development and characterization of a chemically ...4162/datastream... · development and...
TRANSCRIPT
Development and Characterization of a Chemically Crosslinked Polyvinyl Alcohol/Polyethylene Glycol Hydrogel for Injectable Nucleus Pulposus Replacement
A Thesis
Submitted to the Faculty
of
Drexel University
by
Valerie Regina Binetti
in Partial Fulfillment of the
Requirements for the Degree
of
Doctor of Philosophy
in
Materials Science and Engineering
May 2013
ii
© Copyright 2013
Valerie R. Binetti. All rights reserved.
iii
DEDICATIONS
To my parents, Lou and Gloria, and my husband, Jeffrey, for their endless encouragement and support.
iv
ACKNOWLEDGEMENTS
I wish to express my sincere gratitude to my advisor Dr. Anthony Lowman. I will
be forever grateful for his patience and generosity in letting me navigate my own path
through my graduate work, I did not take the shortest path but I learned more than I ever
thought possible. Tony’s excitement as I developed new ideas drove me to push myself to
unprecedented levels. I am forever grateful for his confidence in my work and my
abilities; and for teaching me to have that confidence in myself. I also thank him for the
diverse opportunities he made available to me, in particular the opportunities to present
my research internationally.
I also thank the members of my thesis committee: Dr. Michele Marcolongo, Dr.
Garland Fussell, Dr. Jennifer Vernengo and Dr. Christoper Li. Dr. Marcolongo has been
my mentor and one of my biggest supporters over the last six years; for that I am very
grateful. I also want to thank Dr. Vernengo and Dr. Fussell for being great resources as I
learned organic chemistry and polymer synthesis; in addition, I would like to thank them
for their friendship.
I would like to express my sincere gratitude to my friends and colleagues at Drexel
University. This work would not have been possible without Dr. Sumona Sarkar and Dr.
Matthew Hood; their friendship, their willingness to commiserate when nothing seemed
to be going well and most of all, our daily conversations that pushed each of us to be
better scientists and engineers. I also thank Janah and Steve Szewcyzk for being my
cheerleaders from my first term of graduate school till my last. I want to thank the
v
members of the Biomaterials and Drug Delivery laboratory: in particular Dr. Kristin Kita,
Dr. Julianne Holloway, Dr. Lauren Conova-Grous, Dr. Kara Spiller, Dr. Michael Marks
and Pamela Kubinski. I would like to thank the members of my honorary lab, Dr.
Marcolongo’s Biomaterials laboratory, in particular Dr. Jessica Isaacs, Nandita Ganesh,
David Jamison, and Rob Yucha. I also want to thank Dr. Timothy Himes at the Drexel
Medical School for allowing me to use the cryostat and microscope in his laboratory, and
Dr. George Myer and Steve Peterson at Temple for keeping their diffractometer working
long enough for me to collect my data.
I want to thank my family, my parents Louis and Gloria Binetti, my brother
Louis, and my sisters Stephanie and Laura for their love and support. Finally, I wish to
express my profound gratitude to my husband, Jeffrey Honer, for his initial
encouragement to attend graduate school and his unwavering encouragement and support
everyday since.
vi
TABLE OF CONTENTS 1. INTRODUCTION ................................................................................................ 1 2. BACKGROUND ................................................................................................... 4 2.1. SPINE ANATOMY AND PHYSIOLOGY ............................................................................................................... 4 2.1.1. The Vertebral Column ............................................................................................................................. 4 2.1.2. The Intervertebral Disc .......................................................................................................................... 5 2.1.3. The Nucleus Pulposus ............................................................................................................................. 6 2.1.4. The Annulus Fibrosus .............................................................................................................................. 8 2.1.5. End Plate ................................................................................................................................................... 10
2.2. DEGENERATIVE DISC DISEASE .................................................................................................................... 10 2.2.1. Nucleus Pulposus ................................................................................................................................... 11 2.2.2. Annulus Fibrosus .................................................................................................................................... 13 2.2.3. End Plates ................................................................................................................................................. 14 2.2.4. Relationship to Low Back Pain ........................................................................................................ 15
2.3. CLINICAL TREATMENT OPTIONS ................................................................................................................. 16 2.3.1. Discectomy ................................................................................................................................................ 16 2.3.2. Spinal Fusion ........................................................................................................................................... 17 2.3.3. Total Disc Arthroplasty ....................................................................................................................... 19 2.3.4. Nucleus Replacement and Stabilization Technologies .......................................................... 21
2.4. POLY(VINYL ALCOHOL) ................................................................................................................................. 29 2.5. POLY(VINYL ALCOHOL) HYDROGELS .......................................................................................................... 30 2.6. POLY(ETHYLENE GLYCOL) ............................................................................................................................ 31 2.7. POLY (VINYL ALCOHOL)/POLY (ETHYLENE GLYCOL) HYDROGELS ...................................................... 32 2.7.1. Chemical crosslinking using radiation ......................................................................................... 34 2.7.2. Chemical crosslinking using difunctional agents .................................................................... 34
2.8. BIOCOMPATIBILITY ........................................................................................................................................ 35 2.9. MACROMOLECULAR-‐BASED SOLUTIONS AS SWELLING MEDIA ......................................... 37
3. RESEARCH GOALS .......................................................................................... 64 4. SYNTHESIS OF A CROSSLINKED PVA/PEG HYDROGEL SYSTEM ....... 66 4.1. INTRODUCTION ............................................................................................................................................... 66 4.2. MATERIALS AND METHODS ......................................................................................................................... 67 4.2.1. Materials ................................................................................................................................................... 67 4.2.2. Hydrogel Synthesis ................................................................................................................................ 68 4.2.3. Swelling Mechanics ............................................................................................................................... 70 4.2.4. Mechanical Properties ......................................................................................................................... 72 4.2.5. FTIR ............................................................................................................................................................. 73 4.2.6. Statistical Analysis ................................................................................................................................ 74
4.3. RESULTS AND DISCUSSION ........................................................................................................................... 74 4.3.1. Swelling Mechanics ............................................................................................................................... 74 4.3.2. Mechanical Properties ......................................................................................................................... 76 4.3.3. FTIR ............................................................................................................................................................. 77
4.4. CONCLUSIONS ................................................................................................................................................. 81 5. CHARACTERIZATION OF THE PVA/PVP/PEG-DGE HYDROGEL SYSTEM ................................................................................................................ 101 5.1. INTRODUCTION ............................................................................................................................................. 101 5.2. MATERIALS AND METHODS ....................................................................................................................... 102 5.2.1. Materials ................................................................................................................................................. 102
vii
5.2.2. Hydrogel Synthesis .............................................................................................................................. 103 5.2.3. Swelling Mechanics ............................................................................................................................. 104 5.2.4. Mechanical Properties ....................................................................................................................... 106 5.2.5. FTIR ........................................................................................................................................................... 107 5.2.6. X-ray Diffraction .................................................................................................................................. 108 5.2.7. Optical Microscopy .............................................................................................................................. 110 5.2.8. Statistical Analysis .............................................................................................................................. 111
5.3. RESULTS AND DISCUSSION ......................................................................................................................... 111 5.3.1. Swelling Mechanics ............................................................................................................................. 111 5.3.2. Mechanical Properties ....................................................................................................................... 113 5.3.3. FTIR ........................................................................................................................................................... 114 5.3.4. X-ray Diffraction .................................................................................................................................. 115 5.3.5. Optical Microscopy .............................................................................................................................. 116
5.4. CONCLUSIONS ............................................................................................................................................... 122 6. PURIFICIATION METHOD DEVELOPMENT AND CHARACTERIZATION OF THE PURIFIED PVA/PVP/PEG-DGE HYDROGEL ......................................................................................................... 153 6.1. INTRODUCTION ............................................................................................................................................. 153 6.2. MATERIALS AND METHODS ....................................................................................................................... 154 6.2.1. Materials ................................................................................................................................................. 154 6.2.2. Hydrogel Purification ........................................................................................................................ 155 6.2.3. Swelling Mechanics ............................................................................................................................. 156 6.2.4. Mechanical Properties ....................................................................................................................... 159 6.2.5. FTIR ........................................................................................................................................................... 159 6.2.6. X-ray Diffraction .................................................................................................................................. 160 6.2.7. Optical Microscopy .............................................................................................................................. 161 6.2.8. Cytotoxicity ............................................................................................................................................. 162 6.2.9. Statistical Analysis .............................................................................................................................. 163
6.3. RESULTS AND DISCUSSION ......................................................................................................................... 163 6.3.1. Hydrogel purification ........................................................................................................................ 163 6.3.2. Swelling Mechanics ............................................................................................................................. 165 6.3.3. Mechanical Properties ....................................................................................................................... 166 6.3.4. FTIR ........................................................................................................................................................... 167 6.3.5. X-ray Diffraction .................................................................................................................................. 168 6.3.6. Optical Microscopy .............................................................................................................................. 168 6.3.7. Cytotoxicity ............................................................................................................................................. 169
6.4. CONCLUSIONS ............................................................................................................................................... 174 7. CHARACTERIZATION OF THE SWOLLEN PVA/PVP/PEG-DGE HYDROGEL ......................................................................................................... 199 7.1. INTRODUCTION ............................................................................................................................................. 199 7.2. MATERIALS AND METHODS ....................................................................................................................... 200 7.2.1. Materials ................................................................................................................................................. 200 7.2.2. Mechanical Properties ....................................................................................................................... 201 7.2.3. X-ray Diffraction .................................................................................................................................. 202 7.2.4. Statistical Analysis .............................................................................................................................. 204
7.3. RESULTS AND DISCUSSION ......................................................................................................................... 204 7.3.1. Mechanical Properties ....................................................................................................................... 204 7.3.2. X-ray Diffraction .................................................................................................................................. 206
viii
7.3.3. Discussion ................................................................................................................................................ 207 7.4. CONCLUSIONS ............................................................................................................................................... 211
8. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK ... 232 8.1. CONCLUSIONS ............................................................................................................................................... 232 8.2. RECOMMENDATIONS ................................................................................................................................... 237
ix
LIST OF FIGURES
Figure 2.1: The spinal column [176] ............................................................................... 39
Figure 2.2: Intervertebral Disc [7] ................................................................................... 40
Figure 2.3: The organization of the vertebral endplate [7] .............................................. 41
Figure 2.4: Nonlinear stress/strain curve of collagenous tissues [10] ............................... 43
Figure 2.5: Progression of intervertebral disc degeneration [10] ...................................... 44
Figure 2.6: A cylindrical interbody fusion cage (A, anterior; P, posterior) [177] ............. 46
Figure 2.7: CHARITÉ Artificial Disc [75] ..................................................................... 47
Figure 2.8: DASCAR device [75] ................................................................................... 48
Figure 2.9: PDN-SOLO and HydraFlex devices [75] ..................................................... 49
Figure 2.10: NeuDisc device [75] .................................................................................... 50
Figure 2.11: NuCore Injectable Nucleus Device [75] ...................................................... 51
Figure 2.12: Aquarelle Nucleus, available in two sizes 0.1 cm3 (A) and 0.3 cm3 (B) [91] 52
Figure 2.13: BioDisc Nucleus Pulposus Replacement [75] .............................................. 53
Figure 2.14: Chemical structure of PVA ......................................................................... 54
Figure 2.15: Chemical structure for PVP ......................................................................... 55
Figure 2.16: Interchain hydrogen bonding within a PVA/PVP blend occurs between carbonyl groups on PVP and hydroxyl groups on PVA ........................................... 56
Figure 2.17: Chemical structure for PEG ........................................................................ 57
Figure 2.18: Chemical structure for PEG-DGE ............................................................. 58
Figure 2.19: Schematic of PVA theta-gel formation: (a) PVA-PEG water mixture at 90oC is a uniform solution; (b) as the solution is cooled down phase separation begins and forces the PVA to form crystalline domains; (c) with further cooling to near room temperature, phase separation results in the formation of pores containing water surrounded by PVA rich regions. [98] ........................................................... 59
Figure 2.20: Radiation scission (A) and crosslinking (B) ................................................. 60
x
Figure 2.21: Chemical structure of polymers that degrade or crosslink when exposed to irradiation [121, 122] ............................................................................................... 61
Figure 2.22: Chemical reaction of radiation crosslinking of a polymer ............................ 62
Figure 2.23: Chemical reaction of the crosslinking of poly(vinyl alcohol) with poly(ethylene glycol) diglycidyl ether [178] ............................................................. 63
Figure 4.1: Swelling ratio for 24 hour hydrogel formulation ........................................... 85
Figure 4.2: Swelling ratio of electron beam hydrogel formulations .................................. 87
Figure 4.3: Swelling ratio of hydrogel formulations varying basic catalyst volume .......... 89
Figure 4.4: Swelling ratio of hydrogel formulations varying reaction time ...................... 91
Figure 4.5: Stress versus strain plot for the 24 hour reaction time hydrogel .................... 86
Figure 4.6: Compressive moduli of electron beam hydrogel formulations ....................... 93
Figure 4.7: Compressive moduli of hydrogel formulations varying basic catalyst volume 94
Figure 4.8: Compressive moduli of hydrogel formulations varying reaction time ............ 95
Figure 4.9: FTIR spectra of PVA .................................................................................... 96
Figure 4.10: FTIR spectra of PEG and PEG-DGE ....................................................... 97
Figure 4.11: FTIR of electron beam 14.8% PVA hydrogel formulation; with increased irradiation, from 0 to 20 kGy, the PEG-CH2- symmetric stretch (2851 cm-1) and the PVA-CH2- symmetric stretch (2922 cm-1) increases indicating scission of the polymer chains. ........................................................................................................ 98
Figure 4.12: FTIR of difunctional crosslinked chemically crosslinked hydrogels: (A) varied basic catalyst volume, (B) varied reaction time .............................................. 99
Figure 4.13: Electron beam crosslinked hydrogel formulations with radiation dosages ranging from 0 to 100 kGy (right to left). .............................................................. 100
Figure 5.1: Swelling ratio of the 19% PEG-DGE hydrogel formulation ...................... 125
Figure 5.2: Swelling ratio of 19 to 39% PEG-DGE ...................................................... 126
Figure 5.3: Swelling ratio of 12.6 to 20.0% PVA ........................................................... 128
Figure 5.4: Swelling Ratio of 19% PEG-DGE and PEG-OH ..................................... 130
xi
Figure 5.5: Swelling ratio of 19% PEG-DGE with varying molecular weight (526, 2000 and 4600 Da) ......................................................................................................... 132
Figure 5.6: Compression moduli over 24 hours of 19 to 39% PEG-DGE ................ Error! Bookmark not defined.
Figure 5.7: Stress versus strain plot for 39% PEG-DGE hydrogel formulation ............ 134
Figure 5.8: Compression moduli over 24 hours of 19 to 39% PVA ............................... 135
Figure 5.9: Compression moduli over 24 hours of 12.6 to 20.0% PVA ......................... 136
Figure 5.10: Compressive moduli over 24 hours of 19% PEG-DGE and PEG-OH ... 137
Figure 5.11: Compressive moduli over 24 hours of 19% PEG-DGE with varying molecular weight (526, 2000 and 4600 Da) ........................................................... 138
Figure 5.12: FTIR of 19 and 39% PEG-DGE hydrogel formulations .......................... 139
Figure 5.13: FTIR of 12.6 and 20.0% PVA hydrogel formulations ............................... 140
Figure 5.14: FTIR of 19% PEG-DGE and PEG-OH hydrogel formulations ............. 141
Figure 5.15: 19% PEG-DGE and 19% PEG-OH phase separation ............................ 152
Figure 5.16: FTIR of 19% PEG-DGE with varying molecular weight (526, 2000 and 4600 Da) ................................................................................................................ 142
Figure 5.17: X-ray diffraction patterns of PVA, F/T PVA, PEG and PEG-DGE ...... 143
Figure 5.18: X-ray diffraction patterns of 19% and 39% PEG-DGE hydrogel formulations ........................................................................................................... 144
Figure 5.19: Relative crystallinity for formulations with varying PEG-DGE content .. 145
Figure 5.20: Relative crystallinity for formulations with varying PVA content ............. 146
Figure 5.21: Relative crystallinity for formulations varying PEG functional group (p>0.05) .................................................................................................................. 147
Figure 5.22: Relative crystallinity for formulations varying PEG molecular weight (p<0.05) .................................................................................................................. 148
Figure 5.23: Optical micrographs of 19% PEG-DGE and 39% PEG-DGE. Scale bars indicate 50 µm. ...................................................................................................... 149
Figure 5.24: Optical micrographs of 19% PEG-DGE and 19% PEG-OH. Scale bars indicate 50 µm. ...................................................................................................... 150
xii
Figure 5.25: Optical micrographs of 19% PEG-DGE at varying molecular weights (526 Da, 2000 Da and 4600 Da). Scale bars indicate 50 µm. ........................................ 151
Figure 6.1: Diagram of the purification of the PVA/PEG-DGE hydrogel formulations ............................................................................................................................... 177
Figure 6.2: Compressive moduli of 39% PEG-DGE purified at various osmotic pressures and various periods of time .................................................................................... 178
Figure 6.3: Swelling ratio at four weeks of 39% PEG-DGE purified at various osmotic pressures and various periods of time ..................................................................... 179
Figure 6.4: Swelling ratio of purified 19 to 39% PEG-DGE hydrogel formulations .... 181
Figure 6.5: Swelling ratio of purified 12.6 to 20% PVA hydrogel formulation .............. 183
Figure 6.6: Compressive moduli over 24 hours of purified 19 to 39% PEG-DGE hydrogel formulations ............................................................................................ 185
Figure 6.7: Compressive moduli over 24 hours of purified 12.6 to 20% PVA hydrogel formulations ........................................................................................................... 186
Figure 6.8: FTIR of purified 19 and 39% PEG-DGE hydrogel formulations .............. 187
Figure 6.9: FTIR of purified 12.6 and 20% PVA hydrogel formulations ...................... 188
Figure 6.10: Relative crystallinity for purified formulations with varying PEG-DGE content ................................................................................................................... 189
Figure 6.11: Relative crystallinity for purified formulations with varying PVA content 190
Figure 6.12: Optical micrographs of purified 19% PEG-DGE and 39% PEG-DGE. Scale bars indicate 100 µm. .................................................................................... 192
Figure 6.13: Cytotoxicity of PVA/PVP/PEG and PVA/PVP/PEG-DGE hydrogel extractions in PBS .................................................................................................. 193
Figure 6.14: Cytotoxicity of PVA/PVP/PEG and PVA/PVP/PEG-DGE hydrogel extractions in cottonseed oil ................................................................................... 194
Figure 6.15: Cytotoxicity of PEG-DGE (526 MW), UV sterilized .............................. 195
Figure 6.16: Cytotoxicity of PEG-DGE (526 MW), autoclave sterilized ..................... 196
Figure 6.17: Cytotoxicity of PEG-DGE concentrations with varying molecular weights ............................................................................................................................... 197
Figure 6.18: Cytotoxicity of PEG-OH and PEG-DGE concentrations ....................... 198
xiii
Figure 7.1: Stress versus strain plots for 19% and 39% PEG-DGE hydrogel formulations ............................................................................................................................... 213
Figure 7.2: Compressive moduli of 19 to 39% PEG-DGE hydrogel formulations swollen in 0.2 MPa at 37oC for up to four weeks ............................................................... 214
Figure 7.3: Compressive moduli of 12.6 to 20.0% PVA hydrogel formulations swollen in 0.2 MPa at 37oC for up to four weeks .................................................................... 215
Figure 7.4: Compressive moduli of 19% PEG-DGE and 19% PEG-OH hydrogel formulations swollen in 0.2 MPa at 37oC for up to four weeks ............................. 216
Figure 7.5: Compressive moduli of 19% PEG-DGE with varying molecular weight (526, 2000 and 4600 Da) swollen in 0.2 MPa at 37oC for up to four weeks ................... 217
Figure 7.6: Compressive moduli of purified 19 to 39% PEG-DGE hydrogel formulations swollen in 0.2 MPa at 37oC for up to four weeks ................................................... 218
Figure 7.7: Compressive moduli of purified 12.6 to 20.0% PVA hydrogel formulations swollen in 0.2 MPa at 37oC for up to four weeks ................................................... 219
Figure 7.8: Relative crystallinity for formulations with varying PEG-DGE content swollen at 0.2 MPa and 37oC for four weeks ......................................................... 220
Figure 7.9: Relative crystallinity for formulations with varying PVA content swollen at 0.2 MPa and 37oC for four weeks .......................................................................... 221
Figure 7.10: Relative crystallinity for formulations varying PEG functional group swollen at 0.2 MPa and 37oC for four weeks (p>0.05) ....................................................... 222
Figure 7.11: Relative crystallinity for formulations varying PEG-DGE molecular weight swollen at 0.2 MPa and 37oC for four weeks ......................................................... 223
Figure 7.12: Relative crystallinity for purified formulations with varying PEG-DGE content swollen at 0.2 MPa and 37oC for four weeks ............................................ 224
Figure 7.13: Relative crystallinity for purified formulations with varying PVA content swollen at 0.2 MPa and 37oC for four weeks ......................................................... 225
xiv
LIST OF TABLES
Table 2.1: Linear region moduli of AF tissue [10] .......................................................... 42
Table 2.2: Linear region moduli of nondegenerated (nondeg.) and degenerated (degen.) AF tissue .................................................................................................................. 45
Table 4.1: Sample compositions for radiation crosslinked formulations .......................... 83
Table 4.2: Sample composition for difunctional crosslinked formulation ........................ 84
Table 4.3: Initial water content and mass retention of electron beam hydrogel formulations ............................................................................................................. 88
Table 4.4: Initial water content and mass retention of hydrogel formulations varying basic catalyst volume ......................................................................................................... 90
Table 4.5: Initial water content and mass retention of hydrogel formulations varying reaction time ............................................................................................................ 92
Table 5.1: Sample compositions for 9 to 39% PEG-DGE and 12.6 to 20.0% PVA hydrogel formulations ............................................................................................ 124
Table 5.2: Initial water content and mass retention for 19 to 39% PEG-DGE ............ 127
Table 5.3: Initial water content and mass retention for 12.6 to 20.0% PVA ................. 129
Table 5.4: Initial water content and mass retention for 19% PEG-DGE and 19% PEG-OH ......................................................................................................................... 131
Table 5.5: Initial water content and mass retention of 19% PEG-DGE with varying molecular weight (526, 2000 and 4600 Da) ........................................................... 133
Table 6.1: Initial water content and mass retention of 39% PEG-DGE unpurified and purified at various osmotic pressures and various periods of time .......................... 180
Table 6.2: Initial water content and mass retention of unpurified and purified 19 to 39% PEG-DGE hydrogel formulations ........................................................................ 182
Table 6.3: Initial water content and mass retention of unpurified and purified 12.6 to 20% PVA hydrogel formulations ........................................................................... 184
Table 6.4: Relative Crystallinity for Unpurified and Purified Formulations (p>0.05 is represented by ns) .................................................................................................. 191
Table 7.1: Compressive modulus for unpurified PEG-DGE and PVA formulations before and after swelling in osmotic solution ......................................................... 226
xv
Table 7.2: Water content for unpurified PEG-DGE and PVA formulations before and after swelling in osmotic solution ........................................................................... 227
Table 7.3: Compressive modulus for purified PEG-DGE and PVA formulations before and after swelling in osmotic solution .................................................................... 229
Table 7.4: Water content for purified PEG-DGE and PVA formulations before and after swelling in osmotic solution ........................................................................... 230
Table 7.5: Compressive modulus and relative crystallinity for unpurified PEG-DGE and PVA formulations before and after swelling in osmotic solution ........................... 228
Table 7.6: Compressive modulus and relative crystallinity for purified PEG-DGE and PVA formulations before and after swelling in osmotic solution ........................... 231
xvi
ABSTRACT
Development and Characterization of a Chemically Crosslinked Polyvinyl Alcohol/Polyethylene Glycol Hydrogel for Nucleus Pulposus Replacement
Valerie Regina Binetti
Anthony M. Lowman, Ph.D.
Low back pain caused by intervertebral disc degeneration is one of the most
common spinal disorders among patients seeking medical treatment. The most common
surgical treatments for disc degeneration are spinal fusion and total disc arthroplasty;
both of which are very invasive surgical procedures. Spinal fusion results in a loss of spinal
mobility and increased stress on adjacent intervertebral discs and while total disc
arthroplasty retains spinal mobility it is not FDA approved for multilevel replacement.
Nucleus pulposus replacement is an earlier stage intervention for disc degeneration before
multilevel interventions are necessary. One of the material classes being studied for this
application is hydrogel: a three-dimensional hydrated network of polymer(s), which
mimics the mechanical and physiological properties of the nucleus.
Previous nucleus replacement materials have included the poly(N-
isopropylacrylamide) (PNIPAAm) class of hydrogels and poly(vinyl alcohol) (PVA) and
poly(vinyl pyrrolidone) (PVP) hydrogels; the PNIPAAm hydrogel disadvantages are the
mechanical properties and implant shrinkage. While the PVA/PVP have the desired
mechanical properties, they are molded into string form and injected percutaneously
through a cannula. Due to this implantation method, there are issues with implant
movement and expulsion from the injection site. This is due to the fact that the implant
xvii
is a coiled string. PVA, PVP and poly (ethylene glycol) (PEG) hydrogels have previously
been shown to be great candidate materials for injectable nucleus pulposus replacement,
but have experienced issues with swelling and mass retention. The addition of chemical
crosslinking to the PVA/PVP/PEG hydrogel system will allow tailoring of the swelling,
mechanical, injectability and mass loss properties of the hydrogel network. Two chemical
crosslinking methods were evaluated for the PVA/PVP/PEG hydrogel system, resulting
in the selection of a difunctional crosslinking strategy using PEG functionalized with
terminal epoxide group (PEG diglycidyl ether) (PEG-DGE). The PVA/PVP/PEG-
DGE hydrogel system was characterized by compression and swelling experiments and
then the structure-property relationship was determined with the addition of morphology,
spectroscopy and crystallinity analysis. A purification technique was developed and
optimized to reduce the mass loss of the hydrogel network and then the structure-
property relationship of the new purified gel was investigated due to a change in the
gelation mechanism of the network after purification. The unpurified and purified
hydrogel formulations have mechanical and swelling properties in the desired range for
nucleus replacement, in addition, the purified hydrogel showed low cytotoxicity. Also, the
swelling mechanics of the hydrogel formulations were characterized in model osmotic
solutions to simulate the intradiscal environment.
1
1. INTRODUCTION
Low back pain is the most common among patients seeking medical treatment for
spinal disorders; it will affect 80% of people at some point in their lives [1] [2]. In the
United States 700,000 spine procedures are performed each year; while medical costs, loss
of work, and disability costs exceed fifty billion dollars per year [1] [3]. A major cause of
lower back pain is intervertebral disc degeneration; which is the result of damage or
dehydration to the nucleus pulposus, the inner most portion of the intervertebral disc.
The degeneration of the nucleus pulposus reduces the hydrostatic pressure on the internal
surface of the annulus fibrosus, the outer ring of the intervertebral disc. The reduction of
the hydrostatic pressure results in abnormal compressive stresses on the annulus fibrosus
which, can potentially lead to tears, cracks, and fissures after repeated loadings. As a
result of tears, cracks and fissures of the annulus fibrosus the nucleus tissue can migrate
through the annulus and impinge on the nerve roots causing back pain [1, 4-6]. In
addition, to altering the mechanics of the intervertebral disc, disc degeneration alters disc
height, and mechanics of the entire spinal column potentially adversely affecting the
behavior of the muscles and ligaments in the spine [7].
The most common surgical treatment for disc degeneration is spinal fusion [8], but
this results in loss of spinal mobility and increased stress on adjacent intervertebral discs
which can result in degeneration of these adjacent discs [6, 8]. A viable alternative to
fusion is total disc arthroplasty, which allows for the retention of spinal mobility [8].
Both spinal fusion and total disc arthroplasty are highly invasive surgeries. A less invasive
2
alternative is the replacement of just the nucleus pulposus. Several pre-formed implants
and injectable implants have been investigated. Some designs are currently going to
clinical trial. A concern with injectable implants is expulsion from the injection site; some
of the temperature-transitioning implants have poor mechanical properties. A chemically
crosslinked, cohesive implant is a promising alternative because the implant is injectable
through a small gauge needle. It creates a solid implant upon injection into the nuclear
cavity. The mechanical and swelling properties of the implant can be tailored through the
amount of polymer concentration and amount of crosslinking. This work focuses on
designing an injectable, chemically crosslinked polyvinyl alcohol hydrogel for nucleus pulposus
replacement. It is hypothesized that a candidate from this class of materials can be developed,
which upon implantation could prevent or postpone the annular degeneration process therefore
restoring the healthy biomechanics of the intervertebral disc and alleviate the pain associated
with degenerative disc disease.
The work will begin with the synthesis of injectable, chemically crosslinked PVA
hydrogels. Two crosslinking methods will be investigated: radiation and difunctional
crosslinkers. In addition, to varying the crosslinking method, properties such as PVA
content, radiation dosage, basic catalyst volume, and reaction time will be varied to
measure the limits of the different crosslinking methods. After one crosslinking method
is selected, the polymer content, functional groups, and molecular weight will be varied to
determine the structure- property relationship of the hydrogel system. Finally, the
hydrogel will be swollen in an in vitro model solution to mimic the intradiscal
environment the hydrogel will experience in vivo. The effect on the structure and
3
properties of the hydrogel network will be investigated after swelling and compared to the
unswollen hydrogel systems.
4
2. BACKGROUND
2.1. Spine Anatomy and Physiology
2.1.1. The Vertebral Column
The spinal column is formed from the junction of thirty-three vertebrae, which
provide structural support for the truck and protect the spinal cord (Figure 2.1). The
spinal column functions to transfer loads and bending moments of the head and trunk,
and external loads to the pelvis. The spinal column also allows sufficient physiological
movement and flexibility of the upper body, in addition, to protecting the spinal cord
from danger due to motion and trauma. It also provides protection to other vital internal
organs and is a base of attachment for ligaments, tendons and muscles. Each of the
thirty-three vertebrae are connected by an intervertebral disc (IVD) and are grouped into
five distinct regions: cervical, thoracic, lumbar, fused sacral and fused coccyx [9].
The IVD and spinal ligaments work together in a complex system providing
flexibility and mobility to the spine. In addition to allowing the mobility of the body to
twist, bend forward, bend backward and bend side-to-side; the IVDs and ligaments have
to maintain stability under large spinal loads. The flexibility of the spine must also be soft
enough to allow motion such as bending, torsion, tension, compression, and shear [10].
5
2.1.2. The Intervertebral Disc
Between the cervical, thoracic, and lumbar vertebrae is a fibrocarilaginous joint,
the intervertebral disc (IVD). The IVDs are the largest avascular structure in the body,
largely aneural and sparsely populated with cells. At each of the spinal levels: cervical,
thoracic, and lumbar the size and shape of the IVD varies but the IVD is roughly 7 to 10
mm thick and 4 cm in diameter (Figure 2.2) [7]. The variable in size and shape is to
accommodate the varying mechanical requirements at each level of the spine. What does
not vary with spinal level is the composition and general structure of the IVD. The IVD
is similar to other connective tissues (e.g. ligament, cartilage, tendon) in the body as it
consists of collagen fibers embedded in a highly hydrated extracellular matrix. Though
the composition of the IVD is similar to other connective tissues, the IVDs structure is
unique due to its multidirectional flexibility and large load bearing capacity.
The main components of the IVD are the annulus fibrosus (AF), nucleus
pulposus (NP) and the end plates. The end plates are situated above and below each IVD,
adjacent to the vertebrae. The AF is a fibrous ring that surrounds the NP, the gelatinous
hydrated center. The IVD components interact similar to a thick- walled pressure vessel
and allow the IVD to absorb and transmit the loads experienced by the spine. These
loads can be in complex combinations of torsion, tension, shear, compression and
bending [7, 10-12].
6
2.1.3. The Nucleus Pulposus
The centermost component of the IVD is the NP a translucent, gelatinous, semi-
solid structure. The NP is a loose meshwork of randomly distributed collagen fibers and
radially arranged elastin fibers embedded in a highly hydrated aggrecan containing gel.
Healthy, non-degenerated NP is composed of collagen (20% of the dry weight),
proteoglycans (30-50% of the dry weight), and the remainder of the dry weight is non-
collagenous proteins. Water constitutes 70-80% of the total NP weight. The collagen
consists of 80% type II collagen; the remainder is made up of types V, VI, IX, and XII
[13, 14]. The collagen is helically organized proteins bundled into fibers; these fibers
confer mechanical strength to tissues [15]. The NP is populated at a low density of
chrondrocyte-like cells which predominately produce type II collagen and aggrecan. The
cell density of the NP is 4 x 106 cells/cm3, which is an order of magnitude lower than the
cell densities, 14 x 106 cells/cm3, of other acellular tissues such as cartilage [16].
The proteoglycans within the NP are important for the function of the tissue.
Aggrecan, a proteoglycan present in the NP, is a brush-like structure composed of
multiple glycosaminoglycan molecules with a core protein backbone. Many aggrecan
molecules aggregate together to form a proteoglycan macromolecule. The
glycosaminoglycan molecules that make up the NP contain fixed negative charges, which
attract positively charged ions to achieve electroneutrality within NP tissue. The ion
concentration needed to achieve electoneutrality in the NP tissue results in a higher
concentration than in the surrounding tissue; this concentration gradient attracts water
into the NP tissue due to osmotic pressure within the NP. The high water pressure in the
7
NP is responsible for the largely fluid-like behavior of the tissue [10]. The combination
of the water in the NP, the osmotic imbalance in ion concentration and repulsion of the
negative charges on the glycosaminoglycan molecules pressurizes the NP within the IVD.
This total pressure within the NP is referred to as the swelling pressure; it is constrained
by the type II collagen fiber mesh within the NP and by the surrounding AF and end
plates. This swelling pressure of healthy NP in a recumbent position is 0.1 to 0.2 MPa
and when lifting or standing can reach 1 to 3 MPa [10, 17, 18]. Similar pressures have
been measured in cadaveric motion segments under externally applied loads [19].
The pressurization of the NP allows it to absorb and transmit the compressive
loads of the spine. When the spine is loaded in compression, the pressure in the NP
increases which over time leads to water flow out of the NP to equilibrate the pressure
within the IVD. This fluid flow in and out of the IVD is diurnal; the IVD is loaded in
compression for 16 hours a day which results in a large amount of fluid volume to flow
out of the IVD. The IVD is rehydrated and repressurized overnight during rest; this
repressurization increases by 0.1 to 0.24 MPa, which is between 20-50% of the total IVD
pressure during relaxed standing [18].
The mechanical properties of the NP are isotropic, the same in all directions, is
due to the random organization of the NP tissue. The compressive modulus of the NP
has been measured to be approximately 1 MPa [20] and the shear modulus has been
measured to be roughly 6 kPa [21]. The combination of the mechanical and swelling
properties of the NP suggest that the NP of a healthy IVD is largely fluid-like and loads
are primarily supported due to pressurization.
8
2.1.4. The Annulus Fibrosus
Surrounding the NP is a ring of highly organized fibrocartilage known as the
annulus fibrosus (AF) (Figure 2.2). The AF is composed of 15 to 40 concentric layers of
collagen fibers oriented at alternating angles, embedded in a proteoglycan matrix. The
angled orientation of these fibers contributes to the anisotropic mechanical properties of
the AF. The mechanical function of the IVD is dependent on this layered structure of
the AF to withstand large and complex loads. During compressive loading of the IVD,
the inner portion of the AF is exposed to axial compressive stresses, the outer AF
experiences radial compressive and circumferential tensile stresses from the bulging NP.
When the IVD is loaded in bending or torsion, the fibers of the AF may be loaded
directly in tension. Under typical loading of the spine, the IVD can experience any
combination of these loading scenarios. As previously stated, the angled orientation of
the AF fibers contribute to the anisotropic properties of the AF; the tensile
circumferential modulus is 10-20 times greater than the axial modulus and an order of
magnitude greater than the radial modulus (Table 2.1). The fibers of the AF reorient
during circumferential tensile loading, this reorientation may significantly increase
mechanical properties in this direction [22]. The complex loading conditions of the AF
are accommodated by the anisotropic mechanical properties; such as the tensile loading
occurs primarily in the circumferential direction so the tensile properties of the AF are
greatest in that direction. While the tensile properties are highly anisotropic, the
compressive properties are not suggesting that they are not influenced by collagen fiber
9
direction. The compressive modulus of the AF is 0.6 MPa, water content is thought to
contribute to the compressive properties similar to the NP [23].
The AF also has nonlinear and viscoelastic material properties. When samples of
AF tissue are loaded they exhibit a nonlinear stress/strain curve; the curve shows a “toe”
region where low stresses are observed at low strains and at high strains the material
exhibits high stress in the linear region of the curve, after the linear region the material
fails. The modulus of the AF in the toe region is approximately 2 to 5 MPa
(circumferential direction) and in linear region the modulus is approximately 20 MPa [22,
24]. This material behavior of the AF is similar to other soft tissues such as articular
cartilage and ligaments. Figure 2.4 shows the nonlinear stress/strain curve of collagen
fibers; initially they are wavy, when load is applied they become uncrimped. There are
also interactions between the collagen fibers and the proteoglycan matrix, which
contributes to the nonlinearity. The viscoelastic properties of the AF result in time-
dependent material behaviors (e.g. stress relaxation, creep). Fluid flow through the
permeable matrix of the AF and frictional interactions between collagen fibers and the
proteoglycan matrix may contribute to the viscoelastic behavior of the AF [25, 26].
The inhomogeneous and anisotropic mechanical properties of the AF can also be
attributed to the biochemical composition. From the outer to the inner AF and from the
anterior to the posterior: the water, collagen and proteoglycan content vary. The dry
weight of the outer AF is 60 to 70% collagen, 10% proteoglycans and the remainder is
non-collagenous proteins. The outer AF is dense and fibrous with clearly defined layers
of highly organized fibers. The major collagens in the AF are types I and II, the ratio of
10
type I to type II collagen is very high in the outer AF and moving radially inward the
ratio changes so that type II is predominate in the inner annulus. The composition of the
inner AF is lower in collagen (23 to 30% of dry weight) and contains a higher percentage
of proteoglycans and hydration [13, 27].
The cells in the outer AF region are fibroblast-like: thin, elongated and aligned
parallel to the collagen fibers. The inner AF cells are more oval and chondrocyte-like.
Unique to the cells of the AF and NP, not seen in articular cartilage, are long thin
cytoplasmic projections (> 30 µm in length). The functions of these cells in the disc are
unknown, but it has been suggested that they may act as sensors and communicators of
mechanical strain within the tissue [7, 10].
2.1.5. End Plate
The end plate is a thin horizontal layer of hyaline cartilage, roughly 1 mm thick,
located between the IVD and the vertebral body (Figure 2.3). The end plate collagen
fibers run horizontal and parallel to the vertebral bodies, with fibers continuing into the
disc [7].
2.2. Degenerative Disc Disease
With age, IVDs undergo changes in structure, composition and mechanical
function. In addition to aging, IVD are also susceptible to degenerative disc disease
11
(DDD); the effects of both are very similar and difficult to differentiate. There are many
theories about the onset and progression of DDD including genetics and environmental
influences [28, 29] but scientific evidence to support this are inconclusive [28] due to the
fact that other factors (e.g. socioeconomic status) are difficult to separate from
employment status. It has been suggested that smoking and obesity are thought to
contribute to DDD [30]. IVDs degenerate far earlier than other musculoskeletal tissues,
the first findings of lumbar disc degeneration is seen in the age group of 11 to 16 years.
Twenty percent of people in their teens have mild signs of degeneration; which increases
with age, particularly in males, resulting in 60% of 70-year-old discs being severely
degenerated [7].
The degenerative changes in the IVD result in a loss of separation between the
NP and the AF, loss of disc height, altered loading of the IVD and surrounding tissues,
and a loss in disc height after loading due to dehydration. The responsible factors for
DDD are not known nor is the specific sequence of events. DDD may lead to low back
pain, which is among the top ten reasons for doctors visits in the United States, with
direct costs of $25 billion [10, 31].
2.2.1. Nucleus Pulposus
Early in life, large aggregating proteoglycans in healthy NP tissue begin to break
down, starting the aging and degeneration process, results in a decrease to total
proteoglycan content from 30-50% down to as little as 10% by adulthood [32]. The
12
degraded proteoglycans remains in the NP, affecting the ability to attract and bind water;
resulting in water content decreasing from 90 total NP weight percent in childhood to
less than 70 total NP weight percent in the elderly [13, 33]. In a healthy IVD, there is a
balance between proteoglycan synthesis and degradation, and inflow of nutrients and
outflow of waste products. As the NP tissue degenerates, there is an increased level of
matrix metalloproteinases (MMPS) the enzymes that degrade proteins (e.g.
proteoglycans and collagen). In healthy tissue, a cascade of mechanical and biochemical
factors regulate the MMP production; this cascade is disturbed with age and
degeneration [34]. In addition to the loss of proteoglycans from degradation, as the IVD
degenerates the NP cells are unable to synthesize proteoglycans at the rate they are
destroyed. The production of proteoglycans and collagen in the NP may also be affected
by the decrease in IVD nutrition that occurs with degeneration [10].
As the IVD degenerates, the NP collagen composition and overall structure
change. The total type II collagen amount in the NP decreases, and type I collagen
increases [33, 35]. The NP, which in a healthy IVD is gelatinous and translucent,
becomes firmer and then the color changes from white to yellow or brownish due to
oxidation from poor nutrition and waste product accumulation (Figure 2.5) [27]. As the
NP degenerates and become more of solid than a liquid [36], the material properties of
the NP change to that of a solid-like material. The shear modulus of the NP increases by
up to 80% [21], the swelling pressure decreases from 1-2 MPa to 0.03 MPa or less [17,
18, 20] and the compressive stiffness decreases from 1.0 MPa to 0.4 MPa [10, 20].
13
2.2.2. Annulus Fibrosus
As the IVD degenerates the structure, composition and function of the AF are
affected; it is thought that these changes may be a result from changes initiated in the NP.
Loss of water in the NP prevents the NP tissue from pressurizing, resulting in the
prevention of the NP tissue from absorbing and transferring the compressive loads of the
spine. When the NP is not pressurized the layers of the AF bulge inward in compression,
rather than outward in tension [37, 38]. As the loading of the AF changes due to changes
in the NP, the shear stresses and stress concentrations seen in the AF are increased; this
may lead to cracks, tears or fissures in the tissue or in delamination of the AF layers [10,
39] .
As the AF ages and degenerates, the number of layers through the radial
thickness decrease and each layer becomes thicker [40], in addition to the layers
becoming less distinct and disorganized. These changes to the structure of the AF affect
the mechanical properties of the tissue (Table 2.2). The circumferential linear-region
modulus does not experience significant change, but the circumferential toe-region
modulus in the outer anterior region increases from 2.5 to 5.7 MPa which is likely a
result of water content changes [22]. In addition, the shear modulus of the AF increases
[41] and failure strain decreases [42]. These changes in the mechanical properties of the
AF alter the loading patterns on surrounding tissues including the vertebrae, muscles and
ligaments. These changes in the mechanical environment of the degenerated AF may
result in the biochemical and cellular changes seen in the AF as the IVD degenerates.
The type I collagen content of the AF in degenerated IVDs decreases from 50 to 40% of
14
total collagen content; while the type II collagen content increases from 50 to 60% of
total collagen content [13], this is thought to happen to better withstand the compressive
loads. In addition to the collagen content changing, the ratio of type I to type II collagen
content from the outer to the inner AF changes. The outer AF increases in type II
collagen and the inner AF increases in type I [13]. With degeneration, the collagen
within the AF becomes cross-linked and denatured; these modifications and oxidation of
the collagen cause the discoloration of degenerated AF tissue (Figure 2.5) [43].
The change in AF tissue mechanics might also result in cell death or altered cell
synthesis. Similar to the cells of the NP, the AF cells are subject to density and
nutritional limitations of the changing environment of the degenerated IVD. The low
cell density, cell nutrition and buildup of waste products affect the ability of the AF cells
to synthesize collagen and proteoglycans necessary for proper function of the AF [10].
2.2.3. End Plates
The cartilaginous end plates of the IVD thin, become calcified, and the blood
supply to the end plates diminish as the disc degenerates [44]. During degeneration of
the IVD, there is occlusion of the marrow contact channels in the end plates [45]; this
prevents the transport of glucose and oxygen into the IVD and the removal of waste
products such as lactic acid [46, 47]. The degeneration of the end plates affects the
biochemical environment of the entire IVD, impacting cell metabolism in the NP and
AF. The stiffness of the end plate is not affected by degeneration, but changes in the
15
loading of the AF and NP leads to increased loading of the periphery of the end plates
which can result in end plate fractures in the periphery [10, 48, 49].
2.2.4. Relationship to Low Back Pain
Low back pain may originate from the IVD via various mechanisms. As the IVD
degenerates the loss of disc height and structure may result in pain. The loss of disc height
may contribute to altered loading of the vertebral bodies and facet joints; this altered loading
can result in pain and possible arthritis of the facet joints [12]. If the altered mechanics of the
IVD results in bulging of the disc, this bulging can result in nerve root impingement, which
can cause pain in areas of the body enervated by the impinged nerve. The degenerated IVD
may also release mediators that sensitize nerve endings [50]. Other conditions related to IVD
degeneration include spinal stenosis, a narrowing of the vertebral opening due to thickening
of the ligaments, bones and facet joints adjacent to the space [51]. When the disc herniates,
the NP material protrudes though the weakened AF, which can result in nerve root
impingement. Other potential sources of pain are end plate degeneration and bony
protrusions on the rims of the vertebral bodies. It is not known how any of these conditions,
including DDD itself generate low back pain [10].
16
2.3. Clinical Treatment Options
2.3.1. Discectomy
One of the two categories of surgical interventions for back pain is decompression,
which includes laminotomy, laminectomy and discectomy procedures [52]. These
procedures relieve pressure on the nerve elements by excision of disc, bone or ligament
material [5]. A laminotomy is a small hole in the disc material to free the nerve root to relieve
nerve compression. If lamintomy is unsuccessful, a laminectomy can be performed. A
laminectomy is the removal of a small portion of disc tissue or facet joint impinging on the
nerve. These procedures provide relief as soon as the inflammation subsides [53]. A
discectomy is for herniated discs; in this procedure the portion of the NP, which is
impinging on the nerve root, is removed. A microdiscectomy, which is done with a small
incision, is the gold standard due to no other discectomy technique being able to match or
exceed its outcomes [54].
A study by Wu et al. [55] compared the outcome of percutaneous discectomy to that
of conservative treatment. Discectomy was performed on patients with disc herniation that
were symptomatic for 6 to 12 weeks. A follow up at two years showed that there were no
clinically significant differences in pain or quality of life between the surgical and
conservative treatment groups [55].
17
2.3.2. Spinal Fusion
The second of the two categories of surgical interventions for back pain is
stabilization; which includes procedures such as vertebroplasty, kyphoplasty and spinal fusion.
Vertebroplasty is the injection of bone cement into an area of vertebral compression fracture;
stabilizing the vertebral body by filling the defect in the bone [52]. The injection of the
exothermic poly(methylmethacrylate) bone cement may result in surrounding tissue damage
due to the heat produced from the curing reaction [56]. Spinal fusion uses arthrodesis to
prevent motion across the pain generative disc by the removal of disc material, roughening
the surfaces of the two opposing vertebral bodies and packing with bone material allowing to
fill the gap between the bones in order for them to grow into a single segment [6, 57, 58].
Metal implants can be used to stabilize the vertebrae until the fusion solidifies [52]. The
fused segments can be anteriorly, posteriorly or circumferentially depending on the area of
the defect. The bone graft that is used in a spinal fusion can be an autograft or an allograft;
autografts are bone removed from the patient’s iliac crest, allografts are obtained from a
donor cadaver.
Between 1996 and 2001 fusion rates in the United States rose 77% after the Food
and Drug Administration (FDA) approved intervertebral fusion cages (Figure 2.6) in 1996
[59]. Fusion cages provide stabilization to the fused segment and provide mechanical support
while the bone matures [60]. Due to the high stiffness of some metals used can cause
endplate subsidence [6], titanium and carbon cages are recommended for that reason [60,
61].
18
Spinal fusion provides pain relief for some patients, though its efficacy for treating
DDD remains unclear [60]. Mofidi et al. [60] surveyed 65 patients who received posterior
lumbar fusion with carbon cages; four years after the surgery there was an 84% satisfaction
rate and 61% of the patients were able to return to pre-disease activity level. The UK
Medical Research Council followed 349 patients with a one-year history of chronic low back
pain; half of the patients received a fusion, which varied in surgical approach and
instrumentation, and the other half received intensive rehabilitation for three weeks in
addition to cognitive behavioral therapy. Two years after surgery the surgical group had
slightly lower pain but the groups did not differ in other outcomes (e.g. anxiety, depression
or adverse effects) [62]. Another study in Norway, with 64 patients was done where patients
either received fusion or three weeks of physical exercises and cognitive behavioral treatment
for lumbar degeneration; no differences were found between the groups one year after
surgery [63].
Spinal fusion is associated with surgical complications such as infection [60], nerve
injury [64] and high blood loss [65]. Long-term results of fusion are questionable.
Biomechanical studies have shown that fusion causes increased stress to adjacent spinal
segments to the fusion site [66]; which can promote degeneration of these adjacent segments
and eventually lead to back pain again. Miyakoshi et al. [67] found that disc heights adjacent
to the site of fusion decrease.
19
2.3.3. Total Disc Arthroplasty
Total disc arthroplasty or total disc replacement is a treatment for the advanced
stages of DDD. Disc arthroplasty is to restore pain free motion and mechanical function
to the degenerated spinal unit, which places it at the other end of the spectrum of spinal
fusion. Disc arthroplasty has a theoretical advantage over fusion because it can more
closely mimic the loading and motion characteristics of a healthy spine [68]; in addition
to the hypothesis that preservation of motion will decrease stress on adjacent segments
[69, 70]. Disc arthroplasty is emerging as an alternative to fusion. The benefits of disc
arthroplasty are removal of the painful disc, restoring disc height, improving stability and
restoring a healthy pattern of load bearing to the spine [70].
The CHARITÉ artificial disc (DePuy Spine, Johnson & Johnson, Raynham, MA)
(Figure 2.7) has the longest clinical history of all of the artificial discs; it was developed in
the early 1980s and is a polyethylene core sandwiched between to metal endplates. Since its
original design, it has been through three design revisions in order to minimize
complications such as subsidence and fatigue failure. The CHARITÉ was the first total
disc implant for lumbar spine to gain FDA approval [70]. The implant consists of two
concave cobalt chromium molybdenum alloy endplates, of which the surfaces facing the
vertebral endplates are covered in porous titanium and coated with calcium phosphate to
encourage bonding with the bone. There are six teeth on the surface of the endplate,
which physically anchor the implant into the vertebral body. Between the endplates, there
is a free-floating biconvex sliding core made of ultra high molecular weight polyethylene
20
(UHMWPE) which mimics the major movements of the intervertebral segment: flexion,
extension and translation [71].
An FDA regulated prospective, randomized study was conducted for the purpose of
FDA approval for the CHARITÉ. The study compared the safety and effectiveness of the
CHARITÉ to anterior lumbar fusion for cases with single level lumbar degeneration. 304
patients were followed over 2 years; with 205 patients receiving the CHARITÉ implant
and the remainder receiving a BAK cage and an iliac crest bone graft [72, 73]. Overall
patients in both groups improved after surgery; patients who received the CHARITÉ
recover faster, had lower disability levels, statistically lower pain scores, and had shorter
hospital stays [72]. The CHARITÉ group had a 13.6% increase in mean
flexion/extension ROM and the fusion group had an 82.5% decrease. At the two year
follow up, there was no implant wear or creep found [73]. A major criticism of the study
was the fusion procedure; the BAK cage has a poor clinical history due to its high failure
rate [74]. This cage was chosen because at the time of the study it was the only FDA
approved anterior cage. Despite the improvements in pain scores, 64% of the CHARITÉ
group and 80% of the fusion group remained on narcotics two years after surgery [72].
Additional disc replacement devices are ProDisc-L total disc replacement (Synthes,
West Chester, PA), Mobidic disc prosthesis (LDR, Troyes, France), FlexiCore
intervertebral disc (Stryker Spine, Allendale, NJ), Kineflex, Activ-L artificial disc (Spinal
Motion, Inc., Mountainview, CA), Maverick total disc replacement (Medtronic Sofamor
Danek, Memphis, TN), and Theken eDisc (Theken Disc, Akron, OH) [75].
21
2.3.4. Nucleus Replacement and Stabilization Technologies
Another non-fusion alternative being investigated is the replacement of the
nucleus pulposus. The idea is to use a synthetic material to restore healthy biomechanics
to the spine; the replacement of the NP material will restore the biomechanical function
of healthy NP by applying tension to the AF under compressive loads [5, 76]. In addition,
motion can be preserved and disc height can be restored [1, 5, 76, 77]; disc height
restoration can help lessen compressive forces on facet joints [77, 78]. The implantation
procedure for a nucleus replacement has the potential to be less invasive than either the
total dis replacement or spinal fusion [79]; which avoids the morbidity of those
procedures.
There are various thoughts on what degree of degeneration is acceptable for
nucleus replacement [80]. It has been proposed that nucleus replacement is for people in
earlier stages of degeneration, when the annulus is still intact [81] and there have been no
previous procedures [80]. In these cases, nucleus replacement is seen as an early therapy,
rather than a replacement for total disc replacement or spinal fusion [5]. An
uncompromised annulus is better able to contain the implant, reducing the risk of
implant migration [77]. In addition, for in situ forming nucleus replacement materials, a
competent annulus is necessary to contain the liquid before it becomes a solid implant
[82]. When degeneration progresses to discogenic back pain, the annulus may already be
in later stages of degeneration [80]. It has been proposed that nucleus replacement be an
22
adjunct procedure to discectomy or nucleotomy; with the synthetic material filling the
space to restore function to the disc [81].
When a patient becomes symptomatic with low back pain, nerve impingement
may not be the only cause of pain. In addition to nerve impingement, pain can be
generated by inflammatory mediatiors, which can access and stimulate the nociceptors in
the tears of the AF. In situations like this, just replacing the nucleus will not provide
relief to the patient as the AF is not sufficient [80]; though pain might be able to be
prevented with the use of soft tissue adhesive closing annular fissures.
Material requirements for synthetic nucleus replacement include fatigue, stiffness,
and space filling capability. The material must be able to endure cyclic fatigue without
failure or formation of particle debris [77, 79, 82]. It also needs a stiffness that the load
distribution on the endplates and vertebral bodies will not cause subsidence or stress
shielding, leading to bone resorption [82, 83]. The material must also completely fill the
nuclear cavity, to avoid significant movement [82]. There also needs to be contact
between the material and the inner AF to fully restore the function of the IVD [84, 85].
Hydrogels are the material with the most potential for nucleus replacement
devices. These three dimension hydrated polymer networks are favored for nucleus
replacement due to their ability to mimic natural nucleus material. They have the ability
to exude water under loads and re-imbibe it when unloaded, very similar to that of the
native nucleus [5, 79, 82, 86]. A difference between the nucleus and some hydrogel
nucleus replacement material is their fluid loss under sustained loading: nucleus tissue
23
undergoes a gradual fluid loss where hydrogel material has to be designed with a low
hydraulic permeability [87]. This property is important so that the implant maintains
hydrostatic pressure on the annulus tissue under sustained loading. There are two
categories of hydrogels for nucleus replacement: preformed and in situ curing. Preformed
implants are of a predetermined size and shape; the hydrogel is generally dehydrated to
minimize invasiveness of the implantation procedure. The implants generally swell once
implanted in the nuclear cavity in the presence of physiological fluids [82]. The next
sections will be a brief overview of a selection of nucleus pulposus replacement
technologies.
2.3.4.1. DASCOR
The DASCOR device (Disc Dynamics Inc. Eden Prairie, MN) is a methylene-
diphenyldiisocyanate (MDI) - based polyurethane two-part reactive system (Figure 2.8).
This system is then injected, under controlled pressure, through a catheter into a balloon
located in the nuclectomy space. Upon injection the polymer cures, with an exothermic
temperature of less than 50oC, during this curing process the polymer bonds to the walls
of the balloon forming the final device. A typical compression modulus of the DASCOR
device per ASTM D575 ranges between 4 and 6 MPa, with an ultimate compressive
stress of 25 MPa and strain of greater than 90% [75].
The DASCOR device has been in clinical trials outside of the United States since
2003. DASCOR received CE-Mark approval for commercial sale in the European
Union in July 2005. The U.S. Food and Drug Administration (FDA) did not approve
24
the DASCOR device for a Pivotal clinical trial in the United States, due to the need for
additional data. Disc Dynamics Inc. shut down in 2009 due to inability to secure
additional capital to continue clinical trials [75].
2.3.4.2. PDN-SOLO and HydraFlex
The HydraFlex device (Raymedica, Inc., Minneapolis, MN) is an optimization of
the PDN-SOLO (Raymedica, Inc., Minneapolis, MN), which is a hydrogel-based
technology. There are three components in the device: an inner copolymer hydrogel
pellet, an outer woven jacket of ultra high molecular weight polyethylene (UHMWPE)
fibers, and platinum-iridium wire markers for radiologic identification. The hydrogel core
is a proprietary copolymer of polyacrilonitrile and polyacrylamine which is molded and
then dehydrated, allowing for ease of insertion. The hydrogel formulation for the
HydraFlex absorbs up to 80% of its dry weight in water faster than the hydrogel
formulation in the PDN-SOLO, the HydraFlex rehydration begins immediately after
insertion and swells over 7 to 10 days. The UHMWPE fibers allow for the rapid
rehydration but control the device expansion preventing possible damage to the end
plates. The HydraFlex has shorter tab lengths than the PDN-SOLO, which allows for
greater hydrogel volume resulting in a more compliant implant (Figure 2.9).
The pellet and jacket were subjected to standard fatigue testing for up to 50
million normal range compression and to 10 million compression-translation cycles; the
components passed without deterioration. The terminal burst strength of intact implants,
after prolonged cyclic tests, exceeded the 6 kN limit of the test machine [75].
25
In a prospective multicenter international trial of the PDN-SOLO device, a 74%
reduction of Oswestry Disability Index was observed. Whereas published report for two
fusion cages are 41% and 42%, and 50% for a total disc replacement. The Visual Analog
Scale, a psychometric response scale used in questionnaires, declined from 7.5
(preoperative) to 2.7 at 12 months after surgery. Additionally a patient satisfaction survey
of 300 recipients demonstrated a much better or better response in 87% of patients [75,
76].
2.3.4.3. NeuDisc
NeuDisc (Replication Medical, Cranbury, NJ) is a hydrogel device that mimics
the physiological function of the nucleus pulposus (Figure 2.10). The NeuDisc
proprietary layered hydrogel structure distributes axial loads in the disc and mimics the
osmotic properties of the nucleus. The device is a vertically layered structure of soft
hydrogel between Dacron knitted mesh. The Dacron allows the device to achieve the
necessary stiffness but still allows the device to be soft enough to not damage the end
plate. Upon implantation, the device expands in thickness from 2 to 15 mm, while the
footprint size does not enlarge. Due to swelling in thickness only, there is no undesired
radial pressure on the weakened annulus [75, 88].
The NeuDisc device has been undergoing a two-arm pilot European prospective
longitudinal study since June 2005. The first arm was using an anterolateral transpsoatic
approach (ALPA); the second arm is a posterolateral endoscopic approach. A total of 15
26
implantations have been preformed, two revisions were performed due to infection. Early
clinical outcomes indicate early resolution of symptoms at 12 months [75].
2.3.4.4. NuCore Injectable Nucleus
The NuCore Injectable Nucleus (Spine Wave, Inc., Shelton, CT) device consists
of amino acid sequence blocks derived from silk and elastin structural proteins; more
specifically, it is a sequential block copolymer of two silk blocks and eight elastin blocks
per polymer sequence repeat (Figure 2.11). The NuCore material mimics the protein
content, water content, pH, and complex modulus of the natural nucleus pulposus. The
copolymer is mixed with a diisocyanate-based crosslinking agent and is injectable for
approximately 90-seconds; after which, the material becomes a viscous gel. After 30
minutes, the copolymer plus crosslinker reach near-final mechanical strength [75].
A pilot clinic study of the NuCore Injectable Nucleus device in Switzerland was
conducted on 14 patients (8 male, 6 female) over a period of 17 months. The results of
the study showed that the device is safely implanted into lumbar discs following
microdiscectomy; there were no complications or adverse events related to the material.
Disc height was maintained with the use of NuCore Injectable Nucleus better than
microdiscectomy over 2 years [89]. According to the Spine Wave website, patients are
currently being enrolled in clinic trials in Switzerland, Germany, Australia and the
United States [90].
27
2.3.4.5. Aquarelle
Aquarelle nucleus (Stryker Spine, Allendale, NJ) is a preformed nucleus pulposus
replacement (Figure 2.12). It is the first hydrogel device for NP replacement, with
development starting in 1990 by Howmedica and then acquired Stryker [4]. The
Aquarelle device is made of poly(vinyl alcohol) (PVA) and is physically crosslinked
through repetitive freeze-thaw cycles. The implanted hydrogel contains approximately
80% water. The device has shown biocompatibility, due to the lack of crosslinkers used in
the physically crosslinked gel, and mechanical durability up to 40 million cycles [75, 79].
A baboon study was conducted on twenty-eight skeletally mature male baboons,
which were sacrificed at 1, 3, 6, 12, and 24 months after surgery. In the first twenty-two
animals an anterior approach was used and the discectomy was preformed through a
midline annular incision with a 3-mm trephine. Fifteen of the animals were implanted
with Aquarelle; the other seven received discectomy as a control. The last six animals
received a modified surgical technical so that the trephine was inserted through a
percutaneous portal in the left flank and a 2.5-mm grasper was used to remove the
nucleus. Of the last 6 animals, five received Aquarelle and one was used as a control. Six
out of the twenty implants extruded from the disc. This high extrusion rate might be due
to the limitation of the animal model, surgical technique, the high intradiscal pressure of
the baboon, the low modulus, and slippery surface of the Aquarelle device. After this
clinical result, Stryker halted further development of the device [75, 91].
28
2.3.4.6. BioDisc
The BioDisc Nucleus Pulposus Replacement (CryoLife, Inc., Kennesaw, GA) is
an injectable in situ polymerizing protein hydrogel. Using a controlled delivery device and
delivery tips, where the glutaraldehyde begins crosslinks bovine serum albumin (BSA)
molecules to each other and to the patient’s proteins at the repair site with covalent bonds
(Figure 2.13). This crosslinking creates a flexible, viscoelastic hydrogel, which is
covalently bonded to the patient’s tissues. This material is based on Cryolife’s surgical
adhesive BioGlue. Polymerization of the glutaraldehyde and BSA begin immediately
upon injection into the delivery tip, followed by crosslinking with the patient’s tissues one
the material is injected, full strength of the material is reached within two minutes [75].
Cadaver tests on calf lumbar segments showed that posterior injection of the
BioDisc hydrogel restored disc height and segment stability; in addition, minimal
reduction in height was seen after 10 million loading cycles [92]. Eleven patients have
received the BioDisc device in a safety study; no intraoperative or immediate
postoperative complications have been encountered. Additionally, all patients had relief
of their leg pain immediately; two-year follow up of this study is under way [75].
2.3.4.7. Biomimetic Aggrecan
Current research for nucleus pulposus is to mechanically stabilize the
intervertebral disc in the early stages of degeneration as opposed to replacement of the
nucleus. Sarkar et al. [93, 94] has developed a hybrid-bio/synthetic biomimetic
29
proteoglycan macromolecule for injection into the disc in the early stages of degeneration.
Chondroitin sulfate-marcomolecules which mimic the proteoglycan structure of aggrecan
is produced by incorporating natural chondroitin sulfate chains into a bottle brush
polymer structure. The chondroitin sulfate-marcomolecules will mimic the proteoglycan
structure and function, but will resist enzymatic degradation. Various bottle brush
synthesis techniques have been explored including “grafting-to” and “grafting-through”.
The “grafting-to” strategy immobilized chondroitin sulfate via a terminal primary amine
onto a polyacrylic acid backbone; the “grafting-through” strategy was an epoxy-amine
step-growth polymerization technique resulting in a chondroitin sulfate-macromolecule
with polyethylene glycol backbone segment. The macromolecule synthetized may also be
used to mimic other macromolecules in the proteoglycan family for applications in
regenerative and restorative medicine.
2.4. Poly(vinyl alcohol)
Poly(vinyl alcohol) is a simple chemical structure with a pendant hydroxyl group
(Figure 2.14), produced by the polymerization of vinyl acetate to poly (vinyl acetate)
(PVAc) followed by hydrolysis of PVAc to PVA. The hydrolysis reaction of PVAc to
PVA does not go to completion, resulting in PVA with a degree of hydrolysis that
depends on the extent of the reaction. This degree of hydrolysis affects the chemical
properties, solubility and crystallinity of PVA [95].
30
2.5. Poly(vinyl alcohol) hydrogels
PVA is crosslinked, in a variety of different ways, to create hydrogels; hydrogels
are a hydrophilic crosslinked polymer network that swells in the presence of water or
biological fluids. The high water content and elastic characteristics of hydrogels allow for
their ability to mimic human tissue better than any other class of synthetic biomaterial
[96]. Methods for crosslinking PVA are physical or chemical.
PVA is physically crosslinked though heat treatments [97], phase separation using
theta-solutions [98-103], and freeze-thaw cycling [95, 104, 105] ; all of these methods
results in crystallization of PVA. Physically crosslinked PVA hydrogels, in physiological
environments can undergo dissolution; a process involving a melting of PVA crystallites.
It has been shown that the addition of poly(vinyl pyrrolidone) (PVP) (Figure 2.15)
stabilizes the PVA though hydrogen bonding interactions between the carbonyl group of
the PVP and the hydroxyl group on the PVA (Figure 2.16).
PVA hydrogels are formed, via the freeze/thaw method, by the freezing of the
water in the PVA solution forcing PVA to phase-separate and form concentrated PVA
domains; upon thawing the PVA in the PVA-rich domains hydrogen-bonds to form
crystals resulting in a physically crosslinked gel. In the theta-gel method, the gelling
agent (e.g. low molecular weight PEG) reduces the quality of the solvent (e.g. water) in
the PVA solution. This results in the PVA phase separating into polymer rich regions
and solvent rich regions; within the polymer rich regions the PVA crystallizes without the
need of freeze/thaw cycles.
31
PVA hydrogels can also be formed by syneresis, which is a gradual change in the
structure of the gels accompanying shrinkage of their volume and exclusion of the solvent.
This does not involve shrinkage that is caused by evaporation of the solvent; instead
syneresis is shrinkage caused by inner structural change. This inner structural change can
be from slowly occurring phase separation of the network and/or crystallization, as the gel
when originally formed is not in its equilibrium state [106, 107].
PVA is chemically crosslinked using difunctional crosslinking agents, or radiation
for the formation of a chemically crosslinked PVA hydrogels. Examples of difunctional
agents include: dialdehydes (e.g. glutaraldehyde) [95] and diepoxides (e.g. 1,3-butadiene
diepoxide) [108]. Residual difunctional crosslinking agents, in particular short chain
agents such as glutaraldehyde, can be cytotoxic even at small concentrations, which is a
concern for biomedical applications. Longer chain difunctional crosslinking agents, such
as polymers with functional end groups, are thought to be better options for crosslinking
PVA. The large chains are less likely to diffuse into the body as readily as short chain
crosslinkers such as glutaraldehyde [109, 110]. Chemical crosslinking via radiation, such
as electron beam or γ-irradiation, is preferred for biomedical applications since cytotoxic
agents are not used and possibly left behind. Minimum dose needed for gelation depends
on degree of polymerization and polymer concentration in solution [95].
2.6. Poly(ethylene glycol)
Polyethylene glycol (PEG) is a linear polyether, which contains oxygen in its carbon
backbone (Figure 2.17); it has been called one of the best biocompatible polymers [111].
32
The polyether backbone of PEG is fairly chemically inert. PEG diglycidyl ether is PEG with
terminal epoxide ends (Figure 2.18); it is produced by reacting the terminal hydroxyl groups
of PEG with epichlorohydrin to introduce electrophilic epoxide groups onto the polymer
[111]. The addition of the epoxide end groups allows for covalent bonding between PEG
and functional groups such as hydroxyls and amines.
2.7. Poly (vinyl alcohol)/Poly (ethylene glycol) Hydrogels
A lesser-known way of producing PVA hydrogels is using the theta-gel method
developed by Ruberti and Barithwaite [100-103]; this method produces physically
crosslinked PVA hydrogel that are not created by freeze/thaw cycling. In the theta-gel
method, the gelling agent (e.g. low molecular weight PEG) reduces the quality of the
solvent (e.g. water) in the PVA solution. This results in the PVA phase separating into
polymer rich regions and solvent rich regions; within the polymer rich regions the PVA
crystallizes without the need of freeze/thaw cycles. More specifically a vinyl polymer
(PVA) is dissolved in a first solvent (water), the solution is heated about the melting
point of the physical associations of the vinyl polymer, after which the vinyl polymer
solution is mixed with a gellant. This resulting solution after the addition of the gellant
has a higher Flory interaction parameter than the vinyl polymer solution, which induces
gelation of the mixture of vinyl polymer solution and gellant [100-103, 112, 113].
Inamura et al. [114-119] studied the PVA-PEG-water system and found that
phase separation and gelation of the system are dependent on PEG molecular weight;
33
where phase separation occurred in mixtures with PEG molecular weights greater than
600 Da but did not with molecular weight of 300 Da. Gelation occurred in all systems
investigated by Inamura, independent of PEG molecular weight.
Bodugoz-Senturk et al. [98-100, 120] has investigated the uses of PVA theta-gels
for osteochondral defect repair in human joints using a low molecular weight PEG (400
g/mol). To increase the creep resistance of these hydrogels, after gelation the networks
are dehydrated using different media (e.g. isopropyl alcohol, 5.2 M aqueous NaCl
solution, 100% PEG400); in addition some formulations are dePEGed via the use of
0.9% saline solution before dehydrated using the different media. After immersion in the
dehydration media for at least 24 hours (room temperature, with agitation), the hydrogels
were then immersed in 0.9% saline solution (room temperature, with agitation) to
rehydrate the hydrogel networks. This dehydrate/rehydration procedure was completed
for various numbers of cycles. This method of dehydration and rehydration of the PVA
theta-gels to decrease the equilibrium water content successfully increased the creep
resistance of the PVA theta-gels. Additional work has been done with PEG of varying
molecular weights including 200 and 600 g/mol, all below the limit to form a supernatant
as determine by Inamura [114-119].
Figure 2.19 shows a schematic of the Bodugoz-Senturk et al. [98] PVA theta-gel:
the first image is the PVA-PEG-water mixture at 90oC in a uniform solution, the second
image is as the solution is cooling and phase separation begins forcing PVA to form
crystalline domains, and the final image is further cooling to RT where the phase
separation forms PVA rich regions and pores containing water and PEG. The PVA
34
regions in the last image are hydrogen bonded to itself forming crystallites in addition to
hydrogen bonded water, which is commonly referred to as bound water.
2.7.1. Chemical crosslinking using radiation
Depending on the chemical structure of polymers, they will crosslink when exposed
to irradiation while other polymers degrade. Figure 2.20 shows graphical representation of
scission and crosslinking polymer chains. Polymers of the formulation shown in Figure 2.21
will degrade when irradiated if there is no hydrogen in the α-position (R1≠H, R2≠H) but the
polymer will crosslink if it contains at least one hydrogen in the α-position [121, 122]. PVA,
PVP, and PEG both have hydrogen present in the α-position allowing for these polymers to
crosslink when exposed to irradiation. The chemical reaction shown in Figure 2.22 is the
radiation crosslinking of a polymer. The crosslinking mechanisms may involve the
production of polymer radicals at neighboring sites on adjacent chains, accompanied by the
loss of molecular hydrogen allowing for bonding between chains.
2.7.2. Chemical crosslinking using difunctional agents
Figure 2.23 shows the ring opening reaction that crosslinks PVA with PEG-DGE.
Being a three membered ring, the epoxide has a large amount of ring strain allowing for it to
be subjected to ring opening reactions, in addition due to changes in electronegativity the
oxygen has a partial negative charge and the two carbon atoms have partial positive charges.
Resulting in the electron deficient carbons being prone to nucleophilic attack thus opening
35
the epoxide ring. These ring opening epoxide reactions can be sped up with acid and base
catalyst. The PEG-DGE/PVA crosslinking reaction is base catalyzed; the basic catalyst opens
the epoxide ring of the PEG-DGE allowing it to react with the pendant hydroxyl group of
the PVA.
2.8. Biocompatibility
PVA is known [91] as a well tolerated biomaterial, which produces no systemic or
local toxic effects. PVA hydrogels have excellent biocompatibility and mechanical
properties [123-128] it has been used for meniscal implant [129, 130], nucleus
replacement [91] and is being investigated for articular cartilage repair [131-133] A study
by Kobayashi et al. [134] showed excellent PVA meniscus implant biocompatibility for a
5 rabbit study, with a 2 year follow up. The study also showed comparable mechanical
performance for the PVA implants to native menisci. Oka et al. [133, 135] conducted a
study using PVA hydrogels for artificial articular cartilage, histological studies of the
articular cartilage and synovial membranes around the implanted PVA hydrogel for 8 to
52 weeks showed neither inflammation nor degenerative changes.
PEG is known for its biocompatibility, it has FDA approval for internal
consumption [136], for use in drug formulations [137], and intramuscular injectables
[138]. The biocompatibility of PVA is based on its nonimmunogenicity, nonantigenicity
and protein rejection properties. PEG is a component in several injectable biodegradable
drug delivery systems where PEG is in a block copolymer with poly(DL-lactide-co-
36
glycolide); these systems are effective carriers for both hydrophilic and hydrophobic drugs
[139-142]. Research using PEG-based materials for scaffolds for the support of cell
growth and function has been conducted. Sims et al. [143, 144] developed a technique
for delivering a mixture of bovine chondrocytes and PEG subcutaneously into rats.
Unfortunately the PEG scaffold exhibited rapid dissolution and poor mechanical
properties [144]. PEG-protein hydrogels have been studied for applications in tissue
engineering, where the PEG provides structural integrity to the scaffold [145]. The most
important and extensive use of PEG is in colloidal polymer systems for biological and
pharmaceutical applications [146-148]
The cytotoxicity of diepoxide compounds, including PEG-DGE, was evaluated
by Nishi et al. [149] due to the work done with diepoxy compounds for the modification
of collagen and gelatin [150-153], modification of tissues [154-157] for bioprosthetic
porcine heart valves, [158] valved conduits,[159] and to sterilize vascular grafts [160]. In
addition to reports that show good biochemical properties for diepoxy compounds when
compared with glutaraldehyde (GA) and dialdehyde starches, which have been used as
crosslinking agents [161-167]. Nishi et al. tested a variety of diepoxide compounds,
including PEG-DGE at three molecular weights: 190, 234 and 542 g/mol. The
cytotoxicity of the specimens was expressed as NR50, which was defined as the specimen
concentration at which the cell activity was reduced to 50% of the control cells without
any test specimen. A higher NR50 equals lower toxicity. The NR50 concentration increased
from 27.4 ± 8.1 µg/mL for the 190 g/mol PEG-DGE, to 46.3 ± 9.0 µg/mL for the 234
g/mol PEG-DGE, and to 133.1 ± 21.4 µg/mL for the 542 g/mol PEG-DGE. The
37
decrease in toxicity with molecular weight is due to a reduction in the number of reactive
end groups, in the case epoxide groups, for a certain mass of PEG-DGE due to the
increasing weight of the increasing polymer backbone length.
2.9. Macromolecular-based solutions as swelling media
For biomedical application of synthetic biomaterials, in particular for soft tissue
applications, it is important to adequately model the in vivo environment the biomaterial
will experience. This is particularly important for soft tissue applications of hydrogel,
which have the ability to swell or deswell, and the implant needs to retain its size and
shape after implantation.
The hydration of tissues present in load-bearing joints is effected by the presence
of external forces and osmotic pressure from charged proteoglycans [168]. Swelling
pressures for the IVD range from 0 to 0.4 MPa, and within the hip and knee are between
0.03 to 0.25 MPa [169, 170]. The majority of in vitro studies in literature use phosphate-
buffered saline (PBS) solution as a swelling media; even though PBS does not replicate
the swelling pressure found in load bearing joints. Rapid equilibration of biomaterials is
achieved when PBS is used as the swelling media, due to the small size of the PBS
molecules. Bryant et al. [171] found that hydrogel mechanical properties and their ability
to function as tissue-engineered scaffolds are influenced by the swelling properties;
making it important that an in vitro model adequately represents the swelling pressure of
the tissues the biomaterial will experience in vivo. Lack of an adequate in vitro model can
38
lead to material failures in animal studies, as in a PVA hydrogel baboon study in which
the hydrogels shrunk when implanted into the IVD and resulted in a high rate of
extrusion in vivo [91].
Macromolecule-based solutions, such as PEG and dextran, can create an osmotic
pressure gradient similar to the in vivo environment better than a PBS solution.
Macromolecule solutions have been used to study the swelling behavior of nucleus
pulposus [47, 87, 168, 172-174] and cartilage [169, 175]; but there are limited studies
that apply this technique to biomaterials. Spiller et al. [132] compared the swelling of
PVA hydrogels in macromolecular-based swelling solution, PBS and an ex vivo model.
The swelling behavior was the same in the macromolecular- based solution and the ex
vivo model, which were significantly different than in PBS. The use of macromolecular-
based swelling models lack fluid flow or mechanical loading which make them imperfect
portrayals of in vivo behavior.
39
Figure 2.1: The spinal column [176]
40
Figure 2.2: Intervertebral Disc [7]
41
Figure 2.3: The organization of the vertebral endplate [7]
42
Table 2.1: Linear region moduli of AF tissue [10]
Circumferential Axial Radial Anterior Inner 5.6 – 10 1.0 N/A Outer 17 – 29 0.8 0.4 – 0.5 Posterior Inner 2 – 6 N/A 0.5 Outer 13 - 19 N/A N/A
43
Figure 2.4: Nonlinear stress/strain curve of collagenous tissues [10]
44
Figure 2.5: Progression of intervertebral disc degeneration [10]
45
Table 2.2: Linear region moduli of nondegenerated (nondeg.) and degenerated (degen.) AF tissue
Circumferential Axial Radial
Nondeg. Degen. Nondeg. Degen. Nondeg. Degen. Anterior Inner 5.6 – 10 5.0 1.0 N/A N/A N/A Outer 17 – 29 22 – 29 0.8 N/A 0.4 – 0.5 0.4 Posterior Inner 2 – 6 4.0 N/A N/A 0.5 N/A Outer 13 - 19 8.0 N/A N/A N/A N/A
46
Figure 2.6: A cylindrical interbody fusion cage (A, anterior; P, posterior) [177]
47
Figure 2.7: CHARITÉ Artificial Disc [75]
48
Figure 2.8: DASCOR device [75]
49
Figure 2.9: PDN-SOLO and HydraFlex devices [75]
50
Figure 2.10: NeuDisc device [75]
51
Figure 2.11: NuCore Injectable Nucleus Device [75]
52
Figure 2.12: Aquarelle Nucleus, available in two sizes 0.1 cm3 (A) and 0.3 cm3 (B) [91]
53
Figure 2.13: BioDisc Nucleus Pulposus Replacement [75]
54
Figure 2.14: Chemical structure of PVA
55
Figure 2.15: Chemical structure for PVP
56
Figure 2.16: Interchain hydrogen bonding within a PVA/PVP blend occurs between
carbonyl groups on PVP and hydroxyl groups on PVA [178]
57
Figure 2.17: Chemical structure for PEG
58
Figure 2.18: Chemical structure for PEG-DGE
59
Figure 2.19: Schematic of PVA theta-gel formation: (a) PVA-PEG water mixture at 90oC is a uniform solution; (b) as the solution is cooled down phase separation begins and forces the PVA to form crystalline domains; (c) with further cooling to near room
temperature, phase separation results in the formation of pores containing water surrounded by PVA rich regions. [98]
60
A
B
Figure 2.20: Radiation scission (A) and crosslinking (B) [122]
61
Figure 2.21: Chemical structure of polymers that degrade or crosslink when exposed to irradiation [121, 122]
62
Figure 2.22: Chemical reaction of radiation crosslinking of a polymer [122]
63
Figure 2.23: Chemical reaction of the crosslinking of poly(vinyl alcohol) with poly(ethylene glycol) diglycidyl ether [179]
64
3. RESEARCH GOALS
This work focuses on investigating the properties of a family of injectable
chemically crosslinked PVA hydrogels for the development of a cohesive nucleus
pulposus replacement implant. The first objective in this work was to synthesize a class of
injectable chemically crosslinked PVA/PEG/PVP hydrogels, with the mechanical and
swelling properties similar to the natural nucleus material. Two methods of crosslinking
were investigated: radiation and difunctional crosslinking. For the radiation crosslinked
gels, radiation dosage and PVA content were varied; for the difunctional crosslinked gels
basic catalyst volume and reaction time were varied at a constant PVA and PEG
concentration.
Once a crosslinking method was determined, the hydrogels were characterized
while varying PVA content, PEG-DGE content, PEG functional group and PEG-DGE
molecular weight. The swelling mechanics and mechanical properties were examined, in
addition to spectroscopy, crystallinity and optical microscopy of each formulation to
investigate the structure-property relationships present in this family of hydrogels.
Crosslinking did not resolve the mass retention issue of the PVA theta-gels, to
resolve this issue a purification technique was developed and optimized in order to
continue investigating this material family for nucleus replacement. The swelling
mechanics and mechanical properties of these purified gels were examined varying PVA
content, PEG-DGE content, PEG functional group and PEG-DGE molecular weight;
in addition to spectroscopy, crystallinity and optical microscopy of each formulation to
65
investigate the structure-property relationships present in this family of hydrogels after
purification. Characterization of the purified hydrogel family includes cytotoxicity of the
hydrogels in addition to concentrations of PEG-DGE. The last objective in this project
was to evaluate the effect of swelling on mechanical properties and crystallinity of the
hydrogel in long-term in vitro studies. The specific aims of this work are as follows:
Specific Aim 1: Synthesize an injectable, chemically crosslinked PVA/PEG hydrogel
system
Specific Aim 2: Characterize the hydrogel and investigate of the structure-property
relationships present in this family of hydrogels
Specific Aim 3: Development and optimization of a purification method for the
PVA/PEG hydrogel system, characterization of the purified hydrogel, and investigation
of the structure-property relationships present in this family of purified hydrogels
Specific Aim 4: Characterize the hydrogel in a simulated intradiscal environment
66
4. SYNTHESIS OF A CROSSLINKED PVA/PEG HYDROGEL SYSTEM
4.1. Introduction
Developing a chemically crosslinked hydrogel network, that is injectable after
crosslinking, and will gel within minutes of injection into the nuclear cavity forming a
cohesive implant with similar properties to the native nucleus tissue, meant having to veer
off the path of previously investigated hydrogels. Freeze/thawed PVA-PVP copolymers
which were formed into a string and freeze/thawed before injection or the in situ forming
poly(N-isopropylacrylamide) (PNIPAAm)-based materials could not match the
properties of the nucleus tissue.
Chemical crosslinking of PVA hydrogels has been previously done with radiation
(e.g. gamma radiation and ebeam radiation) [95, 97, 180-186] and difunctional
crosslinking agents (e.g. glutaraldehyde, diepoxides) [187-193]. Chemically crosslinking
of PVA generally eliminates the ability to inject the material: though it can be molded
into a string similarly to PVA/PVP physical network gels but this does not create a
cohesive implant. Physical PVA hydrogels have been produced using the theta-gel
method [98-103, 112, 113, 120] using PEG as the gellant; this method produces an
injectable hydrogel that forms a cohesive implant. The biggest limitations for PVA theta-
gels are high swelling ratios and low mass retention in biological simulated environments.
This work will combine the ideas of using PEG to create theta-gel PVA hydrogels and
chemically crosslinking these hydrogels by either radiation or difunctional crosslinkers to
overcome the swelling and mass retention issues experienced with current PVA theta-gels.
67
In this work, two families of chemically crosslinked hydrogels were synthesized:
one using radiation and one using a difuctional crosslinker. The radiation crosslinked
hydrogels were irradiated using electron beam radiation; PVA content and electron beam
radiation dosage was varied to determine the effects of each variable. For the difunctional
crosslinked gels, PEG-DGE was used to crosslink PVA; basic catalyst volume and
reaction time was varied to determine the effect of each variable on the hydrogel
properties. In the following experiments, the structure-property relationship in these
chemically crosslinked hydrogels was studied. Specifically, swelling mechanics, stiffness,
relative crystallinity and microstructure were investigated as a function of PVA content,
radiation dosage, catalyst volume and reaction time. It is hypothesized that a material
candidate from this family of chemically crosslinked PVA/PEG hydrogels can serve as a
synthetic nucleus pulposus replacement.
4.2. Materials and Methods
4.2.1. Materials
PVA (99.0-99.8% hydrolyzed, molecular weight 145 kDa) and PEG-DGE
(molecular weight 526 Da) were purchased from Aldrich. PVP (molecular weight 58
kDa) was purchased from ISP Technologies. PEG (molecular weight 4.6 kDa) was
obtained from Aldrich and (molecular weight 20 kDa) was obtained from Crescent
Chemical Company. Barium sulfate (BaSO4) with a 1-10 μm particle size was purchased
from J.T. Baker. Sodium hydroxide (reagent grade, ≥ 98%) and sodium chloride (reagent
68
grade, ≥ 99%) were obtained from Sigma-Aldrich. Dialysis tubing, purchased from Fisher
had a nominal molecular weight cut-off of 3,500 g/mol.
4.2.2. Hydrogel Synthesis
4.2.2.1. Radiation crosslinked
PVA/PVP/PEG hydrogel compositions (Table 4.1) were made by preparing an
aqueous PVA/PVP solution (ranging from 9.6 to 14.9% w/w) by mixing PVA, PVP, and
deionized water in a sealed glass bottle and heating to 121°C for 30 minutes in an
autoclave. The ratio of PVA to PVP was 99:1. After the autoclave cycle, solutions were
removed from the autoclave and equilibrated to 75 ± 5°C in a water bath. BaSO4 (7.0%
w/w) was then mixed into the PVA/PVP solution and the mixture was autoclaved again
at 121oC for 30 minutes. Previous work has shown that a 4 to 15% concentration of
BaSO4 in the resultant hydrogel composition was sufficient to make the hydrogels
radiopaque. The addition of barium sulfate was an optional step.
After the second autoclave cycle, solutions were again removed from the autoclave
and equilibrated to 75°C in a water bath, the solution was maintained at 75 ± 5°C during
the addition of PEG (MW=4.6 kDa) by manual stirring. After the addition of PEG, the
mixtures were left to equilibrate at RT for 3 ± 0.25 hours. During the equilibration time,
the solution will separate into a polymer-rich gel and a solvent-rich liquid phase, at this
point the liquid phase is decanted leaving just the polymer-rich gel. The gel is then
autoclaved for a third time at 121oC for 30 minutes, after which the gel is again separated
69
from the additional solvent-rich liquid phase that has formed; the gel is then loaded into
a 60 cc syringe and injected into 15 mL centrifuge tubes. The resulting tubes of hydrogel
were then irradiated at RT with a 10 MeV electron beam to the desired dosages of 15
and 20 kGy at Sterigenics in Salem, North Carolina or reserved as 0 kGy controls.
Following irradiation, the hydrogel is autoclaved again at 121oC for 30 minutes, after
which the gel is loaded into a 60 cc syringe and molded into a 15-mL centrifuge tube or a
test specific mold.
4.2.2.2. Difunctional crosslinked
PVA/PVP/PEG-DGE hydrogel compositions (Table 4.2) were made by
preparing an aqueous PVA/PVP solution (14.4% w/w) by mixing PVA, PVP, and
deionized water in a sealed glass bottle and heating to 121°C for 30 minutes in an
autoclave. BaSO4 was present in hydrogels used for swelling and mechanical testing but it
was not used for other testing due to issues with overpowering the signal of the polymers
in the system. When used, BaSO4 is added before the first autoclave cycle at 7.0 wt %.
After the autoclave cycle, solutions were removed from the autoclave and
equilibrated to 75 ± 5°C in a water bath. PEG-DGE (29.0% of the total solution mass) is
then stirred into the solution with 100 µL of 10 M sodium hydroxide (NaOH) (per 75
gram batch) to form a gel. NaOH, the basic catalyst, was added to create a basic
condition for the ring opening reaction of the PEG-DGE to enable it to crosslink to
PVA. The solution is left to react for 24 hours (if another time is not specified), the
70
supernatant is decanted and the gel is autoclaved again at 121oC for 30 minutes. After
this last autoclave cycle the material is loaded into a 60 cc syringe and molded into a 15-
mL centrifuge tube or a test specific mold.
4.2.3. Swelling Mechanics
Osmotic solutions were made by dissolving PEG (20 kDa) in 0.15 M sodium
chloride to achieve osmotic pressures mimicking the swelling pressure of the IVD.
Equation 4.1 was used to calculate the PEG concentration, c2, required for an osmotic
pressure, Π:
Π = !" !!!!+ !!!! + !!!! +⋯ (4.1)
where R is the universal gas constant, T is the absolute temperature and M2 is the
polymer molecular weight. The second and third virial coefficients, B and C, for 20 kDa
PEG are 2.59 x 10-3 and 13.5 x 10-3, respectively [194]. The osmotic pressure used for
this study is the midpoint of the range (0 to 0.4 MPa) measured by Urban et al. for
cadaver IVDs [168]. To obtain the osmotic pressure of 0.2 MPa, a PEG concentration of
128.2 g/mL is used.
After the final autoclave cycle of the hydrogel synthesis procedure, the hydrogel
was loaded into a 60 cc syringe and injected into poly(vinyl chloride) (PVC) tubing with
71
an inner diameter of 9.5 mm. Cylindrical samples, approximately 0.5 cm3 in volume, were
sliced from the PVC tubing, the tubing was removed and the sample was weighed in air
and heptane to determine the initial density of the hydrogel using Equation 4.2:
!hydrogel = !heptane × mair
mair -‐ mheptane (4.2)
where ρhydrogel is the density of the hydrogel, ρhep is the density of heptane, mair is the mass
of the hydrogel in air, and mheptane is the mass of the hydrogel in heptane [84]. Using the
density and initial mass of each sample the initial volume is calculated by dividing the
mass by the density. The samples were then placed in dialysis tubing; dialysis tubing is
used to prevent uptake of PEG (20 kDa) by the hydrogels. Hydrogels in the dialysis
tubing were placed in the 0.2 MPa PEG solutions for one week at 37oC. The volume of
swelling medium was 100x larger than the volume of hydrogel samples to prevent
significant changes in the pressure of the osmotic solution due to changes in the hydrogel
water content over the length of the study.
Samples were removed from the PEG solution and the dialysis tubing at each
time point (0, 1, 4, 7 days), after which each sample was weighed in air and heptane to
determine the swelling ratio (V/Vo) by comparing the volume of swollen samples to the
volume of the initial samples (Equation 4.3). After swelling, hydrogel samples were dried
in an oven at 50oC. The mass of the dried hydrogel samples was compared to the initial
72
mass to calculate the initial water content (Equation 4.4) and the dry mass compared to
the swollen mass to calculate the equilibrium water content (Equation 4.5). The mass
retention value is calculated using Equation 4.6, where the dry mass of the swollen
Swelling Ratio = swollen volumeinitial volume
(4.3)
Initial Water Content = 1 – dry massinitial mass
× 100% (4.4)
Equilibrium Water Content = 1 – dry massswollen mass
× 100% (4.5)
Mass Retention = dry mass of the swollen gel
initial mass of the swollen gel × dry mass of the unswollen gelinital mass of the unswollen gel
× 100% (4.6)
gel is divided by the product of the initial mass of the swollen gel and the ratio of the dry
mass to the initial mass of the unswollen gel. A sample swelling ratio plot is shown in
Figure 4.1.
4.2.4. Mechanical Properties
Unconstrained, uniaxial compressive modulus was measured to determine if the
hydrogel is a suitable nucleus pulposus replacement material. Testing was conducted
73
using an Instron Materials Testing System Series 4442 (Norwood, MA) bench-top
mechanical testing system with a 50 N load cell.
After the hydrogel was allowed to react for the specified amount of time, the
hydrogel was autoclaved at 121oC for 30 minutes. After the autoclave cycle the hydrogel
was loaded into a 60 cc syringe and injected into poly(vinyl chloride) tubing with an inner
diameter of 9.5 mm. The hydrogel was then removed from the tubing in sections 20 cm
in length, placed in dialysis tubing and swollen for up to four weeks at 37oC. Samples
were swollen in osmotic PEG solutions with an osmotic pressure of 0.2 MPa. After two
weeks of swelling, cylindrical hydrogel samples were sliced (n = 5), to a thickness of 7-8
mm and cut to ensure flat surfaces. The samples were preloaded to 0.1 N and compressed
to a total strain of 30% at a strain rate of 100% min-1, a chord from the initial linear
portion of the stress versus strain curve (10-20% strain) was used to calculate elastic
modulus. At least five independent samples were tested for each set of hydrogels (n = 5).
A samples stress versus strain plot is shown in Figure 4.2.
4.2.5. FTIR
FTIR spectra were collected using a Thermo Nicolet 6700 with DLaTGS
detector, Ge on KBr beamsplitter, and single reflection diamond Attenuated Total
Reflectance (ATR) sample accessory. Spectra were obtained with 64 scans analyzed with
Omnic 8.1.11 software (ThermoFischer Scientific).
74
After the hydrogel was allowed to react for the specified amount of time, the
hydrogel was autoclaved at 121oC for 30 minutes. Following the autoclave cycle the
hydrogel was loaded into a 60 cc syringe and injected onto polyethylene terephthalate
copolymer with cyclohexylene dimethylene segments (PETG) sheeting with spacers of
0.1 mm thickness to control thickness of the film. Samples larger than the diameter of
the ATR crystal were cut from the hydrogel film and then tested (n=3).
4.2.6. Statistical Analysis
All data points are represented as the mean ± one standard deviation for at least
three independent samples. Statistical significance was determined by one-way analysis of
variance (ANOVA) with post-hoc analysis by Bonferroni correction with a 95%
confidence interval. P-values less than 0.05 were considered statistically significant.
4.3. Results and Discussion
4.3.1. Swelling Mechanics
4.3.1.1. Radiation crosslinked
The swelling ratio for the electron beam irradiated hydrogels is shown in Figure
4.3. The 9. 5% PVA hydrogel formulation formed a gel at a dosage of 20 kGy, a gel was
not formed at 0 or 15 kGy. Both the 12.1 and 14.8% PVA hydrogel formulations formed
gels at dosages of 0, 15, and 20 kGy. There is no difference in swelling ratio for the
12.1% PVA formulation at each irradiation dosage (p>0.05), nor for the 14.8% PVA
75
formulation (p>0.05). At 0 kGy dosage, the swelling ratio increases from 12.1 to 14.8%
PVA (p<0.001), due to increased polymer concentration. There is no significant
difference in swelling ratio at 15 kGy between the 12.1 and 14.8% PVA formulations. At
20 kGy the swelling ratio increases from 9.5 to 12.1% PVA (p<0.001), and from 12.1 to
14.8% PVA (p>0.01).
The initial water content and mass retention values for the electron beam
irradiated samples are shown in Table 4.3. For the initial water content of the 12.1%
PVA formulation, there is no significant difference (p>0.05) between 0 and 15 kGy or 15
and 20 kGy, but there is an increase in water content from 0 to 20 kGy (p<0.05). Mass
retention for the 12.6% PVA formulation increases from 0 to 20 kGy (p<0.05), but there
is no significant change between 0 and 15 kGy or 15 and 20 kGy. For the 14.8% PVA
formulation, there is no significant difference (p>0.05) in initial water or mass retention
values between the three irradiation dosages tested: 0, 15 and 20 kGy.
4.3.1.2. Difunctional crosslinked
The swelling ratio of hydrogel formulations with increasing basic catalyst volume
from 0 to 200 µL is shown in Figure 4.4 and the initial water content and mass retention
values are shown in Table 4.4. As catalyst volume is increased for 0 to 200 µL there is an
increase in swelling ratio from 0.96 ± 0.01 to 1.05 ± 0.02 (p<0.01).
The swelling ratio of hydrogel formulations varying reaction time are shown in
Figure 4.5; Table 4.5 shows the initial water content and mass retention of the varying
76
reaction time hydrogels. As reaction time is increased from 1 hour to 72 hours, the
swelling ratio decreases from 1.17 ± 0.01 to 1.01 ± 0.02 (p<0.01). There is no significant
difference in initial water content values as reaction time varies (p>0.05).
4.3.2. Mechanical Properties
4.3.2.1. Radiation crosslinked
The compressive moduli for the radiation crosslinked hydrogels are shown in
Figure 4.6. For the 12.1 and 14.8% PVA formulations there is no change in stiffness
between the 0, 15 and 20 kGy irradiation dosages (p>0.05). The 9.5% PVA hydrogel
formulation only formed a gel at 20 kGy irradiation dosage. There is also no difference
between the compressive moduli of the 12.1 and 14.8% PVA formulations at any dosage.
The only difference measured in electron beam irradiated gel moduli is between the 9.5
and 14.8% PVA formulations at 20 kGy (p<0.01).
4.3.2.2. Difuntional crosslinked
The compressive moduli for the varying catalyst volume hydrogels is shown in
Figure 4.7 and for varying reaction time hydrogels is shown in Figure 4.8. As catalyst
volume increases, from 0 to 200 µL, compressive modulus increases at two weeks
(p<0.001). For the reaction time formulations moduli increases from 1 hour to 24 hours
(p<0.001), from 24 hours to 48 hours (p<0.001) but does not increase from 48 hours to
72 hours (p>0.05).
77
4.3.3. FTIR
The FTIR spectra of PVA is shown in Figure 4.9 and the FTIR spectra of PEG
and PEG-DGE are shown in Figure 4.10. The spectra of PVA highlights the OH
stretch at 3280 cm-1, the CH2 out of plane stretch at 2937 cm-1, the CH stretch at 2906
cm-1 and the C-O stretch at 1023, 1087 and 1142 cm-1. The important peaks of the PEG
spectra are the CH2 stretch at 2880 cm-1, the C-O-C out of phase stretch at 1059, 1093,
and 1145 cm-1 and the C-O-C in phase stretch at 841 cm-1. The only major difference
between the PEG and PEG-DGE spectrums are the peaks for the epoxide group at 760,
844 and 913 cm-1[195].
4.3.3.1. Radiation crosslinked
Figure 4.11 shows the FTIR spectrums for 14.8% PVA at 0, 15 and 20 kGy
irradiation dosages. As irradiation increases, from 0 to 20 kGy, the PEG -CH2-
symmetric stretch (2851 cm-1) and the PVA -CH2- symmetric stretch (2922 cm-1)
increase indicating scission of the polymer chains [195, 196].
4.3.3.2. Difunctional crosslinked
Figure 4.12 shows the FTIR spectrum for basic catalyst volume (A) and reaction
time hydrogel formulations (B). As catalyst volume increases from 0 to 200 µL there is a
decrease in the peaks where the CH stretch of the PVA (2906 cm-1) and the CH2 stretch
78
of the PEG-DGE (2880 cm-1) overlap and where the C-O stretch of the PVA (1023,
1087, 1142 cm-1) and the C-O-C out of phase stretch of the PEG (1059, 1093 and 1145
cm-1) overlap.
There is no difference in spectra for the 1-hour reaction time sample and the 48-
hour reaction time sample. With additional reaction time, there might be an increase in
crosslinking but due to the large number of ethers already present in the system the
relatively small number of new ethers being formed is not visible via FTIR.
4.3.4. Discussion
Hydrogels mimic the behavior of the nucleus pulposus, in particular in their
ability to swell and release water through out the course of the day similar to that of the
natural nucleus material. Proteoglycan and hydration level in the nucleus material are at a
concentration to produce an osmotic pressure between 0.05 MPa and 0.3 MPa [168].
This high pressure arises from the loading conditions within the disc and the spine.
When the material, natural nucleus or hydrogel is at equilibrium there is no net fluid loss
or gain. It has been reported that increases in crosslinking, decrease chain mobility, and
the ability of a matrix to swell [95, 104, 197, 198]. As polymer concentration of the
hydrogel increases, it is expected for the network to experience additional swelling due to
additional polymer chains within the matrix [130, 182, 198, 199]. For the radiation
crosslinked hydrogels we see changes in swelling ratio from a combination of increasing
polymer concentration between formulations, the effects of scission, and the effects of
79
crosslinking. In conjunction with changes in the swelling ratio, we also expect an increase
in mass retention and increase in water content was we increase crosslinking. For the
radiation crosslinked 12.1% PVA formulation, there is an increase in initial water content
and mass retention due to increased crosslinking with increased irradiation dosage. There
is no difference for the radiation crosslinked 14.8% PVA formulations in initial water
content or mass retention between the three dosages. This is due to the increase in
crosslinking and scission that is occurring in this formulation due to the higher polymer
concentration. For the difunctional crosslinked hydrogels, we expect to see a decrease in
swelling and an increase in mass retention with increases in crosslinking [200]. This is
seen in the 1 to 72 hour reaction time gels, as there is additional crosslinking in the
network with additional reaction time.
Normal nucleus tissue properties vary with state of degeneration and have been
described in some classes as fluid [6] or in others as an isotropic solid [36, 84]. Due to the
variation in tissue, it is difficult to match the mechanical properties of the tissue with the
hydrogel material. Cadaver testing and finite element modeling have shown that a
polymeric hydrogel implant should have a compressive modulus of at least 50 kPa at 15%
strain to restore healthy tension in annulus fibers [84, 85]. With increasing polymer
content in the PVA hydrogels, we expect an increase in compressive modulus [130].
There has been significant research in literature suggesting the formation of physical
crosslinks, during the freeze/thaw process, allow for PVA hydrogels to carry increased
load [104, 197, 201]. These physical crosslinks are formed between the PVA chains, as
hydrogen bonds and crystallites, and are also formed during the theta-gel method [98-
80
103, 113] used to create the hydrogels for this thesis work. In addition to the physical
network in the PVA hydrogel, the addition of chemical crosslinks from radiation and
difunctional crosslinkers increases the mechanical properties of the hydrogel network
[185, 187, 202]. The compressive modulus for each of the radiation crosslinked hydrogels
is below 50 kPa [84, 85] and therefore these formulations can not be considered for
potential candidate materials for nucleus replacement. For the difunctional crosslinked
hydrogels, the compressive modulus increases as catalyst volume increases from 0 to 200
µL due to increased crosslinking within the hydrogel network. The gels with varying
reaction time, from 1 to 72 hour, increase in modulus with increasing reaction time; this
is due to increased crosslinking in the hydrogel network. All varying catalyst volume
hydrogels and reaction time hydrogels have moduli values above 50 kPa at all time points.
The lack of change in the swelling and mechanical properties for the radiation
crosslink gels with increased dosage suggest that in addition to crosslinking of the
network the polymers within the hydrogels are experiencing scission. FTIR shows that as
irradiation increases, from 0 to 20 kGy, the PEG -CH2- symmetric stretch (2851 cm-1)
and the PVA -CH2- symmetric stretch (2922 cm-1) increase indicating scission of the
polymer chains [195, 196]. For the difunctional crosslinked hydrogels, an increase in the
C-O-C out of phase stretch of the PEG (1059, 1093 and 1145 cm-1) [195] is expected
with increased crosslinking. Due to the large amount of ethers already present in the
system from the PEG and PEG-DGE, the relatively small number of new ethers being
formed is not visible via FTIR. This results in there being no difference in the spectra for
the 1-hour reaction time sample and the 48-hour reaction time sample. For the catalyst
81
volume hydrogels, as catalyst volume increases from 0 to 200 µL there is a decrease in the
peaks where the CH stretch of the PVA (2906 cm-1) and the CH2 stretch of the PEG-
DGE (2880 cm-1) overlap and where the C-O stretch of the PVA (1023, 1087, 1142 cm-
1) and the C-O-C out of phase stretch of the PEG (1059, 1093 and 1145 cm-1) overlap
[195]. This decrease is attributed to a decrease in PEG-DGE in the resultant hydrogel;
the addition of the catalyst opens the epoxide ring of the PEG-DGE for reaction with
the PVA but it also allows for hydrolysis of the PEG-DGE. This end group change
increases the solubility of the PEG, allowing for it to pull additional water out of the
system into the supernatant that is formed with the addition of PEG or PEG-DGE to a
PVA solution as the PVA solution gels. This conclusion is supported by the decrease in
water content from the 0 µL formulation (51.3 ± 0.1%) to the 200 µL formulation (48.3
± 0.4%), (p<0.05).
4.4. Conclusions
Due to increased scission in the radiation crosslinked hydrogels the compressive
moduli were below 50 kPa, the minimum modulus value necessary to restore healthy
tension in the annulus fibers [84, 85]. Hydrogel samples were irradiated with dosages
ranging from 0 to 100 kGy, but only the formulations from 0 to 20 kGy were flowable
after irradiation (Figure 4.13). The low modulus values measured removed electron beam
irradiation as a potential chemically crosslinking method for the PVA/PEG hydrogels for
injectable nucleus replacement.
82
All of the di-functional crosslinked PVA/PVP/PEG-DGE hydrogels had moduli
values above the necessary 50 kPa, in addition to having low swelling ratios at 0.2 MPa
osmotic pressure. These properties suggest that a family of injectable chemically
crosslinked hydrogels for nucleus replacement was developed. The material properties of
this family of hydrogels will be examined in subsequent chapters.
Similar to the electron beam gels with dosages over 20 kGy shown in Figure 4.13,
catalyst volumes greater than 200 µL resulted in hydrogels that could not be flowed again
after the 24-hour reaction period. This is due to chemical crosslinking and the decreased
water content in the hydrogel as additional catalyst is added. Further testing will use 100
µL (for a 75 g batch) and a 24 hour reaction time; these conditions were chosen to give
the best properties without jeopardizing the injectability of the hydrogels in addition to
reducing the production time for the gels.
83
Table 4.1: Sample compositions for radiation crosslinked formulations
Sample Composition (% w/w)
Components 9.5% PVA 12.1% PVA 14.8% PVA PVA 145 kDa 9.5 12.1 14.8
PVP 58 kDa 0.1 0.1 0.1
Deionized H2O 75.0 70.0 65.0 BaSO4 1-10 µm 7.0 7.0 7.0
PEG 4.6 kDa 8.4 10.7 13.1
84
Table 4.2: Sample composition for difunctional crosslinked formulation
Sample Composition (% w/w) Components 29% PEG-DGE PVA 145 kDa 14.27
PVP 58 kDa 0.13
Deionized H2O 49.60
BaSO4 1-10 µm 7.00
PEG-DGE 526 Da 29.00
85
Figure 4.1: Swelling ratio for 24 hour hydrogel formulation
Swelling Ratio for 24 HourReaction Time Hydrogel
0.0 2.5 5.01.0
1.1
1.2
1.3
1.4
Time, days
86
Figure 4.2: Stress versus strain plot for the 24 hour reaction time hydrogel
24 Hour Reaction Time HydrogelStress vs. Strain
0.0 0.1 0.2 0.3 0.40.00
0.05
0.10
0.15
0.20
Strain
87
0
kGy 9.5 12.1 14.8 15 kGy 9.5 12.1 14.8 20
kGy 9.5 12.1 14.8
9.5 9.5 9.5 12.1 12.1 12.1 *** 14.8 *** 14.8 ns 14.8 *** **
12.1% PVA 0 15 20 14.8%
PVA 0 15 20
0 0 15 ns 15 ns 20 ns ns 20 ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 4.3: Swelling ratio of electron beam hydrogel formulations
Swelling Ratio of Electron BeamRadiation Crosslinked Formulations
0 15 200.0
0.5
1.0
1.59.5% PVA12.1% PVA14.8% PVA
Radiation Dosage, kGy
88
Table 4.3: Initial water content and mass retention of electron beam hydrogel formulations
Radiation Dose Gel Initial Water Content Mass Retention
[%] [%]
9.5% PVA
0 kGy
15 kGy
20 kGy 70.1 ± 4.3 99.3 ± 2.7
12.1% PVA
0 kGy 62.7 ± 0.4 97.5 ± 0.3
15 kGy 63.7 ± 1.1 99.0 ±0.8
20 kGy 64.0 ± 0.2 100.4 ± 1.4
14.8% PVA
0 kGy 56.5 ± 0.5 98.9 ± 2.6
15 kGy 58.8 ± 1.3 100.2 ± 1.5
20 kGy 60.4 ± 1.5 101.0 ± 1.0
Initial Water Content 12.1% PVA 0 15 20 14.8% PVA 0 15 20
0 0 15 ns 15 ns 20 * ns 20 ns ns
Mass Retention 12.1% PVA 0 15 20 14.8% PVA 0 15 20
0 0 15 ns 15 ns 20 * ns 20 ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
89
1 week 0 33 66 100 200
0 33 ns 66 ns **
100 ns ns * 200 ** *** ns ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 4.4: Swelling ratio of hydrogel formulations varying basic catalyst volume
Swelling Ratio Varying Catalyst Volume
0.00.10.20.30.40.50.60.70.80.91.01.1
Catalyst volume, uL
90
Table 4.4: Initial water content and mass retention of hydrogel formulations varying basic catalyst volume
Gel Initial Water Content Mass Retention
[%] [%]
0 µL 51.3 ± 0.1 82.6 ± 5.5
33 µL 49.5 ± 0.6 79.68 ± 3.0
66 µL 51.4 ± 1.7 88.5 ± 1.4
100 µL 49.8 ± 0.4 77.4 ± 0.9
200 µL 48.3 ± 0.4 83.5 ± 3.3
Initial Water Content 0 33 66 100 200
0 33 ns 66 ns ns
100 ns ns ns 200 * ns * ns
Mass Retention 0 33 66 100 200
0 33 ns 66 ns ns
100 ns ns ns 200 ns ns ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
91
1 week 1 24 48 72 1
24 ns 48 ns ns 72 ** ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 4.5: Swelling ratio of hydrogel formulations varying reaction time
Swelling Ratio Varying Reaction Time
1 h 24 h 48 h 72 h0.00.10.20.30.40.50.60.70.80.91.01.11.2
Reaction Time, hours
92
Table 4.5: Initial water content and mass retention of hydrogel formulations varying reaction time
Gel Initial Water Content Mass Retention
[%] [%]
1 hour 49.3 ± 0.1 71.3 ± 0.9
24 hour 49.0 ± 0.3 87.4 ± 1.0
48 hour 48.6 ± 0.7 80.5 ± 4.2
72 hour 48.4 ± 0.8 86.6 ± 3.5
Initial Water Content 1 24 48 72
1 24 ns 48 ns ns 72 ns ns ns
Mass Retention 1 24 48 72
1 24 *** 48 * ns 72 *** ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
93
0
kGy 9.5 12.1 14.8 15 kGy 9.5 12.1 14.8 20
kGy 9.5 12.1 14.8
9.5 9.5 9.5 12.1 12.1 12.1 ns 14.8 ns 14.8 ns 14.8 ** ns
12.1% PVA 0 15 20 14.8%
PVA 0 15 20
0 0 15 ns 15 ns 20 ns ns 20 ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 4.6: Compressive moduli of electron beam hydrogel formulations
Compressive Moduli of ElectronBeam Formulations (at 2 weeks)
0 15 200
10
20
309.5% PVA12.1% PVA14.8% PVA
Radiation Dosage, kGy
94
2 weeks 0 33 66 100 200
0 33 *** 66 *** ***
100 *** *** *** 200 *** *** *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 4.7: Compressive moduli of hydrogel formulations varying basic catalyst volume
Compressive Modulus VaryingCatalyst Volume (at 2 weeks)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
Catalyst volume, uL
95
2 weeks 1 24 48 72
1 24 *** 48 *** *** 72 *** *** ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 4.8: Compressive moduli of hydrogel formulations varying reaction time
Compressive Modulus VaryingReaction Time (at 2 weeks)
1 h 24 h 48 h 72 h0.00
0.25
0.50
0.75
Reaction Time, hours
96
Figure 4.9: FTIR spectra of PVA
97
Figure 4.10: FTIR spectra of PEG and PEG-DGE
98
Figure 4.11: FTIR of electron beam 14.8% PVA hydrogel formulation; with increased irradiation, from 0 to 20 kGy, the PEG-CH2- symmetric stretch (2851 cm-1) and the
PVA-CH2- symmetric stretch (2922 cm-1) increases indicating scission of the polymer chains.
99
Figure 4.12: FTIR of difunctional crosslinked chemically crosslinked hydrogels: (A)
varied basic catalyst volume, (B) varied reaction time
100
Figure 4.13: Electron beam crosslinked hydrogel formulations with radiation dosages
ranging from 0 to 100 kGy (right to left).
101
5. CHARACTERIZATION OF THE PVA/PVP/PEG-DGE HYDROGEL
SYSTEM
5.1. Introduction
There are only two groups publishing work on PVA theta gels formed using PEG,
the difference between the two groups is the molecular weight of PEG that is used [98-
103, 112, 113, 120, 203]. The majority of physical network PVA hydrogel work is done
with freeze/thaw hydrogels [84, 95, 131, 132, 178, 198, 204-208]; PVA theta gels have
increased moduli over the PVA freeze/thaw hydrogels but can have issues with mass
retention in biological simulated environments. The work by Ruberti et al. [100-103] and
Bodugoz-Senturk et al.[98, 99, 120] uses PEG with molecular weights below 600 Da
where the work by Kita et al. [112, 113] uses molecular weights above 600 Da. The PVA
hydrogels formed with PEG below 600 Da do not produce a supernatant, resulting in
PVA rich domains and water/PEG rich domains within the gel network. When a
supernatant is formed the hydrogel network contains PVA rich domains, a water/PEG
rich supernatant and some water/PEG rich domains within the gel whose size varies on
the molecular weight used to create the gel.
This work focuses on the synthesis of hydrogels using PVA and PEG-DGE, the
PVA molecular weight is held constant and the PEG-DGE molecular weight ranges
from 526 to 4600 Da. The majority of the experiments being done with hydrogels made
with 526 Da as it was the only commercially available PEG-DGE at the time this work
was done. It is proposed that crosslinking the PVA formed via the theta gel method can
102
increase the mechanical properties and reduce the swelling issues experienced with some
PVA theta gel formulations while still being able to be injected into the nuclear cavity as
a cohesive implant. In the previous chapter, it was shown that a family of injectable
chemically crosslinked hydrogels was created using diepoxide PEG as a difunctional
crosslinking agent.
In the following experiments, the structure-property relationship in these
chemically crosslinked hydrogels was studied. Specifically, swelling mechanics, stiffness,
relative crystallinity, and microstructure were investigated as a function of PEG-DGE
content, PVA content, PEG functional group and PEG-DGE molecular weight. It is
hypothesized that a material candidate from this family of chemically crosslinked
PVA/PEG-DGE hydrogels can serve as a synthetic nucleus pulposus replacement.
5.2. Materials and Methods
5.2.1. Materials
PVA (99.0-99.8% hydrolyzed, molecular weight 145 kDa) was purchased from
Aldrich. PVP (molecular weight 58 kDa) was purchased from ISP Technologies. PEG
(molecular weights 600 Da) was obtained from Aldrich and (molecular weight 20 kDa)
was obtained from Crescent Chemical Company. PEG-DGE (molecular weights 526
Da, 2 kDa) was obtained from Aldrich, PEG-DGE (molecular weight 4.6 kDa) was
custom synthesized by Advanced Polymer Materials Inc. Barium sulfate (BaSO4) with a
1-10 μm particle size was purchased from J.T. Baker. Sodium hydroxide (reagent grade, ≥
103
98%) and sodium chloride (reagent grade, ≥ 99%) were obtained from Sigma-Aldrich.
Dialysis tubing, purchased from Fisher, had a nominal molecular weight cut-off of 3,500
g/mol.
5.2.2. Hydrogel Synthesis
PVA/PVP/PEG-DGE hydrogel compositions (Table 5.1) were made by
preparing an aqueous PVA/PVP solution (ranging from 12.7 to 20.2% w/w) by mixing
PVA, PVP, and deionized water in a sealed glass bottle and heating to 121°C for 30
minutes in an autoclave. BaSO4 is used in the hydrogel for radiopacity but BaSO4 was
present in hydrogels used for swelling and mechanical testing but it was not used for
other testing due to issues with overpowering the signal of the polymers in the system.
When used, BaSO4 is added before the first autoclave cycle at 7.0 wt%.
After the autoclave cycle, solutions were removed from the autoclave and
equilibrated to 75 ± 5°C in a water bath. PEG-DGE (ranging from 9.0 to 39.0% of the
total solution mass) is then stirred into the solution with 100 µL of 10 M sodium
hydroxide (per 75 gram batch) to form a gel. NaOH was added to create a basic
condition for the ring opening reaction of the PEG-DGE to enable it to crosslink to
PVA. The solution is left to react for 24 hours, the supernatant is decanted and the gel is
autoclaved again at 121oC for 30 minutes. After this last autoclave cycle the material is
loaded into a 60 cc syringe and molded into a 15-mL centrifuge tube or a test specific
mold.
104
5.2.3. Swelling Mechanics
Osmotic solutions were made by dissolving PEG (20 kDa) in 0.15 M sodium
chloride to achieve osmotic pressures mimicking the swelling pressure of the IVD.
Equation 5.1 was used to calculate the PEG concentration, c2, required to for an osmotic
pressure, Π:
Π = !" !!!!+ !!!! + !!!! +⋯ (5.1)
where R is the universal gas constant, T is the absolute temperature and M2 is the
polymer molecular weight. The second and third virial coefficients, B and C, for 20 kDa
PEG are 2.59 x 10-3 and 13.5 x 10-3, respectively. The osmotic pressures used for this
study is the midpoint of the range (0 to 0.4 MPa) measured by Urban et al. for cadaver
IVDs [168]. To obtain the osmotic pressure of 0.2 MPa, a PEG concentration of 128.2
g/mL is used.
After the autoclave cycle of the hydrogel synthesis procedure, the hydrogel was
loaded into a 60 cc syringe and injected into poly(vinyl chloride) (PVC) tubing with an
inner diameter of 9.5 mm. Cylindrical samples, approximately 0.5 cm3 in volume, were
sliced from the PVC tubing, the tubing was removed and the sample was weighed in air
and heptane to determine the initial density of the hydrogel using Equation 5.2:
105
!hydrogel = !heptane × mair
mair -‐ mheptane (5.2)
where ρhydrogel is the density of the hydrogel, ρhep is the density of heptane, mair is the mass
of the hydrogel in air, and mheptane is the mass of the hydrogel in heptane [84]. Using the
density and initial mass of each sample the initial volume is calculated by dividing the
mass by the density. The samples were then placed in dialysis tubing; dialysis tubing is
used to prevent uptake of PEG (20 kDa) by the hydrogels. Hydrogels in tubing were
placed in the 0.2 MPa PEG solutions for up to 28 days at 37oC. The volume of swelling
medium was 100x larger than the volume of hydrogel samples to prevent significant
changes in the pressure of the osmotic solution due to changes in the hydrogel water
content over the length of the study.
Samples were removed from the PEG solution and the dialysis tubing at each
time point (0, 1, 4, 7, 14, 21 and 28 days); after which each sample was weighed in air
and heptane to determine the swelling ratio (V/Vo) by comparing the volume of swollen
samples to the volume of the initial samples (Equation 5.3). After swelling, hydrogel
samples were dried in an oven at 50oC; the mass of the dried hydrogel samples was
compared to the initial mass to calculate the initial water content (Equation 5.4) and the
dry mass compared to the swollen mass to calculate the equilibrium water content
(Equation 5.5). The mass retention value is calculated using Equation 5.6, where the dry
106
Swelling Ratio = swollen volumeinitial volume
(5.3)
Initial Water Content = 1 – dry massinitial mass
× 100% (5.4)
Equilibrium Water Content = 1 – dry massswollen mass
× 100% (5.5)
Mass Retention = dry mass of the swollen gel
initial mass of the swollen gel × dry mass of the unswollen gelinital mass of the unswollen gel
× 100% (5.6)
mass of the swollen gel is divided by the product of the initial mass of the swollen gel and
the ratio of the dry mass to the initial mass of the unswollen gel.
5.2.4. Mechanical Properties
Unconstrained, uniaxial compressive modulus was measured to determine if the
hydrogel is a suitable nucleus pulposus replacement material. Testing was conducted
using an Instron Materials Testing System Series 4442 (Norwood, MA) bench-top
mechanical testing system with a 50 N load cell.
107
After the hydrogel was allowed to react for the specified amount of time, the
hydrogel was autoclaved at 121oC for 30 minutes. After the autoclave cycle the hydrogel
was loaded into a 60 cc syringe and injected into poly(vinyl chloride) tubing with an inner
diameter of 9.5 mm. The hydrogel was left in the tubing at room temperature (RT) until
it was tested at each time point. Sample time points were: 20 minutes, 2 hours, and 24
hours.
At each time point, cylindrical hydrogel samples were sliced (n = 5), to a thickness
of 7-8mm and cut to ensure flat surfaces. The samples were preloaded to 0.1 N and
compressed to a total strain of 30% at a strain rate of 100% min-1, a chord from the initial
linear portion of the stress versus strain curve (10-20% strain) was used to calculate elastic
modulus. At least five independent samples were tested for each set of hydrogels (n = 5).
5.2.5. FTIR
FTIR spectra were collected using a Thermo Nicolet 6700 with DLaTGS
detector, Ge on KBr beamsplitter, and single reflection diamond Attenuated Total
Reflectance (ATR) sample accessory. Spectra were obtained with 64 scans analyzed with
Omnic 8.1.11 software (ThermoFischer Scientific).
After the hydrogel was allowed to react for 24 hours, the hydrogel was autoclaved
at 121oC for 30 minutes. Following the autoclave cycle the hydrogel was loaded into a 60
cc syringe and injected onto polyethylene terephthalate copolymer with cyclohexylene
dimethylene segments (PETG) sheeting with spacers of 0.1 mm thickness to control
108
thickness of the film. Samples larger than the diameter of the ATR crystal were cut from
the hydrogel film and then tested (n=3).
5.2.6. X-ray Diffraction
X-ray diffraction can be used to directly determine the crystallinity of materials.
X-rays primarily interact with the electrons in atoms, so when x-ray photons collide with
atoms, some photons from the incident beam will be deflected from the direction they
were originally traveling. When the wavelength (λ) is held constant, only momentum has
been transferred during this scattering. These scattered x-rays carry information about the
atomic arrangements of the materials; when bombarded on crystals the x-rays are
scattered as per their incidence directions (fixed λ). At certain angles, depending upon
distance between atomic plans and λ, the scattered x-rays reinforce each other to produce
intense beams; this phenomenon is called diffraction, which is stated by Bragg’s law:
! ! = 2! sin! (4.7)
where θ is the scattering angle, d is the distance between atomic planes, λ is the
wavelength of the x-ray, and n is an integer. Diffraction occurs when Equation 4.7 is
satisfied. In powder diffraction, λ is fixed and θ is variable. Intensity of the diffracted x-
ray beam is plotted again the angle of diffraction (2θ) to display the diffraction pattern of
109
the material; on which crystalline solids show as peaks at their respective diffraction angle
and amorphous solids (or liquid), with their lack of periodicity, result in one or two broad
humps. The crystallinity of a polymer, using x-ray diffraction, is expressed in terms of
relative degree of cyrstallinity. Using the wide angle x-ray diffraction (WAXD) method,
the intensity of x-ray scatter from the entire sample is equal to the area under the
diffraction pattern, this includes the crystalline and amorphous. The relative degree of
crystallinity (Q) is calculated using Equation 4.8:
! = !!!!!!!
(4.8)
where Ic is the intensity of the diffraction beam from the crystalline portion (area under
the sharply resolved peaks) and Ia is the intensity of the diffracted beam from the
amorphous part (remaining area under the curve, above the background)[209, 210].
WAXD diffraction patterns were collected using a Rigaku D/max-B Powder X-
ray Diffractometer (Tokyo, Japan) with an operating tube voltage and current of 40 kV
and 30 mA. Spectra were collected at room temperature using a 0.04o 2θ/s scan rate with
a four second dwell time, using a 2θ range of 8o to 35o. The 2θ range was chosen based
on previous work performed by Ricciardi et al. on PVA [104, 211, 212] and indicating
the primary crystalline peak for PVA occurs at 19.4o 2θ [104, 211, 212]. This
corresponds to a d spacing of 4.68 Å and crystalline dimensions in the [101] lattice
110
direction [104, 211, 212]. The diffraction peaks for PEG are at 19.2o and 23.4o 2θ, which
corresponds to d-spacings of 4.6 and 3.8 Å and are attributed to 120 and 231 reflections
of the monoclinic PEG unit cell [213]. In order to prevent drying of the sample during
testing, scan rate and dwell time were chosen to minimize data collection time.
WAXD was performed on 19% PEG-DGE (526, 2000 and 4600 Da), 19%
PEG-OH 29% PEG-DGE, 39% PEG-DGE, 12.6% PVA, 15.1% PVA, 17.5% PVA
and 20.0% PVA hydrogel formulations. Diffraction patterns were smoothed using
Savitzky-Golay's smoothing filter. Relative crystallinity was calculated using equation 4.2:
Relative Crystallinity % = PVA crystalline peak areatotal area
× 100% (4.7)
where PVA crystalline peak area is the area at 2θ =19.4° and total area is the entire area
within the 2θ range of 8°-35°. This relative value of crystallinity can be compared to other
samples using the same testing method; however, it does not represent an exact value for
crystallinity since the equation is not normalized using known crystallinity values.
5.2.7. Optical Microscopy
To accurately capture the microstructure of the hydrogel, the hydrogel was
imaged in a hydrated state. In order to image the hydrogel specimens hydrated, after the
final autoclave cycle the hydrogel was molded into 15 mL centrifuge tubes. Once the
material had gelled, the cylindrical samples were sectioned to 30 µm using a Leica
111
Microsystems cryostat model CM3050 (Bannockburn, IL) in the direction perpendicular
to the cylindrical axis. Immediately after sectioning, samples were images using various
objectives, up to 63x, using a Leica Microsystems DM 5500B Microscope (Bannockburn,
IL). Samples imaged included varying PEG-DGE concentration, varying PVA
concentration, PEG functional group, and PEG-DGE molecular weight.
5.2.8. Statistical Analysis
All data points are represented as the mean ± one standard deviation for at least
three independent samples. Statistical significance was determined by one-way analysis of
variance (ANOVA) with post-hoc analysis by Bonferroni correction with a 95%
confidence interval. P-values less than 0.05 were considered statistically significant.
5.3. Results and Discussion
5.3.1. Swelling Mechanics
A sample swelling ratio plot over 28 days for the 19% PEG-DGE hydrogel
formulation is shown in Figure 5.1. Figure 5.2 shows the swelling ratios of a hydrogel
formulation with increasing PEG-DGE content. Swelling ratio increases at one and two
weeks when PEG-DGE loading is increased from 19 to 29% (p<0.001), but does not
change between 29 and 39% (p>0.05). At weeks three and four, swelling ratio increases
with PEG-DGE content (p<0.01). The initial water content and mass retention of the
hydrogel formulations with increasing PEG-DGE content are shown in Table 5.2. As
112
PEG-DGE content is increased, the initial water content decreases and the mass
retention increases.
Figure 5.3 shows the swelling ratio of the hydrogel formulation with increasing
PVA content. As PVA content increases from 12.6 to 15.1% PVA, swelling ratio does
not increase at one, two, three, or four weeks (p>0.05). As PVA content increases from
15.1 to 17.5%, the swelling ratio increases at one (p<0.05), two (p<0.01), three and four
weeks (p<0.001). At one week there is no change between 17.5 to 20% PVA (p>0.05); at
two (p<0.001), three (p<0.05) and four (p<0.001) week swelling ratio increases with
increased PVA content. Table 5.3 shows the initial water content and mass retention for
the hydrogel formulation with increasing PVA content. As PVA content increases, the
initial water content of the hydrogel decreases (p<0.001); the mass retention increases
with increased PVA content.
The swelling ratio for 19% PEG with varied function group is shown in Figure
5.4. There is no difference between PEG-OH and PEG-DGE at one week (p>0.05), but
at two through four weeks the swelling ratio is higher for the PEG-OH formulation than
for the PEG-DGE formulation (p<0.001 at two and four weeks, p<0.01 at three weeks).
The initial water content and mass retention for the 19% PEG-DGE and 19% PEG-
OH formulations is shown in Table 5.4. There is no significant difference in water
content between the PEG-DGE and PEG-OH hydrogel; the mass retention is higher
for the PEG-OH hydrogel (79.6 ± 0.6%) than for the PEG-DGE hydrogel (64.6 ± 0.5).
113
The swelling ratio for 19% PEG-DGE hydrogel formulation with varying
molecular weight is shown in Figure 5.5. Swelling ratio increases as PEG-DGE
molecular weight increases to 2000 and 4600 Da from 526 Da (p<0.001). This trend is
not as clear between 2000 and 4600 Da, at four week the 4600 Da has a higher swelling
ratio than 2000 Da (p<0.001) but there is no difference at one and three weeks (p>0.05).
Table 5.5 shows the initial water content and mass retention for the 19% PEG-DGE
with increasing molecular weight. As molecular weight is increased from 526 to 2000 and
4600 Da, initial water content decreases but it does not vary significantly (p>0.05)
between 2000 and 4600 Da. Mass retention increases from 526 to 2000 and 4600 Da, but
does not change between 2000 and 4600 Da.
5.3.2. Mechanical Properties
A sample stress versus strain plot for the 39% PEG-DGE hydrogel formulation is
shown in Figure 5.6. The compressive modulus for the family of hydrogels at 20 minutes,
2 hour, and 24 hours with increasing PEG-DGE content is shown in Figure 5.7, with
increasing PVA content is shown in Figure 5.8, with varying PEG functional group is
shown in Figure 5.9, and with varying PEG-DGE molecular weight is shown in Figure
5.10. With increasing PEG-DGE content from 19 to 39%, compressive moduli increases
over the 24 hours testing period (p<0.01). Compressive moduli increases over 24 hours
from 12.6 to 20% PVA (p<0.001), except for 12.6 to 15.1% PVA at 20 minutes where
there was no difference (p>0.05).
114
When varying the PEG functional group, there is no change in modulus at 20
minutes or 2 hours (p>0.05), but at 24 hours the PEG-DGE formulation has a higher
modulus than PEG-OH (p<0.001). Formulations with 19% PEG, are very soft at 20
minutes and 2 hours; there might be a measureable difference in the moduli between
PEG-DGE and PEG-OH but not with the load cell used for this testing. At 24 hours,
the modulus of the hydrogels is within the reliable testing range of the load cell and it
accurately shows the increase in modulus with PEG-DGE over PEG-OH. As PEG-
DGE molecular weight increases, the compressive modulus increases over the 24-hour
testing period (p<0.001).
5.3.3. FTIR
Crosslinking of the PVA and PEG-DGE is not visible via FTIR, the crosslinking
reaction creates an ether which also makes up the entire backbone on the PEG polymer
so a slight increase in ethers is not detectable. FTIR of the PVA and PEG-DGE
hydrogel system can show increases and decreases in PVA and PEG-DGE content. As
explained earlier in this chapter, the hydrogels have a large amount of PEG-DGE
trapped in the pores of the hydrogel, which was eluted during the swelling studies and
resulted in lower than desired mass retention values. This trapped PEG-DGE is visible
in the spectra and masks some of the subtle changes in the hydrogel system as certain
variables are changed.
The FTIR spectrum for 19 and 39% PEG-DGE are shown in Figure 5.11;
Figure 5.12 shows the FTIR spectrum for 12.6 and 20.0% PVA. The spectrum for the
115
39% PEG-DGE hydrogel formulation shows an increase in PVA content over the 19%
PEG-DGE hydrogel formulation. The 12.6 and 20.0% PVA spectrum have basically
identical spectrum. Figure 5.13 shows the FTIR spectrum for 19% PEG-DGE and 19%
PEG-OH. The spectrum of the 19% PEG-OH shows increased PEG and water over
the 19% PEG-DGE spectrum. The FTIR spectrum for 19% PEG-DGE with varying
molecular weights of 526, 2000, and 4600 Da are shown in Figure 5.14. There is no
measureable difference between the spectrum for the 526, 2000 and 4600 Da hydrogel
formulations.
5.3.4. X-ray Diffraction
A large concern with measuring the relative crystallinity of the hydrogels is
maintaining the hydration level of the samples throughout the test. Hydrogel samples
were tested as quickly as possible by changing the 2θ range and the dwell time. Any
reduction in hydration level of the hydrogel would result in crystallization of the PEG-
DGE or PEG in the water rich domains and PVA in the polymer rich domains. The
diffraction pattern for a 30% PVA hydrogel and a 30% PVA hydrogel freeze-thaw cycled
five times (5 F/T) is shown in Figure 5.15; the primary PVA crystalline peak at 19.4o 2θ
is visible in the 30% PVA 5 F/T diffraction pattern. The diffraction patterns for 40%
PEG and 40% PEG-DGE are also shown in Figure 5.15. The 19.2o and 23.4o 2θ are
visible in the PEG and PEG-DGE diffraction patterns. Example diffraction patterns for
19% and 39% PEG-DGE hydrogel formulations are shown in Figure 5.16.
116
Relative crystallinity of hydrogel formulations with increasing PEG-DGE
content is shown in Figure 5.17. Relative crystallinity increases from 19 to 29% (p<0.05)
and from 19 to 39% (p<0.001); but there is no significant difference in relative
crystallinity between 29 and 39% PEG-DGE (p>0.05). Figure 5.18 shows the relative
crystallinity of hydrogel formulations with increasing PVA content. Relative crystallinity
increases from 12.6 to 20.0% PVA (p<0.001), but there is no other significant difference
between the relative crystallinity of the other formulations. The relative crystallinity for a
hydrogel formulation made with 19% PEG varying functional group, epoxide versus
hydroxyl is shown in Figure 5.19. There is no significant difference in relative crystallinity
in the PEG-DGE and PEG-OH hydrogels (p>0.05). Figure 5.20 shows the relative
crystallinity of the 19% PEG-DGE hydrogel formulation made with PEG-DGE of
varying molecular weight: 526 and 2000 Da. Relative crystallinity increases from the 526
Da hydrogel to the 2000 Da hydrogel (p>0.05).
5.3.5. Optical Microscopy
To accurately depict the hydrogel microstructure, the hydrogel must remain in the
hydrated state through out the imaging process to accurately represent the structure of
the hydrogel. The need for the hydrogel sample to stay hydrated does not allow for
imaging with scanning electron microscopy (SEM), as the hydrogel experiences some
level of drying in an environmental scanning electron microscope (ESEM) let alone a
traditional SEM. To minimize the effect of drying, the hydrogel was cryosectioned and
analyzed using optical microscopy.
117
Figure 5.21 shows optical micrographs of the hydrogel formulations 19% PEG-
DGE and 39% PEG-DGE. In the 19% PEG-DGE hydrogel there are very small
water/PEG-DGE regions present; there are considerably larger water/PEG-DGE
regions present in the 39% PEG-DGE. Optical micrographs of the 19% PEG-DGE
(526 Da) and 19% PEG-OH (600 Da) hydrogels are shown in Figure 5.22. The PEG-
DGE phase separates into a polymer rich gel and a water rich supernatant with some
small pores of the water/PEG-DGE rich phase; the PEG-OH hydrogel does not form a
supernatant but instead phase separates over the course of hours and the water/PEG rich
phase is trapped within the PVA hydrogel instead of forming a supernatant. The optical
micrographs of the hydrogel formulations of 19% PEG-DGE with varying molecular
weight (526, 2000 and 4600 Da) are shown in Figure 5.23. The water/PEG-DGE rich
regions are much larger in the 2000 and 4600 Da hydrogels than in the 526 Da hydrogel.
5.3.6. Discussion
It has been reported that increases in crosslinking decrease chain mobility and the
ability of a matrix to swell [95, 104, 197, 198]. The 19% PEG-OH formulation does not
have any chemical crosslinking present, which results in a higher swelling ratio than the
19% PEG-DGE formulation which is chemically crosslinked. As polymer concentration
of the hydrogel increases, it is expected for the network to experience additional swelling
due to additional polymer chains within the matrix [130, 182, 198, 199]. As PEG-DGE
and PVA content is increased for the tested formulations, there is an increase in swelling
118
ratio due to the increased polymer content. It does not appear that effects of crosslinking
of the hydrogel system could be seen over the effect of increased polymer content on
swelling ratio. As additional PEG-DGE is added to the system, more supernatant is
formed which removes more water from the initial PVA/PVP solution. This results in a
denser polymer-rich gel and a larger volume of water-rich supernatant phase. The
removal of PEG-DGE in the supernatant phase and the creation of a denser polymer
rich gel increases mass retention with the addition of more PEG-DGE; the additional
PEG-DGE pulls water out of the system instead of being trapped in PEG-DGE/water
filled pores within the PVA hydrogel. A constant 30% PEG-DGE is added to each of
the PVA formulations, the amount of supernatant removed is roughly the same but this
results in a denser gel as PVA content increases due to feed water content decreasing
from 50.3 to 42.8% as PVA content was increased from 12.6 to 20.0%. For the
formulations varying PEG-DGE molecular weight, it was expected to see an increase in
swelling ratio due to the increase in length of the PEG hydrophilic backbone [108].
As previously stated: cadaver testing and finite element modeling have shown that
a polymeric hydrogel implant should have a compressive modulus of at least 50 kPa at
15% strain to restore healthy tension in annulus fibers [84, 85]. With increasing polymer
content in the PVA hydrogels, we expect an increase in compressive modulus [130].
There has been significant research in literature suggesting the formation of physical
crosslinks, during the freeze/thaw process, allow for PVA hydrogels to carry increased
load [104, 197, 201]. These physical crosslinks are formed between the PVA chains, as
hydrogen bonds and crystallites, and are also formed during the theta-gel method [98-
119
103, 113] used to create the hydrogels for this thesis work. In addition to the physical
network in the PVA hydrogel, the addition of chemical crosslinks increases the
mechanical properties of the hydrogel network [187, 202, 214]. As PEG-DGE content
was increased, the increase in modulus is partly due to increased crosslinking because of
an increase in epoxide functional groups added to the system. The increase in modulus is
also due to an increase in polymer content of each formulation as more PEG-DGE is
added and the increase in relative crystallinity with increased PEG-DGE content. The
addition of more PEG-DGE removes more water from the system in a supernatant
phase, allowing for a more polymer dense gel phase. As more water is pulled out of the
system with the addition of PEG-DGE, it pushes the PVA chains closer together
allowing for increased hydrogen bonding and crystallization. This increased polymer
density, hydrogen bonding and crystallization results in increased compressive moduli. As
PVA content was increased, the number of epoxide groups available in each formulation
stays constant, so there is most likely not increased crosslinking, though there might be
due to an increase in available hydroxyl groups with increased PVA content. The increase
in compressive moduli as PVA content is increased is most likely due to an increase in
overall polymer content and relative crystallinity in the hydrogel network due to the
additional PVA and lower water content in the system.
The 19% PEG-OH and 19% PEG-DGE hydrogels have same feed polymer
content and no significant difference in relative crystallinity, but the 24 hour modulus
value for the PEG-DGE is significantly higher than the PEG-OH. This increase in
modulus is due to chemical crosslinking in addition to the physical network in the PEG-
120
DGE, while the PEG-OH formulation only has the physical network. The 19% PEG-
DGE and PEG-OH formulations (OH vs. epoxide) experience differences in phase
separation (supernatant formation versus no supernatant), the relative crystallinity values
suggest that the same amount of water is removed from the PVA rich domains resulting
in similar relative crystallinity values. When varying the PEG-DGE molecular weight
(526, 2000, 4600 Da) at a constant loading of 19 wt% an increase in modulus is expected.
As the polymer backbone length increases there is an increase in entanglements, which
results in an increase in moduli. As PEG molecular weight is increased from 526 to 2000
Da there is also an increase in crystallinity. This increase is due to the longer PEG chain
removing more water from the PVA rich region, allowing for the PVA chains in the
2000 Da hydrogel to be brought into closer proximity allowing for hydrogen bonding and
crystallization.
The difference in the size of the water/PEG-DGE regions is due to the
additional PEG-DGE in the 39% PEG-DGE hydrogel. As additional PEG-DGE is
added to the PVA solution to make the hydrogel, the Flory interaction parameter, χ,
increases [100-103] resulting in demixing. The demixing with additional PEG-DGE not
only occurs as a supernatant but it also forms larger pores within the PVA hydrogel that
are filled with water and PEG-DGE. The 19% PEG-DGE (526 Da) and 19% PEG-
OH (600 Da) hydrogels only vary in PEG functional group but the mechanism for
gelation differs due to the solubility of the end groups. The PEG-DGE phase separates
into a polymer rich gel and a water rich supernatant with some small pores of the
water/PEG-DGE rich phase; the PEG-OH hydrogel does not form a supernatant but
121
instead phase separates over the course of hours and the water/PEG rich phase is trapped
within the PVA hydrogel instead of forming a supernatant. For the 19% PEG-DGE
formulations varying molecular weight, the water/PEG-DGE rich regions are much
larger in the 2000 and 4600 Da hydrogels than in the 526 Da hydrogel, this is due to
increased demixing of the PVA/water solution with increasing PEG-DGE molecular
weight.
Crosslinking of the PVA and PEG-DGE is not visible via FTIR, the crosslinking
reaction creates an ether which also makes up the entire backbone on the PEG polymer
so a slight increase in ethers is not detectable. FTIR of the PVA and PEG-DGE
hydrogel system can show increases and decreases in PVA and PEG-DGE content. As
explained earlier in this chapter, the hydrogels have a large amount of PEG-DGE
trapped in the pores of the hydrogel, which was eluted during the swelling studies and
resulted in lower than desired mass retention values. This trapped PEG-DGE is visible
in the spectra and masks some of the subtle changes in the hydrogel system as certain
variables are changed. The spectrum for the 39% PEG-DGE hydrogel formulation
shows an increase in PVA content over the 19% PEG-DGE hydrogel formulation. This
increase in PVA concentration is due to additional PEG-DGE content removing water
from the hydrogel system resulting in a more concentrated polymer rich region and a
larger volume of PEG-DGE/water supernatant.
During the synthesis of the 19% PEG-OH hydrogel, the gelation is not
instantaneous as it is with the PEG-DGE formulations, over the course of hours the
PVA solution gels but instead of creating a supernatant it phase separates within the
122
hydrogel, producing PEG and water filled pores within the PVA hydrogel. This phase
separation is show in Figure 5.24; the center image shows the 19% PEG-DGE hydrogel
and the 19% PEG-OH hydrogel. Both gels started at 75 gram batches, but a large
amount of water and PEG-DGE were removed from the PEG-DGE gel as a
supernatant where that volume of PEG and water are trapped within the PEG-OH
hydrogel. The images to the right and left in Figure 5.24 show how the phase separation
within the PEG-OH makes the hydrogel opaque where the PEG-DGE hydrogel is
translucent.
5.4. Conclusions
The initial compression studies suggested the improved mechanical properties
with the addition of PEG-DGE was due to increased crosslinking; the following swelling
mechanics studies suggested that the increase in swelling behavior with PEG-DGE
content was just due to increased polymer content of the hydrogel. The additional FTIR,
XRD and optical microscopy studies provided the additional information to determine
that the PEG-DGE not only chemically crosslinks the hydrogel network but the use of
PEG-DGE with small amounts of NaOH (100 µL per 75 gram batch) alters the gelation
mechanism from the use of PEG-OH.
It was shown that the PVA/PEG-DGE hydrogel are formed almost
instantaneously upon the addition of PEG-DGE and NaOH to the PVA/PVP solution,
resulting in the immediate formation of a water/PEG-DGE rich supernatant phase and a
123
PVA rich hydrogel, which depending on the concentration and molecular weight of the
PEG-DGE had varying size pores of water/PEG-DGE trapped within the polymer rich
PVA hydrogel. The resultant hydrogels had increased crystallinity with increased PEG-
DGE and PVA content due to the ability of the PVA chains to be brought into closer
proximity to promote hydrogen bonding and crystallization. All of the formulations
tested in this chapter were easily injectable through a 10-gauge needle (2.6 mm ID) using
a 60 cc syringe.
The largest obstacle with this family of hydrogels is the mass retention issue for
formulations made with the 526 Da PEG-DGE, the values range from 64.6 ± 0.5% for
the 19% PEG-DGE formulation to 78.2 ± 0.6% for the 20.0% PVA formulation. The
only formulations that did not have gross mass retention issues were the 2000 Da and
4600 Da formulations which had mass retention values of 91.3 ± 1.1% and 93.7 ± 3.0%,
respectively. As previously stated, these molecular weights were not commercially
available during the time period the majority of this work was done. A purification
technique to maximize the mass retention values will be examined in subsequent chapters.
124
Table 5.1: Sample compositions for 9 to 39% PEG-DGE and 12.6 to 20.0% PVA hydrogel formulations
Sample Composition (% w/w) Holding PVA Content Constant Holding PEG-DGE Content
Constant
Components 9%
PEG-DGE
19% PEG-DGE
29% PEG-DGE
39% PEG-DGE
12.6% PVA
15.1% PVA
17.5% PVA
20% PVA
PVA 145 kDa 14.27 14.27 14.27 14.27 12.57 15.06 17.53 20.00
PVP 58 kDa 0.13 0.13 0.13 0.13 0.13 0.14 0.17 0.20
Deionized H2O 69.60 59.60 49.60 39.60 50.30 47.80 45.30 42.80
BaSO4 1-10 µm 7.00 7.00 7.00 7.00 7.00 7.00 7.00 7.00
PEG-DGE 526 Da 9.00 19.00 29.00 39.00 30.00 30.00 30.00 30.00
125
Figure 5.1: Swelling ratio of the 19% PEG-DGE hydrogel formulation
Swelling Ratio of the 19% PEG-DGEHydrogel Formulation
0 5 10 15 20 25 300.60.70.80.91.01.11.21.31.4
Time, days
126
1 week 19 29 39 2 weeks 19 29 39
19 19 29 *** 29 *** 39 *** ns 39 *** ns
3 weeks 19 29 39 4 weeks 19 29 39
19 19 29 *** 29 *** 39 *** ** 39 *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 5.2: Swelling ratio of 19 to 39% PEG-DGE
Swelling Ratio VaryingPEG-DGE Content
1 week 2 weeks 3 weeks 4 weeks0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.419% PEG-DGE29% PEG-DGE39% PEG-DGE
Time, weeks
127
Table 5.2: Initial water content and mass retention for 19 to 39% PEG-DGE
Gel Initial Water Content Mass Retention
[%] [%]
19% DGE 57.5 ± 0.0 64.6 ± 0.5
29% DGE 49.4 ± 0.3 70.8 ± 0.8
39% DGE 44.5 ± 1.3 77.0 ± 0.4
Initial Water Content 19 29 39
19 29 *** 39 *** ***
Mass Retention 19 29 39
19 29 *** 39 *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
128
1 week 12.6 15.1 17.5 20.0 2 weeks 12.6 15.1 17.5 20.0
12.6 12.6 15.1 ns 15.1 ns 17.5 ns * 17.5 *** ** 20.0 * ** ns 20.0 *** *** ***
3 weeks 12.6 15.1 17.5 20.0 4 weeks 12.6 15.1 17.5 20.0 12.6 12.6 15.1 ns 15.1 ns 17.5 *** *** 17.5 *** *** 20.0 *** *** * 20.0 *** *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 5.3: Swelling ratio of 12.6 to 20.0% PVA
Swelling Ratio VaryingPVA Content
1 week 2 weeks 3 weeks 4 weeks0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.412.6% PVA15.1% PVA17.5% PVA20.0% PVA
Time, weeks
129
Table 5.3: Initial water content and mass retention for 12.6 to 20.0% PVA
Gel Initial Water Content Mass Retention
[%] [%]
12.6% PVA 49.9 ± 0.1 72.2 ± 0.2
15.1% PVA 47.8 ± 0.3 71.2 ± 0.6
17.5% PVA 44.0 ± 0.3 74.8 ± 1.0
20% PVA 42.5 ± 0.2 78.2 ± 0.6
Initial Water Content 12.6 15.1 17.5 20.0
12.6 15.1 *** 17.5 *** *** 20.0 *** *** ***
Mass Retention 12.6 15.1 17.5 20.0
12.6 15.1 ns 17.5 * ** 20.0 *** *** **
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
130
1
week DGE PEG 2 weeks DGE PEG
DGE DGE PEG ns PEG ***
3 weeks DGE PEG 4
weeks DGE PEG
DGE DGE PEG ** PEG **
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 5.4: Swelling Ratio of 19% PEG-DGE and PEG-OH
Swelling Ratio VaryingPEG Functional Group
1 week 2 weeks 3 weeks 4 weeks0.0
0.2
0.4
0.6
0.8
1.0
1.2PEG-OHPEG-DGE
Time, weeks
131
Table 5.4: Initial water content and mass retention for 19% PEG-DGE and 19% PEG-OH
Gel Initial Water Content Mass Retention
[%] [%]
19% DGE 57.5 ± 0.0 64.6 ± 0.5
19% PEG 58.8 ± 0.1 79.6 ± 0.6
Initial Water Content DGE PEG
DGE PEG ns
Mass Retention DGE PEG
DGE PEG ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
132
1
week 526 2000 4600 2 weeks 526 2000 4600
526 526 2000 *** 2000 *** 4600 *** ns 4600 *** ***
3 weeks 526 2000 4600 4
weeks 526 2000 4600
526 526 2000 *** 2000 *** 4600 *** ns 4600 *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 5.5: Swelling ratio of 19% PEG-DGE with varying molecular weight (526, 2000 and 4600 Da)
Swelling Ratio VaryingPEG-DGE Molecular Weight
1 week 2 weeks 3 weeks 4 weeks0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6526 PEG-DGE2000 PEG-DGE4600 PEG-DGE
Time, weeks
133
Table 5.5: Initial water content and mass retention of 19% PEG-DGE with varying molecular weight (526, 2000 and 4600 Da)
Gel Initial Water Content Mass Retention
[%] [%]
526 MW 57.5 ± 0.0 64.6 ± 0.5
2000 MW 54.9 ± 0.9 91.3 ± 1.1
4600 MW 52.3 ± 2.3 93.7 ± 3.0
Initial Water Content 526 2000 4600
526 2000 * 4600 *** ns
Mass Retention 526 2000 4600
526 2000 *** 4600 *** ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
134
Figure 5.6: Stress versus strain plot for 39% PEG-DGE hydrogel formulation
Room Temperature Compression of39% PEG-DGE Hydrogel Formulation
0.0 0.1 0.2 0.3 0.40.0
0.1
0.2
0.3
0.4
0.520 minutes2 hours24 hours
Strain
135
20 m 19 29 39 2 h 19 29 39 24 h 19 29 39
19 19 19 29 ** 29 *** 29 *** 39 *** *** 39 *** *** 39 *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 5.7: Compression moduli over 24 hours of 19 to 39% PVA
Room Temperature CompressionVarying PEG-DGE Content
20 minutes 2 hours 24 hours0.0
0.5
1.0
1.5
2.019% PEG-DGE29% PEG-DGE39% PEG-DGE
Time
136
20 m 12.6 15.1 17.5 20.0 2 h 12.6 15.1 17.5 20.0
12.6 12.6 15.1 ns 15.1 *** 17.5 *** *** 17.5 *** *** 20.0 *** *** *** 20.0 *** *** ***
24 h 12.6 15.1 17.5 20.0
12.6 15.1 *** 17.5 *** *** 20.0 *** *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 5.8: Compression moduli over 24 hours of 12.6 to 20.0% PVA
Room Temperature CompressionVarying PVA Content
20 minutes 2 hours 24 hours0.0
0.5
1.0
1.5
2.0
2.512.6% PVA15.1% PVA17.5% PVA20.0% PVA
Time
137
20 m OH DGE 2 h OH DGE 24 h OH DGE
OH OH OH DGE ns DGE ns DGE ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 5.9: Compressive moduli over 24 hours of 19% PEG-DGE and PEG-OH
Room Temperature CompressionVarying PEG Functional Group
20 minutes 2 hours 24 hours0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35PEG-OHPEG-DGE
Time
138
20 m 526 2000 4600 2 h 526 2000 4600 24 h 526 2000 4600
526 526 526 2000 *** 2000 *** 2000 *** 4600 *** *** 4600 *** *** 4600 *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 5.10: Compressive moduli over 24 hours of 19% PEG-DGE with varying molecular weight (526, 2000 and 4600 Da)
Room Temperature CompressionVarying PEG-DGE Molecular Weight
20 minutes 2 hours 24 hours0.00.10.20.30.40.50.60.70.80.91.01.11.21.31.4
526 MW2000 MW4600 MW
Time
139
Figure 5.11: FTIR of 19 and 39% PEG-DGE hydrogel formulations
140
Figure 5.12: FTIR of 12.6 and 20.0% PVA hydrogel formulations
141
Figure 5.13: FTIR of 19% PEG-DGE and PEG-OH hydrogel formulations
142
Figure 5.14: FTIR of 19% PEG-DGE with varying molecular weight (526, 2000 and
4600 Da)
143
Figure 5.15: X-ray diffraction patterns of PVA, F/T PVA, PEG and PEG-DGE
X-ray Diffraction Pattern of30% PVA, 30% PVA 5 F/T cycles
40% PEG, and 40% PEG-DGE
8 18 28
30% PVA30% PVA, 5 F/T cycles
8 18 28
40% PEG
8 18 28
40% PEG-DGE
2Theta, o
144
Figure 5.16: X-ray diffraction patterns of 19% and 39% PEG-DGE hydrogel
formulations
X-ray Diffraction Patterns of19% PEG-DGE and 39% PEG-DGE
Hydrogel Formulations
8 18 28
19% PEG-DGE
8 18 28
39% PEG-DGE
2Theta, o
145
19 29 39
19 29 * 39 *** ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 5.17: Relative crystallinity for formulations with varying PEG-DGE content
Relative Crystallinity of VaryingPEG-DGE Content
0.0
1.0
2.0
3.0
4.0
146
\
12.6 15.1 17.5 20.0
12.6 15.1 ns 17.5 ns ns 20.0 *** ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 5.18: Relative crystallinity for formulations with varying PVA content
Relative Crystallinity of VaryingPVA Content
0.0
1.0
2.0
3.0
4.0
5.0
147
Figure 5.19: Relative crystallinity for formulations varying PEG functional group (p>0.05)
Relative Crystallinity of VaryingPEG Functional Group
0.0
0.5
1.0
1.5
2.0
148
Figure 5.20: Relative crystallinity for formulations varying PEG molecular weight
(p<0.05)
Relative Crystallinity of VaryingPEG Molecular Weight
0.0
0.5
1.0
1.5
2.0
2.5
149
Figure 5.21: Optical micrographs of 19% PEG-DGE and 39% PEG-DGE. Scale bars
indicate 50 µm.
150
Figure 5.22: Optical micrographs of 19% PEG-DGE and 19% PEG-OH. Scale bars
indicate 50 µm.
151
Figure 5.23: Optical micrographs of 19% PEG-DGE at varying molecular weights (526
Da, 2000 Da and 4600 Da). Scale bars indicate 50 µm.
152
Figure 5.24: 19% PEG-DGE and 19% PEG-OH phase separation
153
6. PURIFICIATION METHOD DEVELOPMENT AND
CHARACTERIZATION OF THE PURIFIED PVA/PVP/PEG-DGE
HYDROGEL
6.1. Introduction
When characterizing the PVA/PEG-DGE hydrogel system in Chapter 5 it was
discovered that the hydrogel had mechanical and swelling properties in the range for
nucleus replacement materials, but the mass retention of the hydrogel formulations when
conditioned in an osmotic solution of 0.2 MPa at 37oC lost between 21 to 35% of their
total mass after four weeks. In order to solve this problem and obtain the highest mass
retention possible, a purification method needed to be development that minimally
affected the end mechanical and swelling properties.
The theory behind this purification method is the use of osmotic solutions; the
unbound PEG-DGE could be eluted from the hydrogel and minimizes the increase in
water content of the hydrogel. Osmotic solutions were made using the equation 6.1:
Π = !" !!!!+ !!!! + !!!! +⋯ (6.1)
where Π is osmotic pressure, R is the gas constant, T is absolute temperature and c is
concentration in g/mL solution. Osmotic solutions were made using PEG (20 kDa) at
pressures ranging from 0.55 to 0.85 MPa. Using these osmotic solutions, the hydrogels
were conditioned at different pressures and for different amount of time to elute any
154
unbound polymer while trying to have minimal effect on the water content of the
hydrogel formulations.
After a purification procedure was optimized for a particular osmotic pressure and
length of time, each of the hydrogel formulations varying PEG-DGE content and PVA
content were purified to determine the effect of the purification process on the end
properties of the hydrogel networks. In the following experiments, the structure-property
relationship in these purified chemically crosslinked hydrogels was studied. Specifically,
swelling mechanics, stiffness, relative crystallinity and microstructure were investigated as
a function of PEG-DGE content and PVA content after a common optimized
purification procedure. The overall objective of these studies is to remove the unbound
PEG-DGE from the hydrogel network to increase the mass retention values for the
resultant hydrogel formulations with minimal effect on the end properties of the hydrogel.
6.2. Materials and Methods
6.2.1. Materials
PVA (99.0-99.8% hydrolyzed, molecular weight 145 kDa) was purchased from
Aldrich. PVP (molecular weight 58 kDa) was purchased from ISP Technologies. PEG
(molecular weights 600 Da and 4.6 kDa) was obtained from Aldrich and (molecular
weight 20 kDa) was obtained from Crescent Chemical Company. PEG-DGE (molecular
weights 526 Da, 2 kDa, 6 kDa) was obtained from Aldrich, PEG-DGE (molecular
weight 4.6 kDa) was custom synthesized by Advanced Polymer Materials Inc. Barium
155
sulfate (BaSO4) with a 1-10 μm particle size was purchased from J.T. Baker. Sodium
hydroxide (reagent grade, ≥ 98%) and sodium chloride (reagent grade, ≥ 99%) were
obtained from Sigma-Aldrich. Dialysis tubing, purchased from Fisher, had a nominal
molecular weight cut-off of 3,500 g/mol. Cell viability assay, (CellTiter-Glo®
Luminescent cell viability assay) was purchased from Promega (Madison, Wisconsin).
L929 mouse fibroblast cells were purchases from ATCC (Manassas, Virginia).
6.2.2. Hydrogel Purification
Using the hydrogel synthesis procedure described in Chapter 5, the hydrogel was
synthesized and after the final processing step the hydrogel was molded into 15 mL
centrifuge tubes. After 24 hours the gels were removed from the centrifuge tubes and
placed in 3,500 molecular weight cut off (MWCO) cellulose dialysis tubing and clipped
at the open ends with dialysis clips. The hydrogel/tubing assembly was then placed in the
osmotic solution at 37oC for a set amount of time for purification. After the set amount
of time, the hydrogels were removed from the osmotic solution and dialysis assembly.
The hydrogels were then autoclaved at 121oC for 30 minutes in a sealed media bottle,
after the autoclave cycle the hydrogel is equilibrated to room temperature and loaded into
a 60 cc syringe for molding.
Initially purification was done on the 39% PEG-DGE hydrogel formulation, the
formulation with the largest feed weight percent of PEG-DGE. The 39% PEG-DGE
formulation was purified in osmotic solutions ranging from 0.55 to 0.85 MPa, for 4 to 10
156
day at 37oC. After the purification procedure was optimized with the 39% PEG-DGE
formulation, other formulations were purified with the optimized pressure and time: 19%
PEG-DGE, 19% PEG-OH, 29% PEG-DGE, 12.6% PVA, 15.1% PVA, 17.5% PVA
and 20.0% PVA.
6.2.3. Swelling Mechanics
Osmotic solutions were made by dissolving PEG (20 kDa) in 0.15 M sodium
chloride to achieve osmotic pressures mimicking the swelling pressure of the IVD.
Equation 6.1 was used to calculate the PEG concentration, c2, required to for an osmotic
pressure, Π:
Π = !" !!!!+ !!!! + !!!! +⋯ (6.1)
where R is the universal gas constant, T is the absolute temperature and M2 is the
polymer molecular weight. The second and third virial coefficients, B and C, for 20 kDa
PEG are 2.59 x 10-3 and 13.5 x 10-3, respectively. The osmotic pressures used for this
study is the midpoint of the range (0 to 0.4 MPa) measured by Urban et al. for cadaver
IVDs [168]. To obtain the osmotic pressure of 0.2 MPa, a PEG concentration of 128.2
g/mL is used.
157
After the autoclave cycle of the hydrogel synthesis procedure, the hydrogel was
loaded into a 60 cc syringe and injected into poly(vinyl chloride) (PVC) tubing with an
inner diameter of 9.5 mm. Cylindrical samples, approximately 0.5 cm3 in volume, were
sliced from the PVC tubing, the tubing was removed and the sample was weighed in air
and heptane to determine the initial density of the hydrogel using Equation 6.2:
!hydrogel = !heptane × mair
mair -‐ mheptane (6.2)
where ρhydrogel is the density of the hydrogel, ρhep is the density of heptane, mair is the mass
of the hydrogel in air, and mheptane is the mass of the hydrogel in heptane [84]. Using the
density and initial mass of each sample the initial volume is calculated by dividing the
mass by the density. The samples were then placed in dialysis tubing; dialysis tubing is
used to prevent uptake of PEG (20 kDa) by the hydrogels. Hydrogels in tubing were
placed in the 0.2 MPa PEG solutions for up to 28 days at 37oC. The volume of swelling
medium was 100x larger than the volume of hydrogel samples to prevent significant
changes in the pressure of the osmotic solution due to changes in the hydrogel water
content over the length of the study.
Samples were removed from the PEG solution and the dialysis tubing at each
time point (0, 1, 4, 7, 14, 21 and 28 days); after which each sample was weighed in air
158
and heptane to determine the swelling ratio (V/Vo) by comparing the volume of swollen
samples to the volume of the initial samples (Equation 6.3). After swelling, hydrogel
samples were dried in an oven at 50oC; the mass of the dried hydrogel samples was
compared to the initial mass to calculate the initial water content (Equation 6.4) and the
dry mass compared to the swollen mass to calculate the equilibrium water content
(Equation 6.5). The mass retention value is calculated using Equation 6.6, where the dry
Swelling Ratio = swollen volumeinitial volume
(6.3)
Initial Water Content = 1 – dry massinitial mass
× 100% (6.4)
Equilibrium Water Content = 1 – dry massswollen mass
× 100% (6.5)
Mass Retention = dry mass of the swollen gel
initial mass of the swollen gel × dry mass of the unswollen gelinital mass of the unswollen gel
× 100% (6.6)
mass of the swollen gel is divided by the product of the initial mass of the swollen gel and
the ratio of the dry mass to the initial mass of the unswollen gel.
159
6.2.4. Mechanical Properties
Unconstrained, uniaxial compressive modulus was measured to determine if the
hydrogel is a suitable nucleus pulposus replacement material. Testing was conducted
using an Instron Materials Testing System Series 4442 (Norwood, MA) bench-top
mechanical testing system with a 50 N load cell.
After the hydrogel was allowed to react for the specified amount of time, the
hydrogel was autoclaved at 121oC for 30 minutes. After the autoclave cycle the hydrogel
was loaded into a 60 cc syringe and injected into poly(vinyl chloride) tubing with an inner
diameter of 9.5 mm. The hydrogel was left in the tubing at room temperature (RT) until
it was tested at each time point. Sample time points were: 20 minutes, 2 hours, and 24
hours.
At each time point, cylindrical hydrogel samples were sliced (n = 5), to a thickness
of 7-8mm and cut to ensure flat surfaces. The samples were preloaded to 0.1 N and
compressed to a total strain of 30% at a strain rate of 100% min-1, a chord from the initial
linear portion of the stress versus strain curve (10-20% strain) was used to calculate elastic
modulus. At least five independent samples were tested for each set of hydrogels (n = 5).
6.2.5. FTIR
FTIR spectra were collected using a Thermo Nicolet 6700 with DLaTGS
detector, Ge on KBr beamsplitter, and single reflection diamond Attenuated Total
160
Reflectance (ATR) sample accessory. Spectra were obtained with 64 scans analyzed with
Omnic 8.1.11 software (ThermoFischer Scientific).
After the purification procedure, the hydrogel was autoclaved at 121oC for 30
minutes; upon equilibrating to RT the hydrogel was loaded into a 60 cc syringe and
injected onto polyethylene terephthalate copolymer with cyclohexylene dimethylene
segments (PETG) sheeting with spacers of 0.1 mm thickness to control thickness of the
film. Samples larger than the diameter of the ATR crystal were cut from the hydrogel
film and then tested (n=3).
6.2.6. X-ray Diffraction
WAXD diffraction patterns were collected using a Rigaku D/max-B Powder X-
ray Diffractometer (Tokyo, Japan) with an operating tube voltage and current of 40 kV
and 30 mA. Spectra were collected at room temperature using a 0.04o 2θ/s scan rate with
a four second dwell time, using a 2θ range of 8o to 35o. In order to prevent the drying the
sample during testing, scan rate and dwell time were chosen to minimize data collection
time.
WAXD was performed on 19% PEG-DGE, 29% PEG-DGE, 39% PEG-DGE,
12.6% PVA, 15.1% PVA, 17.5% PVA and 20.0% PVA hydrogel formulations purified at
0.65 MPa for 7 days. Diffraction patterns were smoothed using Savitzky-Golay's
smoothing filter. Relative crystallinity was calculated using equation 4.2:
161
Relative Crystallinity % = PVA crystalline peak areatotal area
× 100% (4.7)
where PVA crystalline peak area is the area at 2θ =19.4° and total area is the entire area
within the 2θ range of 8°-35°. This relative value of crystallinity can be compared to other
samples using the same testing method; however, it does not represent an exact value for
crystallinity since the equation is not normalized using known crystallinity values.
6.2.7. Optical Microscopy
To accurately capture the microstructure of the hydrogel, the hydrogel was
imaged in a hydrated state. In order to image the hydrogel specimens hydrated, after the
purification procedure, the hydrogel was autoclaved at 121oC for 30 minutes, upon
equilibrating to RT the hydrogel was molded into 15 mL centrifuge tubes. Once the
material had gelled, the cylindrical samples were sectioned to 30 µm using a Leica
Microsystems cryostat model CM3050 (Bannockburn, IL) in the direction perpendicular
to the cylindrical axis. Immediately after sectioning samples were images using various
objectives, up to 63x, using a Leica Microsystems DM 5500B Microscope (Bannockburn,
IL). Samples imaged included varying PEG-DGE concentration and varying PVA
concentration.
162
6.2.8. Cytotoxicity
The cytotoxicity of selected PEG-DGE and PEG-OH hydrogel formulations
was tested per ISO 10993-5 and 10993-12. The tested formulations were 19% PEG-
DGE (526 Da, 2 kDa and 4.6 kDa), 19% PEG-OH (600 Da), 29% PEG-DGE (526
Da) and 39% PEG-DGE (526 Da). The hydrogel formulations containing PEG-DGE
(526 Da) and PEG-OH (600 Da) were purified for 7 days at 0.65 MPa, at 37oC.
Extractions of each hydrogel formulation were made using PBS (polar solvent)
and cottonseed oil (nonpolar solvent). Using an extraction ratio of 0.1 g/mL, for
irregularly shaped porous devices, and a largest implant size of 7 mL (density = 1.1580
g/mL). A hydrogel sample of 8.106 ± 10% g was placed in 81.06 mL of solvent for
extraction. The extraction conditions were 37 ± 1oC for 72 ± 2 hours. The hydrogel
samples were removed after the extractions were complete and the solutions were
sterilized in an autoclave cycle (30 minutes at 121oC).
L929 mouse fibroblasts (ATCC CCL-1) cells were cultured in Eagle’s Minimum
Essential Medium (EMEM) (ATCC 30-2003) and 10% horse serum (ATCC 30-2041).
Controls used were natural latex (positive control), polyethylene (negative control), and
PBS or cottonseed oil (reagent control). Cells were plated in all white 96 well plates at a
density of 5 x 103 cells/well (75 µL). Cells were allowed to attach over night, after which
75 µL of each extract was added to the plates (n=3, 5 repetitions per extract). Promega
Cell-Titer Glo assay was used to measure cytotoxicity; this assay determines the number
if viable cells in culture based on a quantitation of the ATP present, which signals the
163
presence of metabolically active cells. The Cell-Titer Glo reagent (150 µL) was added to
each well, the plate was mixed on an orbital shaker for 2 minutes, after shaking it was
incubated at RT for 10 minutes and then each plate was read on a TECAN 200 Pro plate
reader with an integration time of 1 second per well.
Concentrations of PEG-DGE (526 Da, 2 kDa and 4.6 kDa) were also tested;
PEG-DGE was dissolved in PBS (1x) ranging in concentration from 1.56 to 100 mg/mL
(5 repetitions per concentration). The solutions were sterilized in an autoclave cycle (30
minutes at 121oC) unless otherwise noted. The controls for these studies were natural
latex (positive), polyethylene (negative) and PBS (reagent).
6.2.9. Statistical Analysis
All data points are represented as the mean ± one standard deviation for at least
three independent samples. Statistical significance was determined by one-way analysis of
variance (ANOVA) with post-hoc analysis by Bonferroni correction with a 95%
confidence interval. P-values less than 0.05 were considered statistically significant.
6.3. Results and Discussion
6.3.1. Hydrogel purification
The idea behind purification of the hydrogel material, is to remove all of the
unbound PEG from the hydrogel network before implantation into the body but to not
164
significantly change the properties of the hydrogel in the process (Figure 6.1). Initial
purification studies were performed on the 39% PEG-DGE formulation, the formulation
with the largest feed amount of PEG-DGE. The compressive moduli of the purified
hydrogel variations are shown in Figure 6.2; these samples were swollen in 0.2 MPa
osmotic solution at 37oC prior to compressive testing. At one week and two weeks, the
compressive modulus of the sample purified in 0.55 MPa osmotic solution (7 days) had a
lower modulus than the unpurified 39% PEG-DGE, all the other purified samples did
not have moduli significantly different than the 39% PEG-DGE. At four weeks, the
modulus was higher for the samples purified in 0.65 MPa (4 days), 0.75 MPa (7 days),
and 0.85 MPa (7 days) than the unpurified 39% PEG-DGE; all the other purified
samples did not have moduli significantly different than the 39% PEG-DGE.
Samples purified at 0.75 MPa (7 days) and 0.85 MPa (7 days) were very difficult
to inject through a 10-gauge needle (2.6 mm ID), because of this these purification
pressures removed from consideration. Due to the low modulus at one and two weeks,
0.55 MPa (7 days) was also removed from consideration. The remaining purification
conditions left for consideration are 0.65 MPa (4 days), 0.65 MPa (7 days), and 0.65
MPa (10 days); there is no difference in the compressive modulus between these
conditions at one, two or four weeks.
Figure 6.3 shows the swelling ratio of 39% PEG-DGE and the purified variations
of 39% PEG-DGE at four weeks. The swelling ratio is higher for all of the purified
samples than for the unpurified 39% PEG-DGE hydrogel (p<0.001). Between the
165
purification conditions left in consideration: 0.65 MPa (4 days), 0.65 MPa (7 days), and
0.65 MPa (10 days) there is no significant difference in their swelling ratios at four weeks.
Mass retention was the issue with the original formulations that lead to the need
of purification. Table 6.1 shows the initial water content and mass retention of the
unpurified 39% PEG-DGE and each of the purified samples. All of the purified samples
have higher initial water content and lower mass retention than the unpurified 39%
PEG-DGE formulation (p<0.001). Between the purification conditions left in
consideration: 0.65 MPa (4 days), 0.65 MPa (7 days), and 0.65 MPa (10 days) the mass
retention for the 0.65 MPa (7 days) is highest. Due to the highest mass retention of 0.65
MPa (7 days), this purification condition will be used for all further testing.
6.3.2. Swelling Mechanics
All formulations were purified in a 0.65 MPa osmotic solution for seven days at
37oC. Figure 6.4 shows the swelling ratios of purified formulations with varying PEG-
DGE content; Figure 6.5 shows the swelling ratios of purified formulations varying PVA
content. At one week, 19% PEG-DGE had a lower swelling ratio than 29% PEG-DGE
(p<0.01) but there was no significant difference between 19% and 39% or 29% and 39%
PEG-DGE (p>0.05). At four weeks, there is no significant difference in swelling ratio
for 19, 29, and 39% PEG-DGE (p>0.05). At one week, the swelling ratio increases from
12.6% PVA to 15.1, 17.5 and 20.0% PVA but there is no significant difference between
15.1, 17.5 and 20.0% PVA. At four weeks, there is no significant difference in swelling
ratio for 12.6, 15.1, 17.5, and 20.0 PVA (p>0.05).
166
The initial water content and mass retention of each gel before and after
purification is shown in Table 6.2. For unpurified hydrogels the mass retention values
range from 64.6 ± 0.5 to 78.2 ± 0.6%, for purified hydrogels the values range from 82.2 ±
1.7 to 95.0 ± 1.5%; the mass retention increases for each formulation after purification. In
addition to increasing mass retention values, initial water content also increases for each
formulation after purification.
The swelling ratios increase for the purified formulations versus the unpurified
formulations in Chapter 5. At 4 weeks, the unpurified DGE formulations had swelling
ratios ranging from 0.68 to 0.98 while the purified formulations swelling ratios range
from 1.03 to 1.08. There is a similar trend for the PVA formulations, the unpurified
swelling ratios range from 0.89 to 1.13 and the purified swelling ratios range from 1.03 to
1.10.
6.3.3. Mechanical Properties
Figure 6.6 shows the compressive modulus at 20 minutes, 2 hours, and 24 hours
for purified formulations as PEG-DGE content increased from 19 to 39%; Figure 6.7
shows the modulus for purified formulations as PVA content increased from 12.6 to 20%.
Unlike with the unpurified formulations shown in Chapter 5, there is not a difference in
compressive modulus at 20 minutes as PEG-DGE increases from 19 to 39% (p>0.05);
but at 24 hours the result is similar to the unpurified PEG-DGE formulations, the
compressive moduli increases as PEG-DGE content increases (p<0.01). In addition to
167
the trends changing, the modulus for 19, 29 and 39% PEG-DGE formulations are
considerably lower for the purified formulations; this is due to the increase water content
of the purified formulations (Table 6.2). The trends for purified formulations with
increasing PVA content are the same as with increasing PEG-DGE content, including
the decrease in compressive moduli in the purified gels from the unpurified gels.
6.3.4. FTIR
The FTIR spectrum for the purified 19 and 39% PEG-DGE formulations is
shown in Figure 6.8. The PEG peaks are more prominent in the 19% PEG-DGE
formulation spectra than in the 39% PEG-DGE formulation (2880 and 1059 cm-1) and
the PVA peaks are more prominent in the 39% PEG-DGE spectra (2937, 2906, 1142
and 1087 cm-1). The increase in the 1142 cm-1 PVA peak suggests increased crystallinity
[95, 183, 215] in the 39% PEG-DGE formulation than in the 19% PEG-DGE.
Figure 6.9 shows the FTIR spectrum for the purified 12.6 and 20.0% PVA
formulations. There is a decrease in water content from the 12.6 to the 20.0% PVA
formulation (3280 and 1623 cm-1) and an increase in PVA content (2937, 2906 and peaks
centered at 1087 cm-1). The increase in water content in the 12.6% PVA formulation was
also seen in the swelling mechanics experiments.
168
6.3.5. X-ray Diffraction
Relative crystallinity for purified formulations varying PEG-DGE content are
shown in Figure 6.10. For the purified PEG-DGE formulations, there is no significant
difference between 19 and 29% PEG-DGE (p>0.05) or 29 and 39% PEG-DGE
(p>0.05), there is an increase from 19 to 39% PEG-DGE (p<0.05). Figure 6.11 shows
the relative crystallinity for the purified formulations varying PVA content; as PVA
content increases from 12.6 to 20.0% there is no significant difference in relative
crystallinity (p>0.05). The changing of the gelation mechanism and water content
increase result in this change of crystallinity varying between hydrogel formulations,
though Table 6.4 shows that there is no difference between the crystallinity between the
unpurified and purified variation of any formulation.
6.3.6. Optical Microscopy
The change of the hydrogel gelation mechanism after purification is shown in the
optical micrographs of the purified 19% and 39% PEG-DGE formulations shown in
Figure 6.12. The unpurified 19% PEG-DGE micrograph in Chapter 5 is very similar to
the purified 19% PEG-DGE micrograph shown in Figure 6.12. The unpurified 19%
PEG-DGE formulation did not form as many, nor as large, PEG-DGE/water pores
within the hydrogel network as the higher content PEG-DGE formulations did; this
results in similar micrographs with and without the unbound PEG-DGE present in the
unpurified and purified 19% PEG-DGE micrographs. For the unpurified 39% PEG-
DGE shown in Chapter 5, there were larger pores throughout the micrograph of the
169
hydrogel. These pores were formed by the unbound PEG-DGE and water phase
separating from the PVA, due to the lack of unbound PEG-DGE in the purified
hydrogel the larger pores are no longer present.
6.3.7. Cytotoxicity
Cell viability was measured using the Cell-Titer Glo assay; the luminescence
values were divided by luminescence values of wells with just PBS or cottonseed oil to
calculate relative percent cell viability. There was no difference in cell viability (p>0.05)
for the 19% PEG-DGE (526 Da) hydrogel, 19% PEG-OH hydrogel, 39% PEG-DGE
hydrogel, 19% PEG-DGE (2000 Da) hydrogel, and 19% PEG-DGE (6000 Da)
hydrogel formulations extracted in PBS. For extractions in cottonseed oil, the only
difference (p<0.05) was between the 39% PEG-DGE hydrogel and the 19% PEG-DGE
(6000 Da) hydrogel; the 39% had a lower viability than the 6000 Da gel. All other
formulations extracted in cottonseed oil had no difference in their cell viability (p>0.05).
All the formulation extractions in PBS and cottonseed oil had no difference in cell
viability with the negative control, polyethylene and higher viability than the positive
control, latex.
The cell viability of various PEG-DGE concentrations was also tested. Figure
6.15 shows UV sterilized PEG-DGE concentrations of 1.56 to 100 mg/mL in PBS. The
PEG-DGE concentrations were sterilized using 365 nm wavelength UV light for 1 hour.
Figure 6.16 shows autoclave sterilized PEG-DGE concentrations. The cell viability for
the UV sterilized concentrations is 0.01 ± 0.00% for 100 mg/mL and increases to 0.09 ±
170
0.02% for 1.56 mg/mL. The cell viability for the autoclaved sterilized concentrations is
0.03 ± 0.01% for 100 mg/mL and increases to 82.14 ± 5.08% for 1.56 mg/mL. The only
difference between the groups is the sterilization method.
Figure 6.17 shows the cell viability of PEG-DGE concentrations from 1.56 to
100 mg/mL in PBS with varying PEG-DGE molecular weights. At 100 mg/mL the cell
viability increases with molecular weight (p<0.001): 0.03 ± 0.01% for 526 Da, 6.13 ± 1.08
for 2000 Da, and 28.38 ± 2.70 for 6000 Da. At 1.56 mg/mL there is an increase in
viability from 526 to 2000 Da (p<0.05) and 526 to 6000 Da (p<0.001), but there is no
difference between 2000 and 6000 Da.
Figure 6.18 shows the cell viability of PEG concentrations with epoxide and
hydroxyl end groups from 1.56 to 100 mg/mL in PBS. At 1.56 mg/mL there is no
difference between the hydroxyl terminated PEG and the epoxide terminated PEG
(p>0.05). At 100 mg/mL the cell viability is 73.52 ± 3.40% for the hydroxyl terminated
PEG and is 0.03 ± 0.01% for the epoxide terminated PEG, suggesting that cytotoxicity
for the two functional groups is dependent on PEG concentration.
6.3.8. Discussion
The purification process, in addition to increasing mass retention values, increases
the initial water content of the hydrogel formulations after the purification process. This
increase in water content is due to the lack of unbound PEG-DGE in the hydrogel
network. The hydrogel synthesized with the addition of PEG-DGE to the PVA, PVP,
171
DI water solution produced a higher PVA content gel than possible without the addition
of PEG. The removal of unbound PEG-DGE during the purification process resulted in
increased water content in the hydrogel network, which was necessary to push the
unbound PEG-DGE from the network. When the hydrogel was autoclaved after the
purification process, it formed a PVA, PVP, DI water, bound PEG network with
different compositions than the original formulations. This is a result of the additional
water in the system but also due to the unbound PEG-DGE no longer adding to the
gelation of the system by pulling additional water out of the PVA. This results in the
varying hydrogel formulations to have less significant differences in properties due to the
lack of PEG-DGE supernatant and PEG-DGE filled pores within the network, which
had previously encouraged more hydrogen bonding and crystallinity with increased PVA
and PEG-DGE content.
As discussed before, with increasing polymer content in the PVA hydrogels, we
expect an increase in compressive modulus [130]. The removal of the unbound PEG-
DGE in the hydrogel network, results in a lower polymer concentration in each
concentration than before purification and a higher water content. It was expected that
the purified formulations would have a lower compressive modulus than the unpurified
formulations. At 24 hours the compressive modulus for the purified formulations ranged
from 0.05 ± 0.01 to 0.25 ± 0.02 MPa for the 19 and 39% PEG-DGE formulations,
respectively. These values decreased from unpurified 24 hour compressive modulus values
of 0.31 ± 0.01 to 1.65 ± 0.05 MPa for the 19 and 39% PEG-DGE formulations,
respectively.
172
Unlike with the unpurified formulations shown in Chapter 5, there is not a
difference in compressive modulus at 20 minutes as PEG-DGE increases from 19 to 39%
(p>0.05); but at 24 hours the result is similar to the unpurified PEG-DGE formulations,
the compressive moduli increases as PEG-DGE content increases (p<0.01). This is due
to the lack of unbound PEG-DGE in the system causes the immediate phase separation
within the hydrogel network. This immediate phase separation with the formation of the
supernatant results in the forming of PVA rich regimes quickly. Without the formation
of the supernatant the hydrogel phase separates slowly within the hydrogel network
similarly to the 19% PEG-OH hydrogel and the PVA theta-gels in literature made with
low molecular weight PEG [98-103]. The lack of supernatant formulation and the slow
phase separation within the hydrogel network also affects the relative crystallinity of the
purified hydrogel formulations. The increases in crystallinity for the unpurified
formulations was due to the gelation of the PVA from the unbound PEG-DGE pulling
water out of the system as supernatant and in PEG-DGE/water filled pores, allowing for
addition hydrogen bonding and crystallization. This lack of unbound PEG-DGE in the
purified network does not promote additional supernatant and pore formation resulting
in no difference in crystallinity between the 19, 29 and 39% PEG-DGE formulations
and the 12.6, 15.1, 17.5 and 20.0% PVA formulations.
Phase separation for the PVA theta-gels in the formation of the pores and
supernatant is a result of the gellant added to the PVA solution used to form the theta-
gel [98-103, 112, 113]. After the addition of the gellant to the PVA solution, the
resulting solution has a higher Flory interaction parameter than the vinyl polymer
173
solution, which induces gelation of the mixture of vinyl polymer solution and gellant [98-
103, 112, 113]. For the unpurified formulations in this thesis work, the phase separation
is in the form of the PEG-DGE/water supernatant phase and PEG-DGE/water filled
pores within the hydrogel network. The removal of the unbound PEG-DGE in the
purification process changes the phase separation due to the minimal amount of residual
PEG-DGE. The amount of PEG-DGE left after purification varies from formulation to
formulation. Optical micrographs of the unpurified 19% PEG-DGE formulation shows
a minimal number of small pores, while the unpurified 39% PEG-DGE formulation has
increased number of larger pores. These pores are due to the increase in the Flory
interaction parameter by the addition of more gellant [100-103] and additional phase
separation. For the purified formulations, the mass retention values are 82.2 ± 1.7% for
the 19% PEG-DGE formulation and 92.8 ± 0.8% for the 39% PEG-DGE formulation.
These values suggest that more unbound PEG-DGE remains in the 19% PEG-DGE
formulation than in the 39% PEG-DGE formulation. This additional unbound PEG-
DGE in the purified 19% PEG-DGE hydrogel is why there is pore formation in the
optical micrograph. The minimal amount of unbound PEG-DGE in the purified 39%
PEG-DGE hydrogel is why there is no pore formation in the optical micrograph.
Difunctional compounds are know for their cytotoxic nature, due to reactive end
group functionality and low molecular weight [93, 95, 110, 149, 165, 200, 216-224].
Nishi et al. [149] tested a variety of diepoxide compounds, including PEG-DGE at three
molecular weights: 190, 234 and 542 g/mol. The cytotoxicity of the specimens was
expressed as NR50, which was defined as the specimen concentration at which the cell
174
activity was reduced to 50% of the control cells without any test specimen. A higher NR50
equals lower toxicity. The NR50 concentration increased from 27.4 ± 8.1 µg/mL for the
190 g/mol PEG-DGE, to 46.3 ± 9.0 µg/mL for the 234 g/mol PEG-DGE, and to 133.1
± 21.4 µg/mL for the 542 g/mol PEG-DGE. The decrease in toxicity with molecular
weight is due to a reduction in the number of reactive end groups, in the case epoxide
groups, for a certain mass of PEG-DGE due to the increasing weight of the increasing
polymer backbone length. A mass balance of the PVA/PEG-DGE hydrogel system
revealed that after supernatant removal and purification, the amount of PEG-DGE
(bound and unbound) left in the 29% PEG-DGE formulation is 100 mg/mL. The cell
viability for the UV sterilized concentrations is significantly lower than the autoclave
sterilized concentrations due to the presence of epoxide groups in the UV sterilized
samples and mostly hydrolyzed epoxide groups for the autoclave sterilized samples. The
NR50 concentration for the autoclave sterilized PEG-DGE (526 MW) is around 12.50
mg/mL instead of the literature value of 133.1 ± 21.4 µg/mL for 542 g/mol PEG-DGE
(unknown sterilization technique), a significant increase in concentration.
6.4. Conclusions
Using the 39% PEG-DGE formulation, a purification procedure was optimized
that allowed minimal initial water content change but increased the mass retention values
of the hydrogel. The selected purification procedure, of 0.65 MPa for 7 days at 37oC,
allowed for a hydrogel that was still injectable after the removal of the unbound PEG.
The unbound PEG worked as a plasticizer for the hydrogel network [225, 226] the
175
removal of the unbound PEG could have reduced the ability of the hydrogel to be
injected, water also works as a plasticizer in this hydrogel system[227, 228].
After purification at 0.65 MPa for 7 days at 37oC, the 19, 29, and39% PEG-DGE
and 12.6, 15.1, 17.5, and 20.0% PVA hydrogel formulations had increased swelling ratios,
decreased compressive modulus, and no significant change in crystallinity from their
unpurified formulations. This is due to a change in the gelation mechanism for the gels
after purification. When the hydrogels are originally gelled, the PEG-DGE removes
water from the PVA/PVP/water solution causing the solution to phase separate into
water/PEG-DGE regions and PVA regions, which almost instantly gel upon the
addition of PEG-DGE. The PEG-DGE/water phase either separates from the PVA gel
as a supernatant, which is decanted from the gel, or the PEG-DGE/water phase
separates into PEG-DGE and water filled pores. These PEG-DGE water filled pores
result in the low mass retention of the initially formed PVA/PVP/PEG-DGE hydrogels,
as the PEG-DGE is a low molecular weight that easily elutes from the hydrogel network
in an aqueous environment. The purification process removes this unbound PEG-DGE
from the PEG-DGE/water pores, which results in a higher mass retention for the system,
but it also changes the gelation of the hydrogel network as it is heated above 80oC for
injection and then cooled to RT. If the unbound PEG-DGE was still present in the
system, as it is in Chapter 5, the PEG-DGE allows for the formation of additional
supernatant and PEG-DGE/water pores causing more concentrated PVA rich regions,
which promote hydrogen bonding and crystallinity. These PVA rich regions resist
swelling and increase mechanical properties of the hydrogel. After purification, the initial
water content in the hydrogel is higher due to the purification process in addition to the
176
lack of unbound PEG-DGE pulling water out of the PVA and phase separating into
supernatant and PEG-DGE/water pores. This lack of densification of the PVA rich
regions results in lower mechanical properties and increased swelling of the hydrogel.
Cytotoxicity of the hydrogel formulations was determined using a cell viability assay.
The cytotoxicity of the extractions from the formulations in PBS and cottonseed oil
demonstrated no difference in cell viability from polyethylene, a commonly used
biomaterial, and an increase in cell viability from latex, a known cytotoxic agent. This was
extremely important to demonstrate due to the cytotoxic concerns of short chain
diepoxide compounds [149]. The repeated autoclave cycles used in the synthesis of the
hydrogel is thought to be the reason for the low cytotoxicity of the hydrogel extractions,
the heat of the autoclave cycle allows for intact epoxide groups to hydrolyze in the
aqueous environment of the hydrogel.
177
Figure 6.1: Diagram of the purification of the PVA/PEG-DGE hydrogel formulations
178
1 week 39% PEG-DGE
0.55 MPa 7d
0.65 MPa 4d
0.65 MPa 7d
0.65 MPa 10d
0.75 MPa 7d
0.85 MPa 7d
39% PEG-DGE 0.55 MPa 7d * 0.65 MPa 4d ns *** 0.65 MPa 7d ns ns ns
0.65 MPa 10d ns *** ns ns 0.75 MPa 7d ns ns ns ns * 0.85 MPa 7d ns *** ns * ns **
2 weeks 39% 0.55 0.65 0.65 0.65 0.75 0.85 39% PEG-DGE
0.55 MPa 7d *** 0.65 MPa 4d ns *** 0.65 MPa 7d ns *** ns
0.65 MPa 10d ns *** ns ns 0.75 MPa 7d ns *** ns ns ns 0.85 MPa 7d ns *** ns ns ns ns
4 weeks 39% 0.55 0.65 0.65 0.65 0.75 0.85 39% PEG-DGE
0.55 MPa 7d ns 0.65 MPa 4d * *** 0.65 MPa 7d ns ** ns
0.65 MPa 10d ns ns ns ns 0.75 MPa 7d *** *** ns ns ** 0.85 MPa 7d *** *** *** *** *** ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.2: Compressive moduli of 39% PEG-DGE purified at various osmotic pressures and various periods of time
Compressive Moduli of 39% PEG-DGE VaryingPurification Pressure and Length of Time
1 week 2 weeks 4 weeks0.0
0.5
1.0
1.5
2.039% PEG-DGE0.55 MPa 7 Days0.65 MPa 4 Days0.65 MPa 7 Days0.65 MPa 10 Days0.75 MPa 7 Days0.85 MPa 7 Days
Time, weeks
179
4 weeks 39% PEG-DGE
0.55 MPa 7d
0.65 MPa 4d
0.65 MPa 7d
0.65 MPa 10d
0.75 MPa 7d
0.85 MPa 7d
39% PEG-DGE 0.55 MPa 7d *** 0.65 MPa 4d *** ** 0.65 MPa 7d *** *** ns
0.65 MPa 10d *** *** ns ns 0.75 MPa 7d *** ** ns ns ns 0.85 MPa 7d *** *** ns ns ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.3: Swelling ratio at four weeks of 39% PEG-DGE purified at various osmotic pressures and various periods of time
Swelling Ratio of 39% PEG-DGE VaryingPurification Pressure and Length of Time
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
180
Table 6.1: Initial water content and mass retention of 39% PEG-DGE unpurified and purified at various osmotic pressures and various periods of time
Unpurif ied Gels Purif ied Gels
Gel Init ia l Water
Content
Mass Retention
Purif icat ion Solution
Length of Purif icat ion
Init ia l Water
Content
Mass Retention
[%] [%] [MPa] [Days] [%] [%]
39% DGE 44.5 ± 1.3 77.0 ± 0.4
0.55 7 54.5 ± 1.3 92.8 ± 0.7
0.65 4 47.1 ± 0.5 89.9 ± 1.0
0.65 7 50.1 ± 0.2 92.8 ± 0.8
0.65 10 49.0 ± 1.0 88.9 ± 3.0
0.75 7 48.8 ± 0.2 92.3 ± 0.2
0.85 7 47.3 ± 0.4 93.8 ± 0.1
Initial Water Content
39% PEG-DGE
0.55 MPa 7d
0.65 MPa 4d
0.65 MPa 7d
0.65 MPa 10d
0.75 MPa 7d
0.85 MPa 7d
39% PEG-DGE 0.55 MPa 7d *** 0.65 MPa 4d * *** 0.65 MPa 7d *** *** **
0.65 MPa 10d *** *** ns ns 0.75 MPa 7d *** *** ns ns ns 0.85 MPa 7d *** *** ns ns ns ns
Mass Retention
39% PEG-DGE
0.55 MPa 7d
0.65 MPa 4d
0.65 MPa 7d
0.65 MPa 10d
0.75 MPa 7d
0.85 MPa 7d
39% PEG-DGE 0.55 MPa 7d *** 0.65 MPa 4d *** ns 0.65 MPa 7d *** ns ns
0.65 MPa 10d *** * ns * 0.75 MPa 7d *** ns ns ns ns 0.85 MPa 7d *** ns * ns ** ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
181
1
week 19 29 39 2 weeks 19 29 39
19 19 29 ** 29 *** 39 ns ns 39 * ns 3
weeks 19 29 39 4 weeks 19 29 39
19 19 29 * 29 ns 39 ns ns 39 ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.4: Swelling ratio of purified 19 to 39% PEG-DGE hydrogel formulations
Swelling Ratio of Purified FormulationsVarying PEG-DGE Content
1 week 2 weeks 3 weeks 4 weeks1.0
1.1
1.2
1.3
1.419% PEG-DGE29% PEG-DGE39% PEG-DGE
Time, weeks
182
Table 6.2: Initial water content and mass retention of unpurified and purified 19 to 39% PEG-DGE hydrogel formulations
Unpurified Gels Purified Gels
Gel Initial Water
Content
Mass Retention
Purification Solution
Length of Purification
Initial Water
Content
Mass Retention
[%] [%] [MPa] [Days] [%] [%]
19% DGE 57.5 ± 0.0 64.6 ± 0.5 0.65 7 57.9 ± 0.3 82.2 ± 1.7
29% DGE 49.4 ± 0.3 70.8 ± 0.8 0.65 7 54.2 ± 0.4 95.0 ± 1.5
39% DGE 44.5 ± 1.3 77.0 ± 0.4 0.65 7 50.1 ± 0.2 92.8 ± 0.8
Initial Water Content
Unpurified 19 29 39 Purified 19 29 39 19 19 29 *** 29 ns 39 *** *** 39 ns ns
Mass Retention
Unpurified 29 39 Purified 19 29 39 19 19 29 *** 29 *** 39 *** *** 39 *** ***
Initial Water Content
19% Un. Pur. 29% Un. Pur. 39% Un. Pur. Un. Un. Un. Pur. ns Pur. *** Pur. ***
Mass Retention 19% Un. Pur. 29% Un. Pur. 39% Un. Pur. Un. Un. Un. Pur. ** Pur. *** Pur. ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
183
1
week 12.6 15.1 17.5 20.0 2 weeks 12.6 15.1 17.5 20.0
12.6 12.6 15.1 *** 15.1 *** 17.5 * ns 17.5 ** * 20.0 ** ns ns 20.0 *** ns **
3 weeks 12.6 15.1 17.5 20.0 4
weeks 12.6 15.1 17.5 20.0
12.6 12.6 15.1 ns 15.1 ns 17.5 ns ns 17.5 ns ns 20.0 ns ns ns 20.0 ns ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.5: Swelling ratio of purified 12.6 to 20% PVA hydrogel formulation
Swelling Ratio of Purified FormulationsVarying PVA Content
1 week 2 weeks 3 weeks 4 weeks1.0
1.1
1.2
1.3
1.412.6% PVA15.1% PVA17.5% PVA20% PVA
Time, weeks
184
Table 6.3: Initial water content and mass retention of unpurified and purified 12.6 to 20% PVA hydrogel formulations
Unpurified Gels Purified Gels
Gel Initial Water
Content
Mass Retention
Purification Solution
Length of Purification
Initial Water
Content
Mass Retention
[%] [%] [MPa] [Days] [%] [%]
12.6% PVA 49.9 ± 0.1 72.2 ± 0.2 0.65 7 54.6 ± 1.1 91.2 ± 2.1
15.1% PVA 47.8 ± 0.3 71.2 ± 0.6 0.65 7 52.8 ± 0.6 94.4 ± 0.4
17.5% PVA 44.0 ± 0.3 74.8 ± 1.0 0.65 7 52.8 ± 0.4 90.1 ± 1.0
20% PVA 42.5 ± 0.2 78.2 ± 0.6 0.65 7 51.7 ± 0.3 94.7 ± 0.1
Initial Water Content Unpurified 12.6 15.1 17.5 20.0 Purified 12.6 15.1 17.5 20.0
12.6 12.6 15.1 *** 15.1 ns 17.5 *** *** 17.5 ns ns 20.0 *** *** *** 20.0 ns ns ns
Mass Retention Unpurified 12.6 15.1 17.5 20.0 Purified 12.6 15.1 17.5 20.0
12.6 12.6 15.1 ns 15.1 ns 17.5 * ** 17.5 ns * 20.0 *** *** ** 20.0 ns ns **
Initial Water Content 12.6% Un. Pur. 15.1% Un. Pur. 17.5% Un. Pur. 20.0% Un. Pur.
Un. Un. Un. Un. Pur. ** Pur. *** Pur. *** Pur. ***
Mass Retention 19% Un. Pur. 29% Un. Pur. 39% Un. Pur. 39% Un. Pur. Un. Un. Un. Un. Pur. *** Pur. *** Pur. *** Pur. ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
185
20 m 19 29 39 2 h 19 29 39 24 h 19 29 39
19 19 19 29 ns 29 ** 29 *** 39 ns ns 39 *** ns 39 *** **
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.6: Compressive moduli over 24 hours of purified 19 to 39% PEG-DGE hydrogel formulations
Room Temperature Compression of PurifiedFormulations Varying PEG-DGE Content
20 minutes 2 hours 24 hours0.0
0.1
0.2
0.319% PEG-DGE Purified29% PEG-DGE Purified39% PEG-DGE Purified
Time
186
20 m 12.6 15.1 17.5 20.0 2 h 12.6 15.1 17.5 20.0
12.6 12.6 15.1 ns 15.1 *** 17.5 ** ns 17.5 *** ** 20.0 *** *** ns 20.0 *** *** ***
24 h 12.6 15.1 17.5 20.0
12.6 15.1 *** 17.5 *** *** 20.0 *** *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.7: Compressive moduli over 24 hours of purified 12.6 to 20% PVA hydrogel formulations
Room Temperature Compression of PurifiedFormulations Varying PVA Content
20 minutes 2 hours 24 hours0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.712.6% PVA Purified15.1% PVA Purified17.5% PVA Purified20.0% PVA Purified
Time
187
Figure 6.8: FTIR of purified 19 and 39% PEG-DGE hydrogel formulations
188
Figure 6.9: FTIR of purified 12.6 and 20% PVA hydrogel formulations
189
19 29 39
19 29 ns 39 * ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.10: Relative crystallinity for purified formulations with varying PEG-DGE content
Relative Crystallinity of Purified FormulationsVarying PEG-DGE Content
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
190
12.6 15.1 17.5 20.0
12.6 15.1 ns 17.5 ns ns 20.0 ns ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.11: Relative crystallinity for purified formulations with varying PVA content
Relative Crystallinity of Purified FormulationsVarying PVA Content
0.0
1.0
2.0
3.0
191
Table 6.4: Relative Crystallinity for Unpurified and Purified Formulations (p>0.05 is represented by ns)
Hydrogel Formulation
Unpurified Relative Crystallinity
(%)
Purified Relative Crystallinity
(%) P value
19% PEG-DGE 0.65 ± 0.02 1.04 ± 0.29 ns 29% PEG-DGE 2.22 ± 0.19 2.33 ± 0.79 ns 39% PEG-DGE 3.26 ± 0.56 2.58 ± 0.17 ns
12.6% PVA 1.45 ± 0.18 2.17 ± 0.83 ns 15.1% PVA 2.28 ± 0.39 1.91 ± 0.21 ns 17.5% PVA 2.20 ± 0.04 2.20 ± 0.54 ns 20.0% PVA 3.34 ± 0.45 2.13 ± 0.17 ns
192
Figure 6.12: Optical micrographs of purified 19% PEG-DGE and 39% PEG-DGE.
Scale bars indicate 100 µm.
193
19% DGE, 526
19% PEG, 600
39% DGE, 600
19% DGE, 2000
19% DGE, 6000
PBS PE Latex
19% DGE, 526 19% PEG, 600 ns 39% DGE, 600 ns ns
19% DGE, 2000 ns ns ns 19% DGE, 6000 ns ns ns ns
PBS ns ns ns ns ns PE ns ns ns ns ns ns
Latex *** *** *** *** *** *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.13: Cytotoxicity of PVA/PVP/PEG and PVA/PVP/PEG-DGE hydrogel extractions in PBS
Percent Relative Cytotoxicity of PVA/PVP/PEGand PVA/PVP/PEG-DGE Hydrogels
(PBS Extractions)
0102030405060708090
100110120
194
19% DGE, 526
19% PEG, 600
39% DGE, 600
19% DGE, 2000
19% DGE, 6000
Oil PE Latex
19% DGE, 526 19% PEG, 600 ns 39% DGE, 600 ns ns
19% DGE, 2000 ns ns ns 19% DGE, 6000 ns ns * ns
Oil ns ns ns ns ns PE ns ns ns ns ns ns
Latex *** *** *** *** *** *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.14: Cytotoxicity of PVA/PVP/PEG and PVA/PVP/PEG-DGE hydrogel extractions in cottonseed oil
Percent Relative Cytotoxicity of PVA/PVP/PEGand PVA/PVP/PEG-DGE Hydrogels
(Oil Extractions)
0102030405060708090
100110120
195
UV
Sterilized 1.56 3.13 6.25 12.50 25.00 50.00 100.00
1.56 3.13 *** 6.25 *** ns
12.50 *** ns ns 25.00 *** ** ns ns 50.00 *** *** ns ns ns
100.00 *** *** * ns ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.15: Cytotoxicity of PEG-DGE (526 MW), UV sterilized
48h Fibroblast Cell Viability withPEG-DGEs (UV Sterilization)
0.000.010.020.030.040.050.060.070.080.090.100.110.12
PEG-DGE Concentration (526 MW), mg/mL
196
Autoclave Sterilized 1.56 3.13 6.25 12.50 25.00 50.00 100.00
1.56 3.13 *** 6.25 *** ns
12.50 *** * ns 25.00 *** *** *** *** 50.00 *** *** *** *** ns
100.00 *** *** *** *** ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.16: Cytotoxicity of PEG-DGE (526 MW), autoclave sterilized
48h Fibroblast Cell Viability withPEG-DGEs (Autoclave Sterilization)
0102030405060708090
PEG-DGE Concentration (526 MW), mg/mL
197
1.56
mg/mL 526 2000 6000 3.13 mg/mL 526 2000 6000
526 526 2000 * 2000 ** 6000 *** ns 6000 *** * 6.25
mg/mL 526 2000 6000 12.50 mg/mL 526 2000 6000
526 526 2000 ** 2000 *** 6000 *** * 6000 *** ns 25.00
mg/mL 526 2000 6000 50.00 mg/mL 526 2000 6000
526 526 2000 *** 2000 *** 6000 *** ns 6000 *** **
100.00 mg/mL 526 2000 6000
526 2000 * 6000 *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.17: Cytotoxicity of PEG-DGE concentrations with varying molecular weights
Percent Relative Cytotoxicity of PEG-DGEwith Varying Molecular Weights
0
25
50
75
100DGE 526DGE 2000DGE 6000
PEG-DGE Concentration, mg/mL
198
1.56 mg/mL 526 DGE
600 PEG 3.13 mg/mL 526
DGE 600 PEG
526 DGE 526 DGE 600 PEG ns 600 PEG ***
6.25 mg/mL 526 DGE
600 PEG 12.50 mg/mL 526
DGE 600 PEG
526 DGE 526 DGE 600 PEG *** 600 PEG ***
25.00 mg/mL 526 DGE
600 PEG 50.00 mg/mL 526
DGE 600 PEG
526 DGE 526 DGE 600 PEG *** 600 PEG ***
100.00 mg/mL 526 DGE
600 PEG
526 DGE 600 PEG ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 6.18: Cytotoxicity of PEG-OH and PEG-DGE concentrations
Percent Relative Cytotoxicity ofPEG-DGE vs. PEG-OH
0102030405060708090
100110
DGE 526PEG 600
PEG/PEG-DGE Concentration, mg/mL
199
7. CHARACTERIZATION OF THE SWOLLEN PVA/PVP/PEG-DGE
HYDROGEL
7.1. Introduction
Adequate in vitro modeling of the in vivo environment a hydrogel will experience
is essential to avoiding gross failures at the animal study phase of testing a new
biomaterial. As shown by Allen at el. [91] in the baboon study for a PVA hydrogel
nucleus pulposus replacement, where the hydrogels shrunk once implanted in the IVD
resulting in a high rate of extrusion. The osmotic pressure of the baboon disc is higher
than that of humans, which caused the hydrogels to shrink more than expected and
extrude from the IVD.
Macromolecular-based solutions can be used to mimic the swelling pressure of the
intervertebral disc, in addition to measuring the swelling pressure of cadaver tissues [168].
Spiller et al. [132] showed the ability of a macromolecular-based swelling solution to
mimic the swelling behavior seen in an ex vivo articular cartilage model, more accurately
than a PBS swelling model. These macro-molecular based swelling models lack fluid flow
or mechanical loading, so the model is not a perfect portrayal of in vivo behavior.
Swelling ratios for the unpurified and purified hydrogel formulations are shown in
Chapters 5 and 6, but the effect on the structure and properties of the hydrogels as a
result of the swelling and deswelling of the hydrogel networks has not been explored.
Various formulations swell as much as 25 percent and others deswell as much as 32
percent, in addition to water content of the formulations increasing from 3 to 37 percent.
200
In the follow experiments the effect of swelling on the mechanical properties and
crystallization of the unpurified and purified hydrogel formulations are studied.
Specifically, the compressive moduli and relative crystallinity are investigated as a
function of PEG-DGE content, PVA content, PEG functional group and PEG
molecular weight for unpurified and purified formulations.
7.2. Materials and Methods
7.2.1. Materials
PVA (99.0-99.8% hydrolyzed, molecular weight 145 kDa) was purchased from
Aldrich. PVP (molecular weight 58 kDa) was purchased from ISP Technologies. PEG
(molecular weights 600 Da) was obtained from Aldrich and (molecular weight 20 kDa)
was obtained from Crescent Chemical Company. PEG-DGE (molecular weights 526
Da, 2 kDa) was obtained from Aldrich, PEG-DGE (molecular weight 4.6 kDa) was
custom synthesized by Advanced Polymer Materials Inc. Barium sulfate (BaSO4) with a
1-10 μm particle size was purchased from J.T. Baker. Sodium hydroxide (reagent grade, ≥
98%) and sodium chloride (reagent grade, ≥ 99%) were obtained from Sigma-Aldrich.
Dialysis tubing, purchased from Fisher, had a nominal molecular weight cut-off of 3,500
g/mol.
201
7.2.2. Mechanical Properties
Unconstrained, uniaxial compressive modulus was measured to determine if the
hydrogel is a suitable nucleus pulposus replacement material. Testing was conducted
using an Instron Materials Testing System Series 4442 (Norwood, MA) bench-top
mechanical testing system with a 50 N load cell.
Hydrogel samples were synthesized as described in Chapter 5; after the hydrogel
was allowed to react for the specified amount of time, the hydrogel was autoclaved at
121oC for 30 minutes. For unpurified specimens: after the autoclave cycle the hydrogel
samples were loaded into a 60 cc syringe and injected into poly(vinyl chloride) tubing
with an inner diameter of 9.5 mm. For purified specimens: after the autoclave cycle the
hydrogel was loaded into a 60 cc syringe and injected into 15 mL centrifuge tubes. After
24 hours the samples were removed from the centrifuge tubes, placed in 3500 MWCO
dialysis tubing and closed at each end with dialysis bag closures. Each dialysis tubing
assembly was placed in 0.65 MPa osmotic solution for 7 days at 37oC; samples were then
removed from the solutions and dialysis tubing, placed in media bottles and autoclaved at
121oC for 30 minutes. As with the unpurified samples, after the autoclave cycle the
hydrogel samples were loaded into a 60 cc syringe and injected into poly(vinyl chloride)
tubing with an inner diameter of 9.5 mm
The hydrogel was then removed from the tubing in sections 20 cm in length,
placed in dialysis tubing and swollen for up to four weeks at 37oC. Samples were swollen
in osmotic PEG solutions with an osmotic pressure of 0.2 MPa. Sample time points
202
were: 7, 14, and 28 days. At each time point, cylindrical hydrogel samples were sliced (n =
5), to a thickness of 7-8mm and cut to ensure flat surfaces. The samples were preloaded
to 0.1 N and compressed to a total strain of 30% at a strain rate of 100% min-1, a chord
from the initial linear portion of the stress versus strain curve (10-20% strain) was used to
calculate elastic modulus. At least five independent samples were tested for each set of
hydrogels (n = 5).
7.2.3. X-ray Diffraction
WAXD diffraction patterns were collected using a Rigaku D/max-B Powder X-
ray Diffractometer (Tokyo, Japan) with an operating tube voltage and current of 40 kV
and 30 mA. Spectrum were collected at room temperature using a 0.04o 2θ/s scan rate
with a four second dwell time, using a 2θ range of 8o to 35o. In order to prevent the
drying the sample during testing, scan rate and dwell time were chosen to minimize data
collection time.
WAXD was performed on 19% PEG-DGE (526, 2000, and 4600 Da), 19%
PEG-OH, 29% PEG-DGE, 39% PEG-DGE, 12.6% PVA, 15.1% PVA, 17.5% PVA
and 20.0% PVA hydrogel formulations unpurified and purified at 0.65 MPa for 7 days
after four weeks of swelling in 0.2 MPa osmotic solution. Diffraction patterns were
smoothed using Savitzky-Golay's smoothing filter. Relative crystallinity was calculated
using equation 4.2:
203
Relative Crystallinity % = PVA crystalline peak areatotal area
× 100% (4.7)
where PVA crystalline peak area is the area at 2θ =19.4° and total area is the entire area
within the 2θ range of 8°-35°. This relative value of crystallinity can be compared to other
samples using the same testing method; however, it does not represent an exact value for
crystallinity since the equation is not normalized using known crystallinity values.
Hydrogel samples were synthesized as described in Chapter 5; after the hydrogel
was allowed to react for the specified amount of time, the hydrogel was autoclaved at
121oC for 30 minutes. For unpurified specimens: after the autoclave cycle the hydrogel
samples were loaded into a 60 cc syringe and injected onto polyethylene terephthalate
copolymer with cyclohexylene dimethylene segments (PETG) sheeting with spacers of
0.1 mm thickness to control thickness of the film. For purified specimens: after the
autoclave cycle the hydrogel was loaded into a 60 cc syringe and injected into 15 mL
centrifuge tubes. After 24 hours the samples were removed from the centrifuge tubes,
placed in 3500 MWCO dialysis tubing and closed at each end with dialysis bag closures.
Each dialysis tubing assembly was placed in 0.65 MPa osmotic solution for 7 days at
37oC; samples were then removed from the solutions and dialysis tubing, placed in media
bottles and autoclaved at 121oC for 30 minutes. As with the unpurified samples, after the
autoclave cycle the hydrogel samples were loaded into a 60 cc syringe and injected onto
polyethylene terephthalate copolymer with cyclohexylene dimethylene segments (PETG)
sheeting with spacers of 0.1 mm thickness to control thickness of the film
204
The hydrogel was then removed from the sheeting and cut into sections one to
one and a half inches square, each section was placed in dialysis tubing and swollen for
four weeks at 37oC. Samples were swollen in osmotic PEG solutions with an osmotic
pressure of 0.2 MPa. After 4 weeks, each sample was removed from the osmotic solution
and dialysis tubing directly before testing to preventing drying of the sample.
7.2.4. Statistical Analysis
All data points are represented as the mean ± one standard deviation for at least
three independent samples. Statistical significance was determined by one-way analysis of
variance (ANOVA) with post-hoc analysis by Bonferroni correction with a 95%
confidence interval. P-values less than 0.05 were considered statistically significant.
7.3. Results and Discussion
7.3.1. Mechanical Properties
A sample stress versus strain plot for the swollen compression is shown in Figure
7.1. The compressive moduli for hydrogel formulations swollen in 0.2 MPa osmotic
solution at 37oC for up to four weeks with increasing PEG-DGE content are shown in
Figure 7.2. As PEG-DGE content is increases from 19 to 39%, at each time point, the
compressive modulus increases (p<0.001). Figure 7.3 shows the compressive moduli for
hydrogel formulations swollen in 0.2 MPa osmotic solution at 37oC for up to four weeks
205
with increasing PVA content. At one week, compressive moduli increases from 12.6 to
20.0% PVA; but at four weeks, there is no increase from 12.6 to 15.1% (p>0.05), or 15.1
to 17.5% (p>0.05), but moduli does increase from 12.6, 15.1 and 17.5 to 20.0% PVA
(p<0.001).
Figure 7.4 shows the compressive moduli for hydrogel formulations swollen in 0.2
MPa osmotic solution at 37oC for up to four weeks with varying PEG functional group.
There is no significant difference between the compressive moduli for PEG-DGE and
PEG-OH (p>0.05) at one week; but at two and four weeks the compressive moduli is
higher for PEG-DGE than for PEG-OH (p<0.001). The compressive moduli for
hydrogel formulations swollen in 0.2 MPa osmotic solution at 37oC for up to four weeks
with varying PEG-DGE molecular weight is shown Figure 7.5. At one week,
compressive moduli increase with molecular weight from 526 to 2000 Da (p<0.05) and
from 2000 to 4600 Da (p<0.001). At four weeks, compressive moduli continue to
increase with molecular weight (p>0.001). The only hydrogel with a modulus below 50
kPa, the minimum modulus necessary to restore healthy tension in the annulus fibers [84,
85], in the model osmotic solution is the 19% PEG-OH at each time point and the 19%
PEG-DGE at one week.
The compressive moduli for purified hydrogel formulations swollen in 0.2 MPa
osmotic solution at 37oC for up to four weeks with increasing PEG-DGE content are
shown in Figure 7.6 and for increasing PVA content in Figure 7.7. At one week,
compressive modulus increased with increased PEG-DGE loading (p<0.01); and at four
206
weeks, compressive modulus increases from 19 to 29% PEG-DGE (p<0.05) and from 29
to 39% PEG-DGE (p<0.001). At each time point, modulus increases with PVA content.
7.3.2. X-ray Diffraction
The relative crystallinity for hydrogel formulations swollen in 0.2 MPa osmotic
solution at 37oC for four weeks with increasing PEG-DGE content are shown in Figure
7.8; and with increasing PVA content are shown in Figure 7.9. There is no significant
difference in relative crystallinity as PEG-DGE or PVA content increases.
The relative crystallinity for hydrogel formulations swollen in 0.2 MPa osmotic
solution at 37oC for four weeks with varying PEG functional group is shown in Figure
7.10. As with the un-swollen samples shown in Chapter 5, there is no significant
difference in relative crystallinity between the hydroxyl and epoxide terminated PEG
(p>0.05).
Figure 7.11 shows the relative crystallinity for 19% PEG-DGE hydrogel
formulations swollen in 0.2 MPa osmotic solution at 37oC for four weeks with varying
PEG-DGE molecular weight (526 and 2000 Da). Unlike the unswollen formulations in
Chapter 5, there is no difference in relative crystallinity with varying molecular weight
(p>0.05).
The relative crystallinity of purified hydrogel formulations swollen in 0.2 MPa
osmotic solution at 37oC for four weeks with varying PEG-DGE content is shown in
207
Figure 7.12; varying PVA content is shown in Figure 7.13. As with the unpurified
samples shown in Figure 7.8 and Figure 7.9, there is no significant difference in relative
crystallinity (p>0.05) as PEG-DGE or PVA content is increased.
7.3.3. Discussion
As PVA hydrogels are swollen in PBS and macromolecular osmotic solutions we
expect for the compressive modulus to decrease as the water content of the gels increases,
though this is not the case if the hydrogels deswell in PBS and macromolecular osmotic
solutions [130]. PVA hydrogels that contain a small amount of PVP (99:1 PVA to PVP
ratio) do not experience a large amount of mass loss during swelling [178], but as
discussed previously in Chapter 5 the PVA/PEG-DGE hydrogels do experience a large
mass loss of unbound PEG-DGE while swelling in a macromolecular osmotic solution
with a pressure of 0.2 MPa. This loss of polymer content in the hydrogels and the
increase in water content will decrease the compressive modulus [130]. It is expected that
the unpurified hydrogel formulations will have a lower modulus after swelling in
macromolecular osmotic solution with a pressure of 0.2 MPa. The purified hydrogels will
not experience a large mass loss and only increase in water content depending on the
formulation after swelling in macromolecular osmotic solution with a pressure of 0.2
MPa, the effect of swelling and water uptake on the compressive modulus of the purified
gels should be minimal.
208
In addition to swelling and mass loss affecting the mechanical properties of PVA
hydrogel, PVA hydrogels also experience an aging phenomenon that affects the
mechanical properties and crystallinity of the hydrogels. Researchers have documented
this phenomena of aging, for freeze/thawed hydrogels in solution [106, 130] and in
sealed containers not in solution [104, 107, 130, 229, 230]. Temperature and swelling
medium have been observed to have an effect on the aging rate of the PVA [106]. For
samples in sealed containers, increases in crystallization were measured for periods of up
to 500 days [107] and for up to 120 days in solution [106]. In addition to aging of
freeze/thawed PVA hydrogel, aging has been observed to occur in PVA solutions without
any initial crosslinks [231-233]. This aging effect of PVA is an inner structural change
which can either be from slowly occurring phase separation of the network and/or
crystallization, as the gel when originally formed is not in its equilibrium state [106, 107].
The increase in crystallinity (and phase separation) results in an increase in mechanical
properties of the hydrogel network due to the formation of physical crosslinks allowing
the PVA hydrogels to carry increased load [104, 197, 201].
The compressive moduli for the unpurified PEG-DGE and PVA hydrogel
formulations before and after swelling are shown in Table 7.1. The compressive moduli
for the unpurified formulations at 24 hours are higher than the compressive moduli after
two and four weeks swollen in 0.2 MPa. This is due to the mass loss of the unpurified
hydrogel formulations, shown in Chapter 5. At four weeks, between 21 and 35 percent of
the original mass has been lost from the hydrogel network but the volume of the network
is mostly unchanged (19% PEG-DGE experiences some significant deswelling (V/Vo <
209
0.9), 20.0% PVA experiences some significant swelling (V/Vo > 1.1)). This suggests that
the loss of polymer (99.93% PEG-DGE, 0.07 % PVA as determined by 1H NMR) is
replaced in volume with water (Table 7.2), lowering the compressive modulus of the
hydrogel network. The compressive moduli for 19% PEG-DGE (526, 2000 and 4600
Da) and 19% PEG-OH hydrogel formulations before and after swelling are also shown
in Table 7.1. A similar trend is seen in these hydrogels as in the formulations varying
PEG-DGE and PVA content; the compressive modulus at 24 hours is higher than the
compressive moduli after two and four weeks swollen in 0.2 MPa. For the low molecular
weight PEG-DGE (526 Da) and the PEG-OH (600 Da), the issue is two fold as
previously discussed. The hydrogels are losing a large amount of PEG, which is being
replaced with water lowering the compressive modulus. The 2000 and 4600 Da
formulations do not have the mass loss issue that the lower molecular weight PEG
formulations do; their mass loss is between 6 and 8 percent. These formulations do
experience a significant amount of swelling, so instead of losing PEG-DGE and
replacing it with water, these formulations just add additional volume with an influx of
water (Table 7.2). This influx of additional water is due to the hydrophilic nature of PEG,
as PEG molecular weight increases the length of the hydrophilic backbone increases
resulting in increased swelling of the polymer. For the unpurified hydrogels in addition to
the modulus and water content of the hydrogel networks changing with swelling, the
crystallinity of the unpurified formulations increases from 24 hours at room temperature
to four weeks at 0.2 MPa (Table 7.3). This is due to the aging phenomena of PVA.
210
The purified PEG-DGE and PVA formulations have higher compressive moduli
at four weeks swelling in osmotic solution than at 24 hours at room temperature (Table
7.4). These hydrogels have already had the majority of the unbound PEG eluted from the
network and a higher water content as a result; the water content is not as high as the
unpurified samples swollen in 0.2 MPa for 4 weeks since the purification was at 0.65
MPa for 7 days. The large reduction of unbound PEG and increased water result in
lower compressive moduli than the unpurified formulations (Table 7.5). These
formulations still lose a small amount of mass, ranging from 5 to 18 percent, with most
values around 5 to 11 percent. This mass is replaced with water, as the water content for
these formulations starts at 52 to 58 percent and increases to 58 to 62 after 4 weeks. The
increase in compressive moduli after swelling is due to the aging phenomena of PVA.
This phenomena has been well documented and it can either be a result of crystallinity or
slowly occurring phase separation of the network and/or crystallization, as the gel when
originally formed is not in its equilibrium state [106, 107]. In this case, the swelling of
the hydrogel and then deswelling as it reaches equilibrium allowed for reorganization of
the polymer [132]. Larger polymer rich regions increase the compressive modulus of the
network.
Table 7.6 shows the compressive modulus and relative crystallinity of the purified
formulations at 24 hours at room temperature and at four weeks swollen in 0.2 MPa
osmotic solution. The relative crystallinity for 19% PEG-DGE and all the PVA
formulations did not significantly change between before and after swelling. As discussed
in Chapter 6, the gelation mechanism of the hydrogels changes after purification. The
211
lack of unbound PEG-DGE, changes the gelation mechanism, resulting in less dense
PVA rich regions. These less dense regions are less likely to hydrogen bond and
crystallize than the dense regions in the unpurified formulations. The reorganization as
part of syneresis, phase separates the hydrogel but does not pull water out of the system as
successfully as the PEG-DGE theta-gelation, so the crystallinity of the hydrogel does not
significantly increase for these purified, swollen hydrogels. The water introduced to the
hydrogel network in the purification procedure hydrogen bonds to the available hydroxyl
group of the PVA, preventing it from hydrogen bonding to another pendant PVA
hydroxyl group in order for the polymer chains to crystallize.
7.4. Conclusions
Unpurified formulations varying PEG-DGE content from 19 to 39%, PVA
content from 12.6 to 20.0%, PEG functional group and PEG molecular weight from 526
to 4600 g/mol were swollen in 0.2 MPa osmotic solution, the midpoint of the range (0 to
0.4 MPa) measured by Urban et al. for cadaver IVDs [168], for up to four weeks at 37oC.
Compressive modulus for the formulations decreased from the room temperature values
after four weeks of swelling at 0.2 MPa due to the mass loss of the unpurified hydrogel
formulations. Suggesting that the loss of polymer is replaced in volume with water
lowering the compressive modulus of the network, or in the case of the higher molecular
weight PEG-DGE an increase in water content in the network due to swelling. The
crystallinity of these formulations also increases from 24 hours at room temperature to
four weeks at 0.2 MPa due to the aging phenomena of PVA.
212
Hydrogel formulations varying PEG-DGE content from 19 to 39% and PVA
content from 12.6 to 20.0% were purified in 0.65 MPa osmotic solution for seven days
and then swollen in 0.2 MPa osmotic solution, the midpoint of the range (0 to 0.4 MPa)
measured by Urban et al. for cadaver IVDs [168], for up to four weeks at 37oC. The
purified PEG-DGE and PVA formulations have higher compressive moduli at four
weeks swelling in osmotic solution than at 24 hours at room temperature. This is due to
the aging phenomena of PVA. The crystallinity of the purified hydrogels does not change
from the RT values. This is due to the change in the gelation mechanism of the purified
hydrogels; the PVA rich regions are less dense. Resulting in additional water being able
to hydrogen bond to the PVA when swollen preventing additional crystallinity.
213
Figure 7.1: Stress versus strain plots for 19% and 39% PEG-DGE hydrogel formulations
19% PEG-DGE Hydrogel FormulationStress vs. Strain
0.0 0.1 0.2 0.3 0.40.000
0.025
0.050
0.0751 week2 week4 week
Strain
39% PEG-DGE Hydrogel FormulationStress vs. Strain
0.0 0.1 0.2 0.3 0.40.0
0.1
0.2
0.3
0.41 week2 week4 week
Strain
214
1 week 19 29 39 2 weeks 19 29 39 4 weeks 19 29 39
19 19 19 29 *** 29 *** 29 *** 39 *** *** 39 *** *** 39 *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 7.2: Compressive moduli of 19 to 39% PEG-DGE hydrogel formulations swollen in 0.2 MPa at 37oC for up to four weeks
Compressive Moduli VaryingPEG-DGE Content
1 week 2 weeks 4 weeks0.0
0.5
1.0
1.519% PEG-DGE29% PEG-DGE39% PEG-DGE
Time, weeks
215
1 week 12.6 15.1 17.5 20.0 2 weeks 12.6 15.1 17.5 20.0
12.6 12.6 15.1 *** 15.1 ns 17.5 *** *** 17.5 *** ns 20.0 *** *** *** 20.0 *** *** ***
4 weeks 12.6 15.1 17.5 20.0
12.6 15.1 ns 17.5 *** ns 20.0 *** *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 7.3: Compressive moduli of 12.6 to 20.0% PVA hydrogel formulations swollen in 0.2 MPa at 37oC for up to four weeks
Compressive Moduli VaryingPVA Content
1 week 2 weeks 4 weeks0.0
0.5
1.0
1.5
2.0
2.512.6% PVA15.1% PVA17.5% PVA20.0% PVA
Time, weeks
216
1
week DGE OH 2 weeks DGE OH 4
weeks DGE OH
DGE DGE DGE OH ns OH *** OH ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 7.4: Compressive moduli of 19% PEG-DGE and 19% PEG-OH hydrogel formulations swollen in 0.2 MPa at 37oC for up to four weeks
Compressive Moduli VaryingPEG Functional Group
1 week 2 weeks 4 weeks0.0
0.1
0.2
0.319% PEG-DGE19% PEG-OH
Time, weeks
217
1
week 526 2000 4600 2 weeks 526 2000 4600 4
weeks 526 2000 4600
526 526 526 2000 * 2000 ns 2000 *** 4600 *** *** 4600 *** *** 4600 *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 7.5: Compressive moduli of 19% PEG-DGE with varying molecular weight (526, 2000 and 4600 Da) swollen in 0.2 MPa at 37oC for up to four weeks
Compressive Moduli VaryingPEG-DGE Molecular Weight
1 week 2 weeks 4 weeks0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7526 PEG-DGE2000 PEG-DGE4600 PEG-DGE
Time, weeks
218
1 week 19 29 39 2 weeks 19 29 39 4 weeks 19 29 39
19 19 19 29 ** 29 ns 29 * 39 *** ** 39 *** ** 39 *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 7.6: Compressive moduli of purified 19 to 39% PEG-DGE hydrogel formulations swollen in 0.2 MPa at 37oC for up to four weeks
Compressive Moduli of Purified FormulationsVarying PEG-DGE Content
1 week 2 weeks 4 weeks0.000.050.100.150.200.250.300.350.400.45
19% PEG-DGE29% PEG-DGE39% PEG-DGE
Time, weeks
219
1 week 12.6 15.1 17.5 20.0 2 weeks 12.6 15.1 17.5 20.0
12.6 12.6 15.1 * 15.1 ** 17.5 *** *** 17.5 *** * 20.0 *** *** *** 20.0 *** *** ***
4 weeks 12.6 15.1 17.5 20.0
12.6 15.1 *** 17.5 *** ns 20.0 *** *** ***
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 7.7: Compressive moduli of purified 12.6 to 20.0% PVA hydrogel formulations swollen in 0.2 MPa at 37oC for up to four weeks
Compressive Moduli of Purified FormulationsVarying PVA Content
1 week 2 weeks 4 weeks0.00.10.20.30.40.50.60.70.80.9
12.6% PVA15.1% PVA17.5% PVA20% PVA
Time, weeks
220
19 29 39
19 29 ns 39 ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 7.8: Relative crystallinity for formulations with varying PEG-DGE content swollen at 0.2 MPa and 37oC for four weeks
Relative Crystallinity Varying PEG-DGEContent, Swollen at 0.2 MPa for 4 Weeks
0.0
1.0
2.0
3.0
4.0
221
12.6 15.1 17.5 20.0
12.6 15.1 ns 17.5 ns ns 20.0 ns ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 7.9: Relative crystallinity for formulations with varying PVA content swollen at 0.2 MPa and 37oC for four weeks
Relative Crystallinity Varying PVA Content,Swollen at 0.2 MPa for 4 Weeks
0.0
1.0
2.0
3.0
4.0
5.0
6.0
222
Figure 7.10: Relative crystallinity for formulations varying PEG functional group
swollen at 0.2 MPa and 37oC for four weeks (p>0.05)
Relative Crystallinity Varying PEG FunctionalGroup, Swollen at 0.2 MPa for 4 Weeks
0.0
1.0
2.0
3.0
4.0
5.0
6.0
223
Figure 7.11: Relative crystallinity for formulations varying PEG-DGE molecular weight swollen at 0.2 MPa and 37oC for four weeks (p>0.05)
Relative Crystallinity Varying PEG-DGE MolecularWeight, Swollen at 0.2 MPa for 4 Weeks
0.0
1.0
2.0
3.0
4.0
224
19 29 39
19 29 ns 39 ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 7.12: Relative crystallinity for purified formulations with varying PEG-DGE content swollen at 0.2 MPa and 37oC for four weeks
Relative Crystallinity of Purified FormulationsVarying PEG-DGE Content,
Swollen at 0.2 MPa for 4 Weeks
0.0
0.5
1.0
1.5
225
12.6 15.1 17.5 20.0
12.6 15.1 ns 17.5 ns ns 20.0 ns ns ns
P value p>0.05 p<0.05 p<0.01 p<0.001 Summary ns * ** ***
Figure 7.13: Relative crystallinity for purified formulations with varying PVA content swollen at 0.2 MPa and 37oC for four weeks
Relative Crystallinity of Purified FormulationsVarying PVA Content,
Swollen at 0.2 MPa for 4 Weeks
0.0
1.0
2.0
3.0
226
Table 7.1: Compressive modulus for unpurified PEG-DGE and PVA formulations before and after swelling in osmotic solution
Hydrogel Formulation
Compressive Modulus (MPa)
Before Swelling After Swelling in Osmotic Solution (0.2 MPa)
2 hours 24 hours 2 weeks 4 weeks
19% PEG-DGE 0.03 ± 0.00 0.31 ± 0.01 0.11 ± 0.01* 0.25 ± 0.02*
29% PEG-DGE 0.17 ± 0.01 1.28 ± 0.03 0.57 ± 0.03 0.86 ± 0.06
39% PEG-DGE 0.29 ± 0.04 1.65 ± 0.05 0.94 ± 0.08 1.19 ± 0.04
12.6% PVA 0.08 ± 0.00 0.88 ± 0.03 0.39 ± 0.05 0.55 ± 0.09
15.1% PVA 0.20 ± 0.02 1.43 ± 0.06 0.72 ± 0.11 0.92 ± 0.03
17.5% PVA 0.33 ± 0.01 1.59 ± 0.04 1.05 ± 0.06 1.28 ± 0.14
20.0% PVA 0.55 ± 0.02 2.29 ± 0.04 1.90 ± 0.39 1.98 ± 0.35**
19% PEG-OH
600 Da 0.02 ± 0.00 0.17 ± 0.01 0.01 ± 0.00 0.02 ± 0.00
19% PEG-DGE
2000 Da 0.29 ± 0.01 1.01 ± 0.03 0.13 ± 0.02** 0.46 ± 0.04
19% PEG-DGE 4600 Da 0.38 ± 0.01 1.28 ± 0.03 0.30 ± 0.04** 0.64 ± 0.02**
* Deswelling, swelling ratio <0.9 ** Swelling, swelling ratio >1.2
227
Table 7.2: Water content for unpurified PEG-DGE and PVA formulations before and after swelling in osmotic solution
Hydrogel Formulation
Before Swelling After Swelling in Osmotic Solution (0.2 MPa)
Initial Water Content Final Water Content
19% PEG-DGE 58.60 ± 0.10 61.70 ± 0.20*
29% PEG-DGE 49.40 ± 0.30 60.10 ± 0.40
39% PEG-DGE 44.50 ± 1.30 55.20 ± 0.40
12.6% PVA 49.90 ± 0.10 59.20 ± 0.50
15.1% PVA 47.80 ± 0.30 58.60 ± 0.40
17.5% PVA 44.00 ± 0.30 58.70 ± 1.10
20.0% PVA 42.50 ± 0.20 58.30 ± 0.20**
19% PEG-OH
600 Da 58.80 ± 0.60 61.00 ± 3.00
19% PEG-DGE
2000 Da 54.90 ± 0.90 59.50 ± 0.40
19% PEG-DGE 4600 Da 52.30 ± 2.30 62.30 ± 0.90
228
Table 7.3: Compressive modulus and relative crystallinity for unpurified PEG-DGE and PVA formulations before and after swelling in osmotic solution
Hydrogel Formulation
Before Swelling After Swelling in Osmotic Solution (0.2 MPa)
Compressive Modulus
(MPa)
Relative Crystallinity
(%)
Compressive Modulus
(MPa)
Relative Crystallinity
(%)
19% PEG-DGE 0.31 ± 0.01 0.65 ± 0.02 0.25 ± 0.02* 3.18 ± 0.65*
29% PEG-DGE 1.28 ± 0.03 2.22 ± 0.19 0.86 ± 0.06 3.13 ± 0.43
39% PEG-DGE 1.65 ± 0.05 3.26 ± 0.56 1.19 ± 0.04 3.32 ± 0.32
12.6% PVA 0.88 ± 0.03 1.45 ± 0.18 0.55 ± 0.09 3.48 ± 0.47
15.1% PVA 1.43 ± 0.06 2.28 ± 0.39 0.92 ± 0.03 3.57 ± 0.26
17.5% PVA 1.59 ± 0.04 2.20 ± 0.04 1.28 ± 0.14 4.41 ± 0.96
20.0% PVA 2.29 ± 0.04 3.34 ± 0.45 1.98 ± 0.35** 3.67 ± 0.68**
19% PEG-OH
600 Da 0.17 ± 0.01 1.24 ± 0.48 0.02 ± 0.00 4.72 ± 0.47
19% PEG-DGE
2000 Da 1.01 ± 0.03 2.11 ± 0.06 0.46 ± 0.04 2.14 ± 0.58
19% PEG-DGE 4600 Da 1.28 ± 0.03 - 0.64 ± 0.02** -
* Deswelling, swelling ratio <0.9 ** Swelling, swelling ratio >1.2
229
Table 7.4: Compressive modulus for purified PEG-DGE and PVA formulations before and after swelling in osmotic solution
Hydrogel Formulation
Compressive Modulus (MPa)
Before Swelling After Swelling in Osmotic Solution (0.2 MPa)
2 hours 24 hours 2 weeks 4 weeks 19% PEG-DGE
Purified 0.01 ± 0.00 0.05 ± 0.01 0.05 ± 0.00** 0.19 ± 0.01
29% PEG-DGE Purified 0.05 ± 0.01 0.18 ± 0.01 0.08 ± 0.01** 0.26 ± 0.01
39% PEG-DGE Purified 0.08 ± 0.01 0.25 ± 0.02 0.13 ± 0.00** 0.40 ± 0.01
12.6% PVA Purified 0.01 ± 0.00 0.08 ± 0.01 0.13 ± 0.00 0.28 ± 0.00
15.1% PVA Purified 0.09 ± 0.01 0.32 ± 0.01 0.18 ± 0.03** 0.43 ± 0.03
17.5% PVA Purified 0.13 ± 0.01 0.45 ± 0.03 0.20 ± 0.01** 0.42 ± 0.04
20.0% PVA Purified 0.17 ± 0.00 0.62 ± 0.05 0.39 ± 0.02** 0.77 ± 0.05**
* Deswelling, swelling ratio <0.9 ** Swelling, swelling ratio >1.2
230
Table 7.5: Water content for purified PEG-DGE and PVA formulations before and after swelling in osmotic solution
Hydrogel Formulation
Before Swelling After Swelling in Osmotic Solution
(0.2 MPa) Initial Water
Content Final Water
Content 19% PEG-DGE
Purified 57.90 ± 0.33 61.73 ± 1.34
29% PEG-DGE Purified 54.15 ± 0.40 59.00 ± 0.84
39% PEG-DGE Purified 52.50 ± 0.73 58.98 ± 0.90
12.6% PVA Purified 54.60 ± 1.08 60.64 ± 0.76
15.1% PVA Purified 52.84 ± 0.57 58.31 ± 0.55
17.5% PVA Purified 52.83 ± 0.35 57.81 ± 1.08
20.0% PVA Purified 51.75 ± 0.27 57.54 ± 2.56**
231
Table 7.6: Compressive modulus and relative crystallinity for purified PEG-DGE and PVA formulations before and after swelling in osmotic solution
Hydrogel Formulation
Before Swelling After Swelling in Osmotic Solution (0.2 MPa)
Compressive Modulus
(MPa)
Relative Crystallinity
(%)
Compressive Modulus
(MPa)
Relative Crystallinity
(%) 19% PEG-DGE
Purified 0.05 ± 0.01 1.04 ± 0.29 0.19 ± 0.01 1.13 ± 0.23
29% PEG-DGE Purified 0.18 ± 0.01 2.33 ± 0.79 0.26 ± 0.01 1.04 ± 0.07
39% PEG-DGE Purified 0.25 ± 0.02 2.58 ± 0.17 0.40 ± 0.01 1.01 ± 0.14
12.6% PVA Purified 0.08 ± 0.01 2.17 ± 0.83 0.28 ± 0.00 1.51 ± 0.4
15.1% PVA Purified 0.32 ± 0.01 1.91 ± 0.21 0.43 ± 0.03 1.53 ± 0.51
17.5% PVA Purified 0.45 ± 0.03 2.20 ± 0.54 0.42 ± 0.04 2.21 ± 0.70
20.0% PVA Purified 0.62 ± 0.05 2.13 ± 0.17 0.77 ± 0.05** 1.90 ± 0.49**
* Deswelling, swelling ratio <0.9 ** Swelling, swelling ratio >1.2
232
8. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE WORK
8.1. Conclusions
The long-term objective of the project was to develop an injectable, chemically
crosslinked PVA hydrogel for nucleus pulposus replacement, which upon implantation,
could prevent or postpone the annular degeneration process by restoring the healthy
biomechanics of the intervertebral disc. Two chemical crosslinking methods were
investigated, resulting in the down selection of the use of a difunctional poly(ethylene
glycol) as the preferred crosslinking technique due to mechanical property and
injectability limitations of the electron beam crosslinked hydrogel system. The
PVA/PVP/PEG-DGE hydrogel system was developed and characterized. This
PVA/PVP/PEG-DGE hydrodgel system exhibited the swelling, mechanical and
injectability properties necessary for the nucleus pulposus replacement application; but
the hydrogel system had an issue with mass loss. A purification procedure was developed
and optimized to maintain the swelling, mechanical and injectability of the hydrogel
system. The cytotoxicity of the PVA/PVP/PEG-DGE hydrogel system was also
evaluated.
Initially two methods of chemical crosslinking were investigated: radiation
crosslinking and difunctional crosslinking. For the radiation crosslinked hydrogels, PVA
content and irradiation dosage for electron beam radiation were varied to determine if the
swelling, mechanical and injectability requirements could be met with this crosslinking
method. The modulus values were below 50 kPa, , the minimum modulus necessary to
233
restore healthy tension in the annulus fibers [84, 85] removing this from consideration as
a nucleus replacement material. The difunctional crosslinked system was investigated by
varying the basic catalyst volume and reaction time to determine that the necessary
mechanical, swelling and injectability properties could be achieved with this method of
chemical crosslinking.
After determining the desired swelling, mechanical and injectability properties
could be achieved with the chemical crosslinking of PVA with PEG-DGE; PVA content,
PEG-DGE content, PEG functional group and PEG-DGE molecular weight were
varied to measure the range of properties for this family of hydrogels. The structure-
property relationships present in this family of hydrogels was investigated using FTIR, x-
ray diffraction and optical microscopy.
The initial compression studies suggested the improved mechanical properties
with the addition of PEG-DGE was due to increased crosslinking; the following swelling
mechanics studies suggested that the increase in swelling behavior with PEG-DGE
content was just due to increased polymer content of the hydrogel. The additional FTIR,
XRD and optical microscopy studies provided the additional information to determine
that the PEG-DGE not only chemically crosslinks the hydrogel network but the use of
PEG-DGE with small amounts of NaOH (100 µL per 75 gram batch) alters the gelation
mechanism from the use of PEG-OH.
It was shown that the PVA/PEG-DGE hydrogel are formed almost
instantaneously upon the addition of PEG-DGE and NaOH to the PVA/PVP solution,
234
resulting in the immediate formation of a water/PEG-DGE rich supernatant phase and a
PVA rich hydrogel, which depending on the concentration and molecular weight of the
PEG-DGE had varying size pores of water/PEG-DGE trapped within the polymer rich
PVA hydrogel. The resultant hydrogels had increased crystallinity with increased PEG-
DGE and PVA content due to the ability of the PVA chains to be brought into closer
proximity to promote hydrogen bonding and crystallization. All of the formulations
tested in this chapter were easily injectable through a 10-gauge needle (2.6 mm ID) using
a 60 cc syringe.
The largest obstacle with this family of hydrogels is the mass retention issue for
formulations made with the 526 Da PEG-DGE, the values range from 64.6 ± 0.5% for
the 19% PEG-DGE formulation to 78.2 ± 0.6% for the 20.0% PVA formulation. The
only formulations that did not have gross mass retention issues were the 2000 Da and
4600 Da formulations which had mass retention values of 54.9 ± 0.9% and 52.3 ± 2.3%,
respectively. As previously stated, these molecular weights were not commercially
available during the time period the majority of this work was done.
Using the 39% PEG-DGE formulation, a purification procedure was optimized
that allowed minimal initial water content change but increased the mass retention values
of the hydrogel. The selected purification procedure, of 0.65 MPa for 7 days at 37oC,
allowed for a hydrogel that was still injectable after the removal of the unbound PEG.
The unbound PEG worked as a plasticizer for the hydrogel network [225, 226] the
removal of the unbound PEG could have reduced the ability of the hydrogel to be
injected, water also works as a plasticizer in this hydrogel system[227, 228].
235
After purification at 0.65 MPa for 7 days at 37oC, the 19, 29, and39% PEG-DGE
and 12.6, 15.1, 17.5, and 20.0% PVA hydrogel formulations had increased swelling ratios,
decreased compressive modulus, and no significant change in crystallinity from their
unpurified formulations. This is due to a change in the gelation mechanism for the gels
after purification. When the hydrogels are originally gelled, the PEG-DGE removes
water from the PVA/PVP/water solution causing the solution to phase separate into
water/PEG-DGE regions and PVA regions, which almost instantly gel upon the
addition of PEG-DGE. The PEG-DGE/water phase either separates from the PVA gel
as a supernatant, which is decanted from the gel, or the PEG-DGE/water phase
separates into PEG-DGE and water filled pores. These PEG-DGE water filled pores
result in the low mass retention of the initially formed PVA/PVP/PEG-DGE hydrogels,
as the PEG-DGE is a low molecular weight that easily elutes from the hydrogel network
in an aqueous environment. The purification process removes this unbound PEG-DGE
from the PEG-DGE/water pores, which results in a higher mass retention for the system,
but it also changes the gelation of the hydrogel network as it is heated above 80oC for
injection and then cooled to RT. If the unbound PEG-DGE was still present in the
system, it allows for the formation of additional supernatant and PEG-DGE/water pores
causing more concentrated PVA rich regions, which promote hydrogen bonding and
crystallinity. These PVA rich regions resist swelling and increase mechanical properties of
the hydrogel. After purification, the initial water content in the hydrogel is higher due to
the purification process in addition to the lack of unbound PEG-DGE pulling water out
of the PVA and phase separating into supernatant and PEG-DGE/water pores. This
lack of densification of the PVA rich regions results in lower mechanical properties and
236
increased swelling of the hydrogel.
Cytotoxicity of the hydrogel formulations was determined using a cell viability assay.
The cytotoxicity of the extractions from the formulations in PBS and cottonseed oil
demonstrated no difference in cell viability from polyethylene, a commonly used
biomaterial, and an increase in cell viability from latex, a known cytotoxic agent. This was
extremely important to demonstrate due to the cytotoxic concerns of short chain
diepoxide compounds [149]. The repeated autoclave cycles used in the synthesis of the
hydrogel is thought to be the reason for the low cytotoxicity of the hydrogel extractions,
the heat of the autoclave cycle allows for intact epoxide groups to hydrolyze in the
aqueous environment of the hydrogel.
In the last portion of this project, unpurified and purified hydrogel formulations
varying PEG-DGE content, PVA content, PEG functional group, and PEG molecular
weight were swollen in 0.2 MPa osmotic solution to similar the intradiscal environment
for up to four weeks at 37oC. The mechanical properties for the unpurified formulations
decreased from the room temperature values due to mass loss of the network and the
replacing of the PEG-DGE with water. The crystallinity of the unpurified formulations
increases over four weeks due to the aging phenomena of PVA. The purified
formulations have higher compressive moduli at four weeks swelling in osmotic solution
than at 24 hours at room temperature, due to the aging phenomena of PVA. The
crystallinity of the purified hydrogels does not change from the RT values. This is due to
the change in the gelation mechanism of the purified hydrogels; the PVA rich regions are
less dense. Resulting in additional water being able to hydrogen bond to the PVA when
swollen preventing additional crystallinity.
237
All hydrogel formulations tested in this thesis could be injected though a 10 gauge
needle after heating above 80oC either through an autoclave cycle or using a specially
designed delivery device. Above 80oC the physical associations of the PVA are melted out
[100-103], allowing for ability of the hydrogel formulations to be injectable.
8.2. Recommendations
The chemical crosslinking of the PVA/PVP/PEG-DGE hydrogel system was
never quantitatively measured; only determination of chemical crosslinking was the
inability to dissolve the hydrogel network in water or DMSO. FTIR was used to try to
determine the percent crosslinking of the system but due to the large amount of ether
groups already present in the system crosslinking could not be measured. The author
suggests the use of solid state NMR to further characterize this hydrogel system. [234-
236]
More in-depth mechanical and biomechanical studies of the PVA/PVP/PEG-
DGE hydrogel formulations are warranted, in particular the 39% PEG-DGE and 20.0%
PVA formulations. Fatigue testing has been performed on a number of material
candidates being investigated for nucleus replacement [84, 85, 237-239]. In addition to
fatigue testing, Joshi et al. [84, 85] evaluated PVA/PVP hydrogels for NP replacement
implanting the hydrogels in cadaveric specimens to evaluate the compressive behavior of
the implanted discs. Complex loading conditions are also recommended for evaluation of
the PEG/PVP/PEG-DGE hydrogel system.
238
Vernengo et al. [110, 219] developed an injectable, bioadhesive hydrogel system
of branched copolymers of poly(N- isopropylacrylamide) (PNIPAAm) and poly(ethylene
glycol) (PEG) were blended with poly(ethylene imine) (PEI) which with the addition of
an injection of glutaraldehyde into the gel core will adhere the implant to the surrounding
tissues. The epoxide groups of PEG-DGE are reactive with the amines present in the
collagen of the AF tissue. The current synthesis for the PVA/PVP/PEG-DGE hydrogel
hydrolyzed the epoxide groups unreacted with PVA between the basic catalyst and the
repeated autoclave cycles. Using the high molecular PEG-DGE version of the hydrogel,
due to the lack of mass retention issues, it might be possible to lower the temperatures
that the hydrogel experience to preserve the epoxide groups to allow for adhesion of the
hydrogel to the surrounding AF tissues after implantation.
Follow up to the cytotoxicity studies with L929 fibroblast cells, but be cytotoxicity
testing with human annulus fibrosus cells would be necessary [93] including cell
proliferation assays in the presence of PEG-DGE and the PVA/PVP/PEG-DGE
hydrogel. In addition histological and immunohistochemical studies will be warranted to
determine the in vivo inflammatory response to the system [240].
Determinations of effect of aging of the PVA/PVP/PEG-DGE hydrogel system
without the effect of swelling and deswelling from immersion in 0.2 MPa osmotic
solution on the phase separation and crystallinity of the hydrogel system. Aging of the
PVA/PVP/PEG-DGE hydrogel system in sealed containers without solution to
determine the change in crystallinity and phase separation with the effect of swelling of
the hydrogel network is recommended to further understand the changing of the
239
PVA/PVP/PEG-DGE hydrogel system with time as the hydrogel would likely be
manufactured, packaged and have a shelf life in a sealed container or in an container with
an aqueous solution before implantation.
240
REFERENCES
[1] Q. B. Bao and H. A. Yuan, "Artificial disc technology," Neurosurgical focus, vol. 9,
p. e14, 2000.
[2] A. H. Wheeler, "Diagnosis and management of low back pain and sciatica," Am
Fam Phys, vol. 52, pp. 1333-41, 1995.
[3] J. E. Wipf and R. A. Deyo, "Low back pain," Med Clin N Am, vol. 79, pp. 231-46,
1995.
[4] Q. B. Bao and P. A. Higham, "Hydrogel intervertebral disc nuclues," 5,047,055,
1991.
[5] Q. B. Bao, G. M. McCullen, P. A. Higham, J. H. Dumbleton, and H. A. Yuan,
"The artificial disc: Theory, design and materials," Biomaterials, vol. 17, pp. 1157-
1167, 1996.
[6] Q. B. Bao and H. A. Yuan, "Prosthetic disc replacement: The future?," Clinical
Orthopaedics and Related Research, pp. 139-145, 2002.
[7] P. P. Raj, "Intervertebral disc: anatomy-physiology-pathophysiology-treatment,"
Pain practice : the official journal of World Institute of Pain, vol. 8, pp. 18-44, 2008.
[8] S. A. Herring, "Artificial discs and future of spine surgery," Bone and Joint, pp.
53-54, 2005.
[9] A. A. White and M. M. Panjabi, Clinical Biomechanics of the Spine, 2nd ed.:
Lippincott Co. , 1990.
[10] S. M. Kurtz and A. A. Edinin, Spine Technology Handbook. Burlington, MA:
Academic Press, 2006.
241
[11] M. A. Adams, N. Bogduk, K. Burton, and P. Dolan, The Biomechanics of Back
Pain. Edinburgh: Churchill Livingstone, 2002.
[12] I. Stokes and J. C. Iatridis, "Mechanical Conditions That Accelerate
Intervertebral Disc Degeneration: Overload Versus Immobilization," Spine, vol.
29, pp. 2724-2732, 2004.
[13] D. R. Eyre, "Biochemistry of the Intervertebral Disc," Int. Rev. Connect. Tissue
Res. , vol. 8, pp. 227-291, 1979.
[14] J. M. J. Buckwalter, Intervertebral Disk Degeneration and Back Pain: American
Academy of Orthopaedic Surgeons, 1996.
[15] P. J. Roughley, "Biology of Intervertebral Disc Aging and Degeneration," Spine,
vol. 29, pp. 2691-2699, 2004.
[16] T. R. J. Oegema, "Biochemistry of the Intervertebral Disc," Clin. Sports Med., vol.
12, pp. 419-439, 1993.
[17] A. Nachemson and J. M. Morris, "In Vivo Measurements of Intradiscal
Pressure," Journal of Bone and Joint Surgery, vol. 46A, pp. 1077-1092, 1964.
[18] H. J. Wilke, P. Neef, and M. Caimi, "New In Vivo Measurements of Pressures in
the Intervertebral Disc in Daily Life," Spine vol. 24, 1999.
[19] D. S. McNally and M. A. Adams, "Internal Intervertebral Disc Mechanics As
Revealed by Stress Profilometry," Spine, vol. 17, 1992.
[20] W. Johannessen and D. M. Elliott, "Effects of Degeneration on the Biphasic
Material Properties of Human Nucleus Pulposus in Confined Compression,"
Spine, vol. 30, pp. E724-729, 2005.
242
[21] J. C. Iatridis, L. A. Setton, and M. Weidenbaum, "Alterations in the Mechanical
Behavior of the Human Lumbar Nucleus Pulposus with Degeneration and
Aging," J. Orthop. Res., vol. 15, pp. 318-322, 1997.
[22] H. L. Guerin and D. M. Elliott, "Degeneration Affects the Fiber Reorientation
of Human Annulus Fibrosus Under Tensile Load," J. Biomech., vol. 39, pp. 1410-
8, 2006.
[23] J. C. Iatridis, L. A. Setton, and R. J. Foster, "Degeneration Affects the
Anisotropic and Nonlinear Behaviors of Human Annulus Fibrosus in
Compression," J. Biomech., vol. 31, pp. 535-544, 1998.
[24] D. M. Elliott and L. A. Setton, "Anisotropic and Inhomogeneous Tensile
Behavior of the Human Anulus Fibrosus: Experiemental Measurements and
Material Model Predictions," J. Biomech. Eng., vol. 123, pp. 256-263, 2001.
[25] S. L. Woo, G. A. Johnson, and B. A. Smith, "Mathematical Modeling of
Ligaments and Tendons," J. Biomech. Eng., vol. 115, pp. 468-473, 1993.
[26] L. Yin and D. M. Elliott, "A Biphasic and Transversely Isotropic Mechanical
Model for Tendon: Application to Mouse Tail Fascicles in Uniaxial Tension,"
Journal of Biomechanics, vol. 37, 2004.
[27] J. A. Buckwalter, "Aging and degeneration of the human intervertebral disc,"
Spine, vol. 20, pp. 1307-14, 1995.
[28] M. C. Battie, T. Videman, and E. Parent, "Lumar Disc Degeneration:
Epidemiology and Genetic Influences," Spine, vol. 29, pp. 2679-2690, 2004.
243
[29] O. L. Osti and D. E. Cullum, "Occupational Low Back Pain and Intervertebral
Disc Degeneration: Epidemiology, Imaging and Pathology," Clin. J. Pain, vol. 10,
pp. 331-334, 1994.
[30] M. C. Battie, T. Videman, and K. Gill, "1991 Volvo Award in Clinical Sciences.
Smoking and Lumbar Intervertebral Disc Degeneration: An MRI Study of
Indentical Twins," Spine, vol. 16, pp. 1015-1021, 1991.
[31] J. W. Frymoyer and W. L. Cats-Baril, "An Overview of the Incidences and Costs
of Low Back Pain," Orthopedic Clinics Of North America, vol. 22, pp. 263-271,
1991.
[32] B. Johnstone and M. T. Bayliss, "The Large Proteoglycan of the Human
Interverbral Disc: Changes in Their Biosynthesis and Structure with Age,
Topography and Pathology," Spine, vol. 20, pp. 674-684, 1995.
[33] J. Antoniou, R. Steffen, and F. Nelson, "The Human Lumbar Intervertebral
Disc: Evidence for Changes in the Biosynthesis and Denaturation of the
Extracellular Martix with Growth, Maturation, Ageing, and Degeneration," J.
Clin. Invest., vol. 98, pp. 996-1003, 1996.
[34] P. Goupille, M. I. Jayson, and J. P. Valat, "Matric Metalloproteinases: The Clue
to Intervertebral Disc Degeneration," Spine, vol. 23, pp. 1612-1626, 1998.
[35] A. G. Nerlich, E. D. Schleicher, and N. Boos, "1997 Volvo Award Winner in
Basic Science Studies: Immunohistologic Markers for Age-related Changes of
Human Lumbar Intervertebral Discs," Spine, vol. 22, pp. 2781-2795, 1997.
244
[36] J. C. Iatridis, M. Weidenbaum, L. A. Setton, and V. C. Mow, "Is the nucleus
pulposus a solid or a fluid? Mechanical behaviors of the nucleus pulposus of the
human intervertebral disc," Spine, vol. 21, pp. 1174-84, 1996.
[37] J. R. Meakin, T. W. Redpath, and D. W. Hukins, "The Effect of Partial
Removal of the Nucleus Pulposus from the Intervertebral Disc on the Response
of the Human Annulus Fibrosus to Compression," Clinical Biomechanics, vol. 16,
pp. 121-128, 2001.
[38] R. E. Seroussi, M. H. Krag, and D. L. Muller, "Internal Deformations of Intact
and Denucleated Human Lumbar Discs Subjected to Compression, Flexion, and
Extension Loads," J. Orthop. Res., vol. 7, pp. 122-131, 1989.
[39] J. C. Iatridis and I. Gwynn, "Mechanisms for Mechanical Damage in the
Intervertebral Disc Annulus Fibrosus," J. Biomech., vol. 37, pp. 1165-1175, 2004.
[40] F. Marchand and A. M. Ahmed, "Investigation of the Laminate Structure of
Lumbar Disc Annulus Fibrosus," Spine, vol. 15, pp. 402-410, 1990.
[41] J. C. Iatridis, S. Kumar, and R. J. Foster, "Shear Mechanical Properties of
Human Lumbar Annulus Fibrosus," J. Orthop. Res., vol. 17, pp. 732-737, 1999.
[42] E. R. Acaroglu, J. C. Iatridis, and L. A. Setton, "Degeneration and Aging Affect
the Tensile Behavior of Human Lumbar Annulus Fibrosus," Spine, vol. 20, pp.
2690-2701, 1995.
[43] S. E. Hormel and D. R. Eyre, "Collagen in the Aging Human Interverteral Disc:
An Increase in Covalently Bound Fluorophores and Chromophores," Biochimica
et Biophysica Acta, vol. 1078, pp. 243-250, 1991.
245
[44] S. Bernick and R. Cailliet, "Vertebral End-plate Changes with Aging of Human
Vertebrae," Spine, vol. 7, pp. 97-102, 1982.
[45] L. M. Benneker, P. F. Heini, and M. Alini, "2004 Young Investigator Award
Winner: Vertebral Endplate Marrow Contact Channel Occlusions and
Intervertebral Disc Degneration," Spine, vol. 30, pp. 167-173, 2005.
[46] N. Boos, S. Weissbach, and H. Rohrbach, "Classification of Age-related Changes
in Lumbar Intervertebral Discs: 2002 Volvo Award in Basic Science," Spine, vol.
27, 2002.
[47] J. P. Urban, S. Smith, and J. C. Fairbank, "Nutrition of the Intervertbral Disc,"
Spine, vol. 29, pp. 2700-2709, 2004.
[48] J. P. Grant, T. R. Oxland, and M. F. Dvorak, "Mapping the Structural Poperties
of the Lumbosacral Vertebral Endplates," Spine, vol. 26, pp. 889-896, 2001.
[49] O. Perey, "Fracture of the Vertebral End-plate in the Lumbar Spine: An
Experimental Biochemical Investigations," Acta Orthopaedica Scandinavica, pp. 1-
101, 1957.
[50] J. A. Buckwalter, S. D. Boden, and D. R. Eyre, "Intervertebral Disk Aging,
Degeneration, and Herniation," in Orthopaedic Basic Science, J. A. Buckwalter, T.
A. Einhorn, and S. R. Simon, Eds., ed: American Academy of Orthopaedic
Surgeons, 2000.
[51] J. M. Spivak, "Degenerative Lumbar Spinal Stenosis," J. Bone Joint Surg. Am., vol.
80, pp. 1053-1066, 1998.
[52] C. Harvey, "Spinal surgery patient care," Othopaedic Nursing vol. 24, pp. 426-440,
2005.
246
[53] J. Best, "Understand spinal stenosis," Orthopaedic Nursing, vol. 21, pp. 48-55,
2002.
[54] G. Bach and R. D. Lim, "Minimally invasive spine surgery for low back pain,"
Disease-a-month, vol. 51, pp. 34-57, 2005.
[55] W. Xiaotao, S. Zhuang, Z. Mao, and H. Chen, "Microendoscopic discectomy for
lumbar disc herniation surgical technique and outcome in 873 consecutive cases,"
Spine, vol. 31, 2006.
[56] H. J. Wilke, U. Mehnert, L. Claes, J. H. Bierschneider, and B. M. Boszcyk,
"Biomechanical evaluation of vertebroplasty and kyphoplasty with polymethyl
methacrylate or calcium phosphate cement under cyclic loading," Spine, vol. 31,
pp. 2934-2941, 2006.
[57] R. A. Deyo, A. Nachemson, and S. K. Mirza, "Spinal fusion surgery: The case for
restraint," New England Journal of Medicine, vol. 350, 2004.
[58] D. Polly, E. R. G. Santo, and A. A. Mehbod, "Surgical treatment for the painful
motion segment: Matching technology with the indications- posterior lumber
fusion," Spine, vol. 30, pp. 544-551, 2005.
[59] R. A. Deyo, D. T. Gray, W. Kreuter, S. K. Mirza, and B. L. Martin, "United
states trends in lumbar fusion surgery for degenerative conditions," Spine, vol. 30,
pp. 1441-1445, 2005.
[60] A. Mofidi, M. Sedhom, K. O'Shea, E. E. Forgarty, and F. Dowling, "Is high
level of disability an indication for spinal fusion? Analysis of long-term outcome
after posterior lumber interbody fusion using carbon fiber cages.," Journal of spinal
disorders & techniques, vol. 18, 2005.
247
[61] S. Brady and S. Jackson, "Anterior lumbar interbody fusion - advances in spinal
fusion technology," AORN vol. 82, pp. 817-823, 2005.
[62] J. C. Fairbank, H. Frost, J. Wilson-MacDonald, L. M. Yu, and K. Barker,
"Surgical stabilization of the spine improved disability slightly more than an
intensive rehabilitation program in patients with chronic low back pain," The
Journal of Bone and Joint Surgery, vol. 88, p. 453, 2006.
[63] J. Brox, R. Sorensen, and A. Friis, "Lumbar fusion did not differ from cognitive
therapy plus exercise for reducing disability and pain in chronic low back pain. ,"
Spine, vol. 28, pp. 1913-1921, 2003.
[64] F. B. Christensen, K. Thomsen, S. P. Eiskjaer, E. S. Hansen, S. Fruensgaard, J.
Gelinick, and C. E. Bunger, "The effect of pedicle screw instrumentation on
functional outcome and fusion rates in posterolateral lumbar spinal fusion: a
prospective, randomized clinical study," Spine, vol. 22, pp. 2813-2822, 1997.
[65] P. Fritzell, O. Hagg, P. Wessberg, and A. Nordwall, "Chronic low back pain and
fusion: A comparison of three surgical techniques: A prospective multicenter
randomized study from the swedish lumbar spine study group," Spine, vol. 27, pp.
1131-1141, 2002.
[66] C. Lee and N. A. Lagrana, "Lumbosacral spinal fusion: A biomechanical study.,"
Spine, vol. 9, 1984.
[67] N. Miyakoshi, E. Abe, and Y. Shimada, "Outcome of one-level posterior lumbar
interbody fusion for spondylithesis and postoperative intervertebral disc
degeneration adjacent to fusion," Spine, vol. 25, pp. 1837-1842, 2000.
248
[68] S. Hochschuler, D. D. Ohnmeiss, R. D. Guyer, and S. L. Blumenthal, "Artificial
disc: perliminary results of a prospective study in the United States," European
Spine Journal, vol. 11, pp. S106-S110, 2002.
[69] A. Van Ooij, F. C. Oner, and A. J. Verbout, "Complications of artifical disc
replacement," Journal of spinal disorders & techniques, vol. 16, pp. 369-383, 2003.
[70] H. M. Mayer, "Total lumbar disc replacement," British Editorial Society of Bone
and Joint Surgery, vol. 87-B, pp. 1029-1037, 2005.
[71] H. Link, "History, design and biomechanics of the LINK SB Chartie artifical
disc," European Spine Journal, vol. 11, pp. S98-S105, 2002.
[72] S. L. Blumenthal, P. C. McAfee, R. D. Guyer, S. H. Hochschuler, F. H. Geisler,
R. T. Holt, R. Garcia, J. J. Regan, and D. D. Ohnmeiss, "A prospective,
randomized, multicenter food and drug administration investigational device
exemptions study of lumbar total disc replacement with the CHARITETM
artificial disc versus lumbar fusion: Part I: evaluation of clinical outcomes," Spine,
vol. 30, pp. 1565-1575, 2005.
[73] P. C. McAfee, B. Cunningham, G. Holsapple, K. Adams, S. L. Blumenthal, R.
D. Guyer, A. Dmietriev, A. Maxwell, J. J. Regan, and J. Isaza, "A prospective,
randomized multicenter Food and Drug Administration investigational device
exemption study of lumbar total disc replacement with Charite Artificial Disc
versus lumbar fusion," Spine, vol. 30, pp. 1576-1583, 2005.
[74] S. K. Mirza, "Point of view: Commentary on the research reports that led to Food
and Drug Administration approval of an artificial disc," Spine, vol. 30, pp. 1561-
1564, 2005.
249
[75] J. J. Yue, R. Bertagnoli, P. C. McAfee, and H. S. An, Motion Preservation Surgery
of the Spine, 1st ed.: Elsevier, 2008.
[76] P. M. Klara and C. D. Ray, "Artificial nucleus replacement: clinical experience,"
Spine, vol. 27, pp. 1374-1377, 2002.
[77] J. Husson, A. Korge, J. L. Polard, T. Nydegger, S. Kneubuhler, and H. M. Mayer,
"A memory coiling spiral as nucleus pulposus prothesis," Journal of Spinal
Disorders & Techniques, vol. 16, pp. 405-411, 2003.
[78] R. Dunlop, M. A. Adams, and W. C. Hutton, "Disc Space Narrowing and the
Lumbar Facet Joints," Journal of Bone and Joint Surgery, vol. 66, pp. 706-710, 1984.
[79] A. Di Martino, A. R. Vaccaro, J. Y. Lee, V. Denaro, and M. R. Lim, "Nucleus
pulposus replacement - Basic science and indications for clinical use," Spine, vol.
30, pp. S16-S22, 2005.
[80] H. M. Mayer and A. Korge, "Non-fusion Technology in Degenerative Lumbar
Spinal Disorders: Facts, Questions, Challenges.," European Spine Journal, vol. 11,
pp. S85-S91, 2002.
[81] A. Carl, E. Ledet, H. Yuan, and A. Sharan, "New Development in Nucleus
Pulposus Replacement Technology," The Spine Journal, vol. 4, pp. 325S-329S,
2004.
[82] M. Goins, D. W. Wimberly, P. S. Yuan, L. N. Fitzhenry, and A. R. Vaccaro,
"Nucleus Pulposus Replacement: an Emerging Technology," The Spine Journal,
vol. 5, 2005.
[83] J. meakin, J. E. Reid, and D. W. L. Hulkins, "Replacing the nucleus pulposus of
the intervertebral disc," Clinical Biomechanics, vol. 16, pp. 560-565, 2001.
250
[84] A. Joshi, G. Fussell, J. Thomas, A. Hsuan, A. Lowman, A. Karduna, E.
Vresilovic, and M. Marcolongo, "Functional compressive mechanics of a
PVA/PVP nucleus pulposus replacement," Biomaterials, vol. 27, pp. 176-184,
2006.
[85] A. B. Joshi, "Mechanical Behavior of the Human Lumbar Intervertebral Disc
with Polymeric Hydrogel Nucleus Implant: An Experiemental and Finite
Element Study," PhD Phd Thesis, Drexel University, 2004.
[86] V. Goel, K. Nishiyama, J. N. Weinstein, and Y. K. Lui, "Mechanical properties
of lumbar spinal motion segments as affected by partial disc removal," Spine, vol.
11, pp. 1008-1012, 1986.
[87] J. P. G. Urban and J. F. McMullin, "Swelling pressure of the intervertebral disk -
influence of proteoglycan and collagen contents " Biorheology, vol. 22, pp. 145-
157, 1985.
[88] V. Stoy, J. Sabatino, and G. J, "Mechanical testing of a hydrogel nucleus
replacement implant," presented at the Spine Across the Sea 2003, Maui, Hawaii,
2003.
[89] U. Berlemann and O. Schwarzenbach, "An injectable nucleus replacement as an
adjunct to microdiscectomy: 2 year follow-up in a pilot clinical study," European
Spine Journal, vol. 18, pp. 1706-1712, 2009.
[90] (January 23, 2013). NuCore® Injectable Nucleus. Available:
http://www.spinewave.com/products/nucore.html
251
[91] M. J. Allen, J. E. Schoonmaker, T. W. Bauer, P. F. Williams, P. A. Higham, and
H. A. Yuan, "Preclinical evaluation of a poly (vinyl alcohol) hydrogel implant as a
replacement for the nucleus pulposus," Spine, vol. 29, pp. 515-23, 2004.
[92] U. Yuksel, S. Walsh, D. Curd, and K. Black, "Fatigue durabiity of a novel disc
nucleus repair system: in vitro studies in a calf spine model," The Spine Journal, vol.
2, pp. 103-104, 2002.
[93] S. Sarkar, "Synthesis and Characterization of a Chondroitin Sulfate Based Hybrid
Bio/Synthetic Biomimetic Aggrecan Macromolecule," PhD, Biomedical
Engineering, Drexel University, Philadelphia, 2011.
[94] S. Sarkar, S. E. Lightfoot-Vidal, C. L. Schauer, E. Vresilovic, and M.
Marcolongo, "Terminal-end functionalization of chondroitin sulfate for the
synthesis of biomimetic proteoglycans," Carbohydrate Polymers, vol. 90, pp. 431-
440, 2012.
[95] C. M. Hassan and N. A. Peppas, "Structure and applications of poly(vinyl
alcohol) hydrogels produced by conventional crosslinking or by freezing/thawing
methods," Biopolymers/Pva Hydrogels/Anionic Polymerisation Nanocomposites, vol.
153, pp. 37-65, 2000.
[96] A. M. Lowman and N. A. Peppas, "Hydrogels," in Encyclopedia of Controlled Drug
Delivery, E. Mathiowitz, Ed., ed: John Wiley and Sons, 1999, pp. 397-418.
[97] N. A. Peppas and E. W. Merrill, "Differential Scanning Calorimetry of
Crystallized PVA Hydrogels," Journal of Applied Polymer Science, vol. 20, pp.
1457-1465, 1976.
252
[98] H. Bodugoz-Senturk, J. Choi, E. Oral, J. Kung, C. Macias, G. Braithwaite, and
O. Muratoglu, "The effect of polyethylene glycol on the stability of pores in
polyvinyl alcohol hydrogels during annealing," Biomaterials, vol. 29, pp. 141-149,
2008.
[99] H. Bodugoz-Senturk, C. Macias, J. Kung, and O. Muratoglu, "Poly (vinyl
alcohol)-acrylamide hydrogels as load-bearing cartilage substitute," Biomaterials,
vol. 30, pp. 589-596, 2009.
[100] J. W. Ruberti and G. J. C. Braithwaite, "Systems and Methods for Controlling
and Forming Polymer Gels," 7,485,670, 2009.
[101] J. W. Ruberti and G. Braithwaite, "Systems and methods for controlling and
forming polymer gels," 7745532, June 29, 2010, 2010.
[102] J. W. Ruberti and G. Braithwaite, "Systems and methods for controlling and
forming polymer gels," 7619009, November 17, 2009, 2009.
[103] J. W. Ruberti and G. Braithwaite, "Systems and methods for controlling and
forming polymer hydrogels," 7776352, August 17, 2010, 2010.
[104] R. Ricciardi, F. Auriemma, C. Gaillet, C. De Rosa, and F. Laupretre,
"Investigation of the crystallinity of freeze/thaw poly (vinyl alcohol) hydrogels by
different techniques," Macromolecules, vol. 37, pp. 9510-9516, 2004.
[105] X. Liu, G. Fussell, M. Marcolongo, and A. Lowman, "Characterization of
associating hydrogels of poly(vinyl alcohol) and poly(vinyl pyrrolidone)," Journal of
Applied Polymer Science, vol. 112, pp. 541-549, 2009.
253
[106] T. Tanigami, Y. Nakashima, K. Murase, H. Suzuki, K. Yamaura, and S.
Matsuzawa, "High strength and high modulus poly(vinyl alcohol) by the gel
ageing method," J. Mater. Sci., vol. 30, pp. 5110-20, 1995.
[107] T. Tanigami, K. Murase, K. Yamaura, and S. Matsumura, "Aging of poly(vinyl
alcohol) gels prepared from demethylesuloxide/water solutions," Polymer, vol. 35,
pp. 2573-2578, 1994.
[108] G. Odian, Principles of Polymerization, 4th ed. Hoboken, New Jersey: John Wiley
& Sons, Inc. , 2004.
[109] M. A. Slivka and H. Serhan, "Method for Nonsurgical Treatment of the
Intervertebral Disc and Kit Therefor," United States Patent US 7,294,617 B2,
2007.
[110] J. Vernengo, "Injectable Bioadhesive Hydrogels for Nucleus Pulposus
Replacement and Repair of the Damaged Intervertebral Disc," PhD PhD,
Chemical and Biological Engineering, Drexel University, Philadelphia, 2007.
[111] S. Zalipsky, "Functionalized poly(ethylene glycol) for preparation of biologically
relevant conjugates," Bioconjugate Chemistry, vol. 6, pp. 150-165, 1995.
[112] K. B. Kita, "Characterization of In-Situ Curing PVA-PEG Hydrogels for
Nucleus Pulposus Replacement," Ph.D., Materials Science and Engineering,
Drexel University, Philadelphia, 2010.
[113] K. B. Kita, N. G. Smith, A. M. Lowman, G. W. Fussell, and M. Keane, "Phase
Transitioning Hydrogels," 2011.
254
[114] I. Inamura, K. Toki, T. Tamae, and T. Araki, "Effects of molecular weight on the
phase equilibrium of a poly(vinyl alcohol)-poly (ethylene glycol)-water system,"
Polymer Journal, vol. 16, p. 657, 1984.
[115] I. Inamura, "Liquid-liquid phase separation and gelation in the poly (vinyl
alcohol)-poly (ethylene glycol)-water system. Dependence on molecular weight of
poly (ethylene glycol)," Polymer journal, vol. 18, pp. 269-272, 1986.
[116] I. Inamura and Y. Jinbo, "Interaction between poly (vinyl alcohol) and poly
(ethylene glycol) in water studied by viscosity and density," Polymer Journal, vol.
23, pp. 1143-1147, 1991.
[117] I. Inamura, Y. Jinbo, Y. Akiyama, and Y. Kubo, "Effect of a Liquid-Liquid Phase
Separation on the Slopes of Density vs. Concentration Plots," Bulletin of the
Chemical Society of Japan, vol. 68, pp. 2021-2023, 1995.
[118] I. Inamura, K. Akiyama, and Y. Kubo, "Polymer 2-polymer 3 interactions in
water 1/polymer 2/polymer 3 ternary systems," Polymer Journal, vol. 29, pp. 119-
122, 1997.
[119] I. Inamura, M. Kittaka, T. Aikou, K. Akiyama, T. Matsuyama, M. Hiroto, K.
Hirade, and Y. Jinbo, "Relation between Huggins Constant and Compatibility of
Binary Polymer Mixtures in the Aqueous Ternary Systems," Macromolecular
Research, vol. 12, pp. 246-250, 2004.
[120] J. Choi, H. Bodugoz-Senturk, H. J. Kung, A. S. Malhi, and O. K. Muratoglu,
"Effects of solvent dehydration on creep resistance of poly(vinyl alcohol)
hydrogel," Biomaterials, vol. 28, pp. 772-780, 2007.
255
[121] J. Hu, U. Schulze, and J. Pionteck, "Degradation of interpenetrating polymer
networks based on PE and polymethacrylates by electron beam irradiation,"
Polymer, vol. 40, pp. 5279-5284, 1999.
[122] J. E. Wilson, Radiation chemistry of monomers, polymers, and plastics. New York:
Marcel Dekker, 1974.
[123] K. Burczak, T. Fujisato, M. Hitada, and Y. Ikada, "Protein permeation through
poly vinyl alcohol hydrogel membranes," Biomaterials, vol. 15, pp. 231-8, 1994.
[124] T. Noguchi, T. Yamamuro, M. Oka, P. Kumar, Y. Kotoura, S. H. Hyon, and Y.
Ikada, "Poly vinyl alcohol hydrogel as an artificial articular cartilage—evaluation
of biocompatibility. ," J Appl Biomater, vol. 2, pp. 101-7, 1991.
[125] N. Peppas and E. Merrill, "Development of semicrystalline poly vinyl alcohol
hydrogels for biomedical applications," J Biomed Mater Res, vol. 11, pp. 423-34,
1977.
[126] B. Sarti and M. Scandola, "Viscoelastic and thermal properties of collagen/ poly
vinyl alcohol blends," Biomaterials, vol. 16, 1995.
[127] T. H. Young, N. K. Yao, R. F. Chang, and L. W. Chen, "Evaluation of
asymmetric poly vinyl alcohol membranes for use in artificial islets. ," Biomaterials,
vol. 17, pp. 2139-45, 1996.
[128] J. Delecrin, M. Oka, P. Kumar, S. Takahashi, Y. Kotoura, and T. Yamamuro,
"Joint reactions against polymer particles: PVA-H versus UHMWPE," J Jpn
Orthop Assoc, vol. 64, p. S1395, 1990.
256
[129] M. Kobayashi, J. Toguchida, and M. Oka, "Preliminary study of polyvinyl
alcohol-hydrogel (PVA-H) artificial meniscus," Biomaterials, vol. 24, pp. 639-647,
2003.
[130] J. L. Holloway, "Development and Characterization of UHMWPE Fiber-
Reinforced Hydrogels for Meniscal Replacement," Ph.D. , Chemical and
Biological Engineering, Drexel University, Philadelphia, 2012.
[131] K. Spiller, S. Laurencin, D. Charlton, S. Maher, and A. Lowman, "Superporous
hydrogels for cartilage repair: Evaluation of the morphological and mechanical
properties," Acta Biomaterialia, vol. 4, pp. 17-25, 2008.
[132] K. Spiller, S. Laurencin, and A. Lowman, "Characterization of the Behavior of
Porous Hydrogels in Model Osmotically-Conditioned Articular Cartilage
Systems," Journal Of Biomedical Materials Research Part B-Applied Biomaterials, vol.
90B, pp. 752-759, 2009.
[133] M. Oka, K. Ushio, P. Kumar, K. Ikeuchi, S. H. Hyon, K. Nakamura, and H.
Fujita, "Development of artificial articular cartilage," in Proceedings of the
Institution of Mechanical Engineers, Part H: Journal of Engineering in Medicine,
2000, pp. 59-68.
[134] M. Kobayashi, Y. S. Chang, and M. Oka, "A two year in vivo study of polyvinyl
alcohol-hydrogel (PVA-H) artificial meniscus.," Biomaterials, vol. 26, pp. 3243-8,
2005.
[135] M. Oka, T. Noguchi, P. Kumar, K. Ikeuchi, T. Yamamuro, and Y. Ikada,
"Development of an artifical articular cartilage," Clinical Materials, vol. 6, pp. 361-
381, 1990.
257
[136] J. M. Harris, "Introduction to biotechnical and biomedical applications of
poly(ethylene glycol)," in Poly(ethylene glycol) chemistry biotechnical and biomedical
applications., ed New York: Penun, 1992.
[137] S. S. Banerjee, N. Aher, R. Patil, and J. Khandare, "Poly(ethylene glycol)-
Prodrug Conjugates: Concept, Design, and Applications," Journal of Drug
Delivery, vol. 2012, p. 17, 2012.
[138] S. N. S. Alconcel, A. S. Baas, and H. D. Maynard, "FDA-approved poly(ethylene
glycol)-protein conjugate drugs," Polymer Chemistry, vol. 2, pp. 1442-1448, 2011.
[139] M. Quiao, D. Chen, X. Ma, and Y. Liu, "Injectable biodegradable temperature-
responsive PLGA-PEG-PLGA copolymers: Synthesis and effect of copolymer
composition on the drug release from the copolymer-based gels. ," International
Journal of Pharmaceutics, vol. 294, pp. 103-112, 2005.
[140] D. Huynh, W. S. Shim, J. H. Kim, and D. S. Lee, "pH/temperature sensitive
poly(ethylene glycol)-based biodegradable polyester block copolymer hydrogels. ,"
Polymer vol. 47, pp. 7918-7926, 2006.
[141] H. Li, G. E. Yu, C. Price, E. Hect, and H. Hoffman, "Concentrated aqueous
micellar Solutions of diblock copoly(oxyethylene/oxybutylene) E41B8: A study of
phase behavior. ," Macromolecules vol. 30, pp. 1347-1354, 1997.
[142] M. Desai, N. J. Jain, R. Sharma, and P. Bahadur, "Temperature and salt induced
micellization of some block copolymers in aqueous solution. ," Journal of
Surfactants and Detergents vol. 3, pp. 193-199, 2000.
[143] C. Sims, D. Butler, P. E. M. Casanova, R. Lee, B. T. Randolph, M. A. Lee, W.
P. A. Vacanti, and M. J. Yaremchuk, "Injectable Cartilage Using Polyethylene
258
Oxide Polymer Substrates. ," Plastic and Reconstructive Surgery vol. 98, pp. 843-
850, 1996.
[144] J. Elisseeff, K. Anseth, D. Simms, McIntosh, M. Randolph, M. Yaremchuk, and
L. R., "Transdermal Photopolymerization of Poly(Ethylene Oxide)-Based
Injectable Hydrogels for Tissue-Engineered Cartilage. ," Plastic and Reconstructive
Surgery vol. 104, pp. 1014-1022, 1999.
[145] D. Dikovsky, H. Bianco-Peled, and D. Seliktar, "The effect of structural
alterations of PEG-fibrinogen hydrogel scaffolds on 3-D morphology and cellular
migration. ," Biomaterials vol. 27, pp. 1496-1506, 2006.
[146] C. Allen, D. Maysinger, and A. Eisenberg, "Nanoengineering block copolymer
aggregates for drug delivery. ," Colloids and Surfaces B: Biointerfaces vol. 16, pp. 3-
27, 1999.
[147] H. Otsuka, Y. Nagasaki, and K. Kataoka, "PEGylated nanoparticles for biological
and pharmaceutical applications," Advanced Drug Delivery Reviews, vol. 55, pp.
403-419, 2003.
[148] R. Hunter, Foundations of Colloid Science. New York: Oxford University Press,
1991.
[149] C. Nishi, N. Nakajima, and Y. Ikada, "In vitro evaluation of cytotoxicity of
diepoxy compounds used for biomaterial modification," J Biomed Mater Res, vol.
29, 1995.
[150] J. M. Lohre, L. Bacling, E. Wickham, S. Guida, J. Farley, K. Thyagarajan, R. Tu,
and R. C. Quijano, "Evaluation of epoxy ether fixed bovine arterial grafts for
mutagenic potential," Am. Soc. Artif. Intern. Organ. J, pp. 106-113, 1993.
259
[151] R. Tu, C. L. Lu, K. Thyagarajan, E. Wang, H. Nguyen, S. H. Shen, C. Hata,
and R. C. Quijano, "Kinetic study of collagen fixation with polyepoxy fixatives," J
Biomed Mater Res, vol. 27, pp. 3-9, 1993.
[152] E. Wang, K. Thyagarajan, R. Tu, D. Lin, C. Hata, S. H. Shen, and R. C.
Quijano, "Evaluation of collagen modification and surface properties of a bovine
artery via polyepoxy compound fixation," Biomaterials, vol. 16, pp. 530-536, 1993.
[153] K. Tomihata, K. Burczak, K. Shiraki, and Y. Ikada, "Cross-linking and
biodegradation of native and denatured collagen," in Polymers of Biological and
Biomedical Significance, ed Washington, D. C. : American Chemical Society, 1994,
pp. 275-286.
[154] M. Matsuoka and M. Nakamura, "Effect of catalysts on epoxy tannage," Leather
Chem, vol. 30, pp. 223-232, 1985.
[155] C. Nojiri, Y. Noishiki, and H. Koyanagi, "Aortacoronary bypass grafting with
heparinized vascular grafts in dogs," J. Thorac. Cardiovasc. Surg., vol. 93, pp. 867-
877, 1987.
[156] Y. Noishiki, K. Kodaira, M. Furuse, and T. Miyata, "Method of preparing
antithrombogenic medical materials," United States Patent 4,806,595, 1989.
[157] E. Imamura, O. Sawatani, H. Koyanagi, Y. Noishiki, and T. Miyata,
"Anticalcification effects of epoxy compounds as cross-linking agent," J Artif
Organs, vol. 17, pp. 1101-1003, 1988.
[158] H. Sung, S. Shen, R. Tu, E. Wang, C. Hata, D. Lin, H. Nguyan, and R. C.
Quijano, "Development of a polyepoxy compound preserved bioprosthetic heart
valve," Int J Artif Organs, vol. 15, p. 562, 1992.
260
[159] C. Hata, R. Tu, S. Shen, D. Lin, H. Sung, T. Witzel, E. Wang, K. Thyagarajan,
Y. Noishiki, Y. Tomizawa, and R. C. Quijano, "A pliable biological valved
conduit with antithrombogenicity," Int J Artif Organs, vol. 15, p. 562, 1992.
[160] H. B. Chan-Mayers, S. Guida, C. G. Roberts, K. Thyagarajan, R. Tu, and R. C.
Quijano, "Sterilization of small caliber vascular graft with a polyepoxy
compound," Am Soc Artif Intern Organs J, vol. 38, pp. 116-119, 1992.
[161] A. Carpentier, G. Lemaigre, R. Laidlas, S. Carpentier, and C. Dubost, "Biologic
factors affecting long-term results of valvular heterografts," J Thorac Cardivasc
Surg, vol. 55, 1976.
[162] H. Suma, T. Oku, H. Sato, R. Koike, Y. Sawada, and A. Takeuchi, "The bioflow
graft for coronary artery bypass," Texas Heart Inst. J., pp. 103-105, 1990.
[163] D. Robensberg, "Dialdehyde starch tanned bovine heterografts: Development," in
Vascular Grafts, P. N. Sawyer and M. J. Kalpitt, Eds., ed New York: Appleton-
Century-Crofts, 1978, pp. 261-270.
[164] H. Dardik and I. Dardik, "Successful arterial substitution with modified human
umbilical vein," Ann Surg, vol. 183, pp. 252-258, 1976.
[165] J. M. Lee, K. Thyagarajan, C. Pereira, J. McIntyre, and R. Tu, "Cross-linking of
a prototype bovine artery xenograft: Comparison of the effects of glutaraldehyde
and four poly(glycidy1ether) compounds," Artif Organs, vol. 15, p. 303, 1991.
[166] J. McIntyre, R. Tu, E. Wang, C. Hata, M. DeJonge, and R. C. Quijano,
"Biochemical properties of a new biomaterial-a polyepoxy tanned bovine artery,"
Trans Soc Biomater, vol. 14, p. 119, 1991.
261
[167] R. Tu, J. McIntryre, C. Hata, C. L. Lu, and R. C. Quijano, "Dynamic internal
compliance of a vascular prothesis," Trans Amer Soc Artif Intern Organs, vol. 37, pp.
470-472, 1991.
[168] J. P. Urban and J. F. McMullin, "Swelling pressure of the lumbar intervertebral
discs: influence of age, spinal level, composition, and degeneration," Spine, vol. 13,
pp. 179-87, 1988.
[169] J. Urban, A. Maroudas, M. Bayliss, and J. Dillon, "Swelling pressures of
proteoglycans at the concentrations found in cartilaginous tissues," Biorheology, vol.
16, p. 447, 1979.
[170] A. Maroudas, "Biophysical chemistry of cartilaginous tissues with special
reference to solute and fluid transport," Biorheology, vol. 12, pp. 233-248, 1975.
[171] S. J. Bryant and K. S. Anseth, "Hydrogel properties influence ECM production
by chondrocytes photoencapsulated in poly(ethylene glycol) hydrogels," J Biomed
Mater Res A, vol. 59, pp. 63-72, 2002.
[172] J. P. Urban and A. Maroudas, "Swelling of the intervertebral disc in vitro,"
Connective Tissue Research, vol. 9, pp. 1-10, 1981.
[173] J. P. G. Urban and A. Maroudas, "Measurement of fixed charge density and
partition coeffient in the intervertebral disc," Biochimica et Biophysica Acta, vol. 586,
pp. 166-178, 2002.
[174] J. P. G. Urban and S. Roberts, "Degeneration of the intervertebral disc," Arthritis
Research & Therapy, vol. 5, pp. 120-130, 2003.
[175] P. J. Basser, R. Schneiderman, R. A. Bank, E. Wachtel, and A. Maroudas,
"Mechanical properites of the collagen network in the human articular cartilage as
262
measured by osmotic stress technique," Arch Biochem Biophys, vol. 351, pp. 207-
219, 1998.
[176] (Janurary 25, 2013). Bone and Spine. Available:
http://boneandspine.com/musculoskeletal-anatomy/spine-anatomy-overview-
human-spine/
[177] J. Burkus, E. E. Transfeldt, S. H. Kitchel, R. G. Watkins, and R. A. Balderston,
"Clinical and radiographic outcomes of anterior lumbar interbody fusion using
recombinant human bone morphogenetic protein-2," Spine, vol. 27, pp. 2396-
2408, 2002.
[178] J. Thomas, A. Lowman, and M. Marcolongo, "Novel associated hydrogels for
nucleus pulposus replacement," Journal of Biomedical Materials Research Part A, vol.
67A, pp. 1329-1337, 2003.
[179] K. Bergstrom, K. Holmberg, A. Safranj, A. S. Hoffman, M. J. Edgell, A.
Kozlowski, B. A. Hovanes, and J. M. Harris, "Reduction of Fibrinogen
Adsorption on PEG-Coated Polystyrene Surfaces," Journal Of Biomedical
Materials Research, vol. 26, pp. 779-790, 1992.
[180] D. Darwis, "Role of radiation processing in production of hydrogels for medical
applications," Atom Indonesia, vol. 35, pp. 85-104, 2009.
[181] A. Mondino, M. Gonzalez, G. Romero, and E. Smolko, "Physical properties of
gamma irradiated poly (vinyl alcohol) hydrogel preparations," Radiation Physics
And Chemistry, vol. 55, pp. 723-726, 1999.
263
[182] N. Peppas and E. Merrill, "Poly (vinyl alcohol) hydrogels: Reinforcement of
radiation-crosslinked networks by crystallization," Journal of Polymer Science:
Polymer Chemistry Edition, vol. 14, pp. 441-457, 1976.
[183] N. Peppas, "Crystallization of polyvinyl alcohol-water films by slow dehydration,"
European Polymer Journal, vol. 12, pp. 495-498, 1976.
[184] N. Peppas and D. Tennenhouse, "Semicrystalline poly (vinyl alcohol) films and
their blends with poly (acrylic acid) and poly (ethylene glycol) for drug delivery
applications," STP pharma sciences, vol. 14, pp. 291-297, 2004.
[185] Y. Zheng, X. Huang, Y. Wang, H. Xu, and X. Chen, "Performance and
characterization of irradiated poly(vinyl alcohol)/polyvinylpyrrolidone composite
hydrogels used as cartilages replacement," Journal of Applied Polymer Science, vol.
113, pp. 736-741, 2009.
[186] N. A. Peppas and E. Merrill, "Crosslinked poly (vinyl alcohol) hydrogels as
swollen elastic networks," Journal of Applied Polymer Science, vol. 21, pp. 1763-
1770, 1977.
[187] R. A. Bader, "Synthesis and viscoelastic characterization of novel hydrogels
generated via photopolymerization of 1,2-epoxy-5-hexene modified poly(vinyl
alcohol) for use in tissue replacement," Acta Biomaterialia, vol. 4, pp. 967-975,
2008.
[188] N. A. Peppas and R. E. Benner, "Proposed method of intracordal injection and
gelation of poly(vinyl alcohol) solution in vocal cords: polymer considerations,"
Biomaterials, vol. 1, pp. 158-162, 2002.
264
[189] J. Chen, Y. Yang, P. Qian, Z. Ma, W. Wu, P. Sung, X. Wang, and J. Li, "Drug
carrying hydrogel base wound dressing," Radiation Physics And Chemistry, vol. 42,
pp. 915-918, 1993.
[190] R. Hernandez, A. Sarafian, D. Lopez, and C. Mijangos, "Viscoelastic properties
of poly (vinyl alcohol) hydrogels and ferrogels obtained through freezing-thawing
cycles," Polymer, vol. 45, pp. 5543-5549, 2004.
[191] L. Lum and J. Elisseeff, "Injectable hydrogels for cartilage tissue engineering,"
Topics in Tissue Engineering (ebook), 2005.
[192] T. Canal and N. A. Peppas, "Correlation between mesh size and equilibrium
degree of swelling of polymeric networks," Journal of Biomedical Materials Research,
vol. 23, pp. 1183-1193, 1989.
[193] B. Gander, R. Gurny, E. Doelker, and N. A. Peppas, "Effect of polymeric
network structure on drug release from cross-linked poly(vinyl alcohol)
micromatrices," Pharmaceutical research, vol. 6, pp. 578-84, 1989.
[194] E. Edmond and A. G. Ogston, "An approach to the study of phase separation in
ternary aqueous systems," Biochem. J. , vol. 109, pp. 569-576, 1968.
[195] P. J. Larkin, IR and Raman Spectroscopy: Principles and Spectral Interpretation.
Waltham: Elsevier, 2011.
[196] H. S. Mansur, R. L. Orefice, and A. A. P. Mansur, "Characterization of
poly(vinyl alcohol)/poly(ethylene glycol) hydrogels and PVA-derived hybrids by
small-angle X-ray scattering and FTIR spectroscopy," Polymer, vol. 45, pp. 7193-
7202, 2004.
265
[197] V. I. Lozinsky, L. G. Damshkaln, I. N. Kurochkin, and Kurochkin, II, "Study of
cryostructuring of polymer systems: 28. Physicochemical properties and
morphology of poly(vinyl alcohol) cryogels formed by multiple freezing-thawing,"
Colloid Journal, vol. 70, pp. 189-198, 2008.
[198] J. L. Holloway, K. Spiller, A. Lowman, and G. R. Palmese, "Analysis of the in
vitro swelling behavior of poly(vinyl alcohol) hydrogels in osmotic pressure
solution for soft tissue replacement," Acta Biomaterialia, pp. 1-6, 2011.
[199] N. Peppas and S. R. Stauffer, "Reinforced uncrosslinked poly(vinyl alcohol) gels
produced by cyclic freezing-thawing processes: a short review " J. Control Release,
vol. 16, pp. 305-310, 1991.
[200] C. M. Hassan and N. A. Peppas, "Structure and applications of poly(vinyl
alcohol) hydrogels produced by conventional crosslinking or by freezing/thawing
methods," Biopolymers/Pva Hydrogels/Anionic Polymerisation Nanocomposites, vol.
153, pp. 37-65, 2000.
[201] C. M. Hassan and N. Peppas, "Structure and morphology of freeze/thawed PVA
hydrogels," Macromolecules, vol. 33, pp. 2472-9, 2000.
[202] R. A. Bader and W. E. Rochefort, "Rheological characterization of
photopolymerized poly(vinyl alcohol) hydrogels for potential use in nucleus
pulposus replacement," Journal of Biomedical Materials Research Part A, vol. 86A,
pp. 494-501, 2008.
[203] V. R. Binetti, K. B. Kita, G. W. Fussell, A. Lowman, and M. S. Marcolongo,
"Injectable cross-linked hydrogels for biomaterial applications," 2013.
266
[204] J. Thomas, K. Gomes, A. Lowman, and M. Marcolongo, "The effect of
dehydration history on PVA/PVP hydrogels for nucleus pulposus replacement,"
Journal of Biomedical Materials Research Part B: Applied Biomaterials, vol. 69B, pp.
135-140, 2004.
[205] J. D. Thomas, "Development of Injectable Hydrogels for Nucleus Pulposus
Replacement," PhD PhD, Materials Science and Engineering, Drexel University,
Philadelphia, 2006.
[206] D. Charlton, M. Peterson, K. Spiller, A. Lowman, P. Torzilli, and S. Maher,
"Semi-Degradable Scaffold for Articular Cartilage Replacement," Tissue
Engineering Part A, vol. 14, pp. 207-213, 2008.
[207] J. L. Holloway, A. Lowman, and G. R. Palmese, "Mechanical evaluation of
poly(vinyl alcohol)-based fibrous composites as biomaterials for meniscal tissue
replacement," Acta Biomaterialia, pp. 1-9, 2010.
[208] C. M. Hassan, J. Ward, and N. A. Peppas, "Modeling of crystal dissolution of
poly (vinyl alcohol) gels produced by freezing/thawing processes," Polymer, vol. 41,
pp. 6729-6739, 2000.
[209] C. Suryanarayana and M. Grant Norton, X-Ray Diffraction: A Practical Approach.
New York: Plenum Press, 1998.
[210] S. M. Mulla, P. S. Phale, and M. R. Saraf, "Use of X-Ray Diffraction Technique
for Polymer Characterization and Studying the Effect of Optical Accessories,"
presented at the AdMet 2012, 2012.
[211] R. Ricciardi, C. Gaillet, G. Ducouret, F. Lafuma, and F. Laupretre,
"Investigation of the relationships between the chain organization and rheological
267
properties of atactic poly (vinyl alcohol) hydrogels," Polymer, vol. 44, pp. 3375-
3380, 2003.
[212] R. Ricciardi, F. Auriemma, C. De Rosa, and F. Laupretre, "X-ray diffraction
analysis of poly (vinyl alcohol) hydrogels, obtained by freezing and thawing
techniques," Macromolecules, vol. 37, pp. 1921-1927, 2004.
[213] P. T. Mather, H. Qin, J. Wu, and J. Bobiak, "POSS-based polyurethanes: from
degradavle polymers to hydrogels," presented at the Medical Polymers Cologne,
Germany, 2006.
[214] !!! INVALID CITATION !!!
[215] S. K. Mallapragada, N. A. Peppas, and P. Colombo, "Crystal dissolution-
controlled release systems .2. Metronidazole release from semicrystalline
poly(vinyl alcohol) systems," Journal Of Biomedical Materials Research, vol. 36, pp.
125-130, 1997.
[216] A. Bigi, G. Cojazzi, S. Panzavolta, N. Roveri, and K. Rubini, "Stabilization of
gelatin films by crosslinking with genipin," Biomaterials, vol. 23, pp. 4827-32,
2002.
[217] Y. Chang, C. C. Tsai, H. C. Liang, and H. W. Sung, "In vivo evaluation of
cellular and acellular bovine pericardia fixed with a naturally occurring
crosslinking agent (genipin)," Biomaterials, vol. 23, pp. 2447-57, 2002.
[218] W. Furst and A. Banerjee, "Release of glutaraldehyde from an albumin-
glutaraldehyde tissue adhesive causes significant in vitro and in vivo toxicity," The
Annals of thoracic surgery, vol. 79, pp. 1522-8; discussion 1529, 2005.
268
[219] J. Vernengo, G. Fussell, N. Smith, and A. Lowman, "Synthesis and
characterization of injectable bioadhesive hydrogels for nucleus pulposus
replacement and repair of the damaged intervertebral disc," Journal of Biomedical
Materials Research Part B: Applied Biomaterials, pp. 309-317, 2010.
[220] S. Hsing-Wen, S. Jeng-Shiuan, and H. Chin-Sheng, "Crosslinking
characteristics of porcine tendons: Effects of fixation with glutaraldehyde or
epoxy," Journal Of Biomedical Materials Research, vol. 30, pp. 361-367, 1996.
[221] A. Jayakrishnan and S. R. Jameela, "Glutaraldehyde as a fixative in bioprostheses
and drug delivery matrices," Biomaterials, vol. 17, pp. 471-484, 1996.
[222] Y. Murayama, S. Satoh, T. Oka, J. Imanishi, and Y. Noishiki, "Reduction of the
antigenicity and immunogenicity of xenografts by a new cross-linking reagent,"
ASAIO transactions / American Society for Artificial Internal Organs, vol. 34, pp.
546-9, 1988.
[223] D. P. Speer, M. Chvapil, C. D. Eskelson, and J. Ulreich, "Biological Effects of
Residual Glutaraldehyde in Glutaraldehyde-Tanned Collagen Biomaterials,"
Journal Of Biomedical Materials Research, vol. 14, pp. 753-764, 1980.
[224] R. Zeeman, P. J. Dijkstra, P. B. van Wachem, M. J. A. van Luyn, M. Hendriks,
P. T. Cahalan, and J. Feijen, "Crosslinking and modification of dermal sheep
collagen using 1,4-butanediol diglycidyl ether," Journal Of Biomedical Materials
Research, vol. 46, pp. 424-433, 1999.
[225] G. Sun, X. Zhang, and C. Chu, "Effect of the molecular weight of polyethylene
glycol (PEG) on the properties of chitosan-PEG-poly(N-isopropylacrylamide)
269
hydrogels," Journal Of Materials Science-Materials In Medicine, vol. 19, pp. 2865-
2872, 2008.
[226] S. Dhawan, K. Dhawan, M. Varma, and V. R. SInha, "Applications of
Poly(ethylene oxide) in Drug Delivery Systems Part II," Pharaceutical Technology,
pp. 82-96, 2005.
[227] J. Vernengo, G. Fussell, N. Smith, and A. Lowman, "Evaluation of novel
injectable hydrogels for nucleus pulposus replacement," Journal of Biomedical
Materials Research Part B: Applied Biomaterials, vol. 84B, pp. 64-69, 2007.
[228] M. Blanco, R. Olmo, and J. M. Teijon, "Hydrogels," in Encyclopedia of
Pharamaceutical Technology, J. Swarbrick, Ed., ed New York: Marcel Dekker, 2004,
pp. 239-260.
[229] R. Hernandez, A. Sarafian, and D. Lopez, "Viscoelastic properties of poly (vinyl
alcohol) hydrogels and ferrogels obtained through freezing-thawing cycles,"
Polymer, vol. 46, pp. 5543-5549, 2004.
[230] P. Willcox, D. Howie, K. Schmidt-Rohr, D. Hoagland, S. Gido, and S.
Pudjijanto, "Microstructure of poly(vinyl alcohol) hydrogels produced by
freeze/thaw cycling," J. Polym. Sci. Pol. Phys., vol. 37, pp. 3428-54, 1999.
[231] N. Peppas, "Turbidimetric studies of aqueous poly (vinyl alcohol) solutions,"
Makromol Chem, vol. 176, pp. 3433-40, 1975.
[232] S. R. Stauffer and N. A. Peppas, "Poly(vinyl alcohol) hydrogels prepared by
freezing-thawing cyclic processing " Polymer, vol. 33, pp. 3932-3936, 1992.
[233] A. A. Tager, A. A. Anikeyeva, L. N. Adamova, V. M. Andreyeva, T. A.
Kuz'mina, and M. V. Tsilipotkina, Polym. Sci. USSR, vol. 13, p. 751, 1972.
270
[234] R. Hernández, D. López, E. Pérez, and C. Mijangos, "Preparation and
Characterization of Interpenetrating Polymer Hydrogels Based on Poly(acrylic
acid) and Poly(vinyl alcohol)," Macromolecular Symposia, vol. 222, pp. 163-168,
2005.
[235] R. Hernandez, E. Perez, C. Mijangos, and D. Lopez, "Poly(vinyl alcohol)-
poly(acrylic acid) interpenetrating networks. Study on phase separation and
molecular motions," Polymer, vol. 46, pp. 7066-7071, 2005.
[236] G. Cardenas, G. Cabrera, E. Taboada, and S. Miranda, "Chitin characterization
by SEM, FTIR, XRD, and13C cross polarization/mass angle spinning NMR,"
Journal of Applied Polymer Science, vol. 93, pp. 1876-1885, 2004.
[237] C. Ray, B. L. Sachs, B. K. Norton, E. S. Mikkelsen, and N. A. Clausen,
"Prosthetic disc nucleus implants; an update," in Lumbar disc herniation, R.
Gunzberg and M. Szpalski, Eds., ed Philadelphia, Pennsylvania: Lippincott
Wiliams & Wilkins, 2002.
[238] J. Husson, A. Korge, J. L. Polard, T. Nydegger, S. Kneubuhler, and H. M. Mayer,
"A memory coiling spiral as nucleus pulposus prosthesis," Journal of Spinal
Disorders and Techniques, vol. 16, pp. 405-411, 2003.
[239] U. Yuksel, S. Walsh, D. Curd, and K. Black, "Fatigue durability of a novel disc
nucleus repair system: in vitro studies in a calf spine model. ," The Spine Journal,
vol. 2, pp. 47S-128S, 2002.
[240] D. A. Bichara, X. Zhao, N. S. Hwang, H. Bodugoz-Senturk, M. Yaremchuk, M.
Randolph, and O. Muratoglu, "Porous Poly(vinyl alcohol)-Alginate Gel Hybrid
271
Constuct for Neocartilage Formation Using Human Nasoseptal Cells," Journal of
Surgical Research, vol. 163, pp. 331-336, 2010.
272
VITA
Valerie Regina Binetti was born in Pittsburgh, Pennsylvania to Louis and Gloria Binetti
on August 7th, 1980. After completing her secondary education at Corning-Painted Post
West High School in Painted Post, New York she attended Corning Community
College in Corning, New York where she graduated with an Associate of Science in
Math and Science in 1999. Valerie earned her Bachelor of Science in Materials Science
and Engineering in 2002 at Virginia Polytechnic Institute and State University in
Blacksburg, Virginia. Following her undergraduate studies, Valerie started her career at
Lockheed Martin Missiles and Fire Control in Orlando, Florida and then continued
working as an engineer in the aerospace field working for The Boeing Company in
Ridley Park, Pennsylvania. After working in industry for five years, Valerie began her
graduate studies at Drexel University in 2007. Valerie earned a Master of Science
working with Caroline Schauer, Ph.D. and then pursued doctorate research working for
Anthony Lowman, Ph.D.
During her graduate education, Valerie was a National Science Foundation Teaching
Fellow for K-12 education, a Drexel University College of Engineering Dean’s Fellow
and a Drexel University Provost Fellow. Valerie received Honorable Mention in the
National Science Foundation Graduate Research Fellowship.