University of Southern California Dissertations and Theses

Wear of metal-on-metal artificial discs for the lumbar spine

(USC Thesis Other)

Wear of metal-on-metal artificial discs for the lumbar spine

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

WEAR OF METAL-ON-METAL ARTIFICIAL DISCS FOR THE LUMBAR SPINE

by

Jessica Lynn Lee

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirement for the Degree
DOCTOR OF PHILOSOPHY
(BIOMEDICAL ENGINEERING)

August 2007

Copyright 2007 Jessica Lynn Lee
ii
DEDICATION

To my aunt and uncle, who spoiled me like one of their own, and to my mom
and dad, who offered me every opportunity (all I had to do was pick).

iii
ACKNOWLEDGMENTS

I am grateful to the Orthopaedic Hospital Foundation (Los Angeles, CA) and
Abbott Spine (Austin, TX), for providing financial support; to my advisor, Eddie
Ebramzadeh, for challenging me as an advisor while supporting me as a mentor
and friend; to Sophia Sangiorgio, Fabrizio Billi, Bill McGarry, and Jim Alexander,
for helping out with the dirty work; to Paul Benya, Zhen Lu, Harry McKellop, Pat
Campbell, Dave Krueger, and Peter Miller, for sharing their expertise; to my
family, especially my aunt and uncle, mom and dad, for their love and faith in
me; and to my friends, especially Peter, Jarree, and Hilton, for their
encouragement and advice.

iv
TABLE OF CONTENTS

DEDICATION ............................................................................................................ II

ACKNOWLEDGMENTS .......................................................................................... III

LIST OF TABLES ...................................................................................................VII

LIST OF FIGURES....................................................................................................X

ABSTRACT............................................................................................................XIX

CHAPTER 1: INTRODUCTION ................................................................................ 1
1.1 Specific Aims ................................................................................................. 2
1.2 Organization................................................................................................... 4

CHAPTER 2: BACKGROUND.................................................................................. 5
2.1 Literature Review........................................................................................... 5
2.1.1 Artificial Hips.............................................................................................. 5
METAL-ON-POLYETHYLENE............................................................................... 6
CERAMIC-ON-CERAMIC..................................................................................... 7
METAL-ON-METAL ............................................................................................ 9
WEAR SIMULATIONS ....................................................................................... 13
2.1.2 Artificial Discs.......................................................................................... 17
CHARITÉ ........................................................................................................ 18
PRODISC ....................................................................................................... 20
MAVERICK...................................................................................................... 20
SUCCESS RATES............................................................................................ 20
WEAR SIMULATIONS ....................................................................................... 21
2.2 Problem Statements .................................................................................... 28
2.2.1 Implant Geometry.................................................................................... 29
CUP DEPTH AND THICKNESS........................................................................... 29
DIAMETER AND CLEARANCE............................................................................ 31
SURGICAL FEATURES ..................................................................................... 34
2.2.2 Biomechanics.......................................................................................... 35
MOTION AND LOAD ......................................................................................... 35
SLIDING DISTANCE ......................................................................................... 36
LUBRICATION.................................................................................................. 37
CROSSING-PATH MOTION............................................................................... 38

CHAPTER 3: MATERIALS AND METHODS ......................................................... 41
3.1 Analytical Tools ........................................................................................... 41
3.1.1 Finite Element Models............................................................................. 42
3.1.2 Hertzian Contact Equations..................................................................... 42
2.1.3 Elastohydrodynamic Lubrication Equations ............................................ 43
v
3.1.4 Slide Track Analysis................................................................................ 46
3.2 Wear Simulators........................................................................................... 48
3.2.1 Spine Wear Simulator ............................................................................. 49
3.2.2 Hip Wear Simulators ............................................................................... 55
3.3 Specimens.................................................................................................... 58
3.3.1 Artificial Hips............................................................................................ 58
3.3.2 Artificial Lumbar Discs............................................................................. 59
3.4 Testing Profiles............................................................................................ 61
3.4.1 Hip Orbital ............................................................................................... 62
3.4.2 Spine Orbital............................................................................................ 63
3.4.3 Spine Gait................................................................................................ 64
3.4.4 Spine Bends............................................................................................ 65
3.5 Schedule of Wear Tests .............................................................................. 67
3.6 Measurement Methods................................................................................ 69

CHAPTER 4: ANALYTICAL PREDICTIONS OF ARTIFICIAL DISC WEAR......... 72
4.1 Introduction.................................................................................................. 72
4.2 Methods ........................................................................................................ 75
4.3 Results.......................................................................................................... 77
4.4 Discussion and Conclusions...................................................................... 81

CHAPTER 5: DESIGN AND MEASUREMENT OF CLEARANCE OF
ARTIFICIAL DISCS ......................................................................................... 83
5.1 Introduction.................................................................................................. 83
5.2 Methods ........................................................................................................ 85
5.3 Results.......................................................................................................... 87
5.4 Discussion.................................................................................................... 90
5.5 Conclusions ................................................................................................. 95

CHAPTER 6: EFFECTS OF MOTION, LOAD, AND CUP DEPTH ON IN
VITRO WEAR OF METAL-ON-METAL BEARINGS....................................... 97
6.1 Introduction.................................................................................................. 97
6.2 Materials ....................................................................................................... 99
6.3 Methods ...................................................................................................... 102
6.4 Results........................................................................................................ 111
6.5 Discussion.................................................................................................. 120
6.6 Conclusions ............................................................................................... 130

CHAPTER 7: EFFECTS OF CARBON CONTENT, CLEARANCE, AND DISC
FEATURES ON IN VITRO WEAR OF EXPERIMENTAL ARTIFICIAL
DISCS ............................................................................................................ 133
7.1 Introduction................................................................................................ 133
7.2 Materials ..................................................................................................... 134
7.3 Methods ...................................................................................................... 138
7.4 Results........................................................................................................ 141
7.5 Discussion.................................................................................................. 151
7.6 Conclusions ............................................................................................... 160
vi
CHAPTER 8: COMPARISON OF GAIT AND HIGH-MOTION ACTIVITY ON IN
VITRO WEAR OF METAL-ON-METAL BEARINGS AND
EXPERIMENTAL ARTIFICIAL DISCS .......................................................... 162
8.1 Introduction................................................................................................ 162
8.2 Materials ..................................................................................................... 168
8.3 Methods ...................................................................................................... 171
8.4 Results........................................................................................................ 177
8.5 Discussion.................................................................................................. 185
8.6 Conclusions ............................................................................................... 193

CHAPTER 9: SUMMARY AND CONCLUSIONS ................................................. 195

REFERENCES...................................................................................................... 204

APPENDICES ....................................................................................................... 223
A. Chapter 6 Data............................................................................................. 223
B. Chapter 7 Data............................................................................................. 229
C. Chapter 8 Data............................................................................................. 235

vii
LIST OF TABLES

Table 1: Effect of Diameter on Wear ....................................................................... 12
Table 2: Effect of Clearance on Wear ..................................................................... 12
Table 3: Degrees of Freedom of Hip Wear Simulators............................................ 14
Table 4: Summary of Contemporary Hip Wear Simulators ..................................... 15
Table 5: Wear rates of metal-on-metal hips. ........................................................... 16
Table 6: Running-in periods for metal-on-metal hip wear simulations..................... 17
Table 7: Success rates and ranges of motion of TDR............................................. 21
Table 8: In vivo range of motion of the lumbar spine during gait............................. 24
Table 9: In vivo maximum compressive load of the lumbar spine during gait ......... 25
Table 10: Wear simulations of artificial lumbar discs............................................... 28
Table 11: Magnitudes of motion and load in the hip and lumbar spine during
gait.................................................................................................................... 36
Table 12: Input parameters for metal-on-metal artificial hips and discs .................. 44
Table 13: Parameters for testing profiles................................................................. 67
Table 14: Schedule of Wear Tests .......................................................................... 68
Table 15: Schedule of Measurements..................................................................... 69
Table 16: Sliding distances and aspect ratios of slide tracks produced by the hip
and lumbar spine during gait............................................................................ 80
Table 17: Cup depths and diametral clearances analyzed...................................... 86
Table 18: Contact zones for varying clearances. The contact zones were
calculated from the Hertz equations or analyzed with the finite element
model with a cup depth of either 14 or 2 mm. For the Hertz equations, the
diameter (mm) of the contact zone is indicated in the figure. All dimensions
are to scale for comparison. Colors of contour plots are scaled to the
maximum of each specimen for clarity............................................................. 89
Table 19: Schedule of tests for the current study.................................................. 106
viii
Table 20: Results of analytical tools...................................................................... 120
Table 21: Wear rates of artificial lumbar discs....................................................... 121
Table 22: Wear rates of metal-on-metal hips. ....................................................... 122
Table 23: Running-in periods for metal-on-metal hip wear simulations................. 124
Table 24: Schedule of tests for the current study.................................................. 139
Table 25: Wear rates of artificial lumbar discs....................................................... 152
Table 26: Wear rates of metal-on-metal hips. ....................................................... 153
Table 27: Running-in periods for metal-on-metal hip wear simulations................. 155
Table 28: In vivo range of motion of the lumbar spine during gait......................... 166
Table 29: In vivo maximum compressive load of the lumbar spine during gait ..... 167
Table 30: Schedule of tests for the current study.................................................. 173
Table 31: Results of analytical tools...................................................................... 184
Table 32: Wear rates of artificial lumbar discs....................................................... 186
Table 33: Wear rates of metal-on-metal hips. ....................................................... 187
Table 34: Running-in periods for metal-on-metal hip wear simulations................. 189
Table 35: Ball volumetric wear. ............................................................................. 223
Table 36: Cup volumetric wear.............................................................................. 224
Table 37: Total volumetric wear rate. .................................................................... 225
Table 38: Lubrication gap...................................................................................... 226
Table 39: Mediolateral average surface roughness. ............................................. 227
Table 40: Anteroposterior average surface roughness. ........................................ 228
Table 41: Ball volumetric wear. ............................................................................. 229
Table 42: Cup volumetric wear.............................................................................. 230
Table 43: Total volumetric wear rate ..................................................................... 231
Table 44: Lubrication gap...................................................................................... 232
ix
Table 45: Mediolateral average surface roughness. ............................................. 233
Table 46: Anteroposterior average surface roughness. ........................................ 234
Table 47: Ball volumetric wear. ............................................................................. 235
Table 48: Cup volumetric wear.............................................................................. 236
Table 49: Total volumetric wear rate. .................................................................... 236
Table 50: Lubrication gap...................................................................................... 237
Table 51: Mediolateral average surface roughness. ............................................. 237
Table 52: Anteroposterior average surface roughness. ........................................ 238

x
LIST OF FIGURES

Figure 1: In vivo maximum range of motion of the spine......................................... 23
Figure 2: In vivo frequencies and phasing of motion and load during gait. The
solid line represents flexion/extension, the dashed line represents lateral
bending, the dashdot line represents axial rotation, and the dotted line
represents the compressive load. .................................................................... 26
Figure 3: Motion and load waveforms illustrating differences in the (A) hip and
(B) lumbar spine during gait. The solid line represents flexion/extension,
the dashed line represents abduction/adduction of the hip or lateral
bending of the lumbar spine, the dashdot line represents axial rotation, and
the dotted line represents the compressive load.............................................. 36
Figure 4: Custom spine wear simulator. The motor and speed reducer in the
cranial component provide flexion/extension, while those in the the caudal
component provide axial rotation. .................................................................... 51
Figure 5: Custom spine wear simulator. The pinion gears transmit the motion to
three separate stations..................................................................................... 51
Figure 6: Custom spine wear simulator. The cam and linear bearing convert
the planar motion to a pure back-and-forth rotation that oscillates at an
amplitude determined by the offset of the hole in the cam and a frequency
determined by the velocity of the motor shaft................................................... 52
Figure 7: Custom spine wear simulator. The cam and linear bearing convert
the planar motion to a pure back-and-forth rotation that oscillates at an
amplitude determined by the offset of the hole in the cam and a frequency
determined by the velocity of the motor shaft................................................... 52
Figure 8: Custom spine wear simulator. Axial rotation is provided by a motor
and speed reducer in the caudal aspect of the wear simulator, using the
same mechanical concepts as for flexion/extension. ....................................... 53
Figure 9: Frequency and phasing of motion in the orbital wear simulator. The
solid line represents flexion/extension, and the dashed line represents
abduction/adduction of the hip or lateral bending of the lumbar spine............. 56
Figure 10: Hip wear simulator with a total range of motion of A) 46°, as
manufactured, and B) modified with an adapter block to 8.5°.......................... 57
Figure 11: Specimen ball (left), cup with a depth of 14 mm (middle), and cup
with a depth of 2 mm (right).............................................................................. 59
xi
Figure 12: Specimen cup with a depth of 14 mm (left) and 2 mm (right)................. 59
Figure 13: Caudal endplate (left), ball spacer (middle), and cranial cup (right)....... 61
Figure 14: Keeled (upper-left) and un-keeled (upper-right) cups, and notched
(lower-left) and un-notched (lower-right) cups.................................................. 61
Figure 15: Hip Orbital A) motion and load profile and B) slide tracks produced.
The solid line represents flexion/extension, the dashed line represents
abduction/adduction, the dashdot line represents axial rotation, and the
dotted line represents the compressive load.................................................... 63
Figure 16: Spine Orbital A) motion and load profile and B) slide tracks
produced. The solid line represents flexion/extension, the dashed line
represents lateral bending, the dashdot line represents axial rotation, and
the dotted line represents the compressive load.............................................. 64
Figure 17: Spine Gait A) motion and load profile and B) slide tracks produced.
The solid line represents flexion/extension, the dashed line represents
lateral bending, the dashdot line represents axial rotation, and the dotted
line represents the compressive load............................................................... 65
Figure 18: Spine Bends A) motion and load profile and B) slide tracks produced
at phase angles of pi/4, pi/2, 3pi/4, and pi. The solid line represents
flexion/extension, the dashed line represents lateral bending, the dashdot
line represents axial rotation, and the dotted line represents the
compressive load. The phasing of the motion was not synchronized and
therefore not necessarily as shown in the figure.............................................. 66
Figure 19: Centrifugation tube with pre-centrifugation layers labeled. After
centrifugation, the metal particles remained isolated at the bottom of the
tube, while the rest of the sample remained above the CsFm gradient. A
visible layer of white protein at the top of the tube was observed.................... 71
Figure 20: EDAX image of elements present in the sample. The larger peaks
corresponded to the aluminum stub, the polycarbonate filter, and the gold-
sputtering. Smaller, almost indistinguishable peaks corresponded to the
Co and Cr of the metal particles....................................................................... 71
Figure 21: Cup dimensions of A) typically hemispherical artificial hips and B)
shallower artificial discs.................................................................................... 73
Figure 22: Maximum contact pressures of artificial hips and discs. ........................ 78
Figure 23: Lambda ratios and lubrication regimes for artificial hips and artificial
lumbar discs during gait. .................................................................................. 79
xii
Figure 24: Motion waveforms and slide tracks produced for the (A) hip and (B)
lumbar spine during gait. The solid line represents flexion/extension, the
dashed line represents abduction/adduction of the hip or lateral bending of
the lumbar spine, and the dotted line represents axial rotation........................ 80
Figure 25: Maximum contact pressures for varying clearances. The contact
pressures were calculated from the Hertz equations or analyzed with the
finite element model with a cup depth of either 14 or 2 mm. The maximum
contact pressures of the 2-mm model were substantially larger than those
of the 14-mm model; contact pressures from both finite element models
were larger than predictions from the Hertz equations. The contact
pressures for both cup depths increased with increasing clearance until
100 μm for the 14-mm model and 80 μm for the 2-mm model, at which
point the contact pressures decreased; this decrease was in contrast to the
steadily increasing trend predicted by the Hertz equations.............................. 88
Figure 26: Cup radius calculated with four different methods. The nominal
radius is indicated with a dotted line................................................................. 90
Figure 27: Coordinate Axes and Reference Points ................................................. 93
Figure 28: Definition of Radial Angle ( α) and Lubrication gap (L)............................ 93
Figure 29: Comparison of non-surface-fit and surface-fit methods ......................... 95
Figure 30: Specimen ball (left), cup with a depth of 14 mm (middle), and cup
with a depth of 2 mm (right).............................................................................. 99
Figure 31: Specimen cup with a depth of 14 mm (left) and 2 mm (right)............... 100
Figure 32: Hip wear simulator with a total range of motion of A) 46° and B) 8.5°,
as determined by the angle of the platform.................................................... 101
Figure 33: Spine wear simulator constructed by the Biomechanics Laboratory at
Orthopaedic Hospital (Los Angeles, CA). The cranial component
experiences flexion/extension, while the caudal component experiences
axial rotation................................................................................................... 102
Figure 34: Hip Orbital A) motion and load profile and B) slide tracks produced.
For the motion and load profile, the solid line represents flexion/extension,
the dashed line represents abduction/adduction, the dashdot line
represents axial rotation, and the dotted line represents the compressive
load................................................................................................................. 104
xiii
Figure 35: Spine Orbital A) motion and load profile and B) slide tracks
produced. For the motion and load profile, the solid line represents
flexion/extension, the dashed line represents lateral bending, the dashdot
line represents axial rotation, and the dotted line represents the
compressive load. .......................................................................................... 105
Figure 36: Spine Gait A) motion and load profile and B) slide tracks produced.
For the motion and load profile, the solid line represents flexion/extension,
the dashed line represents lateral bending, the dashdot line represents
axial rotation, and the dotted line represents the compressive load. ............. 106
Figure 37: Average volumetric wear and standard deviations for Tests 1 and 2.
The circle data points represent the Hip Orbital specimens, while the
square data points represent the Spine Orbital specimens. The solid lines
represent the balls, while the dashed lines represent the cups. For all
specimens, the cups wore more than the balls. P-values for all tests are
listed at the top of the plot; the top row represents p-values between the
ball and cup for Test 1, while the bottom row represents those of Test 2...... 111
Figure 38: Average volumetric wear and standard deviations for Tests 2 and 3.
The circle data points represent the Spine Orbital specimens, while the
square data points represent the Spine Gait specimens. The solid lines
represent the balls, while the dashed lines represent the cups. For all
specimens, the cups wore more than the balls. P-values for all tests are
listed at the top of the plot; the top row represents p-values between the
ball and cup for Test 2, while the bottom row represents those of Test 3...... 112
Figure 39: Average volumetric wear and standard deviations for Tests 3 and 4.
The circle data points represent the 14-mm specimens, while the square
data points represent the 2-mm specimens. The solid lines represent the
balls, while the dashed lines represent the cups. For all specimens, the
cups wore more than the balls. P-values for all tests are listed at the top of
the plot; the top row represents p-values between the ball and cup for Test
3, while the bottom row represents those of Test 4........................................ 113
Figure 40: Average volumetric wear rates and standard deviations. The Spine
Orbital specimens had a higher wear rate than the Hip Orbital specimens,
the Spine Gait specimens had a lower wear rate than the Spine Orbital
specimens, and the 2-mm specimens had a higher wear rate than the 14-
mm specimens. P-values between tests are listed at the top of the plot; the
top row represents p-values Tests 1 and 2, the middle row represents
those between Test 2 and 3, and the bottom row represents those between
Tests 3 and 4.................................................................................................. 114
Figure 41: Definition of angles. Lubrication gap area is shown in gray. For the
artificial discs of the current study, the lubrication gap was calculated at
60°.................................................................................................................. 115
xiv
Figure 42: Average lubrication gaps and standard deviations. Clearance was
extrapolated from the lubrication gap values and is represented with the y-
axis on the right side. The lubrication gaps of the Spine Orbital and Spine
Gait 14-mm specimens stayed approximately constant, while those of the
2-mm specimens increased and were ultimately higher than those of the
14-mm specimens. P-values between tests are listed at the top of the plot;
the top row represents p-values Test 2 and 3, and the bottom row
represents those between Tests 3 and 4. ...................................................... 116
Figure 43: Average mediolateral surface roughness (Ra) and standard
deviations. All specimens stayed approximately constant, and the only
discernible difference was the higher surface roughness of the Spine Gait
14-mm specimens compared to the 2-mm specimens. P-values between
tests are listed at the top of the plot; the top row represents p-values Test 2
and 3, and the bottom row represents those between Tests 3 and 4. ........... 117
Figure 44: Average anteroposterior surface roughness (Ra) and standard
deviations. All specimens stayed approximately constant, and the only
discernible difference was the higher surface roughness of the Spine Gait
14-mm specimens compared to the 2-mm specimens. P-values between
tests are listed at the top of the plot; the top row represents p-values Test 2
and 3, and the bottom row represents those between Tests 3 and 4. ........... 118
Figure 45: Particle lengths for all tests. The particle lengths were all
approximately 120 nm, with little difference between the three tests............. 119
Figure 46: SEM of metal particles at a magnification of 25,000. The filter has
30 nm holes.................................................................................................... 119
Figure 47: Worn surfaces of the ball after 10
6
cycles for A) Test 2, B) Test 3,
and C) Test 4. For Test 2, the slide tracks were circular, and for Tests 2
and 3, the slide tracks were linear.................................................................. 120
Figure 48: Caudal endplate (left), ball spacer (middle), and cranial cup (right)..... 136
Figure 49: Keeled (upper-left) and un-keeled (upper-right) cups, and notched
(lower-left) and un-notched (lower-right) cups................................................ 137
Figure 50: Spine wear simulator constructed by the Biomechanics Laboratory at
Orthopaedic Hospital (Los Angeles, CA). The cranial cup component
experiences flexion/extension, while the caudal ball component
experiences axial rotation............................................................................... 138
xv
Figure 51: Volumetric wear for Tests 1 and 2a, First- vs. Second-Generation
Specimens. The circle data points represent the first-generation
specimens, while the square data points represent the second-generation
specimens. The solid lines represent the balls, while the dashed lines
represent the cups. The trend for all data points is linear. For the first-
generation specimens, the ball wore significantly more than the cup
(p<0.05 for all intervals). For the second-generation specimens, the cup
wore more than the ball, but not significantly (p>0.05 for all intervals)........... 142
Figure 52: Volumetric wear for Tests 2a and 2b, Notched vs. Un-Notched Cups.
The circle data points represent the notched cup specimens, while the
square data points represent the un-notched cup specimens. The solid
lines represent the balls, while the dashed lines represent the cups. The
trend for all data points is linear. For all specimens, the cups wore more
than the balls; this difference was not significant for the notched specimens
(p>0.05 for all intervals), but significant for the un-notched specimens
(p<0.05 for all intervals).................................................................................. 143
Figure 53: Average volumetric wear rates and standard deviations. The first-
generation specimens had a higher wear rate than all second-generation
specimens. Within the second-generation specimens, the notched cup
specimens had a higher rate than the un-notched cup specimens. The star
(*) indicates p-values less than 0.05 between Tests 1 and 2a, and the
cross (+) indicates p-values less than 0.05 between Tests 2a and 2b........... 144
Figure 54: Definition of angles. Lubrication gap area is shown in gray. For the
artificial discs of the current study, the lubrication gap was calculated at
60°.................................................................................................................. 145
Figure 55: Average lubrication gaps and standard deviations. Clearance was
extrapolated from the lubrication gap values and is represented with the y-
axis on the right side. The first-generation specimens had a decreasing
lubrication gap, opposite to the increasing lubrication gap of all second-
generation specimens, and also had a larger initial lubrication gap. Within
the second-generation specimens, the notched cup specimens had an
increasing lubrication gap, similar in trend and magnitude to the un-
notched cup specimens. The star (*) indicates p-values less than 0.05
between Tests 1 and 2a, and the cross (+) indicates p-values less than
0.05 between Tests 2a and 2b....................................................................... 146
Figure 56: Average mediolateral surface roughness (Ra) and standard
deviations. The first-generation specimens had a lower maximum Ra than
all second-generation specimens. Within the second-generation
specimens, the notched cup specimens had a lower maximum Ra than the
un-notched cup specimens. The star (*) indicates p-values less than 0.05
between Tests 1 and 2a, and the cross (+) indicates p-values less than
0.05 between Tests 2a and 2b....................................................................... 147
xvi
Figure 57: Average anteroposterior surface roughness (Ra) and standard
deviations. The first-generation specimens had a lower maximum Ra than
all second-generation specimens. Within the second-generation
specimens, the notched cup specimens had a lower maximum Ra than the
un-notched cup specimens. The star (*) indicates p-values less than 0.05
between Tests 1 and 2a, and the cross (+) indicates p-values less than
0.05 between Tests 2a and 2b....................................................................... 148
Figure 58: Metal pellets for A) Test 1, B) Test 2a, and C) Test 2b. Test 1 has
more visible metal particles than Test 2a, which has more than Test 2b....... 148
Figure 59: Particle length for all tests. The particle lengths were all
approximately 120 nm, with little difference between the three tests............. 149
Figure 60: SEM of metal particles at a magnification of 25,000. The filter has
30 nm holes.................................................................................................... 150
Figure 61: Worn surfaces of the A) ball and B) cup after 1.0 x 10
6
cycles. The
surfaces of the specimens showed plastic deformation and deep scratches
in the direction of motion, which are common signs of abrasive or fatigue
wear................................................................................................................ 150
Figure 62: SEM of the ball, showing A) gouging and B) pitting, which are
common signs of abrasive or fatigue wear..................................................... 151
Figure 63: In vivo maximum range of motion of the spine..................................... 166
Figure 64: In vivo frequencies and phasing of motion and load during gait. The
solid line represents flexion/extension, the dashed line represents lateral
bending, the dashdot line represents axial rotation, and the dotted line
represents the compressive load. .................................................................. 168
Figure 65: Disc B specimens: caudal endplate (left), ball spacer (middle), and
cranial cup (right)............................................................................................ 169
Figure 66: Disc B specimens: ball (left) and cup (right)......................................... 170
Figure 67: Spine Gait profile. The solid line represents flexion/extension, the
dashed line represents lateral bending, the dashdot line represents axial
rotation, and the dotted line represents the compressive load....................... 172
Figure 68: Spine Bends profile. The solid line represents flexion/extension, the
dashed line represents lateral bending, the dashdot line represents axial
rotation, and the dotted line represents the compressive load. The phasing
of the motion was not synchronized and therefore not necessarily as
shown in the figure. ........................................................................................ 173
xvii
Figure 69: Volumetric wear for Tests 1 and 2, Disc A specimens. The circle
data points represent the Spine Gait specimens, while the square data
points represent the Spine Bends specimens. The solid lines represent the
balls, while the dashed lines represent the cups. For the Spine Gait
specimens, the ball wore marginally more than the cup. For the Spine
Bends specimens, the cup wore substantially more than the cup. P-values
for all tests are listed at the top of the plot; the top row represents p-values
between the ball and cup for Test 1, and the bottom row represents those
for Test 2. ....................................................................................................... 178
Figure 70: Volumetric wear for Tests 3 and 4, Disc B specimens. The circle
data points represent the Spine Gait specimens, while the square data
points represent the Spine Bends specimens. The solid lines represent the
balls, while the dashed lines represent the cups. For all specimens, the
cup wore substantially more than the cup. P-values for all tests are listed
at the top of the plot; the top row represents p-values between the ball and
cup for Test 3, and the bottom row represents p-values between the ball
and cup for Test 4. ......................................................................................... 179
Figure 71: Volumetric wear rates. For both Disc A and Disc B specimens, the
Spine Bends specimens had a higher wear rate than the Spine Gait
specimens. P-values between tests are listed at the top of the plot; the top
row represents p-values between Tests 1 and 2, and the bottom row
represents p-values between Tests 3 and 4. ................................................. 180
Figure 72: Definition of angles. Lubrication gap area is shown in gray. For both
Disc A and Disc B specimens, the lubrication gap was calculated at 60°...... 181
Figure 73: Lubrication gap. Clearance was extrapolated from the lubrication
gap values and is represented on the y-axis on the right side. For both
Disc A and Disc B specimens, the lubrication gaps of all specimens
generally increased. For both Disc A and Disc B specimens, the
lubrication gaps were larger for the Spine Bends specimens than for the
Spine Gait specimens. P-values between tests are listed at the top of the
plot; the top row represents p-values between Tests 1 and 2, and the
bottom row represents p-values between Tests 3 and 4................................ 182
Figure 74: Mediolateral average surface roughness. For Disc A specimens, the
surface roughness for the Spine Bends specimens was substantially higher
than for the Spine Gait specimens, while for Disc B specimens, the surface
roughness for the Spine Gait specimens was marginally higher than for the
Spine Bends specimens. P-values between tests are listed at the top of
the plot; the top row represents p-values between Tests 1 and 2, and the
bottom row represents p-values between Tests 3 and 4................................ 183
xviii
Figure 75: Anteroposterior average surface roughness. For Disc A specimens,
the surface roughness for the Spine Bends specimens was substantially
higher than for the Spine Gait specimens, while for Disc B specimens, the
surface roughness for the Spine Gait specimens was marginally higher
than for the Spine Bends specimens. P-values between tests are listed at
the top of the plot; the top row represents p-values between Tests 1 and 2,
and the bottom row represents p-values between Tests 3 and 4................... 184

xix
ABSTRACT

Metal-on-metal bearings have been established in total hip replacements but,
more recently, have been developed with modifications for artificial discs. The
significance of these modifications is not known. In this study, analytical tools
were used to predict the effects of cup depth, clearance, motion, and load on
wear of metal-on-metal artificial lumbar discs. Then, laboratory wear simulators
were developed. The first series of wear simulations determined the effects of
the differences between artificial hips and discs on wear by comparing the
effects of motion, load, and cup design; the second compared the effects of
metallurgy and cup design; and the third compared the effects of different
motion and load profiles on wear of two designs of discs. From the analytical
tools, the larger maximum contact pressure, lower lubrication, and smaller
cross-path angles predicted more wear for metal-on-metal discs compared to
hips; however, the smaller sliding distance predicted less wear. The majority of
analyses predicted more wear of artificial discs than hips, emphasizing the
importance of considering the effects of dimensional and biomechanical
differences on wear. From the first series of wear simulations, decreasing the
magnitudes of motion and load from hip to lumbar spine increased the wear;
changing the phasing of motion from hip to lumbar spine reduced the wear; and
decreasing the cup depth from artificial hip to disc increased the wear. The
overall increase in wear supported the predictions from the analytical tools that
artificial discs wear more than hips. From the second series of wear

xx
simulations, changing the carbon content of the ball from low to high, eliminating
the anteroposterior keels, and decreasing the initial clearance reduced the wear;
removing the surgical notch further reduced the wear. These reductions in wear
emphasized the importance of considering the effects of individual design
variables on wear. In the third series of wear simulations, changing the motion
and load profile from gait to high-motion activities increased the wear for both
artificial discs; however, the effect on the ball-to-cup wear ratios was different for
each design, emphasizing the importance of considering the effects of the
testing profile in wear simulations.

1
CHAPTER 1: INTRODUCTION

In the United States alone, low-back pain is the reason for 19 million visits to
doctors’ offices per year, exceeded only by the common cold in terms of the
frequency of complaints that are heard by primary care physicians, and has a
total cost of $100 – 200 billion per year
120
. The most common cause of chronic
low back pain is degenerative disc disease. Until recently, the standard surgical
treatment for a degenerated disc was spinal fusion using motion-preventing
internal fixation devices. Unfortunately, spinal fusion alters the spine
biomechanics and has been associated with accelerated degeneration of
adjacent-level discs
37,65,109,110,132,206
.

Artificial discs were introduced to restore the natural biomechanics of the spine,
and to reduce the incidence of adjacent-level disc degeneration. Artificial
lumbar discs are motion-preserving devices that are designed to replace injured
or diseased intervertebral discs that are painful because of decreased disc
height
1,105,124,145,162
or pressure
2,4,62,69,162,167
. One major disadvantage of
restoration of motion is wear of the implant materials, as observed in varying
amounts with virtually all articulating joint replacements. Most notably, in total
hip replacements, it has long been recognized that wear debris from implants
can cause biological reactions
38-40,63,78,80,87,88,90,91,112,148,161,169,191,230
which, in turn,
may lead to osteolysis
98,121,219,224,230
and, eventually, implant loosening.

2
Artificial disc designs resemble artificial hip designs. Like artificial hips, many
articulating artificial discs are ball-and-cup joints. More specifically, artificial
discs incorporate material combinations that, after decades of extensive
research, were developed to be adequately wear-resistant in artificial hips.
However, artificial hips and artificial lumbar discs have substantial differences in
implant geometry, as well as in the motion and load they experience in vivo,
which likely have substantial effects on material wear. Thus, the variables that
were considered critical in the wear of artificial hips, such as the diameter of the
ball, the clearance between the ball and cup, the carbon content of the metal
alloy, or the relative path of motion of the ball and cup, must be re-assessed for
the implant geometry and motion and load of artificial discs. Furthermore, wear
studies specific to artificial lumbar discs are necessary to address other issues
that were not relevant and, therefore, not addressed in hip wear studies.
Examples of these include the addition of keels for fixation of the implant into the
vertebrae or the inclusion of a surgical notch to prevent over-distraction of the
disc space.

1.1 Specific Aims

The current study was performed at Orthopaedic Hospital (Los Angeles, CA), in
collaboration with Abbott Spine (Austin, TX). The specific aims of the current
study were:

3
1. To apply equations for Hertzian contact, lubrication, and slide tracks to
predict or explain wear of metal-on-metal artificial lumbar discs.

2. To create three-dimensional finite element models to relate the contact areas
and contact stresses at the motion interfaces of artificial disc components to
wear.

3. To design and build a machine capable of simulating wear of artificial lumbar
discs by applying dynamic load and motion typical of the lumbar spine during
gait.

4. To modify a machine intended for hip wear simulation to simulate wear of
artificial lumbar discs by applying motion and load typical of the lumbar spine
during gait.

5. To determine the effects of the differences between hips and lumbar discs
on wear of metal-on-metal joints by comparing the effects of motion, load,
and cup design on wear.

6. To compare the effects of metallurgy and cup design on wear of
experimental metal-on-metal artificial lumbar discs, in order to facilitate the
design of artificial lumbar discs.

4
7. To compare the effects of different motion and load profiles on wear of
metal-on-metal artificial lumbar discs, in order to optimize the wear test
parameters for evaluating artificial lumbar discs.

1.2 Organization

The background chapter consists of the literature review and problem
statements formulated from the literature review. The materials and methods
chapter details the analytical tools, wear simulators, specimens, testing profiles,
schedule of wear tests, and measurements methods used in the current study.
The next few chapters present results of the current study:

• Analytical Predictions of Artificial Disc Wear
• Design and Measurement of Clearance of Artificial Discs
• Effects of Motion, Load, and Cup Depth on In Vitro Wear of Metal-on-Metal
Bearings
• Effects of Carbon Content, Clearance, and Disc Features on In Vitro Wear
of Experimental Artificial Discs
• Comparison of Gait and High-Motion Activity on In Vitro Wear of Metal-on-
Metal Bearings and Experimental Artificial Discs

The last chapter consists of a summary of the current study, as well as the
overall impact of the current study on the orthopaedic community.

5
CHAPTER 2: BACKGROUND

To accomplish the specific aims, literature on artificial hips and discs was
reviewed. Problem statements were formulated by comparing and contrasting
artificial hips to discs, citing relevant studies, and emphasizing the limitations of
the studies in establishing wear of metal-on-metal lumbar discs.

2.1 Literature Review

2.1.1 Artificial Hips

The design of articulating artificial discs has relied heavily on that of artificial
hips, particularly in the use of material combinations previously established in
the field of hip arthroplasty. The most widely-used material combination for
artificial hips is a cobalt chromium molybdenum alloy (CoCrMo) surface
articulating against an ultra-high molecular weight polyethylene (UHMWPE)
surface, a material combination commonly known as metal-on-polyethylene.
Ceramic-on-ceramic and metal-on-metal are other widely-used material
combinations; each material combination has unique wear characteristics.

6
METAL-ON-POLYETHYLENE

For artificial hips, metal-on-polyethylene became preferred over metal-on-metal
and ceramic-on-ceramic in the 1970s because of the success of the Charnley
prosthesis, in combination with early reports of impingement or equatorial
clamping of the metal-on-metal McKee-Farrar
3
, and impingement or fracture of
the ceramic-on-ceramic Autophor and Xenophor
21
.

However, with time, metal-on-polyethylene hips were observed to loosen in vivo,
which was attributed in part to polyethylene wear. Polyethylene wear has been
reported in the range of 10 – 95 mm
3
/10
6
cycles
60,68,113,228
, but is typically
between 55 – 95 mm
3
/10
6
cycles
228
, with particle lengths typically in the range of
150 – 1000 nm
7,12,19,39,140,209,220
. Polyethylene wear has been shown to cause
biological reactions
27,30,38-
40,63,78,80,83,87,88,90,91,93,99,100,111,112,139,148,161,166,169,191,192,199,230,235
and, potentially,
osteolysis
98,121,219,224,230
. Because of this, modern alternative bearing surfaces
were developed. Metal-on-metal and ceramic-on-ceramic were revisited due to
improved manufacturing methods, and metal-on-crosslinked polyethylene was
improved upon due to the past experience and success of metal-on-
polyethylene.

Polyethylene can be crosslinked under radiative, thermal, or chemical processes
to decrease the wear rate by 2 – 40 times
45,61,66,67,101,102,106,119,136,144,163,173,176
.

7
Crosslinking polyethylene improved the success rates of metal-on-polyethylene
hips; however, the method of crosslinking may potentially weaken the material.
Clinical explants of crosslinked polyethylene showed surface cracking or
scratching
14,155,156,170
, and impingement tests revealed significantly more surface
damage in the form of pitting, delamination, and crack formation in cups with
higher levels of irradiation
104
. This damage may have been a result of the
decreased fracture toughness
81,82,86
, decreased elongation-to-break
234
,
decreased deformation resistance, and decreased stress and strain at failure
129

that has been observed in crosslinked polyethylene. Despite these
considerations, improvements in crosslinking methods have made metal-on-
crosslinked polyethylene highly successful, and the predominant material
combination for artificial hips today. The Swedish Hip Registry reported low
failure rates of 3% after 5 years, and 7% after 10 years for metal-on-crosslinked
polyethylene hips
118
.

CERAMIC-ON-CERAMIC

For artificial hips, ceramic-on-ceramic bearing surfaces are attractive because of
the high strength and low wear, which has been shown to be 50 – 1000 times
less than that of metal-on-uncrosslinked polyethylene hips
34,68
, resulting in
typical wear rates of less than 0.1 mm
3
/10
6
cycles. The failure rate of ceramic-
on-ceramic hips was reported as 12 of 83 implants (14.5%) in a study with a
12.3-year follow-up period
79
. As a comparison, the Swedish Hip Registry

8
reported failure rates of 10% after 12 years for metal-on-crosslinked
polyethylene hips
118
. However, constant improvements in manufacturing
methods have made ceramic-on-ceramic even more successful. More recently,
the failure rate of ceramic-on-ceramic hips was reported as 6 of 357 implants
(1.7%) in a study with a 3.9-year follow-up period
164
. As a comparison, the
Swedish Hip Registry reported failure rates of 3% after 4 years for metal-on-
crosslinked polyethylene hips
118
.

For artificial hips, another advantage of ceramic-on-ceramic is the inertness and
size of the particles, which at 1 – 90 nm
100,209,211
are smaller than polyethylene
particles. The biological reactions of tissues containing ceramic particles have
been observed to be less than that of tissues containing polyethylene or metal
particles, in the form of smaller inflammatory response and bone resorption,
smaller percentage of apoptotic cells, smaller number of macrophages and giant
cells, and increased cell viability
83,220,230
.

For ceramic-on-ceramic hips, cases of higher wear rates and particle sizes have
been observed. Stripe wear is potentially caused by impingement due to
activities requiring high flexion and load, such as rising from a sitting position.
One study retrieved 16 ceramic-on-ceramic hips from a series of 1588 hip
arthroplasties, and found stripe wear in 11 of 16 implants
227
. These implants
were not retrieved for bearing failure, demonstrating that stripe wear does not
necessarily lead to failure. However, potential effects of stripe wear include

9
wear of the stripe zone up to 35 times that of the unimpinged surfaces
142,160,236
,
larger particles in the range of 50 – 10,000 nm
100,211
, grain pull-out regions
193
, an
increase in surface roughness, and intragranular fractures
159,201
. Even with
stripe wear, ceramic-on-ceramic hips can be highly successful.

The fracture of ceramic-on-ceramic hips is also of concern because of the
potentially catastrophic resulting damage and particulate debris
21
. Fracture was
shown by impact tests to occur at a threshold force of 12 kN
141
, which is quite
high but potentially catastrophic. In the studies mentioned previously, fracture of
ceramic-on-ceramic hips was observed in 3 of 83 (3.6%) implants after 12.3
years; however, constant improvements in manufacturing methods have made
ceramic-on-ceramic highly successful. More recently, fracture of ceramic-on-
ceramic hips was observed in 6 of 357 (1.7%) implants after 3.9 years. In a
larger study of 2963 ceramic-on-ceramic hips with a follow-up of 11 years, the
fracture rate was substantially lower, ranging from 0.06 – 0.4%, depending on
design
76
.

METAL-ON-METAL

For artificial hips, metal-on-metal bearing surfaces are attractive because of the
high strength, self-polishing mechanism of the bearing surfaces
187,210
, and low
wear, which has been shown to be 9 – 400 times less than that of metal-on-
uncrosslinked polyethylene hips
5,34,58,68,85,113,174
, resulting in wear rates of

10
typically less than 1.0 mm
3
/10
6
cycles
31,32,56,57,72,73,85,150,151,187,196,229,232
. Recently,
the failure rate of metal-on-metal hips was reported as 34 of 640 implants (5.3%)
in a study with a 7.1-year follow-up period
153
and 1 of 106 implants (0.9%) in a
study with a 6.4-year follow-up period
177
. As a comparison, the Swedish Hip
Registry reported failure rates of 4% after 7 years for metal-on-crosslinked
polyethylene hips
118
.

For artificial hips, a disadvantage of metal-on-metal is the reactivity of the
particles, which at 30 – 300 nm
25,26,28,29,50,209
are smaller than polyethylene
particles but larger than ceramic particles. The biological reactions of tissues
containing metal particles have been observed to be greater than that of tissues
containing polyethylene particles, in the form of increased cytokine release by
macrophages
191
, increased percentage of apoptotic cells
199
, and increased
fibrogenesis stimulation
192
. Additionally, metal ion levels in the blood and urine
of patients with metal-on-metal hips as compared to those of patients without
artificial hips have been shown to be 2 – 50 times higher in the blood, 10 – 300
times higher in the urine, and up to 50 times higher in the lung, kidney, liver, and
spleen
48,59,146,152
. The effect of these elevated ion levels is not known but
remains of concern.

For metal-on-metal hips, cases of higher wear rates have been attributed to
variables such as carbon content, diameter, and clearance, but were mainly
observed with wear simulations testing these variables.

11

Laboratory hip joint wear simulations have shown that low carbon-content alloys
wear more than high carbon-content alloys in metal-on-metal
hips
56,73,103,172,186,198,210
. Low carbon-content alloys comprise a single-phase
structure, with larger grains and smaller carbides than those in high carbon-
content alloys. In contrast, high carbon-content alloys demonstrate a biphasic
structure, with smaller grains of the alloy surrounded by larger scratch-resistant
carbides
210,229
. One study showed that the ball wore 4 times more than the cup
for low carbon-content ball specimens, but also that the cup wore 4 times more
than the ball for high carbon-content ball specimens
73
. Still, low carbon-content
balls are frequently used with high carbon-content cups because the total wear
is comparable to that of an implant with both high carbon-content components,
but with most of the damage inflicted on the ball, which is modular and more
easily replaced than the cup.

Hip wear simulator studies have shown that smaller ball diameters increase
wear
57,151,171,196
; studies quantifying this effect
57,150,196
are shown in Table 1.
Because of the lower wear rates associated with larger diameters, metal-on-
metal hips usually have diameters larger than 28 mm.

12
Table 1: Effect of Diameter on Wear
Diameter Range (mm) Wear Range
McKellop et al (1996) 35 – 42 3.3 – 1.5 mm
3
/10
6
cycles
Smith et al (2001) 16 – 28 4.9 – 0.5 mm
3
/10
6
cycles
Dowson et al (2004) 54 – 54.5 3.3 – 0.8 mm
3

Hip wear simulator studies have also shown that larger clearance, which is the
difference in diameters of the ball and cup, increases wear of metal-on-metal
total hip replacements
31,32,56,57,103,114,150,151,171,172,186,187
; studies quantifying this
effect
31,57,150,171,172,187
are shown in Table 2.

Table 2: Effect of Clearance on Wear
Clearance Range (µm) Wear Range
McKellop et al (1996) 125 – 400 0.1 – 2.6 mm
3
/10
6
cycles
Chan et al (1999) 30 – 110 0.2 – 1.5 mm
3
/10
6
cycles
Scholes et al (2001) 44 – 80 0.1 – 0.3 mm
3
/10
6
cycles
Dowson et al (2004) 105 – 143 2.3 to 3.5 mm
3

Rieker et al (2004) 70 – 155 2.5 – 27.5 µm/year
Rieker et al (2005) 60 – 275 12.0 – 52.0 µm

Some studies have reported that clearances that are too small lead to equatorial
clamping
114,151,172,186
; thus, an optimum radial clearance for metal-on-metal total
hip replacements is typically on the order of 50 μm. In fact, one of the reasons
for abandoning early versions of metal-on-metal hips developed and implanted
in the 1960s and 1970s, such as the McKee-Farrar, was early reports of
equatorial clamping
3
, presumably due to improper clearances, but also perhaps
due to the acetabular cup thickness
237
.

13
Renewed interest in metal-on-metal hips was sparked by clinical papers
reporting that McKee-Farrar prostheses had favorable long-term wear
properties, presumably from the implants that chanced upon optimum
clearances
15,35,46,47,107,115,117,150,183,184,238
. The success of more recent versions of
metal-on-metal bearing surfaces may be attributed in part to the improvement in
manufacturing methods. For example, low failure rates from loosening of the
modern Metasul are encouraging for the future of metal-on-metal hips
43,51-55
.

WEAR SIMULATIONS

Wear Simulators

Wear of artificial hips is commonly simulated in vitro with pin-on-disc, bi-axial, or
tri-axial wear simulators. The degrees of freedom provided by the different
types of wear simulators are described in Table 3. Rotation around the x-axis is
in the coronal plane (abduction/adduction of the hip, or lateral bending of the
spine), around the y-axis is in the sagittal plane (flexion/extension), and around
the z-axis is in the transverse plane (axial rotation). Pin-on-disc wear simulators
provide translation in the xy-plane and rotation around the z-axis. Bi-axial wear
simulators do not provide translation, but provide rotation around any two axes.
For example, a bi-axial rocking motion (BRM), or orbital, wear simulator provides
rotation around the x- and y-axes. Tri-axial wear simulators do not provide
translation, but provide rotation around all three axes.

14

Table 3: Degrees of Freedom of Hip Wear Simulators
Translation Rotation
Pin-on-Disc
Anteroposterior
Mediolateral
Axial Rotation
Bi-Axial None
Flexion/Extension
Abduction/Adduction
Tri-Axial None
Flexion/Extension
Abduction/Adduction
Axial Rotation

Tri-axial hip wear simulators provide rotations around 3 axes, as is experienced
during hip gait. However, bi-axial hip wear simulators are still commonly used
because of their ability to reproduce the magnitudes and mechanisms of wear
as seen in vivo, despite the limitations of motion. A summary of contemporary
hip wear simulators as presented by Calonius and Saikko
20
is shown in Table 4.
Despite the simplification of wear simulators into bi-axial and tri-axial wear
simulators, the wear simulators within each group still have different methods of
applying the motions and loads.

15
Table 4: Summary of Contemporary Hip Wear Simulators
Neutral Position (°)
Simulator
Direction
of Load
Fixed
Component Ball Cup
Durham MK II V* Ball 45 45
Leeds Mk II V Cup 45 45
ProSim V Ball V 35 Bi-Axial
BRM zero-
offset lever
V Cup V H*
BRM offset
lever
V Cup V H
AMTI V Ball V H
Munich V Ball 45 45
Leeds Mk I Changing Neither 45 45
Tri-Axial
HUT-3 12° to V Cup 45 45
*V = Vertical, H = Horizontal

Wear Rates

For metal-on-metal hips, a period of accelerated run-in wear is typically followed
by a substantial decrease in wear rate tending toward steady-state values. The
wear rates of a representative sample of wear simulations of metal-on-metal
hips that tested variables similar to those in the current
study
31,32,56,57,72,73,85,150,151,187,196,229,232
are shown in Table 5. The steady-state
wear rates were typically less than 1.0 mm
3
/10
6
cycles; outliers of this trend
were attributed to high motion or load
72,232
, small diameters
196
, first-generation
metal-on-metal bearings
150
, sintering heat treatment, or excessive clearances
151
.
Typically, the running-in period of metal-on-metal hips is less than 1.0 x 10
6

cycles, but can be as long as 4.0 x 10
6
cycles
5,31,32,56,57,73,151,171,187,196
, as shown in
Table 6.

16
Additionally, for metal-on-metal hips, stripe wear has been observed at the
impinging edge of the cup
74,232
, which is typically located at the perimeter of the
ball, and has been shown to increase wear by threefold. Rarely, a metal-on-
metal hip exhibits runaway wear, in which the wear does not reach a steady-
state value but instead increases indefinitely to as much as 60.5 mm
3
/10
6

cycles
151
.

Table 5: Wear rates of metal-on-metal hips.

Variable
Tested
Run-In
Wear Rate
(mm
3
/10
6
cycles)
Steady-State
Wear Rate
(mm
3
/10
6
cycles)
Firkins et al (2001) Testing Profile 3.09 1.23
Williams et al (2004) Testing Profile 0.06 – 1.58
Goldsmith (2000) Diameter 0.36 – 0.45
Smith et al (2001) Diameter 1.62 0.54 – 6.30
Scholes (2001) Clearance 0.80 – 0.90 0.10 – 0.50
Dowson et al (2004)
Diameter
Clearance
0.79 – 3.23 0.09 – 0.17
McKellop et al (1996)
Diameter
Clearance
6.00
Medley et al (1996)
Diameter
Clearance
0.09 – 4.67
Chan et al (1996)
Diameter
Clearance
Material
0.20 – 8.00 0.25 – 0.60
Chan et al (1999)
Clearance
Material
0.02 – 1.90 0.03 – 0.21
Dowson et al (2004)
Clearance
Material
1.16 – 3.34 0.17 – 0.25
Firkins et al (2001) Material 0.02 – 0.32
Wang et al (1999) Material 0.05 – 0.32

17
Table 6: Running-in periods for metal-on-metal hip wear simulations
Running-In Period (10
6
Cycles)
Medley et al (1996) 0.1 – 0.5
Chan et al (1996) 0.5
Scholes et al (2001) 0.5
Anissian et al (1999) 0.9
Chan et al (1999) 1.0
Firkins et al (2001) 1.0
Dowson et al (2004) 1.0
Dowson et al (2004) 1.5
Smith et al (2001) 2.0
Rieker et al (2005) 0.5 – 4.0

2.1.2 Artificial Discs

The first known attempt at artificial disc replacement was performed in the 1950s
by Fernström, who inserted steel balls into the intervertebral space to restore
height and motion to the spine segment. In many of these cases, the ball was
excessively compressed and eventually subsided into the vertebral body
10,71
.
This implantation led to an explosion of ideas aimed at restoring height and
motion, ranging from rubber injections, to bladder-like implants, to implants with
springs, hinges, or screws
202
.

More recently, artificial discs have been designed to replicate either the
viscoelastic or kinematic function of the intervertebral disc through the use of
elastic or inelastic designs
10
. Elastic designs attempt to replicate the
viscoelasticity of the natural disc with shock-absorbing materials, while inelastic
designs attempt to replicate the kinematics of the natural disc with rigid

18
articulating surfaces. Some artificial discs have also employed the use of both
designs in an attempt to better simulate the physiology of the natural disc.

Additionally, artificial discs have been designed to replace the intervertebral disc
at any level of the spine. Efforts have been made to separate cervical from
lumbar disc replacements, due to the substantially different motion and load of
the functional spine units.

The current study focuses on inelastic designs of artificial lumbar discs. The
relatively few inelastic artificial lumbar discs that survived the post-Fernström
period of invention and managed to reach the phase of human implantation are
the LINK SB Charité III (DePuy Spine, Inc., Raynham, MA), the ProDisc
(Synthes, Inc., Paoli, PA), and the Maverick Total Disc Replacement prosthesis
(Medtronic Sofamor Danek, Inc., Memphis, TN).

CHARITÉ

The Charité is a metal-on-polyethylene lumbar disc that was first implanted in
1984. The Charité has a three-component set-up, which incorporates a floating
sliding polyethylene core with convex surfaces encased in the concave cavities
of the metal endplates, with the intent of imitating the movement of the nucleus
pulposus within the annulus fibrosus
138
. The polyethylene core of the Charité
shears with translation of the device to provide a mobile axis of rotation, and

19
intentionally impinges against the flared edges of the polyethylene core to
constrain the range of motion of the device and increase the stability of the
functional spine unit.

Reported clinical failure modes of the Charité included plastic deformation,
polyethylene oxidation, polyethylene cracks, polyethylene wear, and disc height
loss
17,24,44,92,126,134,207,217,239
. One case study of one Charité polyethylene core
retrieved after 9.5 years reported fragmentation into six pieces, but no visible
wear or wear debris
207
. Another case study of one Charité polyethylene core
retrieved after 1.6 years reported material failure in the form of transverse cracks
parallel to the central plane of symmetry, as well as pit formation at the interface
between the spherical articulating portion and the flared edges, where the
intentional impingement occurs, but minimal wear
126
.

Case studies of multiple retrievals have also been published. One study of 6
Charité polyethylene cores retrieved after 2.9 – 12.2 years reported adhesive
and abrasive wear; height loss; rim damage, including radial cracking, plastic
deformation, and third body wear; and central core damage, including
transverse and radial cracks. Additionally, histology revealed polyethylene wear
debris in the periprosthetic tissue, along with a local inflammatory reaction in 3/5
patients
127
. Another case study of 21 Charité polyethylene cores retrieved after
1.8 – 16.0 years reported similar results
128
.

20
PRODISC

The ProDisc is another metal-on-polyethylene lumbar disc that was first
implanted in the early 1990s. Unlike the Charité, the ProDisc has a ball-and-cup
design
242
. Reported clinical failure modes of the ProDisc included disc height
loss
8,108,212,213
.

MAVERICK

The Maverick is a metal-on-metal lumbar disc that was first implanted in 2002.
Like the ProDisc, the Maverick has a ball-and-cup design; however, the
Maverick also allows for some translation
147
. One case study of one Maverick
retrieved after 12 months reported highly polished surfaces similar to those
observed on explanted metal-on-metal hips, with microabrasion and focal
microplasticity
125
.

SUCCESS RATES

The success rates and restored range of motion of the Charité, ProDisc, and
Maverick
8,9,17,24,33,41,42,94,108,130,131,133,134,143,194,197,212,213,239,240
are shown in Table 7.
As a comparison, spinal fixation has success rates ranging from 60 –
95%
11,64,65,75,77,132,206
, resulting in an approximate mean success rate of 75%, and
artificial hips, which are the most successful of total joint replacements, have a

21
success rate of 93% after 10 years
118
. Two matched-case clinical studies
comparing patients with the Prodisc and patients with spinal fusion found that
patients with the ProDisc had higher satisfaction rates and greater ranges of
motion than patients with spinal fusion
240,241
.

Table 7: Success rates and ranges of motion of TDR
Implant Article
Number
of
Implants
Follow-
Up Time
(Years)
Success
Rate
(%)
Range of
Motion
(°)
Cinotti et al (1996) 46 3.2 77 9.0
Lemaire et al (1997) 105 4.3 79
Zeegers et al (1999) 50 2.0 70
Sott and Harrison (2000) 14 4.0 86
Buttner-Janz et al (2002) 6.8
Blumenthal et al (2003) 57 1.0 89
Caspi et al (2003) 20 4.0 80
Charité
Lemaire et al (2005) 107 11.3 90 10.3
Bertagnoli and Kumar
(2002)
134 0.3 – 2.0 98
Marnay (2002) 93 8.6 93
Delamarter et al (2003) 10.0
Huang et al (2003) 7.5
Tropiano et al (2003) 53 1.4 90 10.0
Zigler et al (2003) 13.0
Delamarter et al (2005) 56 2.0 93 8.0
Tropiano et al (2005) 8.7 75
Prodisc
Siepe (2006) 108 3.0 83
Le Huec et al (2005) 64 1.5 75 9.4
Maverick
Le Huec et al (2005) 35 1.2 82 7.3

WEAR SIMULATIONS

The appropriate testing profile for the lumbar spine is not generally known. In
contrast, for artificial hips, the ranges of motion and load during gait are
generally known
165
, and have been widely used in hip wear

22
simulations
31,32,56,57,72,73,85,150,151,187,196,229,232
. Most hip wear simulations apply a
gait profile because walking is the predominant daily activity, and have been
successful in accurately reproducing the magnitude and mechanism of wear
observed in vivo. More recently, the modes of wear observed on clinical
retrievals has inspired research advocating the simulation of lower-frequency
events of high motion and load to more accurately depict the wear observed in
vivo; examples of these include microseparation
142,159,200,201,211
, impingement
104
,
and increased patient activity simulations
13
. The simulation of high motion and
load provides insight into the different possible wear mechanisms of artificial
hips, but is usually interspersed with normal gait cycles
13
, thus including both
types of motions in the proportions encountered in vivo.

In contrast, the appropriate testing profile for the lumbar spine is not generally
known; as a result, lumbar disc wear tests vary widely in the application and
interpretation of motion and load
49,147,157,190
. For wear simulation of artificial
lumbar discs, the approved ASTM standard F 2423 – 05 (“Standard Guide for
Functional, Kinematic, and Wear Assessment of Total Disc Prostheses”) has
applied the concept of simulating lower-frequency events of high motion and
load, denoted as “significant bends” in the guide. The intended result is the
simulation of worst-case wear, but without concurrently simulating lower-motion
but higher-repetition gait; the effect of this method of simulation on wear has not
been tested.

23
Motion

The range of motion recommended by the ASTM standard guide is significantly
larger than that experienced by the lumbar spine during gait, and is in fact closer
to the maximum range of motion of the lumbar spine. The ASTM standard guide
recommends 15° of flexion/extension, 12° of lateral bending, and 6° of axial
rotation. The maximum ranges of in vivo motion of the lumbar spine as
presented by Lindh
137
are shown in Figure 1, while the ranges of motion during
gait
70,149,179,203-205,221,222,231
are shown in Table 8; the axial rotation range of
motion during gait was not available. The approximate in vivo range of motion
of flexion/extension during gait is 4°, while that of lateral bending during gait is
8.5°.

0 5 10 15 20 25
L5-S1
L4-5
L3-4
L2-3
L1-2
T12-L1
Functional Spine Unit
Range of Motion (°)
Flexion/Extension
Lateral Bending
Axial Rotation

Figure 1: In vivo maximum range of motion of the spine

24
Table 8: In vivo range of motion of the lumbar spine during gait
Author
Flexion/Extension
(°)
Lateral Bending
(°)
Taylor (1996) 3.2±0.9 12.8±3.0
Willems (1997) 1.8±0.8
McGowan (1998) 5.4±1.1 8.2±2.9
Taylor (1999) 3.2±0.7 3.5±1.3
Sartor (1999) 2.0 12.0
Feipel (2001) 5.0±1.0 9.0±2.0
Vogt (2001) 3.8±1.6 8.1±1.6
Vogt (2002) 4.4 3.9
Taylor (2003) 3.4±1.6 10.2±3.1

Physiologically, an ASTM-defined significant bend would represent maximally
flexing and extending the lumbar spine 125,000 times per year, or approximately
340 times per day, which is likely not representative of patients receiving spinal
disc arthroplasty. Furthermore, the ASTM-recommended profile accounts solely
for loading conditions created by significant bends, even though the lumbar
spine is subjected to far more repetitive loading cycles from walking (1 – 2 x 10
6

cycles per year
154,185,195,214
). Specifically, Morlock et al report “the most frequent
patient activity was sitting (44.3% of the time), followed by standing (24.5%),
walking (10.2%), lying (5.8%), and stair climbing (0.4%)
154
.” To simply estimate
how many significant bends might occur annually, and then calculate how much
wear was produced after that period of time, ignores all the wear that might have
been produced by the lower-motion but higher-repetition daily patient walking.

25
Load

The maximum load recommended by the ASTM standard guide is larger than
that experienced by the lumbar spine during gait. The ASTM standard guide
recommends either a constant 1200 N compressive load on the specimens, or a
sinusoidal compressive load ranging from 600 to 1800 N. The in vivo
compressive load is sinusoidal. The minimum and maximum compressive load
of the lumbar spine during gait
18,22,23,36,84,122,123
are shown in Table 9. For a
person weighing 90 kg, the approximate mean of the reported minimum load is
670 N, while that of the reported maximum load is 1200 N.

Table 9: In vivo maximum compressive load of the lumbar spine during gait
Study
Minimum Load
(times BW)
Maximum Load
(times BW)
Cappozzo (1983) 0.3 – 0.9
Cappozzo (1984) 0.2 – 0.8 1.0 – 2.5
Cromwell (1989) 1.2
Khoo (1994) 1.4 2.1
Khoo (1995) 1.7±0.3
Goh (1998) 1.5
Callaghan (1999)
1.2±0.2
0.4±0.1
2.2±0.5
0.8±0.1

Frequency and Phase Angle

The frequency and phase angle of the coupled motion and load are not specified
in the ASTM standard guide, but have been reported to varying degrees in
published gait studies
158,179-181,223
. The ASTM standard guide does not
recommend a frequency of the sinusoidal load, and recommends that the user

26
determine the phase angle of the motion. In vivo, the frequency of
flexion/extension in the lumbar spine is twice that of the gait cycle frequency,
while the frequency of axial rotation in the lumbar spine is the same as the gait
cycle frequency. Additionally, the frequency of compressive load is the same as
that of flexion/extension. An estimation of the frequency and phasing of motion
and load during gait, based on the gait studies mentioned previously, is shown
in Figure 2.

Figure 2: In vivo frequencies and phasing of motion and load during gait. The solid line
represents flexion/extension, the dashed line represents lateral bending, the dashdot line
represents axial rotation, and the dotted line represents the compressive load.

Wear Rates

There are have been three published wear tests of the metal-on-polyethylene
Charité or Prodisc
49,157,190
, as shown in Table 10. For the phasing, tests with
motion and load in phase produced curvilinear slide tracks, while tests with
motion and load out of phase, or frequency-shifted, produced crossing-path

27
motion. A range of wear rates of metal-on-polyethylene hips
60,68,113,228
was also
included for comparison. The type of polyethylene used for the Charité, Prodisc,
and artificial hips in the table was uncrosslinked polyethylene. The wear rates of
the Charité and Prodisc tested in phase were approximately 35 – 40 times less
than that of metal-on-polyethylene hips, while the wear rates of the Charité and
Prodisc tested out of phase were 3 – 4 less that that of metal-on-polyethylene
hips.

There has been one published wear test of the metal-on-metal Maverick
147
, as
shown in Table 10. A range of wear rates of metal-on-metal
hips
31,32,56,57,72,73,85,150,151,187,196,229,232
was also included for comparison. The wear
rate of the Maverick was 10 times more than that of the Charité and Prodisc
tested in phase, but 10 times less than that of the Charité and Prodisc tested out
of phase; additionally, the wear rate of the Maverick was comparable to that of
metal-on-metal hips.

28
Table 10: Wear simulations of artificial lumbar discs.
Motion (°)
Implant
Load
(N)
Flex./
Ext.
Lat.
Bend.
Rot.
Phase
Wear Rate
(mm
3
/10
6

cycles)
Maverick N/A N/A N/A N/A In 1.2 – 1.4
Metal-on-Metal
Hips <1.0
Charité
900 –
1850
±7.5 0 ±1.5 In 0.12
Charité
900 –
1850
±7.5 0 ±1.5 In 0.14
Charité
200 –
1750
-3,
+6
0 ±1.5 In -0.02
Charité
200 –
1750
-3,
+6
±2 ±1.5 Out 20.37
Prodisc
200 –
1750
-3,
+6
0 ±1.5 In 0.07
Prodisc
200 –
1750
-3,
+6
±2 ±1.5 Out 17.46
Metal-on-
Polyethylene
Hips 55.0 – 95.0

2.2 Problem Statements

The development of artificial hips has been advanced in great part by studies
that have measured the effects of individual independent variables, such as the
carbon content of the implant alloys
31,32,56,73,229
or the diameter and clearance of
the ball and cup
31,32,56,57,85,150,151,187,196
, on implant wear. Such studies have
typically used laboratory hip wear simulators to control the testing conditions
and, therefore, to isolate the effects of each variable of interest on wear. This
type of wear simulation is extensively used and produces results comparable to
those observed in clinical retrievals and radiographic studies.

29
In contrast to artificial hips, wear studies of artificial lumbar discs have not
examined the effects of independent design variables on wear. The existing
artificial lumbar disc wear publications have addressed wear in only the final
version of the product
49,125-127,147,157,190,207
. In these publications, the effects of
independent variables such as material, diameter, or clearance on wear were
not reported, although the importance of such variables has been established in
artificial hip wear studies. Additionally, there are few published retrieval studies
of artificial lumbar discs
125-127,207
against which to validate the results of artificial
disc laboratory wear simulations
49,147,157,190
.

Problem statements were formulated by comparing and contrasting artificial hips
to discs, citing relevant studies, and emphasizing the limitations of the studies in
establishing wear of metal-on-metal lumbar discs.

2.2.1 Implant Geometry

CUP DEPTH AND THICKNESS

¾ Problem Statement: The effects of shallow cup depths and thin cups, both
of which are typical of artificial discs, on wear of metal-on-metal bearing
surfaces have not been established.

30
The differences between the cup depth and thickness of artificial lumbar discs
and hips may have an effect on wear. Artificial hips are typically comprised of a
spherical ball articulating against a hemispherical cup, which is achieved by
reaming the femur and acetabulum to accommodate the implant. Artificial discs
commonly have a similar ball-and-cup design; however, the structure of the
functional spine unit limits the amount of space available for an artificial disc.
Thus, the cup depth or thickness of the implant is commonly decreased to
accommodate the space constraints of the functional spine unit.

The effects of shallow cup depths on wear of artificial discs have not been
established. Decreasing the cup depth may lead to decreased contact surface
area, which is of concern because smaller contact surface area increases
material stresses, which is generally known to increase wear. However, this
potential effect of shallow cup depths on wear has not been established.

The effects of thin cups on wear of artificial discs have not been established.
Decreasing the cup thickness is of concern because the increased flexibility may
lead to decreased lubrication or increased edge loading, both of which have
been shown to increase wear. Conversely, increased flexibility may lead to
larger contact surface area, which is generally known to decrease wear.
However, these potential effects of thin cups on wear have not been
established.

31
In contrast, for artificial hips, a study using three-dimensional finite element
techniques attributed the equatorial clamping of the McKee-Farrar to a
combination of the acetabular cup thickness and the clearance
237
. Another
study using three-dimensional finite element techniques was performed on metal
inlays, which are commonly inserted into polyethylene cups to provide the
shock-absorption of polyethylene with the wear resistance of metal
218
. The
study found that inlays thinner than 1.5 mm, which is on the order of artificial
discs, led to considerable increase in contact area and a reduction in contact
peak stress. However, the authors concluded that thin inlays would be heavily
dependent on unpredictable deformation of the bone, potentially leading to
clamping or excessive deformation, and therefore recommended liners greater
than 5 mm. Thus, the effects of thin cups on wear are unpredictable.
Additionally, the combined effects of shallow and thin cups on wear have not
been established. For example, decreased cup depth and thickness have
contrasting effects on contact surface area.

DIAMETER AND CLEARANCE

¾ Problem Statements: The effects of small diameters typical of artificial discs
on wear of metal-on-metal bearing surfaces have not been established.
Additionally, the definition and design of clearance of metal-on-metal
artificial discs have not been established.

32
The differences between the diameter and clearance of artificial lumbar discs
and hips may have an effect on wear. For artificial discs, the diameter of the
implant is commonly decreased to accommodate the space constraints of the
functional spine unit. Additionally, because clearance has traditionally been
defined for spherical balls and hemispherical cups, the application of this term to
artificial discs is unclear.

The effects of small diameters on wear of artificial discs have not been
established. Decreasing the diameter may result in decreased contact surface
area, which is generally known to increase stresses and wear of the implant.
However, these potential effects of small diameters on wear have not been
established.

In contrast, as discussed previously, hip wear simulator studies on metal-on-
metal hips with diameters ranging from 16 – 54.5 mm have shown that smaller
ball diameters increase wear
57,151,171,196
; thus, diameters larger than 28 mm are
preferable for metal-on-metal hips. For metal-on-metal lumbar discs, new
designs typically use diameters substantially smaller than 28 mm and even, in
some cases, smaller than 16 mm, to accommodate the space constraints of the
functional spine unit. The effects of diameters smaller than 28 mm have only
been tested in one hip wear simulation
196
, and the effects of diameters smaller
than 16 mm on wear have not been established. Additionally, the combined
effects of shallow, thin cups and small diameters on wear have not been

33
established. For example, decreased diameter and thickness have contrasting
effects on contact surface area.

The definition of clearance of artificial discs has not been established. The
definition of clearance is unclear because of the shallow cup depth of artificial
discs. Since clearance traditionally describes the difference in diameters of the
spherical ball and hemispherical cup of an artificial hip, it is also a good
approximation of the size of the gap at the edge of the implant for flow of
lubrication. In contrast, clearance of artificial discs has no physical meaning,
since the size of the lubrication gap at the edge of the implant is much smaller
than the clearance, due to the shallow cup depth of artificial discs.

The design of clearance of artificial discs has not been established. In contrast,
as discussed previously, hip wear simulator studies have shown that smaller
clearances reduce wear of metal-on-metal total hip
replacements
31,32,56,57,103,114,150,151,171,172,186,187
; however, some studies have
reported that clearances that are too small lead to equatorial
clamping
114,151,172,186
; thus, an optimum radial clearance for metal-on-metal total
hip replacements is typically on the order of 50 μm. However, the application of
this optimal clearance to artificial discs may be achieved by multiple methods.
One method of designing clearance is to maintain the difference in diameters of
the ball and cup based on artificial hip data, which will result in a smaller gap at
the edge of the implant for flow of lubrication due to the decreased cup depth

34
and, potentially, decreased lubrication or edge loading. Another method of
designing clearance is to maintain the gap at the edge of the implant based on
artificial hip data, which will result in large differences in diameters and,
potentially, decreased contact surface area. The effects of these methods of
designing clearance of artificial discs have not been established. Additionally,
the combined effects of shallow, thin cups, small diameters, and clearance on
wear have not been established. For example, decreasing the thickness of the
cup may eliminate any clearance between the ball and cup because of the
increased flexibility.

SURGICAL FEATURES

¾ Problem Statement: The effects of surgical features typical of artificial discs
on wear of metal-on-metal bearing surfaces have not been established.

Surgical features are commonly added to artificial discs to facilitate in surgical
technique or implant performance, and may have an effect on wear. One
common application of a surgical feature is the fixation of the implant to the
vertebra, which can be achieved with the addition of a keel. The increased
rigidity of the structure from the keel may increase lubrication, which is known to
decrease wear, but may decrease contact surface area, which is known to
increase wear. However, these potential effects of a keel on wear have not
been established.

35

Another common application of a surgical feature is the prevention of over-
distraction of the disc space during implantation, which can be achieved with the
addition of a notch cut out of the transverse plane of the cup surface. The notch
may decrease contact surface area of the cup, which may result in increased
wear. However, these potential effects of a notch on wear have not been
established.

2.2.2 Biomechanics

MOTION AND LOAD

¾ Problem Statement: The effects of motion and load typical of the lumbar
spine on wear of metal-on-metal bearing surfaces have not been
established.

While artificial hips replace synovial joints that typically experience a large range
of motion, artificial lumbar discs replace amphiarthrodial joints that typically
experience a small range of motion
89
. The differences in motion and load during
gait are shown in Table 11 and Figure 3, and will undoubtedly result in different
types and amounts of wear via sliding distance, lubrication, or crossing-path
motion. The effects of these differences in motion and load on wear of metal-
on-metal lumbar discs have not been established.

36

Table 11: Magnitudes of motion and load in the hip and lumbar spine during gait
Hip Lumbar Spine
Motion Maximum (°) 46 8.5
Load Maximum (N) 2000 1200

A

B

Figure 3: Motion and load waveforms illustrating differences in the (A) hip and (B) lumbar
spine during gait. The solid line represents flexion/extension, the dashed line represents
abduction/adduction of the hip or lateral bending of the lumbar spine, the dashdot line
represents axial rotation, and the dotted line represents the compressive load.

SLIDING DISTANCE

¾ Problem Statement: The effects of sliding distances typical of the lumbar
spine on wear of metal-on-metal bearing surfaces have not been
established.

The differences between magnitudes of motion and load experienced by artificial
lumbar discs and hips during gait may have an effect on wear. The magnitudes
of motion and load are critical because, according to Archard’s equation, the

37
differences in magnitudes of motion and load have a proportional effect on
wear
6
; however, this potential effect on wear of metal-on-metal lumbar discs has
not been established. The equation states that volumetric wear (W) is directly
proportional to load (L) and sliding distance (x), following the equation W = kLX,
where k is a probability factor typical of the combination of materials used,
applying over a range of experimental conditions where the wear process
remains of the same type
6
. Thus, the differences in magnitude of motion and
load have a proportional effect on wear only if one assumes the same wear
coefficient k, thus ignoring the more complex effects on wear via mechanical
interactions between the ball and the cup, including lubrication regimes
96,97
and
slide tracks
175,210,215
.

LUBRICATION

¾ Problem Statement: The effects of lubrication typical of the lumbar spine on
wear of metal-on-metal bearing surfaces have not been established.

The difference between velocity of motion experienced by artificial lumbar discs
and hips during gait may have an effect on wear. The velocity of motion is
critical because it determines the lubrication regime produced between the
articulating surfaces. Lubrication is the separation of the asperities of the
surfaces with a fluid film layer, which has an effect on wear; however, this
potential effect on wear of metal-on-metal lumbar discs has not been

38
established. The three lubrication regimes are boundary lubrication, which is full
asperity contact with no fluid film to separate the articulating surfaces; fluid film
lubrication, which is no asperity contact with a full fluid film layer to separate the
articulating surfaces; and mixed lubrication, which falls between the two. For
artificial hips, previous studies reported that metal-on-metal hips operate under
mixed lubrication
32,151,182,188,189
. Lubrication of metal-on-metal lumbar discs has
not been established.

One useful tool to calculate the lubrication regime is the elastohydrodynamic
lubrication theory. The Hamrock and Dowson elastohydrodynamic lubrication
equations for an equivalent ball-on-plane model of a ball-and-cup joint were
presented in 1977 and 1978
96,97
, and have since been utilized to analyze
lubrication regimes in various types of implants. The equations are presented in
detail in the Materials and Methods chapter.

CROSSING-PATH MOTION

¾ Problem Statement: The effects of crossing-path motion typical of the
lumbar spine on wear of metal-on-metal bearing surfaces have not been
established.

The differences between frequency and phase angles of motion and load
experienced by artificial lumbar discs and hips during gait may have an effect on

39
wear. The relative motion of the ball and the cup during walking produces slide
tracks on the implant. Crossing-path motion describes the amount of crossing of
the slide tracks during motion, and may have an effect on wear; however, this
potential effect on wear of metal-on-metal lumbar discs has not been
established. Slide tracks with no crossing-path motion, or linear motion, are
produced by uni-directional motion along one axis, while slide tracks with
crossing-path motion are produced by multi-directional motion along two or more
axes. Alternatively, crossing-path motion can be time-dependent, with
curvilinear motion produced by slide tracks following the same path in
consecutive cycles, and crossing-path motion produced by slide tracks crossing
over their previous paths in consecutive cycles.

The effects of direction-dependent crossing-path motion on wear of metal-on-
metal lumbar discs have not been established. In contrast, hip wear simulator
studies have shown that uni-directional motion increases the wear of metal-on-
metal hips and decreases the wear of metal-on-polyethylene hips, while multi-
directional motion has the opposite effect
210,215
. This has been attributed to the
multi-directional self-polishing properties of metal, in contrast to the susceptibility
to shearing of linearly-oriented polyethylene. One study quantified this effect for
alloys used in metal-on-metal hips
210
, and showed that uni-axial pin-on-disc tests
produced wear rates of 0.10 – 0.55 mm
3
Km
-1
, while mult-directional tests
produced lower wear rates of 0.10 – 0.18 mm
3
Km
-1
. The effects of uni- and

40
multi-directional motion on wear have not been established for metal-on-metal
lumbar discs.

The effects of time-dependent crossing-path motion on wear of metal-on-metal
lumbar discs have not been established. In contrast, one study on artificial disc
wear has shown that time-dependent crossing-path motion increases the wear
of metal-on-polyethylene lumbar discs
157
. The study tested the Charité and
Prodisc-L metal-on-polyethylene lumbar discs under a curvilinear motion profile,
as well as under a frequency-shifted, “cross-shear” motion profile. The wear
rate of the Charité increased from –0.02 to 20.37 mm
3
/10
6
cycles, while the wear
rate of the Prodisc-L increased from 0.07 to 17.46 mm
3
/10
6
cycles. The effects
of curvilinear and “cross-shear” motion on wear have not been established for
metal-on-metal lumbar discs.

One useful tool to plot slide tracks is the slide track analysis. Saikko and
Calonius developed a computation method based on Euler angles and used to
compute slide tracks for two- and three-axis motion
175
. The equations are
presented in detail in the Materials and Methods chapter.

41
CHAPTER 3: MATERIALS AND METHODS

To accomplish the specific aims, analytical tools were used to predict the effects
of different factors on wear of metal-on-metal lumbar discs. Then, wear
simulators were used to test different metal-on-metal artificial joints under
different testing profiles. The wear simulators, specimens, and testing profiles
were combined in a schedule of wear tests designed to facilitate the design and
testing of artificial lumbar discs.

3.1 Analytical Tools

Four analytical tools were used to predict the effects of different factors on wear,
as stated in Specific Aims #1 and #2:

• Finite element models to predict contact pressure
• Hertzian contact equations to predict contact surface area and contact
pressure
• Elastohydrodynamic lubrication equations to predict lubrication regime
• Slide track analysis to predict sliding distance and aspect ratio of the slide
tracks

42
3.1.1 Finite Element Models

Three-dimensional finite element models were created in ABAQUS (ABAQUS,
Inc., Providence, RI) to analyze the effects of load and implant dimensions on
contact pressure. A static load was applied axially through the center of the ball,
with a bony interface constraining the backside of the cup. The Elastic Modulus
of the ball and cup was 200 GPa, representing Cobalt-Chromium alloy, while
that of bone was 20 GPa. Poisson’s ratio was 0.30 for all materials. Three-
dimensional hex elements were used to seed the models.

3.1.2 Hertzian Contact Equations

The Hertz equations for circular contact were used to analyze the contact
mechanics between the ball and cup, and thus provide a comparison for the
results of the finite element model
95
. The input parameters were the load (L),
radius of the ball and cup (R
1
, R
2
) Elastic Modulus of the ball and cup (E
1
, E
2
),
and Poisson’s Ratio of the ball and cup ( γ
1
, γ
2
). The relative radius and contact
modulus were calculated by the equations:

Equation 1:

1/R
r
= 1/R
1
+ 1/R
2

1/E
c
= (1- γ
1
2
)/E
1
+ (1- γ
2
2
)/E
2

43

Finally, the radius of the contact circle and the maximum contact pressure were
calculated by the equations:

Equation 2:

R
contact
= (3LR
c
/4E
c
)
1/3

P
max
= (1/ π)(6LE
c
2
/R
r
2
)
1/3

2.1.3 Elastohydrodynamic Lubrication Equations

The Hamrock and Dowson elastohydrodynamic lubrication equations were used
to analyze the lubrication between the ball and cup
96,97
. Elastohydrodynamic
lubrication equations have previously been applied to a wrist prosthesis
168
and to
metal-on-metal hip prostheses
116,189,216
. The elastohydrodynamic lubrication
equations for high-elastic modulus materials
97
are described below:

Input parameters for the implant are the ball radius (R
1
), cup radius or radial
clearance (R
2
or c, where c = R
2
– R
1
), semimajor and semiminor axis of contact
ellipse (a,b), average surface roughness value (Ra
1,2
), Poisson’s ratio ( ν
1,2
), and
Elastic Modulus (E
1,2
). Input parameters for the body or simulated environment
are the normal applied load (F), angular velocity ( ω), lubricant viscosity ( η), and

44
asymptotic isoviscous pressure (p
iv,as
), which can be approximated by the
inverse of the pressure viscosity coefficient (p
iv,as
= 1/ α).

The input parameters for the elastohydrodynamic lubrication equations were
based on typical values for metal-on-metal hips, and are shown in Table 12.

Table 12: Input parameters for metal-on-metal artificial hips and discs
Parameter Value
CoCr Modulus of Elasticity 200 GPa
CoCr Poisson’s Ratio 0.3
CoCr Surface Roughness (Ra) 0.012 μm
Lubricant Viscosity 0.01 Pa s

As acknowledged in previous studies
116,168,189,216
, the high value of the viscosity
(0.01 Pa s, as opposed to a more realistic viscosity of 0.002 Pa s) was
necessary to facilitate the numerical solution.

The ball in the EHL model has an effective radius (R
x
), equivalent Elastic
Modulus (E'), surface velocity (u), and compound surface roughness ( σ) that
combines those of the original ball and cup:

45
Equation 3:

R
x
= 1 / ( 1/R
1
– 1/R
2
) = R
1
R
2
/ c
1/E' = 0.5 ( (1- ν
1
2
)/E
1
+ (1- ν
2
2
)/E
2
) )
u = ωR
1
/ 2
σ = √ ( Ra
1
2
+ Ra
2
2
)

Dimensionless parameters were derived by simultaneously solving the elasticity
and Reynolds’ equations in an isothermal elastohydrodynamic lubrication point
contact analysis:

Equation 4:

Ellipticity parameter: k = a/b
Speed parameter: U = ηu/E'R
x

Load parameter: W = F/E'R
x
2

Material parameter: G = E'/p
iv,as

For the ball-on-plate configuration (k = 1) presented in the current study, the
simplified equation for minimum fluid film thickness is:

Equation 5:

h
min
= 1.8R
x
U
0.68
G
0.49
W
-0.073

46
The lambda ratio is the ratio of the minimum fluid film thickness to the compound
surface roughness, thereby representing the height of the film relative to the
height of the asperities of the bearing surfaces:

Equation 6:

λ = h
min
/ σ

The lambda ratio ( λ) was calculated to predict the lubrication regime. A lambda
ratio under 1 indicates boundary lubrication with full asperity contact, while a
ratio above 3 indicates fluid film lubrication with no asperity contact; a ratio
between 1 and 3 indicates a mixed lubrication regime. A higher lambda ratio
indicates more lubrication, which leads to less wear. For artificial hips, previous
studies reported that metal-on-metal hips operate under mixed
lubrication
32,151,182,188,189
.

3.1.4 Slide Track Analysis

The slide track analysis by Saikko and Calonius was used to analyze the
crossing-path motion between the ball and cup
175
. The differences in motion
between the hip and disc during gait were represented by the different motion
waveforms, which were used to calculate the per-cycle distance the cup travels
relative to the ball, as well as to plot the path of motion of the slide tracks.

47
Flexion/extension, abduction/adduction (for the hip) or lateral bending (for the
spine), and axial rotation waveforms were discretized. Flexion/extension
rotations were made about the x-axis at a rotation angle of α
i
, lateral bending
rotations about the y-axis at a rotation angle of β
i
, and axial rotation rotations
about the z-axis at a rotation angle of γ
i
, where i = 1, 2, 3, …, N, and N was the
number of discrete points, in this case 100. A marker point r fixed to the head
was repeatedly rotated from its initial position r
0
to a new position on the slide
track. Each point r
i
of the slide track corresponded to one set of rotation angles
(α
i
, β
i
, γ
i
), as shown in the following equations:

Equation 7:

r
i
= R
xyz
( α
i
, β
i
, γ
i
) r
0

where the rotation matrix R
xyz
( α
i
, β
i
, γ
i
) was:

Equation 8:

[ cos β cos γ - cos β sin γ sin β

(sin α sin β cos γ + cos α sin γ) (-sin α sin β sin γ + cos α cos γ) (-sin α cos β)

(-cos α sin β cos γ + sin α sin γ) (cos α sin β sin γ + sin α cos γ) (cos α cos β) ]

In order to avoid the accumulation of numerical errors, the points were not
computed by using the previous point on the track as a starting point. The track
was drawn by connecting all points defined by r
i
, i = 1, 2, 3, …, N. In the case of

48
two-axis motion, the slide track patterns of the head and the cup were identical,
but their angular positions had a difference of pi/2. Slide tracks and their
distances were calculated at 6 equidistant latitudes and 8 equidistant longitudes.
In the current study, the aspect ratios of the slide tracks were defined as the
ratio of width to length of the slide track, with 0 corresponding to a straight line
and 1 corresponding to a circle. The sliding distance and aspect ratio were then
used together to assess the cross-path angles.

3.2 Wear Simulators

Three different variations of wear simulators were used in the current study, two
of which were built or modified as stated in Specific Aims #3 and #4:

• A custom three-station, bi-axial spine wear simulator that was designed and
constructed for the current study
• An as-manufactured four-station, bi-axial orbital hip wear simulator
manufactured by Shore Western (Shore Western Manufacturing, Monrovia,
CA)
• A modified version of the four-station, bi-axial orbital hip wear simulator that
was modified for the current study to reflect the motion and load of the spine

49
3.2.1 Spine Wear Simulator

A three-station, bi-axial spine wear simulator was designed and constructed by
the Biomechanics Laboratory at Orthopaedic Hospital (Los Angeles, CA). The
machine is capable of applying cyclic flexion/extension and axial rotation,
individually or coupled at any phase angle, under programmable axial load. A
three-station design provides equally distributed loading and stability among the
stations. A ball-joint bearing allows all degrees of rotation for self-adjustment
between the three stations.

Kinematics

Flexion/extension is provided by a motor in the cranial aspect of the wear
simulator. The AC motor (A) is fitted with a speed reducer (B) to provide
maximum power at the appropriate velocity, as shown in Figure 4. The motor
shaft is attached to a bevel gear, which rotates against three pinion gears (C).
The pinion gears transmit the motion of the motor at 90° while branching it out to
the three separate stations, as shown in Figure 5. The shaft from each pinion
gear inserts into a cam (D), which is a cylindrical piece of metal with an offset
hole drilled at a chosen distance from the center of the circular face. A ball
bearing is fitted around the cam, and a collar is fitted around the ball bearing to
provide a surface into which a protruding shaft is inserted at 90° to the shaft
from the pinion gear. The rotation of the shaft from the pinion gear is converted

50
into motion in a plane that is 90° to the shaft from the pinion gear, following the
mechanics of a four-bar linkage. A linear bearing (E) at the distal end of the
protruding shaft eliminates linear motion and therefore converts the planar
motion to a pure back-and-forth rotation that oscillates at an amplitude
determined by the offset of the hole in the cam and a frequency determined by
the velocity of the motor shaft, as shown in Figure 6 and Figure 7.

Axial rotation is provided by a motor in the caudal aspect of the wear simulator,
using the same mechanical concepts as for flexion/extension, as shown in
Figure 8. An H20 optical encoder (BEI Technologies, Inc., Goleta, CA) is
attached to a cam for flexion/extension, and to the shaft of the motor for rotation.
The frequency and phase angle of the motion are synchronized with a two-
signal edge separation program combined with feedback control of the voltage
output of the motors; the program was written in LabView 7.1 (National
Instruments Corporation, Austin, TX).

51

Figure 4: Custom spine wear simulator. The motor and speed reducer in the cranial
component provide flexion/extension, while those in the the caudal component provide
axial rotation.

Figure 5: Custom spine wear simulator. The pinion gears transmit the motion to three
separate stations.

52

Figure 6: Custom spine wear simulator. The cam and linear bearing convert the planar
motion to a pure back-and-forth rotation that oscillates at an amplitude determined by the
offset of the hole in the cam and a frequency determined by the velocity of the motor shaft.

Figure 7: Custom spine wear simulator. The cam and linear bearing convert the planar
motion to a pure back-and-forth rotation that oscillates at an amplitude determined by the
offset of the hole in the cam and a frequency determined by the velocity of the motor shaft.

53

Figure 8: Custom spine wear simulator. Axial rotation is provided by a motor and speed
reducer in the caudal aspect of the wear simulator, using the same mechanical concepts as
for flexion/extension.

Load

The axial load is provided by hydraulic fluid controlled by the TestWare-SX
software (MTS Systems Corporation, Eden Prairie, MN). The load cycle is
triggered by a signal from the optical encoder, which is attached to a cam and
sends a signal to the software whenever the cam reaches a certain point in its
cycle. The load was programmed with either a sinusoidal signal generated with
the Cyclic Command function, or a digitized curve applied with the File Playback
function in the TestWare-SX software. The synchronization of the motion and
load is determined by the position of the optical encoder and the phase angle
specified in the TestWare-SX software.

54
A variety of bearings ensure that the load is only applied through the specimens,
rather than through any part of the assembly. Linear and rotary ball bearings at
the motor shafts and tapered roller bearings at the chamber bases allow for the
required motion of the specimens while keeping the force transfer localized to
the articulating surfaces of the specimens. Shear load is eliminated by allowing
a small degree of self-adjustment in the cranial components of the specimens.

Specimen Chambers

The specimen chambers hold the specimens and fluid in a contained
environment. There are three chambers for the three test stations, as well as
three additional control chambers at the cranial aspect of the wear simulator.
The control specimens experience the same load as the test specimens, but
none of the motion, and may be used to correct for absorption of serum during
loading of polyethylene specimens. The control chambers are also arranged in
a three-station design to provide equally distributed loading and stability among
the control chambers. The weight of the wear simulator is counterbalanced by
springs at the cranial end, which ensures that the control specimens and the test
specimens are loaded simultaneously.

The specimens are molded in polyurethane to minimize micromotion and to
prevent metal exposure. The center of rotation of the specimens is determined
by the height of the polyurethane mold, which is controlled by cutting with a

55
lathe. The polyurethane molds are then secured into the delrin adaptors with set
screws. A polycarbonate hollow cylinder is fit over the polyurethane molds with
an o-ring to create a waterproof chamber for the lubricant of choice.
Polyurethane, delrin, and polycarbonate are used to prevent any metal exposure
to the serum, which may produce undesirable metal transfer or particles.

An intravenous drip device at each chamber replenishes moisture lost through
evaporation. The implant and bulk temperatures are recorded with an 8018
Thermocouple Input Module (SuperLogics, Inc., Waltham, MA) to ensure that
the temperatures do not rise above 50°C, at which point the proteins in the
serum may denature
135
.

3.2.2 Hip Wear Simulators

A four-station, bi-axial rocking motion (BRM), or orbital, hip wear simulator
(Shore Western Manufacturing, Monrovia, CA) was used either as
manufactured, or modified to reflect the motion and load of the spine during gait.
The machine is capable of applying cyclic flexion/extension and lateral bending,
coupled at a set phase angle, under programmable axial load.

56
Kinematics

Flexion/extension and lateral bending are provided by a single motor in the
caudal aspect of the wear simulator. The specimen chambers are separated by
a thrust bearing from the angled platform, which is fixed to a rotating shaft; this
effectively eliminates rotation and couples the flexion/extension and lateral
bending with a phase angle of pi/2, as shown in Figure 9. The magnitude of the
motion oscillates at an amplitude determined by the angle of the platform and a
frequency determined by the velocity of the motor shaft. For the modified
version of the hip wear simulator, the angle of the platform was changed with an
adapter block, as shown in Figure 10.

Figure 9: Frequency and phasing of motion in the orbital wear simulator. The solid line
represents flexion/extension, and the dashed line represents abduction/adduction of the
hip or lateral bending of the lumbar spine.

57

A

B

Figure 10: Hip wear simulator with a total range of motion of A) 46°, as manufactured, and
B) modified with an adapter block to 8.5°.

Load

The axial load was provided by hydraulic fluid controlled by the mmed software
provided by Shore Western. The load cycle is triggered by a magnet, which is
embedded in the test chamber and sends a signal to the software whenever the
chamber reaches a certain point in its cycle.

Specimen Chambers

The specimen chambers hold the specimens and fluid in a contained
environment. There are four chambers for the four test stations. The
specimens are molded in polyurethane to minimize micromotion and to prevent

58
metal exposure. The center of rotation of the specimens is determined by the
height of the polyurethane mold, which is controlled by cutting with a lathe. A
polycarbonate hollow cylinder is fit over the polyurethane molds with an o-ring to
create a waterproof chamber for the lubricant of choice. A lid on each chamber
prevents the evaporation of the testing lubricant.

3.3 Specimens

Five different types of specimens were available for testing:

• artificial hips with the original 14-mm cup depth
• artificial hips with a customized 2-mm cup depth
• first-generation artificial lumbar discs, which had keels on the fixation
surfaces of the endplate and cup, large clearance, and a low carbon-content
ball
• second-generation artificial lumbar discs, which had no keels, small
clearance, and a high carbon-content ball
• second-generation artificial lumbar discs without a notch in the cup surface

3.3.1 Artificial Hips

Four metal-on-metal artificial hip specimens were provided by Sulzer (Sulzer
Ltd., Winterthur, Switzerland). Each specimen consisted of a 28-mm ball and a

59
cup. The cups were tested with either the original cup depth of 14 mm or a
customized cup depth of 2 mm that was modified for the current study by cutting
the original hip specimen down to the desired cup depth with a lathe, as shown
in Figure 11 and Figure 12. The same specimens were used for all tests.

Figure 11: Specimen ball (left), cup with a depth of 14 mm (middle), and cup with a depth of
2 mm (right).

Figure 12: Specimen cup with a depth of 14 mm (left) and 2 mm (right).

3.3.2 Artificial Lumbar Discs

Experimental metal-on-metal lumbar discs were provided by an orthopaedic
manufacturer. Each artificial disc specimen consisted of a caudal endplate, a

60
ball spacer with a diameter of 25.4 mm and a taper that locks into the caudal
endplate, and a cranial cup, as shown in Figure 13. Two generations of
specimens were customized for the current study, with the second generation
consisting of improvements made after analyzing the results of the first
generation. New specimens were used for each test.

The first generation of specimens had balls made of low carbon-content (<0.1%)
CoCrMo alloy. The cups were made of high carbon-content (0.2 – 0.3%)
CoCrMo alloy, and had a notch cut out of the transverse plane to accommodate
the surgical instrumentation for implantation. The clearance between the ball
and the cup was large (125 – 175 μm), and the fixation surfaces of the endplate
and cup had an anteroposterior keel for fixation, as shown in Figure 14.

The second generation of specimens had balls made of high carbon-content
CoCrMo alloy. The cups were also made of high carbon-content CoCrMo alloy,
and were customized with and without a notch cut out of the transverse plane to
facilitate surgical implantation. The clearance between the ball and the cup was
small (10 – 35 μm); additionally, the anteroposterior keel was eliminated from
this generation, as shown in Figure 14.

61

Figure 13: Caudal endplate (left), ball spacer (middle), and cranial cup (right).

Figure 14: Keeled (upper-left) and un-keeled (upper-right) cups, and notched (lower-left)
and un-notched (lower-right) cups.

3.4 Testing Profiles

Although in vivo gait involves the coupling of three motions, the wear simulators
were only able to couple two motions. However, the frequency and phasing of
the bi-axial motions were manipulated so as to produce slide tracks similar to
those produced under tri-axial motions. Using these results and results from
pilot tests, along with data from hip wear simulations, standards for disc wear

62
simulations, and published gait data, four wear test profiles were generated for
the current study:

• Hip Orbital
• Spine Orbital
• Spine Gait
• Spine Bends

3.4.1 Hip Orbital

This profile is a standard method of simulating wear of artificial hips during gait.
The specimen was tested in the orbital hip wear simulator and subjected to a bi-
axial rocking motion with the magnitude of motion the hip experiences during
gait. A discretized Paul curve with the magnitude of compressive load the hip
experiences during gait was applied as the load profile once per gait cycle, as
shown in Figure 15. The sliding distance and aspect ratio of the slide tracks,
also shown in Figure 15, are 29.40 mm and 0.85, and the lambda ratio is 3.64.

63

A

B

Figure 15: Hip Orbital A) motion and load profile and B) slide tracks produced. The solid
line represents flexion/extension, the dashed line represents abduction/adduction, the
dashdot line represents axial rotation, and the dotted line represents the compressive load.

3.4.2 Spine Orbital

This profile is a scaled-down version of the standard method of wear testing of
hips. The specimen was tested in the orbital hip wear simulator and subjected
to a bi-axial rocking motion with the magnitude of motion the lumbar spine
experiences during gait. A scaled-down discretized Paul curve with the
magnitude of compressive load the lumbar spine experiences during gait was
applied as the load profile once per gait cycle, as shown in Figure 16. The
sliding distance and aspect ratio of the slide tracks, also shown in Figure 16, are
5.47 mm and 0.89, and the lambda ratio is 1.20.

64

A

B

Figure 16: Spine Orbital A) motion and load profile and B) slide tracks produced. The solid
line represents flexion/extension, the dashed line represents lateral bending, the dashdot
line represents axial rotation, and the dotted line represents the compressive load.

3.4.3 Spine Gait

This profile was generated from published gait data previously discussed in the
Background section, and applies typical motion and load experienced by the
lumbar spine during gait. The specimen was tested in the spine wear simulator
and subjected to the magnitude of flexion/extension and axial rotation the
lumbar spine experiences during gait, with the frequencies and phasing of bi-
axial motions that most closely produces the slide tracks produced by tri-axial
gait. A scaled-down discretized Paul curve with the magnitude of compressive
load the lumbar spine experiences during gait was applied as the load profile
twice per gait cycle, representing the left and right steps of each gait cycle, as
shown in Figure 17. The sliding distance and aspect ratio of the slide tracks,
also shown in Figure 17, are 7.21 mm and 0.29, and the lambda ratio is 1.45.

65

A

B

Figure 17: Spine Gait A) motion and load profile and B) slide tracks produced. The solid
line represents flexion/extension, the dashed line represents lateral bending, the dashdot
line represents axial rotation, and the dotted line represents the compressive load.

3.4.4 Spine Bends

This profile was generated from the ISO standard, the ASTM standard, and
published gait data, and applies high motion and load representing significant
bends. The specimen was tested in the spine wear simulator and subjected to
the magnitude of flexion/extension and axial rotation the lumbar spine
experiences during significant bends, without synchronization. A sinusoidal
curve was applied as the load profile once per gait cycle, as shown in Figure 18.
The sliding distance and aspect ratio of the slide tracks, also shown in Figure
18, are 6.62 mm and 0.24, and the lambda ratio is 1.37.

66

A

B

Figure 18: Spine Bends A) motion and load profile and B) slide tracks produced at phase
angles of pi/4, pi/2, 3pi/4, and pi. The solid line represents flexion/extension, the dashed
line represents lateral bending, the dashdot line represents axial rotation, and the dotted
line represents the compressive load. The phasing of the motion was not synchronized
and therefore not necessarily as shown in the figure.

A summary of parameters for all tests, including sliding distances, aspect ratios,
and lambda ratios, is shown in Table 13.

67
Table 13: Parameters for testing profiles
Profile
Parameter
Hip Orbital Spine Orbital Spine Gait Spine Bends
Flexion/Extension ±23° ±4.25° ±4.25° ±7.5°
Lateral Bending ±23° ±4.25° None None
Axial Rotation None None ±3° ±3°
Flexion/Extension
Frequency, Axial
Rotation or Lateral
Bending Frequency
1 Hz,
1 Hz
1 Hz,
1 Hz
1 Hz,
0.5 Hz
1 Hz,
1 Hz
Coupling
Flexion/
Extension,
Lateral
Bending
Flexion/
Extension,
Lateral
Bending
Flexion/
Extension,
Axial
Rotation
Flexion/
Extension,
Axial
Rotation
Phase Angle
from
Flexion/Extension
Curve
Pi/2 Pi/2 Pi/4 Random
Compression
Profile
Paul
(140-2000 N)
Paul
(85-1200 N)
Paul
(85-1200 N)
Sinusoidal
(300-1200 N)
Compression
Frequency
1 Hz 1 Hz 1 Hz 1 Hz
Sliding Distance
(mm)
29.40 5.47 7.21 6.62
Aspect Ratio 0.85 0.89 0.29 0.24
Lambda Ratio 3.64 1.20 1.45 1.37

3.5 Schedule of Wear Tests

A series of tests was developed to study the effects of motion and load of the
testing profiles, as well as the geometry and metallurgy of the implants, on wear.
The tests are presented in Table 14.

68
Table 14: Schedule of Wear Tests
Hip Disc
14 mm Cup
Depth
2 mm Cup
Depth
Gen. 2*,
Notch
Gen. 2*,
No Notch
Gen. 1*,
Notch
Hip
Orbital
Test 1
Spine
Orbital
Test 2
Spine
Gait
Test 3 Test 4 Test 6
Spine
Bends
Test 5 Test 7 Test 8 Test 9
*Gen. 1 = Keeled, high clearance, low carbon content (<0.10%)
Gen. 2 = Un-keeled, low clearance, high carbon content (0.20 – 0.30%)

The goals of Tests 1 – 4 were to determine the effects of the differences
between hips and lumbar discs on wear of metal-on-metal joints by comparing
the effects of motion, load, and cup design on wear (Specific Aim #5).

The goals of Tests 7 – 9 were to compare the effects of metallurgy and cup
design on wear of experimental metal-on-metal artificial lumbar discs, in order to
facilitate the design of artificial lumbar discs (Specific Aim #6).

The goals of Tests 4 – 6 were to compare the effects of different motion and
load profiles on wear of metal-on-metal artificial lumbar discs, in order to
optimize the wear test parameters for evaluating artificial lumbar discs (Specific
Aim #7).

69

3.6 Measurement Methods

A schedule of the measurement intervals is shown in Table 15. At each
measurement interval, the specimens were ultrasonically cleaned, following
ASTM Standard F732, prior to measurements.

Table 15: Schedule of Measurements
Measurement Interval (cycles)
Weight Every 0.25 x 10
6

Profilometer
CMM
Light Microscope
0
0.25 x 10
6

0.5 x 10
6

1 x 10
6

2 x 10
6

2.5 x 10
6

SEM End of test

At each measurement interval, the specimens were ultrasonically cleaned,
following ASTM Standard F732, prior to measurements. Specimens were
weighed using an AT261 Precision Balance with an accuracy of 1.0 x 10
-5
g
(Mettler Electronics Corp., Anaheim, CA). Average roughness (Ra) of the balls
were measured with a Perthometer S8P laser profilometer fitted with a non-
contact laser probe and tracing length of 0.070 inches, with an accuracy of 1.0 x
10
-3
μm (Mahr-Perthen-Gottingen, Germany/Feinpruf Corporation, Charlotte,
NC). The topographies of both bearing surfaces were measured in evenly
distributed matrices of 300 to 400 points using a BRT 504 Coordinate
Measurement Machine (CMM) (Mitutoyo USA, Aurora, IL) with a Renishaw TP-

70
200 touch probe and 4-mm ruby stylus for specimens with a cup depth of 14
mm, or a 2-mm ruby stylus for specimens with a cup depth of 2 mm, with an
accuracy of 1.0 x 10
-5
mm. Light microscopy was performed using an MZ8
Optical Microscope (Leica Camera AG, Solms, Germany), in addition to
scanning electron microscopy (SEM). Depending on the distribution of data,
either a Wilcoxin Signed Ranks Test of related samples or an Independent-
Samples T Test (SPSS Inc., Chicago, IL) was used to calculate p-values.

For particle analysis, the serum was enzymatically digested, and metal particles
were isolated by centrifuging the sample through a CsFm gradient, as shown in
Figure 19. Particles were then filtered onto a 30 nm polycarbonate filter and
gold-sputtered. An SEM was used to visualize the particles. Larger particles
were identified with an energy dispersive X-ray (EDAX) instrument. Peaks of Co
and Cr were used to positively identify metal particles, as shown in Figure 20.
Smaller particles were then identified by morphology similar to the positively
identified larger particles. Particle lengths were quantified using Metamorph
4.6r1 digital image analysis (Molecular Devices Corp., Downingtown, PA) of the
scanning electron micrographs.

71

Figure 19: Centrifugation tube with pre-centrifugation layers labeled. After centrifugation,
the metal particles remained isolated at the bottom of the tube, while the rest of the sample
remained above the CsFm gradient. A visible layer of white protein at the top of the tube
was observed.

Figure 20: EDAX image of elements present in the sample. The larger peaks corresponded
to the aluminum stub, the polycarbonate filter, and the gold-sputtering. Smaller, almost
indistinguishable peaks corresponded to the Co and Cr of the metal particles.

72
CHAPTER 4: ANALYTICAL PREDICTIONS OF ARTIFICIAL DISC WEAR

This chapter addresses Specific Aims #1 and #2, which were to create finite
element models and apply equations for lubrication and slide tracks to predict or
explain wear of metal-on-metal artificial lumbar discs.

4.1 Introduction

In recent years, total disc replacement has gained increasing popularity as an
alternative to spinal fusion. Articulating artificial lumbar discs generally use the
same material combinations and constructs that have been proven to be
successful for artificial hips, including metal-on-metal bearings. Although the
material combinations may be the same, the intervertebral space is smaller than
the hip joint, and therefore substantially smaller implant dimensions need to be
used for artificial discs. Furthermore, the motion and load acting on the
intervertebral disc are substantially different than those acting on the hip. The
differences in dimensions, motion, and load between artificial discs and artificial
hips will undoubtedly result in different types and amounts of wear

Wear is dependent on the contact surface area. Artificial hips typically consist of
a ball articulating against a hemispherical cup. On the other hand, the smaller
intervertebral disc height makes it difficult to incorporate a full hemispherical cup
for artificial discs. As a solution, the depth of the cup for artificial discs is typically

73
reduced, as shown in Figure 21. As a result, compared to a hip, the load on an
artificial disc is distributed over a smaller area, creating higher contact pressures
on the surface and potentially increasing wear.

A

B

Figure 21: Cup dimensions of A) typically hemispherical artificial hips and B) shallower
artificial discs.

Wear is also dependent on the magnitude of motion during gait, such as
flexion/extension and rotation, which is substantially larger for the hip than for
each lumbar intervertebral disc. Implants with a larger range of motion produce
a larger worn area and may thus generate greater amounts of wear, according
to a proportional relationship between motion and wear
6
. On the other hand, a
larger range of motion is also typically accompanied by a higher velocity,
potentially creating a lubricating layer of fluid between the hydroplaning surfaces
which would, in fact, decrease wear
96,97
.

Wear is dependent on not only the magnitude of motion, but also the amount of
“criss-crossing” of the paths of interfacial motion on the surfaces of the ball and

74
cup, commonly known as crossing-path motion. For example, in metal-on-metal
bearings, greater amounts of crossing-path motion have been shown to polish
metal surfaces and decrease wear
210,215
. The differences in frequency and
phasing of motion and load between the hip and lumbar spine produce different
amounts of crossing-path motion.

Analytical tools are commonly used in engineering to predict the relative effects
of implant design, load and motion related variables on wear. Ideally, analytical
tools are validated with experimental or clinical data. However, many artificial
discs are being implanted today. Follow-up of these patients will require years
or decades of study, as will obtaining retrieval implants, and even laboratory
wear simulations may take months to perform and, additionally, depend on
retrievals for validation. Therefore, despite their limitations, in the short term,
analytical tools are the only method to infer the wear behavior of artificial discs.
They do not require large investments in time and can highlight specific areas of
concern that need to be further addressed by longer-term studies.

In the current study, we used three analytical tools to assess the wear of metal-
on-metal artificial lumbar discs relative to that of metal-on-metal artificial hips.

75
4.2 Methods

Three-dimensional finite element models were created in ABAQUS (ABAQUS,
Inc., Providence, RI) to analyze the effects of load and implant dimensions on
contact pressure. Two cup models were created. The first, representative of a
hip acetacular cup, was hemispherical with a 14 mm radius (nominal) and cup
depth. The second, representative of a metal-on-metal artificial disc, had the
same radius, but was a spherical segment with a depth of 2 mm, as shown in
Figure 21. Both models included a ball 28 mm in diameter. The diametral
clearance between the ball and each cup was 100 μm, and the cup thickness
was 1 mm. A static load was applied axially through the center of the ball, with a
bony interface constraining the backside of the cup. The Elastic Modulus of the
ball and cup was 200 GPa, representing Cobalt-Chromium alloy, while that of
bone was 20 GPa. Poisson’s ratio was 0.30 for all materials. Three-
dimensional hex elements were seeded at intervals of approximately 0.5 mm.

The lubrication between the surfaces was analyzed with Hamrock and Dowson’s
elastohydrodynamic lubrication equations
97
. Input parameters for the implant
were the ball radius, cup radius, average surface roughness value, Elastic
Modulus, and Poisson’s ratio. Other input parameters were the normal applied
load, angular velocity, lubricant viscosity, and asymptotic isoviscous pressure.
The input parameters were based on typical values for metal-on-metal hips,
except for the high value of the lubricant viscosity (0.01 Pa s), which, as

76
acknowledged by others
116,168,189,216
, was necessary to facilitate the numerical
solution. Based on an equivalent ball-on-plane model of a ball-and-cup joint
with these parameters, the lambda ratio ( λ) was calculated to predict the
lubrication regime. Thresholds for lubrication regimes were 0-1 for boundary
lubrication, 1-3 for mixed lubrication, and greater than 3 for fluid film lubrication.

The crossing-path motion produced on the ball surface was analyzed with the
slide track analysis developed by Saikko and Calonius
175
. Typical magnitudes
and phasing of motion and load of the hip and lumbar spine during gait were
estimated from published biomechanical models and gait
studies
18,22,23,36,70,84,122,123,149,158,165,178-181,203-205,221-223,231
. The differences in motion
between the hip and disc during gait were represented by the different motion
waveforms, which were used to calculate the per-cycle distance the cup travels
relative to the ball, as well as to plot the path of motion of the slide tracks.
Flexion/extension, abduction/adduction (for the hip) or lateral bending (for the
spine), and axial rotation waveforms were digitized. A marker point r fixed to the
head was repeatedly rotated from its initial position to a new position on the slide
track. Each point of the slide track corresponded to one set of rotation angles
defined by the motion waveforms. The slide track was then drawn by connecting
all points. Slide tracks and their distances were calculated at 6 equidistant
latitudes and 8 equidistant longitudes. In the current study, the aspect ratios of
the slide tracks were defined as the ratio of width to length of the slide track, with
0 corresponding to a straight line and 1 corresponding to a circle. The sliding

77
distance and aspect ratio were then used together to assess the cross-path
angles.

4.3 Results

In the current study, we used three analytical tools to predict the wear of artificial
lumbar discs relative to artificial hips: a finite element model to predict the
contact pressure distribution; a slide track analysis to predict the crossing-path
motion distribution; and elastohydrodynamic lubrication equations to predict the
lubrication regime. While these analytical tools are not yet validated with
experimental or clinical data, they highlighted specific areas of concern that
need to be further addressed by longer-term studies.

Wear and material failure are dependent on contact pressure. In the current
study, the contact pressures and, therefore, the potential for wear or material
failure of artificial discs were substantially higher than for hips, as shown in
Figure 22. The higher contact pressures of artificial discs compared to hips
were initially attributed to the reduction in contact surface area from the
reduction in cup depth; however, examination of the contact pressure
distribution revealed a more complex effect. The contact surface area of the
disc was in fact smaller than that of the hip, but the cup was not in full contact,
as initially thought. The contact pressures of the disc were more localized in the
center of the cup, whereas the contact pressures of the hip were also localized

78
in the center but distributed over a greater surface area. This may have been
due to the flexibility of the thin cup.

Maximum Contact Pressures of an
Artificial Hip and Disc
0
50
100
150
200
250
Hip Disc
Maximum Contact Pressure (MPa)

Figure 22: Maximum contact pressures of artificial hips and discs.

Wear is also dependent on the range of motion. The elastohydrodynamic
lubrication equations state that a smaller range of motion results in a slower
velocity and, thus, a thinner film of lubrication between the asperities of the
surfaces, resulting in more wear. In the current study, artificial discs bordered
on boundary lubrication, while artificial hips bordered on fluid film lubrication, as
shown in Figure 23. In previous studies of metal-on-metal hips, lubrication
regimes ranged from boundary lubrication to fluid film lubrication, depending on
factors such as the diameter and clearance, with the majority of publications

79
predicting mixed lubrication
32,57,151,171,182,188,189,196,216
. Therefore, the lower
lubrication of artificial discs predicted more wear of discs than artificial hips.

Figure 23: Lambda ratios and lubrication regimes for artificial hips and artificial lumbar
discs during gait.

On the other hand, smaller sliding distances generally result in less wear and, in
fact, Archard’s equation describes a proportional relationship between sliding
distance and wear
6
. The smaller range of motion of artificial discs would
therefore predict less wear than artificial hips. The combined effect of
contradictory predictions from the elastohydrodynamic lubrication equations and
from Archard’s equation is not known.

Wear is also dependent on the amount of crossing-path motion on the surfaces
of the ball and cup. The path of motion of the slide tracks is critical because

80
greater cross-path angles produce less metal-on-metal wear, but more metal-
on-polyethylene wear
210
. In the current study, artificial discs had substantially
smaller sliding distances but marginally larger aspect ratios than artificial hips,
the combination of which likely forms smaller cross-path angles on artificial discs
than on artificial hips, as shown in Figure 24 and Table 16. Therefore, the
smaller cross-path angles of artificial discs predicted more wear of metal-on-
metal discs than metal-on-metal hips.

Hip

Disc

Figure 24: Motion waveforms and slide tracks produced for the (A) hip and (B) lumbar spine
during gait. The solid line represents flexion/extension, the dashed line represents
abduction/adduction of the hip or lateral bending of the lumbar spine, and the dotted line
represents axial rotation.

Table 16: Sliding distances and aspect ratios of slide tracks produced by the hip and
lumbar spine during gait
Hip Lumbar Spine
Sliding Distance (mm) 20.63 5.23
Aspect Ratio 0.33 0.53

81
4.4 Discussion and Conclusions

In the current study, the larger maximum contact pressure, lower lubrication, and
smaller cross-path angles predicted more wear for metal-on-metal artificial
lumbar discs compared to metal-on-metal artificial hips; however, the smaller
sliding distance predicted less wear. These counter-intuitive results indicate that
the differences in design between artificial discs and artificial hips can have
profound effects on wear. The results caution against the general assumption
that metal-on-metal artificial lumbar discs will be successful only by using the
same material combinations and clearances of successful metal-on-metal
artificial hips, without further design and material considerations.

The analytical tools presented in the current study may aid in designing
comparative studies of metal-on-metal artificial lumbar discs. For example, most
new designs of metal-on-metal artificial lumbar discs use diameters smaller than
28 mm, which were not analyzed in the current study. In previous studies of
metal-on-metal hips, wear of 16 and 22 mm diameter joints experienced 2–10
times the amount of wear than those with diameters in excess of about 28
mm
57,196
; therefore, diameters larger than 28 mm are preferable for metal-on-
metal hips. An analysis using the tools presented in the current study may be
useful in determining the effects of smaller diameters on wear of metal-on-metal
artificial lumbar discs, as a precursor to wear simulator studies or implantation.

82
Wear studies of artificial lumbar discs are scarce, with conflicting results. For
metal-on-polyethylene implants, the wear rates from two published wear
simulations reported that the Charité and Prodisc were 100 times smaller than
those of metal-on-polyethylene artificial hips
49,190
. However, another study
reported that the wear rates can be comparable or even more than those of their
artificial hip counterparts
157
when crossing-path motion is applied to the artificial
discs during the wear simulations. An analysis of the mechanics of wear of
metal-on-polyethylene artificial lumbar discs, similar to the one presented in the
current study, may have predicted this discrepancy in wear rates.

Wear studies of artificial metal-on-metal lumbar discs are even more scarce,
with only one retrieval study
125
and one wear simulation
147
of the metal-on-metal
Maverick. The analytical tools presented in the current study may serve as a
guide to designing the load and motion profiles under which to test metal-on-
metal artificial lumbar discs.

In summary, the contradictory predictions of wear behavior of metal-on-metal
artificial discs emphasize the need for comprehensive wear simulator studies, as
well as clinical and retrieval studies before metal-on-metal artificial lumbar discs
are used on a large scale.

83
CHAPTER 5: DESIGN AND MEASUREMENT OF CLEARANCE OF ARTIFICIAL
DISCS

This chapter addresses Specific Aims #1 and #2, which were to create finite
element models and apply equations for Hertzian contact to predict or explain
wear of metal-on-metal artificial lumbar discs.

5.1 Introduction

Articulating artificial lumbar discs have implemented the metal-on-metal ball-
and-cup design that has been successful for artificial hips, but with modifications
that are necessary because of the anatomical differences between the joints.
Specifically, while the femur and acetabulum are typically reamed to
accommodate the typically spherical ball and hemispherical cup of an artificial
hip, the structure of the functional spine unit limits the amount of space available
for an artificial disc. Thus, the cup depth of artificial discs is commonly
decreased to accommodate the space constraints of the functional spine unit.
The effects of a shallower cup depth on designing and measuring clearance of
artificial discs have not been established.

84
Design of Clearance

A shallow cup depth may have an effect on designing the appropriate clearance
for maximizing lubrication and minimizing contact pressures. Hip wear simulator
studies have shown that smaller clearances reduce wear of metal-on-metal total
hip replacements
31,32,56,57,103,114,150,151,171,172,186,187
; however, some studies have
reported that clearances that are too small lead to equatorial
clamping
114,151,172,186
; thus, an optimum diametral clearance for metal-on-metal
total hip replacements is typically on the order of 100 μm. This clearance may
not be applicable to artificial discs with shallow cups, as the reduced cup depth
influences the lubrication between the surfaces of the specimen, as well as the
distribution of contact pressures in the material.

In the current study, the contact pressures of metal-on-metal joints with
hemispherical cup depths typical of artificial hips were compared to those with
shallower cup depths typical of artificial lumbar discs for five different
clearances. The results may facilitate in the design of clearance of metal-on-
metal joints with shallow cup depths. To facilitate analysis of the results of the
current study, the Hertz equations for circular contact were used to explain the
contact mechanics
95
.

85
Measurement of Clearance

A shallow cup depth may also have an effect on the measurement method for
calculating clearance. For artificial hips, the radius of the ball or cup is typically
calculated by measuring the coordinates of points on the surface, either in an
evenly distributed matrix or in a line, then fitting a surface or curve to the data
points, using regression analysis. Radial clearance is then calculated by
subtracting the radius of the ball from the radius of the cup. This method may
not be applicable to artificial discs with shallow cups, as the reduced cup depth
decreases the surface area available for the typical surface- or curve-fitting
algorithms.

5.2 Methods

Design of Clearance

Three-dimensional finite element models were created in ABAQUS (ABAQUS,
Inc., Providence, RI) to analyze the effects of load and dimensions on contact
pressure. Models were analyzed with cup depths of 14 mm and 2 mm. All
models had a ball diameter of 28 mm and cup thickness of 1 mm. A static load
of 1200 N was applied axially through the center of the ball, with a bony
interface constraining the backside of the cup. The Elastic Modulus of CoCrMo
in the model was 200 GPa, while that of bone was 20 GPa; Poisson’s ratio was
0.30 for CoCrMo and bone. Three-dimensional hex elements were seeded at

86
intervals of approximately 0.5 mm. The dimensions analyzed are shown in
Table 17.

Table 17: Cup depths and diametral clearances analyzed
Cup Depth
(mm)
Diametral Clearance
(μm)
2
14
20
40
60
80
100

To facilitate analysis of the results of the current study, the Hertz equations were
used to explain the contact mechanics. The input parameters were the load (L),
radius of the ball and cup (R
1
, R
2
) Elastic Modulus of the ball and cup (E
1
, E
2
),
and Poisson’s Ratio of the ball and cup ( γ
1
, γ
2
). The relative radius and contact
modulus were calculated by the equations:

1/R
r
= 1/R
1
+ 1/R
2

1/E
c
= (1- γ
1
2
)/E
1
+ (1- γ
2
2
)/E
2

Finally, the radius of the contact circle and the maximum contact pressure were
calculated by the equations:

R
contact
= (3LR
c
/4E
c
)
1/3

P
max
= (1/ π)(6LE
c
2
/R
r
2
)
1/3

87
Measurement of Clearance

The surface-fit method for calculating the radius was performed on the cup of an
artificial disc with a nominal radius of 12.738 mm and a cup depth of 2 mm,
using four independent machines operated by different individuals:

• Mitutoyo CMM with a 2-mm ruby stylus and QUALSTAR software
• Mitutoyo CMM with an 0.5-mm stylus and M Cosmos software
• Zeiss Contourecord
• Mitutoyo FormTracer CS-3000 with a 4 μm 60° conical diamond tip

The CMM used with the QUALSTAR software recorded three-dimensional
measurements in an evenly distributed matrix around the surfaces, while the
other three machines recorded two-dimensional measurements in a single line
through the center of the specimen. The two CMM machines recorded fewer
points of data (~300) than the Contourecord and FormTracer (~30,000).

5.3 Results

Design of Clearance

The maximum contact pressures calculated from the Hertz equations and from
the finite element models with a cup depth of either 14 or 2 mm are shown in
Figure 25, while the contact zones are shown in Table 18.

88

Maximum Contact Pressure vs. Clearance
0
50
100
150
200
250
300
20 40 60 80 100
Clearance (um)
Maximum Contact Pressure (MPa)
Hertz
14 mm
2 mm

Figure 25: Maximum contact pressures for varying clearances. The contact pressures were
calculated from the Hertz equations or analyzed with the finite element model with a cup
depth of either 14 or 2 mm. The maximum contact pressures of the 2-mm model were
substantially larger than those of the 14-mm model; contact pressures from both finite
element models were larger than predictions from the Hertz equations. The contact
pressures for both cup depths increased with increasing clearance until 100 μm for the 14-
mm model and 80 μm for the 2-mm model, at which point the contact pressures decreased;
this decrease was in contrast to the steadily increasing trend predicted by the Hertz
equations.

89
Table 18: Contact zones for varying clearances. The contact zones were calculated from
the Hertz equations or analyzed with the finite element model with a cup depth of either 14
or 2 mm. For the Hertz equations, the diameter (mm) of the contact zone is indicated in the
figure. All dimensions are to scale for comparison. Colors of contour plots are scaled to
the maximum of each specimen for clarity.
Contact Zone Diametral
Clearance
(µm)
Hertz 14 mm 2 mm
20

40

60

80

100

Measurement of Clearance

The resultant radii from the four different measurement methods are shown in
Figure 26.

90

Cup Radius from Four Measurement Methods
12.620
12.640
12.660
12.680
12.700
12.720
12.740
12.760
12.780
12.800
12.820
CMM + QUALSTAR CMM + M Cosmos Contourecord FormTracer
Cup Radius (mm)
Nominal Radius = 12.738 mm

Figure 26: Cup radius calculated with four different methods. The nominal radius is
indicated with a dotted line.

5.4 Discussion

Design of Clearance

The maximum contact pressures of the 2-mm model were substantially larger
than those of the 14-mm model. The larger contact pressures were initially
attributed to the decrease in cup depth and the assumed decrease in contact
surface area; however, further analysis showed that the contact mechanics of
the finite element models presented were more complex than this simple
principle. Specifically, the contact area of the 14-mm model showed evidence of
contact of the entire surface at smaller clearances (20 and 40 μm), and larger

91
contact zones than those predicted by Hertz equations for the remaining
clearances. This may have been due to the 1-mm thickness of the models,
which is quite thin and flexible. However, despite the larger contact zones, the
contact pressures of the 14-mm model were higher than those predicted by the
Hertz equations; this may have been due to the thickness of the implant, which
the Hertz equations do not consider. In contrast, the contact zones of the 2-mm
model were slightly smaller than those predicted by the Hertz equations.
Accordingly, the contact pressures were higher than those predicted by Hertz
equations, as well as those of the 14-mm model.

The contact pressures for both cup depths increased with increasing clearance
until 100 μm for the 14-mm model and 80 μm for the 2-mm model, at which point
the contact pressures decreased; this decrease was not consistent with
predictions from the Hertz equations or results from hip wear simulations.
Further analysis of the contact zones provided more insight into these trends.
Indeed, the contact surface area the 14-mm model increased from 80 to 100
μm, which was consistent with the contact pressures, and perhaps a result of
the thickness of the implant, which the Hertz equations do not consider.
However, the contact surface area of the 2-mm model did not seem to increase
from 60 to 80 μm, which was inconsistent with the contact pressures. Perhaps
the mesh of the models contributed to this discrepancy, as the pattern of the
contact zone is not circular for the 2-mm model at clearances of 60 and 100 μm.

92
Measurement of Clearance

The radii calculated by the four methods varied by 120 μm, with the difference
from the nominal radius ranging from 5 to 69 μm. These differences are
substantial when considering that the nominal radial clearance of the artificial
disc is 38 μm. The results may be different because the cup of the artificial disc
only forms a 60° arc from which data points can be taken, due to the 2 mm cup
depth, while the surface- and curve-fitting algorithms typically require a 90° arc
for accuracy
208
. Thus, the surface- and curve-fitting algorithms typically used to
calculate the radii and clearances of artificial hips may not be accurate when
applied to artificial discs.

In the current study, a new method is proposed to calculate the lubrication gap
at any cup depth, without fitting a surface or curve to the data. The lubrication
gap is defined as the gap between the ball and the cup when the two are making
contact at their apexes. Of particular importance is the lubrication gap at the
edge of the cup, which may determine the flow of lubrication between the
surfaces of the specimen.

For the new non-surface-fit method, the CMM recorded the height (z) in
concentric circles defined in the x-y plane, as shown in Figure 27.

93

Figure 27: Coordinate Axes and Reference Points

Mean heights were plotted for each concentric circle, with the points between
concentric circles connected by linear interpolation. The ball and cup were
plotted with their apexes coinciding at the origin. A radial line was then drawn
from the center of the ball at the desired radial angle ( α) from the origin, which is
directly related to the cup depth. The lubrication gap (L) was defined as the
distance of the radial line segment between the intersection of the ball and the
intersection of the cup, as shown in Figure 28.

Figure 28: Definition of Radial Angle ( α) and Lubrication gap (L)

The new non-surface-fit method is able to calculate the lubrication gap directly
from the data, while calculating clearance from the traditional surface-fit method

94
may be inaccurate for implants with shallow cups, as shown by the results from
four different measurement methods. Additionally, the new non-surface-fit
method is able to calculate the lubrication gap at any cup depth directly from the
data, while the traditional surface-fit method only calculates clearance. Although
the lubrication gap may be calculated from the clearance by linear interpolation,
the deviations at each radial angle due to wear or manufacturing imperfections
are not captured, as they are with the new non-surface-fit method. Thus, the
new non-surface-fit method presented is more accurate for calculating
lubrication gaps of metal-on-metal artificial discs.

A brief analysis of a worn metal-on-metal hip comparing the lubrication gaps
calculated from the new non-surface-fit method and from interpolating from the
traditional surface-fit clearance was performed. The results are shown in Figure
29, and show that the new non-surface-fit method shows deviations at 45° due
to wear, which was not detected by the traditional surface-fit method.

95
Lubrication Gaps at Various Radial Angles
0
10
20
30
40
50
60
70
80
15 30 45 60 75
Radial Angle (degrees)
Lubrication Gap (um)
Non-Surface-Fit
Surface-Fit

Figure 29: Comparison of non-surface-fit and surface-fit methods

5.5 Conclusions

Designing an appropriate clearance of implants with hemispherical cups, such
as artificial hips, is substantially different for implants with shallow cups, such as
artificial discs. The contact mechanics are difficult to predict, as evidenced by
the discrepancies between the finite element models and the predictions from
the Hertz equations. An analysis such as the one presented in the current study
is useful in designing clearance of specific implants.

The traditional method of measuring clearance of artificial hips is inaccurate for
artificial discs, due to the limitations of the curve- and surface-fitting algorithms.
Thus, a new non-surface-fit method of calculating the lubrication gap was

96
proposed in the current study, which may be used to calculate the lubrication
gap at any cup depth directly from the data. A brief analysis showed that the
new method showed deviations in the surface that were unable to be detected
by the traditional method.

97
CHAPTER 6: EFFECTS OF MOTION, LOAD, AND CUP DEPTH ON IN VITRO
WEAR OF METAL-ON-METAL BEARINGS

This chapter addresses the goals of Tests 1 – 4, which were to determine the
effects of the differences between hips and lumbar discs on wear of metal-on-
metal joints by comparing the effects of motion, load, and cup design on wear
(Specific Aim #5). The test labels in this chapter are not consistent with those of
the dissertation.

6.1 Introduction

Articulating artificial lumbar discs have implemented the metal-on-metal ball-
and-cup design that has been successful for artificial hips; however, the
differences between the joints may have substantial effects on wear.
Specifically, artificial lumbar discs and hips are subjected to different magnitudes
and phasing of motion and load during gait. Another difference between artificial
lumbar discs and hips is the cup depth of the implant. While the femur and
acetabulum are typically reamed to accommodate the typically spherical ball and
hemispherical cup of an artificial hip, the cup depth of an artificial disc is
commonly decreased to accommodate the space constraints of the functional
spine unit.

98
In the current study, the effects of the biomechanical and dimensional
differences between artificial lumbar discs and hips on wear were evaluated in a
series of wear simulations. Specifically, the magnitudes motion and load, the
phasing of motion and load, and the cup depth were treated as independent
variables. The separate contribution of each variable on wear was determined
by keeping all other conditions constant while changing each variable
sequentially. To achieve this goal, the motion and load profiles applied by the
wear simulators were independently controlled. Additionally, the same metal-
on-metal specimens were used for all tests, but with different cup depths. A
hemispherical cup depth was representative of artificial hips, while a shallower
cup depth, achieved by cutting the cup with a lathe, was representative of
artificial lumbar discs.

To facilitate analysis of the results of the current study, three analytical tools
were used to explain the mechanics of wear: a finite element model to predict
the contact pressure distribution; a slide track analysis to predict the crossing-
path motion distribution; and elastohydrodynamic lubrication equations to predict
the lubrication regime.

99
6.2 Materials

Specimens

Four metal-on-metal artificial hip specimens (Sulzer Ltd., Winterthur,
Switzerland) were tested, each consisting of a 28-mm ball and a cup. The cups
were tested with either the original cup depth of 14 mm or a shallower cup depth
of 2 mm that was achieved by cutting the cup depth with a lathe (Figure 30 and
Figure 31). The same specimens were used for all tests.

Figure 30: Specimen ball (left), cup with a depth of 14 mm (middle), and cup with a depth of
2 mm (right).

100

Figure 31: Specimen cup with a depth of 14 mm (left) and 2 mm (right).

Wear Simulators

The specimens were tested in either a hip wear simulator (Shore Western
Manufacturing, Monrovia, CA) (Figure 32A), a modified version of the hip wear
simulator that applied the magnitudes of motion and load of the spine (Figure
32B), or a spine wear simulator that was designed and constructed by the
Biomechanics Laboratory at Orthopaedic Hospital (Los Angeles, CA) (Figure
33). The four-station, bi-axial hip wear simulator is capable of applying cyclic
flexion/extension and lateral bending, coupled at a set phase angle, under
programmable axial load. The three-station, bi-axial spine wear simulator is
capable of applying cyclic flexion/extension and axial rotation, individually or
coupled at any phase angle, under programmable axial load. The specimens
were lubricated with alpha calf serum (HyClone, Logan, UT) with 0.2% sodium
azide to retard bacterial degradation, and with 20mM EDTA to prevent non-
physiological precipitation of calcium salts onto the bearing surfaces. For the
spine wear simulator, the flexion/extension and load were run at 1 Hz and

101
synchronized with TestWare-SX (MTS Systems Corporation, Eden Prairie, MN).
The flexion/extension and axial rotation were synchronized with two H20 optical
encoders (BEI Technologies, Inc., Goleta, CA) providing input to a LabView 7.1
program combining two-signal edge separation with feedback control of voltage
output (National Instruments Corporation, Austin, TX). The implant and bulk
temperatures were recorded every 30 minutes with an 8018 Thermocouple Input
Module (SuperLogics, Inc., Waltham, MA) to ensure that the temperatures did
not rise above 50°C, at which point the proteins in the serum may denature.

A

B

Figure 32: Hip wear simulator with a total range of motion of A) 46° and B) 8.5°, as
determined by the angle of the platform.

102

Figure 33: Spine wear simulator constructed by the Biomechanics Laboratory at
Orthopaedic Hospital (Los Angeles, CA). The cranial component experiences
flexion/extension, while the caudal component experiences axial rotation.

6.3 Methods

Testing Profiles

The ranges of motion and load of the hip during gait are generally known
165
, and
have been widely used in hip wear simulations
31,32,56,57,72,73,85,150,151,187,196,229,232
;
typical values of the range of motion and maximum load are 46° and 2000 N,
respectively. The motion profile has been shown to be sinusoidal, while the load
profile has been shown to be a double-peaked curve, commonly known as a
Paul curve. In contrast, the motion and load of the lumbar spine during gait are
not generally known; as a result, artificial lumbar disc wear tests vary widely in

103
the application and interpretation of motion and load
49,147,157,190
. Based on gait
studies of the lumbar spine, the approximate in vivo range of motion of
flexion/extension during gait is 4°, while that of lateral bending during gait is
8.5°
70,149,179,203-205,221,222,231
; the maximum load is approximately 1200
N
18,22,23,36,84,122,123
. For the motion profile, the frequency of flexion/extension in
the lumbar spine has been shown to be twice that of axial rotation. For the load
profile, the frequency of compressive load is the same as that of
flexion/extension
158,179-181,223
.

Three sets of motion and load waveforms were generated for use in two bi-axial
spine wear simulators: the Hip Orbital profile, the Spine Orbital profile, and the
Spine Gait profile. The slide tracks produced on the articulating surfaces by the
relative motion of the ball and cup were calculated as described by the method
based on Euler angles by Saikko and Calonius
175
. The motion waveforms were
used to calculate the per-cycle distance the cup travels relative to the ball, as
well as to plot the path of motion of the slide tracks. Flexion/extension,
abduction/adduction or lateral bending, and axial rotation waveforms were
digitized. A marker point r fixed to the head was repeatedly rotated from its
initial position to a new position on the slide track. Each point of the slide track
corresponded to one set of rotation angles defined by the motion waveforms; the
slide track was drawn by connecting all points. Slide tracks were calculated at 6
equidistant latitudes and 8 equidistant longitudes, and are presented alongside
the profiles.

104

The Hip Orbital profile is a standard method of simulating wear of artificial hips
during gait. The specimen was tested in the orbital hip wear simulator and
subjected to a bi-axial rocking motion with the magnitude of motion (46°) the hip
experiences during gait. A discretized Paul curve with the magnitude of
compressive load (140-2000 N) the hip experiences during gait was applied as
the load profile once per gait cycle (1.1 Hz), as shown in Figure 34.

A

B

Figure 34: Hip Orbital A) motion and load profile and B) slide tracks produced. For the
motion and load profile, the solid line represents flexion/extension, the dashed line
represents abduction/adduction, the dashdot line represents axial rotation, and the dotted
line represents the compressive load.

The Spine Orbital profile is a scaled-down version of the Hip Orbital profile. The
specimen was tested in the modified orbital hip wear simulator and subjected to
a bi-axial rocking motion with the magnitude of motion (8.5°) the lumbar spine
experiences during gait. A scaled-down discretized Paul curve with the
magnitude of compressive load (85-1200 N) the lumbar spine experiences

105
during gait was applied as the load profile once per gait cycle (1.1 Hz), as shown
in Figure 35.

A

B

Figure 35: Spine Orbital A) motion and load profile and B) slide tracks produced. For the
motion and load profile, the solid line represents flexion/extension, the dashed line
represents lateral bending, the dashdot line represents axial rotation, and the dotted line
represents the compressive load.

The Spine Gait profile was generated from published gait
data
18,22,23,36,70,84,122,123,149,158,179-181,203-205,221-223,231
, and applies typical motion and
load experienced by the lumbar spine during gait. The specimen was tested in
the spine wear simulator and subjected to the magnitude of flexion/extension
(8.5°) and axial rotation (6°) the lumbar spine experiences during gait, with the
phasing of bi-axial motions that most closely produces the slide tracks produced
by tri-axial gait. A scaled-down discretized Paul curve with the magnitude of
compressive load (85-1200 N) the lumbar spine experiences during gait was
applied as the load profile twice per gait cycle, representing the left and right
steps of each gait cycle (1 Hz), as shown in Figure 36.

106

A

B

Figure 36: Spine Gait A) motion and load profile and B) slide tracks produced. For the
motion and load profile, the solid line represents flexion/extension, the dashed line
represents lateral bending, the dashdot line represents axial rotation, and the dotted line
represents the compressive load.

Schedule of Tests

A series of tests was developed to study the effects of magnitudes of motion and
load, phasing of motion, and cup depth on wear of metal-on-metal artificial joints
(Table 19), using two spine wear simulators developed for the current study.
Each test had a duration of 1.0 x 10
6
cycles.

Table 19: Schedule of tests for the current study.
Testing Profile
Cup Depth
(mm)
Test
Hip
Orbital
Spine
Orbital
Spine
Gait

1 X 14
2 X 14
3 X 14
4 X 2

107
Tests 1 and 2 examined the effects of the magnitudes of motion and load on
wear. Test 1 tested the specimens under the Hip Orbital profile, to determine a
baseline FDA-approved wear rate of the specimens, while Test 2 tested the
specimens under the Spine Orbital profile, to determine how the wear rate
changes when the magnitudes of the motion and load are changed from that of
the hip to that of the lumbar spine during gait.

Tests 2 and 3 examined the effects of phasing of motion on wear. Test 2 tested
the specimens under the Spine Orbital profile, while Test 3 tested the specimens
under the Spine Gait profile, to determine how the wear rate changes when the
phasing of the motion changes from orbital to that typically experienced by the
lumbar spine during gait.

Tests 3 and 4 examined the effects of cup depth on wear. Test 3 tested
specimens with the hemispherical cup depth of the original hip specimens (14
mm), which is not possible for artificial discs due to space constraints of the
functional spine unit, while Test 4 tested the specimens cut down to a cup depth
typical of artificial discs (2 mm) to determine the effects of cup depth on the wear
rate of the specimens.

108
Measurements

At each measurement interval, the specimens were ultrasonically cleaned,
following ASTM Standard F732, prior to measurements. Specimens were
weighed using an AT261 Precision Balance with an accuracy of 1.0 x 10
-5
g
(Mettler Electronics Corp., Anaheim, CA). Average roughness (Ra) of the balls
were measured with a Perthometer S8P laser profilometer fitted with a non-
contact laser probe and tracing length of 0.070 inches for the ball, or a diamond-
tipped stylus with a tracing length of 0.022 inches for the cup, with an accuracy
of 1.0 x 10
-3
μm (Mahr-Perthen-Gottingen, Germany/Feinpruf Corporation,
Charlotte, NC). The topographies of both bearing surfaces were measured in
evenly distributed matrices of 300 to 400 points using a BRT 504 Coordinate
Measurement Machine (CMM) (Mitutoyo USA, Aurora, IL) with a Renishaw TP-
200 touch probe and 4-mm ruby stylus for specimens with a cup depth of 14
mm, or a 2-mm ruby stylus for specimens with a cup depth of 2 mm, with an
accuracy of 1.0 x 10
-5
mm. Light microscopy was performed using an MZ8
Optical Microscope (Leica Camera AG, Solms, Germany). The serum was
enzymatically digested to isolate the particles, which were quantified using
Metamorph 4.6r1 digital image analysis (Molecular Devices Corp.,
Downingtown, PA) of the scanning electron micrographs. A Wilcoxin Signed
Ranks Test of related samples (SPSS Inc., Chicago, IL) was used to calculate p-
values.

109
Analytical Tools

To facilitate analysis of the results, three analytical tools were used to explain
the mechanics of wear: a finite element model to predict the contact pressure
distribution; a slide track analysis to predict the crossing-path motion distribution;
and elastohydrodynamic lubrication equations to predict the lubrication regime.

Three-dimensional finite element models were created in ABAQUS (ABAQUS,
Inc., Providence, RI) to analyze the effects of load and dimensions on contact
pressure. Models were analyzed with cup depths of 14 mm and 2 mm. All
models had a ball diameter of 28 mm, diametral clearance of 100 μm, and cup
thickness of 3 mm. A static load was applied axially through the center of the
ball, with a bony interface constraining the backside of the cup. The Elastic
Modulus of CoCrMo in the model was 200 GPa, while that of bone was 20 GPa;
Poisson’s ratio was 0.30 for CoCrMo and bone. Three-dimensional hex
elements were seeded at intervals of approximately 1 mm.

The crossing-path motion produced on the ball surface was analyzed with the
slide track analysis developed by Saikko and Calonius
175
. Typical magnitude
and phasing of motion and load of the hip and lumbar spine during gait were
estimated from published biomechanical models and gait
studies
18,22,23,36,70,84,122,123,149,158,165,178-181,203-205,221-223,231
. The differences in motion
between the hip and disc during gait were represented by the different motion

110
waveforms, which were used to calculate the per-cycle distance the cup travels
relative to the ball, as well as to plot the path of motion of the slide tracks.
Flexion/extension, abduction/adduction or lateral bending, and axial rotation
waveforms were digitized. A marker point r fixed to the head was repeatedly
rotated from its initial position to a new position on the slide track. Each point of
the slide track corresponded to one set of rotation angles defined by the motion
waveforms; the slide track was drawn by connecting all points. Slide tracks and
their distances were calculated at 6 equidistant latitudes and 8 equidistant
longitudes. In the current study, the aspect ratios of the slide tracks were
calculated as the ratio of width to length of the slide track, with 0 corresponding
to a straight line and 1 corresponding to a circle. The sliding distance and
aspect ratio together determined the cross-path angles.

The lubrication between the surfaces was analyzed with Hamrock and Dowson’s
elastohydrodynamic lubrication equations
97
. Input parameters for the implant
were the ball radius, cup radius, average surface roughness value, Elastic
Modulus, and Poisson’s ratio. Input parameters for the body or simulated
environment were the normal applied load, angular velocity, lubricant viscosity,
and asymptotic isoviscous pressure. The input parameters were based on
typical values for metal-on-metal hips, except for the high value of the lubricant
viscosity (0.01 Pa s), which, as acknowledged by others
116,168,189,216
, was
necessary to facilitate the numerical solution. Based on an equivalent ball-on-
plane model of a ball-and-cup joint with these parameters, the lambda ratio ( λ)

111
was calculated to predict the lubrication regime. Thresholds for lubrication
regimes were 0-1 for boundary lubrication, 1-3 for mixed lubrication, and greater
than 3 for fluid film lubrication.

6.4 Results

The volumetric wear at each measurement interval is plotted separately for the
ball and cup for each test comparison in Figure 37 through Figure 39.

Volumetric Wear Over Time
0.0
0.5
1.0
1.5
2.0
2.5
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Volumetric Wear (mm^3)
1: Hip Orbital (Ball)
1: Hip Orbital (Cup)
2: Spine Orbital (Ball)
2: Spine Orbital (Cup)
0.32
0.85 0.14 0.07 0.07

Figure 37: Average volumetric wear and standard deviations for Tests 1 and 2. The circle
data points represent the Hip Orbital specimens, while the square data points represent the
Spine Orbital specimens. The solid lines represent the balls, while the dashed lines
represent the cups. For all specimens, the cups wore more than the balls. P-values for all
tests are listed at the top of the plot; the top row represents p-values between the ball and
cup for Test 1, while the bottom row represents those of Test 2.

112
Volumetric Wear Over Time
0.0
0.5
1.0
1.5
2.0
2.5
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Volumetric Wear (mm^3)
2: Spine Orbital (Ball)
2: Spine Orbital (Cup)
3: Spine Gait (Ball)
3: Spine Gait (Cup)
0.28 0.11 0.11 0.11
0.85 0.14 0.07 0.07

Figure 38: Average volumetric wear and standard deviations for Tests 2 and 3. The circle
data points represent the Spine Orbital specimens, while the square data points represent
the Spine Gait specimens. The solid lines represent the balls, while the dashed lines
represent the cups. For all specimens, the cups wore more than the balls. P-values for all
tests are listed at the top of the plot; the top row represents p-values between the ball and
cup for Test 2, while the bottom row represents those of Test 3.

113
Volumetric Wear Over Time
0.0
0.5
1.0
1.5
2.0
2.5
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Volumetric Wear (mm^3)
3: 14 mm (Ball)
3: 14 mm (Cup)
4: 2 mm (Ball)
4: 2 mm (Cup)
0.28 0.11 0.11 0.11
0.11 0.11 0.11 0.11

Figure 39: Average volumetric wear and standard deviations for Tests 3 and 4. The circle
data points represent the 14-mm specimens, while the square data points represent the 2-
mm specimens. The solid lines represent the balls, while the dashed lines represent the
cups. For all specimens, the cups wore more than the balls. P-values for all tests are listed
at the top of the plot; the top row represents p-values between the ball and cup for Test 3,
while the bottom row represents those of Test 4.

The volumetric wear rate was calculated as the difference in total (ball + cup)
volumetric wear divided by the number of cycles in the time interval. Volumetric
wear rates for all tests are shown in Figure 40.

114
Total Volumetric Wear Rate Over Time
(Per Million Cycles)
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
6.0
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Wear Rate
(mm^3 / 1M cycles)
1: Hip Orbital, 14 mm
2: Spine Orbital, 14 mm
3: Spine Gait, 14 mm
4: Spine Gait, 2 mm
0.65
0.28 1.00 0.41 0.59
0.28 0.65 0.11 0.11

Figure 40: Average volumetric wear rates and standard deviations. The Spine Orbital
specimens had a higher wear rate than the Hip Orbital specimens, the Spine Gait
specimens had a lower wear rate than the Spine Orbital specimens, and the 2-mm
specimens had a higher wear rate than the 14-mm specimens. P-values between tests are
listed at the top of the plot; the top row represents p-values Tests 1 and 2, the middle row
represents those between Test 2 and 3, and the bottom row represents those between
Tests 3 and 4.

The radius of the ball or cup of artificial hips is traditionally calculated by
measuring the coordinates of points on the surface, then fitting a surface to the
data points, using regression analysis. Radial clearance is then calculated by
subtracting the radius of the ball from the radius of the cup. The surface-fitting
algorithms typically require a 90° arc of data points for accuracy
208

(hemispherical cups form a 180° arc, as shown in Figure 41).

This method may not be accurate for the artificial disc the current study because
the ball only forms a 74° arc, and the cup only forms a 60° arc. Thus, in the
current study, instead of calculating clearance with the traditional method used

115
for hips, we calculated the lubrication gap, which was defined as the radial
distance between the ball and the cup at the edge of the cup (which, to be
consistent with formerly presented angles, would be at 180° for a hemispherical
cup but was calculated at 60° for all specimens of the current study to facilitate
comparison, as shown in Figure 41) when the apexes of the ball and cup are in
contact, as occurs physiologically. Lubrication gaps and the corresponding
clearances, which were extrapolated from the lubrication gaps, for all tests are
shown in Figure 42. Lubrication gap data was not available for Test 1.

Figure 41: Definition of angles. Lubrication gap area is shown in gray. For the artificial
discs of the current study, the lubrication gap was calculated at 60°.

116
Lubrication Gap Over Time
0
2
4
6
8
10
12
14
16
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Lubrication Gap (um)
2: Spine Orbital, 14 mm
3: Spine Gait, 14 mm
4: Spine Gait, 2 mm
Clearance (um)
0
16
32
48
64
80
0.59
0.59
0.59
0.11
0.11
0.11
0.59
0.11
0.59
0.11

Figure 42: Average lubrication gaps and standard deviations. Clearance was extrapolated
from the lubrication gap values and is represented with the y-axis on the right side. The
lubrication gaps of the Spine Orbital and Spine Gait 14-mm specimens stayed
approximately constant, while those of the 2-mm specimens increased and were ultimately
higher than those of the 14-mm specimens. P-values between tests are listed at the top of
the plot; the top row represents p-values Test 2 and 3, and the bottom row represents those
between Tests 3 and 4.

The mediolateral and anteroposterior surface roughness of the ball at each
measurement interval for all tests are shown in Figure 43 and Figure 44.
Surface roughness data was not available for Test 1.

117
Mediolateral Surface Roughness (Ra)
0.0
0.1
0.2
0.3
0.4
0.5
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Ra (um)
2: Spine Orbital, 14 mm
3: Spine Gait, 14 mm
4: Spine Gait, 2 mm
0.65
0.59
0.28
0.11
1.00
0.28
0.28
0.28
0.59
0.59

Figure 43: Average mediolateral surface roughness (Ra) and standard deviations. All
specimens stayed approximately constant, and the only discernible difference was the
higher surface roughness of the Spine Gait 14-mm specimens compared to the 2-mm
specimens. P-values between tests are listed at the top of the plot; the top row represents
p-values Test 2 and 3, and the bottom row represents those between Tests 3 and 4.

118
Anteroposterior Surface Roughness (Ra)
0.0
0.1
0.2
0.3
0.4
0.5
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Ra (um)
2: Spine Orbital, 14 mm
3: Spine Gait, 14 mm
4: Spine Gait, 2 mm
0.59
0.28
0.28
0.28
1.00
0.11
0.59
0.11
0.28
0.59

Figure 44: Average anteroposterior surface roughness (Ra) and standard deviations. All
specimens stayed approximately constant, and the only discernible difference was the
higher surface roughness of the Spine Gait 14-mm specimens compared to the 2-mm
specimens. P-values between tests are listed at the top of the plot; the top row represents
p-values Test 2 and 3, and the bottom row represents those between Tests 3 and 4.

The particle lengths for all tests are shown in Figure 45, and a sample SEM of
the particles is shown in Figure 46. Additionally, individual larger particles were
analyzed with EDAX and found to have particle lengths of up to 554 nm.
Particle length data was not available for Test 1.

119
Particle Length
0
20
40
60
80
100
120
140
160
180
200
Test 2 Test 3 Test 4
Length (nm)
n = 36
n = 74
n = 128

Figure 45: Particle lengths for all tests. The particle lengths were all approximately 120 nm,
with little difference between the three tests.

Figure 46: SEM of metal particles at a magnification of 25,000. The filter has 30 nm holes.

The surfaces of the specimens showed deep scratches in the direction of motion
and gouging (Figure 47), suggesting low lubrication and abrasive wear.
Additionally, a tribochemical reaction layer was observed.

120

A

B

C

Figure 47: Worn surfaces of the ball after 10
6
cycles for A) Test 2, B) Test 3, and C) Test 4.
For Test 2, the slide tracks were circular, and for Tests 2 and 3, the slide tracks were linear.

The contact pressure, sliding distance, aspect ratio, and lambda ratio calculated
from the analytical tools are shown in Table 20.

Table 20: Results of analytical tools
Test
Contact
Pressure
(MPa)
Sliding
Distance
(mm)
Aspect Ratio Lambda Ratio
1 217.6 29.40 0.85 3.64
2 193.7 5.47 0.89 1.20
3 193.7 7.21 0.29 1.45
4 226.1 7.21 0.29 1.45

6.5 Discussion

The wear rates of the metal-on-metal specimens in the current study ranged
from 0.8 – 2.1 mm
3
/10
6
cycles, which was within the range or slightly higher than
previously reported wear rates
49,147,157,190
of metal-on-metal artificial lumbar
discs, but lower than wear rates of metal-on-polyethylene artificial lumbar discs
tested under crossing-path motion, as shown in Table 21. For the phasing, tests

121
with motion and load in phase produced linear slide tracks, while tests with
motion and load out of phase, or frequency-shifted, produced crossing-path
motion. In the current study, motion and load were in phase.

Table 21: Wear rates of artificial lumbar discs
Material
Combination
Implant
Wear Rate
(mm
3
/10
6
cycles)
Test Conditions
Metal-on-Metal Maverick 1.20 – 1.40 Linear
Charité 0.12 Linear
Charité 0.14 Linear
Charité -0.02 Linear
Charité 20.37 Crossing-Path
Prodisc 0.07 Linear
Metal-on-
Polyethylene
Prodisc 17.46 Crossing-Path

For metal-on-metal hips, a period of accelerated run-in wear is typically followed
by a substantial decrease in wear rate tending toward steady-state values.
Rarely, a metal-on-metal hip exhibits runaway wear, in which the wear does not
reach a steady-state value but instead increases indefinitely to as much as 60.5
mm
3
/10
6
cycles
151
. The wear rates of a representative sample of wear
simulations of metal-on-metal hips that tested variables similar to those in the
current study
31,32,56,57,72,73,85,150,151,187,196,229,232
are shown in Table 22. The steady-
state wear rates were typically less than 1.0 mm
3
/10
6
cycles; outliers of this
trend were attributed to high motion or load
72,232
, small diameters
196
, first-
generation metal-on-metal bearings
150
, sintering heat treatment, or excessive
clearances
151
.

122
The wear rates of the metal-on-metal specimens in the current study (0.8 – 2.1
mm
3
/10
6
cycles) were comparable or slightly higher than the steady-state wear
rates of metal-on-metal hips, but lower than runaway wear rates. Additionally,
the trends of the volumetric wear were generally linear, with slightly shifting
slopes, and thus did not exhibit the same biphasic trend as shown for artificial
hip wear.

Table 22: Wear rates of metal-on-metal hips.

Variable
Tested
Run-In
Wear Rate
(mm
3
/10
6
cycles)
Steady-State
Wear Rate
(mm
3
/10
6
cycles)
Firkins et al (2001) Testing Profile 3.09 1.23
Williams et al (2004) Testing Profile 0.06 – 1.58
Goldsmith (2000) Diameter 0.36 – 0.45
Smith et al (2001) Diameter 1.62 0.54 – 6.30
Scholes (2001) Clearance 0.80 – 0.90 0.10 – 0.50
Dowson et al (2004)
Diameter
Clearance
0.79 – 3.23 0.09 – 0.17
McKellop et al (1996)
Diameter
Clearance
6.00
Medley et al (1996)
Diameter
Clearance
0.09 – 4.67
Chan et al (1996)
Diameter
Clearance
Material
0.20 – 8.00 0.25 – 0.60
Chan et al (1999)
Clearance
Material
0.02 – 1.90 0.03 – 0.21
Dowson et al (2004)
Clearance
Material
1.16 – 3.34 0.17 – 0.25
Firkins et al (2001) Material 0.02 – 0.32
Wang et al (1999) Material 0.05 – 0.32

These quantities are difficult to compare to each other with our limited
knowledge of artificial discs. Even with wear rates wear rates of metal-on-metal
hips as a comparison, the performance of the metal-on-metal specimens of the

123
current study is difficult to ascertain, due to the different biomechanics and
dimensions of the articulating surfaces. Perhaps more applicable descriptions of
the performance of the specimens are the lubrication and mode of wear. For
artificial hips, previous studies reported that metal-on-metal hips operate under
mixed lubrication
32,151,182,188,189
and typically experience abrasion (including
polishing and third-body abrasion), along with fatigue leading to pitting and
tribochemical reactions leading to a surface layer
5,13,16,107,150,198,225,226,229,233
. One
case study of the Maverick, a metal-on-metal lumbar disc, reported highly
polished surfaces similar to those observed on explanted metal-on-metal hips,
with microabrasion and focal microplasticity
125
. In the current study, the type of
wear produced by the specimens, as shown by the surface damage, was
consistent with low lubrication and abrasive wear. Additionally, a tribochemical
reaction layer was observed.

Another indication of the mode of wear is the particle morphology. For metal-on-
metal hips, the particle lengths can range from 30 – 300 nm
25,26,28,29,50,209
, but are
typically 10 – 50 nm
21
. In the current study, the particle lengths of the smaller
particles were approximately 120 nm, which fall in this range, and are thus
consistent with abrasive wear. However, individual larger particle lengths were
shown to be as much as 554 nm, which are larger than typical particles, and are
consistent with fatigue wear.

124
Typically, the running-in period of metal-on-metal artificial hips is less than 1.0 x
10
6
cycles, but can be as long as 4.0 x 10
6
cycles
5,31,32,56,57,73,151,171,187,196
, as
shown in Table 23. In the current study, each test had a duration of 1.0 x 10
6

cycles, which was chosen to incorporate run-in and steady-state wear rates
while retaining the integrity of the specimens for use in multiple tests as well as
maximizing the number of tests performed. The specimens were positioned in
the same orientation after measurements to minimize run-in wear; additionally,
the magnitude, trend, and mode of wear at each measurement interval
suggested that the metal-on-metal specimens of the current study did not
experience running-in wear. However, future tests with a longer duration may
provide more insight into this matter.

Table 23: Running-in periods for metal-on-metal hip wear simulations
Running-In Period (10
6
Cycles)
Medley et al (1996) 0.1 – 0.5
Chan et al (1996) 0.5
Scholes et al (2001) 0.5
Anissian et al (1999) 0.9
Chan et al (1999) 1.0
Firkins et al (2001) 1.0
Dowson et al (2004) 1.0
Dowson et al (2004) 1.5
Smith et al (2001) 2.0
Rieker et al (2005) 0.5 – 4.0

Tests 1 and 2: Hip vs. Spine Orbital Magnitudes

Test 1 tested the specimens under the Hip Orbital profile, to determine a
baseline FDA-approved wear rate of the specimens, while Test 2 tested the

125
specimens under the Spine Orbital profile, to determine how the wear rate
changes when the magnitudes of the motion and load are changed from that of
the hip to that of the lumbar spine during gait. The wear rate of the Spine Orbital
specimens was 2.4 times that of the Hip Orbital specimens, with the increase in
wear arising from an increase in both ball and cup wear. Furthermore, the cup
wore 2.5 times more than the ball for Spine Orbital specimens, and 4.0 times
more than the ball for Hip Orbital specimens, which may suggest different
mechanisms of wear for the Hip Orbital and Spine Orbital specimens.

The wear rates are consistent with the predictions based on cross-path angles
and lambda ratios. Although the aspect ratio of the slide tracks produced by the
Spine Orbital specimens was marginally larger than that produced by the Hip
Orbital specimens, the sliding distance was substantially smaller, indicating
smaller cross-path angles. In metal-on-metal bearings, crossing-path motion
has been shown to polish metal surfaces and decrease wear
210,215
. Thus, the
smaller cross-path angles of the Spine Orbital specimens predicted more wear
for the Spine Orbital specimens, which was consistent with the wear rates.
Additionally, the lambda ratio of the Spine Orbital specimens was substantially
smaller than that of the Hip Orbital specimens. A smaller lambda ratio indicates
a thinner lubricating layer of fluid between the surfaces, potentially increasing
wear. Thus, the smaller lambda ratio of the Spine Orbital specimens predicted
more wear for the Spine Orbital specimens, which was consistent with the wear
rates.

126

The wear rates were contrary to the predictions based on Archard’s equation,
which describes a proportional relationship between motion and wear
6
. The
sliding distance of the slide tracks produced by the Hip Orbital specimens was
larger than that produced by the Spine Orbital specimens, predicting more wear
for the Hip Orbital specimens by Archard’s equation; this was inconsistent with
the wear rates. This inconsistency emphasizes the importance of considering
the path of motion and velocity of the slide tracks, rather than just the
magnitude, and reveals the complexities of the mechanics of wear.

The surface roughness of the specimens did not have any discernible
differences in magnitude or trend. However, the cross-path angles of the slide
tracks produced by the Hip Orbital specimens were larger than those produced
by the Spine Orbital specimens, predicting more self-polishing of the alloy; this
was inconsistent with the surface roughness. This inconsistency may have
been a result of the differences in distribution of wear between the ball and the
cup; the surface roughness was measured only for the ball, while most of the
differences in wear were generated by the cup, for which the surface roughness
was not measured.

Tests 1 and 2 showed that the smaller magnitudes of motion and load
experienced by the Spine Orbital specimens, as compared to the Hip Orbital
specimens, increased wear. The increase in wear was most probably caused

127
by the smaller cross-path angles and decreased lubrication of the Spine Orbital
specimens.

Tests 2 and 3: Spine Orbital vs. Gait Phasing

Test 2 tested the specimens under the Spine Orbital profile, while Test 3 tested
the specimens under the Spine Gait profile, to determine how the wear rate
changes when the phasing of the motion changes from orbital to that typically
experienced by the lumbar spine during gait. The wear rate of the Spine Orbital
specimens was 2.6 times that of the Spine Gait specimens, with the increase in
wear arising from an increase in both ball and cup wear. The cup wore 2.5
times more than the ball for Spine Orbital specimens, and 3.3 times more than
the ball for Spine Gait specimens, which may suggest different mechanisms of
wear for the Spine Orbital and Spine Gait specimens.

The wear rates were consistent with the predictions based on lambda ratios.
Although the resultant lubrication regimes were the same, the lambda ratio of
the Spine Orbital specimens was smaller than that of the Spine Gait specimens,
predicting more wear for the Spine Orbital specimens; this was consistent with
the wear rates.

The wear rates were contrary to the predictions based on Archard’s equation.
The sliding distance of the slide tracks produced by the Spine Gait specimens

128
was larger than that produced by the Spine Orbital specimens, predicting more
wear for the Spine Gait specimens by Archard’s equation; this was inconsistent
with the wear rates. Again, the inconsistency emphasizes the importance of
considering more complexities of the mechanics of wear than just the
magnitude.

The wear rates were difficult to predict based on cross-path angles. Although
the sliding distance of the slide tracks produced by the Spine Gait specimens
was larger than that produced by the Spine Orbital specimens, the aspect ratio
was smaller; the resultant cross-path angles were difficult to determine.

The lubrication gaps and surface roughness of the specimens did not have any
discernible differences in magnitude or trend. Again, the resultant cross-path
angles were difficult to determine based on the sliding distance and aspect ratio.
One possibility, based on the surface roughness, is that the differences may
have balanced out to produce similar cross-path angles on both the Spine
Orbital and Spine Gait specimens.

Tests 2 and 3 showed that the differences in phasing of motion of the Spine
Orbital and Spine Gait profiles generated substantially different wear. The
differences in wear were most probably caused by the increased lubrication of
the Spine Gait specimens.

129
Tests 3 and 4: 14-mm vs. 2-mm Cup Depth

Test 3 tested specimens with the hemispherical cup depth of the original hip
specimens (14 mm), which is not possible for artificial discs due to space
constraints of the functional spine unit, while Test 4 tested the specimens cut
down to a cup depth typical of artificial discs (2 mm), to determine the effects of
cup depth on the wear rate of the specimens. The wear rate of the 2-mm
specimens was 2.5 times that of the 14-mm specimens, with the increase in
wear arising from an increase in both ball and cup wear. The cup wore 1.9
times more than the ball for 2-mm specimens, and 3.3 times more than the ball
for 14-mm specimens, which may suggest different mechanisms of wear for the
14-mm and 2-mm specimens. A finite-element analysis of the contact surface
area showed that the 2-mm specimens had marginally larger contact pressures
than those of the 14-mm specimens, predicting similar or marginally more wear
for the 2-mm specimens; this was consistent with the trend but not the
magnitude of the wear rates.

The lubrication gap of the 14-mm specimens remained relatively constant at 2-4
μm, while that of the 2-mm specimens increased to 8 μm, indicating a deviation
in conformity of the ball and cup. This may have been a result of the larger wear
rates of the 2-mm specimens.

130
The mediolateral and anteroposterior surface roughness of the 14-mm
specimens was higher than that of that of the 2-mm specimens by 0.05 μm.
This is contrary to the difference in wear rates, but may have been a result of the
differences in distribution of wear between the ball and the cup; the surface
roughness was measured only for the ball, while most of the differences in wear
were generated by the cup, which was not measured.

Tests 3 and 4 showed that reduction of cup depth to accommodate the space
constraints of the functional spine unit increased wear. The increase in wear
was most probably caused by the increased contact pressures in the material,
which was shown in the finite-element model, but not to the degree observed in
the wear simulation.

6.6 Conclusions

The wear rates of the metal-on-metal joints in the current study ranged from 0.8
– 2.1 mm
3
/10
6
cycles. Decreasing the magnitudes of motion and load from that
experienced by the hip during gait to that experienced by the lumbar spine
during gait increased the wear from 0.9 to 2.1 mm
3
/10
6
cycles, which showed
that a metal-on-metal joint placed in an environment with a small range of
motion may experience smaller cross-path angles and low lubrication, which
lead to increased wear.

131
Changing the phasing of motion from that of an orbital profile to that of a gait
profile reduced the wear from 2.1 to 0.8 mm
3
/10
6
cycles, which showed that the
phasing of motion in the lumbar spine has a larger effect on wear than that of
the hip. For artificial hips, accurate phasing of motion is not crucial for
reproducing in vivo wear, as evidenced by the use of bi-axial orbital simulators,
which are simpler to manufacture than tri-axial hip gait simulators; however, the
results of the current study showed that the phasing of motion substantially
effect wear of artificial lumbar discs.

Decreasing the cup depth from that of an artificial hip to that of an artificial
lumbar disc increased the wear from 0.8 to 2.1 mm
3
/10
6
cycles, which showed
that the reduction of cup depth to accommodate the space constraints of the
functional spine unit may increase material stresses and, therefore, wear.

P-values for the tests were only significant between Tests 3 and 4, when
reducing the cup depth. The standard deviations for all other tests were quite
high, in part due to the small specimen number; however, the trends between all
other tests were consistent with predictions, and thus require more samples to
make further conclusions.

Although articulating artificial lumbar discs generally use the same material
combinations and constructs that have been proven to be successful for artificial
hips, the biomechanical and anatomical differences between the joints were

132
shown to have an effect on wear. Therefore, careful consideration of the effects
of these differences is necessary during the design and testing of artificial
lumbar discs to avoid increased wear of the implants in vivo and in laboratory
wear simulations.

133
CHAPTER 7: EFFECTS OF CARBON CONTENT, CLEARANCE, AND DISC
FEATURES ON IN VITRO WEAR OF EXPERIMENTAL ARTIFICIAL DISCS

This chapter addresses the goals of Tests 7 – 9, which were to compare the
effects of metallurgy and cup design on wear of experimental metal-on-metal
artificial lumbar discs, in order to facilitate the design of artificial lumbar discs
(Specific Aim #6). The test labels in this chapter are not consistent with those of
the dissertation.

7.1 Introduction

The development of metal-on-metal total hip replacements has been advanced
in great part by studies that have measured the effects of individual independent
variables, such as the carbon content of the implant alloys
31,32,56,73,229
or the
diameter and clearance of the ball and cup
31,32,56,57,85,150,151,187,196
, on implant
wear. Such studies have typically used laboratory hip wear simulators to control
the testing conditions and, therefore, to isolate the effects of each variable of
interest on wear. This type of wear simulation is extensively used, and produces
results comparable to those observed in clinical retrievals and radiographic
studies.

In contrast to artificial hips, wear studies of artificial lumbar discs have not
examined the effects of independent design variables on wear. The existing

134
artificial lumbar disc wear publications have addressed wear in only the final
version of the product
49,125-127,147,157,190,207
. In these publications, the effects of
independent variables such as material, diameter, or clearance on wear were
not reported, although the importance of such variables has been established in
artificial hip wear studies. Additionally, there are few published retrieval studies
of artificial lumbar discs
125-127,207
against which to validate the results of artificial
disc laboratory wear simulations
49,147,157,190
.

The current study addresses some material and geometric design variables that
are applicable to metal-on-metal lumbar discs. Specifically, low and high
carbon-content materials, low and high clearances, keeled and un-keeled
endplates, and notched and un-notched cups were compared using spine wear
simulations of experimental metal-on-metal lumbar discs.

7.2 Materials

Variables Tested

The first variable tested in the current study was the material of the ball. In the
current study, the wear of experimental metal-on-metal lumbar discs with low
and high carbon-content balls was compared.

135
The other variables tested in the current study were geometric aspects of the
cup. The first geometric concern pertains to the fixation of the implant.
Specifically, keels may be added on the fixation surfaces to enhance implant
fixation. In the current study, the wear of experimental metal-on-metal lumbar
discs with and without a keel on the fixation surface of the cup were compared.

Another geometric concern for metal-on-metal implants is the radial clearance,
or difference in radii of the ball and cup. In the current study, the wear of
experimental metal-on-metal lumbar discs with small (10 – 35 μm) and large
(125 – 195 μm) clearances was compared.

Yet another geometric concern is a result of the adaptation of surgical
instrumentation to the space constraints. Specifically, features may be added to
prevent over-distraction of the disc space during implantation. In the current
study, the wear of experimental metal-on-metal lumbar discs with and without a
surgical notch in the cup was compared.

Specimens

Experimental metal-on-metal lumbar discs were provided by an orthopaedic
manufacturer. Each artificial disc specimen consisted of a caudal endplate, a
ball spacer with a diameter of 25.4 mm and a taper that locks into the caudal
endplate, and a cranial cup (Figure 48). Two generations of specimens were

136
customized for the current study, with the second generation consisting of
improvements made after analyzing the results of the first generation.

The first generation of specimens had balls made of low carbon-content (<0.1%)
CoCrMo alloy. The cups were made of high carbon-content (0.2 – 0.3%)
CoCrMo alloy, and had a notch cut out of the transverse plane to accommodate
the surgical instrumentation for implantation. The clearance between the ball
and the cup was large (125 – 175 μm), and the fixation surfaces of the endplate
and cup had an anteroposterior keel for fixation (Figure 49).

The second generation of specimens had balls made of high carbon-content
CoCrMo alloy. The cups were also made of high carbon-content CoCrMo alloy,
and were customized with and without a notch cut out of the transverse plane to
facilitate surgical implantation. The clearance between the ball and the cup was
small (10 – 35 μm); additionally, the anteroposterior keel was eliminated from
this generation (Figure 49).

Figure 48: Caudal endplate (left), ball spacer (middle), and cranial cup (right).

137

Figure 49: Keeled (upper-left) and un-keeled (upper-right) cups, and notched (lower-left)
and un-notched (lower-right) cups.

Wear Simulator

A three-station, bi-axial spine wear simulator was designed and constructed by
the Biomechanics Laboratory at Orthopaedic Hospital (Los Angeles, CA) (Figure
50). The wear simulator is capable of applying cyclic flexion/extension and axial
rotation, individually or coupled at any phase angle, under programmable axial
load. The specimens were lubricated with alpha calf serum (HyClone, Logan,
UT) with 0.2% sodium azide to retard bacterial degradation, and with 20mM
EDTA to prevent non-physiological precipitation of calcium salts onto the
bearing surfaces. The flexion/extension and load were run at 1 Hz and
synchronized with TestWare-SX (MTS Systems Corporation, Eden Prairie, MN).
The flexion/extension and axial rotation were synchronized with two H20 optical
encoders (BEI Technologies, Inc., Goleta, CA) providing input to a LabView 7.1

138
program combining two-signal edge separation with feedback control of voltage
output (National Instruments Corporation, Austin, TX). The implant and bulk
temperatures were recorded every 30 minutes with an 8018 Thermocouple Input
Module (SuperLogics, Inc., Waltham, MA) to ensure that the temperatures did
not rise above 50°C, at which point the proteins in the serum may denature
135
.

Figure 50: Spine wear simulator constructed by the Biomechanics Laboratory at
Orthopaedic Hospital (Los Angeles, CA). The cranial cup component experiences
flexion/extension, while the caudal ball component experiences axial rotation.

7.3 Methods

Schedule of Tests

A series of tests was developed to study the effects of the differences between
the first and second generation of specimens, as well as the effects of cup notch
within the second-generation specimens (Table 24), using the spine wear
simulator designed and constructed for the current study. Tests were labeled

139
with the generation (1 or 2), then alphabetically. Test 1 had a duration of 2.5 x
10
6
cycles, while Tests 2a–2b had a duration of 1.0 x 10
6
cycles.

Table 24: Schedule of tests for the current study.
Variables
Test
Ball Carbon Keel
Radial
Clearance
Cup
Notch
1 Low Keeled
Large
125 – 195 μm
Notched
2a Notched
2b
High Un-Keeled
Small
10 – 35 μm
Un-Notched

Tests 1 and 2a examined the effects of the differences between the first- and
second-generation specimens on wear; specifically, the carbon content,
presence of a keel, and radial clearance were tested. Test 1 tested specimens
with low carbon-content balls, keels on the fixation surfaces of the endplate and
cup, and large clearance, while Test 2a tested specimens with high carbon-
content balls, no keels, and small clearance, to determine the effects on wear.

Tests 2a and 2b examined the effects of implant geometry on wear. Test 2a
tested specimens with a notched cup surface, while Test 2b tested specimens
with an un-notched cup surface, to determine the effects of surgical notches on
wear.

140
Testing Profile

The Spine Bends profile, which is a motion and load waveform intended to
simulate worst-case wear, was generated for use in the bi-axial spine wear
simulator. The Spine Bends profile was generated from the approved ASTM
standard guide F 2423 – 05 (“Standard Guide for Functional, Kinematic, and
Wear Assessment of Total Disc Prostheses”), the approved ISO standard guide
ISO/DIS 18192 – 1 (“Loading and Displacement Parameters for Wear Testing
and Corresponding Environmental Conditions for Test”), and published gait
data, and applies extreme motions and loads representing significant bends.
The specimen was tested in the spine wear simulator and subjected to the
magnitude of flexion/extension (15°) and axial rotation (6°) the lumbar spine
experiences during significant bends, without synchronization. A sinusoidal
curve (300-1200 N) was applied as the load profile once per gait cycle (1 Hz).

Measurements

At each measurement interval, the specimens were ultrasonically cleaned,
following ASTM Standard F732, prior to measurements. Specimens were
weighed using an AT261 Precision Balance with an accuracy of 1.0 x 10
-5
g
(Mettler Electronics Corp., Anaheim, CA). Average roughness (Ra) of the balls
was measured with a Perthometer S8P laser profilometer fitted with a non-
contact laser probe and tracing length of 0.070 inches with an accuracy of 1.0 x
10
-3
μm (Mahr-Perthen-Gottingen, Germany/Feinpruf Corporation, Charlotte,

141
NC). The topographies of both bearing surfaces were measured in evenly
distributed matrices of 300 to 400 points using a BRT 504 Coordinate
Measurement Machine (CMM) (Mitutoyo USA, Aurora, IL) with a Renishaw TP-
200 touch probe and 2-mm ruby stylus with an accuracy of 1.0 x 10
-5
mm. Light
microscopy was performed using an MZ8 Optical Microscope (Leica Camera
AG, Solms, Germany). The serum was enzymatically digested to isolate the
particles, which were quantified using Metamorph 4.6r1 digital image analysis
(Molecular Devices Corp., Downingtown, PA) of the scanning electron
micrographs. An Independent-Samples T Test (SPSS Inc., Chicago, IL) was
used to calculate p-values.

7.4 Results

The volumetric wear at each measurement interval is plotted separately for the
ball and cup for each test comparison in Figure 51 and Figure 52.

142

Volumetric Wear Over Time
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
00.2
5
0.5 0.7
5
11.2
5
1.5 1.7
5
22.2
5
2.5 2.7
5
Time (Millions of Cycles)
Volumetric Wear (mm^3)
1: First Generation (Ball)
1: First Generation (Cup)
2a: Second Generation
(Ball)
2a: Second Generation
(Cup)

Figure 51: Volumetric wear for Tests 1 and 2a, First- vs. Second-Generation Specimens.
The circle data points represent the first-generation specimens, while the square data
points represent the second-generation specimens. The solid lines represent the balls,
while the dashed lines represent the cups. The trend for all data points is linear. For the
first-generation specimens, the ball wore significantly more than the cup (p<0.05 for all
intervals). For the second-generation specimens, the cup wore more than the ball, but not
significantly (p>0.05 for all intervals).

143

Volumetric Wear Over Time
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
00.25 0.5 0.7511.25
Time (Millions of Cycles)
Volumetric Wear (mm^3)
2a: Notched (Ball)
2a: Notched (Cup)
2b: Un-Notched (Ball)
2b: Un-Notched (Cup)

Figure 52: Volumetric wear for Tests 2a and 2b, Notched vs. Un-Notched Cups. The circle
data points represent the notched cup specimens, while the square data points represent
the un-notched cup specimens. The solid lines represent the balls, while the dashed lines
represent the cups. The trend for all data points is linear. For all specimens, the cups wore
more than the balls; this difference was not significant for the notched specimens (p>0.05
for all intervals), but significant for the un-notched specimens (p<0.05 for all intervals).

The volumetric wear rate was calculated as the difference in total (ball + cup)
volumetric wear divided by the number of cycles in the time interval. Volumetric
wear rates for all tests are shown in Figure 53.

144
Total Volumetric Wear Rate Over Time
(Per Million Cycles)
0.0
2.0
4.0
6.0
8.0
10.0
12.0
14.0
16.0
18.0
20.0
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75
Time (Millions of Cycles)
Wear Rate
(mm^3 / 1M cycles)
1: First Generation
2a: Notched
2b: Un-Notched
**+ *+

Figure 53: Average volumetric wear rates and standard deviations. The first-generation
specimens had a higher wear rate than all second-generation specimens. Within the
second-generation specimens, the notched cup specimens had a higher rate than the un-
notched cup specimens. The star (*) indicates p-values less than 0.05 between Tests 1 and
2a, and the cross (+) indicates p-values less than 0.05 between Tests 2a and 2b.

The radius of the ball or cup of artificial hips is traditionally calculated by
measuring the coordinates of points on the surface, then fitting a surface to the
data points, using regression analysis. Radial clearance is then calculated by
subtracting the radius of the ball from the radius of the cup. The surface-fitting
algorithms typically require a 90° arc of data points for accuracy
208

(hemispherical cups form a 180° arc, as shown in Figure 54).

This method may not be accurate for the artificial disc the current study because
the ball only forms a 74° arc, and the cup only forms a 60° arc. Thus, in the
current study, instead of calculating clearance with the traditional method used

145
for hips, we calculated the lubrication gap, which was defined as the radial
distance between the ball and the cup at the edge of the cup (which, to be
consistent with formerly presented angles, would be at 180° for artificial hips and
was at 60° for the artificial discs of the current study, as shown in Figure 54)
when the apexes of the ball and cup are in contact, as occurs physiologically.
Lubrication gaps and the corresponding clearances, which were extrapolated
from the lubrication gaps, are shown for all tests in Figure 55.

Figure 54: Definition of angles. Lubrication gap area is shown in gray. For the artificial
discs of the current study, the lubrication gap was calculated at 60°.

146
Lubrication Gap Over Time
0
5
10
15
20
25
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75
Time (Millions of Cycles)
Lubrication Gap (um)
1: First Generation
2a: Notched
2b: Un-Notched
Clearance (um)
0
40
160
120
80
200
240
**

Figure 55: Average lubrication gaps and standard deviations. Clearance was extrapolated
from the lubrication gap values and is represented with the y-axis on the right side. The
first-generation specimens had a decreasing lubrication gap, opposite to the increasing
lubrication gap of all second-generation specimens, and also had a larger initial lubrication
gap. Within the second-generation specimens, the notched cup specimens had an
increasing lubrication gap, similar in trend and magnitude to the un-notched cup
specimens. The star (*) indicates p-values less than 0.05 between Tests 1 and 2a, and the
cross (+) indicates p-values less than 0.05 between Tests 2a and 2b.

The mediolateral and anteroposterior surface roughness of the ball at each
measurement interval for all tests are shown in Figure 56 and Figure 57.

147
Mediolateral Surface Roughness (Ra)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75
Time (Millions of Cycles)
Ra (um)
1: First Generation
2a: Notched
2b: Un-Notched
*+ *

Figure 56: Average mediolateral surface roughness (Ra) and standard deviations. The first-
generation specimens had a lower maximum Ra than all second-generation specimens.
Within the second-generation specimens, the notched cup specimens had a lower
maximum Ra than the un-notched cup specimens. The star (*) indicates p-values less than
0.05 between Tests 1 and 2a, and the cross (+) indicates p-values less than 0.05 between
Tests 2a and 2b.

148
Anteroposterior Surface Roughness (Ra)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2 2.25 2.5 2.75
Time (Millions of Cycles)
Ra (um)
1: First Generation
2a: Notched
2b: Un-Notched

Figure 57: Average anteroposterior surface roughness (Ra) and standard deviations. The
first-generation specimens had a lower maximum Ra than all second-generation
specimens. Within the second-generation specimens, the notched cup specimens had a
lower maximum Ra than the un-notched cup specimens. The star (*) indicates p-values
less than 0.05 between Tests 1 and 2a, and the cross (+) indicates p-values less than 0.05
between Tests 2a and 2b.

For particle analysis, the metal pellets at the bottom of the centrifugation tubes
are shown in Figure 58.

A

B

C

Figure 58: Metal pellets for A) Test 1, B) Test 2a, and C) Test 2b. Test 1 has more visible
metal particles than Test 2a, which has more than Test 2b.

149
The particle lengths for all tests are shown in Figure 59, and a sample SEM of
the particles is shown in Figure 60. Additionally, individual larger particles were
analyzed with EDAX and found to have particle lengths of up to 311 nm.

Particle Length
0
20
40
60
80
100
120
140
160
180
200
Test 1 Test 2a Test 2b
Length (nm)
n = 51
n = 24
n = 52

Figure 59: Particle length for all tests. The particle lengths were all approximately 120 nm,
with little difference between the three tests.

150

Figure 60: SEM of metal particles at a magnification of 25,000. The filter has 30 nm holes.

The surfaces of the specimens showed plastic deformation, deep scratches in
the direction of motion, gouging, and pitting, as shown in Figure 61 and Figure
62; these are all common signs of abrasive or fatigue wear. Additionally, a
tribochemical reaction layer was observed.

A

B

Figure 61: Worn surfaces of the A) ball and B) cup after 1.0 x 10
6
cycles. The surfaces of
the specimens showed plastic deformation and deep scratches in the direction of motion,
which are common signs of abrasive or fatigue wear.

151

A

B
Figure 62: SEM of the ball, showing A) gouging and B) pitting, which are common signs of
abrasive or fatigue wear.

7.5 Discussion

The wear rates of the experimental artificial discs in the current study ranged
from 5.4 – 17.0 mm
3
/10
6
cycles, which is higher than previously reported wear
rates
49,147,157,190
of metal-on-metal artificial lumbar discs, but within the range of
wear rates of metal-on-polyethylene artificial lumbar discs tested under crossing-
path motion, as shown in Table 25. For the phasing, tests with motions and
loads in phase produced linear slide tracks, while tests with motions and loads
out of phase, or frequency-shifted, produced crossing-path motion. In the
current study, the motions and loads were out of phase.

152
Table 25: Wear rates of artificial lumbar discs
Material
Combination
Implant
Wear Rate
(mm
3
/10
6
cycles)
Test Conditions
Metal-on-Metal Maverick 1.20 – 1.40 Linear
Charité 0.12 Linear
Charité 0.14 Linear
Charité -0.02 Linear
Charité 20.37 Crossing-Path
Prodisc 0.07 Linear
Metal-on-
Polyethylene
Prodisc 17.46 Crossing-Path

For metal-on-metal hips, a period of accelerated run-in wear is typically followed
by a substantial decrease in wear rate tending toward steady-state values.
Rarely, a metal-on-metal hip exhibits runaway wear, in which the wear does not
reach a steady-state value but instead increases indefinitely to as much as 60.5
mm
3
/10
6
cycles
151
. The wear rates of a representative sample of wear
simulations of metal-on-metal hips that tested variables similar to those in the
current study
31,32,56,57,72,73,85,150,151,187,196,229,232
are shown in Table 26. The steady-
state wear rates were typically less than 1.0 mm
3
/10
6
cycles; outliers of this
trend were attributed to high loads or motions
72,232
, small diameters
196
, first-
generation metal-on-metal bearings
150
, sintering heat treatment, or excessive
clearances
151
.

The wear rates of the second generation of experimental artificial discs in the
current study (5.4 – 7.8 mm
3
/10
6
cycles) were higher than the steady-state wear
rates of metal-on-metal hips, but lower than runaway wear rates. Additionally,
the trends of the volumetric wear were linear, and thus did not exhibit the same

153
biphasic trend typical of metal-on-metal hips. In fact, in Test 1, which had a
duration of 2.5 x 10
6
cycles, the wear rate was steady, but then began to
increase after 1.5 x 10
6
cycles, which is in opposition to the trend observed for
metal-on-metal hips. For Test 1, the increase in wear rate was most likely a
result of the impingement of the non-bearing surfaces, which was observed at
approximately this measurement interval.

Table 26: Wear rates of metal-on-metal hips.

Variable
Tested
Run-In
Wear Rate
(mm
3
/10
6
cycles)
Steady-State
Wear Rate
(mm
3
/10
6
cycles)
Firkins et al (2001) Testing Profile 3.09 1.23
Williams et al (2004) Testing Profile 0.06 – 1.58
Goldsmith (2000) Diameter 0.36 – 0.45
Smith et al (2001) Diameter 1.62 0.54 – 6.30
Scholes (2001) Clearance 0.80 – 0.90 0.10 – 0.50
Dowson et al (2004)
Diameter
Clearance
0.79 – 3.23 0.09 – 0.17
McKellop et al (1996)
Diameter
Clearance
6.00
Medley et al (1996)
Diameter
Clearance
0.09 – 4.67
Chan et al (1996)
Diameter
Clearance
Material
0.20 – 8.00 0.25 – 0.60
Chan et al (1999)
Clearance
Material
0.02 – 1.90 0.03 – 0.21
Dowson et al (2004)
Clearance
Material
1.16 – 3.34 0.17 – 0.25
Firkins et al (2001) Material 0.02 – 0.32
Wang et al (1999) Material 0.05 – 0.32

These quantities are difficult to compare to each other with our limited
knowledge of artificial discs. Even with wear rates of metal-on-metal hips as a

154
comparison, the performance of artificial discs is difficult to ascertain, due to the
smaller contact surface area of the articulating surfaces. Perhaps more
applicable descriptions of the performance of artificial discs are the lubrication
and mode of wear. For artificial hips, previous studies reported that metal-on-
metal hips operate under mixed lubrication
32,151,182,188,189
and typically experience
abrasion (including polishing and third-body abrasion), along with fatigue leading
to pitting and tribochemical reactions leading to a surface
layer
5,13,16,107,150,198,225,226,229,233
. One case study of the Maverick, a metal-on-
metal lumbar disc, reported highly polished surfaces similar to those observed
on explanted metal-on-metal hips, with microabrasion and focal
microplasticity
125
. In the current study, the type of wear produced by the
experimental artificial discs, as shown by the surface damage, is consistent with
low lubrication and varying degrees of abrasive or fatigue wear. Additionally, a
tribochemical reaction layer was observed.

Another indication of the mode of wear is the particle morphology. For metal-on-
metal hips, the particle lengths can range from 30 – 300 nm
25,26,28,29,50,209
, but are
typically 10 – 50 nm
21
. In the current study, the particle lengths of the smaller
particles were approximately 120 nm, which fall in this range, and are thus
consistent with abrasive wear. However, individual larger particle lengths were
shown to be as much as 311 nm, which are larger than typical particles, and are
consistent with fatigue wear.

155
Typically, the running-in period of metal-on-metal hips is less than 1.0 x 10
6

cycles, but can be as long as 4.0 x 10
6
cycles
5,31,32,56,57,73,151,171,187,196
, as shown in
Table 27. In the current study, Test 1 had a duration of 2.5 x 10
6
cycles to
determine the running-in period of the specimens. As stated previously, the
wear rate was steady, but then began to increase after 1.5 x 10
6
cycles because
of the impingement of the non-bearing surfaces; no running-in period was
observed. Thus, a duration of 1.0 x 10
6
cycles was chosen for the remaining
tests to incorporate run-in and steady-state wear rates while maximizing the
number of tests performed. The specimens were positioned in the same
orientation after measurements to minimize run-in wear; additionally, the
magnitude, trend, and mode of wear at each measurement interval suggested
that the metal-on-metal specimens of the current study did not experience
running-in wear.

Table 27: Running-in periods for metal-on-metal hip wear simulations
Running-In Period (10
6
Cycles)
Medley et al (1996) 0.1 – 0.5
Chan et al (1996) 0.5
Scholes et al (2001) 0.5
Anissian et al (1999) 0.9
Chan et al (1999) 1.0
Firkins et al (2001) 1.0
Dowson et al (2004) 1.0
Dowson et al (2004) 1.5
Smith et al (2001) 2.0
Rieker et al (2005) 0.5 – 4.0

156
Tests 1 and 2a: First- vs. Second-Generation Specimens

Test 1 tested first-generation specimens with low carbon-content balls, keels on
the fixation surfaces of the endplate and cup, and large clearances, while Test
2a tested second-generation specimens with high carbon-content balls, no
keels, and small clearances, to determine the effects on wear. The wear rate of
the low carbon-content ball specimens was 1.5 times that of high-carbon content
ball specimens after 1.0 x 10
6
cycles (12.0 as opposed to 7.8 mm
3
/10
6
cycles).
This trend is in agreement with the results from metal-on-metal total hip
replacements, in which high carbon-content alloys wear less than low carbon-
content alloys
56,73,103,172,186,198,210
. This may be because low carbon-content
alloys comprise a single-phase structure, with larger grains and smaller carbides
than those in high carbon-content alloys. In contrast, high carbon-content alloys
demonstrate a biphasic structure, with smaller grains of the alloy surrounded by
larger scratch-resistant carbides
210,229
. The absence of larger scratch-resistant
carbides in the low carbon-content ball specimens may result in more wear than
for the high carbon-content ball specimens.

While the cup wore 2.5 times more than the ball for the high carbon-content ball
specimens, the ball wore as much as 24.8 times more than the cup for the low
carbon-content ball specimens. The proportions of ball and cup wear are not in
agreement with the results from the literature on the carbon content of alloys
used in artificial hips, which found that the cup wore 4 times more than the ball

157
for high carbon-content ball specimens, but the ball wore 4 times more than the
cup for low carbon-content ball specimens
73
. Based on the substantially higher
wear ratios of the low carbon-content ball specimens, high carbon-content balls
were used for further tests.

The lubrication gap of the low carbon-content ball specimens decreased over
time, indicating an increase in conformity of the ball and cup with increasing
wear. This may be because the absence of larger scratch-resistant carbides in
the low carbon-content ball may also result in an increase in conformity of the
ball to the cup. The increase in conformity of low carbon-content ball specimens
was contrary to the deviation from conformity of the high carbon-content ball
specimens. Thus, the larger lubrication gap of the worn high carbon-content ball
specimens may have increased lubrication, contributing to the lower wear rates
observed.

The surface roughness of the worn low carbon-content ball was substantially
lower than that of the high carbon-content ball, possibly because of the different
grain and carbide sizes in the alloy. The carbides may have been pulled out of
the alloy, leaving larger pits and resulting in higher surface roughness in the high
carbon-content specimens than in the low carbon-content specimens. Although
the rougher surface of the high carbon-content ball is not desirable because of
decreased lubrication
96,97
, overall the use of high carbon-content balls was
preferred.

158

The presence of keels may have contributed to the wear of the first-generation
specimens. Keels increase the rigidity of the cup, which results in increased
lubrication, which decreases wear. Additionally, the increased rigidity of the cup
may create an unyielding edge, which may lead to stripe wear. For metal-on-
metal hips, stripe wear has been observed at the impinging edge of the cup
74,232
,
which is typically located at the perimeter of the ball, and has been shown to
increase wear by threefold; however, in the current study, the edge of the cup
may cause stripe wear on the apex of the ball due to the geometry of the
artificial disc. Because the lubrication gap of the keeled specimens decreased
over time, indicating an increase in conformity of the ball and cup with increasing
wear, the advantage of including a keel (i.e., increased lubrication) was absent
in the current study, and, thus, outweighed by the possibility of increased stripe
wear. Furthermore, because the lubrication gap of the un-keeled specimens
increased over time, the un-keeled specimens still had a sizeable gap for
lubrication, with the added advantage of decreased stripe wear due to the
decreased rigidity of the specimens; thus, the keel was eliminated for further
tests.

The large initial clearance may have also contributed to the wear of the first-
generation specimens. The lubrication gap of the large-clearance specimens
decreased over time, indicating in increase in conformity of the ball and cup with
increasing wear. Thus, the smaller lubrication gap of the worn large-clearance

159
specimens may have decreased lubrication, contributing to the higher wear
rates observed. Additionally, hip wear simulator wear studies have reported that
smaller initial clearances reduce wear for metal-on-metal
hips
31,32,56,57,103,114,150,151,171,172,186,187
; however, some studies have reported that
clearances that are too small lead to equatorial clamping
114,151,172,186
; thus, an
optimum clearance for metal-on-metal total hip replacements is typically on the
order of 50 μm. Based on the decreased lubrication of the worn large-clearance
specimens, the advocacy of smaller initial clearances in literature on metal-on-
metal hips, and the availability of an improved manufacturing process for the
current study, smaller initial clearances that were closer to those of metal-on-
metal hips were used for further tests.

In summary, the changes from the first to the second generation of specimens
were the increase in carbon content of the ball from low to high, the elimination
of the anteroposterior keels, and the lowering of the initial clearance. Although
the individual effects of each of these variables was not clear, the combined
effects led us to make these changes for all remaining tests.

Tests 2a and 2b: Notched vs. Un-Notched Cup

Test 2a tested specimens with a notched cup surface, while Test 2b tested
specimens with an un-notched cup surface, to determine the effects of notches
on wear. The wear rate of the notched cup specimens was 1.3 times that of the

160
un-notched cup specimens (7.8 as opposed to 5.4 mm
3
/10
6
cycles), with most of
the differences arising from differences in cup wear. Furthermore, the cups of
the specimens wore 2.6 times more than the ball for the notched cup
specimens, and 2.2 times more than the ball for the un-notched cup specimens.
The higher wear rate and wear ratio of the notched cup specimens may be a
result of the smaller contact area, and thus higher material stresses, of the
notched cup. Another possibility is that the notch effectively created a second
edge, which caused stripe wear on the articulating surfaces.

Based on the slightly higher wear rates of the notched cup specimens in the
current study, the increase in accuracy of positioning the implant due to the
surgical feature should be weighed against the increase in wear from the
surgical feature, and further studies on each specific design is needed to make
an informed decision.

7.6 Conclusions

The wear rates of the experimental metal-on-metal lumbar discs in the current
study ranged from 5.4 – 17.0 mm
3
/10
6
cycles. However, changing the carbon
content of the ball from low to high, eliminating the anteroposterior keels, and
lowering the initial clearance reduced the wear from 12.0 to 7.8 mm
3
/10
6
.
Furthermore, removing the surgical notch reduced the wear from 7.8 to 5.4
mm
3
/10
6
cycles. Additionally, since the Spine Bends profile was a simulation of

161
worst-case wear, wear of the implants under spine gait conditions will likely
generate even less wear. These reductions in wear are reminiscent of the
minimization of wear achieved in metal-on-metal total hip replacements after an
extensive number of hip wear simulations that examined the effect of individual
parameters on wear. The results emphasize the need for artificial lumbar disc
wear simulations to examine some of the individual design parameters of
artificial lumbar discs, and for clinical results and retrievals of artificial lumbar
discs to validate these results.

The surface damage was generally consistent with low lubrication and varying
degrees of abrasive and fatigue wear. Although the specimens tested in the
current study were experimental, the results suggest that metal-on-metal lumbar
discs have the potential to produce wear of this magnitude and mechanism in
vivo. Therefore, careful consideration of individual design variables, such as
those considered in the current study, is necessary to avoid production of
excessive wear in artificial lumbar discs.

162
CHAPTER 8: COMPARISON OF GAIT AND HIGH-MOTION ACTIVITY ON IN
VITRO WEAR OF METAL-ON-METAL BEARINGS AND EXPERIMENTAL
ARTIFICIAL DISCS

This chapter addresses the goals of Tests 4 – 6, which were to compare the
effects of different motion and load profiles on wear of metal-on-metal artificial
lumbar discs, in order to optimize the wear test parameters for evaluating
artificial lumbar discs (Specific Aim #7). The test labels in this chapter are not
consistent with those of the dissertation.

8.1 Introduction

The laboratory in vitro simulation of wear of artificial hips has been validated with
extensive clinical and retrieval studies. In contrast, the availability of these
studies is limited for the relatively new artificial lumbar discs, and, thus, the
appropriate method of simulating wear of artificial lumbar discs has not been
established. One attempt at standardizing wear tests of artificial discs has been
the creation of the ASTM standard F 2423 – 05 (“Standard Guide for Functional,
Kinematic, and Wear Assessment of Total Disc Prostheses”); however, the
proposed method of cyclically applying the maximum range of motion of the
spine, which is a departure from the gait simulations typically used for artificial
hips, has not been validated.

163
The appropriate testing profile for the lumbar spine is not generally known. In
contrast, for artificial hips, the ranges of motion and load during gait are
generally known
165
, and have been widely used in hip wear
simulations
31,32,56,57,72,73,85,150,151,187,196,229,232
. Most hip wear simulations apply a
gait profile because walking is the predominant daily activity, and have been
successful in accurately reproducing the magnitude and mechanism of wear
observed in vivo. More recently, the modes of wear observed on clinical
retrievals has inspired research advocating the simulation of lower-frequency
events of high motion and load to more accurately depict the wear observed in
vivo; examples of these include microseparation
142,159,200,201,211
, impingement
104
,
and increased patient activity simulations
13
. The simulation of high motion and
load provides insight into the different possible wear mechanisms of artificial
hips, but is usually interspersed with normal gait cycles
13
, thus including both
types of motions in the proportions encountered in vivo.

In contrast, the appropriate testing profile for the lumbar spine is not generally
known; as a result, lumbar disc wear tests vary widely in the application and
interpretation of motion and load
49,147,157,190
. For wear simulation of artificial
lumbar discs, the approved ASTM standard guide has applied the concept of
simulating lower-frequency events of high motion and load, denoted as
“significant bends” in the guide. The ASTM-recommended profile accounts
solely for loading conditions created by significant bends, even though the
lumbar spine is subjected to far more repetitive loading cycles from walking (1 –

164
2 x 10
6
cycles per year
154,185,195,214
). Specifically, Morlock et al report “the most
frequent patient activity was sitting (44.3% of the time), followed by standing
(24.5%), walking (10.2%), lying (5.8%), and stair climbing (0.4%)
154
.” The
intended result of the ASTM standard guide is the simulation of worst-case
wear, but without concurrently simulating lower-motion but higher-repetition gait;
the effect of this method of simulation on wear has not been established.

The testing profile has been shown to make a difference on wear of metal-on-
polyethylene artificial lumbar discs. One study on artificial disc wear testing
profiles
157
tested the Charité and Prodisc-L metal-on-polyethylene artificial
lumbar discs under a frequency shifted, “cross-shear” motion profile as well as
under a curvilinear motion profile. The wear rate of the Charité increased from –
0.02 to 20.37 mm
3
/10
6
cycles, while the wear rate of the Prodisc-L increased
from 0.07 to 17.46 mm
3
/10
6
cycles; the results emphasize the importance of
testing profile on wear of artificial discs.

In the current study, the wear produced by gait simulation was compared to that
produced by a simulation of based on but modified from the ASTM definition of
significant bends, using two different metal-on-metal artificial lumbar discs.
Metal-on-metal artificial hips with a cup depth cut to that representative of
artificial lumbar discs were tested under gait conditions and under modified
significant bends, and experimental metal-on-metal artificial lumbar discs were
tested under the same two testing profiles.

165

To facilitate analysis of the results of the current study, two analytical tools were
used to explain the mechanics of wear: a slide track analysis to predict the
crossing-path motion distribution, and elastohydrodynamic lubrication equations
to predict the lubrication regime.

Range of Motion

The range of motion recommended by the ASTM is significantly larger than that
experienced by the lumbar spine during gait, and is in fact closer to the
maximum range of motion of the lumbar spine. Physiologically, an ASTM-
defined significant bend would represent maximally flexing and extending the
lumbar spine 125,000 times per year, or approximately 340 times per day, which
is likely not representative of patients receiving spinal disc arthroplasty. The
ASTM standard guide recommends 15° of flexion/extension, 12° of lateral
bending, and 6° of axial rotation. The maximum ranges of in vivo motion of the
lumbar spine
137
are shown in Figure 63, while the ranges of motion during
gait
70,149,179,203-205,221,222,231
are shown in Table 28; the axial rotation range of
motion during gait was not available. The approximate in vivo range of motion
of flexion/extension during gait is 4°, while that of lateral bending during gait is
8.5°.

166
0 5 10 15 20 25
L5-S1
L4-5
L3-4
L2-3
L1-2
T12-L1
Functional Spine Unit
Range of Motion (°)
Flexion/Extension
Lateral Bending
Axial Rotation

Figure 63: In vivo maximum range of motion of the spine

Table 28: In vivo range of motion of the lumbar spine during gait
Author
Flexion/Extension
(°)
Lateral Bending
(°)
Taylor (1996) 3.24±0.95 12.84±3.00
Willems (1997) 1.81±0.79 N/A
McGowan (1998) 5.44±1.07 8.20±2.94
Taylor (1999) 3.21±0.68 3.46±1.32
Sartor (1999) 2 12
Feipel (2001) 5±1 9±2
Vogt (2001) 3.77±1.61 8.08±1.63
Vogt (2002) 4.4 3.9
Taylor (2003) 3.4±1.6 10.2±3.1

Load

The maximum load recommended by the ASTM standard guide is larger than
that experienced by the lumbar spine during gait. The ASTM standard guide
recommends either a constant 1200 N compressive load on the specimens, or a
sinusoidal compressive load ranging from 600 to 1800 N. The in vivo
compressive load is sinusoidal. The minimum and maximum compressive load

167
of the lumbar spine during gait
18,22,23,36,84,122,123
are shown in Table 29. The
approximate mean of the reported minimum load is 670 N, while that of the
reported maximum load is 1200 N.

Table 29: In vivo maximum compressive load of the lumbar spine during gait
Study
Minimum Load
(times BW)
Maximum Load
(times BW)
Cappozzo (1983) 0.3-0.9
Cappozzo (1984) 0.2-0.8 1.0-2.5
Cromwell (1989) 1.2
Khoo (1994) 1.4 2.1
Khoo (1995) 1.7±0.3
Goh (1998) 1.5
Callaghan (1999)
1.2±0.2
0.4±0.1
2.2±0.5
0.8±0.1

Frequency and Phase Angle

The frequency and phasing of the motion and load are not specified in the
ASTM standard guide, but have been reported to varying degrees in published
gait studies. The ASTM standard guide does not recommend a frequency of the
sinusoidal load, and recommends that the user determine the phase angle of the
motion. In vivo, the frequency of flexion/extension in the lumbar spine is twice
that of the gait cycle frequency, while the frequency of axial rotation in the
lumbar spine is the same as the gait cycle frequency. Additionally, the
frequency of compressive load is the same as that of flexion/extension
158,179-
181,223
. An estimation of the frequency and phasing of motion and load during
gait, based on gait studies, is shown in Figure 64.

168

Figure 64: In vivo frequencies and phasing of motion and load during gait. The solid line
represents flexion/extension, the dashed line represents lateral bending, the dashdot line
represents axial rotation, and the dotted line represents the compressive load.

8.2 Materials

Specimens

Two types of experimental metal-on-metal specimens were tested in the current
study, denoted as Disc A and Disc B. Disc A was an experimental artificial
lumbar disc made by an orthopaedic manufacturing company. Disc B was made
by an orthopaedic manufacturing company, and was originally designed as an
artificial hip, but modified by Orthopaedic Hospital (Los Angeles, CA) to function
as an artificial lumbar disc by cutting the cup depth.

Six specimens of Disc A were provided by an orthopaedic manufacturing
company. Each specimen consisted of a caudal endplate, a ball spacer with a
diameter of 25.4 mm and a taper that locks into the caudal endplate, and a

169
cranial cup with a cup depth of 2 mm (Figure 65), all made of high carbon-
content CoCrMo alloy. The cups had a notch cut out of the transverse plane to
facilitate surgical implantation. New specimens were used for each test.

Figure 65: Disc B specimens: caudal endplate (left), ball spacer (middle), and cranial cup
(right).

For Disc B specimens, four artificial hip specimens were provided by Sulzer
(Sulzer Ltd., Winterthur, Switzerland), and modified to function as artificial
lumbar discs by Orthopaedic Hospital by cutting the original cup depth to 2 mm
with a lathe. Each specimen consisted of a 28-mm ball and a cup, both made of
high carbon-content CoCrMo alloy (Figure 66). The same specimens were used
for all tests.

170

Figure 66: Disc B specimens: ball (left) and cup (right).

Wear Simulator

The specimens were tested in a spine wear simulator that was designed and
constructed by the Biomechanics Laboratory at Orthopaedic Hospital (Los
Angeles, CA). The three-station, bi-axial spine wear simulator is capable of
applying cyclic flexion/extension and axial rotation, individually or coupled at any
phase angle, under programmable axial load. The specimens were lubricated
with alpha calf serum (HyClone, Logan, UT) with 0.2% sodium azide to retard
bacterial degradation, and with 20mM EDTA to prevent non-physiological
precipitation of calcium salts onto the bearing surfaces. The flexion/extension
and load were run at 1 Hz and synchronized with TestWare-SX (MTS Systems
Corporation, Eden Prairie, MN). The flexion/extension and axial rotation were
synchronized with two H20 optical encoders (BEI Technologies, Inc., Goleta,
CA) providing input to a LabView 7.1 program combining two-signal edge
separation with feedback control of voltage output (National Instruments
Corporation, Austin, TX). The implant and bulk temperatures were recorded

171
every 30 minutes with an 8018 Thermocouple Input Module (SuperLogics, Inc.,
Waltham, MA) to ensure that the temperatures did not rise above 50°C, at which
point the proteins in the serum may denature
135
.

8.3 Methods

Testing Profiles

Two sets of motion and load waveforms were generated for use in a bi-axial
spine wear simulator: the Spine Gait profile and the Spine Bends profile. The
Spine Gait profile was generated from published gait
data
18,22,23,36,70,84,122,123,149,158,179-181,203-205,221-223,231
, and applies typical motion and
load experienced by the lumbar spine during gait. The specimen was tested in
the spine wear simulator and subjected to the magnitude of flexion/extension
(8.5°) and axial rotation (6°) the lumbar spine experiences during gait, with the
phasing of bi-axial motions that most closely produces the slide tracks produced
by tri-axial gait. A scaled-down discretized Paul curve with the magnitude of
compressive load (85-1200 N) the lumbar spine experiences during gait was
applied as the load profile twice per gait cycle, representing the left and right
steps of each gait cycle (1 Hz), as shown in Figure 67.

172

Figure 67: Spine Gait profile. The solid line represents flexion/extension, the dashed line
represents lateral bending, the dashdot line represents axial rotation, and the dotted line
represents the compressive load.

The Spine Bends profile was generated from the approved ASTM standard
guide, the approved ISO standard ISO/DIS 18192 – 1 (“Loading and
Displacement Parameters for Wear Testing and Corresponding Environmental
Conditions for Test”), and the published gait data mentioned previously, and
applies high motion with unsynchronized phasing, representing significant
bends. The specimen was tested in the spine wear simulator and subjected to
the magnitude of flexion/extension (15°) and axial rotation (6°) the lumbar spine
experiences during significant bends, without synchronization. A sinusoidal
curve (300-1200 N) was applied as the load profile once per gait cycle (1 Hz), as
shown in Figure 68; however, the phasing of the motion was not synchronized
and therefore not necessarily as shown in the figure.

173

Figure 68: Spine Bends profile. The solid line represents flexion/extension, the dashed line
represents lateral bending, the dashdot line represents axial rotation, and the dotted line
represents the compressive load. The phasing of the motion was not synchronized and
therefore not necessarily as shown in the figure.

Schedule of Tests

A series of tests was developed to study the effects of the test profiles on wear
of metal-on-metal artificial joints (Table 30), using a spine wear simulator
developed for the current study. Each test had a duration of 1.0 x 10
6
cycles.

Table 30: Schedule of tests for the current study.
Testing Profile
Test Specimen Spine
Gait
Spine
Bends
1 A X
2 A X
3 B X
4 B X

Tests 1 and 2 will examine the effects of the magnitudes, frequencies, and
phasing of motion and load on wear of Disc A specimens. Test 1 tested the
Disc A specimens under the Spine Gait profile, while Test 2 tested them under

174
the Spine Bends profile, to determine how the wear rate changes when the
magnitudes, frequency, and phasing of the motion and load are changed from
those typically experienced by the lumbar spine during gait to those experienced
by the lumbar spine during significant bends.

Tests 3 and 4 will examine the effects of the magnitudes, frequency, and
phasing of motion and load on wear of Disc B specimens. Test 3 tested the
Disc B specimens under the Spine Gait profile, while Test 4 tested them under
the Spine Bends profile, to determine how the wear rate changes when the
magnitudes, frequency, and phasing of the motion and load are changed from
those typically experienced by the lumbar spine during gait to those experienced
by the lumbar spine during significant bends.

Measurements

At each measurement interval, the specimens were ultrasonically cleaned,
following ASTM Standard F732, prior to measurements. Specimens were
weighed using an AT261 Precision Balance with an accuracy of 1.0 x 10
-5
g
(Mettler Electronics Corp., Anaheim, CA). Average roughness (Ra) of the balls
were measured with a Perthometer S8P laser profilometer fitted with a non-
contact laser probe and tracing length of 0.070 inches for the ball, or a diamond-
tipped stylus with a tracing length of 0.022 inches for the cup, with an accuracy
of 1.0 x 10
-3
μm (Mahr-Perthen-Gottingen, Germany/Feinpruf Corporation,

175
Charlotte, NC). The topographies of both bearing surfaces were measured in
evenly distributed matrices of 300 to 400 points using a BRT 504 Coordinate
Measurement Machine (CMM) (Mitutoyo USA, Aurora, IL) with a Renishaw TP-
200 touch probe and 4-mm ruby stylus for specimens with a cup depth of 14
mm, or a 2-mm ruby stylus for specimens with a cup depth of 2 mm, with an
accuracy of 1.0 x 10
-5
mm. Light microscopy was performed using an MZ8
Optical Microscope (Leica Camera AG, Solms, Germany). An Independent-
Samples T Test (SPSS Inc., Chicago, IL) was used to calculate p-values of Disc
A specimens, while a Wilcoxin Signed Ranks Test of related samples was used
to calculate p-values of Disc B specimens.

Analytical Tools

To facilitate analysis of the results, two analytical tools were used to explain the
mechanics of wear: a slide track analysis to predict the crossing-path motion
distribution, and elastohydrodynamic lubrication equations to predict the
lubrication regime.

The crossing-path motion produced on the ball surface was analyzed with the
slide track analysis developed by Saikko and Calonius
175
. Typical magnitude
and phasing of motion and load of the hip and lumbar spine during gait were
estimated from published biomechanical models and gait
studies
18,22,23,36,70,84,122,123,149,158,165,178-181,203-205,221-223,231
. The differences in motion

176
between the hip and disc during gait were represented by the different motion
waveforms, which were used to calculate the per-cycle distance the cup travels
relative to the ball, as well as to plot the path of motion of the slide tracks.
Flexion/extension, abduction/adduction or lateral bending, and axial rotation
waveforms were digitized. A marker point r fixed to the head was repeatedly
rotated from its initial position to a new position on the slide track. Each point of
the slide track corresponded to one set of rotation angles defined by the motion
waveforms; the slide track was drawn by connecting all points. Slide tracks and
their distances were calculated at 6 equidistant latitudes and 8 equidistant
longitudes. In the current study, the aspect ratios of the slide tracks were
calculated as the ratio of width to length of the slide track, with 0 corresponding
to a straight line and 1 corresponding to a circle. The sliding distance and
aspect ratio together determined the cross-path angles.

The lubrication between the surfaces was analyzed with Hamrock and Dowson’s
elastohydrodynamic lubrication equations
97
. Input parameters for the implant
were the ball radius, cup radius, average surface roughness value, Elastic
Modulus, and Poisson’s ratio. Input parameters for the body or simulated
environment were the normal applied load, angular velocity, lubricant viscosity,
and asymptotic isoviscous pressure. The input parameters were based on
typical values for metal-on-metal hips, except for the high value of the lubricant
viscosity (0.01 Pa s), which, as acknowledged by others
116,168,189,216
, was
necessary to facilitate the numerical solution. Based on an equivalent ball-on-

177
plane model of a ball-and-cup joint with these parameters, the lambda ratio ( λ)
was calculated to predict the lubrication regime. Thresholds for lubrication
regimes were 0-1 for boundary lubrication, 1-3 for mixed lubrication, and greater
than 3 for fluid film lubrication.

8.4 Results

The volumetric wear at each measurement interval is plotted separately for the
ball and cup for each test comparison in Figure 69 and Figure 70.

178
Volumetric Wear Over Time
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Volumetric Wear (mm^3)
1: Spine Gait (Disc A Ball)
1: Spine Gait (Disc A Cup)
2: Spine Bends (Disc A
Ball)
2: Spine Bends (Disc A
Cup)
0.18 0.14 0.06
0.53 0.57 0.52

Figure 69: Volumetric wear for Tests 1 and 2, Disc A specimens. The circle data points
represent the Spine Gait specimens, while the square data points represent the Spine
Bends specimens. The solid lines represent the balls, while the dashed lines represent the
cups. For the Spine Gait specimens, the ball wore marginally more than the cup. For the
Spine Bends specimens, the cup wore substantially more than the cup. P-values for all
tests are listed at the top of the plot; the top row represents p-values between the ball and
cup for Test 1, and the bottom row represents those for Test 2.

179

Volumetric Wear Over Time
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
00.25 0.5 0.7511.25
Time (Millions of Cycles)
Volumetric Wear (mm^3)
3: Spine Gait (Disc B Ball)
3: Spine Gait (Disc B Cup)
4: Spine Bends (Disc B
Ball)
4: Spine Bends (Disc B
Cup)
0.28 0.11 0.11
0.11 0.11 0.11
0.28
0.11

Figure 70: Volumetric wear for Tests 3 and 4, Disc B specimens. The circle data points
represent the Spine Gait specimens, while the square data points represent the Spine
Bends specimens. The solid lines represent the balls, while the dashed lines represent the
cups. For all specimens, the cup wore substantially more than the cup. P-values for all
tests are listed at the top of the plot; the top row represents p-values between the ball and
cup for Test 3, and the bottom row represents p-values between the ball and cup for Test 4.

The volumetric wear rate was calculated as the difference in total (ball + cup)
volumetric wear divided by the number of cycles in the time interval. Volumetric
wear rates for all tests are shown in Figure 71.

180
Total Volumetric Wear Rate Over Time
(Per Million Cycles)
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Wear Rate
(mm^3 / 1M cycles)
1: Spine Gait, Disc A
2: Spine Bends, Disc A
3: Spine Gait, Disc B
4: Spine Bends, Disc B
0.11 0.11 0.11 0.11
0.02 0.00 0.00

Figure 71: Volumetric wear rates. For both Disc A and Disc B specimens, the Spine Bends
specimens had a higher wear rate than the Spine Gait specimens. P-values between tests
are listed at the top of the plot; the top row represents p-values between Tests 1 and 2, and
the bottom row represents p-values between Tests 3 and 4.

The radius of the ball or cup of artificial hips is traditionally calculated by
measuring the coordinates of points on the surface, then fitting a surface to the
data points, using regression analysis. Radial clearance is then calculated by
subtracting the radius of the ball from the radius of the cup. The surface-fitting
algorithms typically require a 90° arc of data points for accuracy
208

(hemispherical cups form a 180° arc, as shown in Figure 72).

This method may not be accurate for the artificial disc the current study because
the ball only forms a 74° arc, and the cup only forms a 60° arc. Thus, in the
current study, instead of calculating clearance with the traditional method used
for hips, we calculated the lubrication gap, which was defined as the radial

181
distance between the ball and the cup at the edge of the cup (which, to be
consistent with formerly presented angles, would be at 180° for a hemispherical
cup but was calculated at 60° for all specimens of the current study to facilitate
comparison, as shown in Figure 72) when the apexes of the ball and cup are in
contact, as occurs physiologically. Lubrication gaps and the corresponding
clearances, which were extrapolated from the lubrication gaps, for all tests are
shown in Figure 73. Lubrication gap data was not available for Test 1.

Figure 72: Definition of angles. Lubrication gap area is shown in gray. For both Disc A and
Disc B specimens, the lubrication gap was calculated at 60°.

182
Lubrication Gap Over Time
0
2
4
6
8
10
12
14
16
18
20
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Lubrication Gap (um)
1: Spine Gait, Disc A
2: Spine Bends, Disc A
3: Spine Gait, Disc B
4: Spine Bends, Disc B
Clearance
0
32
64
96
128
160
0.02
0.11
1.00
0.11
0.51
0.11 1.00
0.41
0.11

Figure 73: Lubrication gap. Clearance was extrapolated from the lubrication gap values
and is represented on the y-axis on the right side. For both Disc A and Disc B specimens,
the lubrication gaps of all specimens generally increased. For both Disc A and Disc B
specimens, the lubrication gaps were larger for the Spine Bends specimens than for the
Spine Gait specimens. P-values between tests are listed at the top of the plot; the top row
represents p-values between Tests 1 and 2, and the bottom row represents p-values
between Tests 3 and 4.

183
The mediolateral and anteroposterior surface roughness of the ball at each
measurement interval for all tests are shown in Figure 74 and Figure 75.

Mediolateral Surface Roughness (Ra)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Ra (um)
1: Spine Gait, Disc A
2: Spine Bends, Disc A
3: Spine Gait, Disc B
4: Spine Bends, Disc B
0.00
0.59
0.20
0.41
0.01
0.11 0.28
0.02
0.28

Figure 74: Mediolateral average surface roughness. For Disc A specimens, the surface
roughness for the Spine Bends specimens was substantially higher than for the Spine Gait
specimens, while for Disc B specimens, the surface roughness for the Spine Gait
specimens was marginally higher than for the Spine Bends specimens. P-values between
tests are listed at the top of the plot; the top row represents p-values between Tests 1 and
2, and the bottom row represents p-values between Tests 3 and 4.

184
Anteroposterior Surface Roughness (Ra)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0 0.25 0.5 0.75 1 1.25
Time (Millions of Cycles)
Ra (um)
1: Spine Gait, Disc A
2: Spine Bends, Disc A
3: Spine Gait, Disc B
4: Spine Bends, Disc B
0.01
0.59
0.26
0.11
0.06
0.28 0.28
0.59
0.59

Figure 75: Anteroposterior average surface roughness. For Disc A specimens, the surface
roughness for the Spine Bends specimens was substantially higher than for the Spine Gait
specimens, while for Disc B specimens, the surface roughness for the Spine Gait
specimens was marginally higher than for the Spine Bends specimens. P-values between
tests are listed at the top of the plot; the top row represents p-values between Tests 1 and
2, and the bottom row represents p-values between Tests 3 and 4.

The sliding distance, aspect ratio, and lambda ratio calculated from the
analytical tools are shown in Table 31.

Table 31: Results of analytical tools

Sliding Distance
(mm)
Aspect Ratio Lambda Ratio
Spine Gait 7.21 0.29 1.45
Spine Bends 6.62 0.24 1.37

185
8.5 Discussion

The wear rates of the metal-on-metal specimens in the current study ranged
from 1.5 – 7.8 mm
3
/10
6
cycles, which was slightly to substantially higher than
previously reported wear rates
49,147,157,190
of metal-on-metal artificial lumbar
discs, but lower than wear rates of metal-on-polyethylene artificial lumbar discs
tested under crossing-path motion, as shown in Table 32. For the phasing, tests
with motion and load in phase produced curvilinear slide tracks, while tests with
motion and load out of phase, or frequency-shifted, produced crossing-path
motion. In the current study, motion and load were in phase for the Spine Gait
profile, but out of phase for the Spine Bends profile.

186
Table 32: Wear rates of artificial lumbar discs
Motion (°)
Implant
Load
(N)
Flexion/
Extension
Lateral
Bending
Rotation
Phasing
Wear
Rate
(mm
3
/10
6

cycles)
Mathews
et al
(2004)
Maverick N/A N/A N/A N/A In 1.2 – 1.4
Serhan
et al
(2005)
Charité
900 –
1850
±7.5 0 ±1.5 In 0.12
Dooris et
al (2005)
Charité
900 –
1850
±7.5 0 ±1.5 In 0.14
Charité
200 –
1750
-3, +6 0 ±1.5 In -0.02
Charité
200 –
1750
-3, +6 ±2 ±1.5 Out 20.37
Prodisc
200 –
1750
-3, +6 0 ±1.5 In 0.07
Nechtow
et al
(2006)
Prodisc
200 –
1750
-3, +6 ±2 ±1.5 Out 17.46

For metal-on-metal hips, a period of accelerated run-in wear is typically followed
by a substantial decrease in wear rate tending toward steady-state values.
Rarely, a metal-on-metal hip exhibits runaway wear, in which the wear does not
reach a steady-state value but instead increases indefinitely to as much as 60.5
mm
3
/10
6
cycles
151
. The wear rates of a representative sample of wear
simulations of metal-on-metal hips that tested variables similar to those in the
current study
31,32,56,57,72,73,85,150,151,187,196,229,232
are shown in Table 33. The steady-
state wear rates were typically less than 1.0 mm
3
/10
6
cycles; outliers of this
trend were attributed to high motion or load
72,232
, small diameters
196
, first-
generation metal-on-metal bearings
150
, sintering heat treatment, or excessive
clearances
151
.

187

The wear rates of the metal-on-metal specimens in the current study (1.5 – 7.8
mm
3
/10
6
cycles) were higher than the steady-state wear rates of metal-on-metal
hips, but lower than runaway wear rates. Additionally, the trends of the
volumetric wear were generally linear, with slightly shifting slopes, and thus did
not exhibit the same biphasic trend as shown for artificial hip wear.

Table 33: Wear rates of metal-on-metal hips.

Variable
Tested
Run-In
Wear Rate
(mm
3
/10
6
cycles)
Steady-State
Wear Rate
(mm
3
/10
6
cycles)
Firkins et al (2001) Testing Profile 3.09 1.23
Williams et al (2004) Testing Profile 0.06 – 1.58
Goldsmith (2000) Diameter 0.36 – 0.45
Smith et al (2001) Diameter 1.62 0.54 – 6.30
Scholes (2001) Clearance 0.80 – 0.90 0.10 – 0.50
Dowson et al (2004)
Diameter
Clearance
0.79 – 3.23 0.09 – 0.17
McKellop et al (1996)
Diameter
Clearance
6.00
Medley et al (1996)
Diameter
Clearance
0.09 – 4.67
Chan et al (1996)
Diameter
Clearance
Material
0.20 – 8.00 0.25 – 0.60
Chan et al (1999)
Clearance
Material
0.02 – 1.90 0.03 – 0.21
Dowson et al (2004)
Clearance
Material
1.16 – 3.34 0.17 – 0.25
Firkins et al (2001) Material 0.02 – 0.32
Wang et al (1999) Material 0.05 – 0.32

These quantities are difficult to compare to each other with our limited
knowledge of artificial discs. Even with wear rates wear rates of metal-on-metal
hips as a comparison, the performance of the metal-on-metal specimens of the

188
current study is difficult to ascertain, due to the different biomechanics and
dimensions of the articulating surfaces. Perhaps more applicable descriptions of
the performance of the specimens are the lubrication and mode of wear. For
artificial hips, previous studies reported that metal-on-metal hips operate under
mixed lubrication
32,151,182,188,189
and typically experience abrasion (including
polishing and third-body abrasion), along with fatigue leading to pitting and
tribochemical reactions leading to a surface layer
5,13,16,107,150,198,225,226,229,233
. One
case study of the Maverick, a metal-on-metal lumbar disc, reported highly
polished surfaces similar to those observed on explanted metal-on-metal hips,
with microabrasion and focal microplasticity
125
. In the current study, the type of
wear produced by Disc A specimens was consistent with low lubrication and
varying degrees of abrasive and fatigue wear, while the type of wear produced
by Disc B specimens was consistent with low lubrication and moderate abrasive
wear.

Typically, the running-in period of metal-on-metal artificial hips is less than 1.0 x
10
6
cycles, but can be as long as 4.0 x 10
6
cycles
5,31,32,56,57,73,151,171,187,196
, as
shown in Table 34. In the current study, each test had a duration of 1.0 x 10
6

cycles, which was chosen to incorporate run-in and steady-state wear rates
while retaining the integrity of Disc B specimens for use in multiple tests as well
as maximizing the number of tests performed. The specimens were positioned
in the same orientation after measurements to minimize run-in wear;
additionally, the magnitude, trend, and mode of wear at each measurement

189
interval suggested that the metal-on-metal specimens of the current study did
not experience running-in wear. However, future tests with a longer duration
may provide more insight into this matter.

Table 34: Running-in periods for metal-on-metal hip wear simulations
Running-In Period (10
6
Cycles)
Medley et al (1996) 0.1 – 0.5
Chan et al (1996) 0.5
Scholes et al (2001) 0.5
Anissian et al (1999) 0.9
Chan et al (1999) 1.0
Firkins et al (2001) 1.0
Dowson et al (2004) 1.0
Dowson et al (2004) 1.5
Smith et al (2001) 2.0
Rieker et al (2005) 0.5 – 4.0

Tests 1 and 2: Spine Gait and Bends, Disc A Specimens

Test 1 tested Disc A specimens under the Spine Gait profile, while Test 2 tested
them under the Spine Bends profile, to determine how the wear rate changes
when the magnitudes, frequency, and phasing of the motion and load are
changed from those typically experienced by the lumbar spine during gait to
those experienced by the lumbar spine during significant bends. The wear rate
of the Spine Bends specimens was 1.8 times that of the Spine Gait specimens,
with the increase in wear mainly attributed to the increase in cup wear. The cup
wore 2.6 times more than the ball for the Spine Bends specimens, while the ball
wore 1.1 times more than the cup for the Spine Gait specimens, which may

190
suggest different mechanisms of wear for the Spine Gait and Spine Bends
specimens.

The wear rates were consistent with the predictions based on the slide track
analysis and elastohydrodynamic lubrication equations. The sliding distance
and aspect ratio of the slide tracks produced by the Spine Bends specimens
were smaller than those of the Spine Gait specimens, indicating smaller cross-
path angles. In metal-on-metal bearings, crossing-path motion has been shown
to polish metal surfaces and decrease wear
210,215
. Thus, the smaller cross-path
angles of the Spine Bends specimens predicted more wear for the Spine Bends
specimens, which was consistent with the wear rates. Additionally, the lambda
ratio of the Spine Bends specimens was smaller than that of the Spine Gait
specimens. A smaller lambda ratio indicates a thinner lubricating layer of fluid
between the surfaces, potentially increasing wear. Thus, the smaller lambda
ratio of the Spine Bends specimens predicted more wear for the Spine Bends
specimens, which was consistent with the wear rates. Thus, the increase in
wear from the Spine Gait to Spine Bends testing profiles was most probably due
to the smaller cross-path angles and lower lubrication produced by the Spine
Bends profile.

The wear rates were contrary to the predictions based on Archard’s equation,
which describes a proportional relationship between motion and wear
6
. The
sliding distance of the slide tracks produced by the Spine Gait specimens was

191
larger than that produced by the Spine Bends specimens, predicting more wear
for the Spine Gait specimens by Archard’s equation; this was inconsistent with
the wear rates. This inconsistency emphasizes the importance of considering
the path of motion and velocity of the slide tracks, rather than just the
magnitude, and reveals the complexities of the mechanics of wear.

The lubrication gap of the worn Spine Bends specimens was marginally larger
than that of the Spine Gait specimens. Additionally, the maximum surface
roughness of the Spine Bends specimens was substantially higher than that of
the Spine Gait specimens. This may be a result of the higher amount of wear of
the Spine Bends specimens, which reduced the sphericity of the ball and cup,
leading to larger lubrication gaps, as well as roughened their surfaces, leading to
larger surface roughness.

Tests 3 and 4: Spine Gait and Bends, Disc B Specimens

Test 3 tested Disc A specimens under the Spine Gait profile, while Test 4 tested
them under the Spine Bends profile, to determine how the wear rate changes
when the magnitudes, frequency, and phasing of the motion and load are
changed from those typically experienced by the lumbar spine during gait to
those experienced by the lumbar spine during significant bends. The wear rate
of the Spine Bends specimens was 3.2 times that of the Spine Gait specimens,
with the increase in wear arising from an increase in both ball and cup wear.

192
The cup wore 2.1 times more than the ball for the Spine Bends specimens and
1.9 times more than the ball for the Spine Gait specimens, which may suggest
similar mechanisms of wear for the Spine Gait and Spine Bends specimens.

The wear rates were consistent with the predictions based on the slide track
analysis and elastohydrodynamic lubrication equations. The sliding distance
and aspect ratio of the slide tracks produced by the Spine Bends specimens
were smaller than those of the Spine Gait specimens, indicating smaller cross-
path angles and, therefore, predicting more wear for the Spine Bends
specimens, as shown by Tipper et al
210
; this was consistent with the wear rates.
Additionally, the lambda ratio of the Spine Bends specimens was smaller than
that of the Spine Gait specimens, predicting more wear for the Spine Bends
specimens; this was consistent with the wear rates. Thus, the increase in wear
from the Spine Gait to Spine Bends testing profiles was most probably due to
the smaller cross-path angles and lower lubrication produced by the Spine
Bends profile.

The wear rates were contrary to the predictions based on Archard’s equation.
The sliding distance of the slide tracks produced by the Spine Gait specimens
was larger than that produced by the Spine Bends specimens, predicting more
wear for the Spine Gait specimens by Archard’s equation; this was inconsistent
with the wear rates. This inconsistency emphasizes the importance of

193
considering the path of motion and velocity of the slide tracks, rather than just
the magnitude, and reveals the complexities of the mechanics of wear.

The lubrication gap of the worn Spine Bends specimens was substantially larger
than that of the Spine Gait specimens. This may be a result of the higher
amount of wear of the Spine Bends specimens, which reduced the sphericity of
the ball and cup, leading to larger lubrication gaps. The trend of the lubrication
gaps was similar to that of Disc A specimens, but with a more pronounced
difference.

The maximum surface roughness of the Spine Gait specimens was higher than
that of the Spine Bends specimens, despite the lower amount of wear; however,
this may be because the difference was marginal. The trend of the surface
roughness was opposite that of Disc A specimens.

8.6 Conclusions

Two types of metal-on-metal artificial lumbar discs were tested in the current
study. The wear rates of the two types of metal-on-metal specimens in the
current study ranged from 1.5 – 7.8 mm
3
/10
6
cycles. For Disc A specimens,
changing the testing profile from Spine Gait to Spine Bends increased the wear
from 4.4 to 7.8 mm
3
/10
6
cycles, but with substantially different ball-to-cup wear
ratios. For Disc B specimens, changing the testing profile from Spine Gait to

194
Spine Bends increased the wear from 2.1 to 6.6 mm
3
/10
6
cycles, with similar
ball-to-cup wear ratios. The increase in wear from the Spine Gait to Spine
Bends testing profiles was most probably due to the smaller cross-path angles
and lower lubrication produced by the Spine Bends profile.

For both Disc A and Disc B specimens, the substitution of the Spine Bends
profile for the Spine Gait profile is not recommended because the wear
generated by the Spine Gait profile is substantial and should not be ignored and,
furthermore, the Spine Bends profile was shown to produce different magnitudes
of wear than the Spine Gait profile. Moreover, the Spine Bends profile can
produce different mechanisms of wear than the Spine Bends profile, as
evidenced by the different ball-to-cup wear ratios of Disc A specimens.
However, for these same reasons, the simulation of the Spine Gait profile
interspersed with the Spine Bends profile may more accurately depict the wear
generated in vivo. To generate a more accurate testing profile, further clinical or
ergonomic research is needed to determine the proportion of Spine Bends
encountered in vivo.

Testing both Disc A and Disc B provided independent confirmation that the
Spine Bends profile produced 2-3 times more wear and larger lubrication gaps
than the Spine Gait profile. Additionally, testing both designs revealed that
differences in ball-to-cup wear ratios and surface roughness may arise from
differences in design.

195
CHAPTER 9: SUMMARY AND CONCLUSIONS

The current study was performed at Orthopaedic Hospital (Los Angeles, CA), in
collaboration with Abbott Spine (Austin, TX). The specific aims of the current
study were:

1. To apply equations for Hertzian contact, lubrication, and slide tracks to
predict or explain wear of metal-on-metal artificial lumbar discs.

2. To create three-dimensional finite element models to relate the contact areas
and contact stresses at the motion interfaces of artificial disc components to
wear.

3. To design and build a machine capable of simulating wear of artificial lumbar
discs by applying dynamic load and motion typical of the lumbar spine during
gait.

4. To modify a machine intended for hip wear simulation to simulate wear of
artificial lumbar discs by applying motion and load typical of the lumbar spine
during gait.

196
5. To determine the effects of the differences between hips and lumbar discs
on wear of metal-on-metal joints by comparing the effects of motion, load,
and cup design on wear.

6. To compare the effects of metallurgy and cup design on wear of
experimental metal-on-metal artificial lumbar discs, in order to facilitate the
design of artificial lumbar discs.

7. To compare the effects of different motion and load profiles on wear of
metal-on-metal artificial lumbar discs, in order to optimize the wear test
parameters for evaluating artificial lumbar discs.

To accomplish the specific aims, literature on artificial hips and discs was
reviewed. Problem statements were created by comparing and contrasting
artificial hips to discs, citing relevant studies, and emphasizing the limitations of
the studies in establishing wear of metal-on-metal lumbar discs. Analytical tools
were used to predict the effects of different factors on wear of metal-on-metal
lumbar discs. Then, wear simulators were designed and built or modified
(Specific Aims #3 and #4) to test different metal-on-metal artificial joints under
different testing profiles. The following conclusions from the analyses and
simulations were produced by these efforts:

197
Analytical Predictions of Artificial Disc Wear

This study addressed Specific Aims #1 and #2, which were to create three-
dimensional finite element models and apply equations for lubrication and slide
tracks to predict or explain wear of metal-on-metal lumbar discs.

In this study, the larger maximum contact pressure, lower lubrication, and
smaller cross-path angles predicted more wear for metal-on-metal artificial
lumbar discs compared to metal-on-metal artificial hips; however, the smaller
sliding distance predicted less wear. These counter-intuitive results indicate that
the differences in design between artificial discs and artificial hips can have
profound effects on wear. The results caution against the general assumption
that metal-on-metal artificial lumbar discs will be successful only by using the
same material combinations and clearances of successful metal-on-metal
artificial hips, without further design and material considerations. Additionally,
the contradictory predictions of wear behavior of metal-on-metal artificial discs
emphasize the need for comprehensive wear simulator studies, as well as
clinical and retrieval studies before metal-on-metal artificial lumbar discs are
used on a large scale.

198
Design and Measurement of Clearance of Artificial Discs

This study addressed Specific Aims #1 and #2, which was to create three-
dimensional finite element models and apply equations for Hertzian contact to
predict or explain wear of metal-on-metal lumbar discs.

In this study, it was concluded that designing an appropriate clearance of
implants with hemispherical cups, such as artificial hips, is substantially different
for implants with shallow cups, such as artificial discs. The contact mechanics
are difficult to predict, as evidenced by the discrepancies between the finite
element models and the predictions from the Hertz equations. An analysis such
as the one presented in this study is useful in designing clearance of specific
implants.

It was also concluded that the traditional method of measuring clearance of
artificial hips is inaccurate for artificial discs, due to the limitations of the curve-
and surface-fitting algorithms. Thus, a new non-surface-fit method of calculating
the lubrication gap was proposed in this study, which may be used to calculate
the lubrication gap at any cup depth directly from the data. A brief analysis
showed that the new method showed deviations in the surface that were unable
to be detected by the traditional method.

199
Effects of Motion, Load, and Cup Depth on In Vitro Wear of Metal-on-Metal Bearings

This study addressed Specific Aim #5, which was to determine the effects of the
differences between hips and lumbar discs on wear of metal-on-metal joints by
comparing the effects of motion, load, and cup design on wear.

In this study, the wear rates of the metal-on-metal joints ranged from 0.8 – 2.1
mm
3
/10
6
cycles. Decreasing the magnitudes of motion and load from that
experienced by the hip during gait to that experienced by the lumbar spine
during gait increased the wear from 0.9 to 2.1 mm
3
/10
6
cycles, which showed
that a metal-on-metal joint placed in an environment with a small range of
motion may experience smaller cross-path angles and low lubrication, which
lead to increased wear.

Changing the phasing of motion from that of an orbital profile to that of a gait
profile reduced the wear from 2.1 to 0.8 mm
3
/10
6
cycles, which showed that the
phasing of motion in the lumbar spine has a larger effect on wear than that of
the hip. For artificial hips, accurate phasing of motion is not crucial for
reproducing in vivo wear, as evidenced by the use of bi-axial orbital simulators,
which are simpler to manufacture than tri-axial hip gait simulators; however, the
results of this study showed that the phasing of motion substantially effect wear
of artificial lumbar discs.

200
Decreasing the cup depth from that of an artificial hip to that of an artificial
lumbar disc increased the wear from 0.8 to 2.1 mm
3
/10
6
cycles, which showed
that the reduction of cup depth to accommodate the space constraints of the
functional spine unit may increase material stresses and, therefore, wear.

P-values for the tests were only significant between Tests 3 and 4, when
reducing the cup depth. The standard deviations for all other tests were quite
high, in part due to the small specimen number; however, the trends between all
other tests were consistent with predictions, and thus require more samples to
make further conclusions.

Although articulating artificial lumbar discs generally use the same material
combinations and constructs that have been proven to be successful for artificial
hips, the biomechanical and anatomical differences between the joints were
shown to have an effect on wear. Therefore, careful consideration of the effects
of these differences is necessary during the design and testing of artificial
lumbar discs to avoid increased wear of the implants in vivo and in laboratory
wear simulations.

201
Effects of Carbon Content, Clearance, and Disc Features on In Vitro Wear of
Experimental Artificial Discs

This study addressed Specific Aim #6, which was to compare the effects of
metallurgy and cup design on wear of experimental metal-on-metal artificial
lumbar discs, in order to facilitate the design of artificial lumbar discs.

In this study, the wear rates of the experimental metal-on-metal lumbar discs
ranged from 5.4 – 17.0 mm
3
/10
6
cycles. However, changing the carbon content
of the ball from low to high, eliminating the anteroposterior keels, and lowering
the initial clearance reduced the wear from 12.0 to 7.8 mm
3
/10
6
. Furthermore,
removing the surgical notch reduced the wear from 7.8 to 5.4 mm
3
/10
6
cycles.
Additionally, since the Spine Bends profile was a simulation of worst-case wear,
wear of the implants under spine gait conditions will likely generate even less
wear. These reductions in wear are reminiscent of the minimization of wear
achieved in metal-on-metal total hip replacements after an extensive number of
hip wear simulations that examined the effect of individual parameters on wear.
The results emphasize the need for artificial lumbar disc wear simulations to
examine some of the individual design parameters of artificial lumbar discs, and
for clinical results and retrievals of artificial lumbar discs to validate these
results.

The surface damage was generally consistent with low lubrication and varying
degrees of abrasive and fatigue wear. Although the specimens tested in this

202
study were experimental, the results suggest that metal-on-metal lumbar discs
have the potential to produce wear of this magnitude and mechanism in vivo.
Therefore, careful consideration of individual design variables, such as those
considered in this study, is necessary to avoid production of excessive wear in
artificial lumbar discs.

Comparison of Gait and High-Motion Activity on In Vitro Wear of Metal-on-Metal
Bearings and Experimental Artificial Discs

This study addressed Specific Aim #7, which was to compare the effects of
different motion and load profiles on wear of metal-on-metal artificial lumbar
discs, in order to optimize the wear test parameters for evaluating artificial
lumbar discs.

In this study, two types of metal-on-metal artificial lumbar discs were tested.
The wear rates of the two types of metal-on-metal specimens in this study
ranged from 1.5 – 7.8 mm
3
/10
6
cycles. For thin specimens, changing the testing
profile from Spine Gait to Spine Bends increased the wear from 4.4 to 7.8
mm
3
/10
6
cycles, but with substantially different ball-to-cup wear ratios. For thick
specimens, changing the testing profile from Spine Gait to Spine Bends
increased the wear from 2.1 to 6.6 mm
3
/10
6
cycles, with similar ball-to-cup wear
ratios. The increase in wear from the Spine Gait to Spine Bends testing profiles
was most probably due to the smaller cross-path angles and lower lubrication
produced by the Spine Bends profile.

203

For both thin and thick specimens, the substitution of the Spine Bends profile for
the Spine Gait profile is not recommended because the wear generated by the
Spine Gait profile is substantial and should not be ignored and, furthermore, the
Spine Bends profile was shown to produce different magnitudes of wear than
the Spine Gait profile. Moreover, the Spine Bends profile can produce different
mechanisms of wear than the Spine Bends profile, as evidenced by the different
ball-to-cup wear ratios of the thin specimens. However, for these same reasons,
the simulation of the Spine Gait profile interspersed with the Spine Bends profile
may more accurately depict the wear generated in vivo. To generate a more
accurate testing profile, further clinical or ergonomic research is needed to
determine the proportion of Spine Bends encountered in vivo.

Testing both thin and thick specimens provided independent confirmation that
the Spine Bends profile produced 2-3 times more wear and larger lubrication
gaps than the Spine Gait profile. Additionally, testing both thin and thick
revealed that differences in ball-to-cup wear ratios and surface roughness may
arise from differences in design.

204
REFERENCES

1. Adams MA, Dolan P. Time-dependent changes in the lumbar spine's resistance
to bending. Clin Biomech (Bristol, Avon) 1996;11:194-200.
2. Adams MA, Freeman BJ, Morrison HP, et al. Mechanical initiation of
intervertebral disc degeneration. Spine 2000;25:1625-36.
3. Amstutz HC, Grigoris P. Metal on metal bearings in hip arthroplasty. Clin Orthop
Relat Res 1996:S11-34.
4. Andersson GB, Schultz AB. Effects of fluid injection on mechanical properties of
intervertebral discs. J Biomech 1979;12:453-8.
5. Anissian HL, Stark A, Gustafson A, et al. Metal-on-metal bearing in hip prosthesis
generates 100-fold less wear debris than metal-on-polyethylene. Acta Orthop Scand
1999;70:578-82.
6. Archard JF. Contact and rubbing of flat surfaces. J of Applied Physics
1953;24:981-8.
7. Benz EB, Federman M, Godleski JJ, et al. Transmission electron microscopy of
intracellular particles of polyethylene from joint replacement prostheses: size
distribution and cellular response. Biomaterials 2001;22:2835-42.
8. Bertagnoli R, Kumar S. Indications for full prosthetic disc arthroplasty: a
correlation of clinical outcome against a variety of indications. Eur Spine J 2002;11
Suppl 2:S131-6.
9. Blumenthal SL, Ohnmeiss DD, Guyer RD, et al. Prospective study evaluating total
disc replacement: preliminary results. J Spinal Disord Tech 2003;16:450-4.
10. Bono CM, Garfin SR. History and evolution of disc replacement. Spine J
2004;4:145S-50S.
11. Bono CM, Lee CK. Critical analysis of trends in fusion for degenerative disc
disease over the past 20 years: influence of technique on fusion rate and clinical
outcome. Spine 2004;29:455-63; discussion Z5.
12. Bos I, Johannisson R. Foreign body reactions in lymph nodes of oncology
patients with joint prostheses--light-, electron microscopic and immunohistological
investigations. Pathol Res Pract 2004;200:189-96.

205
13. Bowsher JG, Shelton JC. A hip simulator study on the influence of patient
activity level on the the wear of crosslinked polyethylene under smooth and
roughened femoral conditions. Wear 2001;250:167-79.
14. Bradford L, Baker DA, Graham J, et al. Wear and surface cracking in early
retrieved highly cross-linked polyethylene acetabular liners. J Bone Joint Surg Am
2004;86-A:1271-82.
15. Brown SR, Davies WA, DeHeer DH, et al. Long-term survival of McKee-Farrar
total hip prostheses. Clin Orthop Relat Res 2002:157-63.
16. Buscher R, Tager G, Dudzinski W, et al. Subsurface microstructure of metal-on-
metal hip joints and its relationship to wear particle generation. J Biomed Mater Res
B Appl Biomater 2005;72:206-14.
17. Buttner-Janz K, Hahn S, Schikora K, et al. [Basic principles of successful
implantation of the SB Charite model LINK intervertebral disk endoprosthesis].
Orthopade 2002;31:441-53.
18. Callaghan JP, Patla AE, McGill SM. Low back three-dimensional joint forces,
kinematics, and kinetics during walking. Clin Biomech (Bristol, Avon) 1999;14:203-
16.
19. Calonius O, Saikko V. Analysis of polyethylene particles produced in different
wear conditions in vitro. Clin Orthop Relat Res 2002:219-30.
20. Calonius O, Saikko V. Slide track analysis of eight contemporary hip simulator
designs. J Biomech 2002;35:1439-50.
21. Campbell P, Shen FW, McKellop H. Biologic and tribologic considerations of
alternative bearing surfaces. Clin Orthop Relat Res 2004:98-111.
22. Cappozzo A. Compressive loads in the lumbar vertebral column during normal
level walking. J Orthop Res 1984;1:292-301.
23. Cappozzo A. The forces and couples in the human trunk during level walking. J
Biomech 1983;16:265-77.
24. Caspi I, Levinkopf M, Nerubay J. Results of lumbar disk prosthesis after a
follow-up period of 48 months. Isr Med Assoc J 2003;5:9-11.
25. Catelas I, Bobyn JD, Medley JB, et al. Size, shape, and composition of wear
particles from metal-metal hip simulator testing: effects of alloy and number of
loading cycles. J Biomed Mater Res A 2003;67:312-27.

206
26. Catelas I, Bobyn JD, Medley JB, et al. Effects of digestion protocols on the
isolation and characterization of metal-metal wear particles. I. Analysis of particle
size and shape. J Biomed Mater Res 2001;55:320-9.
27. Catelas I, Campbell PA, Dorey F, et al. Semi-quantitative analysis of cytokines
in MM THR tissues and their relationship to metal particles. Biomaterials
2003;24:4785-97.
28. Catelas I, Huk OL, Petit A, et al. Flow cytometric analysis of macrophage
response to ceramic and polyethylene particles: effects of size, concentration, and
composition. J Biomed Mater Res 1998;41:600-7.
29. Catelas I, Medley JB, Campbell PA, et al. Comparison of in vitro with in vivo
characteristics of wear particles from metal-metal hip implants. J Biomed Mater Res
B Appl Biomater 2004;70:167-78.
30. Catelas I, Petit A, Vali H, et al. Quantitative analysis of macrophage apoptosis
vs. necrosis induced by cobalt and chromium ions in vitro. Biomaterials
2005;26:2441-53.
31. Chan FW, Bobyn JD, Medley JB, et al. The Otto Aufranc Award. Wear and
lubrication of metal-on-metal hip implants. Clin Orthop Relat Res 1999:10-24.
32. Chan FW, Bobyn JD, Medley JB, et al. Engineering issues and wear
performance of metal on metal hip implants. Clin Orthop Relat Res 1996:96-107.
33. Cinotti G, David T, Postacchini F. Results of disc prosthesis after a minimum
follow-up period of 2 years. Spine 1996;21:995-1000.
34. Clarke IC, Good V, Williams P, et al. Ultra-low wear rates for rigid-on-rigid
bearings in total hip replacements. Proc Inst Mech Eng [H] 2000;214:331-47.
35. Clarke MT, Darrah C, Stewart T, et al. Long-term clinical, radiological and
histopathological follow-up of a well-fixed Mckee-Farrar metal-on-metal total hip
arthroplasty. J Arthroplasty 2005;20:542-6.
36. Cromwell R, Schultz AB, Beck R, et al. Loads on the lumbar trunk during level
walking. J Orthop Res 1989;7:371-7.
37. de Kleuver M, Oner FC, Jacobs WCH. Total disc replacement for chronic low
back pain: background and a systematic review of the literature. Eur Spine J
2003;12:108-16.
38. Dean DD, Lohmann CH, Sylvia VL, et al. Effect of polymer molecular weight and
addition of calcium stearate on response of MG63 osteoblast-like cells to UHMWPE
particles. J Orthop Res 2001;19:179-86.

207
39. Dean DD, Schwartz Z, Liu Y, et al. The effect of ultra-high molecular weight
polyethylene wear debris on MG63 osteosarcoma cells in vitro. J Bone Joint Surg
Am 1999;81:452-61.
40. DeHeer DH, Engels JA, DeVries AS, et al. In situ complement activation by
polyethylene wear debris. J Biomed Mater Res 2001;54:12-9.
41. Delamarter RB, Bae HW, Pradhan BB. Clinical results of ProDisc-II lumbar total
disc replacement: report from the United States clinical trial. Orthop Clin North Am
2005;36:301-13.
42. Delamarter RB, Fribourg DM, Kanim LE, et al. ProDisc artificial total lumbar disc
replacement: introduction and early results from the United States clinical trial.
Spine 2003;28:S167-75.
43. Delaunay CP. Metal-on-metal bearings in cementless primary total hip
arthroplasty. J Arthroplasty 2004;19:35-40.
44. Dickerman R, Zigler JE, Sachs BL, et al. Stretch neuropraxia after anterior
lumbar interbody fusion or artificial disc placement. The Spine Journal 2005;5:1S-
189S.
45. Digas G, Karrholm J, Thanner J, et al. The Otto Aufranc Award. Highly cross-
linked polyethylene in total hip arthroplasty: randomized evaluation of penetration
rate in cemented and uncemented sockets using radiostereometric analysis. Clin
Orthop Relat Res 2004:6-16.
46. Djerf K, Wahlstrom O. Total hip replacement comparison between the McKee-
Farrar and Charnley prostheses in a 5-year follow-up study. Arch Orthop Trauma
Surg 1986;105:158-62.
47. Djerf K, Wahlstrom O, Hammerby S. Loosening 5 years after total hip
replacement--a radiological study of the McKee-Farrar and Charnley prostheses.
Arch Orthop Trauma Surg 1986;105:339-42.
48. Dobbs HS, Minski MJ. Metal ion release after total hip replacement. Biomaterials
1980;1:193-8.
49. Dooris AP, Ares PJ, Gabriel SM, et al. Wear characterization of an artificial disc
using ASTM guidelines. 51st Annual Meeting of the Orthopaedic Research Society
2005;Poster 1335.
50. Doorn PF, Campbell PA, Worrall J, et al. Metal wear particle characterization
from metal on metal total hip replacements: transmission electron microscopy study
of periprosthetic tissues and isolated particles. J Biomed Mater Res 1998;42:103-
11.

208
51. Dorr LD, Hilton KR, Wan Z, et al. Modern metal on metal articulation for total hip
replacements. Clin Orthop Relat Res 1996:108-17.
52. Dorr LD, Long WT. Metal-on-metal: articulations for the new millennium. Instr
Course Lect 2005;54:177-82.
53. Dorr LD, Long WT, Sirianni L, et al. The argument for the use of Metasul as an
articulation surface in total hip replacement. Clin Orthop Relat Res 2004:80-5.
54. Dorr LD, Wan Z, Longjohn DB, et al. Total hip arthroplasty with use of the
Metasul metal-on-metal articulation. Four to seven-year results. J Bone Joint Surg
Am 2000;82:789-98.
55. Dorr LD, Wan Z, Sirianni LE, et al. Fixation and osteolysis with Metasul metal-
on-metal articulation. J Arthroplasty 2004;19:951-5.
56. Dowson D, Hardaker C, Flett M, et al. A hip joint simulator study of the
performance of metal-on-metal joints: Part I: the role of materials. J Arthroplasty
2004;19:118-23.
57. Dowson D, Hardaker C, Flett M, et al. A hip joint simulator study of the
performance of metal-on-metal joints: Part II: design. J Arthroplasty 2004;19:124-30.
58. Dumbleton JH, Shen C. A study of the wear of some materials in connection
with total hip replacement. Wear 1974;29:163-71.
59. Dunstan E, Sanghrajka AP, Tilley S, et al. Metal ion levels after metal-on-metal
proximal femoral replacements: a 30-year follow-up. J Bone Joint Surg Br
2005;87:628-31.
60. Edidin AA, Kurtz SM. Influence of mechanical behavior on the wear of 4
clinically relevant polymeric biomaterials in a hip simulator. J Arthroplasty
2000;15:321-31.
61. Edidin AA, Pruitt L, Jewett CW, et al. Plasticity-induced damage layer is a
precursor to wear in radiation-cross-linked UHMWPE acetabular components for
total hip replacement. Ultra-high-molecular-weight polyethylene. J Arthroplasty
1999;14:616-27.
62. el-Bohy AA, Yang KH, King AI. Experimental verification of facet load
transmission by direct measurement of facet lamina contact pressure. J Biomech
1989;22:931-41.
63. Endo M, Tipper JL, Barton DC, et al. Comparison of wear, wear debris and
functional biological activity of moderately crosslinked and non-crosslinked
polyethylenes in hip prostheses. Proc Inst Mech Eng [H] 2002;216:111-22.

209
64. Errico TJ. Why a mechanical disc? Spine J 2004;4:151S-7S.
65. Errico TJ, Gatchel RJ, Schofferman J, et al. A fair and balanced view of spine
fusion surgery. Spine J 2004;4:S129-38.
66. Essner A, Klein R, Bushelow M, et al. The effect of sagittal conformity on knee
wear. Wear 2003;255:1085-92.
67. Essner A, Schmidig G, Wang A. The clinical relevance of hip joint simulator
testing: In vitro and in vivo comparisons. Wear 2005;259:882-6.
68. Essner A, Sutton K, Wang A. Hip simulator wear comparison of metal-on-metal,
ceramic-on-ceramic and crosslinked UHMWPE bearings. Wear 2005;259:992-5.
69. Farooq N, Park JC, Pollintine P, et al. Can vertebroplasty restore normal load-
bearing to fractured vertebrae? Spine 2005;30:1723-30.
70. Feipel V, De Mesmaeker T, Klein P, et al. Three-dimensional kinematics of the
lumbar spine during treadmill walking at different speeds. Eur Spine J 2001;10:16-
22.
71. Fernstrom U. Arthroplasty with intercorporal endoprothesis in herniated disc and
in painful disc. Acta Chir Scand Suppl 1966;357:154-9.
72. Firkins PJ, Tipper JL, Ingham E, et al. Influence of simulator kinematics on the
wear of metal-on-metal hip prostheses. Proc Inst Mech Eng [H] 2001;215:119-21.
73. Firkins PJ, Tipper JL, Saadatzadeh MR, et al. Quantitative analysis of wear and
wear debris from metal-on-metal hip prostheses tested in a physiological hip joint
simulator. Biomed Mater Eng 2001;11:143-57.
74. Fischer A, Wimmer MA. Differences in wear mechanisms between main contact
areas and stripe wear regions in metal-on-metal hip joints. J of Biomechanics
2006;39:S139.
75. Frelinghuysen P, Huang RC, Girardi FP, et al. Lumbar total disc replacement
part I: rationale, biomechanics, and implant types. Orthop Clin North Am
2005;36:293-9.
76. Fritsch EW, Gleitz M. Ceramic femoral head fractures in total hip arthroplasty.
Clin Orthop Relat Res 1996:129-36.
77. Fritzell P, Hagg O, Wessberg P, et al. 2001 Volvo Award Winner in Clinical
Studies: Lumbar fusion versus nonsurgical treatment for chronic low back pain: a
multicenter randomized controlled trial from the Swedish Lumbar Spine Study
Group. Spine 2001;26:2521-32; discussion 32-4.

210
78. Galvin AL, Tipper JL, Ingham E, et al. Nanometre size wear debris generated
from crosslinked and non-crosslinked ultra high molecular weight polyethylene in
artificial joints. Wear 2005;259:977-83.
79. Garcia-Cimbrelo E, Martinez-Sayanes JM, Minuesa A, et al. Mittelmeier
ceramic-ceramic prosthesis after 10 years. J Arthroplasty 1996;11:773-81.
80. Garrigues GE, Cho DR, Rubash HE, et al. Gene expression clustering using
self-organizing maps: analysis of the macrophage response to particulate
biomaterials. Biomaterials 2005;26:2933-45.
81. Gencur SJ, Rimnac C, Kurtz SM. Fatigue crack propogation resistance of virgin
and highly crosslinked, thermally treated ultra-high molecular weight polyethylene.
Biomaterials 2006;27:1550-7.
82. Gencur SJ, Rimnac CM, Kurtz SM. Failure micromechanisms during uniaxial
tensile fracture of conventional and highly crosslinked ultra-high molecular weight
polyethylenes used in total joint replacements. Biomaterials 2003;24:3947-54.
83. Germain MA, Hatton A, Williams S, et al. Comparison of the cytotoxicity of
clinically relevant cobalt-chromium and alumina ceramic wear particles in vitro.
Biomaterials 2003;24:469-79.
84. Goh JH, Thambyah A, Bose K. Effects of varying backpack loads on peak forces
in the lumbosacral spine during walking. Clin Biomech (Bristol, Avon) 1998;13:S26-
S31.
85. Goldsmith AA, Dowson D, Isaac GH, et al. A comparative joint simulator study of
the wear of metal-on-metal and alternative material combinations in hip
replacements. Proc Inst Mech Eng [H] 2000;214:39-47.
86. Gomoll A, Wanich T, Bellare A. J-integral fracture toughness and tearing
modulus measurement of radiation cross-linked UHMWPE. J Orthop Res
2002;20:1152-6.
87. Granchi D, Amato I, Battistelli L, et al. Molecular basis of osteoclastogenesis
induced by osteoblasts exposed to wear particles. Biomaterials 2005;26:2371-9.
88. Granchi D, Ciapetti G, Amato I, et al. The influence of alumina and ultra-high
molecular weight polyethylene particles on osteoblast-osteoclast cooperation.
Biomaterials 2004;25:4037-45.
89. Gray H. Anatomy of the Human Body. 20th ed. Philadelphia: Lea & Febiger,
1918.

211
90. Green TR, Fisher J, Matthews JB, et al. Effect of size and dose on bone
resorption activity of macrophages by in vitro clinically relevant ultra high molecular
weight polyethylene particles. J Biomed Mater Res 2000;53:490-7.
91. Green TR, Fisher J, Stone M, et al. Polyethylene particles of a 'critical size' are
necessary for the induction of cytokines by macrophages in vitro. Biomaterials
1998;19:2297-302.
92. Griffith SL, Shelokov AP, Buttner-Janz K, et al. A multicenter retrospective study
of the clinical results of the LINK SB Charite intervertebral prosthesis. The initial
European experience. Spine 1994;19:1842-9.
93. Gutwein LG, Webster TJ. Increased viable osteoblast density in the presence of
nanophase compared to conventional alumina and titania particles. Biomaterials
2004;25:4175-83.
94. Guyer RD, McAfee PC, Hochschuler SH, et al. Prospective randomized study of
the Charite artificial disc: data from two investigational centers. Spine J
2004;4:252S-9S.
95. Hale LC. Principles and techniques for designing precision machines. MIT PhD
Thesis 1999:417-26.
96. Hamrock BJ, Dowson D. Elastohydrodynamic lubrication of elliptical contacts for
materials of low elastic modulus I - fully flooded conjunction. J of Lubrication
Technology 1978;100:236-45.
97. Hamrock BJ, Dowson D. Isothermal elastohydrodynamic lubrication of point
contacts: Part III - fully flooded results. J of Lubrication Technology 1977;99:264-76.
98. Han CD, Choe WS, Yoo JH. Effect of polyethylene wear on osteolysis in
cementless primary total hip arthroplasty: minimal 5-year follow-up study. J
Arthroplasty 1999;14:714-23.
99. Hatton A, Nevelos JE, Matthews JB, et al. Effects of clinically relevant alumina
ceramic wear particles on TNF-alpha production by human peripheral blood
mononuclear phagocytes. Biomaterials 2003;24:1193-204.
100. Hatton A, Nevelos JE, Nevelos AA, et al. Alumina-alumina artificial hip joints.
Part I: a histological analysis and characterisation of wear debris by laser capture
microdissection of tissues retrieved at revision. Biomaterials 2002;23:3429-40.
101. Heisel C, Silva M, dela Rosa MA, et al. Short-term in vivo wear of cross-linked
polyethylene. J Bone Joint Surg Am 2004;86-A:748-51.

212
102. Hermida JC, Bergula A, Chen P, et al. Comparison of the wear rates of twenty-
eight and thirty-two-millimeter femoral heads on cross-linked polyethylene
acetabular cups in a wear simulator. J Bone Joint Surg Am 2003;85-A:2325-31.
103. Hernandez-Rodriguez MAL, Mercado-Solis RD, Perez-Unzueta AJ, et al. Wear
of cast metal-metal pairs for total replacement hip prostheses. Wear 2005;259:958-
63.
104. Holley KG, Furman BD, Babalola OM, et al. Impingement of acetabular cups in
a hip simulator: comparison of highly cross-linked and conventional polyethylene. J
Arthroplasty 2005;20:77-86.
105. Holmes AD, Hukins DW. Analysis of load-relaxation in compressed segments
of lumbar spine. Med Eng Phys 1996;18:99-104.
106. Hopper RH, Jr., Young AM, Orishimo KF, et al. Correlation between early and
late wear rates in total hip arthroplasty with application to the performance of
marathon cross-linked polyethylene liners. J Arthroplasty 2003;18:60-7.
107. Howie DW, McCalden RW, Nawana NS, et al. The long-term wear of retrieved
McKee-Farrar metal-on-metal total hip prostheses. J Arthroplasty 2005;20:350-7.
108. Huang RC, Girardi FP, Cammisa Jr FP, et al. Long-term flexion-extension
range of motion of the prodisc total disc replacement. J Spinal Disord Tech
2003;16:435-40.
109. Huang RC, Girardi FP, Lim MR, et al. Advantages and disadvantages of
nonfusion technology in spine surgery. Orthop Clin North Am 2005;36:263-9.
110. Huang RC, Wright TM, Panjabi MM, et al. Biomechanics of nonfusion implants.
Orthop Clin North Am 2005;36:271-80.
111. Huk OL, Catelas I, Mwale F, et al. Induction of apoptosis and necrosis by metal
ions in vitro. J Arthroplasty 2004;19:84-7.
112. Ingram JH, Stone M, Fisher J, et al. The influence of molecular weight,
crosslinking and counterface roughness on TNF-alpha production by macrophages
in response to ultra high molecular weight polyethylene particles. Biomaterials
2004;25:3511-22.
113. Jacobs MA, Schmidt MB, Farrar R. Comparison of wear rates for metal-on-
polyethylene and metal-on-metal hips. AAHKS Sixth Annual Meeting 1997.
114. Jacobs MA, Schmidt MB, Farrar R. The effect of clearance and diameter on
debris generation in a metal-on-metal hip. J Arthroplasty 1998;13.

213
115. Jacobsson SA, Djerf K, Wahlstrom O. Twenty-year results of McKee-Farrar
versus Charnley prosthesis. Clin Orthop Relat Res 1996:S60-8.
116. Jagatia M, Jin ZM. Analysis of elastohydrodynamic lubrication in a novel metal-
on-metal hip joint replacement. Proc Inst Mech Eng [H] 2002;216:185-93.
117. Jantsch S, Schwagerl W, Zenz P, et al. Long-term results after implantation of
McKee-Farrar total hip prostheses. Arch Orthop Trauma Surg 1991;110:230-7.
118. JRU. Swedish National Hip Arthroplasty Register [Department of
Orthopaedics], 2004.
119. Kang PH, Nho YC. The effect of gamma-irradiation on ultra-high molecular
weight polyethylene recrystallized under different cooling conditions. Radiat Phys
Chem 2001;60:79-87.
120. Katz JN. Lumbar disc disorders and low-back pain: socioeconomic factors and
consequences. J Bone Joint Surg Am 2006;88 Suppl 2:21-4.
121. Kesteris U, Jonsson K, Robertsson O, et al. Polyethylene wear and synovitis in
total hip arthroplasty: a sonographic study of 48 hips. J Arthroplasty 1999;14:138-
43.
122. Khoo BC, Goh JC, Bose K. A biomechanical model to determine lumbosacral
loads during single stance phase in normal gait. Med Eng Phys 1995;17:27-35.
123. Khoo BC, Goh JC, Lee JM, et al. A comparison of lumbosacral loads during
static and dynamic activities. Australas Phys Eng Sci Med 1994;17:55-63.
124. Koeller W, Muehlhaus S, Meier W, et al. Biomechanical properties of human
intervertebral discs subjected to axial dynamic compression--influence of age and
degeneration. J Biomech 1986;19:807-16.
125. Kurtz SM, McCombe P, Ciccarelli L, et al. Retrieval analysis of short-term
implanted Maverick total disc arthroplasty. Proceedings of the NASS 20th Annual
Meeting: The Spine Journal 2005;5:1S-189S.
126. Kurtz SM, Peloza J, Siskey R, et al. Analysis of a retrieved polyethylene total
disc replacement component. Spine J 2005;5:344-50.
127. Kurtz SM, van Ooij A, John P, et al. Central core and rim damage in retrieved
polyethylene total disc replacement components. Proceedings of the NASS 20th
Annual Meeting: The Spine Journal 2005;5:1S-189S.
128. Kurtz SM, van Ooij A, Ross R, et al. Polyethylene wear and rim fracture in total
disc arthroplasty. Spine J 2007;7:12-21.

214
129. Kurtz SM, Villarraga ML, Herr MP, et al. Thermomechanical behavior of virgin
and highly crosslinked ultra-high molecular weight polyethylene used in total joint
replacements. Biomaterials 2002;23:3681-97.
130. Le Huec J, Basso Y, Mathews H, et al. The effect of single-level, total disc
arthroplasty on sagittal balance parameters: a prospective study. Eur Spine J
2005;14:480-6.
131. Le Huec JC, Mathews H, Basso Y, et al. Clinical results of Maverick lumbar
total disc replacement: two-year prospective follow-up. Orthop Clin North Am
2005;36:315-22.
132. Lee CK, Langrana NA. A review of spinal fusion for degenerative disc disease:
need for alternative treatment approach of disc arthroplasty? Spine J 2004;4:173S-
6S.
133. Lemaire JP, Carrier H, Sari Ali el H, et al. Clinical and radiological outcomes
with the Charite artificial disc: a 10-year minimum follow-up. J Spinal Disord Tech
2005;18:353-9.
134. Lemaire JP, Skalli W, Lavaste F, et al. Intervertebral disc prosthesis. Results
and prospects for the year 2000. Clin Orthop Relat Res 1997:64-76.
135. Liao YS, McKellop H, Lu Z, et al. The effect of frictional heating and forced
cooling on the serum lubricant and wear of UHMW polyethylene cups against
cobalt-chromium and zirconia balls. Biomaterials 2003;24:3047-59.
136. Liao Y-S, McNulty D, Hanes M. Wear rate and surface morphology of
UHMWPE cups are affected by the serum lubricant concentration in a hip simulation
test. Wear 2003;255:1051-6.
137. Lindh M. Chapter 10: Biomechanics of the lumbar spineed. Philadelphia: Lea
and Febiger, 1989.
138. Link HD. History, design and biomechanics of the LINK SB Charite artificial
disc. Eur Spine J 2002;11 Suppl 2:S98-S105.
139. Luo L, Petit A, Antoniou J, et al. Effect of cobalt and chromium ions on MMP-1,
TIMP-1, and TNF-alpha gene expression in human U937 macrophages: a role for
tyrosine kinases. Biomaterials 2005;26:5587-93.
140. Mabrey JD, Afsar-Keshmiri A, McClung GA, 2nd, et al. Comparison of
UHMWPE particles in synovial fluid and tissues from failed THA. J Biomed Mater
Res 2001;58:196-202.

215
141. Maher SA, Lipman JD, Curley LJ, et al. Mechanical performance of ceramic
acetabular liners under impact conditions. J Arthroplasty 2003;18:936-41.
142. Manaka M, Clarke IC, Yamamoto K, et al. Stripe wear rates in alumina THR--
comparison of microseparation simulator study with retrieved implants. J Biomed
Mater Res B Appl Biomater 2004;69:149-57.
143. Marnay T. Lumbar disc replacement: 7 to 11-year results with Prodisc. The
Spine Journal 2002;2:94.
144. Martell JM, Verner JJ, Incavo SJ. Clinical performance of a highly cross-linked
polyethylene at two years in total hip arthroplasty: a randomized prospective trial. J
Arthroplasty 2003;18:55-9.
145. Martinez JB, Oloyede VO, Broom ND. Biomechanics of load-bearing of the
intervertebral disc: an experimental and finite element model. Med Eng Phys
1997;19:145-56.
146. Masse A, Bosetti M, Buratti C, et al. Ion release and chromosomal damage
from total hip prostheses with metal-on-metal articulation. J Biomed Mater Res B
Appl Biomater 2003;67:750-7.
147. Mathews HH, Lehuec JC, Friesem T, et al. Design rationale and biomechanics
of Maverick Total Disc arthroplasty with early clinical results. Spine J 2004;4:268S-
75S.
148. McEvoy A, Jeyam M, Ferrier G, et al. Synergistic effect of particles and cyclic
pressure on cytokine production in human monocyte/macrophages: proposed role in
periprosthetic osteolysis. Bone 2002;30:171-7.
149. McGowan B, Callaghan J, McGill S. The effects of cadence on lumbar spine
kinematics during gait. North American Congress on Biomechanics 1998.
150. McKellop H, Park SH, Chiesa R, et al. In vivo wear of three types of metal on
metal hip prostheses during two decades of use. Clin Orthop Relat Res 1996:S128-
40.
151. Medley JB, Chan FW, Krygier JJ, et al. Comparison of alloys and designs in a
hip simulator study of metal on metal implants. Clin Orthop Relat Res 1996:S148-
59.
152. Milosev I, Pisot V, Campbell P. Serum levels of cobalt and chromium in
patients with Sikomet metal-metal total hip replacements. J Orthop Res
2005;23:526-35.

216
153. Milosev I, Trebse R, Kovac S, et al. Survivorship and retrieval analysis of
Sikomet metal-on-metal total hip replacements at a mean of seven years. J Bone
Joint Surg Am 2006;88:1173-82.
154. Morlock M, Schneider E, Bluhm A, et al. Duration and frequency of every day
activities in total hip patients. J Biomech 2001;34:873-81.
155. Muratoglu OK, Greenbaum ES, Bragdon CR, et al. Surface analysis of early
retrieved acetabular polyethylene liners: a comparison of conventional and highly
crosslinked polyethylenes. J Arthroplasty 2004;19:68-77.
156. Muratoglu OK, Wannomae K, Christensen S, et al. Ex vivo wear of
conventional and cross-linked polyethylene acetabular liners. Clin Orthop Relat Res
2005;438:158-64.
157. Nechtow W, Hintner M, Bushelow M, et al. IVD replacement mechanical
performance depends strongly on input parameters. 52nd Annual Meeting of the
Orthopaedic Research Society 2006;Paper No: 0118.
158. Nester CJ, van der Linden ML, Bowker P. Effect of foot orthoses on the
kinematics and kinetics of normal walking gait. Gait Posture 2003;17:180-7.
159. Nevelos J, Ingham E, Doyle C, et al. Microseparation of the centers of alumina-
alumina artificial hip joints during simulator testing produces clinically relevant wear
rates and patterns. J Arthroplasty 2000;15:793-5.
160. Nevelos JE, Ingham E, Doyle C, et al. Analysis of retrieved alumina ceramic
components from Mittelmeier total hip prostheses. Biomaterials 1999;20:1833-40.
161. Niki Y, Matsumoto H, Otani T, et al. Flow cytometric technique for the detection
of phagocytosed wear particles in patients with total joint arthroplasty. Biomaterials
2003;24:3715-24.
162. Niosi CA, Oxland TR. Degenerative mechanics of the lumbar spine. Spine J
2004;4:202S-8S.
163. Ohta M, Hyon S-H, Tsutumi S. Control of crystalline orientation to enhance the
wear resistance of ultra-high molecular weight polyethylene crystallization cups for
artificial joints. Wear 2003;255:1045-50.
164. Park YS, Hwang SK, Choy WS, et al. Ceramic failure after total hip arthroplasty
with an alumina-on-alumina bearing. J Bone Joint Surg Am 2006;88:780-7.
165. Perry J. Gait analysis: normal and pathological functioned. Thorofare, NJ:
Slack, Inc., 1992.

217
166. Pizzoferrato A, Vespucci A, Ciapetti G, et al. Biocompatibility testing of
prosthetic implant materials by cell cultures. Biomaterials 1985;6:346-51.
167. Pollintine P, Przybyla AS, Dolan P, et al. Neural arch load-bearing in old and
degenerated spines. J Biomech 2004;37:197-204.
168. Pylios T, Shepherd DE. Prediction of lubrication regimes in wrist implants with
spherical bearing surfaces. J Biomech 2004;37:405-11.
169. Rao S, Shirata K, Furukawa KS, et al. Evaluation of cytotoxicity of UHMWPE
wear debris. Biomed Mater Eng 1999;9:209-17.
170. Rieker CB, Konrad R, Schon R, et al. In vivo and in vitro surface changes in a
highly cross-linked polyethylene. J Arthroplasty 2003;18:48-54.
171. Rieker CB, Schon R, Konrad R, et al. Influence of the clearance on in-vitro
tribology of large diameter metal-on-metal articulations pertaining to resurfacing hip
implants. Orthop Clin North Am 2005;36:135-42, vii.
172. Rieker CB, Schon R, Kottig P. Development and validation of a second-
generation metal-on-metal bearing: laboratory studies and analysis of retrievals. J
Arthroplasty 2004;19:5-11.
173. Ries MD, Scott ML, Jani S. Relationship between gravimetric wear and particle
generation in hip simulators: conventional compared with cross-linked polyethylene.
J Bone Joint Surg Am 2001;83-A Suppl 2 Pt 2:116-22.
174. Saikko V. A 12-station anatomic hip joint simulator. Proc Inst Mech Eng [H]
2005;219:437-48.
175. Saikko V, Calonius O. Slide track analysis of the relative motion between
femoral head and acetabular cup in walking and in hip simulators. J Biomech
2002;35:455-64.
176. Saikko V, Calonius O, Keranen J. Effect of counterface roughness on the wear
of conventional and crosslinked ultrahigh molecular weight polyethylene studied with
a multi-directional motion pin-on-disk device. J Biomed Mater Res 2001;57:506-12.
177. Saito S, Ryu J, Watanabe M, et al. Midterm results of metasul metal-on-metal
total hip arthroplasty. J Arthroplasty 2006;21:1105-10.
178. Sartor C, Alderink G, Greenwald H, et al. Critical kinematic events occurring in
the trunk during walking. Hum Mov Sci 1999;18:669-79.
179. Sartor C, Alderink G, Greenwald H, et al. Critical kinematic events occurring in
the trunk during walking. Human Movement Science 1999;18:669-79.

218
180. Schache AG, Blanch P, Rath D, et al. Three-dimensional angular kinematics of
the lumbar spine and pelvis during running. Hum Mov Sci 2002;21:273-93.
181. Schache AG, Wrigley TV, Blanch PD, et al. The effect of differing Cardan angle
sequences on three dimensional lumbo-pelvic angular kinematics during running.
Med Eng Phys 2001;23:493-501.
182. Schey JA. Systems view of optimizing metal on metal bearings. Clin Orthop
Relat Res 1996:S115-27.
183. Schmalzried TP, Peters PC, Maurer BT, et al. Long-duration metal-on-metal
total hip arthroplasties with low wear of the articulating surfaces. J Arthroplasty
1996;11:322-31.
184. Schmalzried TP, Szuszczewicz ES, Akizuki KH, et al. Factors correlating with
long term survival of McKee-Farrar total hip prostheses. Clin Orthop Relat Res
1996:S48-59.
185. Schmalzried TP, Szuszczewicz ES, Northfield MR, et al. Quantitative
assessment of walking activity after total hip or knee replacement. J Bone Joint Surg
Am 1998;80:54-9.
186. Schmidt M, Weber H, Schon R. Cobalt chromium molybdenum metal
combination for modular hip prostheses. Clin Orthop Relat Res 1996:S35-47.
187. Scholes SC, Green SM, Unsworth A. The wear of metal-on-metal total hip
prostheses measured in a hip simulator. Proc Inst Mech Eng [H] 2001;215:523-30.
188. Scholes SC, Unsworth A, Goldsmith AA. A frictional study of total hip joint
replacements. Phys Med Biol 2000;45:3721-35.
189. Scholes SC, Unsworth A, Hall RM, et al. The effects of material combination
and lubricant on the friction of total hip prostheses. Wear 2000;241:209-13.
190. Serhan HA, Dooris AP, Ares PJ, et al. Wear characterization of the Charite
Artificial Disc using ASTM guidelines. Proceedings of the NASS 20th Annual
Meeting: The Spine Journal 2005;5:1S-189S.
191. Sethi RK, Neavyn MJ, Rubash HE, et al. Macrophage response to cross-linked
and conventional UHMWPE. Biomaterials 2003;24:2561-73.
192. Shanbhag AS, Jacobs JJ, Black J, et al. Effects of particles on fibroblast
proliferation and bone resorption in vitro. Clin Orthop Relat Res 1997:205-17.

219
193. Shishido T, Clarke IC, Williams P, et al. Clinical and simulator wear study of
alumina ceramic THR to 17 years and beyond. J Biomed Mater Res B Appl
Biomater 2003;67:638-47.
194. Siepe CJ, Mayer HM, Wiechert K, et al. Clinical results of total lumbar disc
replacement with ProDisc II: three-year results for different indications. Spine
2006;31:1923-32.
195. Silva M, Shepherd EF, Jackson WO, et al. Average patient walking activity
approaches 2 million cycles per year: pedometers under-record walking activity. J
Arthroplasty 2002;17:693-7.
196. Smith SL, Dowson D, Goldsmith AA. The effect of femoral head diameter upon
lubrication and wear of metal-on-metal total hip replacements. Proc Inst Mech Eng
[H] 2001;215:161-70.
197. Sott AH, Harrison DJ. Increasing age does not affect good outcome after
lumbar disc replacement. Int Orthop 2000;24:50-3.
198. St John KR, Zardiackas LD, Poggie RA. Wear evaluation of cobalt-chromium
alloy for use in a metal-on-metal hip prosthesis. J Biomed Mater Res B Appl
Biomater 2004;68:1-14.
199. Stea S, Visentin M, Granchi D, et al. Apoptosis in peri-implant tissue.
Biomaterials 2000;21:1393-8.
200. Stewart T, Tipper J, Streicher R, et al. Long-term wear of HIPed alumina on
alumina bearings for THR under microseparation conditions. J Mater Sci Mater Med
2001;12:1053-6.
201. Stewart TD, Tipper JL, Insley G, et al. Long-term wear of ceramic matrix
composite materials for hip prostheses under severe swing phase microseparation.
J Biomed Mater Res B Appl Biomater 2003;66:567-73.
202. Szpalski M, Gunzburg R, Mayer M. Spine arthroplasty: a historical review. Eur
Spine J 2002;11 Suppl 2:S65-84.
203. Taylor NF, Evans OM, Goldie PA. Angular movements of the lumbar spine and
pelvis can be reliably measured after 4 minutes of treadmill walking. Clin Biomech
(Bristol, Avon) 1996;11:484-6.
204. Taylor NF, Evans OM, Goldie PA. The effect of walking faster on people with
acute low back pain. Eur Spine J 2003;12:166-72.
205. Taylor NF, Goldie PA, Evans OM. Angular movements of the pelvis and lumbar
spine during self-selected and slow walking speeds. Gait Posture 1999;9:88-94.

220
206. Tehranzadeh J, Ton JD, Rosen CD. Advances in spinal fusion. Semin
Ultrasound CT MR 2005;26:103-13.
207. Thierry D. Revision of a Charite artificial disc 9.5 years in vivo to a new Charite
artificial disc: case report and explant analysis. Eur Spine J 2005.
208. Tingle E, Sellner R. Personal communication with Mitutoyo employees. City of
Industry, CA, 2006.
209. Tipper JL, Firkins PJ, Besong AA, et al. Characterisation of wear debris from
UHMWPE on zirconia ceramic, metal-on-metal and alumina ceramic-on-ceramic hip
prostheses generated in a physiological anatomical hip joint simulator. Wear
2001;250:120-8.
210. Tipper JL, Firkins PJ, Ingham E, et al. Quantitative analysis of the wear and
wear debris from low and high carbon content cobalt chrome alloys used in metal on
metal total hip replacements. J Mater Sci Mater Med 1999;10:353-62.
211. Tipper JL, Hatton A, Nevelos JE, et al. Alumina-alumina artificial hip joints. Part
II: characterisation of the wear debris from in vitro hip joint simulations. Biomaterials
2002;23:3441-8.
212. Tropiano P, Huang RC, Girardi FP, et al. Lumbar total disc replacement. Seven
to eleven-year follow-up. J Bone Joint Surg Am 2005;87:490-6.
213. Tropiano P, Huang RC, Girardi FP, et al. Lumbar disc replacement: preliminary
results with ProDisc II after a minimum follow-up period of 1 year. J Spinal Disord
Tech 2003;16:362-8.
214. Tudor-Locke C, Bassett DR, Swartz AM, et al. A preliminary study of one year
of pedometer self-monitoring. Ann Behav Med 2004;28:158-62.
215. Turell M, Wang A, Bellare A. Quantification of the effect of cross-path motion
on the wear rate of ultra-high molecular weight polyethylene. Wear 2003;255:1034-
9.
216. Udofia IJ, Jin ZM. Elastohydrodynamic lubrication analysis of metal-on-metal
hip-resurfacing prostheses. J Biomech 2003;36:537-44.
217. van Ooij A, Oner FC, Verbout AJ. Complications of artificial disc replacement. J
of Spinal Disorders and Techniques 2003;16:369-83.
218. Verdonschot N, Vena P, Stolk J, et al. Effects of metal-inlay thickness in
polyethylene cups with metal-on-metal bearings. Clin Orthop Relat Res 2002:353-
61.

221
219. Vidovszky TJ, Cabanela ME, Rock MG, et al. Histologic and biochemical
differences between osteolytic and nonosteolytic membranes around femoral
components of an uncemented total hip arthroplasty. J Arthroplasty 1998;13:320-30.
220. Visentin M, Stea S, Squarzoni S, et al. A new method for isolation of
polyethylene wear debris from tissue and synovial fluid. Biomaterials 2004;25:5531-
7.
221. Vogt L, Pfeifer K, Banzer W. Comparison of angular lumbar spine and pelvis
kinematics during treadmill and overground locomotion. Clin Biomech (Bristol, Avon)
2002;17:162-5.
222. Vogt L, Pfeifer K, Portscher, et al. Influences of nonspecific low back pain on
three-dimensional lumbar spine kinematics in locomotion. Spine 2001;26:1910-9.
223. Vogt L, Portscher M, Brettmann K, et al. Cross-validation of marker
configurations to measure pelvic kinematics in gait. Gait Posture 2003;18:178-84.
224. von Knoch M, Jewison DE, Sibonga JD, et al. The effectiveness of
polyethylene versus titanium particles in inducing osteolysis in vivo. J Orthop Res
2004;22:237-43.
225. Walker PS, Gold BL. The tribology (friction, lubrication, and wear) of all-metal
artificial hip joints. Wear 1971;17:285-99.
226. Walker PS, Salvati E, Hotzler RK. The wear on removed McKee-Farrar total
hip prostheses. J Bone Joint Surg Am 1974;56:92-100.
227. Walter WL, Insley GM, Walter WK, et al. Edge loading in third generation
alumina ceramic-on-ceramic bearings: stripe wear. J Arthroplasty 2004;19:402-13.
228. Wang A, Essner A, Schmidig G. The effects of lubricant composition on in vitro
wear testing of polymeric acetabular components. J Biomed Mater Res B Appl
Biomater 2004;68:45-52.
229. Wang A, Yue S, Bobyn JD, et al. Surface characterization of metal-on-metal
hip implants tested in a hip simulator. Wear 1999;225-229:708-15.
230. Warashina H, Sakano S, Kitamura S, et al. Biological reaction to alumina,
zirconia, titanium and polyethylene particles implanted onto murine calvaria.
Biomaterials 2003;24:3655-61.
231. Willems PC, Nienhuis B, Sietsma M, et al. The effect of a plaster cast on
lumbosacral joint motion. An in vivo assessment with precision motion analysis
system. Spine 1997;22:1229-34.

222
232. Williams S, Stewart TD, Ingham E, et al. Metal-on-metal bearing wear with
different swing phase loads. J Biomed Mater Res B Appl Biomater 2004;70:233-9.
233. Wimmer MA, Sprecher C, Hauert R, et al. Tribochemical reaction on metal-on-
metal hip joint bearings: A comparison between in-vitro and in-vivo results. Wear
2003;255:1007-14.
234. Wu JJ, Buckley CP, O'Connor JJ. Mechanical integrity of compression-
moulded ultra-high molecular weight polyethylene: effects of varying process
conditions. Biomaterials 2002;23:3773-83.
235. Yagil-Kelmer E, Kazmier P, Rahaman MN, et al. Comparison of the response
of primary human blood monocytes and the U937 human monocytic cell line to two
different sizes of alumina ceramic particles. J Orthop Res 2004;22:832-8.
236. Yamamoto T, Saito M, Ueno M, et al. Wear analysis of retrieved ceramic-on-
ceramic articulations in total hip arthroplasty: Femoral head makes contact with the
rim of the socket outside of the bearing surface. J Biomed Mater Res B Appl
Biomater 2005;73:301-7.
237. Yew A, Jagatia M, Ensaff H, et al. Analysis of contact mechanics in McKee-
farrar metal-on-metal hip implants. Proc Inst Mech Eng [H] 2003;217:333-40.
238. Zahiri CA, Schmalzried TP, Ebramzadeh E, et al. Lessons learned from
loosening of the McKee-Farrar metal-on-metal total hip replacement. J Arthroplasty
1999;14:326-32.
239. Zeegers WS, Bohnen LM, Laaper M, et al. Artificial disc replacement with the
modular type SB Charite III: 2-year results in 50 prospectively studied patients. Eur
Spine J 1999;8:210-7.
240. Zigler JE. Clinical results with ProDisc: European experience and U.S.
investigation device exemption study. Spine 2003;28:S163-6.
241. Zigler JE. Lumbar spine arthroplasty using the ProDisc II. Spine J
2004;4:260S-7S.
242. Zigler JE, Burd TA, Vialle EN, et al. Lumbar spine arthroplasty: early results
using the ProDisc II: a prospective randomized trial of arthroplasty versus fusion. J
Spinal Disord Tech 2003;16:352-61.

223
APPENDICES

A. Chapter 6 Data

The plots in Figure 37 through Figure 44 were generated from the following
tables, which show measurements for each specimen, measurement interval,
and test.

Table 35: Ball volumetric wear.
Ball Volumetric Wear (mm
3
)

Test 1 Test 2 Test 3 Test 4
0M
0.00000
0.00000

0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000

0.00000
0.00000
0.00000

0.25M

0.02401
0.14806
0.17607
0.06803
0.10004
0.08803
0.02401

0.14406
0.18007
0.02801

0.5M

0.15606
0.29612
0.52021
0.09204
0.11605
0.22409
0.05202

0.30012
0.46419
0.03201

0.75M

0.33613
0.40816
0.73229
0.07603
0.13205
0.35214
0.04402

0.41216
0.79232
0.04402

1M
0.17207
0.18407

0.58824
0.53621
0.83633
0.06002
0.15206
0.48820
0.06403

0.50820
1.24050
0.05202

224
Table 36: Cup volumetric wear.
Cup Volumetric Wear (mm
3
)

Test 1 Test 2 Test 3 Test 4
0M
0.00000
0.00000

0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000

0.00000
0.00000
0.00000

0.25M

0.21609
0.23209
-0.01601
0.05202
0.35614
0.43217
-0.00800

0.41617
0.30412
0.11204

0.5M

0.50820
0.60024
0.55222
0.08804
0.38015
0.91236
0.07203

0.69228
0.64426
0.14406

0.75M

0.97239
1.11244
1.19248
0.09204
0.44818
1.41657
0.06002

1.08844
1.13645
0.17207

1M
0.17207
1.26050

1.67267
1.55262
1.72869
0.06803
0.47219
1.81673
0.06803

1.43657
1.74470
0.20808

225
Table 37: Total volumetric wear rate.
Total Volumetric Wear Rate (mm
3
/10
6
cycles)

Test 1 Test 2 Test 3 Test 4
0M

0.25M

0.96038
1.52061
0.64026
0.48019
1.82473
2.04882
0.06403

2.24090
1.93677
0.56022

0.5M

1.69668
2.06483
3.64946
0.24010
0.16006
2.49700
0.43217

1.72869
2.49700
0.14406

0.75M

2.57703
2.49700
3.40936
-0.04802
0.33613
2.52901
-0.08003

2.03281
3.28131
0.16006

1M
0.34414
1.44458

3.80952
2.27291
2.56102
-0.16006
0.17607
2.14486
0.11204

1.77671
4.22569
0.17607

226
Table 38: Lubrication gap.
Lubrication Gap ( μm)

Test 1 Test 2 Test 3 Test 4
0M

1.67
1.01
7.40
4.52
1.27
1.36
5.31

0.16
2.97
3.14

0.25M

0.65
1.48
4.62
6.06
0.02
2.76
2.41

2.85
4.19
5.12

0.5M

0.35
2.06
3.55
1.17
1.06
7.91
2.67

2.30
9.88
8.77

0.75M

0.36
1.95
0.36
1.23
1.17
1.05
3.70

4.39
5.24
9.24

1M

1.27
1.36
4.40
5.31
0.16
2.97
3.14

2.71
6.04
15.01

227
Table 39: Mediolateral average surface roughness.
Mediolateral Surface Roughness ( μm)

Test 1 Test 2 Test 3 Test 4
0M

0.118
0.195
0.343
0.209
0.187
0.081
0.174
0.209
0.125
0.182
0.261

0.25M

0.182
0.058
0.099
0.112
0.150
0.212
0.305

0.105
0.125
0.244

0.5M

0.261
0.080
0.241
0.198
0.192
0.113
0.253

0.127
0.175
0.181

0.75M

0.192
0.077
0.253
0.264
0.126
0.102
0.235

0.129
0.091
0.138

1M

0.187
0.081
0.174
0.209
0.125
0.182
0.261

0.172
0.150
0.112

228
Table 40: Anteroposterior average surface roughness.
Anteroposterior Surface Roughness ( μm)

Test 1 Test 2 Test 3 Test 4
0M

0.123
0.196
0.315
0.230
0.130
0.069
0.156
0.229
0.113
0.199
0.260

0.25M

0.115
0.068
0.094
0.090
0.101
0.186
0.323

0.125
0.158
0.246

0.5M

0.293
0.096
0.233
0.205
0.155
0.155
0.268

0.146
0.149
0.170

0.75M

0.239
0.043
0.210
0.265
0.149
0.105
0.261

0.136
0.102
0.163

1M

0.130
0.069
0.156
0.229
0.113
0.199
0.260

0.217
0.117
0.119

229
B. Chapter 7 Data

The plots in Figure 51 through Figure 57 were generated from the following
tables, which show measurements for each specimen, measurement interval,
and test.

Table 41: Ball volumetric wear.
Ball Volumetric Wear (mm
3
)

Test 1 Test 2a Test 2b
0M
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.25M
2.81673
3.48659
3.15526
1.21289
0.24650
0.17727
0.29412
0.41817
0.65906
0.5M
5.36935
6.56703
6.09444
2.19008
0.64386
0.48139
0.66186
0.72709
1.29372
1M
12.50860
11.05082
11.48980
3.79992
1.32173
1.31373
1.74030
1.54742
2.29532
1.5M
20.16487
15.26931
17.36415
2M
27.61224
22.18607
24.35934
2.5M
35.93437
30.41777
32.70188

230
Table 42: Cup volumetric wear.
Cup Volumetric Wear (mm
3
)

Test 1 Test 2a Test 2b
0M
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.25M
0.05882
0.06643
0.36174
0.73870
1.32453
2.06323
1.24690
1.33453
1.54142
0.5M
0.00804
0.16767
0.47099
1.46499
2.81473
3.59344
2.17327
2.48699
2.48179
1M
0.22449
0.64026
0.85474
3.56423
6.33733
6.58583
3.67627
4.44618
4.23369
1.5M
0.33653
1.21208
1.29652
2M
0.50820
1.49140
1.50620
2.5M
0.67587
1.69468
1.71709

231
Table 43: Total volumetric wear rate
Total Volumetric Wear Rate (mm
3
/10
6
cycles)

Test 1 Test 2a Test 2b
0M

0.25M
11.50220
14.21208
14.06803
7.80632
6.28411
8.96198
6.16407
7.01080
8.80192
0.5M
10.32733
12.72669
12.19368
6.81393
7.55022
7.33733
5.17647
5.84554
6.30012
1M
14.55142
9.91277
11.55822
7.41817
8.40096
7.64946
5.16287
5.55902
5.50700
1.5M
15.53661
9.58063
12.63225
2M
15.23810
14.39216
14.40976
2.5M
16.97959
16.86995
17.10684

232
Table 44: Lubrication gap.
Lubrication Gap ( μm)

Test 1 Test 2a Test 2b
0M
24.09
15.60
22.92
3.16
1.15
3.26
4.21
1.90
3.73
0.25M
5.34
6.81
12.62
9.33
13.56
14.68
5.07
9.44
13.34
0.5M
9.02
8.36
11.32
15.19
11.62
16.50
7.84
17.62
15.08
1M
8.08
8.51
12.93
11.35
20.90
13.87
3.08
16.51
20.91
1.5M

2M
7.51
5.17
11.98
2.5M
5.22
1.50
0.09

233
Table 45: Mediolateral average surface roughness.
Mediolateral Surface Roughness ( μm)

Test 1 Test 2a Test 2b
0M
0.100
0.100
0.110
0.100
0.100
0.110
0.100
0.090
0.100
0.25M
0.160
0.110
0.150
0.520
0.220
0.180
0.220
0.150
0.100
0.5M
0.320
0.190
0.290
0.750
0.630
0.450
0.270
0.140
0.200
1M
0.210
0.330
0.180
0.720
0.790
0.380
0.720
0.140
1.730
1.5M

2M
0.260
0.210
0.360
2.5M
0.205
0.218
0.178

234
Table 46: Anteroposterior average surface roughness.
Anteroposterior Surface Roughness ( μm)

Test 1 Test 2a Test 2b
0M
0.110
0.100
0.090
0.090
0.100
0.100
0.100
0.100
0.100
0.25M
0.160
0.140
0.240
0.890
0.270
0.230
0.260
0.130
0.100
0.5M
0.330
0.210
0.300
0.930
0.370
0.490
0.210
0.190
0.100
1M
0.170
0.190
0.230
0.130
0.370
0.380
0.890
0.110
1.090
1.5M

2M
0.160
0.210
0.240
2.5M
0.205
0.245
0.186

235
C. Chapter 8 Data

The plots in Figure 69 through Figure 75 were generated from the following
tables, which show measurements for each specimen, measurement interval,
and test.

Table 47: Ball volumetric wear.
Ball Volumetric Wear (mm
3
)

Test 1 Test 2 Test 3 Test 4
0M
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.25M
0.45578
0.64266
0.47859
1.21289
0.24650
0.17727
0.14406
0.18007
0.02801
0.15606
0.69228
0.01601
0.5M
1.04842
1.19848
0.86715
2.19008
0.64386
0.48139
0.30012
0.46419
0.03201
0.36415
1.58864
0.08804
0.75M
1.78872
1.77751
1.43697
0.41216
0.79232
0.04402
0.90836
2.45298
0.19208
1M
2.30652
2.27771
1.94998
3.79992
1.32173
1.31373
0.50820
1.24050
0.05202
1.68467
3.23329
0.27211

236
Table 48: Cup volumetric wear.
Cup Volumetric Wear (mm
3
)

Test 1 Test 2 Test 3 Test 4
0M
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.00000
0.25M
0.71469
0.33653
0.17927
0.73870
1.32453
2.06323
0.41617
0.30412
0.11204
1.10044
0.64426
0.33213
0.5M
1.28852
0.82833
0.56142
1.46499
2.81473
3.59344
0.69228
0.64426
0.14406
2.55702
1.59664
0.88435
0.75M
1.84874
1.40096
1.19408
1.08844
1.13645
0.17207
3.50540
2.42497
1.76471
1M
2.26010
2.01401
1.94798
3.56423
6.33733
6.58583
1.43657
1.74470
0.20808
4.45778
3.35734
3.20128

Table 49: Total volumetric wear rate.
Total Volumetric Wear Rate (mm
3
/10
6
cycles)

Test 1 Test 2 Test 3 Test 4
0M

0.25M
4.68187
3.91677
2.63145
7.80632
6.28411
8.96198
2.24090
1.93677
0.56022
5.02601
5.34614
1.39256
0.5M
4.66587
4.19048
3.08283
6.81393
7.55022
7.33733
1.72869
2.49700
0.14406
6.65866
7.39496
2.49700
0.75M
5.20208
4.60664
4.80992
2.03281
3.28131
0.16006
5.97039
6.77071
3.93758
1M
3.71669
4.45298
5.06763
7.41817
8.40096
7.64946
1.77671
4.22569
0.17607
6.91477
6.85074
6.06643

237
Table 50: Lubrication gap.
Lubrication Gap ( μm)

Test 1 Test 2 Test 3 Test 4
0M
0.16
2.97
3.14
2.71
6.04
15.01
4.21
1.90
3.73
3.16
1.15
3.26
0.25M
2.85
4.19
5.12
5.52
14.10
15.41
5.07
9.44
13.34
9.33
13.56
14.68
0.5M
2.30
9.88
8.77
7.89
14.57
20.31
7.84
17.62
15.08
15.19
11.62
16.50
0.75M
4.39
5.24
9.24
5.51
2.18
9.60
1M
2.71
6.04
15.01
11.30
21.07
19.34
3.08
16.51
20.91
11.35
20.90
13.87

Table 51: Mediolateral average surface roughness.
Mediolateral Surface Roughness ( μm)

Test 1 Test 2 Test 3 Test 4
0M
0.075
0.075
0.068
0.100
0.100
0.110
4.910
7.170
10.280
6.770
5.890
4.390
0.25M
0.170
0.144
0.107
0.520
0.220
0.180
4.120
4.910
9.610
3.170
5.160
9.370
0.5M
0.119
0.119
0.167
0.750
0.630
0.450
5.000
6.900
7.140
3.690
4.430
4.520
0.75M
0.087
1.103
0.130
5.070
3.600
5.430
2.810
3.630
4.330
1M
0.167
0.125
0.100
0.720
0.790
0.380
6.770
5.890
4.390
3.330
3.110
5.220

238
Table 52: Anteroposterior average surface roughness.
Anteroposterior Surface Roughness ( μm)

Test 1 Test 2 Test 3 Test 4
0M
0.080
0.076
0.066
0.090
0.100
0.100
4.430
7.840
10.250
8.540
4.610
4.670
0.25M
0.116
0.260
0.153
0.890
0.270
0.230
4.940
6.230
9.670
4.730
6.130
7.450
0.5M
0.170
0.164
0.170
0.930
0.370
0.490
5.740
5.860
6.680
4.210
2.720
8.060
0.75M
0.157
0.108
0.133
5.370
4.030
6.410
3.200
2.960
7.230
1M
0.403
0.150
0.115
0.130
0.370
0.380
8.540
4.610
4.670
2.810
3.270
6.230

Abstract (if available)

Abstract Metal-on-metal bearings have been established in total hip replacements but, more recently, have been developed with modifications for artificial discs. The significance of these modifications is not known. In this study, analytical tools were used to predict the effects of cup depth, clearance, motion, and load on wear of metal-on-metal artificial lumbar discs. Then, laboratory wear simulators were developed. The first series of wear simulations determined the effects of the differences between artificial hips and discs on wear by comparing the effects of motion, load, and cup design

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Investigation of preclinical testing methods for total ankle replacements

PDF

Modeling human regulation of momentum while interacting with the environment

PDF

The effects of the vitreous state on the ocular pharmacokinetics following intravitreal drug delivery: an evaluation with quantitative magnetic resonance imaging

PDF

Feasibility of using a lined portion of the Los Angeles River for artificial recharge

PDF

The influence of hip muscle performance on postural stability & ankle biomechanics: implications for ankle injury

PDF

Developing efficient methods for the manufacture and analysis of composites structures

PDF

Atomistic modeling of the mechanical properties of metallic glasses, nanoglasses, and their composites

Asset Metadata

Creator Lee, Jessica Lynn (author)

Core Title Wear of metal-on-metal artificial discs for the lumbar spine

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Biomedical Engineering

Publication Date 06/29/2007

Defense Date 03/23/2007

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag artificial disc,lumbar spine,metal-on-metal,OAI-PMH Harvest,wear simulator

Language English

Advisor Ebramzadeh, Edward (committee chair), Baker, Lucinda L. (committee member), D'Argenio, David (committee member), Davoodi, Rahman (committee member), McNitt-Gray, Jill L. (committee member)

Creator Email jessicll@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m569

Unique identifier UC1216946

Identifier etd-Lee-20070629 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-512552 (legacy record id),usctheses-m569 (legacy record id)

Legacy Identifier etd-Lee-20070629.pdf

Dmrecord 512552

Document Type Dissertation

Rights Lee, Jessica Lynn

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu