ORIGINAL ARTICLE
A Biomechanical Study of 3 Different Types of Sublaminar
Wire Used for Constructs in the Thoracic Spine
Hideki Murakami, MD, PhD,*w z Ken Yamazaki, MD, PhD,* Emad S. Attallah-Wasif, MD,w z
Kai-Jow Tsai, MD,w z Tadashi Shimamura, MD, PhD,* and William C. Hutton, DScw z
Abstract: A biomechanical study was carried out on 3 different
types of sublaminar wire used in constructs to secure the
thoracic spine: stainless steel monofilament wire (steel wire),
titanium cable (cable), and ultra-high molecular weight poly-
ethylene tape (tape). Two experiments were carried out.
Experiment 1: Thirty-one fresh human thoracic vertebrae classi-
fied as osteoporotic (bone mineral density of <0.8 g/cm2) were
used. The steel wire, cable, or tape was placed sublaminarly and
a tensile force was applied until the steel wire, cable, or tape cut
5 mm through the lamina, and the force at this point was noted.
Experiment 2: Seven fresh human thoracic spines (T7–T10) were
biomechanically tested as follows: axial compression (250 N),
flexion (7.5 Nm), extension (7.5 Nm), left lateral bending
(7.5 Nm), right lateral bending (7.5 Nm), left axial torsion
(10 Nm), and right axial torsion (10 Nm). This sequence was
applied to the intact spine. The spine was then de-stabilized and
then restabilized using one or other of the 3 different types of
sublaminar wires. The biomechanical testing was then repeated
on the restabilized spine and stiffness curves were generated. In
the laminar cut-through test, the cut-through force for tape was
higher than that for either steel wire or cable. In the
biomechanical stiffness testing, there was no significant differ-
ence between the 3 different sublaminar wiring constructs in any
of the loading modes tested. The results of both experiments
suggest that tape is as good, if not better, than steel wire or cable
as a sublaminar wiring construct material.
(J Spinal Disord Tech 2006;19:442–446)
S pinal instrumentation is generally necessary when
surgery is required to treat deformity and instability in
osteoporotic patients. Posterior segmental sublaminar
Received for publication April 18, 2005; accepted October 17, 2005.
From the *Department of Orthopaedic Surgery, Iwate Medical
University School of Medicine, Morioka, Japan; wAtlanta Veteran’s
Affairs Medical Center, Decatur; and zDepartment of Orthopaedics,
Emory University School of Medicine, Atlanta, GA.
No funds were received in support of this work.
The device(s)/drug(s) that is/are the subject of this manuscript is/are not
FDA-approved for this indication and is/are not commercially
available in the United States.
Reprints: Hideki Murakami, MD, PhD, Department of Orthopaedic
Surgery, Iwate Medical University School of Medicine, 19-1
Uchimaru, Morioka, Iwate 020-8505, Japan (e-mail: hmura@iwate-
med.ac.jp).
Copyright r 2006 by Lippincott Williams & Wilkins
442
wiring is rarely indicated in the lumbar spine because
pedicle screw systems are more rigid.1 However, because
of the difficulty of using pedicle screws in the thoracic
spine, sublaminar wiring remains a valuable method,
especially in the osteoporotic spine.2 Hu et al3 reported
that sublaminar wires that can be repeatedly tightened
have a role in deformity correction in the osteoporotic
spine.
The surgical treatment of the osteoporotic spine is
complicated by the weak bone, which may not support
instrumentation. Sometimes, sublaminar wire cuts
through the lamina in the osteoporotic spine, especially
if the lamina is decorticated. Implant loosening or failure
can lead to nonunion or loss of correction. The major
factors determining the fixation strength of a spine
instrumented with sublaminar wire is the properties of
the vertebral bone, the thickness of the laminar cortex,
and the sublaminar wiring implant itself (eg, material and
configuration).
Two experiments were carried out to compare the
biomechanical properties of 3 different types of sublami-
nar wire. In Experiment 1, the laminar cut-through force
was measured using 3 different types of sublaminar wires.
In Experiment 2, the biomechanical stiffness of constructs
secured by these 3 different types of sublaminar wires was
measured. These results are given to provide information
to aid in the selection of appropriate sublaminar wire for
constructs, especially for osteoporotic patients.
MATERIALS AND METHODS
Experiment 1: Laminar Cut-through Test
Five fresh human thoracic spines (T6 to T12) were
harvested from cadavers (mean age 83 y, range 76 to 87 y,
3 females, 2 males). The specimens were examined grossly
and radiographically to rule out malignancy or fractures
that might interfere with the results. The bone mineral
density (BMD) was measured for each vertebral segment
using dual-energy x-ray absorptiometry (GE Lunar, GE
Medical Systems, Madison, WI) with the specimen in a
water bath. Using the World Health Organization
definition of osteoporosis as BMD<0.8 g/cm2,4 we
classified 31 vertebrae in our test group were classified
as osteoporotic. All of the specimens were then frozen at
 20 1C until the day of testing. After thawing to room
temperature for 24 hours, the surrounding soft tissue and
muscles were dissected meticulously and the spines
J Spinal Disord Tech  Volume 19, Number 6, August 2006
J Spinal Disord Tech  Volume 19, Number 6, August 2006
separated into individual vertebrae. Three different types
of sublaminar wire were compared: (1) stainless steel
monofilament wire 0.9 mm diameter (steel wire)
(SUS316L, Mizuho, Tokyo, Japan) (Fig. 1A), (2)
titanium cable 1.0 mm diameter (cable) (Atlas cable,
826-211, Medtronic Sofamor Danek, Memphis, TN) (Fig.
1B), and (3) ultra-high molecular weight polyethylene
tape, 5 mm wide (tape) (Tekmilon tape, NCS5, Alfresa
pharma, Osaka, Japan) (Fig. 1C). Nine vertebrae were
used for steel wire, 10 vertebrae were used for cable, and
12 vertebrae were used for tape. The vertebrae were
randomly assigned to one of the 3 groups. Throughout
the text we used the term ‘‘sublaminar wire’’ to describe
all 3 types of wire, and ‘‘steel wire’’ to distinguish the
stainless steel monofilament wire 0.9 mm diameter.
One lamina was decorticated in its caudal portion
and the other lamina was kept intact. Each vertebra was
placed in a pot containing dental cement and clamped on
a materials testing machine (MTS 858 Mini-Bionix Test
System, Minneapolis, MN) in the upright position
(Fig. 2A). The steel wire, cable, or tape was placed
sublaminarly and then threaded through a ring attached
to the movable ram of the testing machine. The 2 ends
were then made into a loop by twisting or knotting the
ends together. A tensile force was applied at a constant
speed of 5 mm/min until the steel wire, cable, or tape
migrated 5 mm through the lamina, and the force at this
point was noted (Fig. 2B). This laminar cut-through test
was carried out on the intact lamina. The test was then
repeated on the opposite decorticated lamina. After this,
FIGURE 1. Sublaminar wire. A, Stainless steel monofilament
wire 0.9 mm diameter (steel wire); B, Titanium cable 1.0 mm
diameter (cable); C, Ultra-high molecular weight polyethylene
tape 5 mm wide (tape).
r 2006 Lippincott Williams & Wilkins
Constructs in the Thoracic Spine
FIGURE 2. Experiment 1: Laminar cut-through test. A, Each
vertebra was placed in a pot containing dental cement and
clamped on a materials testing machine in the upright
position. The steel wire, cable, or tape was placed sublami-
narly and then threaded through a ring attached to the
movable ram of the testing machine. The 2 ends were then
made into a loop by twisting or knotting the ends together. B,
A tensile force was applied at a constant speed of 5 mm/min
until the steel wire, cable, or tape cut 5 mm through the
lamina, and the force at this point was noted.
the laminae were cut off using a bone saw and the
thickness of the cortical bone at the 5-mm cut-through
position was measured using a digital micrometer.
This was found to be a difficult experiment to carry
out on weak osteoporotic vertebrae. In Experiment 1, the
decorticated side was tested first. It was reasoned that the
cut-through force on the decorticated side would gen-
erally be lower than that on the intact side, and that that
the neural arch would be subjected to the lower force first.
Even using this method the neural arch was sometimes
weakened to such an extent after cutting through the
decorticated lamina, that the cut-through force on the
intact side was sufficient to fracture the neural arch at the
pedicles. Thus from the 31 osteoporotic vertebrae that we
started with, we obtained results for only 15. Five for steel
wire, 5 for cable, and 5 for tape. The statistical analysis
was carried out on these 15.
Statistics: A 2-way analysis of variance was carried
out on the log transformation of the cut-through force,
followed by Tukey’s pairwise comparisons. The 2 factors
were: the different types of sublaminar wires (cable, tape,
or wire) and whether or not the lamina was intact or
decorticated. The overall level of significance was set at a
P<0.05. A log transform was applied to the cut-through
force to improve normality and homogeneity assump-
tions. The results are shown as a mean ± standard
deviation.
Experiment 2: Biomechanical Stiffness Test
Seven fresh human thoracic spines (T7 to T10) were
harvested from cadavers (mean age 83 y, range 76 to 87 y,
4 females, 3 males). The specimens were examined grossly
and radiographically to rule out malignancy or fractures
443
Murakami et al
that might interfere with the results. All the specimens
were then frozen at  20 1C until the day of testing. After
thawing to room temperature for 24 hours, the surround-
ing soft tissue and muscles were dissected meticulously to
preserve the osseous and ligamentous structures before
testing. The specimens were kept moist throughout the
entire testing procedure.
The intact specimens, which were placed in pots
containing dental cement, were aligned in the cement to
make the cranio-caudal ends of the pots parallel. These
were then clamped on a biomechanical testing machine
(MTS 858 Mini-Bionix Test System, Minneapolis, MN)
in the upright position. The intact specimen was first
cyclically preloaded by axial compression (150 ± 50 N at
1 Hz for 15 min) to remove excess fluid from the discs and
ligaments. This preconditioning is a normal procedure in
biomechanical testing of cadaveric spines in our labora-
tory; it allows the disc height to get back to the normal
physiologic height and also to yield reproducible stiffness
curves during subsequent loading. Biomechanical testing
of the intact specimen was then carried out according to
the following sequence: axial compression (250 N), flexion
(7.5 Nm), extension (7.5 Nm), left lateral bending
(7.5 Nm), right lateral bending (7.5 Nm), left axial torsion
(10 Nm), and right axial torsion (10 Nm). This biomecha-
nical testing procedure is explained in more detail
elsewhere.5 Destabilization was then performed by
sectioning the supraspinatous ligament, interspinous
ligament, ligamentum flavum, facet joint capsules, poster-
ior longitudinal ligament, and the posterior part of the
anulus fibrosus.6 The specimens were then restabilized
using a Hartshill rectangle (with rod 5 mm diameter)
(Surgicraft, Redditch, UK) plus 1 or other of the same 3
different types of sublaminar wires as described in
Experiment 1. The implants used with the Hartshill
rectangle were constructed as follows: (1) Single steel
wires from T7 to T10 were placed sublaminarly and fixed
to the rectangle by twisting the ends together (Fig. 3A),
(2) Cable from T7 to T10 was placed sublaminarly and
fixed to the rectangle and tightened to 200 N of tension
FIGURE 3. Experiment 2: Biomechanical stiffness test. The
specimens were restabilized from T7 to T10 using a Hartshill
rectangle plus sublaminar wire. A, sublaminar single steel
wires fixed to the rectangle by twisting the ends together;
B, Sublaminar cable fixed to the rectangle and tightened
to 200 N of tension using a special tightening device; C,
Sublaminar double tape fixed to the rectangle and tightened
to 200 N of tension using a special tightening device.
444
J Spinal Disord Tech  Volume 19, Number 6, August 2006
using a special device. The tensioner (Alfresa pharma,
Osaka, Japan) had a graded scale to apply tension of
200 N to the cable (Fig. 3B), (3) Double tape from T7 to
T10 was placed sublaminarly and fixed to the rectangle
and tightened to 200 N of tension using a special device
(Fig. 3C). Every test was carried out twice and successive
load/deformation curves were nearly identical (this is
usually the case because the specimens were precondi-
tioned). From the load/deformation curves, stiffness
values were calculated by measuring the slope of the
linear part of the curve. This method is explained more
fully elsewhere.7 Normalized stiffness values were calcu-
lated by dividing the stiffness obtained from the experi-
mental specimen by the stiffness obtained from the intact
specimen (ie, ratio of instrumented to intact).
Statistics: A 1-way repeated measures analysis of
variance was used to compare the results of the intact
specimens and those fixed by steel wire, cable, or tape in
each loading sequence, followed by Tukey’s multiple
comparison procedure. The overall level of significance
was set at a P<0.05.
RESULTS
Experiment 1: Laminar Cut-through Test
On the basis of the results from the 2-factor analysis
of variance, there was no statistically significant interac-
tion between wire type and whether the laminar was
decorticated. Overall, the average cut-through force
whether the lamina was decorticated or not was margin-
ally insignificant (P = 0.06). However, there was a
statistically significant difference in the average cut-
through force depending on the steel wire (P = 0.002).
On the basis of Tukey’s pairwise comparison using a 0.05
error rate, steel wire was significantly different from both
cable and tape. There was no significant difference
between cable and tape.
The average BMD, thickness of the laminar cortex,
and the average value of force at the point of 5 mm cut-
through on the lamina in each of the 3 groups tested using
the 3 types of sublaminar wires are shown in Table 1. In
the test using steel wire, the average BMD for the 5
vertebrae was 0.79 ± 0.04 g/cm2. The average thickness of
the 5 laminar cortices was 3.72 ± 0.72 mm, and the
average cut-through force when the laminar cortex was
intact was 566.8 ± 82.2 N, and 501.0 ± 55.4 N when the
laminar cortex was decorticated.
In the test using cable, the average BMD for the
5 vertebrae was 0.79 ± 0.03 g/cm2. The average thickness
of the 5 laminar cortices was 3.96 ± 0.48 mm, and the
average cut-through force when the laminar cortex was
intact was 808.6 ± 144.4 N, and 610.8 ± 153.8 N when
the laminar cortex was decorticated.
In the test using tape, the average BMD for the
5 vertebrae was 0.78 ± 0.05 g/cm2. The average thickness
of the 5 laminar cortices was 3.67 ± 0.72 mm, and the
average cut-through force when the laminar cortex was
intact was 1164.7 ± 57.6 N, and 911.3 ± 7.0 N when the
laminar cortex was decorticated.
r 2006 Lippincott Williams & Wilkins
J Spinal Disord Tech  Volume 19, Number 6, August 2006 Constructs in the Thoracic Spine

TABLE 1. Results for Experiment 1

Wire Cable Tape
BMD (g/cm2) 0.79 ± 0.05 0.79 ± 0.06 0.78 ± 0.08
Thickness of laminar cortex (mm) 3.72 ± 0.72 3.96 ± 0.48 3.67 ± 0.72
Cut-through force (N): intact lamina 566.8 ± 82.2b 817.7 ± 203.0 1157.0 ± 79.2a,b
Cut-through force (N): decorticated lamina 501.0 ± 55.4c 610.8 ± 153.8d 911.3 ± 7.0a,c,d
The BMD, thickness of the laminar cortex, and the value of tensile force at the 5-mm cut-through position on the
lamina, for each of the 3 types of sublaminar wire. There is no significant difference in the BMD and thickness of the
laminar cortex between the 3 groups. There is a significant difference in the values of cut-through force between the test
using steel wire and the test using tape, steel wire, and cable (P<0.01). Data are shown as mean ± standard deviation,
n = 5, 5, and 5 for steel wire, cable, and tape, respectively. The superscripts a–d are used to identify comparison between
two numbers: for example the two numbers with ‘a’ superscript (1157.0 and 911.3) are statistically different.

Experiment 2: Biomechanical Stiffness Test
In none of the experiments did the bone or the
sublaminar wire break. All data are shown normalized as
a relative ratio to the results obtained from the intact
specimens (Fig. 4). In axial compression, the average
normalized values of stiffness were all less than intact
(0.88 steel wire, 0.76 cable, and 0.99 tape). There is no
significant difference between any of these 3 values.
In flexion, the average normalized values of stiffness
were 1.11 steel wire, 1.20 cable, and 1.04 tape. There was
no significant difference between any of these 3 values.
In extension, the average normalized values of
stiffness were 1.11 steel wire, 1.05 cable, and 1.07 tape.
There was no significant difference between any of these 3
values.
In lateral bending, the average normalized values of
stiffness were 1.08 steel wire, 0.98 cable, and 1.0 tape.
There was no significant difference between any of these 3
values. In lateral bending, we averaged the results for
right and left.
In axial torsion, the average normalized values of
stiffness were 1.76 steel wire, 2.36 cable, and 1.48 tape.
There was no significant difference between any of these 3
FIGURE 4. Results for Experiment 2. All data are shown
normalized as a relative ratio to the results obtained from the
intact specimens.
values. In axial torsion, we averaged the results for right
and left.
DISCUSSION
The main results to emerge from these experiments
are as follows: (1) Experiment 1; in the laminar cut-
through test, the cut-through force for steel wire was
significantly lower than that for either tape or cable. The
BMD and thickness of the laminar cortex was similar for
each of 3 experimental groups. Further, whereas the
average cut-through force for the intact lamina was
higher than that for the decorticated lamina for all 3 types
of wires it was not statistically significant at the 0.05 level.
(2) Experiment 2; in the biomechanical stiffness test, there
was no significant difference between the 3 different
sublaminar wiring constructs in any of the loading modes
tested.
Butler et al8 reported that BMD correlated with
strength and stiffness of the sublaminar wire construct in
osteoporotic spines. Andingoz et al9 reported that there
was positive linear relationship between BMD and
laminar strength in sublaminar wiring. Biomechanical
studies have reported that the strength of the decorticated
lamina was less than that of the intact lamina in ‘‘pull-
through’’ force using sublaminar wire.9,10 Actually, in the
setting of osteoporotic bone, especially in the decorticated
lamina, migration of sublaminar wire through soft
laminar bone occurs frequently. We considered BMD
and the thickness of laminar cortex as major factors
determining the fixation strength of a sublaminar wiring
construct. Normalizing the properties of the vertebrae
(eg, BMD and thickness of the laminar cortex) as we did
in each of the 3 groups is necessary to make fair
comparisons between the 3 different types of sublaminar
wiring constructs. Thus, the results from the cut-through
test (Experiment 1) suggest that tape has the advantage in
securing the osteoporotic lamina, as compared with steel
wire or cable. The narrow contact surface between the
steel wire and cable produces a high-local stress on the
lamina. This high-local stress results in lower values of
cut-through force. Tape, on the other hand, is broad and
so distributes the force better (ie, lower local stress), thus
allowing the lamina to resist a higher cut-through
force. The results from the biomechanical stiffness test

r 2006 Lippincott Williams & Wilkins 445

Murakami et al
(Experiment 2) revealed no difference (in terms of
stiffness) between the 3 types of sublaminar wire. That
may be a true result or may be because of the lack of
power with such a small sample. It is difficult to obtain
large numbers of cadaver spines that are suitable for
experiments such as these. Further, it could be that
stiffness may not be a good indicator for showing
differences between spinal constructs, unless the differ-
ences are large.
The following is the best information we could
obtain on the mechanical properties of the 3 different
types of sublaminar wires. Dickman et al11 reported that
polyethylene cable exhibited creep at loads that were
much lower than the static failure loads, namely
polyethylene cable 1.0 mm diameter stretched 43% at
800 N force, although it had similar strength to stainless
steel cable in the tensile testing. Also they reported that
titanium cable 1.2 mm diameter stretched 6% at 800 N
forces, and monofilament steel wire 0.8 mm diameter
stretched 9% at 250 N. Creep can cause a sublaminar wire
construct to loosen. We used tape made of woven ultra-
high molecular weight polyethylene fibers. This tape
allows little stretch and creep under tensile force, as
compared with standard polyethylene cable; the material
properties for the ultra-high molecular weight polyethy-
lene tape were supplied by Alfresa pharma. Furthermore,
the tape we used is soft and flexible; this property helps to
avoid neural damage while passing the tape under the
lamina.
CONCLUSIONS
In the laminar cut-through test, the cut-through
force for tape was significantly higher than for wire in
these osteoporotic vertebrae, but no significant difference
446
J Spinal Disord Tech  Volume 19, Number 6, August 2006
between tape and cable could be established. In the
biomechanical stiffness test, there was no significant
difference between the 3 different sublaminar wiring
configurations in any of the loading modes tested. The
results of both experiments suggest that tape is as good, if
not better, than steel wire and, perhaps, cable as a
sublaminar wiring implant material.
REFERENCES
1. Eysel P. Biomechanical principles of ventral and dorsal instrumen-
tation correction in scoliosis. Orthopade. 2000;29:507–517.
2. Yazar T, Gurkan I, Yilmaz C. A new approach to scoliosis. Eur
Spine J. 1999;8:86–92.
3. Hu SS. Internal fixation in the osteoporotic spine. Spine.
1997;22:43S–48S.
4. 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.
5. Murakami H, Boden SD, Hutton WC. Anterior lumbar interbody
fusion using a barbell-shaped cage: a biomechanical comparison.
J Spinal Disord. 2001;14:385–392.
6. Abumi K, Panjabi MM, Kramer KM, et al. Biomechanical
evaluation of lumbar spinal stability after graded facetectomies.
Spine. 1990;15:1142–1147.
7. Volkman T, Horton WC, Hutton WC. Transfacet screws with
lumbar interbody reconstruction: biomechanical study of motion
segment stiffness. J Spinal Disord. 1996;9:425–432.
8. Butler TE Jr, Asher MA, Jayaraman G, et al. The strength and
stiffness of thoracic implant anchors in osteoporotic spines. Spine.
1994;19:1956–1962.
9. Aydingoz O, Bilsel N, Botanlioglu H, et al. Effect of decortication
on laminar strength during sublaminar wiring: an experimental
study. J Spinal Disord Technol. 2004;17:498–504.
10. Crawford RJ, Sell PJ, Ali MS, et al. Segmental spinal instrumenta-
tion. A study of the mechanical properties of materials used for
sublaminar fixation. Spine. 1989;14:632–635.
11. Dickman CA, Papadopoulos SM, Crawford NR, et al. Comparative
mechanical properties of spinal cable and wire fixation systems.
Spine. 1997;22:596–604.
r 2006 Lippincott Williams & Wilkins