- Email: [email protected]
A novel anchoring system for use in a nonfusion scoliosis correction device
Accepted Manuscript A novel anchoring system for use in a non-fusion scoliosis correction device M. Wessels, PhD J.J. Homminga, PhD E.E.G. Hekman, MSc G.J. Verkerke, PhD PII:
S1529-9430(14)00461-6
DOI:
10.1016/j.spinee.2014.04.028
Reference:
SPINEE 55879
To appear in:
The Spine Journal
Received Date: 5 March 2013 Revised Date:
19 March 2014
Accepted Date: 16 April 2014
Please cite this article as: Wessels M, Homminga J, Hekman E, Verkerke G, A novel anchoring system for use in a non-fusion scoliosis correction device, The Spine Journal (2014), doi: 10.1016/ j.spinee.2014.04.028. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title A novel anchoring system for use in a non-fusion scoliosis correction device
RI PT
Authors
Wessels M, PhD,* Homminga JJ, PhD,* Hekman EEG, MSc,* Verkerke GJ, PhD,*†
Academic affiliation
SC
* Laboratory of Biomechanical Engineering, Faculty of Engineering Technology, University of Twente,
M AN U
The Netherlands
† Department of Rehabilitation Medicine, University of Groningen, University Medical Center Groningen, The Netherlands
Corresponding author
HR W 207, PO box 217
TE D
Martijn Wessels
EP
7500 AE Enschede, The Netherlands
AC C
Email: [email protected]
Tel: 0031 53489 2351
ACCEPTED MANUSCRIPT
Abstract
2
Background context
RI PT
1
Insertion of a pedicle screw in the mid and high thoracic region has a serious risk of facet joint
4
damage. Since flexible implant systems require intact facet joints, we developed an enhanced fixation
5
that is less destructive to spinal structures. The XSFIX is a posterior fixation system that uses cables that
6
are attached to the transverse processes of a vertebra.
7
Purpose
9
M AN U
8
SC
3
Purpose of this study is to determine whether a fixation to the transverse process using the XSFIX is strong enough to withstand the loads applied by the XSLATOR (a novel, highly flexible non-fusion implant system) and thus whether it is a suitable alternative for pedicle screw fixation.
11
Study Design/Setting
12
TE D
10
The strength of a novel fixation system using transverse process cables was determined and compared to the strength of a similar fixation using polyaxial pedicle screws on different vertebral levels.
14
Methods
EP
13
Each of 58 vertebrae, isolated from four adult human cadavers was instrumented with either a
16
pedicle screw anchor (PSA) system or a prototype of the XSFIX. The PSA consisted of two poly-axial
17
pedicle screws and a 5 mm diameter rod. The XSFIX prototype consisted of two bodies that were fixed to
18
the transverse processes, interconnected with a similar rod. Each fixation system was subjected to a
19
lateral or an axial torque.
AC C
15
1
ACCEPTED MANUSCRIPT
1 2
Results The PSA demonstrated fixation strength in lateral loading and torsion higher than required for use in the XSLATOR. The XSFIX demonstrated high enough fixation strength (in both lateral loading and
4
torsion), only in the high and mid thoracic region (T10-T12).
5
Conclusions
6
RI PT
3
This experiment showed that the fixation strength of XSFIX is sufficient for use with the XSLATOR only in mid and high thoracic region. For the low thoracic and lumbar region, the PSA is a more rigid
8
fixation. Since the performance of the new fixation system appears to be favorable in the high and mid
9
thoracic region, a clinical study is the next challenge.
11
M AN U
10
SC
7
Introduction
Pedicle screws are commonly used for posterior anchoring of spinal implants.1-3 Although pedicle screws are now commonly used on the lumbar and thoracic spine,2 the shape of the pedicles in the
13
thoracic region is inconsistent, especially in spinal deformities such as scoliosis.4,5 As a result of this,
14
pedicle screw placement is very challenging in this region.6-9 Currently, pedicle screws are inserted by
15
using either the free hand technique (as in the Roy-Camille procedure), the open lamina technique
16
where a partial laminectomy is performed, or under application of an image-guidance system. In each of
17
these techniques, required accuracy is very high.10,11 The pedicle entry point for a high or mid thoracic
18
pedicle screw is situated at the base of the superior facet.12,13 Even a well-placed pedicle screw may still
19
violate the facet joint. Particularly in the thoracic region, the size and position of the screw head can
20
result in obstruction of movements of the facet joints. Facet joint violation in pedicle screw anchoring in
21
the thoracic and lumbar region is a common occurrence in spinal surgery.14,15 The effect of limited
AC C
EP
TE D
12
2
ACCEPTED MANUSCRIPT
motion of the joints, such as in spondylodesis is often desirable , but particularly troublesome in the
2
recent trend toward dynamic, non-fusion systems that aim to maintain the flexibility of the spine.
3
Consequently, a demand for an enhanced fixation, that is less destructive to spinal structures, has grown.
4
RI PT
1
For these reasons, a novel fixation system, referred to as XSFIX, was developed; a bridge
construction that consists of two anchors, with a bridge in between. Each of these anchors consists of a
6
connector that is attached to a transverse process by means of a cable. This new fixation system can be
7
applied for non-fusion spinal implant systems that will correct spinal deformities e.g. scoliosis. The
8
XSLATOR, a non-fusion scoliosis correction system which is currently being developed (Figure 1), is able
9
to lengthen (cranially-caudally) to allow for spinal growth and daily motion (including flexion/extension)
M AN U
SC
5
by making use of sliding bearings. The axial forces (caused by all spinal motion including growth) and
11
sagittal forces (caused by flexion/extension) remain extremely low, while lateral force and axial torque
12
are significant but low.16 To consider the XSFIX as a solid fixation, the anchor must be able to transfer the
13
forces and moments forces delivered by the XSLATOR, which are (including a design safety factor of 2) a
14
maximal lateral force of 200 N and a maximal axial torque (around the local spinal longitudinal axis) of 7
15
Nm. 16 These loads are smaller than the loads applied by fusion systems, because the non-fusion implant
16
systems use the visco-elastic properties of the spine to correct the deformity.
EP
TE D
10
Purpose of this study is to determine whether our new XSFIX fixation system is strong enough to
18
withstand the loads applied by the XSLATOR and thus whether it is a suitable alternative for a similar
19
bridge construction that uses pedicle screws.
20
Materials and Methods
AC C
17
3
ACCEPTED MANUSCRIPT
1
60 Vertebrae (42 thoracic and 18 lumbar; table 2) were isolated from four adult human cadavers (aged between 84 and 98 years) and cleaned of all soft tissues. All specimens were kept frozen until
3
testing. Prior to the experiment, all vertebrae were DXA-scanned to determine the bone mineral density
4
scores. All spines were identified as osteoporotic (Table 1); T-scores were well below -2.5.17
5
RI PT
2
Each vertebra was instrumented with either a pedicle screw anchor (PSA) system or a prototype of the XSFIX system (Table 2). The PSA consisted of two poly-axial pedicle screws (one in each pedicle) that
7
were rigidly interconnected with a 5 mm diameter rod. The XSFIX prototype consisted of two stainless
8
steel bodies that were each fixed to the left and right transverse processes of the vertebrae with black
9
Nylon 6-6 cable ties (5mm wide, tensile strength 250 N), and were rigidly interconnected with a 5 mm
M AN U
SC
6
10
diameter rod (Figure 2). Tension in the cable ties was consistently applied (at 80 N) using a cable tie
11
installation tool.
12
Before testing, each vertebra was embedded with PMMA bone cement in a metal housing that was rigidly clamped onto the testing setup. Each fixation system was subjected to one of two loading
14
conditions until failure.
TE D
13
A lateral force (in the local coordinate system) was applied to the bridge, with a loading rate of
16
1.0 N/s. An axial torque (in the local coordinate system) was applied to the bridge, with a loading rate of
17
0.05 Nm/s (Figure 3). In the test setup a narrow or a wide fork construction was used to generate torque,
18
depending on the available space (Figure 4).
20
AC C
19
EP
15
A hydraulic tensile/compression machine (MTS, Berlin, Germany) was used to generate the appropriate loading on the bridge. Time and loads were recorded digitally with a sample rate of 2 Hz. The
4
ACCEPTED MANUSCRIPT
1
vertebrae from the four spines were alternately assigned to a type of fixation and to a loading condition
2
(Table 2).
Failure was defined as the point at which the displacement suddenly increased (as shown by the
4
rotation-time graphs (Figure 5a and Figure 5b). The absolute maximal force for each construct was also
5
measured, which represented the point of complete separation of the fixation system from the vertebra.
6
However, as clinically relevant failure clearly started long before that point, the initiation of failure seems
7
a more relevant strength criterion.
SC
RI PT
3
For statistical analysis, no difference in vertebral strength between spines was assumed. Results
9
were analyzed based on four regions: high thoracic (HT): T1-T5; mid thoracic (MT): T6-T10; low thoracic
10
(LT): T11-T12 and lumbar (L): L1-L5. The anatomy of T11 and T12 was markedly different from the other
11
thoracic vertebrae by poorly articulated transverse processes (which are essential for attaching the XSFIX
12
system). Because of this, T11 and T12 represent a separate group. Fixation failure strengths for torque
13
and lateral force were compared between vertebral levels and between fixation techniques using
14
ANOVA (with significance set at p=0.05).
TE D
M AN U
8
The overall performance of XSFIX and PSA was assessed by assigning three grades to each spinal
16
region. The grades ‘poor’, ‘good’ and ‘very good’ were used to evaluate on fixation strength, facet joint
17
damage, and placement. Fixation strength was assessed using the mean (Mean) and standard error (SE)
18
values for each VB level.
AC C
EP
15
19
For each loading condition, the fixation strength was graded:
20
•
‘poor’ if Mean – SE < RV, 5
ACCEPTED MANUSCRIPT
1
•
‘good’ if 1 ·RV ≤ Mean - SE < 2·RV, and
2
•
‘very good’ if 2·RV ≤ Mean – SE,
where RV is the required value (RV = 200 N for lateral force and RV = 7 Nm for torsion).
RI PT
3
For all four VB levels, the lowest grades of PSA and XSFIX, which represent the limiting factor in
5
fixation strength, were selected. For example, if a grade for XSFIX at the LT level was ‘poor’ for torsion
6
and ‘sufficient’ for lateral force, then fixation strength was rated ‘poor’.
SC
4
Facet joint damage was graded:
8
•
‘poor’ if the facet joint was intrinsically damaged,
9
•
‘good’ if the facet joint remained unharmed but with minor risk of joint obstruction, and
10
•
‘very good’ if no damage or possible obstruction occurred.
11
Placement of the fixation was graded:
12
•
‘poor’ if fixation was not possible,
13
•
‘good’ if placement was possible and straightforward but with risk of structural damage, and
14
•
‘very good’ if placement was very easy with very low risk of structural damage.
16
TE D
EP
AC C
15
M AN U
7
Results
The PSA typically showed slow migration of the screws until the structures ruptured completely. The
17
XSFIX generally showed large deformation of the vertebral arc before actual rupture of the structures.
18
Both fixation types typically show a point of sudden increase in rotation (or displacement), see Figure 5a
6
ACCEPTED MANUSCRIPT
1
and Figure 5b, indicating an early collapse of bone structures, defining the clinically relevant point of
2
fixation failure.
No differences in fixation strength for either spinal region were found between spines loaded with
4
lateral force (p>0.445) or torsion (p>0.13). General type of failure for XSFIX in all spines was transverse
5
process failure in axial torsion and pedicle failure in lateral loading. General types of failure for PSA in all
6
spines were pedicle screw migration and vertebral body rupture (Figure 6). Anchoring the XSFIX system
7
to T11 and T12 was problematic and sometimes impossible as the cables slipped off the transverse
8
processes when tightened. Although mean strength difference between XSFIX and PSA in both lateral
9
loading and in axial torsion appeared large, differences showed no significance due to the small sample
11
SC
M AN U
10
RI PT
3
size in the LT region.
In lateral loading, both the PSA and the XSFIX demonstrated fixation strengths above 200 N that are typically expected for a non-fusion implant system such as the XSLATOR. The fixation strength of the
13
XSFIX is significantly lower than that of the PSA in both the high thoracic region (p=0.019) and the lumbar
14
region (p=0.045). In the mid and low thoracic levels the mean strength of XSFIX is not significantly
15
different from the strength of the PSA system (p=0.52, Table 3 and Figure 7). The small sample size of the
16
LT region level causes the mean difference to be not significant (Figure 7a).
EP
In all spinal regions, the PSA demonstrated fixation strengths in torsion above 7 Nm, which was
AC C
17
TE D
12
18
estimated to be maximally generated by the non-fusion implant system. The XSFIX demonstrated fixation
19
strengths in torsion above 7 Nm only in the high and mid thoracic regions. In the lumbar region, the
20
fixation strength of the XSFIX significantly lower than that of the PSA (p=0.015) and lower than 7 Nm. In
21
the high and mid thoracic levels the mean strength of the XSFIX in torsion is not significantly different 7
ACCEPTED MANUSCRIPT
from the strength of the PSA system (p=0.989). The fixation strength (torsion) of the XSFIX was
2
significantly lower in the lumbar region compared to both the high thoracic (p=0.024) and mid thoracic
3
region (p=0.041, Table 3 and Figure 7b). In the mid thoracic region, mean fixation strength difference (in
4
torsion) is large, however not significant caused by small sample size (Figure 7b).
5
In Table 4, a schematic overview is given of the assessment of the XSFIX and PSA in respect to the
6
different spinal regions. Figure 8 shows the cumulated scores of both fixation systems for each vertebral
7
level. Evidently, the scores of the PSA are higher than the XSFIX for the low thoracic and lumbar regions.
8
The XSFIX shows higher accumulated scores than the PSA for the high and mid thoracic regions.
9
Discussion
SC
M AN U
10
RI PT
1
Some limitations apply to this study. First, all tested vertebrae were osteoporotic; the results may differ for healthy vertebrae. However, both the PSA and the XSFIX obtain their strength from bone
12
structures of the vertebral arc, which are relatively unaffected by osteoporosis.18 Furthermore, this test
13
can be seen as a worst-case scenario. It is reasonable to assume that the relative performance of the
14
new fixation system in relation to a similar fixation using pedicle screws is comparable to the
15
performance in adolescent bone.
EP
Second, both the XSFIX and the PSA were placed without the physical limitations that are present
AC C
16
TE D
11
17
during in vivo surgery. Pedicle screws were placed using visual cues not available during surgery. The
18
XSFIX was also easier to attach, due to the removal of the ribs. If the clinical reality prevents proper
19
placement of these systems, failure strengths may be different.
8
ACCEPTED MANUSCRIPT
1
Third, in this experiment the pedicle screw fixation (PSA) consisted of a bridge construction. Clinically, pedicle screws are used as stand-alone systems. A single screw will transfer a torque directly to
3
the vertebra, resulting in high peak pressures and increased risk of early fixation failure. Our test results
4
for the PSA cannot be compared to single pedicle screw arrangements; though it can be stated our
5
results will give much higher strength values.
6
RI PT
2
We defined failure strength as the point at which a sudden increase in displacement (rotation) took place. Because the PSA showed migration of the screws and the XSFIX showed large unrecoverable
8
deformation of the vertebral arc before actual rupture of the structures, we attribute the sudden
9
increase in deformation to microscopic bone failure. Nonetheless, the failure strength criterion we used
11
M AN U
10
SC
7
seems clinically most relevant.
In some cases, premature XSFIX failure occurred in the small ratchet of a cable tie. An improved version of the ratchet that consists of high strength materials will prevent early fixation failure. In
13
addition, the material aspect needs further attention because the material used in our ties (Polyamide
14
6/6) is not suitable for implantation due to issues with creep, relaxation, and biocompatibility. Replacing
15
the material for a creep resistant and biocompatible material such as carbon fiber reinforced PEEK
16
(polyetheretherketone) may well be considered.19,20
EP
TE D
12
In the high and mid thoracic regions the torsion strength of the XSFIX is higher than the required
18
value of 7 Nm for use with the non-fusion implant system such as the XSLATOR. In the low thoracic and
19
lumbar regions, the torsion strength of the XSFIX is lower than the required 7 Nm, but in these regions,
20
the PSA is strong enough. Lateral fixation strength of the XSFIX meets the requirement of 200 N for use
21
with the non-fusion implant system for all spinal regions. However, the anchoring of the XSFIX to T11 and
AC C
17
9
ACCEPTED MANUSCRIPT
T12 is problematic, due to the poorly pronounced transverse processes of those vertebrae.
2
Consequently, the XSFIX is not very suitable for those vertebrae. Similarly, the lumbar vertebrae have
3
long transverse processes that are very slender and fragile, making the lumbar vertebrae not suited for
4
use of the XSFIX. In fact, some of the lumbar transverse processes broke during tensioning of the cable
5
ties. Fortunately, pedicle screws can be inserted easily into these vertebrae without obstructing or
6
damaging the facet joints.
RI PT
1
The highly flexible implant XSLATOR requires three vertebral levels for fixation.16 Either PSA or XSFIX
8
must be used for anchoring. The force requirements for the fixation are based on the maximal forces on
9
the center vertebra. The actual force on the other two vertebrae is only 50% of the referred maximal
10
force. The XSLATOR therefore cannot be compared to other (fusion or non-fusion) scoliosis correction
11
constructs.
M AN U
SC
7
The grading system used in table 4 is rather subjective. However, the grading system was used
13
merely to illustrate our conclusions. Despite the subjectivity of the grading, we think the outcome
14
regarding comparison between both systems will be similar. Nevertheless, Table 4 shows that the XSFIX
15
is an excellent fixation for the high and mid thoracic regions. Especially in this region, the application of
16
pedicle screws is risky, because of the small diameter of the pedicles. Therefore, the XSFIX could play an
17
important role in the fixation of devices to the spine. In contrast, grades for fixation strength and
18
placement of XSFIX in the low thoracic and lumbar region are valuated much lower, whereas PSA
19
produced decent grades for those regions.
20
With this experiment, we have a strong indication that the XSFIX is comparable to the pedicle screw
21
performance (strength, placement and joint damage) in the mid and high thoracic region and that for the
AC C
EP
TE D
12
10
ACCEPTED MANUSCRIPT
lumbar region the pedicle screw is the method of choice. With this outcome, animal experiments can be
2
started to gather statistical evidence that the XSFIX is indeed preferred for fixation in the mid and high
3
thoracic region. Therefore, the experiments described in this manuscript contributed to a restriction of
4
animal experiments by giving these experiments a specific focus.
5
RI PT
1
In summary, this experiment showed that the XSFIX is a solid fixation with strength that is
comparable to the PSA strength in torsion, when placed on a vertebra in the T2-T10 region. Fixation
7
strength of XSFIX is sufficient for use with a non-fusion implant system, such as the XSLATOR, in mid and
8
high thoracic regions. For the low thoracic and lumbar region, the PSA is a more rigid fixation. Since the
9
performance of the new fixation system appears to be favorable in the high and mid thoracic region, a
10
clinical study is the next challenge.
11
References
1. Steffee AD, Biscup, RS, Sitkowskj DJ. Segmental spine plates with pedicle screw fixation - a new
TE D
12
M AN U
SC
6
internal fixation device for disorders of the lumbar and thoracolumbar spine. Clin Orthop Relat Res
14
1986;203:45-53.
16
17 18
19 20
2. Gaines RW Jr. The use of pedicle-screw internal fixation for the operative treatment of spinal disorders. J Bone Joint Surg Am 2000;82:1458-76.
AC C
15
EP
13
3. Cotrel Y, Dubousset J, Guillaumat M. New universal instrumentation in spinal surgery. Clin Orthop Relat Res 1988;227:10-23.
4. Zeiller S, Lee J, Lim M, Vaccaro A. Posterior thoracic segmental pedicle screw instrumentation: Evolving methods of safe and effective placement. Neurology India 2005;53:458-65. 11
ACCEPTED MANUSCRIPT
3 4
5 6
7
instrumentation: critical appraisal of current "state-of-art". Neurology India 2009;57:715-21.
6. Vaccaro AR, Rizzolo SJ, Allardyce TJ, et al. Placement of pedicle screws in the thoracic spine. Part I:
RI PT
2
5. Mattei T, Meneses M, Milano J, et al. "Free-hand" technique for thoracolumbar pedicle screw
Morphometric analysis of the thoracic vertebrae. J Bone Joint Surg Am 1995;77:1193-9.
7. Vaccaro A.R., et al. Placement of pedicle screws in the thoracic spine. Part II: An anatomical and radiographic assessment. J Bone Joint Surg Am 1995;77:1200-6.
SC
1
8. Xu R, Ebraheim NA, Ou Y, et al. Anatomic considerations of pedicle screw placement in the thoracic spine. Roy-Camille technique versus open-lamina technique. Spine 1998;23:1065-8.
9
9. Lonstein JE, et al. Complications associated with pedicle screws. J Bone Joint Surg Am
13 14
15 16
17 18
TE D
12
10. Rampersaud YR, Simon DA, Foley KT. Accuracy requirements for image-guided spinal pedicle screw placement. Spine 2001;26:352-9.
11. Papadopoulos EC, Girardi FP, Sama A, et al. Accuracy of single-time, multilevel registration in image-guided spinal surgery. Spine J 2005;5:263-7.
EP
11
1999;81:1519-28.
12. Chung KJ, Suh SW, Desai S, Song HR. Ideal entry point for the thoracic pedicle screw during the
AC C
10
M AN U
8
free hand technique. Int Orthop 2008;32(5): 657-662.
13. Modi HN, Suh SW, Fernandez H, et al. Accuracy and safety of pedicle screw placement in neuromuscular scoliosis with free-hand technique. Eur Spine J 2008;17(12):1686-1696.
12
ACCEPTED MANUSCRIPT
1
14. Chung KJ, Suh SW, Swapnil K, et al. Facet joint violation during pedicle screw insertion: a cadaveric study of the adult lumbosacral spine comparing the two pedicle screw insertion techniques. Int
3
Orthop 2007;31:653-656.
4
RI PT
2
15. Chen Z, Zhao J, Xu H, et al. Technical factors related to the incidence of adjacent superior
segment facet joint violation after transpedicular instrumentation in the lumbar spine. Eur Spine J
6
2008;17(11):1476-1480.
8
9 10
16. Wessels M, Hekman EEG, Verkerke GJ. Mechanical behavior of a novel non-fusion scoliosis correction device. J Mech Behav Biomed 2013;27:107-114.
M AN U
7
SC
5
17. WHO. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis. Report of a WHO Study Group. World Health Organ Tech Rep Ser 1994;843:1-129.
18. Riggs BL. Overview of osteoporosis. West J Med 1991;154:63-77.
12
19. Kurtz SM, Devine JN. PEEK biomaterials in trauma, orthopedic, and spinal implants. Biomaterials
14
2007;8:4845-69.
20. Toth JM, et al. Polyetheretherketone as a biomaterial for spinal applications. Biomaterials 2006;27:324-34.
16
Figure Legends
17
AC C
15
EP
13
TE D
11
Figure 1: The XSLATOR consists of two implants that are posteriorly fixed to the vertebrae. The three
18
bridges can relatively slide axially keeping axial forces extremely low. The bridges are fixed to the spine
19
using transverse process cables (XSFIX; right, top) or 2 pedicle screws (right, bottom). 13
ACCEPTED MANUSCRIPT
2
3 4
5
Figure 2: The two different fixations were attached to the vertebrae and the vertebrae were fixed in cups with PMMA. (Left) Vertebra with the XSFIX. (Right) Vertebra with the PSA in setup.
Figure 3: Loading with a vertical (lateral) force on XSFIX (left), PSA (middle), and torque around a
RI PT
1
vertical axis on PSA (right) was performed using a hydraulic tensile/compression machine.
Figure 4: The cemented cups were clamped into a framework, which was clamped onto a table. Torque was applied either between the cables (left) using a narrow rotating fork or on the ends of the
7
rod (right) using a wide rotating fork.
Figure 5a: Typical rotation-time curve of a XSFIX fixation. The rotation-time curve shows a sudden
M AN U
8
SC
6
9
increase in rotation at time=385 s, indicating a collapse of structures. At time=488 s one of the
10
transverse process is separated from the vertebral arc resulting in a sharp increase in rotation.
11
Figure 5b: Typical rotation-time curve of a PSA fixation. Also here, at time=998 s, a sudden increase of rotation can be seen in the curve (defined as the failure point). At time=1851 s, the vertebral body
13
ruptured.
15
Figure 6: (Left) Transverse process rupture, indicated with a white arrow, in torsion test with XSFIX.
EP
14
TE D
12
(Right) Vertebral body rupture in torsion test with PSA (black arrow).
Figure 7: High-low plots with mean failure loads in lateral loading (a) and axial torsion (b) at the
17
different vertebral levels. The symbol * indicates significant difference. Dashed lines indicate lower limit
18
values required for use in the non-fusion implant system.
19 20
AC C
16
Figure 8: The accumulated scores (for individual scores, see table 4 ) of the XSFIX are higher for the high and mid thoracic level but lower for the low thoracic and lumbar level. 14
ACCEPTED MANUSCRIPT
Table 1: T-scores taken from DXA scans show osteoporosis in all spines. T-score
Gender
Age
80
T6-T12
L1-L5
1
-4.4
-3.6
Male
2
-3.3
-4.1
Male
3
-4.8
-4.3
Male
4
-6.4
-5.5
Female
RI PT
Spine
84
AC C
EP
TE D
M AN U
SC
98
92
ACCEPTED MANUSCRIPT
Table 2: Loading types (‘F’ for lateral force and ‘T’ for torque) and fixations (XS-FIX/PSA) were altered between different spines and spinal levels. Four different spinal levels were defined for comparison: high thoracic (HT), mid thoracic (MT), low thoracic (LT), and lumbar (L).
S1
F
S2
T
S3
T4
T6
T7
T8
T
T
T
F
F
T
F
T
S4
T5
F F
T9
T10
T
T
T11
T
F
F
F
S3
T
T
F T
F
F
T
T
T
F
EP
T
T
F
F
L3
L4
L5
F
F
T
T
T
TE D
S4
F
L2
T
T F
L1
T
F
F
S2
T12
T T
S1
AC C
PSA
T3
L
M AN U
XS-FIX
T2
LT
RI PT
T1
MT
SC
HT
T T
T F T
T
F
F F
T
F
T T F
ACCEPTED MANUSCRIPT
Table 3: Mean fixation strengths and standard errors of the two fixations at all different vertebral levels. HT
L
Mean
SE
Mean
SE
Mean
SE
Mean
SE
XS-FIX
341.3
61.3
402.2
97.8
317.5
112.5
451.0
79.1
PSA
702.7
92.1
506.7
60.7
594.6
28.8
631.8
14.1
XS-FIX
13.9
2.2
12.3
1.8
5.0
1.8
4.5
1.5
PSA
9.6
1.9
13.1
3.4
14.0
4.0
16.2
2.8
AC C
EP
TE D
M AN U
SC
Torque [Nm]
LT
RI PT
Force [N]
MT
ACCEPTED MANUSCRIPT
Table 4: Evaluation of fixations on basis of three criteria (strength, facet joint damage, and placement) shows that the XS-FIX scores well on all criteria in the high and mid thoracic regions. The PSA scores best in the low thoracic and lumbar regions. Grades used for assessment are ‘poor’ (− or -1), ‘good’ (+ or 1) and ‘very good’ (++ or 2).
Fixation strength
+(1)
+(1)
−(-1)
−(-1)
Facet joint damage
+(1)
+(1)
+(1)
+(1)
Placement
+(1)
+(1)
−(-1)
+(1)
Fixation strength
+(1)
+(1)
+(1)
+(1)
Facet joint damage
−(-1)
−(-1)
+(1)
+(1)
Placement
+(1)
+(1)
+(1)
++(2)
SC
M AN U
TE D
L
RI PT
LT
EP
PSA
MT
AC C
XS-FIX
HT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S1529-9430(14)00461-6
DOI:
10.1016/j.spinee.2014.04.028
Reference:
SPINEE 55879
To appear in:
The Spine Journal
Received Date: 5 March 2013 Revised Date:
19 March 2014
Accepted Date: 16 April 2014
Please cite this article as: Wessels M, Homminga J, Hekman E, Verkerke G, A novel anchoring system for use in a non-fusion scoliosis correction device, The Spine Journal (2014), doi: 10.1016/ j.spinee.2014.04.028. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title A novel anchoring system for use in a non-fusion scoliosis correction device
RI PT
Authors
Wessels M, PhD,* Homminga JJ, PhD,* Hekman EEG, MSc,* Verkerke GJ, PhD,*†
Academic affiliation
SC
* Laboratory of Biomechanical Engineering, Faculty of Engineering Technology, University of Twente,
M AN U
The Netherlands
† Department of Rehabilitation Medicine, University of Groningen, University Medical Center Groningen, The Netherlands
Corresponding author
HR W 207, PO box 217
TE D
Martijn Wessels
EP
7500 AE Enschede, The Netherlands
AC C
Email: [email protected]
Tel: 0031 53489 2351
ACCEPTED MANUSCRIPT
Abstract
2
Background context
RI PT
1
Insertion of a pedicle screw in the mid and high thoracic region has a serious risk of facet joint
4
damage. Since flexible implant systems require intact facet joints, we developed an enhanced fixation
5
that is less destructive to spinal structures. The XSFIX is a posterior fixation system that uses cables that
6
are attached to the transverse processes of a vertebra.
7
Purpose
9
M AN U
8
SC
3
Purpose of this study is to determine whether a fixation to the transverse process using the XSFIX is strong enough to withstand the loads applied by the XSLATOR (a novel, highly flexible non-fusion implant system) and thus whether it is a suitable alternative for pedicle screw fixation.
11
Study Design/Setting
12
TE D
10
The strength of a novel fixation system using transverse process cables was determined and compared to the strength of a similar fixation using polyaxial pedicle screws on different vertebral levels.
14
Methods
EP
13
Each of 58 vertebrae, isolated from four adult human cadavers was instrumented with either a
16
pedicle screw anchor (PSA) system or a prototype of the XSFIX. The PSA consisted of two poly-axial
17
pedicle screws and a 5 mm diameter rod. The XSFIX prototype consisted of two bodies that were fixed to
18
the transverse processes, interconnected with a similar rod. Each fixation system was subjected to a
19
lateral or an axial torque.
AC C
15
1
ACCEPTED MANUSCRIPT
1 2
Results The PSA demonstrated fixation strength in lateral loading and torsion higher than required for use in the XSLATOR. The XSFIX demonstrated high enough fixation strength (in both lateral loading and
4
torsion), only in the high and mid thoracic region (T10-T12).
5
Conclusions
6
RI PT
3
This experiment showed that the fixation strength of XSFIX is sufficient for use with the XSLATOR only in mid and high thoracic region. For the low thoracic and lumbar region, the PSA is a more rigid
8
fixation. Since the performance of the new fixation system appears to be favorable in the high and mid
9
thoracic region, a clinical study is the next challenge.
11
M AN U
10
SC
7
Introduction
Pedicle screws are commonly used for posterior anchoring of spinal implants.1-3 Although pedicle screws are now commonly used on the lumbar and thoracic spine,2 the shape of the pedicles in the
13
thoracic region is inconsistent, especially in spinal deformities such as scoliosis.4,5 As a result of this,
14
pedicle screw placement is very challenging in this region.6-9 Currently, pedicle screws are inserted by
15
using either the free hand technique (as in the Roy-Camille procedure), the open lamina technique
16
where a partial laminectomy is performed, or under application of an image-guidance system. In each of
17
these techniques, required accuracy is very high.10,11 The pedicle entry point for a high or mid thoracic
18
pedicle screw is situated at the base of the superior facet.12,13 Even a well-placed pedicle screw may still
19
violate the facet joint. Particularly in the thoracic region, the size and position of the screw head can
20
result in obstruction of movements of the facet joints. Facet joint violation in pedicle screw anchoring in
21
the thoracic and lumbar region is a common occurrence in spinal surgery.14,15 The effect of limited
AC C
EP
TE D
12
2
ACCEPTED MANUSCRIPT
motion of the joints, such as in spondylodesis is often desirable , but particularly troublesome in the
2
recent trend toward dynamic, non-fusion systems that aim to maintain the flexibility of the spine.
3
Consequently, a demand for an enhanced fixation, that is less destructive to spinal structures, has grown.
4
RI PT
1
For these reasons, a novel fixation system, referred to as XSFIX, was developed; a bridge
construction that consists of two anchors, with a bridge in between. Each of these anchors consists of a
6
connector that is attached to a transverse process by means of a cable. This new fixation system can be
7
applied for non-fusion spinal implant systems that will correct spinal deformities e.g. scoliosis. The
8
XSLATOR, a non-fusion scoliosis correction system which is currently being developed (Figure 1), is able
9
to lengthen (cranially-caudally) to allow for spinal growth and daily motion (including flexion/extension)
M AN U
SC
5
by making use of sliding bearings. The axial forces (caused by all spinal motion including growth) and
11
sagittal forces (caused by flexion/extension) remain extremely low, while lateral force and axial torque
12
are significant but low.16 To consider the XSFIX as a solid fixation, the anchor must be able to transfer the
13
forces and moments forces delivered by the XSLATOR, which are (including a design safety factor of 2) a
14
maximal lateral force of 200 N and a maximal axial torque (around the local spinal longitudinal axis) of 7
15
Nm. 16 These loads are smaller than the loads applied by fusion systems, because the non-fusion implant
16
systems use the visco-elastic properties of the spine to correct the deformity.
EP
TE D
10
Purpose of this study is to determine whether our new XSFIX fixation system is strong enough to
18
withstand the loads applied by the XSLATOR and thus whether it is a suitable alternative for a similar
19
bridge construction that uses pedicle screws.
20
Materials and Methods
AC C
17
3
ACCEPTED MANUSCRIPT
1
60 Vertebrae (42 thoracic and 18 lumbar; table 2) were isolated from four adult human cadavers (aged between 84 and 98 years) and cleaned of all soft tissues. All specimens were kept frozen until
3
testing. Prior to the experiment, all vertebrae were DXA-scanned to determine the bone mineral density
4
scores. All spines were identified as osteoporotic (Table 1); T-scores were well below -2.5.17
5
RI PT
2
Each vertebra was instrumented with either a pedicle screw anchor (PSA) system or a prototype of the XSFIX system (Table 2). The PSA consisted of two poly-axial pedicle screws (one in each pedicle) that
7
were rigidly interconnected with a 5 mm diameter rod. The XSFIX prototype consisted of two stainless
8
steel bodies that were each fixed to the left and right transverse processes of the vertebrae with black
9
Nylon 6-6 cable ties (5mm wide, tensile strength 250 N), and were rigidly interconnected with a 5 mm
M AN U
SC
6
10
diameter rod (Figure 2). Tension in the cable ties was consistently applied (at 80 N) using a cable tie
11
installation tool.
12
Before testing, each vertebra was embedded with PMMA bone cement in a metal housing that was rigidly clamped onto the testing setup. Each fixation system was subjected to one of two loading
14
conditions until failure.
TE D
13
A lateral force (in the local coordinate system) was applied to the bridge, with a loading rate of
16
1.0 N/s. An axial torque (in the local coordinate system) was applied to the bridge, with a loading rate of
17
0.05 Nm/s (Figure 3). In the test setup a narrow or a wide fork construction was used to generate torque,
18
depending on the available space (Figure 4).
20
AC C
19
EP
15
A hydraulic tensile/compression machine (MTS, Berlin, Germany) was used to generate the appropriate loading on the bridge. Time and loads were recorded digitally with a sample rate of 2 Hz. The
4
ACCEPTED MANUSCRIPT
1
vertebrae from the four spines were alternately assigned to a type of fixation and to a loading condition
2
(Table 2).
Failure was defined as the point at which the displacement suddenly increased (as shown by the
4
rotation-time graphs (Figure 5a and Figure 5b). The absolute maximal force for each construct was also
5
measured, which represented the point of complete separation of the fixation system from the vertebra.
6
However, as clinically relevant failure clearly started long before that point, the initiation of failure seems
7
a more relevant strength criterion.
SC
RI PT
3
For statistical analysis, no difference in vertebral strength between spines was assumed. Results
9
were analyzed based on four regions: high thoracic (HT): T1-T5; mid thoracic (MT): T6-T10; low thoracic
10
(LT): T11-T12 and lumbar (L): L1-L5. The anatomy of T11 and T12 was markedly different from the other
11
thoracic vertebrae by poorly articulated transverse processes (which are essential for attaching the XSFIX
12
system). Because of this, T11 and T12 represent a separate group. Fixation failure strengths for torque
13
and lateral force were compared between vertebral levels and between fixation techniques using
14
ANOVA (with significance set at p=0.05).
TE D
M AN U
8
The overall performance of XSFIX and PSA was assessed by assigning three grades to each spinal
16
region. The grades ‘poor’, ‘good’ and ‘very good’ were used to evaluate on fixation strength, facet joint
17
damage, and placement. Fixation strength was assessed using the mean (Mean) and standard error (SE)
18
values for each VB level.
AC C
EP
15
19
For each loading condition, the fixation strength was graded:
20
•
‘poor’ if Mean – SE < RV, 5
ACCEPTED MANUSCRIPT
1
•
‘good’ if 1 ·RV ≤ Mean - SE < 2·RV, and
2
•
‘very good’ if 2·RV ≤ Mean – SE,
where RV is the required value (RV = 200 N for lateral force and RV = 7 Nm for torsion).
RI PT
3
For all four VB levels, the lowest grades of PSA and XSFIX, which represent the limiting factor in
5
fixation strength, were selected. For example, if a grade for XSFIX at the LT level was ‘poor’ for torsion
6
and ‘sufficient’ for lateral force, then fixation strength was rated ‘poor’.
SC
4
Facet joint damage was graded:
8
•
‘poor’ if the facet joint was intrinsically damaged,
9
•
‘good’ if the facet joint remained unharmed but with minor risk of joint obstruction, and
10
•
‘very good’ if no damage or possible obstruction occurred.
11
Placement of the fixation was graded:
12
•
‘poor’ if fixation was not possible,
13
•
‘good’ if placement was possible and straightforward but with risk of structural damage, and
14
•
‘very good’ if placement was very easy with very low risk of structural damage.
16
TE D
EP
AC C
15
M AN U
7
Results
The PSA typically showed slow migration of the screws until the structures ruptured completely. The
17
XSFIX generally showed large deformation of the vertebral arc before actual rupture of the structures.
18
Both fixation types typically show a point of sudden increase in rotation (or displacement), see Figure 5a
6
ACCEPTED MANUSCRIPT
1
and Figure 5b, indicating an early collapse of bone structures, defining the clinically relevant point of
2
fixation failure.
No differences in fixation strength for either spinal region were found between spines loaded with
4
lateral force (p>0.445) or torsion (p>0.13). General type of failure for XSFIX in all spines was transverse
5
process failure in axial torsion and pedicle failure in lateral loading. General types of failure for PSA in all
6
spines were pedicle screw migration and vertebral body rupture (Figure 6). Anchoring the XSFIX system
7
to T11 and T12 was problematic and sometimes impossible as the cables slipped off the transverse
8
processes when tightened. Although mean strength difference between XSFIX and PSA in both lateral
9
loading and in axial torsion appeared large, differences showed no significance due to the small sample
11
SC
M AN U
10
RI PT
3
size in the LT region.
In lateral loading, both the PSA and the XSFIX demonstrated fixation strengths above 200 N that are typically expected for a non-fusion implant system such as the XSLATOR. The fixation strength of the
13
XSFIX is significantly lower than that of the PSA in both the high thoracic region (p=0.019) and the lumbar
14
region (p=0.045). In the mid and low thoracic levels the mean strength of XSFIX is not significantly
15
different from the strength of the PSA system (p=0.52, Table 3 and Figure 7). The small sample size of the
16
LT region level causes the mean difference to be not significant (Figure 7a).
EP
In all spinal regions, the PSA demonstrated fixation strengths in torsion above 7 Nm, which was
AC C
17
TE D
12
18
estimated to be maximally generated by the non-fusion implant system. The XSFIX demonstrated fixation
19
strengths in torsion above 7 Nm only in the high and mid thoracic regions. In the lumbar region, the
20
fixation strength of the XSFIX significantly lower than that of the PSA (p=0.015) and lower than 7 Nm. In
21
the high and mid thoracic levels the mean strength of the XSFIX in torsion is not significantly different 7
ACCEPTED MANUSCRIPT
from the strength of the PSA system (p=0.989). The fixation strength (torsion) of the XSFIX was
2
significantly lower in the lumbar region compared to both the high thoracic (p=0.024) and mid thoracic
3
region (p=0.041, Table 3 and Figure 7b). In the mid thoracic region, mean fixation strength difference (in
4
torsion) is large, however not significant caused by small sample size (Figure 7b).
5
In Table 4, a schematic overview is given of the assessment of the XSFIX and PSA in respect to the
6
different spinal regions. Figure 8 shows the cumulated scores of both fixation systems for each vertebral
7
level. Evidently, the scores of the PSA are higher than the XSFIX for the low thoracic and lumbar regions.
8
The XSFIX shows higher accumulated scores than the PSA for the high and mid thoracic regions.
9
Discussion
SC
M AN U
10
RI PT
1
Some limitations apply to this study. First, all tested vertebrae were osteoporotic; the results may differ for healthy vertebrae. However, both the PSA and the XSFIX obtain their strength from bone
12
structures of the vertebral arc, which are relatively unaffected by osteoporosis.18 Furthermore, this test
13
can be seen as a worst-case scenario. It is reasonable to assume that the relative performance of the
14
new fixation system in relation to a similar fixation using pedicle screws is comparable to the
15
performance in adolescent bone.
EP
Second, both the XSFIX and the PSA were placed without the physical limitations that are present
AC C
16
TE D
11
17
during in vivo surgery. Pedicle screws were placed using visual cues not available during surgery. The
18
XSFIX was also easier to attach, due to the removal of the ribs. If the clinical reality prevents proper
19
placement of these systems, failure strengths may be different.
8
ACCEPTED MANUSCRIPT
1
Third, in this experiment the pedicle screw fixation (PSA) consisted of a bridge construction. Clinically, pedicle screws are used as stand-alone systems. A single screw will transfer a torque directly to
3
the vertebra, resulting in high peak pressures and increased risk of early fixation failure. Our test results
4
for the PSA cannot be compared to single pedicle screw arrangements; though it can be stated our
5
results will give much higher strength values.
6
RI PT
2
We defined failure strength as the point at which a sudden increase in displacement (rotation) took place. Because the PSA showed migration of the screws and the XSFIX showed large unrecoverable
8
deformation of the vertebral arc before actual rupture of the structures, we attribute the sudden
9
increase in deformation to microscopic bone failure. Nonetheless, the failure strength criterion we used
11
M AN U
10
SC
7
seems clinically most relevant.
In some cases, premature XSFIX failure occurred in the small ratchet of a cable tie. An improved version of the ratchet that consists of high strength materials will prevent early fixation failure. In
13
addition, the material aspect needs further attention because the material used in our ties (Polyamide
14
6/6) is not suitable for implantation due to issues with creep, relaxation, and biocompatibility. Replacing
15
the material for a creep resistant and biocompatible material such as carbon fiber reinforced PEEK
16
(polyetheretherketone) may well be considered.19,20
EP
TE D
12
In the high and mid thoracic regions the torsion strength of the XSFIX is higher than the required
18
value of 7 Nm for use with the non-fusion implant system such as the XSLATOR. In the low thoracic and
19
lumbar regions, the torsion strength of the XSFIX is lower than the required 7 Nm, but in these regions,
20
the PSA is strong enough. Lateral fixation strength of the XSFIX meets the requirement of 200 N for use
21
with the non-fusion implant system for all spinal regions. However, the anchoring of the XSFIX to T11 and
AC C
17
9
ACCEPTED MANUSCRIPT
T12 is problematic, due to the poorly pronounced transverse processes of those vertebrae.
2
Consequently, the XSFIX is not very suitable for those vertebrae. Similarly, the lumbar vertebrae have
3
long transverse processes that are very slender and fragile, making the lumbar vertebrae not suited for
4
use of the XSFIX. In fact, some of the lumbar transverse processes broke during tensioning of the cable
5
ties. Fortunately, pedicle screws can be inserted easily into these vertebrae without obstructing or
6
damaging the facet joints.
RI PT
1
The highly flexible implant XSLATOR requires three vertebral levels for fixation.16 Either PSA or XSFIX
8
must be used for anchoring. The force requirements for the fixation are based on the maximal forces on
9
the center vertebra. The actual force on the other two vertebrae is only 50% of the referred maximal
10
force. The XSLATOR therefore cannot be compared to other (fusion or non-fusion) scoliosis correction
11
constructs.
M AN U
SC
7
The grading system used in table 4 is rather subjective. However, the grading system was used
13
merely to illustrate our conclusions. Despite the subjectivity of the grading, we think the outcome
14
regarding comparison between both systems will be similar. Nevertheless, Table 4 shows that the XSFIX
15
is an excellent fixation for the high and mid thoracic regions. Especially in this region, the application of
16
pedicle screws is risky, because of the small diameter of the pedicles. Therefore, the XSFIX could play an
17
important role in the fixation of devices to the spine. In contrast, grades for fixation strength and
18
placement of XSFIX in the low thoracic and lumbar region are valuated much lower, whereas PSA
19
produced decent grades for those regions.
20
With this experiment, we have a strong indication that the XSFIX is comparable to the pedicle screw
21
performance (strength, placement and joint damage) in the mid and high thoracic region and that for the
AC C
EP
TE D
12
10
ACCEPTED MANUSCRIPT
lumbar region the pedicle screw is the method of choice. With this outcome, animal experiments can be
2
started to gather statistical evidence that the XSFIX is indeed preferred for fixation in the mid and high
3
thoracic region. Therefore, the experiments described in this manuscript contributed to a restriction of
4
animal experiments by giving these experiments a specific focus.
5
RI PT
1
In summary, this experiment showed that the XSFIX is a solid fixation with strength that is
comparable to the PSA strength in torsion, when placed on a vertebra in the T2-T10 region. Fixation
7
strength of XSFIX is sufficient for use with a non-fusion implant system, such as the XSLATOR, in mid and
8
high thoracic regions. For the low thoracic and lumbar region, the PSA is a more rigid fixation. Since the
9
performance of the new fixation system appears to be favorable in the high and mid thoracic region, a
10
clinical study is the next challenge.
11
References
1. Steffee AD, Biscup, RS, Sitkowskj DJ. Segmental spine plates with pedicle screw fixation - a new
TE D
12
M AN U
SC
6
internal fixation device for disorders of the lumbar and thoracolumbar spine. Clin Orthop Relat Res
14
1986;203:45-53.
16
17 18
19 20
2. Gaines RW Jr. The use of pedicle-screw internal fixation for the operative treatment of spinal disorders. J Bone Joint Surg Am 2000;82:1458-76.
AC C
15
EP
13
3. Cotrel Y, Dubousset J, Guillaumat M. New universal instrumentation in spinal surgery. Clin Orthop Relat Res 1988;227:10-23.
4. Zeiller S, Lee J, Lim M, Vaccaro A. Posterior thoracic segmental pedicle screw instrumentation: Evolving methods of safe and effective placement. Neurology India 2005;53:458-65. 11
ACCEPTED MANUSCRIPT
3 4
5 6
7
instrumentation: critical appraisal of current "state-of-art". Neurology India 2009;57:715-21.
6. Vaccaro AR, Rizzolo SJ, Allardyce TJ, et al. Placement of pedicle screws in the thoracic spine. Part I:
RI PT
2
5. Mattei T, Meneses M, Milano J, et al. "Free-hand" technique for thoracolumbar pedicle screw
Morphometric analysis of the thoracic vertebrae. J Bone Joint Surg Am 1995;77:1193-9.
7. Vaccaro A.R., et al. Placement of pedicle screws in the thoracic spine. Part II: An anatomical and radiographic assessment. J Bone Joint Surg Am 1995;77:1200-6.
SC
1
8. Xu R, Ebraheim NA, Ou Y, et al. Anatomic considerations of pedicle screw placement in the thoracic spine. Roy-Camille technique versus open-lamina technique. Spine 1998;23:1065-8.
9
9. Lonstein JE, et al. Complications associated with pedicle screws. J Bone Joint Surg Am
13 14
15 16
17 18
TE D
12
10. Rampersaud YR, Simon DA, Foley KT. Accuracy requirements for image-guided spinal pedicle screw placement. Spine 2001;26:352-9.
11. Papadopoulos EC, Girardi FP, Sama A, et al. Accuracy of single-time, multilevel registration in image-guided spinal surgery. Spine J 2005;5:263-7.
EP
11
1999;81:1519-28.
12. Chung KJ, Suh SW, Desai S, Song HR. Ideal entry point for the thoracic pedicle screw during the
AC C
10
M AN U
8
free hand technique. Int Orthop 2008;32(5): 657-662.
13. Modi HN, Suh SW, Fernandez H, et al. Accuracy and safety of pedicle screw placement in neuromuscular scoliosis with free-hand technique. Eur Spine J 2008;17(12):1686-1696.
12
ACCEPTED MANUSCRIPT
1
14. Chung KJ, Suh SW, Swapnil K, et al. Facet joint violation during pedicle screw insertion: a cadaveric study of the adult lumbosacral spine comparing the two pedicle screw insertion techniques. Int
3
Orthop 2007;31:653-656.
4
RI PT
2
15. Chen Z, Zhao J, Xu H, et al. Technical factors related to the incidence of adjacent superior
segment facet joint violation after transpedicular instrumentation in the lumbar spine. Eur Spine J
6
2008;17(11):1476-1480.
8
9 10
16. Wessels M, Hekman EEG, Verkerke GJ. Mechanical behavior of a novel non-fusion scoliosis correction device. J Mech Behav Biomed 2013;27:107-114.
M AN U
7
SC
5
17. WHO. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis. Report of a WHO Study Group. World Health Organ Tech Rep Ser 1994;843:1-129.
18. Riggs BL. Overview of osteoporosis. West J Med 1991;154:63-77.
12
19. Kurtz SM, Devine JN. PEEK biomaterials in trauma, orthopedic, and spinal implants. Biomaterials
14
2007;8:4845-69.
20. Toth JM, et al. Polyetheretherketone as a biomaterial for spinal applications. Biomaterials 2006;27:324-34.
16
Figure Legends
17
AC C
15
EP
13
TE D
11
Figure 1: The XSLATOR consists of two implants that are posteriorly fixed to the vertebrae. The three
18
bridges can relatively slide axially keeping axial forces extremely low. The bridges are fixed to the spine
19
using transverse process cables (XSFIX; right, top) or 2 pedicle screws (right, bottom). 13
ACCEPTED MANUSCRIPT
2
3 4
5
Figure 2: The two different fixations were attached to the vertebrae and the vertebrae were fixed in cups with PMMA. (Left) Vertebra with the XSFIX. (Right) Vertebra with the PSA in setup.
Figure 3: Loading with a vertical (lateral) force on XSFIX (left), PSA (middle), and torque around a
RI PT
1
vertical axis on PSA (right) was performed using a hydraulic tensile/compression machine.
Figure 4: The cemented cups were clamped into a framework, which was clamped onto a table. Torque was applied either between the cables (left) using a narrow rotating fork or on the ends of the
7
rod (right) using a wide rotating fork.
Figure 5a: Typical rotation-time curve of a XSFIX fixation. The rotation-time curve shows a sudden
M AN U
8
SC
6
9
increase in rotation at time=385 s, indicating a collapse of structures. At time=488 s one of the
10
transverse process is separated from the vertebral arc resulting in a sharp increase in rotation.
11
Figure 5b: Typical rotation-time curve of a PSA fixation. Also here, at time=998 s, a sudden increase of rotation can be seen in the curve (defined as the failure point). At time=1851 s, the vertebral body
13
ruptured.
15
Figure 6: (Left) Transverse process rupture, indicated with a white arrow, in torsion test with XSFIX.
EP
14
TE D
12
(Right) Vertebral body rupture in torsion test with PSA (black arrow).
Figure 7: High-low plots with mean failure loads in lateral loading (a) and axial torsion (b) at the
17
different vertebral levels. The symbol * indicates significant difference. Dashed lines indicate lower limit
18
values required for use in the non-fusion implant system.
19 20
AC C
16
Figure 8: The accumulated scores (for individual scores, see table 4 ) of the XSFIX are higher for the high and mid thoracic level but lower for the low thoracic and lumbar level. 14
ACCEPTED MANUSCRIPT
Table 1: T-scores taken from DXA scans show osteoporosis in all spines. T-score
Gender
Age
80
T6-T12
L1-L5
1
-4.4
-3.6
Male
2
-3.3
-4.1
Male
3
-4.8
-4.3
Male
4
-6.4
-5.5
Female
RI PT
Spine
84
AC C
EP
TE D
M AN U
SC
98
92
ACCEPTED MANUSCRIPT
Table 2: Loading types (‘F’ for lateral force and ‘T’ for torque) and fixations (XS-FIX/PSA) were altered between different spines and spinal levels. Four different spinal levels were defined for comparison: high thoracic (HT), mid thoracic (MT), low thoracic (LT), and lumbar (L).
S1
F
S2
T
S3
T4
T6
T7
T8
T
T
T
F
F
T
F
T
S4
T5
F F
T9
T10
T
T
T11
T
F
F
F
S3
T
T
F T
F
F
T
T
T
F
EP
T
T
F
F
L3
L4
L5
F
F
T
T
T
TE D
S4
F
L2
T
T F
L1
T
F
F
S2
T12
T T
S1
AC C
PSA
T3
L
M AN U
XS-FIX
T2
LT
RI PT
T1
MT
SC
HT
T T
T F T
T
F
F F
T
F
T T F
ACCEPTED MANUSCRIPT
Table 3: Mean fixation strengths and standard errors of the two fixations at all different vertebral levels. HT
L
Mean
SE
Mean
SE
Mean
SE
Mean
SE
XS-FIX
341.3
61.3
402.2
97.8
317.5
112.5
451.0
79.1
PSA
702.7
92.1
506.7
60.7
594.6
28.8
631.8
14.1
XS-FIX
13.9
2.2
12.3
1.8
5.0
1.8
4.5
1.5
PSA
9.6
1.9
13.1
3.4
14.0
4.0
16.2
2.8
AC C
EP
TE D
M AN U
SC
Torque [Nm]
LT
RI PT
Force [N]
MT
ACCEPTED MANUSCRIPT
Table 4: Evaluation of fixations on basis of three criteria (strength, facet joint damage, and placement) shows that the XS-FIX scores well on all criteria in the high and mid thoracic regions. The PSA scores best in the low thoracic and lumbar regions. Grades used for assessment are ‘poor’ (− or -1), ‘good’ (+ or 1) and ‘very good’ (++ or 2).
Fixation strength
+(1)
+(1)
−(-1)
−(-1)
Facet joint damage
+(1)
+(1)
+(1)
+(1)
Placement
+(1)
+(1)
−(-1)
+(1)
Fixation strength
+(1)
+(1)
+(1)
+(1)
Facet joint damage
−(-1)
−(-1)
+(1)
+(1)
Placement
+(1)
+(1)
+(1)
++(2)
SC
M AN U
TE D
L
RI PT
LT
EP
PSA
MT
AC C
XS-FIX
HT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT