Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
A preliminary investigation to determine the effects of a crosslinking reagent on the fatigue resistance of the posterior annulus of the intervertebral disc
(USC Thesis Other)
A preliminary investigation to determine the effects of a crosslinking reagent on the fatigue resistance of the posterior annulus of the intervertebral disc
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A PRELIMINARY INVESTIGATION TO DETERMINE THE
EFFECTS OF A CROSSLINKING REAGENT ON THE FATIGUE
RESISTANCE OF THE POSTERIOR ANNULUS OF THE
INTERVERTEBRAL DISC
by
Dean Michael Gray
A Thesis Presented to the
FACULTY OF THE SCHOOL OF ENGINEERING
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment o f the
Requirements for the Degree
MASTER OF SCIENCE
(BIOMEDICAL ENGINEERING)
May 2002
Copyright 2002 Dean Michael Gray
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
This thesis, written by
Dean Michael Gray
under the guidance of his/her Faculty Committee and
approved by all its members, has been presented to and
accepted by the School of Engineering in partial
fulfillm ent of the requirements for the degree of
Master of Science
L i £ 2 m e d i c a l _ _ E r i a i j Q £ . e j : i n i a
Date: udnuaxy 2002
Faculty Committee
C hairm an
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
D & d ic f tt ip n
Dedicated to Vanessa, for her endless patience and support
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T able of Contents
Dedication ii
Acknowledgements jji
List o f Tables v
List o f Figures vi
Abstract ix
Chapter 1 Low Back Pain, Repetitive Loading and Interest in the 1
Intervertebral Disc
Chapter 2 Non-Destructive Testing Methods used to Determine 10
the Material Properties o f Intervertebral Disc A n n u la r
Tissue
Chapter 3 The Response o f the Annulus Fibrosis to Repetitive 33
Loading
Chapter 4 The Impact o f Crosslinking on the Collagenous Tissue 43
o f the Annulus Fibrosis
Chapter 5 A Pilot Study Designed to Determine the Impact o f a 53
Biochemical Treatment on the Fatigue Response o f the
Annulus Fibrosis
Chapter 6 Interpretation o f the Results o f the Pilot Study 66
Chapter 7 Outstanding Questions and Recommendations 97
Bibliography 108
Appendix 114
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
L ist of Tables
Table 2-1: Regional variation test statistics. 30
Table 5-1: Pilot study mechanical test protocol 58
Table 7-1: Fatigue induced disc dehydration study test protocol 101
Table 7-2: Fatigue induced disc dehydration study results. 101
Table 7-3: Fatigue induced disc dehydration study statistics. 102
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
L ist o f Figures
Figure 1-1: Flexion-compression o f spine - axial and bending
components.
7
Figure 1-2: Flexion-compression o f spine - annulus stresses. 8
Figure 2-1: Schematic stress-strain curve. 11
Figure 2-2: Indentation nomenclature. 14
Figure 2-3: Creep deformation curve — conventional materials. 18
Figure 2-4: Creep deformation curve — biological soft tissues. 19
Figure 2-5: Long duration creep test results. 22
Figure 2-6: Repeated viscoelastic measurement test results. 25
Figure 2-7: Repeated elastic-plastic measurement test results. 26
Figure 2-8: Lateral effects testing results. 27
Figure 2-9: Regional variation test results. 29
Figure 2-10: Regional variation test results — improved technique. 30
Figure 3-1: Fatigue loading of posterior annulus — tensile test results. 38
Figure 3-2: Fatigue loading of posterior annulus - indentation test
results.
39
Figure 3-3: Fatigue loading of posterior annulus — combined results. 40
Figure 4-1: Age related changes in crosslink concentration within
the intervertebral disc.
47
Figure 4-2: Age related changes in low back pain incidence. 49
Figure 5-1: Depth o f penetration o f genipin treatment. 55
Figure 5-2: Indentation testing apparatus. 60
Figure 5-3: Fatigue testing apparatus. 61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 5-4: Fatigue testing schematic. 62
Figure 5-5: Fatigue testing mechanical model. 63
Figure 6-1: Pilot study alternate test procedure stress relaxation
results.
66
Figure 6-2: Pilot study alternate test procedure creep deformation
results.
67
Figure 6-3: Pilot study alternate test procedure hardness results. 67
Figure 6-4: Stress relaxation data sample. 70
Figure 6-5: Stress relaxation data sample — with regression curves. 71
Figure 6-6: Calculation o f the coefficient o f determination. 73
Figure 6-7: Pilot study stress relaxation results. 74
Figure 6-8: Creep deformation data sample. 76
Figure 6-9: Creep deformation data sample — with regression curve. 77
Figure 6-10: Creep deformation data sample — controller overshoot
and error.
79
Figure 6-11: Creep deformation data sample — controller drift. 79
Figure 6-12: Pilot study creep deformation results. 81
Figure 6-13: Pilot study creep deformation results — creep rate. 82
Figure 6-14: Hardness data sample. 85
Figure 6-15: Hardness data sample — with regression curves. 86
Figure 6-16: Hardness data sample — linear force ramp. 87
Figure 6-17: Hardness data sample — non-linear force ramp. 88
Figure 6-18: Pilot study hardness results. 89
Figure 6-19: Pilot study stress relaxation results — effect o f genipin
concentration.
92
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A bstract
The posterior annulus has been found to be the primary site of injury to the
intervertebral disc and a likely source o f low-back pain. Epidemiological studies
implicate repetitive compression-flexion loading in the acceleration o f disc
degeneration and the onset o f low-back symptoms. This loading condition places
the posterior annulus in tension, and prior studies in our laboratory have shown a
correlation between fatigue tensile loading and a degradation o f mechanical
properties o f the posterior annulus tissue. This combination o f factors motivates the
examination o f the mechanical properties o f the tissue o f the posterior annulus and
their response to repetitive compression-flexion loading.
It is our objective to investigate the possibility that the degree o f crosslinlring
in the posterior annulus may be manipulated to positively modify the material
characteristics o f the tissue. Prior research on human articular tissues has
demonstrated that alterations in the degree o f crosslinking modify the macroscopic
material properties o f collagen. Furthermore, epidemiological data regarding the
incidence o f low back pain, when considered with the progression o f crosslink
formation in the intervertebral disc, may suggest that crosslinking is a stabilizing
factor, reducing the generation o f tissue damage. Genipin, a naturally occurring
plant extract, was selected as the crosslinking agent in this study, having been shown
to form stable crosslinks, but avoid the cytotoxicity and calcification problems o f
conventional synthetic chemical fixation.
ix
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In this study, pseudo-physiologic loading is simulated by compression-flexion
applied to individual calf spine motion segments. Non-destructive indentation
testing methods are utilized to measure elastic-plastic properties (hardness) and
viscoelastic properties (creep deformation and stress relaxation) of the posterior
annulus tissue, with the benefit o f allowing the testing o f intact motion segments at
multiple times during a prescribed loading regimen.
The pilot study results indicate that treatment o f the posterior annulus tissue
with genipin leads to a decrease in viscoelastic behavior, an increase in elastic-plastic
strength properties, and a potential resistance to the degradation o f material
properties due to pseudo-physiologic repetitive loading. These results strongly
suggest that the proposed use o f a genipin treatment to prevent material degradation
o f the intervertebral disc in vivo merits further research and study.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
are described in Chapter 5. Analysis o f the pilot study results, outstanding issues
and future directions are presented in Chapters 6 and 7.
The Burden of Lav Back Disorders
Low back pain is one o f the most prevalent medical problems impacting
western society. The large percentage o f the general population impacted and the
corresponding cost demand attention to better understand the root pathology in a
continuous attempt to improve treatment
The scope o f the problem in the United States is clearly seen in
epidemiological studies. In 1983 Frymoyer compiled a study o f 1221 men, age 18 to
55, who had visited a model family-practice facility in the U.S. from 1975 to 1978.
This was the only medical facility serving a community o f 8000, so that it may be
viewed as a good cross section representing a wide variety o f occupational
backgrounds. The results indicated that 46.3% had, or were having moderate low-
back pain, defined as mild, discomforting, or distressing. In addition, 23.6% had or
were having severe low-back pain, defined as horrible or excruciating. O f the 69.9%
who reported moderate to severe back pain, a significant number also exhibited
symptoms in the lower limbs, which may indicate nerve root involvement (Frymoyer
et aL 1983).
Epidemiological studies also highlight the burden o f low-back problems on
the medical care system. Between 4.7 % (according to medical records) and 9.4%
(according to population surveys) o f the US population seeks professional health
care annually (1990) for treatment associated with low-back problems (Waddell
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1996). Similarly, Waddell reports that between 8.2% and 12.5% o f the population in
the U.K. seeks medical cate annually (1993). The problem extends beyond
congested waiting rooms at local clinics. Intervertebral disc disorders led to over
400,000 hospitalizations in the U.S. annually in 1988 (Andersson 1995).
The direct economic cost o f this health care burden is immense. Frymoyer
estimated a direct health care cost o f $33.6 billion annually in the U.S. in 1994
(Frymoyer and Durett 1997). Andersson provides a somewhat lower estimate o f
$23.5 billion annually for the year 1990 (Andersson 1995). O f course, the problem is
not confined to the U.S. Direct costs o f $1 billion annually are reported for the U.K.
(Waddell 1996).
Indirect costs in the form o f reduced output due to the cessation or
reduction o f work activity are generally believed to be at least o f the same magnitude
as the direct costs. Andersson estimates 175 million lost workdays annually in the
U.S. for the period 1985 through 1988 (Andersson 1995).
Interestingly, results o f national surveys in Finland indicated that the
prevalence rates o f low-back symptoms remained constant from 1978 through 1992
(Leino, Berg and Puska 1994). This suggests, despite the efforts o f the spine health
care community, that the general public is not seeing an improvement in the
occurrence o f symptoms. There is definitely room to advance the understanding and
treatment o f low-back problems and cut the enormous social and economic costs to
the global community.
3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The Mechanical Structure ofthe Intervertebral Disc and Mechanisms o f Degradation
The human spine consists o f the 24 articulating vertebrae o f the cervical,
thoracic, and lumbar regions and the typically fused complex o f the sacrum and
coccyx. The vertebral column o f vertebrae joined by intervertebral discs, ligaments
and the associated muscle groups provides flexibility o f motion, protection of the
spinal cord, and distribution o f body forces. This paper will focus on the
articulations o f the lumbar region.
The intervertebral disc (TVD) connects the cartilaginous end plates o f two
adjacent vertebrae. The disc has two primary components, the annulus fibrosis
consisting o f concentric layers, or lamellae, o f tough fibrocartilage and the central,
gel-like nucleus pulposus. The annulus fibrosus consists o f Type I or II collagen
fibrils and type VI microfibrils that interact with proteoglycans to form a fiber-
reinforced matrix (Vang et aL 1994). The matrix fibers o f adjacent layers are oriented
obliquely, at 30 degrees to the horizontal in a criss-cross pattern (Miely et al. 1990).
Biomechanically, the nucleus serves to transmit compression and provide shock
dispersion, while the annulus fibrosis is principally loaded in circumferential tension
to contain the compressed, pressurized nucleus. Note that the orientation o f the
collagen fibers at 30% from the horizontal favors circumferential tensile loading
relative to longitudinal loads.
Also o f interest to this study, the posterior longitudinal ligament (PLL)
extends down the length o f the vertebral column on the posterior aspect o f the
vertebrae and IVD’s, facing the spinal canaL In die lumbar region, the PLL takes on
4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In addition to association with nuclear degeneration, tears of the peripheral
annulus may be linked with pain. Yoshizawa analyzed the IVD for nerve fibers in a
study of 26 degenerate discs removed during fusion procedures and 5 normal IYD’s
removed at routine necropsy. The anterior and posterior ligaments and the outer half
o f the annulus fibrosus are similarly supplied with a “profuse” axonal network, with
abundant free nerve terminals in both the degenerate and control specimens. The
morphology o f these free nerve terminals is o f the type that is generally accepted to
be a nociceptor. In addition, the nucleus and inner annulus were found to be free o f
axons in both the degenerate and control specimens. They found no evidence o f in
growth o f axonal networks into degenerate zones o f the nucleus. This implies that
the degenerate zones o f the nucleus and inner annulus cannot be a direct source o f
low-back pain. Importantly, this suggests that the nerve networks in the outer
annulus and PLL could be a source o f pain when placed in excessive tension or
compression (Yoshizawa et aL 1980). One may even hypothesize that the PLL,
which has been shown to carry very litde load, with a tensile strength o f only 90N
(Adams, Green and Dolan 1994), may function primarily as a sensor to detect
excessive strain or pressure.
The development o f damage to the posterior annulus, combined with the
presence o f pain receptors in the outer annulus leads the focus o f these studies to the
material properties o f the posterior annulus fibrosus.
6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
o f the motion segment anteriorly and caudally relative to the inferior vertebrae. This
motion applies compressive force to the anterior annulus and longitudinal tension in
the tissues o f the posterior annulus (Figure 1-2). The tensile forces indicated by this
simple model have been confirmed empirically in prior studies by Adams (Adams
and Hutton 1982) and Hedman (Hedman and Femie 1997).
Posterior Annulus
in Tension Ifttts..
Anterior
Annulus in
Compression
Figure 1-2: Flexion-com pression o f spine - annulus stresses. Sagittal view o f a simplified
schematic model o f spinal vertebrae in flcxioa-comprcsskm showing representative forces on the
annulus fibrosus tissue. Compression is represented by arrows directed toward the specimen, while
tension is represented by arrows directed away from the specimen. Note that the stress profile is an
approximate representation for illustrative purposes. The actual stress profile is non-linear and
dependent on disc structure and the ratio o f bending moment to compression load.
Research indicates that the bending moment is the more important
component o f flexion-compression. Relatively low compressive force can trigger
prolapse when combined with a large bending moment. Conversely, compression in
the absence o f a bending moment cannot damage the soft tissues prior to the
8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vertebrae (Adams, Green and Dolan 1994). The importance o f the bending
moment, and therefore posterior annulus tension, in the mechanism o f injury to the
IVD is another factor driving interest in the properties o f the posterior annulus in
tension.
As outlined above, the posterior annulus is the primary site o f injury to the
intervertebral disc and a likely source o f low-back pain. Meanwhile, epidemiological
studies implicate repetitive compression-flexion loading in the acceleration o f disc
degeneration and the onset o f low-back symptoms. Additionally, this loading
condition places the posterior annulus in high tension. This combination o f factors
motivates the examination of the mechanical properties o f the tissue o f the posterior
annulus and their response to repetitive compression-flexion loading.
9
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C hapter 2 - N on-D estructive T esting M ethods used to Determ ine
the M aterial Properties of Intervertebral D isc Annular Tissue
Destructive Testing Methods
Scientists interested in determining a material’ s mechanical properties
commonly evaluate the elastic-plastic behavior o f the material. Traditionally,
materials may be prepared in a strip and pulled in tension until fractured, measuring
the resulting force and displacement, in what is referred to as a tensile test.
This destructive type of experiment yields several meaningful indications o f
the material’ s mechanical properties. When placed in tension, an elastic material,
such as soft tissue, will initially undergo a period o f displacement that is linearly
proportional to the tensile stress applied. In this region, if the tensile load is released,
the specimen will return to its original shape, with no permanent deformation.
However, if sufficient load is applied, and the material is not brittle, the specimen
will enter a region of plastic deformation, where the displacement is no longer
linearly related to the applied force, and the specimen will exhibit permanent, plastic
deformation when the load is released. Above this point, the cross sectional area o f
the specimen continues to decrease, while increasingly smaller additions in stress are
required for plastic deformation. A point is reached where the load carried by the
specimen reaches a maximum point, referred to as the ultimate, or tensile, strength.
Beyond this point, the load supported by the specimen decreases while plastic
deformation continues to increase until the point o f fracture.
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
J________________________
S>juU ^ultim ate
Strain
Figure 2-1: Schem atic stress-fltrain curve. A schematic tensile test stress-sttain curve showing the
yield strength (S^tu), uldmate strength (S u k m w e), and die modulus o f resilience (Ur )-
Several meaningful measurements can be determined from a tensile test
experiment (Figure 2-1). The point at which the force-deformadon relationship
deviates from a straight line, usually at a 0.2% offset, is referred to as the yield point,
which is an indication o f the stress to which a material can be subjected prior to
permanently deforming. The ultimate strength is the highest normalized load that
the material will absorb prior to the onset o f failure. It is similarly possible to
calculate the energy absorbed by the material at each o f these points with the
following relationship.
Strain Energy = (j-d e (2.1)
Where: er — Stress
£ — Strain
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Note that this definition o f strain energy corresponds to the area under the
true stress-strain curve. The strain energy absorbed at the yield point is known as the
modulus o f resilience.
Non-Destructive Testing Methods
Destructive tensile testing is not the only way to determine elastic-plastic
properties. Indentation testing provides a non-destructive means o f determining
material properties. In addition, non-destructive techniques have the key advantage
o f being able to record data at several points along a well-defined repetitive loading
history.
Hardness
Indentation testing can be used to determine the elastic-plastic property
known as hardness. Hardness is simply defined is a material’ s resistance to
penetration. In metals, hardness measurements are based on plastic deformations,
typically the diameter o f the dent left in the material, which in turn is related to the
depth o f the permanent indentation. This plastic deformation occurs at a critical
shear stress that is a characteristic o f the material. This critical shear stress has been
shown empirically to be related to the yield stress, a fundamental measure o f strength
(Tabor 1996).
12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
r = Y 0e ” (2.2)
Where: Y = yield stress
Y0 — critical shear stress
e — logarithmic strain
n — empirical value ranging between 0 and V i
In this way, indentation hardness provides an indication o f yield stress. The
relationship presented above applies to metals and other crystalline materials that do
not work harden.
As indicated above, conventional hardness measurement techniques as
applied to metals are rooted in the measurement of plastic deformation in order to
find the related strength properties. Soft tissue testing is fundamentally different in that
displacement measurements are recorded during indentation. Therefore, we are
measuring elastic deformation in addition to plastic deformation. So it is necessary
to identify the unique characteristics o f the relationship between the force and
deformation measurements, and the material’s strength properties.
As hardness techniques have been applied to softer materials, specifically
polymers, more recent approaches to hardness testing address the distinction
between elastic hardness and plastic hardness. That is to say that as an indenter is
depressed into a material surface, one component o f the displacement is due to
elastic deformation, and a second component o f displacement is due to plastic
deformation. Remembering that conventional hardness measurements describe
permanent deformations (after the load has been removed), it is the measurement o f
plastic deformation that allows us to correlate the plastic hardness to yield strength.
13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Therefore, in our measurements o f hardness, it would be desirable to find the plastic
hardness independently.
Neideck, et aL have provided a method to distinguish the elastic and plastic
components for indentation testing on polymers. In their method, it is necessary to
initially calculate the real contact penetration depth, defined as the distance between
the indenter tip and the surface o f the sample, during any point o f an indentation test
(Figure 2-2).
hc = h - a ( F / S ) (2.3)
Where: hc — contact penetration depth
b — penetration depth o f the sphere
a — geometrical factor
F = applied load
S — stiffness o f the sample
Initial Surface
Surface Profile
Under Load
Figure 2-2: Indentation nom enclature. Schematic showing polymer indentation and the
nomenclature adopted by Neideck.
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Here, the sample stiffness, S, is found as follows.
„ 2
(2.4)
Where: — projected impression surface
E r — reduced elasticity modulus
The reduced elasticity modulus, E„ is defined as follows.
1 (l-v ,2) (l-vf)
— = ^-----—+ ------^ (2.5)
Er Es E,
Where: v , — Poisson ratio (specimen)
Et — elasticity modulus (specimen)
v . = Poisson ratio (indenter)
E, = elasticity modulus (indenter)
After calculating the real contact penetration depth, Neideck then introduces
the plastic hardness number.
H ^ ^ F I A ^ ^ F l b D h ' ) (2.6)
In this way it is possible to separate the elastic and plastic properties and use
the plastic hardness index as a relative indicator of yield strength when dealing with
polymers (Neideck, Franzel and Grau 1999).
It is critical to note that Neideck’s method does not account for viscoelastic
effects (see definition below), which are significant in biological soft tissue. To our
knowledge, a suitable model has not been developed. For this reason it is necessary
to adopt an alternate method for the current investigation.
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In this case it is reasonable to adopt a simplified hardness index,
H im k x = ^ in d tn ttr ! ^p tm tra tkm (2-7)
Where: R _,_ = applied indenter force
H; = penetration depth
By adopting this simplified hardness index it is understood that the hardness
measurement will incorporate elastic as well as plastic effects. Therefore the
recorded hardness will be directly related to a combination of two material
properties, the yield stress and the elastic modulus. In this way the hardness index
incorporates indications of material strength and elastic stiffness.
Viscotlasticity
The elastic-plastic properties described above serve as an adequate
description o f the mechanical properties o f conventional metals. These properties
are assumed to be time independent and remain constant in unvarying conditions, in
most applications. However, when dealing with soft materials, including biological
tissues, there is an additional factor o f viscoelasticity. Viscoelastic materials are
characterized by time and rate dependence o f material properties as evidenced by the
phenomena o f creep displacement and stress relaxation.
Creep deformation refers to an increase o f strain, due to a fixed stress, with
time. Similarly, stress relaxation is a reduction in the stress required to maintain a
fixed deformation over time. In metals, strain rate dependence is observable, but
negligible, at room temperature. Viscoelasticity in metals is only significant at
absolute temperatures nearing Vz the melting point o f the material (Crandall, Dahl
and Lardner 1978). In metals, viscoelastic behavior is due to “slip occurring along
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
crystallographic directions in the individual crystals, together with some flow o f the
grain-boundary material,” (MacGregor and Symonds 1978) a high energy process
seen only at elevated temperatures. In polymers, viscoelasticity is due to the
continuous rearrangement o f the secondary bonds between the long-chain
molecules, which is biased by an applied stress at room temperatures (Crandall, Dahl
and Lardner 1978).
Similar to polymers, biological soft tissues exhibit viscoelasticity under
ambient conditions, however the mechanisms are significandy different Perhaps the
best-understood mechanism is the biphasic viscoelastic behavior o f articular cartilage
in compression, which is shared by the nucleus and inner annulus o f the
intervertebral disc (Buckwalter et aL 2000). The biphasic mechanism results from a
balance o f flow dependent properties and the material properties o f the proteoglycan
(PG) matrix. As load is applied to articular cartilage, fluid is exuded from the
porous-permeable solid matrix o f the tissue. The drag force o f this interstitial fluid
flow and the associated fluid pressure are the primary mechanisms o f load support,
relative to compressive stress carried by the PG matrix. With continued time, fluid
exudation decreases and the compressive stress is carried by the proteoglycan matrix
at the point o f equilibrium (Mow, Proctor and Kelly 1989; Mow and Ratdiffe 1997).
Note, that the mechanism o f viscoelasticity in cartilaginous materials, as
outlined above, results in a pattern o f creep displacement that is unlike the creep
pattern observed for conventional materials. Conventional materials exhibit 3
distinct phases o f creep displacement (Figure 2-3). Primary creep is characterized by
a decreasing creep rate while the secondary creep stage is marked by a constant,
17
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
minimum creep rate. Tertiary creep represents a period o f increasing creep rate
followed by fracture.
< 2
< U
o<
V
a
u
creep rate = constant
Phase 3
Phase 2
Phase 1
Tim e
Figure 2-3: Creep deform ation curve — conventional m aterials. Creep deformation pattern for
conventional viscoelastic engineering materials. Note 3 distinct phases followed by fracture.
Alternately, biological soft tissues exhibit a creep pattern o f rapid
displacement foQowed by a continuous decrease in the rate o f creep, leading to an
equilibrium deformation rate, which may equal zero for low load levels (Figure 2-4).
The approach o f the equilibrium point marks the transition o f primary load support
from fluid flow effects to the proteoglycan matrix. Note that a significant time is
required to reach the equilibrium point. For tests performed on a plug o f
intervertebral disc in compression, the equilibrium point is reached after
approximately 104 seconds (Mow and Ratdiffe 1997). The predominantly type I
collagenous tissues o f the tendons and ligaments exhibit similar viscoelastic behavior
18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
in tension, resulting in a similar creep deformation pattern (Buckwalter et aL 2000).
The anticipated behavior o f the annulus tissue in the current study is discussed
below.
Equilibrium
Tim e
Figure 2-4: Creep deform ation curve — biological to ft tissues. Creep deformation pattern for
biological soft tissues. Note for tests performed on a plug o f intervertebral disc in confined
compression, the equilibrium point is reached after approximately 10* seconds (Mow and Ratdiffe
1997). Also note that the equilibrium rate o f deformation may be greater than zero for sufficiently
large loads.
Determination of tbe Viscoelastic Properties o f the Posterior Annnlsu via Indentation Testing
There are multiple methods to measure viscoelastidtic properties. Much o f
the prior work with cartilaginous soft tissues has focused on confined compression
experiments, compressing cartilage against a test fixture or native bone. Neither
technique can be applied to an intact FVD motion segment Consequently, as
highlighted earlier, die current study utilizes non-destructive indentation testing
19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
methods that allow us to test intact specimens at multiple times during a prescribed
loading regimen.
Indentation testing can be used to determine the viscoelastic properties o f
stress relaxation and creep deformation in two separate experiments. In a stress
relaxation experiment, the indenter is loaded on the surface of the tissue up to a
target load and subsequently held at that displacement for a fixed period o f time,
while recording the resulting decrease in stress, referred to as the stress relaxation.
Similarly, creep data is gathered by ramp loading an indenter to a target load. The
target load is then maintained for a fixed duration while recording the resulting
increase in displacement, referred to as creep deformation.
Indentation testing is commonly applied to bulk soft tissues o f the body
(Kawchuk and Elliott 1998; Pathak et aL 1998; Vannah and Childress 1996). In
these cases, the skin surface is indented in an attempt to characterize properties of
the sub-dermal tissue structures. This type o f test is typically carried out in vivo to
determine the properties o f gross tissue structures, such as muscle groups. For the
purpose o f this research, we are interested in applying indentation testing techniques
to specific soft tissues. While this application is not as widespread as bulk tissue
studies, other researchers have used this technique. As an example, Scapino used
direct indentation on the temporomandibular joint disc o f rabbits using 4.8mm
diameter spherical steel indenters to obtain stress relaxation data (Scapino et aL
1996). The growth o f indentation testing is also evidenced in an interest in modeling
indentation tests (Suh and Spilker 1994) and the shift to testing at increasingly
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
sm a lle r scales in micro-indentation (Costa and Yin 1999). However indentation
testing has not been applied to the connective tissues o f the spine.
It is important to note that indentation o f the posterior annulus involves
mixed mechanisms o f indentation resistance. One component o f resistance is the
biphasic compression mechanism described above, while another component is
tensile resistance o f the predominantly type I collagen annulus fibers attached to the
vertebral end plates. Recall that nucleus and inner annulus tissue in compression,
and type I collagenous tissue in tension, exhibit similar viscoelastic behavior. It is
therefore reasonable to expect the posterior annulus to exhibit an indentation creep
pattern similar to Figure 2-4.
This expectation is supported by a long-duration indentation creep
experiment carried out in our laboratory. In this case, a motion segment was
prepared as described in Chapter 5, and indented on the posterior annulus with a 10
N load for 15,000 seconds. The resulting creep deformation followed the expected
pattern, reaching an equilibrium deformation rate at approximately 3000 seconds
(Figure 2-5).
21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1.4
1.2 - ^ - 12
10
- 6
- 4
0.2
5000 10000 15000
Tim e (s)
D isplacem ent Force
Figure 2-5: L ong duration creep teat results. Note that creep deformation reaches an equilibrium
rate at approximately 3000 seconds.
Tens i l e vs . C o mp r e s s i o n Pr o p e r t i e s
By indentation testing the soft tissue o f the posterior annulus, we are
attempting to look for changes in the material properties of the tissue. More
specifically, we are looking for changes in the material properties that affect the
performance o f the tissue in tension, since tensile stress in the posterior annulus has
been implicated in the onset o f low back pain and degeneration.
As mentioned above, annulus tissue subjected to an indentation hardness test
resists indentation through tensile stress provided by the annulus fibers attached to
the vertebral end plates, and through resistance to compression o f the annulus tissue.
A study was designed in our laboratory to determine the relative contribution o f
these two mechanisms
22
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Two groups of specimens were indentation tested. The first test group
consisted o f intact discs potted in urethane. Specimens from this test group provide
tensile and compressive resistance to indentation. The second test group consisted
o f discs that had been removed from the vertebra, and lightly compressed between
two plates during indentation testing. The discs are not attached to the plates, and
therefore cannot provide tensile resistance. However, the plates do constrain the
annulus and nucleus tissue so that the specimen does provide compressive resistance
to indentation.
Specimens were indentation tested at the center o f the posterior annulus with
a ramp displacement to 0.4 mm, while recording the resulting indenter load. The
force required for the intact specimens represents the tensile plus the compressive
components o f resistance. While the force required for the dissected discs
represents compressive resistance only. In this way, it is possible to approximately
determine the relative contributions o f tensile and compressive indentation
resistance.
The results indicate that the intact specimen provides significantly greater
resistance to indentation. Intact specimens developed an average of 6.43 N o f
resistance (st dev. 1.52 N, n=12) to 4 mm indentation, relative to 0.45 N o f
resistance (st. dev. 0.21 N, n=6) for dissected specimens. This 93% decrease for
dissected specimens was found to be statistically significant by a Mann Whitney non-
parametric rank sum test (p=0.001).
These results suggest that the contribution o f tensile resistance is relatively
greater than compressive resistance in the case o f indentation testing of the posterior
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
annulus. It is therefore reasonable to suggest that changes in indentation
measurements are primarily indicative o f changes in the tensile properties o f the
annulus tissue.
Re f i nement o f I ndent at i on Tes t i ng Me t hods
It has been shown above that indentation testing can be used to determine
both elastic-plastic and viscoelastic material properties. However, it is important to
note that elastic and viscoelastic properties are inter-related, and measurement
techniques beneficial for one can negatively impact the other. Therefore, prior to
further use, a series o f experiments was carried out in order to determine a
measurement protocol that minimized the overlapping effects o f viscoelastic and
elastic properties.
For each o f these experiments, calf spine lumbar motion segments were
separated by cutting adjacent vertebrae midway in the transverse plane. Muscle,
tendon, and miscellaneous soft tissue were removed along with the vertebral arch, in
order to provide access to the posterior annulus. The posterior longitudinal
ligament, which is integrated with the posterior annulus, was left intact. A complete
description o f the indentation testing methods is presented with the pilot study
design (reference Chapter 5).
Re pe at e d Vi scoel as t i c Me as ur e me nt Tes t i ng
One test was designed to determine the impact o f repeated indentation at a
fixed tissue site on recorded viscoelastic measurements. Specimens were indentation
24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
tested at the mid-point o f the posterior disc, recording stress relaxation and creep
displacement. The indentation test was repeated 3 times at the same location, at
five-minute intervals, without moving the specimen. The five-minute interval was
intended to allow tissue recovery while minimizing the effects o f ambient exposure.
R epeated V iscoelastic Property M easurem ent Results
■Stress Relaxation ■Creep Deformation
Figure 2-6: R epeated viscoelastic m easurem ent teat results. Results o f testing to show the
effects o f repeated indentation testing at the same test location on measured viscoelastic properties.
Error bats indicate the standard deviation.
The test results indicate a decrease in measured viscoelastic properdess with
repeated indentation (Figure 2-6). Creep measurements exhibited a 37.5% decrease
(Friedman statistic, p=0.005), with stress relaxation measurements decreasing by
11.3% (Friedman statistic, p=0.0169). This decrease in measured viscoelastic
properties must be accounted for in the design o f an indentation test
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Re pe at e d El ast i c- Pl ast i c Me as ur e me nt Tes t i ng
An additional experiment was designed to determine the impact o f repeated
indentation at a fixed tissue site on recorded elastic-plastic measurements. In this
case, a 2.5 mm diameter probe was repeatedly indented (15 cycles) into the center o f
the posterior annulus to a uniform depth o f 1mm, recording the resulting force. It
was anticipated that the load achieved at 1mm displacement would decrease with
repeated indentation and approach an asymptotic value, due to the gradual reduction
in viscoelastic effects with repeated indentation.
R epeated Elastic-Plastic Property M easurem ent Typical R esult
- 12
I
M
25 - 10
20
15 - 6
10 - 4
- 2
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Cycle
■ Hardness Index ♦ % Reduction (relative to prior cycle)
Figure 2-7: R epeated elastic-plastic m easurem ent test results. Typical result o f testing to show
the effects o f repeated indentation testing at the same test location on measured elastic-plastic
properties.
The results o f this experiment indicate that the recorded hardness generally
decreases with the number o f ramp loading cycles (Figure 2-7). The decrease is
asymptotic, approaching a plateau at around 10 cycles. In all test cases, the percent
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The results signal that for measured creep deformation, the mean at the pre-
repeated-indentadon site is equal to the mean at the post-repeated-loading site
(Figure 2-8). A Wilcoxan non-paramettic paired test yielded a p-value o f 0.5, which
lends support to the null hypothesis that the pre and post-indentation cycling data
creep averages are the same.
Likewise, for stress relaxation data, the mean at the pre-repeated-loading site
appears to be equal to the mean at the post-repeated-loading site. In this case, the
Wilcoxan test indicated that there is not a significant difference in the means with a
p-value o f 0.686, which lends support to the null hypothesis that the pre and post
indentation cycling stress relaxation averages are the same.
The significance o f these results is limited by the sample size o f 5. For both
creep and stress relaxation, the high variance relative to the desired detectable
difference in means (5% o f the mean) yields a statistical power less than 0.55. In
other words, the results lend support to the conclusion that the pre and post
indentation cycling creep and stress relaxation averages are the same, with a greater
than 45% probability o f error.
Also note that the absolute values o f stress relaxation are lower in this case
than the results presented in Figure 2-6. This is due to the fact that the size o f the
indenter was decreased, and the magnitude o f the applied force was reduced
accordingly, in order to generate equivalent stress. Therefore, for equivalent stress,
the relative changes in stress relaxation are comparable.
28
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
'Re gi on al Vari at i on Tes t i ng
One additional round o f testing was designed to determine if there is a
central region o f the posterior annulus where differences in repeated viscoelastic
property data due to varying test location are small or insignificant. For 13
specimens, Creep and Stress Relaxation measurements were recorded at the center of
the disc, then at four additional test locations, each incrementally shifted 3mm
laterally.
Regional V ariation T est R esults
33 0.7
0.6
2.5 0.5
0.4
0.3
0mm 3mm 6mm 9mm 12mm
a
a.
Indentation T est Position (L ateral Shift Relative to C enter o f the
Posterior A nnulus)
■Stress Relaxation 1 # Creep Deformation
Figure 2-9: Regional variation test results. Complete results o f test to determine the regional
variation o f viscoelastic measurements on the face o f the posterior annulus. Error bars represent the
standard deviation.
The recorded viscoelastic data remained relatively constant within the central
region o f the posterior annulus (Figure 2-9). A comparison o f the data recorded at
the center and lateral measurements suggests that there is no significant difference in
means for creep and stress relaxation measurements within 3 mm o f the center o f
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the posterior annulus (Table 2-1). Outside o f this region, the measurements are seen
to diverge from the values measured in the central region. Importandy, these results
imply that it is possible to collect repeatable viscoelastic measurements from the
region within 3 mm of the center of the posterior annulus.
Test Comparison z-scorc Two-tailed p value
Stress Relaxation Center vs. 3mm Lateral Shift 0.784 0.433
Center vs. 6mm Lateral Shift 2.353 0.019
Center vs. 9mm Lateral Shift 0.784 0.433
Center vs. 12mm Lateral Shift 1.02 0308
Creep Center vs. 3mm Lateral Shift 1.098 0.272
Center vs. 6mm Lateral Shift 0392 0.695
Center vs. 9mm Lateral Shift 2.746 0.006
Center vs. 12mm Lateral Shift 2.746 0.006
T able 2 - 1 : R egional variation test statistics. Mann-Whitney non-parametric rank sum test results
comparing data recorded at the center o f the posterior annulus with lateral measurements. Note that
the calculated p values suggest that stress relaxation measurements are uniform within 3 mm o f the
center, and creep measurements are uniform within 6 mm.
Regional V ariation T est R esults - Im proved Technique
0.7
- 0.6
- 0.5
' 0 4 I
-- 0 3 J
c 7 > 0.5 - 0.1
Omm
3mm 9mm
Indentation T est Position (L ateral Shift Relative to Center o f the
Posterior Annulus)
Stress Relaxation — • — Creep Deformation
F igure 2 - 1 0 : R egional variation test results — im proved technique. Subset o f results o f regional
variation testing including data obtained with an improved force feedback signaL N ote the
significantly reduced standard deviations relative to Figure 2-6. Error bars represent the standard
deviation.
30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Note that a subset o f 5 out o f 12 total specimens in the regional variation
testing was tested with an improved technique. In these cases, the MTS controller
was supplied with a force feedback signal from an external load cell, rather than the
internal MTS load cell. The external load cell provides a significandy less noisy
feedback signal, improving controller performance. This modification improved the
testing repeatability, particularly for the stress relaxation test, as evidenced by the
consistendy reduced standard deviations (Figure 2-10 vs. Figure 2-9).
This subset o f results leads to two conclusions. The reduction in variance
indicates that this control scheme provides increased repeatability, and therefore was
adopted for all subsequent testing. In addition, these results reinforce our
conclusions regarding regional variation on the face o f the posterior annulus, with
greater precision.
Summar y o f Ref i nement o f I ndent at i on Tes t i ng Met hods
The series of tests presented above suggest three general principles for an
effective, repeated non-destructive indentation testing protocol First, repetitive
indentation loading should precede hardness measurements in order to avoid
undesired effects o f the c h a n g in g viscoelastic properties, specifically lack o f
repeatability, on recorded hardness. Second, the decrease o f viscoelastic effects with
repeated cycling indicates that viscoelastic data should be collected from previously
un-tested tissue. Fortunately, the lateral effects testing indicates that viscoelastic
measurements may be made at neighboring locations on the face o f the posterior
annulus and avoid the undesired effects o f repeated indentation at the center o f the
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 3 - T he Response of the Annulus Fibrosis to Repetitive
Loading
Pr i or R e s e a r c h o n Fat i gue Loadi ng o f t he Spi ne
Other researchers have examined loading o f spine segments and
intervertebral discs, but with some important differences in techniques and
objectives.
Prior studies have focused on extreme loading conditions with the objective
o f qualitatively assessing failure modes, rather than quantifying changes in material
properties. Hansson cyclically loaded human lumbar spine segments in axial
compression to the extent that the vertebrae fractured under load. He was then able
to quantify the fracture modes, and correlate the failure mode with the degree o f disc
degeneration, age, and bone mineral content (Hansson, Keller and Spengler 1987).
In another experiment, Lotz loaded intervertebral discs o f a mouse tail in v i v o in an
attempt to study changes in the biological activity within the IVD and the
biomechanical performance o f the motion segments. It is important to note that
these discs were loaded for one week at magnitudes up to 10 times the animal’s body
weight (Lotz et al. 1998). Similarly, extreme loading conditions have been used by
Adams to simulate “a days heavy labour” (Adams and Hutton 1983). In this case
spine segments were loaded in flexion-compression to 1500-6000 N (337-1349 lbf),
depending on the physical characteristics o f the cadaver, at a frequency o f 0.67 Hz
for four hours. The tested discs were examined to determine the degree o f disc
degeneration and the failure mode o f disc fracture or disc prolapse, if applicable.
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
These three examples are indicative o f the tendency for prior research to focus on
qualitative evaluations o f failure modes due to extreme loading conditions.
In contrast, the current study is focused on quantitatively evaluating the
material properties o f the IVD tissue in response to sub-traumatic, pseudo-
physiologic repetitive loading.
St udy t o De t e r mi ne t he De gr ad at i o n o f M at eri al Pr ope r t i e s oft he Pos t er i or Annul us d u e t o P s e u d o-
Pby s i ol ogc Fat i gue Loadi ng
An earlier study in this laboratory has established a correlation between
repetitive flexion-compression loading and the degradation o f the material properties
o f the posterior annulus (Gray and Hedman 2001). This study was designed to
utilize both conventional destructive testing in addition to non-destructive
indentation testing to record material properties.
De s t r uc t i v e Tes t i ng Pr o t o c o l
The destructive test protocol utilized in this study was modeled after a well-
controlled study by Hoshaw (Hoshaw et aL 1997), who measured a decrease in
canine proximal femoral strength and stiffness due to fatigue damage. In Hoshaw’ s
work, the control femur o f a matched pair was loaded to failure, while recording the
ultimate load. The contralateral femur was then cyclically loaded to 50% o f the
control’s ultimate load. After 3600 cycles the contralateral femur was loaded to
failure. A comparison was then made between the control and post-fatigue
mechanical properties.
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A test paradigm similar to Hoshaw’s was employed in this study to
determine the effects o f repetitive, non-traumatic loading on the mechanical
properties o f posterior disc dssue. Destructive testing was carried out on human
cadaveric specimens as follows. Lumbar motion segments were separated by
cutting adjacent vertebrae in the transverse plane. Muscle, tendon, and
miscellaneous soft tissue were removed, however, the posterior longitudinal
ligament (PLL), which is integrated with the posterior annulus, was left intact. The
spinous and transverse processes were removed by cutting the pedicles in the
frontal plane. Each motion segment was cut in the sagittal plane three times to
divide the segment into four test specimens o f approximately equal width. The
mid-sagittal cut approximately divided the PLL at the midpoint. Therefore, each
motion segment yielded two matched pairs o f test specimens, medial and lateral.
For each specimen, the anterior annulus and nucleus pulposus were
removed. To ensure that the specimen failed at the disc tissue, rather than the
adjacent bone, the remaining posterior annulus was significandy necked anteriorly,
and necked to a lesser degree in the medial-lateral profile. The cross sectional areas
ranged from 24.5 mm2 to 84.3 mm2 , (avg. 52.7 mm2 , std. dev. 19.0 mm2 ) The
superior and inferior bone segments were potted in urethane with 1mm wire drilled
through the cortical bone, to ensure fixation during testing.
Matched specimens were tested in pairs. From each pair, one randomly
selected specimen was designated as a control. This specimen was allowed to stand
at room temperature for 8000 seconds, to match the exposure to the ambient
conditions experienced by the fatigue specimen during cycling. The specimen was
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Following initial indentation testing, the specimen was fatigue loaded at 200
N fot 2000 cycles at 0.25 Hz. The load was applied perpendicularly to the
transverse plane, 40 mm. anterior to the mid-point o f the specimen in the transverse
plane. The specimen was then indentation tested following fatigue cycling, as
described above.
This procedure was followed for a total o f three fatigue loading cycles.
During all testing, the specimens were wrapped in saline wetted gauze.
Resul t s
Destructive testing indicated a reduction in linear elastic properties following
2000 cycles o f tensile fatigue (Figure 3-1). Yield Stress decreased 28% (Wilcoxon
statistics: 2=2.49, p=0.013), while the strain energy absorbed prior to yield, or
modulus o f resilience, decreased 42% (Wilcoxon statistics: z=2.50, p=0.013). In
this case, a Wilcoxon non-parametric paired test was utilized to confirm the effect
due to fatigue.
37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
700
600
-- 500
- 400 4
300 (A
200
100
Pre-Fatigue Post 2000 Cycles Tensile Fatigue
i ' ♦ Yield Stress ■ O Energy Absorbed Prior to Yield i i
Figure 3 - 1 ; Fatigue loading o f posterior annulus - tensile test results. Destructive tensile
testing results, showing a decrease in linear elastic properties due to 2000 cycles of repetitive tensile
loading. Yield stress decreased from 3.408 MPa to 2.445 MPa (Wilcoxon p=0.0l3) and the energy
absorbed prior to yield decreased from 472 kj/m 3 to 274 kj/m 3 (Wilcoxon p=0.013).
Similarly, non-destructive testing indicated a change in viscoelastic and
elastic-plastic properties following fatigue flexion (Figure 3-2). Specimens exhibited
a 25% increase in creep following 6000 cycles o f fatigue loading (Friedman: Chi-
square =9.1 5, p=0.027; Spearman: p=0.0067). Similarly, specimens exhibited a 25%
increase in stress relaxation following 6000 cycles o f fatigue loading (Friedman: Chi-
square=15.85, p— 0.0012; Spearman: p=0.00l4). A 22% decrease in calculated
hardness was observed, but was not found to be statistically significant (Friedman:
Chi-squarc=1.95, p=0.58; Spearman: p=0.20). T n this case, Friedman non-
parametric repeated measures analysis was utilized to confirm the effect due to
fatigue and Spearman’ s non-parametric correlation analysis was used to confirm the
38
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
correlation between the number o f fatigue cycles and the change in material
properties.
140 0.7
120 0.6
100 -
80 0.4
0.2
20 0.1
| 0 2000 4000 6000
i
N um ber o f Fatigue Cycles
i
| 1 A Stress Relaxation • Creep j |
Figure 3-2: Fatigue loading o f posterior annulus — indentation test results. Non destructive
indentation testing results, showing the response of viscoelastic properties at 2000 cycle intervals of
repetitive compression/flexion loading. Creep displacement increased from 0.274 mm to 0.344 mm
(Friedman p=0.027,Spearman p=0.0067) and stress relaxation increased from 89.1 N to III J N
(Friedman p=0.0012, Spearman p=0.0014) following 6000 fatigue cycles.
Viewed together, the destructive and non-destructive results indicate a
general trend. The elastic-plastic properties o f the annulus tissue decrease with
repetitive stress, while viscoelastic properties show a corresponding increase (Figure
3-3).
39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
130%
Non-destructive
a. .7|Xest_Rcsultj,.. _.
■9 I Viscoelastic
Properties........
120% -
110% -
Non-destrucdve
100% -
90% -
Properties
80% -
Destructive Test
Results, Elastic-—
Plastic Properties
70% -
60% -
50%
Cycles
2000
• Stress Relaxation
•Yield Stress
•Creep
• Energy Absorbed Prior to Yield
•Hardness
• Ultimate Stress
Figure 3-3: Fatigue loading of posterior annulus — com bined results. Combined results of non
destructive and destructive testing following 2000 cycles of fatigue loading showing the general
increase in the viscoelastic properties creep (Friedman p=0.034) and stress relaxation (Friedman
p=0.008), and the elastic-plastic properties hardness (Not statistically significant), yield stress
(Wilcoxon p=0.013), energy absorbed prior to yield (Wilcoxon p=0.0l3), and ultimate stress
(Wilcoxon p=0.074). Data shown is normalized relative to the pre-fatigue measurement.
Degradation of viscoelastic properties was observed to plateau after 2000 to
4000 fatigue cycles. The greatest increase in creep and stress relaxation followed
initial fatigue cycling, with this increase accounting for 65% and 80% of the overall
total increase for creep and stress relaxation, respectively. Note that fatigue cycling
was controlled as a function o f force, rather than deformation, so that plastic
deformation would not explain a reduced effect o f later fatigue cycles. Also note
that control specimens subjected to indentation testing at time intervals
corresponding to sets of 2000 fatigue cycles showed no trends in creep, stress
relaxation, or hardness (Friedman’s non-parametric repeated measures analysis
p=0.241, 0.392, 0.896; Spearman’s non-parametric correlation analysis p=0.172,
40
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
viscoelastic and elastic-plastic data from indentation tests may be viewed as an
indication o f conventional measures o f material properties. In addition, non
destructive materials testing techniques provide the important benefit of determining
elastic-plastic and viscoelastic material properties at multiple points along a
prescribed time or loading history.
Second, the material properties o f the posterior annulus tissue degrade due to
repetitive sub-traumatic loading. It is reasonable to hypothesize that the degradation
o f tissue properties over time can leave the posterior annulus vulnerable to injury or
prolapse due to progressively less traumatic loading conditions. Quantification of
the degradation due to fatigue loading is used as a principle measurement o f the
effectiveness o f biochemical treatment in the present study.
42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C hapter 4 - T he Im pact of Crosslinking on the Collagenous Tissue
of the Annulus Fibrosis
The St ruct ure o f Co l l a g e t t
Collagen, while present in a multitude o f connective tissues o f the body in
more than a dozen distinct principle types, has some fundamental common
characteristics. The simplified description that follows is applicable to Type I and
Type II collagens, the primary structural protein fibers o f the intervertebral disc.
A collagen molecule consists o f three polypeptide chains arranged in a triple
helix. This complex arrangement is defined by four levels o f structural order. The
primary structure designates the sequence of amino adds composing each
polypeptide chain, with glycine accounting for 1/3 of the residues and proline and
hydroxyproline composing V * of the residues, The second order structure refers to
the configuration o f the individual polypeptide chain resulting from the
stereochemical angle and bonding characteristics o f the peptide residues. Third
order structure denotes the joining o f polypeptide chains to form the characteristic
triple helix. Joining o f the polypeptide chains is achieved by crosslinks formed as
condensation products o f a reaction involving lysine and hydroxylysine residues.
Fourth order structure describes the arrangement o f these helical collagen molecules
in a staggered, quasi-hexagonal lattice forming a repeating supermolecular unit
structure, the tnicrofibnl (Yannas 1996). Therefore, both intramolecular crosslinks
between the polypeptide chains (third order structure) and intermolecular crosslinks
43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
between the molecular helices (fourth order structure) are important for structural
integrity.
C o l l a g e n Crossl i nks and M at eri al Pr o p e r t i e s
It is has been shown that biochemical crosslinking impacts the macroscopic
material properties of collagen tissue. Chen et al. studied patellofemoral cartilage
from bovine knees to invesdgate the biomechanical effects o f increased crosslinking.
Their group o f 175 samples was divided into 3 treatment groups: normal controls,
ribose treated to induce AGE formation, and a sham treatment. They were able to
successfully increase the degree of crosslinking in the ribose treatment group and
found, via biomechanical tensile testing, a corresponding increase in ultimate
strength and tensile stiffness, and a decrease in the elongadon at the point of ultimate
strength (Chen et al. 2001).
In separate studies, Sung et al investigated the biomechanical impact o f
fixation, or introduction o f crosslinks, on bovine and porcine heart valve tissue,
respectively. Utilizing tensile test methods they determined that fixation significantly
impacted material properties such as stress relaxation, ultimate tensile strength,
elastic modulus, strain at fracture, and fracture toughness, but these effects were
generally dependent upon the orientation and pressure o f fixation of the tissue
specimen. Pericardial tissue is anisotropic with the collagen fibers arranged with a
preferential orientation. Sung hypothesized that intramolecular crosslinks within the
collagen fiber impact tensile properties in the preferred orientation, while
intennolecular cro sslinks join adjacent fibers and impact tensile properties
44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
perpendicular to the preferred orientation. Therefore, different crosslinking agents
had varying effects, dependent on direction. In this case, the treatment pressure
plays a role due to the treatment techniques. The pericardial tissue was treated by
filling the intact aorta of the donor specimen with crosslinking reagent. Therefore,
pressurized treatment places the collagen fibers in tension, modifying the fiber
geometry during crosslink formation (Sung et al. 1999a, 1999b).
The use o f crosslinking agents already reaches beyond research into clinical
application. Fixation techniques have long been applied to pericardial bioprosthcses
in order to increase resistance to enzymatic degradation and reduce antigenicity
(Nimni et al. 1988). Glutaraldehyde has been the principle crosslinking agent used in
the chemical modification of biological tissues (Valente et al. 1992), however
formaldehyde, dialdehyde starch, and epoxy compound have also been used (Sung et
aL 1999a). Each o f these crosslinking agents is synthetic in nature.
Co l l a g e n Mat urat i on and Agi ng
In the last 12 years, researchers have arrived at a greater understanding of
crosslinking activity in collagenous tissue. In development, immature collagen
undergoes enzyme-mediated crosslinlting resulting in the formation o f reducible
crosslinks. These reducible crosslinks arc enzymatically converted to irreducible,
stable crosslinks in the first ten years o f maturation. [Confirm 10 years] In the
intervertebral disc, articular cartilage, and mature bone, lysyl oxidase-mediated
pyridinoline intermolecular crosslinks predominate (Pokhama and Phillips 1998).
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Once mature, the number o f irreducible crosslinks remains approximately constant
with age.
With age, a separate mechanism, nonenzymatic glycosylation, generates
additional crosslinks in the collagen matrix. These crosslinks are formed by the
nonenzymatic modification o f proteins by reducing sugars. Collagen molecules in
the proteoglycan matrix o f mature articular cartilage (> 20 years o f age) have an
“exceptionally long life-dme (> 200 years) making them susceptible to non
enzymatic glycation (NEG)” (Bank ct al. 1998). Nonenzymatic glycosylation
generates many advanced glycation end-products in collagen, but only pcntosidinc
has been shown to be a crosslink (Duance et al. 1998). Pentosidine is a pentose-
mediated crosslink between lysine and arginine, and is used as an advanced
glycosylation end product (AGE) marker to measure the degree o f crosslinking
(Pokhama and Phillips 1998). In addition, Pokhama has found evidence that
nonenzymatic glycosylation can result in AGE crosslinks to aggrecan proteoglycans
within the aggrecan matrix (Pokhama and Pottenger 2000), potentially further
impacting the material properties o f the cartilage.
The progression outlined above has been confirmed in human collagen by
multiple researchers. In 1991, Uchiyama found that levels o f pyridinoline in human
articular cartilage were constant in mature tissue and not correlated with age. His
team also found that the amount of pentosidine per collagen increased linearly with
age (Uchiyama et al. 1991). This work by Uchiyama expanded on the
groundbreaking work by Sell and Monneir, which initially identified the pentosidine
crosslink and implied the connection to aging (Sell and Monnier 1989). Bank, Chen
46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and Takahashi have also confirmed the overall crosslinldng progression (Bank et al.
1998; Chen et al. 2001; Takahashi et al. 1994).
Recently, researchers have turned their attention to crosslinldng in the
intervertebral disc. In separate studies Duance and Pokhama have investigated
various aspects o f crosslinldng in IVD tissue and their findings have been in general
agreement with the expected crosslinldng progression. In each case, pentosidine
levels were found to increase with age (Duance et al. 1998; Pokhama and Phillips
1998) (Figure 4-1). In addition, Pokhama found a slight decrease in pyridinoline
levels with age. Collagen of the IVD is particularly susceptible to crosslinldng via
nonenzytnic glycosylation because the large, avascular volume leads to low oxygen
tension that favors these reactions (Duance et al. 1998; Hormel and Eyre 1991).
2 0 -4 0 40-50 50-80
AGE (YEARS)
Figure 4 - 1 : Age related changes in crosslink concentration w ithin th e intervertebral disc.
Age-related changes in the content o f pyridinoline and pentosidine in lumbar intervertebral discs.
Note the significant increase in pentosidine crosslinks with age. Error bars indicate standard error of
the mean. (From: Pokhama and Phillips 1998. Reprinted with permission.)
47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I mpl i cat i ons
These findings raise interesting suggestions when combined with
epidemiological data on the onset o f low back pain. In order to clearly understand
the epidemiology, it is necessary to make a clear distinction between the concepts o f
prevalence and incidence. P r e v a l e n c e refers to the number o f subjects with a given
characteristic during a specified time period, whereas, i n c i d e n c e is limited to the
number o f subjects who d e v e l o p a given characteristic during a specified time period.
With these definitions in mind, we can review the epidemiological data
available regarding the occurrence o f back pain. This data has been conveniently
summarized in a review article by Andersson. The prevalence o f back pain as a
function o f age indicates a consistent pattern. The prevalence steadily increases with
age from the mid-twenties to the mid-fifties, where the rate o f prevalence peaks.
The prevalence then begins to decline with age. This pattern is consistent for
prevalence reported for time periods specified as cumulative lifetime, one year, and
point prevalence (at the time o f the survey) (Andersson 1995).
Alternately, back pain i n c i d e n c e data reveals a different pattern. Data indicates
that the incidence rate for males dramatically increases to a peak in the early twenties
(Figure 4-2). This peak is then followed by a steady decline with advancing age. For
females, the rate o f change o f incidence is more gradual, reaching a peak in the early
thirties, followed by a steady decline with age (Andersson 1995).
48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
w
V
JC
ts
*
Moles Females
o
o
1.5-
w
tl
a.
E
2
0
1 1.0-
O
5
< r
UJ
« .5-
ui
o
u
z
Under 20*24 25*29 30*94 35-44 45*54 5 5 -6 4 65 +
A G E GROUPS
Figure 4-2: Age related changes in low back pain incidence. Incidence o f significant low back
pain as determined from worker’ s compensation claims. (From: Klein, Jensen and Sanderson 1984.
Reprinted by permission.)
Therefore, epidemiology suggests that that the first occurrence o f significant
low-back pain typically occurs in the 20’ s to 30’s, pri or to the significant increase in
crosslinking due to the age related accumulation o f pentosidine (Figure 4-1 and 4-2).
The rate o f back pain incidence then steadily decreases as pentosidine crosslinks
accumulate. It has been suggested by Hedman that young discs, prior to the increase
in AGE crosslink content, are perhaps more susceptible to damage, leading to the
49
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
onset o f back pain. In addition, the age related increase in crosslinking might
stabili2e the disc tissue, and therefore reduce the generation o f new dssue damage
and the corresponding incidence of back pain. This progression of events suggests
that the increased degree of crosslinking with age may actually be a stabilizing,
beneficial factor (Hedman 2000).
An additional result o f the study by Duance also lends support to the
concept o f beneficial crosslinks. This study indicates that the degree o f crosslinking
increased with the degree o f disc degeneration for low to moderate degeneration.
However, the degree o f crosslinking was found to be significantly l owe r for severely
degenerated discs (Duance et al. 1998). Once again, Hedman suggests that it is
possible that the relatively low level o f crosslinking may contribute to make the tissue
more vulnerable to mechanical damage (Hedman 2000).
It is important to note that similar research by Pokhama did not yield the
same results. Pokhama found a decrease in pyridinoline and an increase in
pentosidine corresponding with increasing severity o f disc degeneration. However,
Pokhama also notes that severe degeneration tended to occur in older specimens, so
that the effects o f aging and disc degeneration could not be separated (Pokhama and
Phillips 1998).
Duance suggests that the reduction in pentosidine crosslinks in severely
degenerated discs observed in his study may be seen as an indicator o f increased
collagen regeneration (Duance et aL 1998). Recall that it is the “exceptionally long
lifetime” o f collagen molecules in the normal, mature extracellular matrix that makes
them susceptible to AGE. crosslinldng (Bank et al. 1998). Increased regeneration,
50
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
either in immature dssue or abnormal mature dssue, does not allow significant
accumulation o f AGE crosslinks. The view that severely degenerated discs may be
undergoing increased collagen regeneration is also supported by an observed
presence o f reducible crosslinks (Duance et al. 1998). However, Duance also found
reducible crosslinks to be present for all levels o f disc degeneration, and the quantity
did not increase for severely degenerated discs. Therefore Duance’ s assertion should
be considered unresolved.
It is important to note that the view o f crosslinldng as beneficial is novel, and
contradicts the current general opinion that crosslinldng is detrimental. Bank
hypothesizes that “a stiffer and more crosslinked collagen network may become
more brittle and more prone to fatigue,” and therefore subject to damage and
degeneration (Bank et al. 1998). Multiple researchers have offered similar
hypotheses (Chen et al. 2001; Duance et al. 1998; Hormcl and Eyre 1991; Pokhama
and Phillips 1998; Yang et a t 1994). The association with crosslinking and
degenerative disc disease is also supported by Takahashi, who found increased
pentosidine content in articular cartilage in the case o f bone and joint disorders
(Takahashi et aL 1994).
It is equally important to note that this general opinion is currently a
hypothesis, and that a cause and effect relationship between crosslinking and tissue
degeneration has not been established.
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Summary o f Crosslink Data
In summary, multiple factors are driving the study of the biomechanical
impact of biochemically inducing crosslinking in the posterior annulus dssue.
Crosslinks are vital to the structure of collagen. Prior research on human articular
tissues has demonstrated that alterations in the degree of crosshnking modify the
macroscopic material properties of collagen.
Furthermore, epidemiological data regarding the incidence of low back pain,
when considered with the progression of crosslink formation in the IVD, may
suggest that crossErtiting is a stabilizing factor, reducing the generation of tissue
damage.
The use of crosslinking agents in a biomedical appEcation is not
unprecedented. Crosslinking is commonly manipulated in the field of bioprostheses,
particularly for bovine and porcine pericardial tissues used as implants. While in
these cases the intent is somewhat different, resistance to enzymatic degradation and
reduced antigenicity rather than manipulation of tissue strength and durabiEty, the
precedent has clearly been set for the manipulation of collagen crosslinks in order to
modify the macroscopic tissue properties.
It is therefore reasonable to investigate the possihiEty that the degree of
crosslinking in the posterior annulus may be manipulated to positively modify the
material characteristics o f the tissue.
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C hapter 5 — A Pilot Study D esigned to Determ ine the Im pact o f a
Biochem ical Treatm ent on the Fatigue Response of the Annulus
Fibrosis
The cutrent pilot study was designed to investigate the ability of a
crosslinking agent to affect the response of annulus tissue to repetitive stress.
Following is a description of the pilot test protocol and supporting background.
Crosslinking Reagent Selection
As recommended by Hedman, a commercially available crosslinking agent,
genipin was selected (Hedman 2000). Genipin is a naturally occurring plant extract,
from the fruits of Gardenia jasminoides Ellis, as opposed to the conventional, synthetic
crosslinking agents. Genipin has a history of use in herbal medicine (Akao, Kobashi
and Aburada 1994), and reacts with proteins to form deep blue pigments that have
been used in the manufacture of food dyes (Sung et al. 1999b). Genipin is
considered as an ideal crosslinking candidate having been shown to form stable
crosslinks, but avoid the cytotoxicity and calcification problems of synthetic chemical
fixation (Sung et al. 2000). Prior test results that genipin is significantly less cytotoxic
than the most commonly utilized crosslinking reagent, glutaraldehyde (Huang et aL
1998; Sung, Huang et al. 1999; Tsai et al. 2000).
Treatment Method Selection
The reagent exposure method was determined in collaboration with
biochemists in the field of collagen research (Nimni and Han 2001). A qualitative
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
penetration study was carried out for the purpose of determining a method of
delivery for in vitro testing. The relative effectiveness of reagent injection vs. bath
treatment was examined. In this study, one group of dissected IVD’s was injected
with genipin solution (0.33%). It was determined that injection with conventional
hypodermic needles direcdy into the outer layers of the annulus was ineffective.
There was high resistance to injection due to the dense fibrous nature of the tissue
and reagent was simply expelled through the annular space around the needle
insertion. For this reason, crosslinking reagent was injected into the area of
transition between the inner annulus and the nucleus pulposus.
Injected specimens were allowed to stand for 36 hours at ambient
temperature while sealed in plastic to prevent dehydration. Specimens were then
dissected in the transverse plane through the approximate midline of the disc to
visually observe the penetration pattern of the crosslinking reagent. In these cases
the genipin was found to have diffused 3 — 4 mm from the injection needle tip, as
indicated by deep blue discoloration.
Additional specimens were treated in a 500 ml bath of 0.33% genipin
solution for 36 hours. These specimens were also dissected in the transverse plane
through the approximate midlinc of the disc to visually observe the genipin
penetration pattern. For these specimens a 3-4 mm band of deep blue discoloration
was observed around the outer layers of the annulus at the time of dissection.
Following dissection and exposure to air, the specimens were sealed and refrigerated
overnight. Subsequent examination revealed that the blue discoloration was
observable throughout the IV D , including the nucleus pulposus and the in n e r and
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
outer annulus (Figure 5-1). The blue discoloration extended from the exposed tissue
surface deep into the tissues adjacent to the end plate. The blue discoloration was
interpreted as an indication of the thorough penetration o f the crosslinldng agent
Figure 5-1: D epth o f penetration o f genipin tre a tm e n t Transverse cross section of an
intcrvertebraL disc showing depth of penetration from genipin bath treatment
Based on the above observations, bath treatment was selected for the in vitro
pilot testing. Note that it is recognized that this method of treatment is not
appropriate for use in vivo, so that further research beyond the scope of this study
will be required to solve the issue of delivery.
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the posterior annulus, was left intact. The spinous and transverse processes were
removed by cutting the pedicles in the frontal plane. On the anterior and lateral
faces of the vertebral body, a 2 to 3 mm band of periosteum was removed to allow
for better adhesion of the urethane potting material to the cortical bone. A 5 to 10
mm band of periosteum remained intact above and below the disc endplates in order
to insure the integrity of the intervertebral disc.
Chemical Treatment
Prepared segments were then selected for biochemical treatment. Thirteen
specimens were selected for treatment with 0.33 Mol % genipin in phosphate
buffered saline (P.B.S.). Specimens were treated in a 0.5 L bath o f soludon for 36
hours (+/-1 hour) at ambient room temperature. The 36 hour treatment period is
consistent with prior studies involving the fixation of porcine heart valves with
genipin (Sung et al. 1999b). Seven additional specimens were treated with 0.033 Mol
% genipin in the same manner.
Similarly, seven control specimens were exposed to a PBS bath, without
genipin. Note that an IVD in a saline bath, in the absence of a compressive load, will
imbibe fluid due to Donnan osmotic pressure. Care was taken in specifying the
control bath to prevent this excess hydration, by including 1 M NaCL This is based
on unpublished results from Han that indicate that the amount of tissue swelling
steadily decreases with increasing N a d concentration, and is negligible at 1 M NaCl
(Han 2001). Based on the results o f these seven control specimens, explained in
57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter 6, thirteen additional controls were prepared without exposure to the PBS
bath.
The selection of treatments was evenly distributed by spine and level, as
much as practical (Appendix A-l).
Following treatment, the specimens were removed from the treatment bath
and rinsed in saline. Vertebra ends were fixed in urethane to allow clamping to the
mechanical test fixtures.
Mechanical Testing Protocol
Indentation testing and compression/flexion fatigue cycling were carried out
in the sequence outlined in Table 5-1.
Measurement
Sequence
Measurement Location
1 Stress Relaxation Center of the Posterior Annulus
2
Creep Displacement Center of the Posterior Annulus
3 Hardness Center of the Posterior Annulus
3000 Compression/Flexion Fatigue Cycles
4 Stress Relaxation 4 mm Lateral to Center
5 Creep Displacement 4 mm Lateral to Center
6 Hardness Center of the Posterior Annulus
3000 Compression/Flexion Fatigue Cycles
7 Stress Relaxation 4 mm Lateral to Center (Opposite Side)
8 Creep Displacement 4 mm Lateral to Center (Opposite Side)
9 Hardness Center of the Posterior Annulus
T able 5-1: Pilot study m echanical test protocol.
Note that this test protocol was based on the principles established in
preliminary indentation testing refinement studies, as described in Chapter 2. Recall
that it was found that viscoelastic effects asymptotically decrease with repeated
loading. As a result, hardness measurements are sensitive to the loading history of
the tissne. However this effect becomes negligible following 10 loading cycles.
58
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
maatM
F iguic 5-2: Indentation testing apparatus.
All indentation tests were earned out with a 2.5 mm hemispherical indenter
attached to an MTS 858.02 biaxial, tabletop, 10 kN capacity servo-hydraulic materials
60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
test station, with the MTS Test Star data acquisition system (Figure 5-2). A 50 N
capacity external load cell was installed in order to provide more precise force
measurements and control feedback, relative to the internal MTS load cell.
Fatigue Loading P r o c e d u r e
Figure 5-3: Fatigue testing apparatus.
The fatigue cycling procedure is as follows. The specimen was clamped to
the MTS test table and loaded sinusoidally in flexion/compression from 20 N to 200
N for 3000 cycles at 0.5 Hz. The load was applied perpendicularly to the transverse
plane, 40 mm. anterior to the mid-point of the specimen in the transverse plane
(Figures 5-3 and 5-4). Note that the load is applied vertically downward, such that
the moment applied to the motion segment is equal to the force multiplied by the
61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
40 mm
M oment A nn
Resistance
Moment
Anterior Posterior
Figure 5-5: Fatigue testing m echanical m odel. Mechanical model of the lumbar motion segment
subjected to fatigue loading. Note that the intervertebtal disc is modeled as a simple hinge with a
constant center of rotation and an inherent resistance moment.
Alternate Mechanical Test P r o c e d u r e
Two specimens from each of the treatment and control test groups were
selected to undergo an alternate test procedure. In addidon to the Standard Test
Procedure outlined above, these specimens were subjected to indentation testing
before and after biochemical treatment, but before potting in urethane. In this way
it was possible to qualitatively review the impact of the biochemical bath on the test
specimens.
63
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Test specimens were prepared in the same manner as indicated above. Prior
to biochemical treatment, the specimen was clamped direcdy to the test table, and
subsequently subjected to indentation testing. Likewise, the specimen was tested
following the bath treatment. Note that each instance of viscoelastic property
measurement (creep and stress relaxation) was carried out at a new location on the
posterior annulus, and each elastic-plastic measurement (hardness) was recorded at
the center.
Tissue Dehydration
During each of the preparation, loading and testing procedures outlined
above, specimens were managed to minimize specimen dehydration. Specimens
were wrapped in plastic and frozen upon receipt from the supplier. Specimens were
allowed to thaw following removal of the motion segment from the donor spine,
after which they were not re-frozen. Specimens were placed into the treatment or
control bath immediately following specimen preparation. Following bath treatment,
the specimens were potted in urethane and tested within 24 hours. Specimens were
sealed in plastic and refrigerated overnight prior to testing in many cases. In all
cases, the specimens were wrapped in saline wetted gauze and/or sealed in plastic
whenever the specimen was not being handled or tested. The specimen was also
sprayed and wrapped with saline wetted gauze during the 100-minute fatigue loading
sessions.
Pflaster has investigated the impact of specimen preparation on disc
hydration and found that a similar specimen preparation protocol, including
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
additional freeze-thaw cycles, did not lead to a significant change in disc hydration
(Pflaster et al. 1997).
The specimen preparation procedures given above address the issue of
dehydration due to exposure to the ambient environment However, these measures
do not address disc dehydration from fluid exudation through the annulus and the
vertebral end plates due to compressive loading. Refer to Chapter 7 for additional
information regarding the potential of load-induced dehydration.
In Vitro Model Validity
In reviewing this type of in vitro study, it is important to address the
applicability of attributing in vitro results to expected in vivo behavior. In this case, the
intervertebral disc is a primarily avascular structure, which relies on diffusion through
the cndplates to provide nutrients to its limited number of viable cells (Boden et al.
1991). Additionally, the steady accumulation of AGE pentosidine crosslinks in
mature IVD tissue indicates a small rate of collagen regeneration (Duance et al.
1998). These factors indicate that the IVD has a limited regenerative capability, and
suggest that regularly occurring tissue damage would accumulate with time.
Consequently, while it is acknowledged that this study does not incorporate
biological healing mechanisms, it is reasonable to suggest that the in vitro model is
indicative of the tissue reaction in vi vo .
65
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C hapter 6 — Interpretation o f the Results o f the Pilot Study
Control Treatment
The strategy for control treatment was evaluated and modified in the course
of the current study due to the findings of the alternate mechanical testing protocol
Recall that 6 specimens, including 2 controls, were indentation tested before and
after biochemical treatment. The controls in this case were subject to exposure to a
PBS bath for 36 hours.
'lhe alternate mechanical test results revealed qualitatively that the control
bath was having a noticeable impact on the material properties. The viscoelastic
properties (creep and stress relaxation) were observed to increase by 99% and 49%,
respectively (Figure 6-1, 6-2). Meanwhile, the elastic-plastic property (hardness) did
not show a significant trend (Figure 6-3).
e
■ S
et
X
V
a
§
in
7
6
5
4
3
Post Treatment Pre-Treatment
i — O — Control (Bath) Average
| A 033% Geneptn Treatment Average
■0.033% Genepin Treatment Average
Figure 6-1: Pilot study alternate teat procedure stress relaxation results. Note that due to small
sample size and significant variance, the results may only be considered qualitatively.
66
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0.3
0.15
0.1
Pre-Treatment Po st Treatment
Control (Bath) Average
0.33% Gencpin Treatment Average
— O— 0.033% Genepin Average
I
Figure 6-2: Pilot study alternate test procedure creep deform ation results. Note that due to
small sample size and significant variance, the results may only be considered qualitatively.
Pre-Treatment Post Treatment
Control (Bath) Average
033% Genepin Treatment Average
0.033% Gencpin Treatment Average
Figure 6-3: P ilot study alternate test procedure hardness results. Note that due to small sample
size and significant variance, the results may only be considered qualitatively.
It is important to realize the limitations of these test results. The test
specimens were clamped directly to the MTS test table without benefit o f urethane
potting. In this scenario, the blunt point of the anterior face of the disc is the point
67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
but 4 data sets are non-normal (Appendix A-2). For this reason, a Mann-Whitney
non-parametnc rank sum test is utilized for hypothesis testing.
For the results to follow, it is important to view the data within the context
of the statistical significance and the statistical power of the experiment. Recall from
fundamental statistics that the statistical significance (a) represents the probability of
a “false positive.” Similarly, (3 represents the probability of a “false negative.” The
statistical power is defined as P=(l-P), and represents the probability of not
accepting a false null hypothesis. In the design of hypothesis tests, there is a
relationship between the appropriate sample size, the standard deviation of the
measurements, and the desired statistical significance and statistical power of the
result Lieber has presented the following reiterative equation: (Lieber 1990)
*^a,v+^2(l-P),yY (6* 1)
Where: n = sample size
cr = population standard deviation
8 — difference to be detected
ta v = the t value corresponding to or, v
P = the desired statistical power
In this study, this relationship is used in a reiterative calculation of the
statistical power for each result, based on the available sample size. The difference
to be detected, 8, is consistendy selected to be 20% of the control measurement,
unless otherwise noted. Likewise, the desired significance level was selected to be
0.05.
69
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In this way, the results may be interpreted with the available statistical power
and significance level in mind. In general, the null hypothesis will be rejected if the
calculated p value is less than the desired significance level.
Pilot Study Stress Relaxation Analysis
Determination of the stress relaxation results is illustrated by the following
representative sample (Figure 6-4).
0.7
0.6
10 -- 0.5
0.4 S
0.3
- 0 .2 Q
0.1
Time (s)
Force ~ ■1 Displacement
Figure 6-4: Stress relaxation data sam ple. Representative sample of the data recorded during
stress relaxation testing.
The initial 30 seconds represents the ramp loading of the indenter into the
posterior surface of the annulus tissue. Subscqucndy, the displacement is held
constant for 60 seconds and the stress required to maintain that displacement is
recorded. The stress relaxation represents the change in force required to maintain a
constant displacement during that 60-second period.
70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The stress relaxation curves displayed a consistent pattern o f asymptotic
decrease. Figure 6-4 is typical and indicates that after 60 seconds of stress relaxation,
the rate of change is approaching zero. This may be confirmed by examining the
slope of the regression line approximating the final 5% of data (Figure 6-5). In this
sample case the slope of the regression line is -0.0211, indicating that the rate of
change has dropped to approximately 0.5% per second. The results for all
specimens indicated that the absolute value of the slope of the regression line is
consistendy less than 1% per second after 60 seconds of stress relaxation. This
indicates that measurements from all specimens arc approaching a steady state and
that the data is comparable.
0.7
12 - 0.6
= 0.859
10 - 0.5
- 0.4 c
0.2
y = -0.021 lx + 5.9357
R2 = 0.0161
0.1
30 75 0 15 45 60 90 105
Time (s)
Figure 6-5: Stress relaxation data sample — with regression curves. Representative sample of
the data recorded during stress relaxation testing, with regression curves at the ends o f the test period.
The coefficients of determination are also shown.
As seen in the figure, the displacement signal is dean, however, the force
signal is observed to contain significant noise. In order to minimize the impact of
71
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
noise on the stress relaxation measurement, regression techniques were applied to
the initial and final 5% of the force data during the stress relaxation period. It was
decided not to apply non-linear or pseudo-linear regression to the entire test data set
in order to avoid errors at the end points, which are our points of interest. In other
words, non-linear or pseudo-linear methods applied to the entire curve may result in
a good representation of the entire curve, but at the expense of accuracy at the end
points. Consequently, the resulting stress relaxation calculation is based upon the
difference between the initial and final force values as estimated by first order
regression curve approximations.
In calculating an appropriate regression curve, the creep deformation data
was fitted to a linear function of the form j — a *x + b, where the coefficients a and b
are determined by calculation o f the least squares fit.
The effectiveness o f the regression estimate of the initial and final 5% of the
force data may be evaluated by examining the coefficient of determination, R2 . The
average coefficient of determination for linear regression of the initial 5% of the
force data for all samples is 0.792 (st. dev. 0.145), indicating acceptable reliability.
The coefficient of determination for the final 5% of the force data is a very
low 0.023 (st. dev. 0.020). Recall from statistics that the coefficient of determination
is defined as the quotient o f the regression sum of squares (SSR) divided by the total
sum of squares (‘ TSS). As seen in figure 6-6, as the regression line approaches the
horizontal, the SSR, and consequendv R2 , approach zero.
72
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
y
%
y - p \
¥
y
y
S S E ^ ( y - f i ) 2
t s s = Y d( y - ¥ ) 2 = s s r +s s e
r2 = s s r
TSS
x
Figure 6-6: Calculation of the coefficient of determination. Schematic showing the calculation
of the coefficient o f determination for linear regression o f an imaginary data set. The regression line
is represented by the solid line, fi , while the mean y-value is represented by the dashed line, y
(Freund and Wilson 1997).
Therefore, in this case, the low value of R2 does not signal low reliability, but
rather is a result of the very small rate of change of the mean measured force in this
region. Therefore, the linear regression estimation for the final 5% is acceptable
despite a low value of R2 .
Summary ofStress Relaxation Results
A 2-way ANOVA applied to the stress relaxation measurements strongly
indicates an effect due to crosslinking treatment (2-way ANOVA p = l.09x10'®).
73
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Response to Fatigue Loading
Also apparent is a reduction, due to crosslinldng treatment, in the rate of
change in stress relaxation with repetitive loading. Stress relaxation for treated
specimens increased by 0.474 N (st. dev.: 1.006 N) following 6000 cycles of
repetitive loading, 69.0% lower relative to the 1.530 N (st. dev.: 0.809 N) increase
observed for control specimens (Mann-Whitney p=0.044).
Note that in this case, due to an exceptionally large difference in
measurements between the control and treatment groups, the desired detectable
difference was selected to be 65% of the control value. As a result, the experiment
design provides an available significance level of 0.05 and statistical power of 0.7. In
other words, there is a 5% probability of finding a “false positive,” and a 30%
probability of finding a false negative. Therefore, we can conclude that there is an
effect due to the crosslinking treatment, with an acceptable 5% probability of error.
Pilot Study Cree p Analysis
Determination of the pilot study creep results is illustrated by the following
representative sample (Figure 6-8).
75
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
-r 0.9
0.8
- 0.7
- 0.6
14
12 -
-- 0.4 «
- 0 .2 Q
0.1
90
Time (s)
Force Displacement
Figure 6-8: C reep deform ation data sam ple. Representative sample of the data recorded during
creep testing.
Analogous to the stress relaxation test, the initial 30 seconds represents the
ramp loading of the indenter into the posterior surface of the annulus tissue. In this
case, the applied force is then held constant for 60 seconds and the resulting increase
in displacement is recorded. The creep deformation represents the change in
displacement during the period of constant force application.
The pattern of creep deformation is consistently represented by a continuous
increase in displacement. The rate of displacement is not linear, but decreases with
time. However, the creep displacement does not reach an asymptotic steady state
within the 60-second period.
This implies that, unlike stress relaxation, more than one calculation is
required to present a meaningful picture of creep behavior. Calculating the creep
deformation within the 60-second test period may be a good indication o f the
material behavior in the earliest stage of creep, but it does not provide insight into
76
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the creep equilibrium state. However, additional information in that initial 60-
second period may provide clues as to the equilibrium state. Specifically, if
regression techniques are applied to the individual creep curves it is possible to
determine the rate of creep deformation at the end of the 60-sccond test period
(Figure 6-9). The rate of change at this point may be viewed as a relative indication
of proximity to the equilibrium region of constant creep deformation seen in Figure
2-5. In this way, the 60-second test interval results may be used to compare both the
initial creep displacement rate and the relative rate at which the specimens are
approaching the equilibrium state.
0.9
16
14
- 0.8
7 = 0.0982Ln(x) + 0.3991
0.7
R = 0.9906
0.6
0.4
- 0.2
0.1
Time (s)
F igure 6-9: Creep deform ation d ata sam ple— w ith regression curve. Representative sample of
creep testing data with regression curve. The regression curve equation and coefficient of
determination are shown.
In calculating an appropriate regression curve, the creep deformation data
was fitted to a natural logarithm ic function of the form y — a */n(x)+b, where the
coefficients a and b are determined by calculation of the least squares fit. The
77
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
coefficient of determination, R2 , is calculated in ordet to confirm the goodness of fit
of the regression curve.
Note that two problems are seen in the measured force data. As mentioned
previously, the force feedback signal is contaminated by high frequency noise. This
noise may be attributed to the external load cell sensor rather than a representation
of noise in the actual applied load, as evidenced by the presence of high frequency
noise during no-load conditions. Therefore, it is possible to minimize the effects of
the high frequency noise, where required in data analysis, through the application of
filters or regression techniques.
However, as seen to a small degree in Figure 6-8, and to a larger degree in
Figures 6-10 and 6-11, the mean applied force exhibited low frequency drift (Figure
6-10, 6-11). This low frequency drift represents error in the applied load, rather than
sensor noise. As seen in figures 6-10 and 6-11, this error takes the form of
overshoot (13% in figure 6-10), and error relative to the target force throughout the
test and at the end points of the test interval. In this way, error and variation arc
introduced into the experiment
A possible contribution to this error in applied load is the high frequency
noise in the force feedback signal. Note that the creep displacement test is
performed under force control The noisy input signal is likely to impact the
performance of the MTS controller.
78
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
30 60 90
Time (s)
Force •Displacement j
Figure 6-10: C reep deform ation data sam ple — controller overshoot and error. Representative
sample of the data recorded during creep testing. Nodcc the large overshoot and absolute error.
o
u.
18
16
14
12
1 0
8
6
4
2
0
-r 0.9
0.8
0.7
0.6
0.5
0.4
0.3
- 0.2
0.1
0 20 40 60 80 too
e
u
S
u
u
J3
£ •
.2
Time (s)
Force •Displacement
Figure 6- 11: C reep deform ation data sam ple — controller drift. Representative sample of the data
recorded during creep testing. Notice the low frequency drift, and downward shift of the applied force
at the end o f the creep testing interval.
This error and variation contribute to a relative increase in variance observed
for creep displacement measurements reladve to stress relaxation. The average
variance for stress relaxation, when expressed as a percent o f the mean, is 15.7%,
79
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
compared to 21.1% for creep deformation. As a consequence, the trends observed
in the creep results are generally rendered statistically insignificant. Note that these
average calculations include pre-fatigue, post 3000 cycle, and post 6000 cycle
measurements and arc weighted according to the relative number of treatment
specimens.
Summary of Creep Results
A 2-way ANOVA applied to the creep deformation results strongly indicates
a general effect due to crosslinking treatment (2-way ANOVA p=5.51xl0"7 ).
The summary results of the creep test are seen in Figures 6-12 and 6-13.
(Refer to Appendices A-6, A-7, A-8 for complete results.)
Analogous to stress relaxation, there are two potential effects on creep
deformation to be evaluated, an immediate impact on the measured creep
deformation prior to fatigue cycling and an impact on how the measured creep
deformation changes with repetitive loading.
Pre-Fatigue Impact
There appears to be a downward shift in creep measurements due to the
crosslinking treatment, prior to fatigue loading (Figures 6-12 and 6-13). The 60-
second creep deformation for treated specimens was recorded at 0.096 mm (st. dev.
0.034 mm) prior to repetitive loading. This is 19.3% lower relative to the 0.119 mm
(st. dev. 0.036 mm) recorded for control specimens (Mann-Whitney p=0.092). As
outlined above, this potentially indicates a change in the initial creep behavior.
80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In this case, the results fail to satisfy our desired significance level of 0.05,
therefore the null hypothesis could not be confidendy rejected. The statistical power
was calculated to be less than 0.55 for these results. In other words, there is a greater
than 45% probability that we found a “false negative.” Therefore, wc can draw only
a marginal conclusion that there is an effect due to the crosslinking treatment, with a
9.2% probability of error.
0.2 j
0.18
0.16
0.14
0.12
0.06
0.04
0.02
Pre-Fatigue Post 3000 Fatigue Cycles Post 6000 Fatigue Cycles
Control Average 0-33% Gcncpin Average
Figure 6- 12: Pilot study creep deform ation results. Pilot test results showing 60-second creep
deformation. This may be viewed as an indication of the material behavior in the initial stage of
creep.
Similarly, there is an apparent change in the rate of creep deformation at the
end of the 60-second test period, prior to fatigue cycling. The final creep rate for
treated specimens is 8.99x1c4 mm/s (st dev.: 3.47x1c4 mm/s), compared to
1.13xl0'3 mm/s (st. dev.: 3.57x10^ mm/s) for control specimens. This represents a
20.8% drop in the final creep rate due to crosslinking treatment. This suggests that
at the end of the 60-second test period, the treated specimens are relatively closer to
81
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
creep equilibrium (Mann-Whitney p=0.082). This supports the assertion that the
treated specimens exhibit reduced viscoelasticity.
In this case, the results fail to satisfy our desired significance level of 0.05,
therefore the null hypothesis could not be confidently rejected. The statistical power
was calculated to be 0.75 for these results. In other words, there is a 25% probability
that we found a “false negative.” Therefore, we can draw only a marginal conclusion
that there is an effect due to the crosslinking treatment, with an 8.2% probability of
error.
0.002
0.0018
^ 0.0016
I | 0.0014
I V 0.0012
o.ooi
e u
g 0.0008
75 0.0006
.2
“« 0.0004
0.0002
0
Figure 6- 13: Pilot study creep deform ation results — creep rate. Instantaneous creep rate at the
end of the 60-second test period, as determined by the slope of the corresponding regression curve.
This may be viewed as an indication of the relative proximity to the equilibrium state, Le., lower creep
rates indicate that the tissue is closer to the point of equilibrium.
Response to Fatigue Loading
In order to evaluate the impact on how the measured creep deformation
changes with repetitive loading it is necessary to determine the change in creep
behavior from pre-fatigue to post-6000 cycle measurements. 60-second creep
82
i
Post 6000 Cycles Post 3000 Cycles
i
, |
Control 1 A 0.33% Genepin Treatment i
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
deformation for treated specimens increased by 0.041 mm (st. dev.: 0.022 mm)
following 6000 cycles of repetitive loading, 18.0% lower relative to the 0.050 mm (st.
dev.: 0.036 mm) increase observed for control specimens (Mann-Whitney p=0.786).
This result was not found to be statistically significant due to the high
standard deviations (average 0.029 mm) recorded for this data relative to the desired
detectable difference (0.010 mm).
In order to evaluate the creep behavior it is also important to examine the
change in the final creep rate at the end of the 60-second test period. The final creep
rate for treated specimens increased 2.83 x 10"4 mm/s (st. dev.: 2.32 x 1 0 "* mm/s)
after 6000 cycles of repetitive loading. This is 35.5% lower than the increase in the
final creep rate, 4.39x1 O '4 mm/s (st. dev.: 3.26xl04 mm/s), observed for control
specimens (Mann-Whitncy p=0.172).
In this case as well, the result was not found to be statistically significant due
to the high standard deviations (average 2.79x1 O '4 mm/s) recorded relative to the
desired detectable difference (9.2x10‘ 5 mm/s).
Summary of Ch an ges in Viscoelastic Behavior
The pilot study results suggests that crosslinking treatment may have two
principle effects on the viscoelastic properties of the posterior annulus tissue. There
appears to be an immediate downward shift in viscoelastic properties due to genipin
treatment, prior to any repetitive loading. This assertion is well supported by the
stress relaxation data, and marginally supported by the 60-second creep deformation
and final creep rate data.
83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12 0.8
0.7
10
0.6
8
6
4
- 0.2
2
0.1
0
i
0 5 10 15 20 25 30 35
i
I Time(s) j
i . !
J Force Displacement I j
Figure 6-14: H ardness data sam ple. Representative sample o f the data recorded during hardness
testing.
The 30-second test period represents the ramp loading of the indenter into
the posterior surface of the annulus tissue. The calculated hardness index is the
change in indenter load divided by the depth of indentation.
Two considerations must be accounted for in making this calculation. First,
there is a small amount of “slack” in the test specimen. In other words, there is a
period of indenter displacement with no corresponding load on the indenter. Note
that this is due to residual deformation from the pre-cycling that precedes the
hardness test. In order to account for this “slack,” the initial force and displacement
measurements are taken at the point that the indenter load reaches 0.25 N. The final
force and displacement measurements are taken at the end of the 30-second test
period.
85
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Second, the force signal contains high fre q u e n c y noise, as observed in the
stress relaxation test. Sim ilarly to the case of stress relaxation, we minimize the
impact of this noise by applying regression techniques to the end points of the test
period (Figure 6-15). Specifically, the initial point is considered the instant that the
indenter load reaches 0.25N, and is determined by curve fitting to the 5% of the data
set preceding and following that point. In calculating an appropriate regression
curve, the data was fitted to a polynomial function of the £atm y— a * * ? + b*x? + c*x
+ d, where the coefficients a, b, c and d are determined by least squares fit calculation.
y = 0.3478x : 0.4573
R2 = 0.7959
y = -0.0182xJ + 0.5224x - 4.4854x + 12.159
R2 = 0.9418
T 0.8
-- 0.7
0.6 |
0.5
• 0.4
u
e
u
0.3 J
£•
0.2 Q
0.1
0
10 20 30
Hme(s)
Figure 6-15: H ardness d ata sam ple — w ith regression curves. Representative sample o f the data
recorded during hardness testing showing regression curves at the test period end points. The
corresponding regression curve equations and coefficients o f determination arc also shown.
The final data point is determined by curve fitting to the final 10% of the
data set. In this case, the data was fitted to a linear function of the form j — a *x + b,
where the coefficients a and b are determined by calculation of the least squares fit
The effectiveness of the regression estimate of the initial and final 5% of the
force data may be evaluated by examining the coefficient of determination, R2 . The
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1.4
1.2
+ 0.8 a
Time(s)
Force Displacement I
Figure 6-17: H ardness d ata sam ple — non-linear force ram p. Representative sample of the data
recorded during hardness testing showing a highly non-linear force ramp.
Summary o f Hardness Results
A 2-way ANOVA applied to the hardness measurements strongly indicates
an effect due to crosslinking treatment (2-way ANOVA p=6.41xl0'®).
The summary results of the hardness testing are seen in Figures 6-18. (Refer
to Appendices A-9, A-10, and A-ll for complete results.)
Analogous to the viscoelastic properties, there are two potential effects on
hardness to be evaluated, an immediate impact on the measured hardness index prior
to fatigue cycling and an impact on how the measured hardness index changes with
repeddve loading.
88
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In this case, the results fail to satisfy our desired significance level of 0.05,
therefore the null hypothesis could not be confidendy rejected. The statistical power
was calculated to be 0.80 for these results. In other words, there is a 20% probability
that we found a "false negative.” Therefore, the calculated p value for this data lends
support to the assertion that there is no detectable difference in measured hardness
between treated and untreated specimens following repetitive loading, with a 20%
probability of error.
Response to Fatigue Loading
As seen in Figure 6-18, there appears to be no difference in the response of
treated and untreated specimens to fatigue loading, following the elimination of the
initial pre-fatigue shift. Between 3000 cycles and 6000 cycles of repetitive loading,
the hardness index for treated specimens increased by 0.94 N/mm (st. dev.: 3.07
N/mm), compared to the 1.01 N/mm (st. dev.: 1.42 N/mm) increase observed for
control specimens (Mann-Whitney p=0.328).
In this case, the high standard deviation, 2.25 N/m m (average), relative to
the desired difference to be detected, 0.20 N/mm, does not allow us to achieve
meaningful significance or statistical power. The statistical power for these results is
significantly less than 0.55, implying that the probability of finding a "false negative”
is significantly greater than 45%. Therefore, we are unable to draw conclusions
regarding the assertion that from 3000 cycles to 6000 cycles of repeated loading,
there is no detectable difference in rate of change o f measured hardness between
treated and untreated specimens.
90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Summary ofChanges in Elastic-Plastic ’ behavior
The above results indicate that there is a significant increase in hardness due
to crosslinking treatment prior to fatigue loading. However, this increase appears to
be temporary. Following 3000 cycles of loading, there is no detectable difference
between the hardness recorded for treated and untreated specimens. There appears
to be no distinction between the rates of change of hardness from 3000 to 6000
cycles for treated and untreated specimens, however confidence in this result is low.
Genipin Concentration and the Observed Results
The preceding analysis of stress relaxation, creep deformation and hardness
has focused on the effects of treatment with 0.33 % genipin. This concentration was
selected to agree in order of magnitude with the concentrations utilized in prior
studies (0.250%, 0.625%, 1.000%) (Sung et al. 1999a, Sung et al. 2000) In addition to
this standard treatment, a limited number (#) of specimens received a dilute
treatment o f0.033 % genipin.
The results for the dilute genipin are distinct from the 0.33% genipin
treatment, and not as potentially beneficial.
In the case of stress relaxation, an initial shift prior to fatigue cycling is found
to b e o f sim ila r magnitude to the shift resulting from 0 .3 3 % genipin treatment
(Figure 6 -1 9 ). The stress relaxation for dilute treated specimens was recorded at
3 .4 0 N (st. dev.: 0 .2 9 N) prior to repetitive loading. This is a 2 5 .6 % reduction
relative to the 4 .5 7 N (st. dev.: 0 .9 7 N) recorded for control specimens. (Mann-
91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
found to be of similar magnitude to the shift resulting from 0.33% genipin
treatment The 60-second creep deformation for dilute treated specimens was
recorded at 0.091 mm (st dev.: 0.007 N) prior to repetitive loading. This is a 23.5%
reduction relative to the 0.119 mm (st dev.: 0.036) recorded for control specimens.
(Mann-Whitney p=0.lll). This is similar in magnitude to the 19.6% downward shift
recorded for the 0.33% genipin treated specimens.
0.18
0.16
' S ' 0.14
T 0.12
0.08
0.06
0.04
Pre-Fatigue Post 3000 Fatigue Cycles Post 6000 Fatigue Cycles
0.033% Genepin Average 0.33% Genepin Average j j
Figure 6-20: Pilot study creep deform ation results - effect o f genipin concentration. Summary
o f the pilot test 60-second creep deformation results including the dilute genipin treatment.
Similar to the 0.33% treatment specimens, the cate of increase in 60-second
creep defonnation due to repeated cycling for the dilute treatment specimens is
similar to the control specimens. Dilute treatment specimens showed a 0.048 mm
(st. dev.:0.016 mm) increase in 60-second creep deformation following 6000 cycles o f
repetitive loading, which is comparable to the 0.050 mm (st. dev.: 0.036 mm)
increase in the control specimen data. A Mann-Whitney non-parametric rank sum
93
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
test yielded p=0.640, which lends support to the assertion that the 60-second creep
deformation for control and dilute treated specimens are increasing with fatigue
cycling at a similar rate.
There is not an apparent significant effect on the recorded hardness due to
the dilute genipin treatment (Figure 6-21). The pre-fatigue hardness index
measurement for dilute treated specimens was recorded at 17.59 N/m m (st. dev.:
1.87 N/mm). This is a 6.5% reduction relative to the 18.82 N/mm (st. dev.: 2.82
N/mm) recorded for control specimens (Mann-Whitney p=0.512). This is in
contrast to the 16.5% upward shift recorded for the 0.33% genipin treated
specimens. The relatively high p value lends support to the assertion that there is no
significant effect on the pre-fatigue hardness index due to the dilute treatment
16
Post 3000 Fatigue Cycles Post 6000 Fatigue Cycles Pte-Fatigue
0.033% Genepin Average 033% Genepin Average
Figure 6-21; Pilot test hardness results — effect o f genipin concentration. Summary o f the pilot
test hardness results including dilute genipin treatment.
94
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Similarly, the rate of change of hardness due to fatigue cycling is comparable
for the dilute treatment and control specimens. Dilute treatment specimens showed
a 3.17 N/m m (st dev.: 1.19 N/mm) decrease in hardness index following 6000
cycles of repetitive loading, which is comparable to the 3.99 N/m m (st dev.: 2.16
N/mm) decrease in the control specimen data. A Mann-Whitney non-parametnc
rank sum test yields p=0.461, which lends support to the assertion that the hardness
index for control and dilute treated specimens are decreasing with fadgue cycling at a
similar rate.
Summary of P ikt Study Results
The pilot study results indicate that there is a temporary increase in recorded
hardness due to genipin treatment. The eliminadon o f this effect with repeddve
loading raises interesting quesdons. Recall that the hardness index measurement
utilized in this study is an indication of both elastic and plastic properties. In this
way, the hardness index is a combined indication of the elastic modulus and the yield
strength. Note that an increase in the elastic modulus, or stiffness, may be associated
with increased brittleness. For this reason, increased hardness may be viewed as a
negative effect on soft tissue if the increase in hardness represents an increase in the
clastic modulus.
Considering that the relative impact of the elastic modulus and the yield
strength on the hardness index measurement is not known, it is not possible to make
a conclusion regarding material properties based upon the hardness results alone.
95
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C hapter 7 — O utstanding Q uestions and Recom m endations
Fatigue Loading Magnitude
It is important to note that the 200 N (45.0 lbf) fatigue loading condition
used in the current study was intended to represent light loading conditions
experienced during routine activities. In order to put this into perspective, it is
necessary to review prior investigations of the loads experienced by the spine. It has
long been understood that the loads experienced by the spine are significantly greater
than the supported weight due to additional compression from muscle contraction.
It has been suggested that the spine is subject to loads 6 to 8 times greater than the
weight lifted by a subject (Adams and Hutton 1983). Schultz, in collaboration with
Nachemson and others, has presented estimated loads based on a biomechanical
model, as supported by myoelectric and intradiscal pressure measurements. This
study estimated compression forces at the third lumbar level ranging from 380 N
(85.4 lbf) during relaxed sitting, to 440 N (98.9 lbf) during relaxed standing, up to
2350 N (528 lbf) for standing while flexed at the hips 30 degrees with arms extended
forward holding 8 kg (17.6 lb) (Schultz et al. 1982). These values are in agreement
with the frequently referenced work of Nachemson (Nachemson 1966).
The flexion-compression load magnitude impacts the stress experienced by
the posterior annulus in two ways. The axial compression component increases
pressure in a healthy nucleus, which in turn places the surrounding annulus fibers in
tangential tension. In his fundamental study, Nachemson indicates that the
tangential tensile stress in the posterior annulus is 4 to 5 times the applied external
97
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
magnitude in agreement with the widely accepted estimates of load levels in order to
more accurately simulate the nuclear pressurization and resulting increase in
tangential stress. This increase in axial load should be coupled with a corresponding
decrease in the moment arm, in order to keep the applied bending moment in a
reasonable range.
With this in mind, it is particularly important to note that the crosslinking
treatment in the current investigation showed significant fatigue resistant effects in
the response to repetitive loading o f a very low magnitude. It is anticipated that the
results may be augmented when subjected to the larger load magnitudes referenced
above.
Hydration Loss Due to " R e p e ti t iv e Loading
An important issue when loading spine motion segments in compression is
the factor of disc hydration. As indicated in Chapter 5, careful specimen preparation
procedures are able to prevent significant dehydration due to exposure to the
ambient environment. However, these measures do not address disc dehydration
from fluid exudation through the vertebral end plates due to compressive loading.
This mechanism of dehydration may have a potential impact on material
properties measured following 3000 and 6000 load cycles. If this dehydration
mechanism is significant, it may be partially responsible for the tendency of
indentation data to exhibit larger changes from pre-fatigue to post-3000 cycle
measurements relative to the changes from post-3000 to post-6000 cycle
measurements (Figures 3-2,6-19,6-20,6-21).
99
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Note that this argument is countered by the lack of large 0-3000 cycle
changes in the viscoelastic measurements for 0.33% genipin treated specimens
(Figure 6-19, 6-20). However, this may in part be due to a reduced ability of the
treated disc to exude fluid due to the effect of increased crosslinking on the
microporous structure.
An indicadon that other researchers are concerned with this question is
evidenced in adapted test procedures to address the issue of load related disc
dehydration. Adams utilizes a preliminary creep test which is intended to force the
water content of the disc specimens into a uniform physiologic range prior to cyclic
loading (Adams, Green and Dolan 1994). Adams and others have immersed the disc
during fatigue cycling so that the swelling pressure of the disc counterbalances the
outflow of liquid due to fatigue loading (Adams and Hutton 1983, Hedman and
Femie 1997).
Consequently, it is important that the existence of a dehydration effect on
indentation measurements be confirmed or disproved, so that indentation
measurements may be interpreted as representative of tissue material properties
rather than the disc hydration state. For this reason, Syed has carried out a
preliminary study in our laboratory in order to study the impact of repetitive loading
on disc hydration (Syed 2001).
In this study, specimens from a control group and a 0.33% genipin treatment
group were subjected to dehydration testing in order to determine water content
before cycling, following 3000 load cycles, and following 6000 load cycles (Table 7-
1 ).
100
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
value is greater than 0.05 in all cases, supporting the suggestion that there is no
significant relationship. The results from this preliminary study suggest that the
changes in indentation measurements due to fatigue loading represent changes in
tissue material properties rather than changes in disc hydration.
C hange in Disc Hydration due to Bath Treatment
However, the preliminary hydration study results presented above do indicate
that the bath treatment does impact the water content of the disc tissue. This effect
is particularly noticeable in the outer annulus. A Mann-Whitney non-parametric rank
sum test was applied to the data to test the hypothesis that there is no change in disc
hydration for like specimens due to genipin treatment (Table 7-3). Calculated p-
values of 0.05 for the outer posterior annulus tissue allow us to conclude that there is
a significant increase in the outer annulus water content due to the genipin bath
treatment.
Nucleus Inner Posterior Annulus O uter Posterior Annulus
0
Cycles
3000
Cycles
6000
Cycles
0
Cycles
3000
Cycles
6000
Cycles
0
Cycles
3000
Cycles
6000
Cycles
Mann Whitney Test
Result - Control vs.
Treated Specimens (p) 0.275 0.127 0.827 0.827 0.05 0.827 0.05 0.05 0.05
Table 7*3: Fatigue induced d u e dehydration study statistics. Mann-Whitney test utilized to test
the hypothesis that there is no change in disc hydration for like specimens due to genipin treatment.
This finding is significant due to the implication that some of the material
property effects attributed to the genipin treatment in the pilot study may be due to
changes in the water content of the outer annulus. Future testing can address this
issue by incorporating control specimens subject to a modified saline bath treatment
in order to reach a similar hydration levels. Alternately, additional testing may be
102
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
carded out to determine what, if any, effects on material properties may be attributed
to like changes (5%) in hydration content alone.
Indentation Testing and High Variances
As seen in the results presented in Chapter 6, the results of the current test
are often limited in significance by the tendency for the measured standard deviation
to be large relative to the desired detectable change. Preparation of the current test
protocol incorporated refinements resulting from the preliminary testing described in
Chapter 2 in an effort to reduce measured variance. However, it will be valuable for
future studies to make additional modifications to the test protocol to further reduce
variance.
One existing source of potential v a ria tio n results from the indentation testing
of un-damped specimens. In the current study it was assumed that the low
indentation loads would be unable to generate a moment sufficient to rotate the
urethane block and therefore impact the indentation measurement (Figure 7-1).
However, in practice plastic deformation from repetitive loading may deform the
motion segment such that one of the urethane blocks does not rest flat on the test
table (Figure 7-2). A test specimen in this condition may be susceptible to error due
to the movement of the urethane block.
103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Indenter
Anterior
Test Table
Figure 7-1: Indentation test schem atic. Schematic showing indentation test specimen resting
flush on the test table. The small moment applied by the indenter is insufficient to rotate the
urethane block.
For this reason, it may be beneficial to consider clamping the indentation test
specimens in future experiments.
104
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Indenter
Anterior
Test Table
Figure 7-2: Indentation test schem atic — plastic deform ation. Schematic showing indentation
test specimen following fatigue cycling. Plastic deformation may prevent the urethane blocks from
resting flat on the test table. Note that the defotmation is exaggerated for the purpose o f this
schematic. In this case, it may be possible for the moment applied by the indenter to rotate the
urethane block.
jQuantification o f Crosslinks
The current study docs not incorporate the quantification of the degree of
crosslinking due to the genipin treatment. This type of testing will be necessary in
order to establish a direct correlation with the quantity of crosslinks and the tissue
strength properties. In addition it will be possible to examine if there is any change
in the quantity of crosslinks due to repetitive loading. This information will be
critical in considering crosslinking treatment for clinical application.
105
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
At the USC Orthopaedic Research Lab, Dr. Hedman is planning to carry out
histological analysis in order to determine the degree of crosslinking for 4 test
groups.
1. Control specimen, non-fatigucd.
2. Treated specimen, non-fadgued.
3. Control specimen, fadgued.
4. Treated specimen, fadgued.
This test will allow confirmation of a shift in the degree of crosslinking due
to genipin treatment, and provide insight into any correladon between a change in
the degree of crosslinking and repeddve loading.
Specifically, this would determine if crosslinks are being lost or broken when
subjected to repeddve loading, and if the creadon of additional crosslinks would be
beneficial. Once again considering the long-term goal of clinical application, this
data will be critical in determining whether repeated treatments would be beneficial.
In Vivo Testing
The use of crosslinking reagent to treat the annulus tissue of the
intervertebral disc is at the earliest stages of testing. In addition to the significant in
vitro testing which remains, in vivo models will need to be evaluated. Sung has
warned, “there is good evidence that the host environment can significandy change
the mechanical properties of chemically modified biological tissues” (Sung et al.
1999b). In vivo testing always serves as an indicator of the biocompatibility of a
treatment or procedure. In this case, in vivo testing will also be critical to demonstrate
106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Bibliography
Adams MA, Hutton WC. 1982. Prolapsed intervertebral disc: a hyperflexion injury.
Spine. 7:184-191.
Adams MA, Hutton WC. 1983. The effect of fatigue on the lumbar intervertebral
disc. The Journal ofBone and Joint Sutgpry. 65-B(2):l99-203.
Adams MA, Dolan P, Hutton WC. 1986. The stages of disc degeneration as
revealed by discograms. The Journal o f Bone and Joint Surgery. 65-B(l):36-41.
Adams MA, Green TP, Dolan P. 1994. The strength in anterior bending of lumbar
intervertebral discs. Spine. 19(19):2197-2203.
Akao T, Kobashi K, Aburada M. 1994. Enzymatic studies on the animal and
intestinal bacterial metabolism of geniposide. Biological and Pharmaceutical Bulletin.
17:1573-1576.
Anderson DD, Adams DJ, Hale JE. 2000. Mechanical effects of forces acting on
bone, cartilage, ligaments, and tendons. In Biomechanics and Bi ology of Movement, ed.
Nigg BM, Macintosh BR, Mester J. Champaign, IL: Human Kinetics.
Andcrsson, G. 1995. Epidemiology. In Essentials o f the Spine, cd. Weinstein JN,
Rydcvik BL, Sonntag V. New York: Raven Press, 1-10.
Bank RA, Bayliss MT, Lafeber FPJG, Maroudas A, teKoppele JM. 1998. Ageing
and zonal variation in post-transitional modification of collagen in normal human
articular cartilage. Biochemistry Journal. 330:345-351.
Boden SD, Wiesel SW, Laws ER, Rothman RH. 1991. The Aging Spine; Essentials of
Pathophysiology, Diagnosis, and Treatment. Philadelphia: W.B. Saunders Co., 28.
Buckwalter JA, Mow VC, Boden SD, Eyre DR, Weidenbaum M. 2000.
Intervertebral disk structure, composition, and mechanical function. In Orthopaedic
Basic Science; Biology and Biomechanics o f the Musculoskeletal System, ed. Buckwalter JA,
Einhom TA, Simon SR. Rosemont IL: American Academy of Orthopaedic
Surgeons, 551-555,585-588.
Chen AC, Temple MM, Ng DM, Richardson CD, DeGroot J, Verzijl N, teKoppele
JM, Sah RL. 2001. Age-Related Crosslinking Alters Tensile Properties of Articular
Cartilage. Paper presented at the 47th Annual Meeting of the Orthopaedic Research
Society, San Fransisco, CA.
108
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Huang LLH, Sung HW, Tsai CC, Huang DM. 1998. Biocompadbility study of a
biological tissue fixed with a naturally occurring crosslinking reagent Journal o f
Biomedical Materials Re sea rch . 42(4):568-576.
Kawchuk GN, Elliott PD. 1998. Validation of displacement measurements
obtained from ultrasonic images during indentation testing. Ultrasound in Medicine and
Biology. 24(1):105-111.
Klein BP, Jensen RC, Sanderson LM. 1984. Assessment of workers’ compensation
claims for back strains/sprains. Journal o f Occupational Medicine. 26(6):443-448.
Leino PI, Berg MA, Puska P. 1994. Is back pain increasing? Results from national
surveys in Finland during 1978/9-1992. Scandinavian Journal of Rheumatology. 23:269-
276.
Lieber RL. 1990. Statistical significance and statistical power in hypothesis testing.
Journal o f Orthopaedic Rese arch . 8:304-309.
Lotz JC, Colliou OK, Chin JR, Duncan NA, Liebcnbcrg E. 1998. Compression-
induced degeneration of the intervertebral disc: An in vivo mouse model and finite-
element study. Spine. 23(23):2493-2506.
MacGregor CW, Symonds J. 1978. In M arks' Standard Handbook for Mechanical
Engineers, ed. Baumeister T, Avallone EA, Baumeister III T. New York: Me Graw-
Hill Book Company, 5.10-5.11.
Miely WR, McLain R, Weinstein JN, Goel VK, Found EM Jr. 1990. Anatomy of
the lumbar spine. In Biomecbanics o f the Spine: Clinical and Surgical Persp ec tive , ed. Goel
VK, Weinstein JN. Boca Raton, Florida: CRC Press.
Mow VC, Proctor CS, Kelly MA. 1989. Biomcchanics of articular cartilage. In Basic
Biomechanics o f the Musculoskeletal System, ed. Nordin M, Frankel VH. Philadelphia: Lea
& Febiger, 39-40.
Mow VC, Ratdiffe A. 1997. Structure and function of articular cartilage and
meniscus. In Basic Orthopaedic Biomecbanics, ed. Mow VC, Hayes WC. Philadelphia:
Lippincott-Raven Publishers, 151-159.
Nachemson A. 1966. The load on lumbar disks in different posidons of the body.
Clinical Orthopaedics and Related Research. 45:107-122.
Nachemson A. 1975. Towards a better understanding of low-back pain: A review
of the mechanics of the lumbar disc. Rheumatology and Rehabilitation. 14(3) :129-43.
110
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Neideck K, Franzel W, Grau P. 1999. Dynamic ball hardness tests on polymers.
Journal o f Macromolecular Scien ce — Physics. B38(5-6) :669-680.
Nimni ME, Cheung D, Strates B, Kodama M, Sheikh K. 1988. Bioprosthesis
derived from cross-linked and chemically modified collagenous tissues. In Collag en,
VoL m , ed. Nimni ME. Boca Raton, FL: CRC Press. 1-38.
Nimni ME, Han B. 2001. Conversation with author.
Osti OL, Vcmon-Roberts B, Moore R, Fraser RD. 1992. Annular tears and disc
degeneration in the lumbar spine. A post-mortem study of 135 discs. The Journal of
Bone and Joint Surgery. 74-B:678-682.
Pathak AP, Silver-Thom MB, Thierfelder CA, Prieto TE. 1998. A rate-controlled
indentor for in vivo analysis of residual limb tissues. IE E E Transactions on Rehabilitation
Engineering. 6(l):l2-20.
Pflaster DS, Krag MH, Johnson CC, Haugh LD, Pope MH. 1997. Effect of test
environment on intervertebral disc hydration. Spine. 22(2):133-139.
Pokhama HK, Phillips FM. 1998. Collagen crosslinks in human lumbar
intervertebral disc aging. Spine. 23(15):1645-1648.
Pokhama HK, Pottenger f .A. 2000. Glycation induced crosslinking of link proteins,
in vivo and in vitro. Journal o f Surgical Rese ar ch. 94(l):35-42.
Scapino RP, Canham PB, Finlay HM, Mills DK. 1996. The behavior of collagen
fibres in stress relaxation and stress distribution in the jaw-joint disc of rabbits.
Archives o f Oral Biology. 41(11):1039-1052.
Schultz A, Andersson G, Ortengren R, Haderspeck K, Nachemson A. 1982. Loads
on the lumbar spine. Validation of a biomcchanical analysis by measurements of
intradiscal pressures and myoelectric signals. Journal of Bone and Joint Surgery.
64(5):713-720.
Sell DR, Monnier VM. 1989. Structure elucidation of a senescence cross-link from
human extracellular matrix. Implication of pentoses in the aging process. Journal o f
Biological Chemistry. 264(36):21597-21602.
Suh JK, Spilker RL. 1994. Indentation analysis of biphasic articular cartilage: Non
linear phenomena under finite deformation. Journal of Biomechanical Engineering.
116(l):l-9.
I l l
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Sung HW, Chang Y, Chiu CT, Chen CN, Liang HC. 1999(a). Crosslinking
characteristics and mechanical properties of a bovine pericardium fixed with a
naturally occurring crosslinking agent. Journal o f Biomedical Materials Rese ar ch. 47:116-
126.
Sung HW, Chang Y, Chiu CT, Chen CN, Liang HC. 1999(b). Mechanical properties
of a porcine aortic valve fixed with a naturally occurring Crosslinking agent.
Biomaterials. 20:1759-1772.
Sung HW, Chang Y, Liang LL, Chang WH, Chen YC. 2000. Fixation of biological
tissues with a naturally occurring crosslinking agent: Fixation rate and effects of pH,
temperature, and initial fixative concentration. Journal o f Biomedical Materials Re se a rch .
52:77-87.
Sung HW, Huang RN, Huang LLH, Tsai CC. 1999. In atm evaluation of
cytotoxicity of a naturally occurring cross-linking reagent for biological tissue
fixation. Journal of Biomaterials S ci en ce — Polymer Edition. 10(l):63-78.
Syed B. 2001. Load induced intervertcbral disc dehydration study. University of
Southern California.
Tabor D. 1996. Indentation hardness: Fifty years on — A personal view.
Philosophical Magazine A — Physics of Condensed Matter Structure Defects and Mechanical
Properties. 74(5):1207-1212.
Takahashi M, Kushida K, Ohishi T, Kawana K, Hoshino H, Uchiyama A, Inoue T.
1994. Quantitative analysis of crosslinks pyridinoline and pentosidine in articular
cartilage of patients with bone and joint disorders. Arthritis and Rheumatism.
37(5):724-728.
Tsai CC, Huang RN, Sung HW, Liang HC. 2000. In vitro evaluation of the
genotoxidty of a naturally occurring crosslinking agent (genipin) for biologic tissue
fixation. Journal of Biomedical Materials Re sea rc h. 52(l):58-65.
Uchiyama A, Ohishi T, Takahashi M, Kushida K, Inoue T, Fujie M, Horiuchi K.
1991. Fluorophores from aging human articular cartilage. Journal o f Biochemistry
(Tokyo). 110(5):714-718.
Valente M, Minarini M, Maizza AF, Bortootti U, Thiene G. 1992. Heart valve
bioprosthesis durability: A challenge to the new generation o f porcine valves.
European Journal o f Cardio-Tboracic Surgery. 6:S82-90.
Vannah WM, Childress DS. 1996. Indentor tests and finite element modeling of
bulk muscular tissue in vi vo . Journal i f Rehabilitation Research and Development.
33(3)^39-252.
112
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Videman T, Nutminen M, Troup JDG. 1990. Lumbar spinal pathology in cadaveric
material in relation to history of back pain, occupation, and physical loading. Spine.
15(8):728-738.
Waddell, G. Low back pain: 1996. A twentieth century health care enigma. Spine.
21(24) :2820-2825.
Yang C, Mosler S, Rui H, Baetge B, Notbohm H, Muller PK. 1994. Structural and
functional implications of age-related abnormal modifications in collagen II from
intervertebral disc. M atrix Biology. 14(8):643-651.
Yannas IV. 1996. Natural Materials. In Biomaterials Sci enc e: A n Introduction to Materials
in Medicine, ed. Ratner BD, Hoffman AS, Schoen FJ, Lemons JE. San Diego:
Academic Press, 84-92.
Yoganandan N, Cusick JF, Pintar FA, Droese K, Reinartz J. 1994. Cyclic
compression-flexion loading of the human lumbar spine. Spine. 19(7):784-790.
Yoshizawa H, O’ Brien JP, Smith WT, Trumpcr M. 1980. The neuropathology of
intervertebral discs removed for low-back pain. Journal of Pathology. 132:95-104.
113
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Appendix
Treatment Test Procedure
PTl LI L2 Control (Bath) Standard
PT1 L2 L3 0.033% Gcncpin Standard
PTl L3 L4 0.33% Gcncpin Standard
PTl L5 SI 0.033% Gcncpin Alternate
PT2 LI L2 0.033% Genepin Alternate
P H L2 L3 Control (Bath) Standard
PT2 L3 L4 0.033% Genepin Standard
PT2 L4 L5 033% Genepin Standard
F1'3 L2 L3 0.33% Gcncpin Alternate
PT3 L3 L4 Control (Bath) Standard
PT3 L4 L5 0.033% Genepin Standard
PT3 L5 SI 033% Genepin Standard
PT4 LI L2 0.33% Genepin Standard
PT4 L3 L4 033% Genepin Alternate
PT4 L4 L5 Control (Bath) Standard
PT4 L5 SI 0.033% Genepin Standard
PT5 Lt 12 . 0.033% Genepin Standard
PT5 L2 L3 033% Genepin Standard
PT5 L3 L4 Control (Bath) Standard
PT5 L4 L5 Control (Bath) Alternate
Spare LI L2 Control (Bath) Standard
Spare L2 L3 Control (Bath) Standard
PT6 LI L2 Control (Bath) Alternate
PT6 L2 L3 Control Standard
PT6 L3 L4 033% Genepin Standard
PT6 L4 L5 Control Standard
PT7 L2 L3 Control Standard
PT7 L3 L4 033% Genepin Standard
PT7 L5 SI Control Standard
PT8 LI L2 Control Standard
PT8 L2 L3 Control Standard
PT8 L3 L4 Control Standard
PT8 L4 L5 Control Standard
Pt'8 L5 SI 0.33% Genepin Standard
PT9 LI L2 Control Standard
PT9 L2 L3 033% Genepin Standard
PT9 L3 L4 Control Standard
PT9 L4 L5 Control Standard
PT9 L5 ; Sl 033% Genepin Standard
PT10 LI L2 Control Standard
A ppendix A -l: Table o f specimen treatment selection.
114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Test Measurement Treatment Shapiro-Wilk's \ Probability
Stress Relaxation Pre-Fatigue Control 1 1 0.926 0.317
G1 1.026
G2 0.864 0.042
Post 3K. Cycles Control II 0.903 0.164
G1 1.005
G2 0.925 0.281
Post 6K Cycles Control II 0.937 0.431
G1 1.001
G2 0.972 0.877
0-6K Deltas Control II 0.930 0.358
G1 1.041
G2 0.964 0.770
Creep Pre-Fatigue Control II 0.972 0.880
G l 1.004
G2 0.924 0.271
Post 3K Cycles Control II 0.956 0.674
G l 0.954 0.318
G2 0.923 0.268
Post 6K Cycles Control II 0.976 0.855
G l 0.833 0.985
G2 0.908 0.169
0-6K Deltas Control II 0.848 0.032
G l 0.946 0.320
G2 0.938 0.409
Hardness Pre-Fatigue Control II 0.944 0.510
G l 0.967 0318
G2 0.910 0.177
Post 3K Cycles Control II 0.924 0.301
G l 0.737 0.000
G2 0.830 0.015
Post 6K Cycles Control II 0.924 0.303
G l 0.976 0321
G2 0.980 0.954
0-6K. Deltas Controls 0.875 0.072
G l 0.102 0.000
G2 0.947 0.519
A ppendix A-2: Resuits o f a Shapiro Wilk’ s test applied to each data set in order to determine
normality. Only 4 data sets are found to be Gaussian.
115
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
' S - 5
! e
* §
i t
-P T 6 L 2 /L 3 ]|
-PT 6 L 4 /L 5 li
- P T 7 L 2 /L 3 II
- P T 7 L 5 /S 1 ||
- P T 8 L 1 /L 2 I!
-P T 8 L 2 /L 3 II
- P T 8 L 3 /L 4 j!
- P T 8 L 4 /L 5 I]
- P T 9 U / L 2 it
- P T 9 L 3 /L 4 1 1
- P T 9 L 4 /L 5 !;
-P T 1 0 L A /L 5 ]|
-C ontrol Average 1 1
Pre-Fatigue Post 3K Fatigue Cycles Post 6fC Fatigue Cycles
A ppendix A-3: Pilot test stress relaxation results for the Control treatment group.
7
6
5
4
3
1
0
■ P T IL 2 /L 3
-PT2 LI / L2
-P T 2 L 3 /L 4
-PT3 U / L5
-P T 4 L 5 /S 1
-PT5 LI / L2
*0.033% Genepin Average j
1 1
Pre-Fatigue Post 3K Fatigue Cycles Post 6K Fatigue Cycles
A ppendix A-4: Pilot test stress relaxation results for the 0.033% gcnipin treatment group.
116
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
' c
' -I 4
P T 1 L 3 /L A
P T 2 L 4 /L 5
P T 3 L 2 /L 3
P T 3 L 5 /S I
PT4 LI / L2
P T 4 L 3 /L 4
P T 5 L 2 /L 3
P T 6 L 3 /L 4
PT7L3 /L 4
P T 8 L 5 /S 1
P T 9 L 2 /L 3
PT9 L5 / SI
PT10 L2 / L3
I I
-4
■0.033% Gcncpin Avenge 1 1
Pre-Fatigue Post 3K Fatigue Cycles Post 6K Fatigue Cycles
I
A ppendix A-5: Pilot test stress relaxation results for the 0.33% genipin treatment group.
•=> 0-15
fr
B
u 0.1
0.05
Post 3K Fatigue Cycles Post 6K Fatigue Cycles
■ P T 6 L 2 /L 3
-PT6 L 4 /L 5
- F 1 7 L 2 /L 3
-P T 7 L 5 /S 1
-FF 8 LI / L2
- P T 8 L 2 /L 3 i j
-P T 8 L 3 /L 4 j |
- P T 8 L 4 /L 5 'I
- P T 9 L 1 /L 2 j j
- P T 9 L 3 /L 4 |j
- P T 9 L 4 /L 5 ii
■ PT10 L4 / L5 | j
-C ontrol Average , I
A ppendix A-6: Pilot test creep deformation results for the control treatment group.
117
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
0.18
0.16
t
0.14
0.12
0.1
U 0.08
I
0.06
I
i
0.04
0.02
i
Pre-Fatigue Post 3K Fatigue Post 61C Fatigue
-P T l L 2 /L 3 j j
-P T 2 L 1 /L 2 i,
• P T 2 L 3 /L 4 |i
-PT3 L4 / L5 * j
-PT4 L5 / SI | j
• P T 5 L 1 /L 2 |!
■0.033% Genepin Average j 1
Cycles Cycles
A ppendix A-7: Pilot test creep deformation results for the 0.033% genipin treatment group.
0.2
0.15
8 -
e
o 0.1
0.05
Post 3K Fatigue
Cycles
Post 6K Fatigue
Cycles
1 3 / L4
■ !
L 4 /L 5
i
L 2 /L 3
it
L 5 /S 1
II
it
LI /L 2
n
L 3 /L 4 :!
L 2 /L 3 n
L 3 /L 4
!i
L 3 /L 4 ij
L 5 /S I
i|
L 2 /L 3 II
I I
L 5 /S 1
I I
-PT10 L 2 /L 3
“033% Gcncpin Average
A ppendix A-8: Pilot test creep deformation results for the 033% genipin treatment group.
118
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Pre-Fatigue Post 31C Fatigue Cycles Post 6K Fatigue Cycles
-PT6 L 2 /L 3
-PT6 L4 / L5
- P T 7 L 2 /L 3
-PT7 L5 / SI
-P T 8 L I / L2
- P T 8 L 2 /L 3
- P T 8 L 3 /L 4
-P T 8 L 4 / L5
- P T 9 L 1 /L 2
- P T 9 L 3/LA
-P T 9 L 4 /L 5
-PT tO L4 / L5
-C ontrol Average
A ppendix A-9: Pilot test hardness results for control treatment group.
25
20
15
10
5
0
Post 3K Fatigue Cycles Post 6K Fatigue Cycles Pte- Fatigue
t i «
■ P T 1L 2/L 3 |
■PT2L1 / L2 j
•PT2L3 flA j
-P T 3 L 4 /L 5 j
•P T 4 L 5 /S 1 |
-PT5 LI / L2 !
‘ 0.033% Genepin Average |
A ppendix A - 10: Pilot hardness results for the 0.033% genipin treatment group.
119
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3 5
30
25
J 20
.5
■ »
s
•f 15
X
to
5
0
Pie-Fadguc Poat 3K Fatigue Cycles Pose 6K Fadguc Cycles
I i
-P T 1 L 3 /L 4 |
•P T 2 L 4 /L 5 I
-P T 3 L 2 /L 3 '
-PT3 L 5 /S 1 !
-PT4 LI / L2
-P T 4 L 3 /L 4
-P T 5 L 2 /L 3
-PT6 L 3 /L 4
-P T 7 L 3 /L 4
-P T 8 L 5 /S 1
- P T 9 L 2 /L 3
-H T 9 L 5 /S 1
-P T 1 0 L 2 /L 3
■ 0.33% Gcncpin Average
A ppendix A -ll: Pilot test creep deformation results for the 033% genipin treatment gtoup.
120
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Destructive and non-destructive approaches for quantifying the effects of a collagen cross-linking reagent on the fatigue resistance of human intervertebral disc
PDF
Head injury biomechanics: Quantification of head injury measures in rear-end motor vehicle collisions
PDF
Biological materials investigation by atomic force microscope (AFM)
PDF
Effects of prenatal cocaine exposure in quantitative sleep measures in infants
PDF
Bayesian estimation using Markov chain Monte Carlo methods in pharmacokinetic system analysis
PDF
Comparisons of deconvolution algorithms in pharmacokinetic analysis
PDF
Characteristic acoustics of transmyocardial laser revascularization
PDF
A model of upper airway dynamics in obstructive sleep apnea syndrome
PDF
Dynamics of the newly formed neuromuscular synapse
PDF
Comparing signal processing methods for spectral bio-imaging
PDF
A user interface for the ADAPT II pharmacokinetic/pharmacodynamic systems analysis software under Windows 2000
PDF
Cellular kinetic models of the antiviral agent (R)-9-(2-phosphonylmethoxypropyl)adenine (PMPA)
PDF
Development of ceramic-to-metal package for BION microstimulator
PDF
A finite element model of the forefoot region of ankle foot orthoses fabricated with advanced composite materials
PDF
A fluorescence microscopy study of quantum dots as fluorescent probes for brain tumor diagnosis
PDF
Computation and validation of circulating blood volume with the indocyanine green dilution technique
PDF
Assessment of minimal model applicability to longitudinal studies
PDF
Cardiorespiratory interactions in sleep apnea: A comprehensive model
PDF
English phoneme and word recognition by nonnative English speakers as a function of spectral resolution and English experience
PDF
Propofol Effects On Eeg And Levels Of Sedation
Asset Metadata
Creator
Gray, Dean Michael
(author)
Core Title
A preliminary investigation to determine the effects of a crosslinking reagent on the fatigue resistance of the posterior annulus of the intervertebral disc
School
School of Engineering
Degree
Master of Science
Degree Program
Biomedical Engineering
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
engineering, biomedical,health sciences, rehabilitation and therapy,OAI-PMH Harvest
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
D'Argenio, David (
committee chair
), [illegible] (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c16-294465
Unique identifier
UC11341393
Identifier
1411786.pdf (filename),usctheses-c16-294465 (legacy record id)
Legacy Identifier
1411786.pdf
Dmrecord
294465
Document Type
Thesis
Rights
Gray, Dean Michael
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
engineering, biomedical
health sciences, rehabilitation and therapy