-parison among load-, ROM-, and displacement-controlled methods used in the lum-bosacral nonlinear finite-element analysis Spine 38, E276–E285.

Corniola, M.V., Jägersberg, M., Stienen, M.N., Gautschi, O.P., 2015 Complete cage migra-tion/subsidence into the adjacent vertebral body after posterior lumbar interbody fu-sion J Clin Neurosci 22, 597–598.

Fig 5 The schematic diagrams to illustrate the load-transferring mechanism of the

Galbusera, F., Bellini, C.M., Anasetti, F., Ciavarro, C., Lovi, A., Brayda-Bruno, M., 2011 Rigid

and flexible spinal stabilization devices: a biomechanical comparison Med Eng Phys.

33, 490–496.

Highsmith, J.M., Tumialán, L.M., Rodts Jr., G.E., 2007 Flexible rods and the case for

dynam-ic stabilization Neurosurg Focus 15, E11.

Hikata, T., Kamata, M., Furukawa, M., 2014 Risk factors for adjacent segment disease after

posterior lumbar interbody fusion and efficacy of simultaneous decompression

sur-gery for symptomatic adjacent segment disease J Spinal Disord Tech 27, 70–75.

Hudson, W.R., Gee, J.E., Billys, J.B., Castellvi, A.E., 2011 Hybrid dynamic stabilization with

posterior spinal fusion in the lumbar spine SAS J 5, 36–43.

James, J.Y., William, L., 2015 Full endoscopic spinal surgery techniques: advancements,

indications, and outcomes Int J Spine Surg 9, 17.

Kim, C.H., Chung, C.K., Jahng, T.A., 2011 Comparisons of outcomes after single or

multilev-el dynamic stabilization: effects on adjacent segment J Spinal Disord Tech 24,

60–67.

Kim, Y.H., Jung, T.G., Park, E.Y., Kang, G.W., Kim, K.A., Lee, S.J., 2015 Biomechanical efficacy

of a combined interspinous fusion system with a lumbar interbody fusion cage Int.

J Precis Eng Manuf 16, 997–1001.

Kwon, D.W., Kim, K.H., Park, J.Y., Chin, D.K., Kim, K.S., Cho, Y.E., Kuh, S.U., 2013 Clinical

outcomes and considerations of the lumbar interbody fusion technique for lumbar

disk disease in adolescents Childs Nerv Syst 29, 1339–1344.

Lawrence, B.D., Wang, J., Arnold, P.M., Hermsmeyer, J., Norvell, D.C., Brodke, D.S., 2012.

Predicting the risk of adjacent segment pathology after lumbar fusion: a systematic

review Spine 37, S123–S132.

Lee, J.C., Kim, Y., Soh, J.W., Shin, B.J., 2014 Risk factors of adjacent segment disease requir-ing surgery after lumbar spinal fusion: comparison of posterior lumbar interbody fu-sion and posterolateral fufu-sion Spine 39, E339–E345.

Lequin, M.B., Verbaan, D., Bouma, G.J., 2014 Posterior lumbar interbody fusion with stand-alone trabecular metal cages for repeatedly recurrent lumbar disc herniation and back pain J Neurosurg Spine 20, 617–622.

Lu, K., Liliang, P.C., Wang, H.K., Liang, C.L., Chen, J.S., Chen, T.B., Wang, K.W., Chen, H.J.,

2015 Reduction in adjacent-segment degeneration after multilevel posterior lumbar interbody fusion with proximal DIAM implantation J Neurosurg Spine 23, 190–196.

Nakashima, H., Kawakami, N., Tsuji, T., Ohara, T., Suzuki, Y., Saito, T., Nohara, A., Tauchi, R., Ohta, K., Hamajima, N., Imagama, S., 2015 Adjacent segment disease after posterior lumbar interbody fusion: based on cases with a minimum of 10 years of follow-up Spine 40, E831–E841.

Niesche, M., Juratli, T.A., Sitoci, K.H., Neidel, J., Daubner, D., Schackert, G., Leimert, M., 2014.

Percutaneous pedicle screw and rod fixation with TLIF in a series of 14 patients with recurrent lumbar disc herniation Clin Neurol Neurosurg 124, 25–31.

Oh, K.W., Lee, J.H., Lee, J.H., Lee, D.Y., Shim, H.J., 2016 Jun 28 The correlation between cage subsidence, bone mineral density, and clinical results in posterior lumbar interbody fusion Clin Spine Surg (Epub ahead of print).

Okuda, S., Oda, T., Yamasaki, R., Maeno, T., Iwasaki, M., 2014 Repeated adjacent-segment degeneration after posterior lumbar interbody fusion J Neurosurg Spine 20, 538–541.