Intervertebral Biomechanics of Locking Compression Plate

Monocortical Fixation of the Canine Cervical Spine

Kimberly A. Agnello1, DVM, MS, Diplomate ACVS, Amy S. Kapatkin2,3, DVM, MS, Diplomate ACVS,
Tanya C. Garcia3, MS, Kei Hayashi2,3, DVM, PhD, Diplomate ACVS, Anja T. Welihozkiy3, DVM, and
Susan M. Stover3,4, DVM, PhD, Diplomate ACVS
1

William R. Pritchard Veterinary Medical Teaching Hospital, School of Veterinary Medicine, University of California-Davis, Davis, CA, 2Department of
Surgical and Radiological Sciences, School of Veterinary Medicine, University of California-Davis, Davis, CA, 3JD Wheat Veterinary Orthopedic
Research Laboratory, School of Veterinary Medicine, University of California-Davis, Davis, CA and 4Department of Anatomy, Physiology and Cell
Biology, School of Veterinary Medicine, University of California-Davis, Davis, CA

Corresponding Author
Dr. Kimberly A. Agnello, DVM, MS,
Diplomate ACVS, Matthew J. Ryan
Veterinary Hospital, Department of Clinical
Studies, School of Veterinary Medicine,
University of Pennsylvania, 3900 Delancey
Street, Philadelphia, PA 19104-6010.
E-mail: kagnello@vet.upenn.edu
Submitted September 2009
Accepted July 2010
DOI:10.1111/j.1532-950X.2010.00755.x

Objective: To evaluate the use of a locking compression plate (LCP) with monocortical screw purchase for stabilization of the canine cervical spine.
Study Design: Experimental study.
Animals: Cadaveric canine cervical spine specimens (n = 7).
Methods: Flexion and extension bending moments were applied to canine cadaveric specimens (C3C6) in 4-point bending, before and after creation of a ventral
slot at C4C5, and after xation with a 5 hole, 3.5 mm LCP with monocortical
screw placement. Screw placement and penetration into the vertebral canal were
determined by radiography. Range of motion, stiffness, and energy for passive
physiologic loads were determined for the C3C4, C4C5, and C5C6 vertebral
motion units (VMU). Monotonic failure properties were determined for cervical
extension. Effects of treatments on biomechanical variables were assessed using
repeated measures analysis of variance and least square means (P  .05).
Results: The ventral slot procedure increased range of motion at the treated
VMU. Plate xation decreased range of motion, increased stiffness, and decreased
energy at the treated VMU. No changes were observed at adjacent VMUs. None
of the screws penetrated the vertebral canal. Mean (  SD) yield bending moment
of plate stabilized, slotted spines was 15.6  4.6 N m.
Conclusion: LCP xation with monocortical screws stabilized the canine cervical
spine.

Vertebral fusion to stabilize acute fractures and luxations

of the vertebral column and chronic instabilities (eg, canine
cervical spondylomyelopathy [CCSM]) is performed in
dogs. CCSM encompasses multiple different disease entities, including chronic degenerative disc disease; congenital
osseous malformations; vertebral tipping; hypertrophy of
the ligamentum avum and vertebral arch malformations;
and hourglass compressions.13 In the dynamic form of
CCSM, compression of the spinal cord is induced by instability from exion and extension.1,2,4
Surgery is commonly recommended for dogs with
the dynamic form of CCSM to relieve spinal cord
compression, stabilize the vertebral column, and optimize
neurologic function.3,57 Steinmann pins and polymethylmethacrylate (pinPMMA) techniques are one of the

Presented in part at the 32nd Annual Conference of the Veterinary Orthopedic Society in Steamboat Springs, CO, February
28March 7, 2009.

common methods of xation.3,7,8 Despite reasonable clinical success, complications associated with pinPMMA xation of cervical vertebrae include implant failure before
bony union from pin migration, pin pull out, and failure at
the pin to PMMA interface; transverse or vertebral foraminae pin penetration with injury to the vertebral artery
and spinal cord; and soft tissue injuries (eg, thermal damage to the esophagus).2,713 PMMA is also likely to have a
higher incidence of complications than metallic implants
because of PMMAs inferior mechanical properties to
metal alloys (eg, low ductility), heat generation, release of
methylmethacrylate monomer into circulation, and hypersensitivity reactions.13
In canine lumbar vertebra, plate stabilization with
bicortical screws provided superior strength and rigidity
when compared in vitro with other forms of vertebral stabilization, including pinPMMA.14 Although, the current
study focuses on stabilization of the cervical vertebra,
similarities likely exist at these different spinal locations.
Analogous to the use of bicortical pins in the pinPMMA

c Copyright 2010 by The American College of Veterinary Surgeons

Veterinary Surgery 39 (2010) 9911000

991

Locking Compression Plate Monocortical Fixation of the Canine Cervical Spine

Agnello et al.

Specimens were wrapped in saline (0.9% NaCl)-solutionsoaked towels and plastic bags, and stored at 221C.
Specimens were thawed to room temperature (221C)
the day before testing, and mounted in a custom 4-point
bending xture. The C3 and C6 vertebral bodies were
secured to their respective ends of the xture by embedding half of each vertebra containing 2 orthogonal, transversely oriented, 2.4-mm-diameter Steinmann pins in
PMMA within larger xture pots. The xture conned motion to exion and extension of the vertebral column in the
sagittal plane. Dorsoventral and lateral radiographs were
performed to ensure adequate placement of the pins and
PMMA pots. The potted preparation was then placed in
saline-solution-soaked towels and plastic bags, and refrigerated overnight at 21C until testing the next day.

technique, bicortical screws used in plate stabilization of

the cervical vertebra likely have a similar risk of inadvertent vertebral canal penetration,12 because both cortices
need to be engaged for implant stability within the vertebra. The introduction of locking compression plating systems allows use of monocortical screw purchase, because of
the ability of the screw to rigidly engage the plate, which
will minimize motion and optimize stability of the screw
within the vertebra. Therefore, locking compression plate
(LCP) systems provide an attractive method of cervical
spinal stabilization to potentially decrease spinal canal penetration and subsequent spinal cord compromise. In the
human cervical spine, anterior plating systems have been
shown to be effective forms of stabilization,1519 and have
revealed signicant improvement in spinal fusion.20 In veterinary clinical studies, good outcome has been reported
after use of locking plate systems on the cervical spine,2123
but there is limited data examining the biomechanics of
these systems on the canine cervical spine.
Our purpose was to determine if use of a LCP with
monocortical screw placement can effectively stabilize cervical spinal segments in the dog. We hypothesized that a
LCP with monocortical screws will provide stabilization of
the canine cervical spine. Locking compression-platestabilized cervical vertebra are likely to minimize clinical
complications, such as vertebral foramen penetrations, implant failure before bone fusion, and PMMA-associated
morbidity.

A ventral slot procedure was performed at the C4C5 intervertebral disc space. Briey, the paired longus coli
muscles were separated over the disc space and a highspeed burr (Dremels rotary tool, Racine, WI) was used to
create the slot. The intervertebral disc, cortical and cancellous bone of the adjacent endplates, and underlying dorsal
longitudinal ligament were removed to achieve slot dimensions that correspond to removal of 1/3 of the length and
width of the C4 and C5 vertebral bodies.24 Radiographs
were used to conrm slot location and size.

MATERIALS AND METHODS

LCP Fixation Technique

Study Design
Cervical vertebral specimens were collected from 7 canine
cadavers. Each specimen was exed and extended in
4-point bending as an intact spine, after creation of a ventral slot at the C4C5 intervertebral space, and after C4C5
xation with a 5-hole 3.5 mm LCP with monocortical
screws (Synthess, West Chester, PA). The effect of treatment (intact, slot, plate) on C3C4, C4C5, and C5C6
intervertebral biomechanics was assessed using ANOVA.
Screw penetration into the vertebral foramen was assessed
using radiography.

Ventral Slot Technique

A 5-hole 3.5 mm LCP was applied, uncontoured, to the

ventral surface of the C4C5 vertebral bodies (Fig 1). Four
3.5 mm locking screws were inserted, the 1st screw in the
most cranial plate hole engaged the cranial aspect of C4,
the 2nd screw in the most caudal plate hole engaged in the
caudal aspect of C5, the 3rd screw engaged the caudal aspect of C4, and the 4th screw engaged the cranial aspect of
C5. A screw was not placed in the central plate hole, positioned over the intervertebral disc space. All screws were

Specimens
Vertebral specimens consisting of the 3rd, 4th, 5th, and 6th
cervical vertebra (C3C6) and surrounding musculature
were collected from 7 mature canine cadavers, weighing
2740 kg. A sample of 7 was chosen based on the ndings
of a similar study12 that found signicant differences between treatments for within specimen (n = 7) comparisons.
Dogs were euthanatized for reasons unrelated to this study
and had no known history indicative of spinal cord or vertebral column disease. None of the specimens had evidence
of orthopedic disease or vertebral physes on dorsoventral
and lateral radiographs (74 kVP, 0.75 mAs, focal lm
distance 36 in.; Sound-EklinTM Mark III, Carlsbad, CA).

992

Figure 1 Illustration with the locking compression plate fixed to the

cervical vertebral bodies of C4 and C5 with monocortical screw purchase.

c Copyright 2010 by The American College of Veterinary Surgeons

Veterinary Surgery 39 (2010) 9911000

Agnello et al.

Locking Compression Plate Monocortical Fixation of the Canine Cervical Spine

placed monocortically in the cis cortex and were tightened

to 1.5 N m of torque (Synthess torque limiting attachment,
Synthes). All screws penetrated 4 50% the vertebral body
dorsoventral dimension. Radiographs were taken after
plate xation to ensure plate placement over the C4C5
disc space, screw placement in the appropriate vertebral
body, and 4 50% screw purchase into the C4 and C5 vertebral bodies.

Mechanical Testing
A servohydraulic materials testing system (Model 319.25A/
T: MTS Corporation, MN) was used to apply 4-point
bending moments in exion and extension to the C3C6
spinal preparation. Reective markers were attached to C4
and to C5 using a bar attached to a 1/8th in. diameter
negatively, partially threaded pin (IMEX Veterinary Inc.,
Longview, TX) inserted into each pedicle (Fig 2). Correct
pin placement was veried using radiographs. Two reective markers were xed to each xture pot to track movement of the C3 and C6 vertebrae. Marker positions, and
thus vertebral body positions, were tracked at 60 Hz during
all tests using a high-resolution video camera (Fastcm PCI,
Photron USA Inc., San Diego, CA).
Specimens were tested for each condition (intact, slot,
plate) through 5 full exion and extension cycles at a rate of
0.1 Hz using 4-point bending under actuator displacement
control to achieve a physiologic passive bending moment.12
Preconditioning was performed under displacement

Figure 2 Illustration of the material testing system with a cervical spine

preparation, in the 4-point bending custom designed fixture arms. The
distance between the inner and outer supports was 5.9 cm and the distance between the inner supports was 29.3 cm. Two reflective markers
were attached to a bar inserted in the vertebral pedicle of C4 and of C5 to
track motion of these vertebrae and 2 reflective markers were attached
to each fixture pot to track movement of the C3 and C6 vertebral bodies.

control at a rate of 0.1 cycles/s for 5 cycles before each test

to a maximum exion and extension bending moment of
3 Nm. Specimens were loaded in the intact condition under
displacement control to end levels that approximated 531
exion and 681 extension of the entire specimen (ie, all
3 vertebral motion units [VMUs]). Because the stiffness of
the specimens were markedly different for each treatment
condition, and the xture had a displacement range limit,
slotted specimens were tested to the same displacement
range as that of their respective intact condition, whereas
plated specimens were tested to the displacement range that
produced a bending moment range similar to the bending
moments of the intact condition. However, each specimen
achieved estimated physiologic bending moments without
visible or biomechanical evidence of damage. The specimens were removed from the test xture during slot creation and plate xation.
After cyclic testing for all treatment conditions, plated
specimens were loaded in extension in a single load to
failure at 10 mm/s to the displacement limit of the bending
xture. After the test, specimens were radiographed to
determine if specimen failure occurred and corresponding
mode of failure.
Mechanical and Kinematic Data Reduction
Data from the 5 consecutive cycles of loading during a test
were examined to verify that specimen stiffness and peak
loads did not deteriorate during the test because deterioration would be consistent with a damaged specimen. Data
from the 3rd test cycle were analyzed to avoid potential
artifacts from initiation and suspension of cycling during
the rst and last cycles of each test (an incomplete cycle for
exion or extension), and the 3rd cycle appeared representative of the 2nd through 4th cycles. Angular deformations
of the video markers in each VMU were determined using
digitization and processing software (Motus, VICON/
Peak, Denver, CO). Marker angular deformations were
transformed to VMU angular deformations by correcting
for any angle between a line drawn through the vertebral
markers relative to a line along the ventral oor of the
respective vertebral body canal using lateral radiographs of
marked specimen preparations.
Bending moments were determined from the axial
loads and geometry of the 4-point bending xture using
the equation: M = (P/2)  W; where M is the bending moment (Nm), P is the applied load (N), and W is the distance
(m) between the inner and outer swing supports. Bending
moment versus angular deformation curves were constructed from the resulting data.
As described previously,12 the neutral zone (NZ) is the
region of the bending moment versus angular deformation
curve between 0.4 N m of exion and extension. This region
of the curve is where small bending moments cause large
angular deformations. The stiffness zone (SZ) is dened as
the region of the curve between 0.4 and 4 N m for exion
and extension. The SZ corresponds to the region where
high bending moments cause small angular deformations.

c Copyright 2010 by The American College of Veterinary Surgeons

Veterinary Surgery 39 (2010) 9911000

993

Locking Compression Plate Monocortical Fixation of the Canine Cervical Spine

Agnello et al.

Figure 3 Modified from Koehler et al.12 Bending moment versus angular deformation curve derived from mechanical testing of the intact (untreated)
C3C6 vertebral column of cadaveric dogs. The neutral zone (NZ) is defined as the region of the curve between 0.4 Nm of extension (NZe) and flexion
(NZf), and the stiffness zone (SZ) is the region of the curve between 0.4 and 4 Nm for extension (SZe) and flexion (SZf). The radiographic images
illustrate extension and flexion corresponding to the extension and flexion portions of the curve. The 1801 vertical line represents the position of the
spine at which the dorsal boarders of the vertebral bodies are aligned. The dashed lines indicate the SZ stiffness for SZe and SZf.

A custom program (Matlab, The Mathworks, Natick,

MA) was used to determine NZ and SZ, and calculate stiffness and energy. Briey, VMU stiffness was calculated as
the least square means slope of the respective portions (NZ,
SZ for exion and extension) of the bending moment
angular deformation curves. Corresponding energies were
calculated as the area under the respective portions of the
curve. Range of motion was dened as the angular deformation for each zone for exion and extension (Fig 3).

examined using ANOVA (separate for each VMU) that

accounted for repeated measures within specimens (PROC
MIXED, SAS 9.1, SAS Institute Inc., Cary, NC). Treatment was considered a xed effect. Pairwise comparisons
between treatments were examined using least square
means. Differences are reported when the P-value for a
statistical test was  .05. Trends are reported when the
P-value for statistical tests were marginally insignicant
(.05 o P  .10).

Screw Position
Screw positions within a vertebrae after plate application
were categorized by vertebral purchase ( o 50% or 4 50%
of vertebral body dorsoventral depth), vertebral canal
penetration (present/absent), and intervertebral space
penetration (present/absent).
Failure Analysis
Bending moment versus angular deformation curves were
constructed for the C4C5 VMU resulting from the single
load-to-failure test. Because all specimens demonstrated
approximately linear bending moments versus angular deformation curves at the region between 5 and 8 Nm, a least
squares linear t was calculated through that data to dene
a linear elastic region. The yield point was determined
using the 0.1% angular displacement offset criteria from
the linear region.
Statistical Analysis
The effects of treatment (intact, slot, plate) on mechanical
properties in the neutral and SZs of the C3C4, C4C5,
and C5C6 VMUs during exion and extension were

994

RESULTS
C4C5 VMU
Ventral Slot. Creation of a ventral slot affected predominantly the C4C5 VMU (Table 1). Total ROM increased
by 58% (211), NZROM by 57% (101), and extension
ROM by 71% (151), when compared with the intact specimen (Fig 4). The ventral slot also increased the mean range
of the extension NZ by 63% (61). A similar trend of lesser
degree with an increase of 47% (3.61) occurred through
exion in the NZ. There were no statistically signicant
differences in stiffness (Fig 5). The slot increased the energy
required to move the C4C5 VMU through extension in
the NZ (Fig 6).
LCP xation. Fixation of the C4C5 VMU with an LCP
plate after ventral slot creation signicantly decreased total
ROM by 70% (401), the NZROM by 96% (261), exion
ROM by 85% (191), and extension ROM by 60% (221)
compared with the slotted VMU (Fig 4). LCP xation increased NZ stiffness of the slotted spine by 3420% in exion and 3080% in extension (Fig 5). Energy to deform the

c Copyright 2010 by The American College of Veterinary Surgeons

Veterinary Surgery 39 (2010) 9911000

SZ

NZ

Flex
Ext
Flex
Ext

Flex
Ext
Flex
Ext

c Copyright 2010 by The American College of Veterinary Surgeons

Veterinary Surgery 39 (2010) 9911000

0.069  0.089
0.022  0.044
0.643  0.594
0.531  0.188
2.14  2.34
2.63  1.05
16.6  17.0
22.3  7.7
46.2  11.1
19.0  5.3
18.7  8.7
27.5  10.1

1.95  1.33
2.42  1.00
15.4  5.7
19.8  5.2

45.6  9.0
22.5  6.3
20.1  7.7
25.6  6.8

6.4  3.3
12.6  6.0
10.8  4.7
16.4  8.0

Ventral Slot

0.034  0.019
0.036  0.039
0.600  0.297
0.501  0.129

9.5  4.9
12.9  5.5
10.5  5.1
12.7  2.4

Intact

43.9  8.5
19.1  3.8
17.3  8.5
26.6  10.9

1.75  1.14
2.11  0.94
10.7  2.9
18.5  2.8

0.051  0.029
0.036  0.039
1.082  0.343 
0.558  0.100

7.6  3.7
12.3  5.2
9.0  5.0
14.7  6.1

Plate

36.6  7.7
17.5  6.6
15.6  7.1
21.0  5.7

1.56  1.21
1.80  0.68
11.9  3.2
16.7  3.8

0.046  0.028
0.045  0.046
0.731  0.278
0.622  0.217

7.60  5.2
9.9  4.5
8.3  3.1
10.9  2.0

Intact

57.7  12.1
27.3  7.2
21.7  11.2
36.0  12.6

3.72  3.74
3.58  1.95
15.1  11.7
20.9  11.3

0.026  0.019
0.025  0.058
1.236  1.087
0.992  0.606

11.2  7.4
16.1  8.8
10.4  6.6
20.0  9.7

Ventral Slot

C4C5 VMU

17.6  24.3
1.2  1.5
3.2  3.0
14.4  24.4

0.26  0.34
0.16  0.33
26.8  55.0
8.6  10.0

0.889  1.226
0.770  1.119
1.825  1.481
1.566  0.834

0.9  1.3
0.2  0.6
12.4  22.5
6.3  7.8

Plate

w,n

,w,n
,w,n
,n

,n,w

,n
,n



,n
,n,w

39.1  11.1
17.5  6.6
16.9  8.7
22.1  7.6

1.47  1.11
1.9  1.0
15.3  10.0
18.3  3.4

0.421  1.146
0.045  0.047
0.736  0.659
0.549  0.142

7.2  5.2
10.3  5.3
9.9  5.3
11.7  2.5

Intact

42.2  21.0
17.5  9.3
16.7  11.7
25.6  13.2

1.54  2.24
2.2  1.6
20.8  18.7
19.5  6.9

0.042  0.076
0.026  0.042
0.861  0.707
0.585  0.159

5.6  5.8
10.5  6.1
10.8  7.0
14.0  9.0

Ventral Slot

C5C6 VMU

32.2  14.0
13.9  4.9
13.8  10.1
18.5  8.7

1.33  1.14
1.5  0.8
11.1  5.2
14.9  8.0

0.070  0.033
0.085  0.129
1.534  1.530
0.719  0.441

5.4  3.1
9.0  3.9
8.8  5.7
9.6  5.6

Plate

Symbols in the column to the right of summary data for the C3C4 and C4C5 VMUs indicate statistical significance (P o 0.05) among treatments for the row variable. The intact specimen values
are significantly different than plated specimens (), the intact specimen values are significantly different than the slotted values (w), and the plated specimens are significantly different than the
slotted values (n). There are no statistically significant differences among treatments for the C5C6 VMU.
VMU, vertebral motion units.

Flex
Ext
SZ
Flex
Ext
Total range of motion (1)
SZ
NZ
Flex
Ext

Energy (N m1)
NZ

SZ

Stiffness (N m/1)
NZ

Range (1)

Variable Zone Motion (1)

C3C4 VMU

Table 1 Mean ( SD) Mechanical Variables for 7 Canine Cervical Cadaveric Specimens for the C3C4, C4C5, and C5C6 Vertebral Motion Units After No Treatment (Intact), Ventral Slot Treatment, and Locking Compression Plate Fixation

Agnello et al.
Locking Compression Plate Monocortical Fixation of the Canine Cervical Spine

995

Locking Compression Plate Monocortical Fixation of the Canine Cervical Spine

Figure 4 Flexion (negative values) and extension (positive values)

range of motion mean ( SD) data for the intact, ventral slot, and plate
fixation treatments. The light gray portion of the bars represents the
neutral zone (NZ), and the dark gray portion represents the stiffness
zone (SZ).

VMU was also signicantly decreased by LCP xation

of the slotted spine for exion and extension in the NZ
(Fig 6).
Compared with the intact specimen, LCP xation resulted in a 52% (141) signicant decrease in total ROM, a
93% (161) signicant decrease in total NZROM, and an
80% (201) decrease in exion ROM (Fig 4). Plate application as compared with the intact spine signicantly
increased stiffness of the NZ by 1930% in exion and by
1710% in extension; and in the SZ by 250% in exion and
by 252% in extension (Fig 5). Plate xation compared with
the intact spine signicantly decreased NZ energy in exten-

996

Agnello et al.

Figure 5 Flexion (negative values) and extension (positive values)

stiffness mean ( SD) data for the intact, ventral slot, and plate fixation
treatments. The light gray bars represent the neutral zones (NZ), and the
dark gray bars represent the stiffness zone (SZ).

sion, and a trend toward increased energy was noted in the

SZ with the spine in extension (Fig 6).

Adjacent VMUs
An 80% signicant increase in SZ stiffness in exion at the
C3C4 VMU was noted with plate application when compared with the intact specimen (Fig 5). No other signicant
changes in range of motion, stiffness, and energy were
noted at the adjacent C3C4 and C5C6 VMUs (Table 1).

c Copyright 2010 by The American College of Veterinary Surgeons

Veterinary Surgery 39 (2010) 9911000

Agnello et al.

Locking Compression Plate Monocortical Fixation of the Canine Cervical Spine

bral body through an outer screw hole. Mean yield bending

moment for all specimens was 15.6  4.6 Nm (n = 7).

DISCUSSION

Figure 6 Flexion (negative values) and extension (positive values) energy mean ( SD) data for the intact, ventral slot, and plate fixation treatments. The light gray portion of the bars represents the neutral zone
(NZ), and the dark gray portion represents the stiffness zone (SZ).

Screw Position
All screws had 4 50% purchase into the vertebral body.
None of the screws penetrated the vertebral canal. All
screws fully engaged the vertebral bodies without penetration into the intervertebral disc space.

Tests to Failure
Only 2 specimens failed because the test xture had insufcient ROM for the remaining specimens to reach failure. In
both constructs, failure occurred by fracture of the verte-

We found that the ventral slot procedure resulted in an

increase in ROM at the C4C5 VMU and plate stabilization resulted in a marked decrease in ROM, an increase in
stiffness, and a decrease in the energy required to deform
the construct at the C4C5 VMU. No implant penetration
into the vertebral canal was identied with plate xation.
LCP with monocortical screw xation offers a safe and
easy method for stabilization of the cervical spine.
The ventral slot resulted in marked passive instability
at the treated VMU similar to other studies.12,24,25 Expansion of the NZ, SZ, and total ROMs occurred, particularly
in cervical spinal extension. In in vitro studies, passive mechanical instability occurs with the ventral slot procedure,
although it is unclear if instability occurs clinically.26 In live
animals, active muscle contraction may compensate for
compromised passive intervertebral motion restraints.
However, in vivo canine studies have shown immediate
postoperative increases in total range of motion (ie, instability) after facetectomy, with subsequent temporal
increases in stability.27,28 In vitro study results may represent the worst case clinical scenario after ventral slot procedure. Healing and adaptation of the active and passive
stabilizers may contribute to at least partial recovery of stability, but even initial instability may warrant stabilization
post ventral slot procedure for optimal patient outcome.
Most studies that have examined spinal instability
evaluated total ROM, but not ROM of the NZ.2628 In
people, clinical instability has been dened as the inability
to maintain the intervertebral NZs within physiologic limits, which results in neurologic dysfunction, deformities,
and/or pain29,30 Therefore, NZROM is likely a better
indicator of vertebral instability than total physiologic
ROM.31 The ideal cervical vertebral stabilization technique
would be one that stabilizes the NZROM31 and recreates
the biomechanical behavior of the intact healthy VMU.
Unfortunately, the implants currently available for veterinary applications are too stiff to allow normal intervertebral motion. It is unclear how normal vertebral motion
would be regained without engineering an intervertebral
disc that recreates the geometry, location, and mechanical
behavior of the normal canine intervertebral disc. In the
meantime, regaining intervertebral distraction and stabilization to prevent impingement of the spinal cord by extruded disc, hypertrophied ligamentous structures, and
subluxation of the vertebral bodies is critical for pain-free
optimal cervical motion without neurologic complications.
Therefore, the goals of spinal stabilization at this point are
to stabilize the NZROM, facilitate vertebral fusion, minimize excessive bone formation that could impinge on the
spinal canal and associated nerve roots, decrease or eliminate pain, and eliminate spinal cord compromise. The LCP
with monocortical xation, signicantly reduced all aspects

c Copyright 2010 by The American College of Veterinary Surgeons

Veterinary Surgery 39 (2010) 9911000

997

Locking Compression Plate Monocortical Fixation of the Canine Cervical Spine

of ROM at the treated VMU, and increased stiffness and

decreased the NZ energy needed for exion and extension
of the treated VMU. As stated earlier, the NZ appears to be
the best indicator available of assessing spinal stability31
and in the LCP plated spines, the largest decrease in ROM
(96%), increase in stiffness (97%), and decrease in energy
were seen in the NZ. The LCP imparted a similar amount
of cervical stability to that of pins-PMMA.12
Clinical complications are likely to be lower with LCP
vertebral stabilization than other commonly used methods. The use of a LCP allowed placement of monocortical
screws in the cervical vertebra because of the xed angle
construct design (Synthess). Screw engagement can be
conservative, because the screw rigidly engages the plate,
which minimizes motion and optimizes stability of the
screw in the vertebra. None of the screws in this study penetrated the vertebral foramina. Use of a conventional plate
or pins-PMMA, which require engagement of the screws
and pins in the vertebral body for screw stability in the
vertebra, has greater potential for implant impingement of
the spinal cord. In an in vitro study similar to the current
study, the use of pins-PMMA for stabilization of the cervical vertebra resulted in protrusion of 29% of pins into
the vertebral canal and a combined 41% protrusion into
either the vertebral or transverse foraminae.12 Use of a
plate also eliminates some of the other reported complications with the use of pins-PMMA, such as pin migration
with subsequent neurologic deterioration,9 and risks inherent to PMMA including thermal heat injury, release of
methylmethacrylate into circulation, and hypersensitivity
reactions.13
Interestingly in our study, creation of the ventral slot
at C4C5 and application of the LCP did not alter the mechanics of the adjacent VMUs, except for an increase in SZ
stiffness in exion at the C3C4 VMU. This differs from
other reports and likely reects technique differences between studies. First, our ventral slot may have been smaller
than that created by Koehler et al.12 Our slot dimensions
were standardized for specimen size according to clinical
recommendations for slot width and length.24 To be consistent with clinical practice, the bers of the dorsal longitudinal ligament were also disrupted to expose the spinal
cord. Slot length in the current study (o 1/3rd the vertebral
body length) was slightly larger than used by Koehler
et al. study12 (o 1/4th vertebral body length). Unfortunately, because of the subjective nature of this criteria and
the different surgeons who performed the procedure in the
2 studies, it is difcult to determine actual variation in slot
size between the 2 studies. However, the increase in
NZROM at the slotted VMU in our study (57%) was
smaller than that of the Koehler et al.12 (98%), which
would be consistent our slot being smaller. Second, specimen dissection may have been different between the 2 studies. Perivertebral soft tissues could provide passive support
to cervical specimens. More soft tissue support might protect adjacent VMUs from treatment effects at the C4C5
VMU. Although the method for assessing vertebral body
motion differed between studies, the differences among

998

Agnello et al.

treatments should have been consistent within each study.

The trend in the mean values among treatments at adjacent
VMUs observed in the Koehler study12 was not observed in
our study. Consequently, the discrepancy in treatment
effects on adjacent VMU biomechanics are more likely related to differences in specimen dissection, slot technique,
and/or stabilization technique than differences in study design. Seemingly, variation in ventral slot size within the
current recommended slot dimensions along with extent of
perivertebral soft tissue dissection may affect the biomechanics of the treated and untreated VMUs; however, the
clinical consequences of these differences are unknown at
this time.
The bending tests of cervical spine constructs presented several challenges with corresponding limitations.
Cervical spine geometry is neither simple nor uniform, and
range of motion is large. Ideally, specimens would have
been tested using load control, as occurs naturally through
a range corresponding to physiologic exion and extension
bending moments at a physiologic, constant angular deformation testing rate. In order for testing machines using
closed loop feedback to maintain load control (corresponding to bending moment control), the machine must be
tuned to expected specimen stiffness so that actuator displacement will result in an expected specimen load. Specimen stiffness must be similar or predictable so that the
machine can be tuned appropriately for each test, or unexpected specimen stiffness must be compensated for by
machine experience and feedback during sequential cycles
of testing. However, stiffness of the spinal preparations is
unknown, markedly different among the 3 treatment conditions within specimens, and would degrade with the cyclic testing needed for compensation because of creep of
supporting soft tissues. Consequently, the tests were run
using actuator displacement control, which is a straightforward procedure for mechanical test machines. The challenge is that specimens with different treatments would
have different displacements for the desired range of physiologic bending moments. For example, the plated specimen would have failed (at supraphysiologic loads) if it had
been tested to the same displacement range as that of the
intact specimen because the plated specimen was much
stiffer than the intact specimen. An additional factor was
the physical range of the xture, which limited the range
that the specimen with a slot could have exed and extended. It is likely that the paravertebral muscles and soft
tissues that were removed in vitro for testing purposes,
would have also restricted cervical spine motion in vivo.
Consequently, the displacement ranges were standardized
to either the displacement range of the intact condition (for
the slotted condition) or to the displacement range that
produced a bending moment range similar to the bending
moments of the intact condition (for the plated condition).
However, each specimen achieved estimated physiologic
bending moments without visible or biomechanical evidence of damage as would occur with overloading; and
mechanical properties were compared among treatments at
standardized bending moments.

c Copyright 2010 by The American College of Veterinary Surgeons

Veterinary Surgery 39 (2010) 9911000

Agnello et al.

Locking Compression Plate Monocortical Fixation of the Canine Cervical Spine

Second, because biological tissues are viscoelastic, results are load rate dependent, so the loading rate should be
constant within and among tests. However, the load rates
among test conditions and within a test condition differed.
Because all specimens were cycled at 0.1 Hz and the displacement range was lower for plated specimens than for
intact and slot specimens, the load rate was slower for
plated specimens. Additionally, linear actuator displacement translates into nonlinear angular deformation. The
angular deformation rate is faster as the angle increases,
and thus faster nearer the extremes of linear actuator displacement. The effects of different angular deformation
rates among treatment conditions and within tests are
unknown; but because the load rate was relatively slow for
all tests and the effects of different treatments are large, it is
assumed that effects of load rate on the results of the study
are relatively small and negligible.
A third challenge is to prevent undesirable tension
along the longitudinal axis of the specimen during bending.
As the specimen bends, the length of the arc of curvature
between the load supports increases. Similar to a dog that
can move the head closer to the body as the neck exes, the
slotted xture arms allow for movement of the specimen
relative to the load supports to allow lengthening of the
preparation between the load supports during bending
while preventing artifactual longitudinal stresses within
the vertebrae.
Preconditioning was used to minimize unrepeatable
mechanical behavior that occurs with initial loading at low
loads. The source of this behavior is unknown, but it is
thought to be related to irregularities in attachment of the
specimen to the xture that collapse at low loads, tissues
with low stiffness that yield at low loads (eg, remaining
strands of muscle on the specimen after dissection), or
unrecoverable creep of specimen tissues.
The study capitalized on a repeated measures study
design, because of high expected individual (ie, interspecimen) variation. Standard deviations are high relative to the
magnitude of the means. However, comparisons among the
repeated measures for the treatments within specimens
were consistent and yielded statistically apparent effects of
treatment condition for most biomechanical properties for
the treated VMU. High variability in energy variables
might be attributed to the reliance of energy calculations
(area under the bending moment versus angular deformation curves) on not only the endpoints of the range of data
used, but also on the shape of the curve between endpoints.
High variability likely reects complicated mechanical behavior between vertebral bodies especially where low loads
are less effective in constraining specimen behavior than
high loads. There may also be complicated behavior between vertebral bodies with complicated geometries and
multiple articulations (intervertebral joint and bilateral
facets) that we have yet to understand. If the sample size
were larger, values for more energy variables might be statistically different between treatment conditions, but biological relevance of the magnitude of the difference should
still be considered. The power of the ndings of no statis-

tically signicant differences are likely to be low reecting

the lack of condence in the nding that there is no effect of
treatment on respective variables.
One limitation to the use of this plate was that plate
size and hole conformation must be compatible with vertebral size. The plate we used was selected because of availability in most locking plate sets, which eliminates the need
for inventory of specialty plates; however, the plate was too
large for 2 of the original specimens in this study resulting
in one of the outer screws placed into the intervertebral disc
space. The affected specimens failed during testing by pullout of the sole remaining screw in the vertebral body. Data
from these tests were excluded from the current study. The
advent of various sized locking plates and screws may eliminate this constraint to plate usage.
Specimens in the current study were only tested in exion and extension in the sagittal plane. Clinically, spinal
cord compression occurs with exion and extension, and
ROM is greatest in exion and extension compared with
lateral bending and rotation.25 Also, exion and extension
at the C4C5 VMU are uncoupled motions, whereas lateral
bending and rotation are strongly coupled.32 However, the
effects of a ventral slot and subsequent stabilization on
biomechanics of the LCP xation in lateral bending and
rotation could be biologically relevant, but are unknown.
In conclusion, the LCP with monocortical screws
offers stability to the canine cervical spine after the ventral
slot procedure. The decrease in ROM at the treated VMU
was similar to that imposed by the pins-PMMA xation.12
Therefore, the use of a LCP with monocortical screws may
offer a safe and easy stabilization method for cervical spinal stabilization that will decrease the incidence of spinal
canal impingement and the eliminate complications associated with the use of PMMA.

ACKNOWLEDGMENT
Funded in part by the Veterinary Orthopedic Society, HohnJohnson Research Award, and Synthess, West Chester, PA.

REFERENCES
1. Seim HB, Withrow SJ: Pathophysiology and diagnosis of
caudal cervical spondylo-myelopathy with emphasis on the
Doberman Pinscher. J Am Anim Hosp Assoc 1982;18:241251
2. McKee WM, Sharp NJH: Cervical spondylomyelopathy, in
Slatter D (ed): Textbook of Small Animal Surgery (ed 3).
Philadelphia, PA, W. B. Saunders, 2003, pp 11801188
3. Seim HB: Wobbler syndrome, in Fossum TW (ed): Small
Animal Surgery (ed 2). St Louis, MO, Mosby, 2002, pp
12371249
4. Olsson SE, Stavenborn M, Hoppe F: Dynamic compression
of the cervical spinal cord. Acta Vet Scand 1982;23:6578
5. VanGundy T: Canine wobbler syndrome. Part II. Treatment.
Compend Contin Educ Pract Vet 1989;11:269284
6. Hurov LI: Treatment of cervical instability in the dog. J Am
Vet Med Assoc 1979;175:278285

c Copyright 2010 by The American College of Veterinary Surgeons

Veterinary Surgery 39 (2010) 9911000

999

Locking Compression Plate Monocortical Fixation of the Canine Cervical Spine

7. Bruecker KA, Seim HB, Blass CE: Caudal cervical

spondylomyelopathy: decompression by linear traction and
stabilization with Steinmann pins and polymethylmethacrylate. J Am Anim Hosp Assoc 1989;25:677683
8. Bruecker KA, Seim HB, Withrow SJ: Clinical evaluation of
three surgical methods for treatment of caudal cervical
spondylomyelopathy of dogs. Vet Surg 1989;18:197203
9. Blass CE, Seim HB: Spinal xation in dogs using Steinmann
pins and methylmethacrylate. Vet Surg 1984;3:203210
10. Bruecker KA, Seim HB, Blass CE: Caudal cervical
spondylomyelopathy: decompression by linear traction and
stabilization with Steinmann pins and polymethylmethacrylate. J Am Anim Hosp Assoc 1989;27:677683
11. Sharp NJH, Wheeler SJ: Cervical Spondylomyelopathy: Small
Animal Spinal Disorders (ed 2). Edinburgh, Elsevier-Mosby,
2005, pp 211246
12. Koehler CL, Stover SM, LeCouteur RA, et al: Effect of a
ventral slot procedure and of smooth or positive-prole
threaded pins with polymethylmethacrylate xation on
intervertebral biomechanics at treated and adjacent canine
cervical vertebral motion units. Am J Vet Res 2005;66:
678687
13. Smith GK: Orthopedic biomaterials, in Newton W,
Nunamaker DM (eds): Textbook of Small Animal
Orthopedics. Philadelphia, PA, Lippincott, 1985, pp 231241
14. Walter MC, Smith GK, Newton CD: Canine lumbar spinal
internal xation techniques. A comparative biomechanical
study. Vet Surg 1986;15:191198
15. Aebi M, Zuber K, Marchesi D: Treatment of cervical spine
injuries with anterior plating: indications, techniques, and
results. Spine 1991;16(Suppl): 3845
16. Bohler
J, Ganderhak T: Anterior plate stabilization for

fracture-dislocations of the lower cervical spine. J Trauma

1980;20:203205
17. Brown JA, Havel P, Ebraheim N, et al: Cervical stabilization
by plate and bone fusion. Spine 1988;13:236240
18. Smith SA, Lindsey RW, Doherty BJ, et al: An in-vitro
biomechanical comparison of the Orosco and AO locking
plates for anterior cervical spine xation. J Spinal Disord
1995;8:220223
19. DiAngelo DJ, Foley KT, Vossel KA: Anterior cervical
plating reverses load transfer through multilevel strut-grafts.
Spine 2000;25:783795

1000

Agnello et al.

20. Kaiser MG, Haid RW, Subach BR, et al: Anterior cervical
plating enhances arthrodesis after discectomy and fusion with
cortical allograft. Neurosurgery 2002;50:229238
21. Trotter EJ: Cervical spine locking plate xation for treatment
of cervical spondylotic myelopathy in large breed dogs. Vet
Surg 2009;38:705718
22. Bergman RL, Levine JM, Coates JR, et al: Cervical spinal
locking plate in combination with cortical ring allograft for a
one level fusion in dogs with cervical spondylotic myelopathy.
Vet Surg 2008;37:530536
23. Voss K, Steffen F, Montavon PM: Use of the ComPact
unilock system for ventral stabilization procedures of
the cervical spine. Vet Comp Orthop Tramatol 2006;19:
2128
24. Fauber AE, Wade JA, Lipka AE, et al: Effect of width of disk
fenestration and a ventral slot on biomechanics of the canine
C5C6 vertebral motion unit. Am J Vet Res 2006;67:
18441848
25. Macy NB, Les CM, Stover SM, et al: Effect of disk
fenestration on sagittal kinematics of the canine C5C6
intervertebral space. Vet Surg 1999;28:171179
26. Panjabi MM, Abumi K, Duranceau J, et al: Spinal stability
and intersegmental muscle forces: a biomechanical model.
Spine 1989;14:194200
27. Crisco JJ, Panjabi MM, Wang E, et al: The injured canine
cervical spine after six months of healing: an in vitro threedimensional study. Spine 1990;15:10471052
28. Buff HU, Panjabi MM, Sonu CM, et al: Functional stability
of the canine cervical spine after injury: a three-month in vivo
study. Spine 1990;15:10401046
29. White AA, Panjabi MM (eds): Clinical Biomechanics of the
Spine (ed 2). Philadelphia, PA, Lippincott, 1990
30. Panjabi MM: The stabilizing system of the spine. Part II.
Neutral zone and instability hypothesis. J Spinal Disord
1992;5:390396
31. Panjabi MM: Clinical spinal instability and low back pain.
J Electomyography Kinesiol 2003;13:371379

32. Hofstetter M, Gedet

P, Doherr M, et al: Biomechanical
analysis of the three-dimensional motion paternal of the
canine cervical spine segment C4C5. Vet Surg 2009;38:
4958

c Copyright 2010 by The American College of Veterinary Surgeons

Veterinary Surgery 39 (2010) 9911000