Reach Us +44-7447-215064
Implications of Decompressive Surgical Procedures for Lumbar Spine Stenosis on the Biomechanics of the Adjacent Segment: A Finite Element Analysis | OMICS International
ISSN: 2165-7939
Journal of Spine
Make the best use of Scientific Research and information from our 700+ peer reviewed, Open Access Journals that operates with the help of 50,000+ Editorial Board Members and esteemed reviewers and 1000+ Scientific associations in Medical, Clinical, Pharmaceutical, Engineering, Technology and Management Fields.
Meet Inspiring Speakers and Experts at our 3000+ Global Conferenceseries Events with over 600+ Conferences, 1200+ Symposiums and 1200+ Workshops on Medical, Pharma, Engineering, Science, Technology and Business

Implications of Decompressive Surgical Procedures for Lumbar Spine Stenosis on the Biomechanics of the Adjacent Segment: A Finite Element Analysis

Francesco Travascio1, Shihab Asfour1*, Joseph Gjolaj2, Loren L. Latta2, Shady Elmasry1 and Frank Eismont2
1Biomechanics Research Lab, Department of Industrial Engineering, University of Miami, Coral Gables, USA
2Department of Orthopaedics, Miller School of Medicine, University of Miami, Miami, FL, USA
Corresponding Author : Shihab Asfour
Department of Industrial Engineering
College of Engineering, University of Miami
Coral Gables, FL 33124-0621, USA
Tel: (305) 284-2367
Fax: (305) 284-4040
E-mail: [email protected]
Received February 12, 2015; Accepted March 17, 2015; Published March 19, 2015
Citation: Travascio F, Asfour S, Gjolaj J, Latta LL, Elmasry S (2015) Implications of Decompressive Surgical Procedures for Lumbar Spine Stenosis on the Biomechanics of the Adjacent Segment: A Finite Element Analysis. J Spine 4:220. doi:10.4172/2165-7939.1000220
Copyright: © 2015 Travascio F, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Visit for more related articles at Journal of Spine


Surgeries for Lumbar Spinal Stenosis (LSS) aim at decompressing spinal nerves and relieving symptoms of radiculopathy or myelopathy. Frequently after surgery, stenosis may progress in adjacent spinal segments, but the etiology of adjacent segment degeneration is still unclear. It is hypothesized that surgical approaches for LSS may alter the normal biomechanics of adjacent segments, eventually contributing to the development of stenosis. This study investigated implications of established decompressive surgical approaches on adjacent segments biomechanics. A realistic finite element model of a L1-L5 human lumbar spine was used for assessing changes in spine segments’ biomechanics due to laminotomy and laminectomy surgeries. First, the model was validated by comparing its predictions to previously reported spine kinematic data obtained after multi-level laminotomy and laminectomy. Subsequently, using a hybrid loading protocol, segments’ kinematics, intradiscal pressure, and stress in flexionextension were investigated simulating single level (L4-L5) laminotomy and laminectomy procedures. Alterations of spine segments biomechanics due to laminotomy were minimal. In contrast, after laminectomy, the L3-L4 range of motion, intradiscal pressure, and stress increased up to 50%, 20%, and 120%, respectively. These results suggest that laminotomy represents a better approach than laminectomy for reducing risks of spine instability or mechanically-accelerated disc degeneration in adjacent segments.

Lumbar stenosis; Laminotomy; Laminectomy; Adjacent segment degeneration; Finite element analysis
With a prevalence of approximately 20% in individuals older than 60 years, and up to 80% in those older than 70 years, Lumbar Spinal Stenosis (LSS) is exerting a greater clinical impact as the population ages [1,2]. The clinical presentation of LSS, defined as radiculopathy or myelopathy, is characterized by lower extremity pain, paresthesias and weakness and may also contribute to low back pain [3-5].
The pathogenesis of LSS is attributable to bone remodeling or overgrowth, intervertebral disc (IVD) protrusion, spondylolisthesis, or any combination of these [3]. Bone overgrowth is either initiated or accelerated by the degenerative process affecting facet joints and IVD [6]. Remodeling of the bone is either a reaction to the excessive joint motion or a physiologic attempt for local arthrodesis, eventually resulting in decreased segmental mobility. This loss of mobility in one segment creates abnormal forces and stresses on adjacent spinal segments, which then degenerate at an accelerated rate [3].
The LSS is often surgically treated. The objective of surgery is decompression of the spinal nerves without causing spinal instability [7]. In the past 60 years, a myriad of surgical techniques have been developed for achieving spinal nerve decompression. Among them, lumbar laminectomy with or without fusion are well-established approaches [8]. However, longitudinal studies on surgically treated patients report the occurrence of adjacent segment degeneration (e.g., disc herniation, spondylolisthesis, newly developed stenosis, etc., at adjacent spinal segments) after fusion or laminectomy [9-14]. Consequently, a large proportion of those patients require additional procedures to address the adjacent segment degeneration (ASD), especially if they experience symptoms of recurrent stenosis related to the ASD, which has been clinically defined as adjacent segment disease [11,13,15-17]. All surgical treatments for LSS involve alteration of the bony and soft tissue anatomy in the affected portion of the spine. The particular alterations to the musculoskeletal anatomy generated by each of these procedures may alter the normal physiological biomechanics of untreated segments of the spine [18,19]. Such alterations might have implications for the development of the adjacent segment disease.
Laminectomy or laminotomy are the preferred surgical approaches when there are no indications of pre-operative spinal instability [20-22]. The implications of such surgical approaches on the biomechanical behavior of the spine have been investigated via clinical [20-27] in vitro, [19,28-33] and numerical studies [34-40]. However, information on the specific alterations of adjacent spinal segment biomechanics due to these surgical procedures is still incomplete. Hence, the objective of this study was to provide new insights on the implications of laminectomy and laminotomy on the mechanical behavior of adjacent spinal segments.
A realistic computational model was developed to describe the biomechanical behavior of lumbar spine undergoing common surgical procedures such as unilateral laminotomy, bilateral laminotomy, and facet-sparing laminectomy. An additional procedure that is not typically performed clinically, laminectomy with complete facetectomy, was included for biomechanical comparison purposes. First, the model was validated by comparing its predictions of spinal segments motion to experimental data reported in an in vitro study [28]. Subsequently, the model was utilized for assessing and comparing post-operative changes of adjacent segment biomechanics in terms of segment kinematics (range of motion), intradiscal fluid pressure, and stress fields in IVD. Details on methods and procedures are reported below.
Lumbar spine computational model
A three-dimensional nonlinear finite element model of the lumbar spine was developed. It consisted of the L1 to L5 vertebrae, associated IVDs, intact facet joints, and all major ligaments of lumbar spine. The geometry of the computational domain was obtained from a CT of a normal, non-pathological spine. Vertebrae were modeled as rigid bodies. The IVDs were constituted by two distinct anatomic regions: the annulus fibrosus (AF) and the nucleus pulposus (NP). Both AF and NP were considered as biphasic media [41,42] constituted by a solid phase embedded in a fluid phase. More specifically, the solid phase of AF was modeled as a fiber reinforced hyperelastic composite: collagen fibers were modeled as tension-only elements [43] and arranged in a total of four concentric layers enclosing the NP with alternating ±30° orientation [44] the ground substance of AF was modeled as a Mooney- Rivlin material [45]. The solid phase of NP was isotropic elastic, with mechanical properties taken from a previous study [46]. Water volumetric fractions and hydraulic permeability for both NP and AF were those reported in the literature [47-49]. Each facet joint had a gap of 0.5 mm 44 and two cartilaginous layers which were modeled as elastic isotropic materials [50]. The ligaments were represented by linear elastic tension-only spring elements, and their stiffness was that reported by Pintar and co-workers [51]. A summary of the material properties used in this model is reported in Table 1.
Both IVD and cartilaginous layers at facet joints were modeled with 8-node hexahedral elements (~3200 elements for each IVD, and ~1000 element for each cartilage layer). Non-commercial software FEBio (FEBio 1.8.0, Musculoskeletal Research Laboratory, University of Utah, Salt Lake City, UT) was used to solve the set of governing equations defining the computational model. The FEBio software suite is a nonlinear implicit finite element framework designed specifically for analysis in computational solid biomechanics, whose accuracy and the robustness have been documented [52,53].
Simulated surgical procedures
Surgical procedures simulated in this study include unilateral laminotomy, bilateral laminotomy, facet-sparing laminectomy, and laminectomy with complete facetectomy. In this study, the ligamentum flavum at each spinal segment was modeled as composed of two spring elements (one element for each operative side). Accordingly, for unilateral laminotomy, only the spring element corresponding to the operative side was removed, together with part of the vertebral lamina. In contrast, for bilateral laminotomy, the entire ligamentum flavum connecting the two vertebral bodies (i.e., both spring elements) was removed. When simulating facet-sparing laminectomy, the entire lamina of the vertebra was removed, together with the connecting flavum, interspinous, and supraspinous ligaments. For the comparison case of laminectomy with complete facetectomy, in addition to all the steps performed in the case of facet-sparing laminectomy, the facet joints (including cartilaginous layers and capsular ligaments) were also removed. The spine models resulting from these procedures are shown in Figure 1.
Model validation
A preliminary validation was performed by comparing model predictions to experimental data reporting the effects of laminotomy and laminectomy on lumbar spine kinematics. Several experimental characterizations of spine biomechanics after decompressive surgeries for lumbar stenosis have been reported [19,29-32]. Each of these studies uses a different testing protocol, and addresses a specific subset of surgical approaches (e.g., facetectomy and laminectomy, graded facetectomy, bilateral laminotomy and laminectomy, etc.). Hence, quantitative information on spine biomechanics suitable for validating the model adopted in this study is fragmented. To the authors’ best knowledge, the in vitro analysis developed in Lee et al. [28] is the only one to report information on spine biomechanics after bilateral laminotomy and laminectomy in human spine, which is the standard for biomechanical evaluation. Accordingly, the experimental conditions used by Lee and co-workers were replicated in the simulations. More specifically, in the investigated cases, the inferior endplate of L5 was fixed (equivalent to potting the lumbar spine at L5), and a pure flexion/ extension moment was applied at the superior endplate of L1 (8 Nm in flexion and 6 Nm in extension, respectively) with a frequency of 1 Hz. In addition, a follower load of 400 N was applied to the spine as previously described [54]. Both laminotomy and laminectomy procedures were performed on L2-L5 segments (Figures 1b and 1c). The ranges of motion (rotations in the sagittal plane) of L2-L3, L3-L4, and L4-L5 segments were evaluated and compared to the in vitro results of Lee et al. [28]. In order to improve the agreement with the experiments, the initially chosen elastic moduli of some discs were slightly modified in the computational model within their physiological range. For all the cases investigated, it was found that the predicted range of motion of the model followed the same trend of in vitro data, and their differences were always less than one standard deviation (Figure 2).
Analysis of spine segments biomechanics
In this analysis, the surgical procedures of unilateral laminotomy, bilateral laminotomy, facet-sparing laminectomy, and laminectomy with complete facetectomy were performed at L4-L5, since this was assumed to be the spine level affected by pathology (Figures 1d-1g). The post-operative changes in range of motion (i.e., rotation in the sagittal plane, anteroposterior translation, and axial translation), intradiscal pressure, and normal and shear stress in both AF and NP were evaluated at all spine levels. A hybrid test method55 was adopted as a protocol for spine loading conditions. More specifically, the ‘intact’ spine was tested with the same loading conditions used for validation, and its total range of motion was computed. When testing the spine for each surgical procedure, the pure flexion/extension moment applied at L1 was varied in order to make the total range of motion equal to that attained in the ‘intact’ case.
The total range of motion of the spine (L1-L5) in the sagittal plane resulting from loading the intact model was 11.46° for flexion and 14.3° for extension. The moments required to produce the same range of motion after performing the surgical procedures are shown in Table 1. Moments changed during flexion, decreasing up to 42% for the case of laminectomy with facetectomy. Conversely, minimal changes were found during extension for all procedures investigated.
Post-operative alterations of spinal segments biomechanics during extension were minimal (<5%) and are not reported. The post-operative motion redistribution during flexion for the individual spine segments is reported in Figure 3, and compared to the ‘intact’ case. For all procedures, sagittal rotations increased at L4-L5 and L3- L4, and decreased at L2-L3 and L1-L2. Major changes were found after laminectomies, with increments up to 18% and 23% (at L3-L4 and L4-L5, respectively), and reductions up to 15% and 39% (at L2- L3 and L1-L2, respectively). In contrast, post-operative changes after either unilateral or bilateral laminotomy were minimal (<5%), (Figure 3a). Similar trends were observed in the anteroposterior translations: after all the procedures, translations increased at L4-L5 and L3-L4, and decreased at the above segments. The only exception was found at L2- L3, where unilateral and bilateral laminotomy caused anteroposterior translation to increase up to 2.52 mm (18%) and 2.46 mm (15%), respectively (Figure 3b). For all the procedures investigated, increments in the axial compression did not exceed 0.2 mm (Figure 3c).
Post-operative alterations of spinal segments kinematics were reflected in changes of intradiscal pressure and stresses in the IVDs. After laminectomy procedures, intradiscal pressure increased in both NP and AF at L3-L4 (up to 20%) and L4-L5 (up to 10%). Conversely, at L2-L3 and L1-L2, pressure reduced up to 35% and 31%, respectively (Figure 4). After either unilateral or bilateral laminotomy, pressure changes were minor at all spine levels, with the exception of L3-L4, whose fluid pressure in AF dropped up to 30% (Figure 4b). Changes in the normal stresses were similar to those found in intradiscal pressure: after laminectomy procedures, stress in both NP and AF increased one-fold in L4-L5 and L3-L4, and decreased up to 30% to 35% in L2- L3 and L1-L2, respectively, (Figure 5a and 5b). Again, after unilateral and bilateral laminotomy, no major changes from the ‘intact’ case were observed for all the spine levels. Major changes in shear stress were only observed in the NP of L3-L4 after laminectomy procedures, increasing up to 120% the value attained in the ‘intact’ case (Figure 5c and 5d).
In this study, we adopted a realistic three-dimensional finite element model of human lumbar spine to investigate the implications of surgical procedures for lumbar stenosis on the biomechanics of the adjacent segments. Specifically, the model was implemented to simulate biomechanical tests on a L1-L5 spinal column undergoing unilateral laminotomy, bilateral laminotomy, facet-sparing laminectomy, and laminectomy with facetectomy at L4-L5 to yield changes in kinematics, intradiscal pressure, and disc stress at all spine levels. Such metrics are especially relevant when investigating the etiology of ASD since altered range of motion of spine segments is believed to increase the risk of spinal instability, eventually leading to spondylolisthesis and LSS [3]. Besides, abnormal levels of fluid pressure or stress may suggest ongoing IVD degeneration, which also contributes to the development of stenosis [3,6,55-57].
The post-operative changes of spinal segments biomechanics were tested during flexion/extension. In agreement with both in vitro [28,30,32] and numerical studies, [34,39,40] no changes were observed during extension. The kinematic analysis carried out in this study shows that the largest increase in post-operative spine motion is attained at the operated level (L4-L5) and the immediate adjacent one (L3-L4) after laminectomies are performed (Figure 3). This is in agreement with previously reported in vitro [28-30] and numerical biomechanical analyses [34,35,40]. Moreover, these results are also consistent with clinical studies observing that laminotomy generates a lower level of instability when compared to laminectomy [20,25]. Post-laminotomy alterations of the kinematics at the levels L3-L4 and above are caused by the reduction of stiffness at L4-L5. In contrast, laminectomy of L4 also entails the removal of flavum, interspinous, and supraspinous ligaments, which connect this vertebra to both L3 and L5. These anatomic changes directly affect the stiffness of the adjacent segment L3-L4. It has been reported that over 60% of the flexion movement of the spine is taken up by the posterior ligaments. Among them, interspinous and supraspinous ligaments withstand the highest tensile force [58]. This would explain the fact that: (1) after laminotomy, alterations of L4-L5 kinematics were minor; (2) after laminectomy, the model predicted increase in motion during flexion and almost negligible changes during extension.
Changes in intradiscal pressure and stress can lead to altered metabolism within the disc, with potential long-term disc degeneration [56,57,59,60]. Minor changes are found after either unilateral or bilateral laminotomy (<10%). In contrast, after laminectomy, variations of fluid pressure (up to 20%) and stress (up to 120%) occur in NP and anterior AF (Figures 4 and 5). However, these changes occurred at the operated level (L4-L5) and its immediate adjacent level L3-L4, while the other spine levels experienced reduction of both intradiscal pressure, and normal and shear stresses. These results are in agreement with an in vitro study on calf spine reporting that, after laminotomy, intradiscal pressure changes at the operated level did not statistically differ from those found in intact spine. In contrast, after laminectomy, significant pressure increase was found in the anterior portion of IVD [29]. Similar findings were also reported in a recent computational study showing intradiscal pressure increase up to 50% after laminectomy, and minor changes after laminotomy [34].
It has been historically reported that laminectomy with complete facetectomy induces excessive spinal instability, so that the more conservative facet-sparing laminectomy is typically performed to treat LSS [27,32,36-38]. This study confirms that, compared to facetsparing laminectomy, the complete facetectomy model yielded a larger increase in spine kinematics and, consequently, larger intradiscal pressure and stress at the operated level L4-L5. However, at the immediate adjacent level L3-L4, facet-sparing laminectomy caused the largest biomechanical alterations (Figures 4 and 5). Hence, according to model’s predictions, the two laminectomy procedures are similarly detrimental for spine health, with laminectomy with facetectomy mostly altering the biomechanics of the operated level L4-L5, while facet-sparing laminectomy mainly affecting the adjacent segment L3-L4.
The advancements in minimally-invasive spine surgery have been promoted as a potential way to decrease the rate of ASD by diminishing paraspinal muscle damage and avoiding disruption of the midline structures that provide stability. While this topic has been recently studied in the context of spinal fusion surgery (open versus minimally invasive spinal fusion techniques), there is little known about the comparative rate of ASD in open versus minimally invasive spinal decompression surgery alone [61-64]. Multiple authors have reviewed outcomes of open versus minimally-invasive fusion transforaminal lumbar interbody fusion (TLIF) and have found a lower rate of ASD in the minimally-invasive groups, presumably due to less soft tissue dissection (e.g., less paraspinal muscle stripping) [63,64]. Radcliff and colleagues have reviewed the rate of ASD amongst various lumbar interventions and noted a rate of ASD of 2-3% per year [61]. The same authors performed a subsequent study of their own patients who underwent anterior lumbar interbody fusion (ALIF) and supplemental posterior instrumentation performed either open or percutaneously. The results of their study did not show a difference in rate of ASD between the two groups [62]. So while there is a strong theoretical advantage to minimally invasive spine surgical techniques, the clinical evidence that it reduces the rate of ASD is still somewhat equivocal. There is clearly a need for more comparative clinical studies reviewing this topic. Future directions for our present study will include biomechanical comparison using the same modeling techniques to determine the difference in adjacent level motion when various spinal stabilization/fusion techniques are applied, (i.e., posterior pedicle screw and rod instrumentation, interbody placement, etc).
Some limitations of this study must be noted. The model schematizes vertebrae as rigid bodies, so that the only deformable structures in the spine are the soft tissues (i.e., intervertebral discs, cartilage at facet joints, and ligaments). Such simplification may have affected the results of both kinematic and stress analyses hereby reported. However, the stiffness of the soft tissues in the spine is about two and four orders of magnitude lower than those of cancellous and cortical bone in vertebrae, respectively [65]. Accordingly, one would expect that, for the surgical procedures and loading conditions investigated in this study, spine strains mostly occur in the soft tissues. Also, spine ligaments were modeled as linear elastic elements, whose stiffness corresponded to the slope of the most linear portion of the force-deformation curve experimentally determined by Pintar and co-workers [51]. Ligaments linear behavior is considered the normal (physiologic) response of the tissue to routine external stimuli [66,67]. Accordingly, a linear behavior may be used as an initial approximation of ligament characteristics in computational models [51]. Moreover, another factor potentially affecting spine stability is the extent of paraspinal muscle damage associated to the specific surgical procedure performed. However, the contribution of muscles to spine biomechanical stability was not accounted for in the present finite element model, and its inclusion will be addressed in our future studies. Also, laminotomy and laminectomy are characterized by a similar degree of paraspinal muscle dissection. Finally, the computational model used in this study was validated through kinematic data from an in vitro study only reporting spine kinematics during flexion/extension [28]. Accordingly, the results reported in this study are only relevant for the case of flexion/extension spine motion, since other physiologically relevant movements (e.g., axial rotation, lateral bending, etc.) were not studied, and will be addressed in the future upon further model validation.
For the loading conditions investigated in this analysis, our results suggest that laminotomy, whether unilateral or bilateral, represents a superior technique in terms of potential risk reduction for developing either spine instability or mechanically-accelerated disc degeneration in the adjacent segment. However, additional tests, under different and more complex physiologically relevant loading conditions, should be performed in order to confirm our findings. Moreover, it is recognized that surgical decision-making must take into account many other factors, among which the severity of the stenosis. While laminotomy has been recommended for cases of moderate or unilateral stenosis, and it might not allow for adequate decompression of severe central or bilateral stenosis [25] in which case laminectomy may represent a better surgical solution despite the increase in instability shown in our study.
Study supported by funds donated to Biomechanics Research Group of the University of Miami. The authors have no conflict of interest to disclose.
Select your language of interest to view the total content in your interested language
Post your comment

Share This Article

Article Usage

  • Total views: 12885
  • [From(publication date):
    April-2015 - Feb 25, 2020]
  • Breakdown by view type
  • HTML page views : 9040
  • PDF downloads : 3845