novel associated hydrogels thesis

8/13/2019 Novel Associated Hydrogels Thesis

1/122

Novel Associated PVA/PVP Hydrogels for Nucleus Pulposus Replacement

A Thesis

Submitted to the Faculty

of

Drexel University

by

Jonathan D. Thomas

in partial fulfillment of the

requirements for the degree

of

Master of Science

in

Materials Engineering

September 2001


2/122

ii

Acknowledgments

This work represents two years of research. During this time, I have been blessed

with two advisors: Michele Marcolongo and Anthony Lowman. Their confidence in my

abilities and encouragement with my research has helped me grow as a researcher. I

would like to also acknowledge and thank Thomas Dziubla, Xinyin Liu, Robert Murray,

and Arvind Sivasubramanian for their friendship and the knowledge I gained from many

hours of discussing experiments and results. Lastly, I would like to thank my Mother,

Father, and Brother for their love and support they have always shown me.


3/122

iii

Table of Contents

List of Tables....................................................................................................................... vList of Figures ....................................................................................................................viAbstract ............................................................................................................................... x

1 Introduction ................................................................................................................. 12 Background ................................................................................................................. 32.1 Intervertebral Disc............................................................................................... 3

2.1.1 Structure ...................................................................................................... 32.1.2 Mechanics.................................................................................................... 5

2.1.2.1 Disc Kinematics ...................................................................................... 52.1.2.2 Static Material Properties ........................................................................ 62.1.2.3 Viscoelastic Behavior.............................................................................. 8

2.2 Degenerative Disc Disease................................................................................ 152.3 Treatment Options............................................................................................. 19

2.3.1 Conservative Treatments........................................................................... 19

2.3.2 Surgical Procedures................................................................................... 192.4 Intervertebral Disc Prostheses........................................................................... 222.4.1 Total Disc Replacements........................................................................... 232.4.2 Nucleus Pulposus Replacement ................................................................ 282.4.3 Tissue Engineering Approach ................................................................... 34

2.5 PVA/PVP Hydrogels......................................................................................... 343 Experimental - Materials Synthesis and Testing Methods........................................ 38

3.1 Hydrogel Synthesis ........................................................................................... 383.2 Equilibrium Swelling Analysis ......................................................................... 393.3 Chemical Analysis............................................................................................. 403.4 Network Characterization with Rubber Elasticity Theory................................ 403.5 Thermal Analysis .............................................................................................. 423.6 Statistical Analysis ............................................................................................ 42

4 Results ....................................................................................................................... 444.1 Swelling Characterization ................................................................................. 444.2 Dissolution of Hydrogels .................................................................................. 47

4.2.1 Mass Loss Analysis................................................................................... 474.2.2 Chemical Analysis..................................................................................... 49

4.3 Mechanical Analysis ......................................................................................... 554.3.1 Youngs Modulus...................................................................................... 554.3.2 Network Characterization with Rubber Elasticity Theory........................ 58

4.4 Thermal Analysis .............................................................................................. 63

4.5 Effects of wM on Hydrogel Properties............................................................. 684.5.1 Equilibrium Swelling Characterization..................................................... 684.5.2 Mass Loss Analysis................................................................................... 694.5.3 Youngs Modulus...................................................................................... 694.5.4 Network Characterization with Rubber Elasticity Theory........................ 704.5.5 Thermal Analysis ...................................................................................... 71

5 Discussion ................................................................................................................. 875.1 Equilibrium Swelling ........................................................................................ 87


4/122

iv

5.2 Dissolution of Hydrogels .................................................................................. 875.2.1 Mass Loss Analysis................................................................................... 875.2.2 Chemical Analysis..................................................................................... 88

5.3 Mechanical Analysis ......................................................................................... 885.3.1 Youngs Modulus...................................................................................... 88

5.3.2 Network Characterization with Rubber Elasticity Theory........................ 895.4 Thermal Analysis .............................................................................................. 89

5.5 Effects of wM on Hydrogel Properties.............................................................. 905.6 Summary ........................................................................................................... 93

6 Conclusions ............................................................................................................... 957 Recommendations for Further Research ................................................................... 97

7.1 Processing of Hydrogels.................................................................................... 977.2 Quantitative Analysis of Hydrogen Bonding.................................................... 977.3 New Criteria Needed for Nucleus Pulposus Replacements .............................. 987.4 Drug Delivery Capabilities of PVA/PVP Hydrogel Blends ............................. 99

List of References............................................................................................................ 100


5/122

v

List of Tables

Table 1. Projections for discectomy and fusion surgeries in the United States forthe year 2000 [32] .............................................................................................. 22

Table 2. PVA/PVP polymer blends that were prepared................................................... 38

Table 3. Molecular weight series ..................................................................................... 68

Table 4. Comparison of the properties of blends prepared with 1% PVP ....................... 92

Table 5. Design chart for PVA/PVP hydrogel blends...................................................... 94


6/122

vi

List of Figures

Figure 1. The anatomy of the lumbar spine. Each of the intervertebral discs arespaced between 2 vertebral bodies. The discs are named accordingly[34]. ................................................................................................................. 10

Figure 2. A lateral cross section of a lumbar segment (vertebrae / disc / vertebrae)shows the position of the nucleus pulposus and the annulus layers [34]. ....... 11

Figure 3. An anterior / posterior cross section of an intervertebral disc shows theregions of the annulus fibrosus and nucleus pulposus [13]............................. 12

Figure 4. Axial compression causes hydrostatic pressure within the nucleus toplace the annulus layers under tension [15]. ................................................... 13

Figure 5. The position of the nucleus pulposus changes under lateral bending and

flexion/extension motions. Note the opposing tensile/compressiveforces present in each case [15]....................................................................... 14

Figure 6. The nerve roots exit the spinal cord through the neuroforamina whichare openings between the facet joints of 2 adjacent vertebral bodies[34]. ................................................................................................................. 17

Figure 7. The nucleus depicted in this cross section of an intervertebral disc hasmigrated into the annulus layers posteriorlaterally resulting in anannular bulge and nerve irritation [34]............................................................ 18

Figure 8. (a) Vertebral spinal unit (b) with detail of intervertebral disc.Prosthetic disc design by: (c) Fernstrom, (d) Hedman and Kostuik, (e)Patil, (f) Buttner-Janz, (g) Steffee et al., (h) Lee and Parsons, (i)Froning, (j) Ray [32]. ...................................................................................... 32

Figure 9. Radiographic images of SB Charite disc designed by Buttner-Janz showthe position of the polyethylene core between 2 Co-Cr-Mo endplates.Core position can be determined by the radiographically opaque metalring that surrounds the core [15]. .................................................................... 33

Figure 10. Interchain Hydrogen bonding within a PVA/PVP blend occursbetween carbonyl groups on PVP and hydroxyl groups on PVA. .................. 37

Figure 11. The device that was used to cut PVA/PVP films into tensile strips wasformed from four Feather Disposable Scalpels that were gluedtogether. A natural rubber cutting surface was used. ..................................... 43


7/122

vii

Figure 12. Mass swelling coefficient as a function of time of immersion. Blendsprepared with higher portions of PVP swelled to a greater extenthowever blends prepared with excess PVP suffered dissolution andfailed to reach a stable equilibrium swelling................................................... 46

Figure 13. Polymer mass loss after 120 days swelling as a function of PVPcomposition. Polymer mass loss reached a minimum at 0.75% and 1%PVP.................................................................................................................. 48

Figure 14. Typical ATR-FTIR spectra for PVA/PVP blend. The spectra shownwas prepared with 95% PVA / 5% PVP. The four noted absorbancepeaks are labeled in cm-1. ................................................................................ 51

Figure 15. ATR-FTIR Spectra of PVA/PVP blends (143K PVA / 10K PVP) priorto swelling. Intensity of the free C=O peak increases with PVP%.The hydrogen bonded C=O peak is not detectable above 5% PVP due

to its relative intensity to the larger and broader free C=O peak. ................... 52

Figure 16. Change in peak height ratio of blend spectra over 56 days immersionin vitro. PVP not incorporated within the polymer network is free todissolve into solution. An equilibrium value of free C=O in each gel isreached after 56 days immersion..................................................................... 53

Figure 17. Fractional loss of free carbonyl bonds as a function of PVPcomposition. The fraction loss was determined by the change in freeC=O peak height ratio with ATR-FTIR spectra analysis................................ 54

Figure 18. Youngs Modulus (E) was calculated by determining the slope for theplot of tensile stress as a function of tensile strain. ......................................... 56

Figure 19. Youngs Modulus of PVA/PVP polymer blend thin films as a functionof PVP composition after various times of immersion in vitro....................... 57

Figure 20. Sample plot showing how tensile modulus (G) was determined. Theslope was measured between 0 and 8.3% strain.............................................. 60

Figure 21. Effective molecular weight between crosslinks ( cM ) after 14, 28, and56 days of immersion for 143K PVA / 10K PVP blend series. ...................... 61

Figure 22. Crosslinking density (X) after 14, 28, and 56 days of immersion for143K PVA / 10K PVP blend series................................................................. 62

Figure 23. DSC scans of PVA/PVP blends (143K PVA / 10K PVP) showingchanges in Tmwith PVP composition. ............................................................ 64


8/122

viii

Figure 24. Typical analysis of melting in a DSC scan with integration performedwith a changing baseline using Universal Analysis software provided

by TA Instruments. ...................................................................................... 65

Figure 25. Melting points (Tm) of PVA/PVP blends (143K PVA / 10K PVP) prior

to immersion as a function of PVP composition............................................. 66

Figure 26. Changes of enthalpy at melting (Hmelt) of PVA/PVP blends (143KPVA / 10K PVP) prior to immersion as a function of PVPcomposition. .................................................................................................... 67

Figure 27. Mass swelling coefficient (q) for pure PVA hydrogels as a function ofimmersion time................................................................................................ 72

Figure 28. Mass swelling coefficient (q) for 99% PVA / 1% PVP hydrogel blendsas a function of immersion time...................................................................... 73

Figure 29. Mass swelling coefficient (q) for 95% PVA / 5% PVP hydrogel blendsas a function of immersion time...................................................................... 74

Figure 30. Mass swelling coefficient (q) for 90% PVA / 10% PVP hydrogelblends as a function of immersion time. ......................................................... 75

Figure 31. Mass swelling coefficient (q) for 75% PVA / 25% PVP hydrogelblends as a function of immersion time. ......................................................... 76

Figure 32. Mass swelling coefficient (q) for 50% PVA / 50% PVP hydrogel

blends as a function of immersion time. ......................................................... 77

Figure 33. Polymer mass loss after 120 days swelling as a function of PVP

composition. Polymer mass loss was not dependent on PVP wM . ............... 78

Figure 34. Influence of PVP wM on Youngs Modulus for 143,000 wM PVAafter 56 days immersion. ................................................................................. 79

Figure 35. Influence of PVP wM on Youngs Modulus for 95,000 wM PVAafter 56 days immersion. ................................................................................. 80

Figure 36. Influence of PVA wM on Youngs Modulus for 10,000 wM PVPafter 56 days immersion. ................................................................................. 81

Figure 37. Influence of PVA wM on Youngs Modulus for 40,000 wM PVPafter 56 days immersion. ................................................................................. 82


9/122

ix

Figure 38. Effective molecular weight between crosslinks ( cM ) after 56 days ofimmersion as a function of PVP composition................................................. 83

Figure 39. Crosslinking density (X) after 56 days of immersion as a function ofPVP composition............................................................................................. 84

Figure 40. Melting Point (Tm) for PVA/PVP blends as a function of PVPcomposition. .................................................................................................... 85

Figure 41. Enthalpy change due to melting (Hmelt) for PVA/PVP blends as afunction of PVP composition. ......................................................................... 86


10/122

x

Abstract

Novel Associated PVA/PVP Hydrogels for Nucleus Pulposus ReplacementJonathan D. ThomasMichele Marcolongo

Anthony Lowman

Degenerative disc disease in the lumbar spine is marked by a dehydration of the

intervertebral disc and loss of biomechanical function of the spinal unit. Since the

current surgical procedures are ineffective in restoring natural biomechanical function

back to the diseased disc, researchers have looked to replace the intervertebral disc.

These designs are flawed in that they either dont restore natural movement back to the

spinal unit, require surgeries that are highly invasive, or they further promote disc

degeneration of adjacent spinal levels. Recently, researchers have sought to only replace

the central portion of the disc called the nucleus pulposus. A potential nucleus

replacement could mimic a healthy nucleus pulposus in restoring healthy biomechanical

function to the spinal unit. A hydrogel, poly(vinyl alcohol) (PVA), has been investigated

to serve as a nucleus replacement. However, semicrystalline PVA suffers dissolution

under physiological conditions.

Blends of PVA and poly(vinyl pyrrolidone) (PVP) may provide a material that is

a suitable nucleus pulposus replacement. Through interchain hydrogen bonding, they

possess greater stability than pure PVA hydrogels. This research is an examination of the

stability of these gels under physiological conditions. Polymer dissolution and stability

were studied with equilibrium swelling analysis over 120 days immersion, ATR-FTIR

analysis over 56 days immersion, and mechanical tensile testing over 56 days immersion.

Rubber elasticity theory was used to combine mechanical results with swelling data to


11/122

xi

calculate network characteristics such as the molecular weight between crosslinks and

density of crosslinks. DSC studies after polymer blend preparation were used to illustrate

relative degrees of physical crosslinking. Properties were examined as a function of

PVA/PVP composition as well as PVA wM and PVP wM .

Results indicate that PVA/PVP blends prepared with moderate amounts of PVP

(0.5-5%) result in a polymer network stabilized through interchain hydrogen bonding

between hydroxyl groups on PVA chains and carbonyl groups on PVP chains. Most

notably, a significant decrease in percentage of polymer mass loss was seen for blends

prepared with 143K wM PVA. Additions of larger amounts of PVP to the blends

resulted in weaker gels that had more open network structures suffering from higher

amounts of polymer dissolution. ATR-FTIR results indicate that PVP unincorporated in

the network structure through hydrogen bonding suffers significant dissolution out of the

polymer network and into solution. wM of PVA and PVP are shown to have a

significant influence on the blends network properties. Lower wM PVA resulted in a

more stable blend containing a higher density of crosslinks. However blends prepared

with a higher wM PVA showed superior polymer network stability in dissolution studies.

Higher PVP wM was shown to have a better stabilizing influence on the higher wM

PVA that was tested. The blend that had the best combination of network stability under

physiological conditions and a relatively tight, stable, and crosslinked network was

prepared with 99% PVA (143K) and 1% PVP (40K).


12/122

xii


13/122

1

1 Introduction

Lower back pain is the most common and costly disabling musculoskeletal

ailment in the United States and is the second leading cause of lost work days next to

upper respiratory tract illness [1-3]. Studies have estimated that between 50-90% of the

adult population suffers from lower back pain [4, 5]. Treatment and compensation of the

ailment have been estimated to cost more than $50 billion per year in the United States

[6]. The impact of lower back pain on quality of life has also been considered

extensively. In many cases, pain usually subsides in a matter of weeks or months.

However, chronic lower back pain can become a permanent disabling symptom for some

patients. While the causes of lower back pain often remain unclear, it is estimated that

75% of the cases are associated with degenerative disc disease [7].

Each of the intervertebral discs, which are located between the spinal vertebrae,

are composed of a fluid-like region called the nucleus pulposus surrounded by oriented,

concentric layers of tissue collectively called the annulus fibrosus. The process of disc

degeneration is marked by a dehydration of the nucleus pulposus. The ability of the

nucleus to handle applied loads is decreased with dehydration causing the load supported

by the annulus to increase. Further damage can result in the nucleus migrating through

and delaminate the annulus layers. Complete penetration through the annulus fibrosus is

referred to as a herniation. Pain can originate from degenerative disc material that

stimulates the nerves exiting the spine.

There are many conservative treatment options for lower back pain. They share

the common goal of reducing the pain that stems from inflammation and nerve

impingement. Conservative options do not aim to restore biomechanical function to the


14/122

2

disc. Surgical options, including discectomy and lumbar fusion, can be performed when

lower back pain is not successfully relieved with more conservative treatments. These

procedures can offer pain relief but they do not restore natural biomechanical function to

the spine. Both of the surgical procedures can result in accelerated degeneration of

adjacent discs.

The goal of this project is to investigate a PVA/PVP blend as a potential material

for a nucleus pulposus substitute. The effects of PVA / PVP composition and wM on the

hydrogel structure and stability will be studied in vitro. A device designed with these

blends has the potential to return biomechanical function to the degenerated spinal unit

and relive pain for the patient. Under compression, the hydrogel nucleus substitute is

designed to expand radially, placing the annulus layers under tension, similar to the way a

healthy nucleus mechanically functions. These hydrogel blends are stable within the

physiological environment due to physical crosslinks consisting of intermolecular

hydrogen bonds between PVA and PVP and intramolecular hydrogen bonds within PVA

crystals. Both PVA and PVP have proven biocompatibility. A hydrogel design allows

for the arthroscopic insertion of the dehydrated nucleus substitute and subsequent

rehydration, once in the body.


15/122

3

2 Background

2.1 Intervertebral Disc

2.1.1 Structure

The human spine consists of vertebral bodies separated by intervertebral discs.

Figure 1 shows the lumbar section of the spine. Discs are labeled with respect to the two

vertebral levels which they are located between. For example, the L4/L5 disc is located

between the 4thand 5thlumbar vertebrae. S1 represents the sacrum, which is the vertebral

body located most inferior along the spinal column. The sacrum is located below the

lumbar section of the spine. The discs allow motion of the vertebral bodies and distribute

applied loads over the total area of the vertebral bodys cartilaginous endplates.

Intervertebral discs have a fluid-like region called the nucleus pulposus that is surrounded

by the slightly more fibrous annulus fibrosus as shown in Figure 2. The nucleus, in a

young healthy spine, is a gelatinous material with a consistency similar to toothpaste [8].

A healthy nucleus contains between 70-90% water by weight which is responsible for its

fluid-like properties [8, 9]. Proteoglycans make up for 65% of the dry mass of the

nucleus [8]. These macromolecules consist of sulfated glycosaminoglycan side-chains

covalently bonded to core proteins. Proteoglycan units can link together and form larger

aggregate molecules. These molecules have the ability to attract and retain water due to

ionic carbonyl and sulphate groups on the glycosaminoglycan chains [8, 10, 11].

Collagen, which constitutes 5-20% of the dry mass of the nucleus, holds the proteoglycan

aggregates together [8, 9]. The remaining dry portion of the nucleus consists of elastic

fibers and non-collagenous proteins.


16/122

4

The annulus fibrosus is an arrangement of concentric layers of oriented fibrous

tissue. Depending on location within the disc, the fibers are connected to the vertebral

endplates or directly to the vertebral body. Water accounts for 60-70% of the mass of the

annulus [8]. Within each annular layer, fibers lie parallel to each other at an orientation

that is 60-70with respect to the spine axis (20-30with respect to radial direction) as

shown in Figure 3 [9, 10, 12]. Fibers in adjacent layers in the annulus are oriented in

alternating directions, e.g. (...+65/-65/+65/-65...) [13]. Collagen accounts for 50-60%

of the dry mass of the annulus [8]. Proteoglycans, contributing about 20% to the dry

mass in the annulus, are located between collagen fibers in adjacent annular layers and

serve to bind them together [8]. Other components of the annulus, elastin fibers,

contribute about 10% to the dry mass of the annulus. They are concentrated towards the

superior and inferior portions of the annulus that contact the vertebral end plates [8, 9].

While the nucleus pulposus and annulus fibrosus appear to be very similar

chemically, differences are found in their structural organization. Collagen in the annulus

is mostly type I while the majority of collagen found in the nucleus is type II.

Additionally, the microstructure of the annulus and nucleus is dependent on location

within the disc. For example, the annulus layers adjacent to the nucleus and near the

edge of the disc exhibit different tensile properties. Also, the water content of the

nucleus varies as a function of distance from the center of the disc.

The vertebral endplates serve to separate the nucleus and annulus from the

vertebral bodies. The cartilage found on these endplates resembles the chemical structure

of the adjacent portion of disc [8]. The center portions of the endplates have a higher

water and proteoglycan content but less collagen content than the outer regions of the


17/122

5

endplates. The permeability of the central portion of the endplates plays a major role in

the diffusional transport of water and nutrients to the nucleus pulposus [14]. Water and

nutrients can reach the nucleus through the annulus as well but to a much lesser extent

than the endplate route [14].

2.1.2 Mechanics

2.1.2.1 Disc Kinematics

In a functional spinal unit, the nucleus, annulus, and vertebral bodies work

together to withstand the different load conditions that are placed on the spine. The basic

modes of movement of the spine are axial compression, flexion / extension, lateral

bending, and torsion. When the spinal unit is axially compressed as shown in Figure 4,

hydrostatic stresses cause the nucleus to expand in the radial direction, increasing the

tensile stress in the annular fibers. The tensile stresses in the annulus are greatest

adjacent to the nucleus and decrease gradually away from the nucleus. Axial

compression causes discs to bulge in the radial direction. Stress states within the disc are

more complex for bending or twisting movements. The annulus is subjected to

compressive stresses in the direction of the applied stress and tensile stresses in the

opposite direction. The nucleus will actually move to the direction of the tensile forces

(extension) in order to distribute the load across the disc as shown in Figure 5 [15].

Annular bulge preferentially occurs in the region of the disc that is under compressive

stresses [15]. Torsion affects each region of the annulus fibrosus differently. Anterior

and lateral regions undergo tensile stresses in annular layers in which the fibers are in the

direction of the axial torque while layers in the opposite direction are subjected to no

stress [15]. Annular layers in the posterior spine undergo tensile stresses when the disc is


18/122

6

subjected to torsion regardless of fiber orientation [12, 15]. Loading of the spine is often

a coupling of different movements.

2.1.2.2 Static Material Properties

Biomechanical research has focused on determining the material properties of the

intervertebral disc. Research has also concentrated on quantifying the stresses the disc is

subject to in vivoand ex vivo. The wide range of reported values for disc properties and

stress states are due in part to variations in analytical techniques and variations within

disc structure that are dependent on location within the disc [16-18]. Furthermore, disc

properties are dependent on patient age, disc location within the spine, and the degree of

disc degeneration [12, 15, 19, 20].

Variations in tensile modulus of the annulus have been reported to be as low as

0.2 MPa and as high as 645 MPa [18, 21]. Ebara et al. [18] found values ranging from 5

MPa to 50 MPa in their study observing location dependent variations in tensile

properties when they tested single annular layers from different radial and circumferential

positions within the disc. Similarly, they saw variations in failure stress and strain with

sample position. Keller et al. [16] found a similar pattern of dispersity in ultimate

compressive strength and compressive modulus when they tested the properties of

vertebral endplates. Umehara, et al. [22] used an indentation technique to find

compressive moduli for regions of the annulus and nucleus within the intervertebral disc.

The indentation results correlated to modulus values that ranged from 5 kPa to 210 kPa

[22]. However, indentation tests on the fibrous annular layers and fluid-like nucleus are

ineffective ways to determine modulus values. Variations in modulus seen with their

indentation test are more likely due to variations in the surfaces of the nucleus and the


19/122

7

annulus fibers. Indentation testing of fibrous tissue can leads to results that are error-

prone and cannot be correlated to a compressive modulus.

Previous researchers [12, 19, 23] have made attempts to characterize the

mechanical properties of the complete disc. The compressive strength of lumbar

intervertebral discs has been measured to be from 5,000 N to 8,000 N, however the

strength is a function of patient weight, disc position, and level of degeneration [12, 19].

One has to cautiously use these data because they are also dependent on a discs cross-

sectional area which varies with disc position and among a population of patients. Using

these values to determine the stress state within the nucleus is not advised because the

nucleus has been reported to take up anywhere between 30-60% of the cross-sectional

area of the disc [24]. Researchers have also measured a value called the stiffness

coefficient to characterize discs. It is defined as the load per instantaneous displacement

of the disc [23]. Stiffness coefficient values are subject to variation with disc geometry

and are not a complete representation of the disc material.

Intradiscal pressure is the measure of the hydrostatic pressure that is present in the

nucleus. Under a simple movement like compression of the intervertebral disc, a

hydrostatic pressure is created in the nucleus resulting in tension in the annulus layers. It

is an important measurement of the state of stress within a disc. Nachemson [25] was the

first to measure the intradiscal pressure within an intervertebral disc. Pressures were

measured with a transducer design of a needle equipped with a miniature pressure gauge

at its tip. He found that most discs have an internal hydrostatic pressure that is 1.3-1.5

times axial stress applied to the spinal unit [25]. When he used this pressure transducer

in vivo, Nachemson was able to determine the dependence that posture had on intradiscal


20/122

8

pressure [26]. Normal intranuclear hydrostatic pressure for a healthy L5/S1 disc has been

measured to be about 0.5 MPa [9, 27]. With physical exertion, the pressure was shown to

almost triple [9, 27]. The pressure within the disc is estimated to be 50% higher for

degenerated discs [28].

In biomechanical mathematical models, it has been assumed that the nucleus is an

incompressible, inviscid fluid that has a bulk modulus value of 2,200 MPa [29, 30].

However some have assumed that the nucleus is a poroelastic solid with a Youngs

modulus between 4.5-1,500 kPa and a Poissons ratio between 0.1 and 0.45 [21, 30].

Models based on inaccurate material property assumptions will only yield unreliable

results. Proper determination of properties of the materials within a function spinal unit

is important because it gives mathematical models their validity. Tadano et al. made the

observation that stress and strain states in the intervertebral disc calculated with a

mathematical model like finite element analysis are highly dependent on the chosen

properties of the materials [31].

2.1.2.3 Viscoelastic Behavior

There has long been an uncertainty in whether the nucleus is a solid or a fluid.

This debate stems from the viscoelastic nature of the disc. The nucleus of the disc acts as

a compressible fluid having a 2-4 hour time constant [32]. This time period is used for

the transport of fluid into and out of the disc [33]. Viscoelasticity allows the spinal unit

to be highly flexible under low loads and more rigid under higher loads [15]. It has been

suggested that collagen fibers take up an instantaneous load but if a load is applied over a

period of time, fluid transport will change the content of the nucleus [15]. The time


21/122

9

dependent creep properties of the disc are crucial to withstanding prolonged loads. Under

slow creep, the disc will eventually reach a final deformation under constant load.


22/122

10

Figure 1. The anatomy of the lumbar spine. Each of the intervertebral discs are spacedbetween 2 vertebral bodies. The discs are named accordingly [34].


23/122

11

Figure 2. A lateral cross section of a lumbar segment (vertebrae / disc / vertebrae) showsthe position of the nucleus pulposus and the annulus layers [34].


24/122

12

Figure 3. An anterior / posterior cross section of an intervertebral disc shows the regionsof the annulus fibrosus and nucleus pulposus [13].


25/122

13

Figure 4. Axial compression causes hydrostatic pressure within the nucleus to place theannulus layers under tension [15].


26/122

14

Figure 5. The position of the nucleus pulposus changes under lateral bending andflexion/extension motions. Note the opposing tensile/compressive forces present in eachcase [15].


27/122

15

2.2 Degenerative Disc Disease

The efficient transfer of stress from the nucleus to the annulus begins to change

around the age of 30 with a change in the concentration and types of proteoglycans in the

disc as well as a loss in the discs overall water content [11, 20, 35-37]. The normal

aging process causes the water content in the nucleus to fall below 70% [13]. The

distinction between the nucleus and annulus becomes unclear, most likely as a result of

the increase in collagen content and decreases in the number of proteoglycan side chains

and proteoglycan aggregates within the nucleus [13, 27]. The vertebral endplates become

sclerotic and loose their permeability reducing water and nutrient transport to and from

the disc. Metabolic products like lactic acid have no route to leave the disc leading to a

decrease in pH [20, 35]. Frymoyer and Moskowitz [20] put forth that loss of the

proteoglycan matrix in degenerative disc disease stems from cell death due to lower pH

levels within the disc.

Nucleus dehydration results in a decrease in the load carried by the nucleus and an

increase in load on the annulus [35]. In a degenerated spinal unit, stresses are transferred

to the outer regions of the vertebral endplates and intervertebral disc [15]. Tears, cracks,

and fissures can occur within the annular layers allowing the migration of the nucleus

through the annular layers and causing pain by stimulating the sinu-vertebral nerve [15,

35]. The spinal canal and spinal cord are located posterior to the intervertebral disc.

Nerves exit the spinal canal through openings called neuroforamina and travel to the

extremities as shown in Figure 6. Nuclear migration can also lead to disc bulge reducing

the height of the disc. Loss of disc height is a significant problem because it leads to a

narrowing of the nerve root openings and potential buckling in the surrounding ligaments


28/122

16

[13, 20, 35, 38]. Pain also originates from the increased motion stemming from laxity

that accompanies a reduction in disc height.

With continued degeneration, a portion of the nucleus can penetrate through the

annular layer in what is referred to as a disc herniation. This nuclear material or the

bulge created by the nuclear material can impinge on sensitive nerves surrounding the

disc as shown in Figure 7 and cause pain in the back or lower extremities. Furthermore,

the herniated material elicits an inflammatory response because the nucleus of a healthy

intervertebral disc is avascular [35]. The nuclei of intervertebral discs are the largest

avascular structures within the body. This means they receive nutrients solely from

diffusion transport mainly through the surrounding vertebral endplates rather than

through vascularized tissue [32]. The body attacks the herniated material because it has

become part of the vascular system. The reduction in disc volume leads to instability and

as a result other parts of the spine including the bones, ligaments and endplates thicken

and grow to compensate for this instability [35]. The reduction in space surrounding the

nerves is referred to as spinal stenosis [35]. Inflammations can also occur at the facet

joints, which are located posterior to the discs, because of wear generated by increased

compressive stresses [35].

Approximately 95% of nuclear protrusions occur either at the L4/L5 and L5/S1

levels due to the high stress states found in these discs [20, 39]. Water loss is also highest

in these two levels [11]. Degeneration has been described as a cascade process in

which vertebral segments adjacent to the diseased disc are placed under increased stresses

and as a result undergo an accelerated degenerative process [15, 35].


29/122

17

Figure 6. The nerve roots exit the spinal cord through the neuroforamina which areopenings between the facet joints of 2 adjacent vertebral bodies [34].


30/122

18

Figure 7. The nucleus depicted in this cross section of an intervertebral disc hasmigrated into the annulus layers posteriorlaterally resulting in an annular bulge and nerveirritation [34].


31/122

19

2.3 Treatment Options

2.3.1 Conservative Treatments

Pain can occur at any point during the degeneration process. A commonly

recommended initial treatment for lower back pain is bed rest. Bed rest is used to treat

lower back pain because the prone position naturally reduces intradiscal pressure. Also,

prolonged activity can cause inflammation to worsen. However, in acute conditions, bed

rest resulted in no significant pain relief when compared to a control group having no bed

rest [15]. Furthermore, prolonged bed rest can lead to muscle atrophy and cardiovascular

deconditioning.

Analgesics, anti-inflammatory drugs, muscle relaxants, sedatives, and

neurotrophic medications have been used to ease lower back pain. These drugs are

therapeutic in that they reduce inflammation and relieve pain, but they do not address the

deterioration of disc function. Other conservative treatments that have been used include

electrotherapy, magnetic fields, ultrasound, chiropractic manipulation, acupuncture, and

traction. However, the effectiveness of these treatments is still in question. Conservative

treatments cannot reverse the degenerative process within the disc. Their aim is to reduce

inflammation stemming from disc material pressing on the nearby nerves. Conservative

treatment options do not sufficiently provide pain relief for 33% of lower back pain

sufferers [15]. The only remaining option for these patients suffering from debilitating

chronic lower back pain is surgery.

2.3.2 Surgical Procedures

Surgery is considered when conservative options fail to relieve pain or when a

risk of nerve damage exists. Decompression of the intervertebral disc can involve cutting


32/122

20

away disc, bone, and ligament material that is applying pressure to spinal nerves. The

traditional discectomy surgery involves removing herniated disc material posteriorly.

This surgery may require removal of a part of the annulus and surrounding ligaments [5,

15, 35]. Endoscopic discectomy procedures, which became popular in the 1980s, involve

the use of X-Ray fluoroscopy imaging to remove herniated disc material through

posteriorlaterally placed cannulas [40, 41]. These minimally invasive procedures can be

performed on an outpatient basis without general anesthesia [41]. Operation through a

small cannula (4.5-6.0 mm) lessens the damage to the annulus, bone, and ligaments

during surgery [35]. Discectomy has been very successful in reducing pain attributed to

disc herniation because it involves removal of the disc material that is causing nerve root

compression [15, 35, 40]. However, pain relief in discectomy does not come from any

biomechanical improvements to the spine. In fact, it has been found that discectomy

leads to a high occurrence of disc height reduction [42]. Success rates have been found

to be very unpredictable. Hanley et al. [42] found that 14% of patients experienced

disabling lower back pain after discectomy surgery. Lee et al. [43] found that as many as

50% of patients have continued lower back pain after disc excision. The offending

portion of the disc is removed but biomechanical instability is introduced as well [44].

Discectomy does not restore function to the vertebral joint. The degeneration process

continues, often making further surgery necessary.

Lumbar interbody fusion is a procedure that is performed for sufferers of chronic

lower back pain stemming from an advanced instability of the intervertebral disc joint.

This highly invasive surgery involves complete removal of the painful disc and

replacement with a bone graft. Posterior fusion is another method used in fusing a spinal


33/122

21

unit. It involves the fixation of the posterior facets and lamina which are located

posterior to the disc while leaving the degenerated disc in place. Synthetic internal

fixation devices are sometimes used to stabilize the spinal level. Pain relief as a result of

lumbar fusion comes from the stabilization achieved through elimination of motion.

Studies show a wide range in success rates of spinal fusion, from 32% to 98%

[35]. This is in part due to a poorly defined definition of success after back surgery.

Spinal fusion often leads to newer problems such as bone graft donor site pain,

pseudoarthosis, and spinal stenosis [45]. Fusing two vertebral bodies alters the normal

biomechanical properties of the spinal unit by increasing motion and stresses on the

adjacent intervertebral joints [6, 35, 43]. The degeneration rate may be accelerated in

adjacent discs due to these increased stresses, creating more instability and pain [46].

Highly rigid internal fixation devices have been suspected of causing disuse osteoporosis

and stress shielding, which results in muscle atrophy and decrease in bone density [47,

48].

Ray [32] made projections for the number and cost of discectomy and fusion

surgeries for the year 2000 by combining census data with statistics from the National

Center for Health Statistics. These projections are found in Table 1.


34/122

22

Table 1. Projections for discectomy and fusion surgeries in the United States for the year2000 [32]

Surgery Discectomy Fusion

Number 254,000 65,000

% of Population 0.094 % 0.024 %

Cost / Surgery $40,000 $80,000

Failures 20,000 (8%) 9,800 (15%)

Total Cost $8.8 billion $6 billion

These data suggest that 1 out of 1064 members of the general U.S. population had

discectomy surgery while 1 out of 4167 underwent fusion surgery during the year 2000.

The total cost associated with each surgery includes the cost of procedure and post

operative care. The cost to repeat a failed procedure was included as well. Rays cost

projections do not include work compensation and other procedures. Bao and Yuan [6]

state that 700,000 spine procedures are performed in the United States each year. Based

on these figures the total cost caused by back pain and treatment is approximately $50

billion per year.

2.4 Intervertebral Disc Prostheses

Problems with the currently available surgical techniques have led researchers to

investigate the idea of a prosthetic intervertebral disc. Other joints susceptible to

degeneration, like the hip, knee, and knuckle, can be successfully replaced with

prostheses [43]. Artificial joint designs replaced the popular fusion procedures that were

once commonly performed for diseased knee and hip joints.


35/122

23

For any prosthetic intervertebral disc design, there are general requirements. The

design must involve biocompatible materials that will not cause local tissue reactions,

carcinogenesis, or organ toxicity [43, 48]. The wear resistance of the materials used in

the design need to be taken into consideration because wear particles can lead to a foreign

body response [48]. The fatigue properties of the material need to be sufficient to

withstand the number of loading cycles to which the intervertebral disc is subjected. The

mechanical properties of a prosthetic intervertebral disc, including stiffness, yield and

fatigue strength, and viscoelastic properties, should match or exceed that of a normal

healthy disc [43]. Kinematically, the joint should be stable, that is, not susceptible to

allowing movement that exceeds that allowed by a healthy functional spinal unit [2, 43,

48]. Finally, the prosthetic disc design should allow for a safe surgical implantation.

2.4.1 Total Disc Replacements

Total disc replacement has been considered for discs that have undergone

extensive degeneration. The prosthesis would serve as a preferable treatment to fusion

because it allows for biomechanical movement between two adjacent vertebral bodies.

The benefit of replacing both the nucleus and annulus is that the effectiveness of surgery

is not dependent on the integrity of the annulus or degree of degeneration [6]. However,

total intervertebral discs replacements, which are mostly multicomponent designs, are

susceptible to problems such as interfacial bonding and wear [6]. Total intervertebral

disc implants must also have adequate fixation to the vertebral endplates.

For comparative purposes, Figure 8a shows an image of a healthy functional

intervertebral disc between two vertebral bodies. Figure 8b shows the position of a

native intervertebral disc between outlines of the vertebral bodies. One of the earliest


36/122


37/122

25

by multiple stainless steel springs [52]. The design was to direct stresses away from the

intervertebral disc with cup-shaped endplates. Lee et al. [43] suggested that this design

may present many problems such as wear debris and fatigue failure of the springs. Bao

and Yuan report that short term testing of an artificial disc made with two Ti-6Al-4V

springs in a sheep model showed no tissue ingrowth into the springs or hinges [6]. Lee et

al. [43] state that spring system designs such as those proposed by Kostuik and Patil

grossly oversimplify spinal motion and are subject to tissue interpenetration that can

potentially disrupt motion.

The SB Charite disc, developed by German orthopedic surgeons, Buttner-Janz

and Schellnack in conjunction with Link has been the design that has undergone the

most clinical trials [53-55]. The latest generation of this disc is shown in Figure 8f.

Radiographic images of this device are shown in Figure 9. The multicomponent device

consists of an ultra-high molecular weight polyethylene (UHMWPE) sliding core

positioned between two Co-Cr-Mo endplates. The device does not attempt to mimic the

natural discs mechanical stiffness. Kostuik [48] claims that the SB Charite disc design

is unconstrained with a center of rotation that is too anterior and projects that the

polyethylene core will undergo cold flow in about 4 years. Lee et al. [43] suggests that

dislocation in vivois probable due to the unconstrained nature of the sliding polyethylene

core of the SB Charite disc. Early clinical tests on the first SB Charite discs showed high

occurrence of problems such as core or plate dislodgement and core fracturing [32]. The

tests were halted and the device was redesigned. Over 2000 SB Charite prosthesis have

been implanted in patients since 1984 [6]. Griffith et al. [56] performed the most

extensive clinical trial on the third generation SB Charite disc and found significant


38/122

26

decreases in patient back and leg pain following implantation of the disc on an analog

scale. European clinical trials over 2 to 5 years show satisfactory results in 63% of the

patients with the prosthesis [57]. In one clinical study, problems with the implant seemed

to be centered around the endplate/implant interface [58]. David [58] claims that the SB

Charite disc does not match the anatomical nature of the vertebral endplates. It appears

that no studies have been performed to characterize the wear that potentially can occur

between the surfaces of the metal endplates and polyethylene core. However,

polyethylene wear particles generated in UHMWPE artificial hips have been shown to

cause inflammation and osteolysis [59].

Researchers began to evaluate soft polymeric materials due to their mechanical

similarity to the natural disc material. It is generally believed that better tissue tolerance

will exist with implant materials that are closer in mechanical properties to the natural

disc [43]. A design by Stubstad and Urbaniak [60, 61] consisted of three layers, a

silicone core sandwiched between two layers of Dacronmesh embedded silicone. The

purpose of the Dacroncovering on the superior and inferior caps was to promote tissue

ingrowth and implant fixation [60, 61]. The disc was tested in the chimpanzee model,

where infection and resorbtion of the adjacent bone was reported [61]. Furthermore,

dislodgement of the prosthesis occurred if there was not an exact fit [61]. Edeland [62]

proposed a silicon design that was capped with polyethylene base plates. Another design

by Edeland [63] involved a silicone nucleus design constrained by silicone or

polyethylene annulus replacements on the bases and outer circumference.

The Acroflexdisc, developed by Steffee [64, 65] in collaboration with Acromed

Corporation, has been implanted in humans in the United States. The implant, shown in


39/122

27

Figure 8g, is composed of a hexane-based rubber bonded to two Ti-6Al-4V endplates that

are sintered with Ti beads [64, 65]. Out of 6 patients who underwent replacement

surgery with the Acroflexdisc, 4 had satisfactory results [66, 67]. The two discs that

failed suffered fractures of the rubber core [66, 67]. The Acroflexdisc was the first

intervertebral disc prosthesis approved by the FDA to undergo clinical trials under the

Investigational Device Exemption [35]. However, in 1990, 2-mercaptobenzothiazole, a

chemical used in the vulcanization process of the rubber core, was found to be

carcinogenic in rats and clinical use was stopped [66, 67].

Another polymeric disc, designed by Lee and Parsons [43, 45, 68, 69], is shown

in Figure 8h. They designed a device that would replace all of the disc except for the

outermost annular layers. Their composite design was composed of a polysiloxane

modified styrene-ethylene/butylenes copolymer (C-Flex) core to mimic the softness of

the nucleus [68, 70, 71]. They surrounded the core with a stiffer polyurethane reinforced

with Dacron fiber lamina to mimic the annulus [68, 70, 71]. They used stiffer

polyurethane endplates coated with a woven jacket to frame the core [68, 70, 71]. The

jacket has also been coated with hydroxyapatite to promote bone ingrowth from the

vertebral endplates [68, 70, 71]. Bao and Huan state that a lack of fixation between the

implant and the vertebral bodies is what has prevented the Lee device from being tested

clinically [6].

Ambrosio et al. investigated the potential use of poly(2-

hydroxyethylmethacrylate) / polycaprolactone (PHEMA/PCL) embedded with

poly(ethylene therephthalate) (PET) fibers [72]. When studied under compression,


40/122

28

increases in PCL and PET content increased the implants mechanical strength and

stiffness, which they found to be close to that of a canine intervertebral disc [72].

2.4.2 Nucleus Pulposus Replacement

Researchers began to realize the potential benefits of only replacing the

dehydrated nucleus. The approach entails a less invasive posterior surgery as well as a

return of the annulus to its healthy natural tension [5]. This approach would not be

suitable for cases in which the disc is in the later stages of degeneration. Because these

devices are not intended to be fixed to the vertebrae, no problems with endplate fixation

exist for a nucleus replacement. Also, the surgical time required for a nucleus

replacement procedure should approach the time required for a discectomy in contrast to

the longer time required for more invasive complete disc replacements using the

prostheses described previously [6].

Garcia showed that a polyurethane elastomer can be polymerized in situwithin a

nucleus cavity at physiological temperatures rather than at the traditional processing

temperature of 100C [73]. Since polymerization is an exothermic process, localized

heating of surrounding tissue needs to be considered. In addition, injection pressure

needs to be closely monitored to make sure the nuclear cavity is filled and that the

monomer and oligomer solution doesnt herniate through the annulus prior to curing.

Schulmann [74] saw this problem with his work with in situ polyurethane

polymerization.

Hou [75] tested a silicone rubber as a nucleus replacement in vitro and in

monkeys. The results from the monkey model showed no adverse reactions from

surrounding tissue [75]. Froning worked on designs involving fluid filled bladders that


41/122

29

more closely mimicked the nucleuss fluid properties [76]. One of his designs is shown

in Figure 8i. However, fluid filled bladders pose the problem of critical rupture, as is

often seen with Silicone and Saline breast implants, and therefore would not be suitable

nucleus prostheses.

In the late 1980s, Ray and Corbin [32, 77-79] designed cylindrical implants to be

inserted after removal of the nucleus. The side by side implants are shown in Figure 8j.

They consisted of fiber-weaved shells of biodegradable poly(glycolic acid) (PGA) that is

filled with a thixotropic gel such as hyaluronic acid once it is positioned during surgery

[32, 77-79]. With swelling, the shell degrades as tissue begins to penetrate and grow.

The implants have been investigated for controlled release of therapeutic agents such as

anti-inflammatory drugs [79]. It was believed that difficulty with sealing the fluid into

the capsules caused them to abandon this design [6]. In the mid 1990s, Ray et al. [80, 81]

embarked on another design that involved side by side implants positioned in the medial-

lateral position rather than anterior-posterior like the earlier design. A polyacrylonitrile

(PAN) hydrogel surrounded by flexible but inelastic extended polyethylene fiber was

proposed [80, 81]. The size of the implant requires a hole to be cut in the annulus larger

than incisions commonly used for discectomy [6]. It has been reported that out of 101

patients who received the implant, 17 suffered implant extrusion [6]. They reported that

for the patients that did not suffer implant extrusion, pain relief was common [6]. One

potential problem with implants consisting of multiple cylindrical components is that

they are free to move around within the area left by the removal of the native nucleus

until tissue ingrowth occurs. Improper positioning of these two swollen cylinders can

change the biomechanical behavior. Their shape characteristics do not ensure the transfer


42/122

30

of compressive stresses into tensile stresses in the annulus fibrosus, as it occurs in a

healthy intervertebral disc.

Bao and Higham developed a poly(vinyl alcohol) (PVA) nucleus pulposus

replacement that restores function to the intervertebral disc by mimicking both the

mechanical and physiological properties of the disc [6, 35, 82-87]. The hydrogel

material, containing 70% water, acts similarly to the nucleus in that it absorbs and

releases water depending on the applied load [6, 35, 83]. Bao [87] has proposed methods

for dehydrating the hydrogel nucleus substitutes to facilitate a less invasive insertion into

the spine. Bao and Higham have proposed forming a nucleus replacement from two

cylindrical segments similar to Rays design to ease insertion and reduce the size of the

hole needed in the annulus [86]. A baboon test model of the PVA nucleus showed no

adverse local or systematic tissue reaction [6]. They report the PVA nucleus substitute to

have a compressive modulus greater than 4 MPa and a compressive strength greater than

1 MPa [83]. Prosthetic implants using PVA should not be considered stable within the

physiological environment due to the fact that PVA is a semicrystalline, hydrophilic

polymer that can undergo dissolution. The dissolution process involves an unfolding of

PVA crystal chains that join the amorphous region of the polymer, disentangle, and

eventually dissolve [88, 89]. Polymer chain dissolution results in a network with a

decreased mechanical stiffness resulting from a larger network mesh size. Larger

crystals, which undergo a slower dissolution process, are found in semicrystalline PVA

hydrogels that have higher PVA molecular weights [88].

Felt et al. [90, 91] have investigating a disc design involving the injection of

curing polyureathane in situ in conjunction with a balloon catheter and delivery balloon.


43/122

31

The design was tested with a human cadaver model and showed that the device is able to

restore disc height and disc modulus [6]. Bao and Yuan have also proposed the use of an

aperture sealing device in conjunction with an implantable PVA nucleus replacement

[92].


44/122

32

(a) (b)

(c) (d) (e) (f)

(g) (h) (i) (j)

Figure 8. (a) Vertebral spinal unit (b) with detail of intervertebral disc. Prosthetic discdesign by: (c) Fernstrom, (d) Hedman and Kostuik, (e) Patil, (f) Buttner-Janz, (g) Steffeeet al., (h) Lee and Parsons, (i) Froning, (j) Ray [32].


45/122

33

Figure 9. Radiographic images of SB Charite disc designed by Buttner-Janz show theposition of the polyethylene core between 2 Co-Cr-Mo endplates. Core position can bedetermined by the radiographically opaque metal ring that surrounds the core [15].


46/122

34

2.4.3 Tissue Engineering Approach

Stone [93] has attempted to regenerate the intervertebral disc by constructing a

scaffold of biocompatible and bioresorbable glycosaminoglycan fibers. Cells penetrate

and grow into the scaffold while the scaffold dissolves within the body. A tissue

engineering approach to repairing the intervertebral disc has also been developed by Gan

and Ducheyne [94, 95]. They implanted nucleus pulposus cells onto PLGA (polylactide-

co-glycolide) and bioactive glass substrates [94]. Their results suggest that both the

PLGA and bioactive glass substrates had cell adhesion and proliferation. The bioactive

glass with a calcium phosphate rich layer was found to induce cellular activity much

better than the PLGA [94]. Contrary to their promising results, the avascular

characteristics of the nucleus region of the intervertebral disc make the tissue-engineering

approach impractical. Cell migration, adhesion and growth onto a scaffold are not likely

to occur in a nuclear cavity that is devoid of vascularized tissue.

2.5 PVA/PVP Hydrogels

PVA is a polymer that has been studied extensively for potential biomedical

applications. The swelling, chemical, and mechanical properties of PVA were

determined in the late 1970s by Peppas and Merrill [96, 97]. Oka et al. [98] found no

inflammatory or degenerative changes in the articular cartilage or synovial membrane

surrounding their artificial PVA cartilage after 8-52 weeks. PVA hydrogels can be

produced from solution via repeated freezing and thawing cycles that increase the order

of the crystals, changing the dissolution properties, mesh size, and diffusion properties of

the polymer [99-101]. Mondino et al. [102] recently showed that gamma irradiation

treatments increase the tensile strength of PVA hydrogels. The ability of PVA to release


47/122

35

therapeutic drugs from its polymer network has been studied extensively [89, 103-105].

PVA has recently been investigated as a potential keratoprostheses [106] and as a

promising material for a bioartificial pancreas design [107].

Stammen et al. [108] has proposed using a freeze-thawed PVA hydrogel in a

number of applications including cartilage replacement and spine disc replacement. They

characterized the mechanical properties of these hydrogels crosslinked by the freeze-thaw

process. They saw an increase in tangent compressive modulus between 1-18 MPa from

10-60% strain [108]. They also found shear tangent modulus in the range of 0.1-0.4

MPa, depending on strain magnitude [108].

PVP is a hydrogel that has been used for a number of biomedical applications. It

was used as a colloidal plasma substitute in World War II [109] and has been used in soft

contact lenses [59]. It has recently been investigated by Kao et al. for applications such

as single-layer hydrogel wound dressings and tissue adhesives [110]. Risbud and Bhat

[111] evaluated the biocompatibility of PVP / -chitosan hydrogel membranes. They

found the membranes to be biocompatible. Risbud and Bhat [111] also found that the

additions of PVP gave the hydrogel polymer network increased strength due to the

viscoelastic properties of polymer [111].

Characterization of PVA/PVP copolymer blends has shown that a higher PVP

molecular weight within the blend will lead to more interactions between PVA and PVP

and higher blend crystallinity [112]. The single glass transition temperature (Tg) value

observed for PVA/PVP blends verifies the blends miscibility [113]. Interactions

between PVA and PVP occur through interchain hydrogen bonding between the carbonyl

group of PVP and the hydroxyl group on PVA as shown in Figure 10. These interactions


48/122

36

have been studied with FTIR and NMR [114, 115]. Hydrogen bonding interactions

involving carbonyl and hydroxyl groups have also been seen in ethyl isobutyrate (EIB) /

4-ethylphenol (EPh) mixtures, poly(4-vinyl phenol) (PVPh) / poly(methyl methacryalate)

PMMA blends, and poly(4-vinyl phenol) (PVPh) / poly(vinyl pyrrolidone) PVP blends

[116-119].


49/122

37

Figure 10. Interchain Hydrogen bonding within a PVA/PVP blend occurs betweencarbonyl groups on PVP and hydroxyl groups on PVA.

O

H

O O O

H

O

H

N

N

N

N

O

O

O

N

N

O

O

O

PVA

PVP

InterchainHydrogenBonding


50/122


51/122

39

Dried polymers were weighed using a Denver InstrumentsM-120 balance with

a sensitivity of 0.0001g. Polymer solutions (10% w/w) were prepared by dissolving the

polymers in DI water. The beakers were sealed with DuraSealand placed in an oven at

90C for 12-24 hours. The polymer solutions were removed from the oven and stirred

for 10 minutes. This was followed by 10 minutes of sonication to remove air bubbles.

Each of the solutions was caste into polyethylene petri dishes and 8in x 12in

rectangular PMMA molds. The solutions were placed back in the oven set at 37C to

evaporate the DI water. Evaporation took between 36-48 hours to complete. Upon

drying, residual water was removed from the polymer films in a vacuum oven at 35C

with an absolute pressure of 5 in Hg.

3.2 Equilibrium Swelling Analysis

The polymer blends were swollen in DI water for 1-2 hours to form gels. Circular

discs were cut from the gels with a 8.35 mm ID brass hole punch and then dried in an

oven at 37C to evaporate the DI water that was introduced in swelling. The circular

discs were further dried in a vacuum oven overnight at 35C with an absolute pressure of

5 in Hg to further remove water from the network. The dry mass of the circular discs

(n=3) for each polymer was measured. The discs were then swelled with 30 ml of DI

water in glass jars. The discs were kept at 37C in a shaker bath shaking at 60 RPM for

the first 4 days. The discs were then kept at 37C in a waterbath. The masses of each

swollen gel were measured after 1 hr, 2 hr, 4 hr, 8 hr, 1 day, 2 days, 4 days, and 1 week

swelling and weekly thereafter up to 120 days. Solutions were changed after each

measurement. After 120 days of swelling in DI water, the circular thin film discs were


52/122

40

dried in an oven at 37C and then in a vacuum oven at 35C around an absolute pressure

of 5 in Hg. Dry masses of the discs were then measured.

The density of each of the polymers was calculated prior to swelling by

measuring the weight of the three discs in heptane. The mass of the disc measured in

heptane was compared to the mass of the dry polymer discs in air prior to swelling. The

following equation which is based on the buoyancy of the polymer in heptane was used

to calculated the density of the polymer blends:

( )

heptane drypolymer

dry heptane

! m!

m m

=

(3.1)

3.3 Chemical Analysis

Attenuated Total Reflectance - Fourier Transform Infrared Spectroscopy (ATR-

FTIR) tests were conducted to examine the chemistry of the blends using a Nicolet

(Madison, WI) model 560 with ATR attachment. Evaporated polymer blends were

placed on a ZnSe crystal for analysis. Samples were examined as prepared (n=3) as well

as following 8 weeks of swelling in Phosphate buffered saline solution (n=1). The

spectra of the as-prepared films were obtained with 4 cm-1 resolution (1.928 cm-1 data

spacing) and 1024 scans. The spectra of the samples that were swollen for 8 weeks and

then dried were obtained with 1 cm-1resolution (0.482 cm-1data spacing) and 1024 scans.

Spectra analysis was performed with the Nicolet OMNICsoftware package.

3.4 Network Characterization with Rubber Elasticity Theory

Six 60 mm x 90 mm sheets were cut out of the larger thin films that were caste for

each copolymer blend. The sheets were swollen in PBS (pH=7.4) at 37C in 500 ml

glass jars. The jars were shaken at 60 RPM in a 37C shaker water bath for 2 days. After


53/122

41

2 days, the jars were moved to a 37C water bath. After 2 days, 1 week, 2 weeks, 4

weeks, and 8 weeks of immersion in vitro, thin strips were cut from the swollen

hydrogels on a natural rubber surface with a cutting tool shown in Figure 11. The tool

was prepared by gluing four Feather Disposable Scalpels together with glue. The width

of the sample, determined by the distance between the scalpel blades, varied from 4.25

mm to 4.65 mm. The lengths of the tensile samples were all greater than 75 mm. Widths

of the tensile samples were dependent on the film thickness that remained after the caste

polymer blend was dried. The dimensions of the swollen strips adhered to the ASTM

Standard Test Method for Tensile Properties of Thin Plastic Sheeting [120]. Sample

width and thickness were measured with digital veneer calipers having a resolution of

0.01 mm.

Tensile tests were performed on the thin film strips for each of the blends after 2

days, 1 week, 2 weeks, 4 weeks, and 8 weeks of swelling in vitro. Tensile tests were run

on each blend (n=5) on an Instron Series Materials Testing System Series 4442

(Canton, MA) with a 500 N load cell. The distance between the tensile grips at the start

of each test was 50 mm. The thin film samples were extended 12.5 mm (25% extension)

at a rate of 5 mm/min as proposed by the ASTM standard [120]. Axial load and

displacement values were recorded at a rate of 20 pts/sec with a computer interface.

The tensile data was used to calculate the tensile modulus of the gels. Also, the

rubber elasticity theory was used to relate the tensile modulus and degree of swelling of

the gels with the characteristics of the hydrogel network structure, namely, average

molecular weight between crosslinks, cM , and crosslink density, X.


54/122

42

3.5 Thermal Analysis

The PVA/PVP blends were studied using a differential scanning calorimeter

(DSC) (TA Instruments (Newark, DE)). DSC tests were performed on samples that

were cut from the polymer film that was caste after the water had evaporated. The

experiments involved heating the polymer from 25C to 250C at a rate of 5C/min. The

mass of each sample (usually between 4.0 mg and 6.0 mg) was measured, encapsulated

within an aluminum pan and lid, and sealed with a standard die press. The melting point

and enthalpy change with melting were determined with Universal Analysissoftware.

3.6 Statistical Analysis

Selected results from the previously described tests were compared statistically

using a two sided analysis of variance and Students ttest. The sample sizes for each of

the characterization tests were stated in the previous sections. Compared results that

were found to be statistically different at the 95% confidence level are annotated with a

* signifying that p


55/122

43

Figure 11. The device that was used to cut PVA/PVP films into tensile strips wasformed from four Feather Disposable Scalpels that were glued together. A natural rubbercutting surface was used.


56/122

44

4 Results

4.1 Swelling Characterization

Swelling experiments involve measuring the mass and/or volume of a hydrogel

over an immersion time. A mass swelling coefficient, q, can be calculated by the

following equation:

dry

mq

m= (4.1)

mdry is the initial dry mass of the polymer disc and m is the mass as a function of time. A

plot of q vs. time for the blends having a PVA wM of 143,000 and a PVP wM of 10,000

is shown in Figure 12.

The pure PVA hydrogel and blends prepared with up to 25% PVP reached a

maximum in swelling after 1-2 days of immersion. Equilibrium swelling, which was

reached shortly thereafter, was sustained for 120 days in vitro (q=2.3 2.6). The range

seen in mass swelling coefficients corresponds to 57 62% water w/w. Addition of PVP

resulted in an increase in swelling coefficient. Blends prepared with 50% and 75% PVP

had a significantly higher maximum swelling; q=3.3 and q=4.6 respectively. However,

these 2 blends failed to reach a stable equilibrium over the immersion time. Blends

prepared with these higher percentages of PVP proved to be more brittle in the dry state

and more prone to tear when being handled during weighing after immersion.

The densities of the gels ranged from 1.25 1.29 g/ml. Density values of the dry

polymers were used to convert mass swelling coefficients to volume swelling coefficients

with the following equation:

2,rQ 1 (q 1) != + (4.2)


57/122

45

The reciprocal of the volume swelling coefficient, Q, is the polymer volume fraction, 2,s.

Polymer volume fractions are required in the calculation of network characteristics in

Section 4.3.2.


58/122

46

Figure 12. Mass swelling coefficient as a function of time of immersion. Blends

prepared with higher portions of PVP swelled to a greater extent however blendsprepared with excess PVP suffered dissolution and failed to reach a stable equilibriumswelling.

1

1.5

2

2.5

3

3.5

4

4.5

5

0 20 40 60 80 100 120

Time (days)

q

0

0.5

0.75

1

5

10

25

50

75

PVP %

50 % H20

75 % H20


59/122

47

4.2 Dissolution of Hydrogels

4.2.1 Mass Loss Analysis

The dry masses of the circular discs that were immersed in vitro for 120 days,

m120, were compared to the initial dry mass of the discs, mdry, to determine the extents of

dissolution within the polymer networks. Percentage of polymer mass loss was

calculated with the following equation:

dry 120

dry

m m% Polymer Mass Loss 100%

m

=

(4.3)

Figure 13shows a plot of polymer mass loss for the 143K PVA / 10K PVP blend.

Discs made from pure PVA (%PVP=0) had 6.9% mass loss over 120 days in

vitro. Small additions of PVP to the polymer network resulted in a decrease in mass loss

with the minimum percentage mass loss suffered by the blend prepared with 99% PVA

and 1% PVP (2.8%). Blends with 0.75% PVP and 1% PVP showed statistically

significant less dissolution than pure PVA (p < 0.05).


60/122

48

Figure 13. Polymer mass loss after 120 days swelling as a function of PVP composition.Polymer mass loss reached a minimum at 0.75% and 1% PVP.

*

*

0

10

20

30

40

50

6070

0 0.5 0.75 1 5 10 25 50 75

% PVP

%P

olymerMassLoss


61/122

49

4.2.2 Chemical Analysis

Three dry thin film samples of each polymer blend were tested with ATR-FTIR to

determine the content of the films after being crosslinked by evaporation. A typical

absorbance spectrum obtained for the PVA/PVP blends is shown in Figure 14. The

notable absorbance peaks of the blend, with respect to this study, are identified in Figure

14. There is a proportional relationship between the height of the absorbance peak

around 1142 cm-1and the degree of crystallinity of the PVA network [88, 96, 97]. The

peak results from a C-O stretching mode that is influenced by hydrogen bonding. The

1142 cm

-1

band is a shoulder on the larger peak centered around 1090 cm

-1

. This peak in

PVA is the characteristic alcohol peak due to C-O stretching.

Carbonyl bonds, which are attached to the pyrrolidone rings, result in stretching

modes seen with the absorbance peaks at 1657 cm -1 and 1567 cm-1. These two peaks

represent the free and hydrogen bonded carbonyl stretching modes for the PVA / PVP

blend respectively. It is assumed that a hydrogen bonded carbonyl is an indication of an

intermolecular secondary interaction between the carbonyl oxygen on a PVP chain and a

hydroxyl group along a PVA chain. Li and Brisson [116, 117] analyzed similar twin

carbonyl peaks for PVPh / PMMA blends, which have a very similar functional group

pairing to the PVA / PVP blend. They found carbonyl peaks for the PVPh / PMMA

blend to be at 1733 cm-1and 1707 cm-1.

The addition of PVP to a pure PVA polymer network causes many changes to the

spectra, the most notable of which are due to the carbonyl double bond as shown in

Figure 15. Other smaller changes were seen in the 1200-1500 cm-1range but these peaks

were not specifically identified. A large increase in the height of the 1657 cm-1peak was


62/122

50

seen when the proportion of PVP in the blend was increased. However, the hydrogen

bonded carbonyl peak at 1567 cm-1showed small increases in relative intensity up to the

blend containing 95% PVA and 5% PVP. For blends that were prepared with higher

fractions of PVP (10%, 17.5%, and 25%), the hydrogen bonded carbonyl peak at 1567

cm-1became shadowed and enveloped behind the larger free carbonyl peak at 1657 cm-1.

Distinguishing between the two peaks above 5% PVP was impossible as seen in Figure

15.

The heights of the free carbonyl peak at 1657 cm-1and the alcohol peak at 1090

cm

-1

were determined from spectra from three different films of each of the blends. The

baseline chosen for each spectra was used to determine the heights of both peaks. The

ratio of the peaks on each of the spectra was calculated and an average peak height ratio

(1657 cm-1/ 1090 cm-1) was calculated for each blend composition. For the samples that

had just been crosslinked via evaporation, an increase in peak height ratio was seen with

increasing content of PVP in the blend. After 56 days of immersion in vitro and

subsequent drying, one film from each blend was analyzed again with ATR-FTIR and the

peak height ratio was calculated. Figure 16 shows the results from this dissolution study.

A percentage decrease in peak height ratio over the 56 days immersion can be

calculated that roughly estimates the fractional loss of free carbonyl groups from the

polymer network. Figure 17 shows that blends prepared with PVP percentages 5% and

greater have a fractional loss of free carbonyl groups between 70-90%, indicating a

significant dissolution of PVP out of the polymer network.


63/122

51

Figure 14. Typical ATR-FTIR spectra for PVA/PVP blend. The spectra shown wasprepared with 95% PVA / 5% PVP. The four noted absorbance peaks are labeled in cm-1.

1657

1567

1142

1090


64/122

52

Figure 15. ATR-FTIR Spectra of PVA/PVP blends (143K PVA / 10K PVP) prior toswelling. Intensity of the free C=O peak increases with PVP%. The hydrogen bondedC=O peak is not detectable above 5% PVP due to its relative intensity to the larger andbroader free C=O peak.

1657 1567

% PVP


65/122

53

Figure 16. Change in peak height ratio of blend spectra over 56 days immersion in vitro.PVP not incorporated within the polymer network is free to dissolve into solution. Anequilibrium value of free C=O in each gel is reached after 56 days immersion.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0.5 1 5 10 17.5 25

PVP%

P

eakHeightRatio(1657/1090)

Prior to immersion (n=3)

After 56 days immersion (n=1)


66/122

54

Figure 17. Fractional loss of free carbonyl bonds as a function of PVP composition. Thefraction loss was determined by the change in free C=O peak height ratio with ATR-FTIR spectra analysis.

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25 30

PVP%

Fra

ctionallossoffreeC=O


67/122

55

4.3 Mechanical Analysis

4.3.1 Youngs Modulus

Youngs Modulus values were calculated by measuring the slopes of plots of

tensile stress versus tensile strain between 0 and 8.3% strain. The plots remained

relatively linear within this strain range. A sample plot showing how Youngs Modulus

values were measured is shown in Figure 18. Values were calculated for each of the

PVA/PVP blends after 2, 7, 14, 28, and 56 days immersion and are plotted in Figure 19.

For each of the blends, only slight changes were seen over the immersion time. After 2

days of immersion, the films had reached maximum swelling and were approaching

equilibrium swelling. The film prepared with pure PVA showed a decrease in modulus

between 2 days and 7 days immersion that remained the same throughout the duration of

the study. Films that were prepared with small additions of PVP to the polymer network

showed some deviations in modulus over time. Tests of blends containing 0.5% and 1%

PVP showed that maximum stiffness was reached within 7-14 days, after which a

decrease in modulus occurred with continued immersion. Blends prepared with 17.5%

and 25% PVP showed stable Youngs Moduli

novel associated hydrogels thesis

Documents