University of Southern California Dissertations and Theses

Destructive and non-destructive approaches for quantifying the effects of a collagen cross-linking reagent on the fatigue resistance of human intervertebral disc

(USC Thesis Other)

Destructive and non-destructive approaches for quantifying the effects of a collagen cross-linking reagent on the fatigue resistance of human intervertebral disc

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content NOTE TO USERS This reproduction is the best copy available. ® UMI R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. DESTRUCTIVE AND NON-DESTRUCTIVE APPROACHES FOR QUANTIFYING THE EFFECTS OF A COLLAGEN CROSS-LINKING REAGENT ON THE FATIGUE RESISTANCE OF HUMAN INTERVERTEBRAL DISC by Baber R. Syed A Thesis Presented to the FACULTY OF THE SCHOOL OF ENGINEERING UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree MASTER OF SCIENCE (BIOMEDICAL ENGINEERING) December 2004 Copyright 2004 Baber R. Syed R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. UMI Number: 1424233 Copyright 2004 by Syed, Baber R. All rights reserved. INFORMATION TO USERS The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. ® UMI UMI Microform 1424233 Copyright 2005 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Dedication To my parents, and my wife Sania for their love and support. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Acknowledgements It is not often that you meet someone who impacts your life profoundly. Dr. Tom Hedman is that person. I am grateful for the opportunity to be able to research and learn from him. I would like to thank my colleagues at the USC Orthopedic Research Laboratory, Dean Gray, Chuong Vo, and Brendan Chuang for their friendship and support. I would also like to extend my sincere appreciation to the members o f my committee, Dr. David D ’ Argenio and Dr. Stanley Yamashiro, for making my USC experience a memorable one. I would also like to recognize Ryan Garcia, Vinko Zlomislic, and Nissim Benbassat for the countless hours they have put into this research. Last but not least, I am thankful to my wife Sania, whose love, patience and encouragement helped me through my studies. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Table of Contents Dedication ii Acknowledgements iii List of Tables v List o f Figures vi Abstract viii Chapter 1 Origin o f Back Pain 1 Chapter 2 Non-destructive Testing o f Human Intervertebral Discs 9 Chapter 3 A New Approach to Non-destructive Testing 50 Chapter 4 Effects o f Disc Hydration on Material Properties 73 Chapter 5 Destructive Testing o f Human Intervertebral Disc 93 Chapter 6 Recommendations with Regards to In Vitro Testing o f 120 Human Cadaver Discs Bibliography 126 Appendix 130 iv R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. List of Tables Table 2-1: Calf pilot study test protocol. 30 Table 2-2: Degeneration grade for human cadaver discs. 34 Table 2-3: Specimen information. 36 Table 2-4: Creep deformation results prior to fatigue cycling. 43 Table 3-1: Multiple vs. single point indentation test protocol. 55 Table 3-2: Comparison of two indentation methods performed on 57 the same specimen. Table 3-3: Alternate test protocol using single point indentation 65 method. Table 4-1: Treatment and loading condition of specimens. 79 Table 4-2: Effect of fatigue loading on water content of disc outer 81 annulus. Table 4-3: Effect of fatigue loading on water content of disc 82 nucleus. Table 4-4: Effect of fatigue loading on water content of disc inner 83 annulus. Table 4-5: Results of comparison of hydration level of four 84 control treatments with Genipin treatment. Table 4-6: Results of comparison of hydration level of four 86 control treatments with Genipin treatment. Table 4-7: Results of comparison of hydration level of four 86 control treatments with Genipin treatment. Table 4-8: Average hydration content for each treatment. 87 Table 5-1: Mean destructive material properties of moderately 115 degenerated human cadaver discs. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. List of Figures Figure 1-1: Human spine. 4 Figure 1-2: Human disc. 5 Figure 2-1: Loads on the disc during lifting. 9 Figure 2-2: Genipin treated samples. 12 Figure 2-3: Motion segment. 15 Figure 2-4: Schematic of loads applied to disc specimens. 16 Figure 2-5: Stress vs. strain graph for a viscoelastic material. 19 Figure 2-6: Indentation tests sites. 22 Figure 2-7: Theoretical viscoelastic material response to load over 23 time. Figure 2-8: Stress relaxation data sample. 26 Figure 2-9: Static creep data sample. 27 Figure 2-10: Human specimen cross-section. 35 Figure 2-11: Human cadaver study stress relaxation for all four specimens. 38 Figure 2-12 A severely degenerated sample. 40 Figure 2-13: Force vs. deformation curves sample. 45 Figure 3-1: Multiple vs. single point indentation test protocol. 59 Figure 3-2: Comparison of recovery techniques. 62 Figure 3-3: Effect of fatigue cycling on viscoelastic material property. 64 Figure 3-4: Preliminary repeated stress relaxation results. 67 Figure 3-5: Preliminary repeated stress relaxation results. 69 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 3-6: Preliminary repeated stress relaxation results. 70 Figure 4-1: Section of the disc. 77 Figure 4-2: Effect of hydration on viscoelastic properties. 88 Figure 4-3: Hydration content of saline treated cadaver discs. 90 Figure 4-4: Hydration content of Genipin treated cadaver discs. 91 Figure 4-5: Change in hydration content of the disc with age. 92 Figure 5-1: Typical stress vs. strain plot. 97 Figure 5-2: Schematic of a disc segment. 101 Figure 5-3: Sectioning the disc into segments. 102 Figure 5-4: Marking the depth of the posterior annulus. 103 Figure 5-5: Slicing the remaining tissue. 103 Figure 5-6: Prepared posterior annulus segment. 104 Figure 5-7: Schematic of laser reflectance system. 107 Figure 5-8: Area measurement of the posterior annulus disc 108 segment using the laser reflectance system. Figure 5-9: Completely restrained specimen ready for potting. 109 Figure 5-10: Change in modulus of elasticity with location on the 114 posterior annulus of pre-fatigued specimens. Figure 5-11: Test results comparing the change in tissue property 115 with type of treatment in pre-fatigue specimens. Figure 5-12: Change in modulus of elasticity with fatigue cycling of 116 posteriomedial annulus samples. Figure 5-13: Change in modulus of elasticity with fatigue cycling of 117 posteriolateral annulus samples. Figure 6-1: X-Y table with precise linear motion. 122 vii R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Abstract The posterior annulus has been implicated as a primary region of degenerative disc disease leading to low-back pain. Prior in vitro studies in our lab, using an animal model, have shown the posterior annulus o f the lower lumbar discs to degrade following repetitive compress-flexion loading. In these studies measurement o f the viscoelastic material properties o f the disc annulus prior to and following fatigue cycling where used to quantify degradation o f the annulus. In an effort to improve the fatigue resistance o f intervertebral discs (IVD), Hedman proposed a cross-linking reagent to strengthen the material properties of the IVD collagen matrix (Hedman 2001). In addition, non-destructive testing using calf intervertebral discs cross-linked with this reagent have demonstrated a decrease in viscoelastic behavior, and a potential increase in resistance to degradation of material properties due to repetitive loading. It is the objective of this investigation to develop non-destructive and destructive test methods to show the effectiveness o f the cross-linking reagent in improving the fatigue resistance o f human cadaver disc tissue by measuring changes in the tissue’s material properties prior to and following cyclical fatigue loading. The results o f these studies indicated that previously developed non destructive indentation measurement techniques could not be used with human cadaver discs. This is primarily due to the fact that calf discs have a homogeneous, morphologically pristine disc structure, while human cadaver discs have years of damage in the discs tissue structure, making each sample unique. This resulted in viii R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. large variances within the localized posterior annulus test region, limiting the utility of these techniques. Consequendy, the focus o f this research shifted towards creating alternative non-destructive testing procedures, and improving the accuracy o f destructive testing of human cadaver discs. Preliminary studies have indicated a decrease in variance and feasibility of these procedures in demonstrating the change in non-destructive material properties with fatigue cycling. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Chapter 1 - Understanding Low Back Pain, and the Role of the Intervertebral Disc Prevalence o f Law Back Pain in Modem Society Low back pain affects millions in the United States each year, making it the second most common symptom related reason for visiting a physician (Deyo et al. 2001). Researchers have estimated that about two thirds o f adults suffer from idiopathic low back pain. While most cases are nonspecific low back pain, where recovery from pain is rapid, there is a burden o f cost on society associated with this epidemic that is non-recoverable. An estimated $20 billion dollars annually is lost in income by businesses or charged to taxpayers in the form o f worker’s compensation, making it one o f the costliest health problems in the United States. Similar economic costs o f low back pain affect other industrialized nations. In the U.K., an estimated 13 million working days are lost every year to back pain related absences. Studies have shown that low back pain may originate from many spinal structures, including ligaments, facet joints, the annulus fibrosis, and spinal nerve roots, to name a few. Perhaps some of the most common sources o f back pain are musculoligamentous injuries and age related degeneration o f the intervertebral disc, which account for 70% and 10% o f all the adult patients, respectively (Deyo et al. 2001). O n a smaller scale, problems such as spinal stenosis and disc herniation account for less than 7% o f all cases. 1 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Low back pain affects men and women most often between the ages of 30 and 50 years (Deyo et al. 2001). Risk factors include heavy lifting and twisting, bodily vibration, obesity, and poor conditioning. More specifically, epidemiological and morphological studies have linked repetitive lifting with higher incidences o f disc degeneration, disc herniation, and low back pain (Magora 1972, Kelsey 1975, Frymoyer 1983, Videman 1990). However, individuals may also be genetically predisposed to disc degeneration disease, resulting in accelerated rate of disc degradation. These individuals are especially susceptible to disc annulus tears or herniation early in their lives as a consequence o f mechanically weak discs. The debilitating impact o f degenerated discs on daily lives o f individuals, and the financial burden on society, has prompted vigorous research for solutions to low back pain. The arrays o f solutions available today are based on the seriousness of the injury to the disc. While most low back pain can be classified as strains and sprains, which require a combination o f bed rest along with specialized exercises to reduce pain, degenerated discs on the other hand may require major spine surgery. Surgical procedure is normally required for severely degenerated discs, as well as for spinal stenosis and herniated discs. Minor disc herniation may be treated surgically by removing tissue that is impinging on the nerve roots. Additionally, greater muscle strength is vital for prevention o f back pain, according to physicians. While this is not a permanent solution, it will relieve back pain until the discs degenerate further. Furthermore, serious surgical procedures may be postponed until much later in life. Reduction of physically demanding activities such as heavy lifting and challenging athletics may additionally prolong the life o f weakened discs. 2 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. In recent years, physicians have also been able to treat minor herniation and protrusions of tissue non-surgically by ablating disc tissue using percutaneus procedures. However, this procedure benefits a limited segment o f the patient population, and may not prevent recurrence of symptoms or reduce the rate o f disc degeneration. Serious herniation o f the disc and repeated spine surgery may eventually require fusion o f the vertebrae. While this may relieve pain, flexibility of the spine is gready limited, and it normally accelerates degeneration in the upper disc levels o f the spine. The present solutions to severe low back pain do not go beyond relieving the symptoms. Surgical solutions (including minimally invasive) are motivated by the removal of tissue from the injured disc, and are generally ineffective in slowing the progression of disc degeneration. Hedman has proposed a solution motivated by the idea o f keeping the degenerating disc intact. This may be accomplished by strengthening the existing disc matrix, as an alternative approach to relieving recurring symptoms that cause low back pain. The proposed strengthening of the disc matrix will be accomplished by forming cross-links between collagen fibers to improve the durability o f the intervertebral discs (Hedman 2001). If successful, this treatment may reduce annular fatigue damage and help maintain nuclear pressurization and integrity o f the entire disc. Increase in fatigue resistance o f discs, along with prescribed aerobic conditioning by physicians to strengthen the back and leg muscles, may reduce the frequency of recurrence o f low back pain. 3 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The Intervertebral Disc The spine is a complex structure of vertebrae and intervertebral discs, formed to protect the spinal cord, transfer load to the pelvis, and provide flexibility to the back. The spine is divided into three segments, cervical, thoracic, and the ones in three-dimensional motion, are joined together by intervertebral discs, ligaments, and musculature. The vertebral bodies, which are designed to bear mainly compressive loads, are progressively larger caudally. The lumbar region of the spine, the area most often associated with back pain, contains 5 of the 24 vertebrae and the related intervertebral discs. lumbar region (Figure 1-1). Twenty-four vertebrae, which articulate with adjacent Figure 1-1: Human spine. Lower lumber section shown. Re-sketched from Gray’s Anatomy (Clemente 1985). 4 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The intervertebral disc is formed of collagenous material with a unique composition in the inner and outer structures (Figure 1-2). The inner portion o f the disc, known as the nucleus pulposus, is a gelatinous mass containing glycosaminoglycans. This water binding molecule is in abundance in the young adult nucleus, and enables the nucleus to store energy and uniformly distribute compressive loads. The outer portion of the disc surrounding the nucleus, the annulus fibrosis, is composed primarily o f type I collagen fibrils. The crisscross arrangement of collagen fiber bundles within the fibrocartilage form a composite structure, optimized to withstand large circumferential forces transmitted by the pressurized nucleus (Figure 1-2). The disc itself is attached to the endplates, separating the vertebral bodies. Nucleus Figure 1-2: H um an disc. Disc shown with transverse processes intact. Re-sketched from Gray’s Anatomy (Clemente 1985). The focus of this paper will be on the material properties of the intervertebral disc, and in particular the discs o f the lumbar region of the spine. 5 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Evaluation of disc material characteristics A critical aspect in proving the efficacy o f the proposed cross-linker reagent in slowing the progression of disc degeneration is the development o f fundamental models o f the mechanism by which the tissue degrades, and tests to measure the amount o f degradation. Earlier, numerous mechanical loading conditions were cited as risk factors associated with low back problems. Research into the causes o f disc degeneration has implicated repetitive loading of the tissue as a likely means by which the tissue is weakened (Hedman 2001). Fatigue causes tissue subjected to repetitive loading to fail at stresses well below their ultimate level. In the intervertebral disc, degradation o f the disc tissue is frequently observed in the posterior annulus (Osti 1992). Since the intervertebral disc is primarily an avascular structure, degradation o f tissue normally dominates repair (Hedman 2001). This suggests that fatigue is a dominant process in the posterior annulus, more specifically repetitive circumferential and tensile loading o f the annular tissue (Adams 1994). This mechanism of degradation was reproduced in our lab by Hedman et al. through repetitively loading intervertebral motion segments to simulate flexion- compression o f the spine (Hedman and Femie 1997). Their assertion was that one loading mechanism responsible for tissue deterioration in the posterior annulus involved combined axial compression and sagital plane bending (Hedman 2001). 6 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. While this represented a reasonable experimental model of disc degeneration, quantification o f the amount o f damage due to fatigue was carried out by targeted tests to record the changes in material properties o f the posterior annulus. N on destructive indentation test methods were utilized to measure viscoelastic properties of the disc. Meanwhile elastic-plastic measurements were obtained using both destructive and non-destructive test methods. Preliminary experiments by Hedman et al. using destructive and non destructive mechanical testing techniques have shown degradation o f elastic-plastic and viscoelastic material properties of the disc tissue when subjected to non- traumatic fatigue loading (Hedman 2001). Moreover, extensive fatigue testing on bovine intervertebral discs has given preliminary indication o f the effectiveness of the cross-linker reagent in improving the fatigue resistance o f the disc tissue (Gray et al. 2001). Organisation The basis for this thesis work is the application of previously developed non destructive and destructive test methods to human cadaveric IVDs. However, our ultimate goal is to establish in vitro tests that, in the near future, will be used to show efficacy o f the cross-linking reagent to reduce fatigue induced degradation of moderately degenerated human cadaveric IVD tissue. In addition, these tests may provide helpful indication in refining the biochemistry o f the cross-linker. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The thesis is organized with two main topics, destructive and non-destructive testing of human tissue. Chapter 2 describes the non-destructive test methods (creep, stress relaxation, and hardness) and their application to human cadaveric IVDs. The results of the experiments reported in this chapter were significant, and form the basis for chapter 3, which explores an alternative way to evaluate viscoelastic material properties of human IVDs. Chapter 4 evaluates the role hydration level in the tissue plays in determining viscoelastic material properties. Chapter 5 describes the refinement o f destructive tests, such as the tensile test. Some of the refinement studies include improving the precision o f area measurement techniques for soft tissue specimen, as well as attachment of the disc samples to tensile testing instrumentation. Outstanding issues and future directions are presented in chapter 6. 8 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Chapter 2 — Non-destructive Testing of Human Intervertebral Disc Back ground: Early calf pilot study The objective of the early fatigue resistance pilot study using calf intervertebral joints conducted by Gray and Hedman et al was to examine the relationship between collagen cross-linking and degradation o f the calf intervertebral disc (TVD), due to repetitive loading. They cycled the intervertebral joint in flexion, generating a sagital plane bending moment as seen in the diagram below (Lindh 1989). This type o f loading has been linked to low back pain and degeneration of the posterior annulus o f the lower lumbar discs. In some cases, back pain may even result from the degenerated posterior annulus impinging on the spinal cord. — ^ 1 r ~ J Disc ’ I / " t t 2MN Figure 2-1: Loads on the disc during lifting. This schematic shows the loads on the lower lumber while lifting, in this case a 200 N object. The load on the disc is not simply the compressive weight of the object, but also the bending moment due to the object being lifted at a distance from the disc. This bending moment causes high tensile loads on the posterior annulus (Lindh 1989). 9 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Early investigations by Hedman et al. confirmed, through in vitro testing, that degradation o f elastic-plastic and viscoelastic material properties occurs when a disc is subjected to non-traumatic cyclical loading. This led to an investigation in exploring possible ways to improve fatigue resistance of the disc, specifically the posterior annulus, which has been implicated as a possible cause o f back pain. Previous studies show that collagen cross-links play a critical role in strengthening the adult connective tissues. O ne of the main goals o f the pilot fatigue resistance study was to examine the role o f collagen cross-links in stabilizing the degeneration of disc tissue (Gray et al. 2002). In order to achieve the goals of this study specific methods were developed by Hedman and Gray et al. to test the relationship between cross-linking and degeneration o f the posterior annulus. Some o f these included methods for producing cross-links in the IVD matrix, degradation o f the posterior annulus through non-traumatic cyclical loading, and quantification of the amount of degradation by measuring the material properties o f the posterior annulus. The critical question that needs to be answered is, whether cross-linking can slow down the degeneration process of the human intervertebral disc, thereby relieving symptoms that cause back pain. The methods developed in the aforementioned pilot study will be employed to human cadaver discs to show if degradation o f the posterior annulus o f a cadaver disc is significandy decreased due to cross-linking. 10 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The following sections provide a brief description o f treatments, and then lead into test methods that were developed by Gray for the calf study. A brief summary o f the fatigue resistance pilot study will be presented, followed by the main topic of this chapter, application of these methods to human cadaver discs. Treatment Methods Developed With the hypothesis that cross-linking will improve disc fatigue resistance by increasing collagen cross-links, a treatment and application method was developed (Hedman 2001). An important criterion for a choice of a cross-linker was that it must be biocompatible. To form cross-links in the discs, an off the shelf reagent, Genipin, was recommended by Hedman. Genepin has been used in the food industry and was known to be potentially biocompatible. It was also possible to visually confirm the reaction o f Genipin within the IVD. When Genipin treated IVD samples are exposed to air, they form a deep purple discoloration on the surface (Figure 2-2). By dissecting the disc in half across the transverse plane, it was possible to assess the amount of penetration o f the cross-linker within the disc. A dark purple edge, normally 3-4 mm in thickness, was observed across the annulus fibrosis o f the disc. And a gradual change in hue from dark to light purple was observed across the cross section towards the nucleus pulpous. This was a visual confirmation of the chemical reaction of the cross-linker with the collagen matrix. Although most samples had good penetration o f the cross-linker, a method for 11 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. quantitatively discerning the number o f cross-links formed as a result o f the treatment is currendy under development. Figure 2-2: Genipin treated samples. The photograph above shows a control sample (left) along side a Genipin treated sample (right). Note the dark edges formed around the posterior annulus of the treated disc and the penetration within the annulus and nucleus region as evidenced by the change in color. This method o f treatment by soaking was used in the calf pilot study of Gray and human pilot study o f Syed. The samples above are from a single human spine. After experimenting with several different methods for cross-linking, a 36- hour soak method was chosen for maximum coverage (Gray 2001). An injection method was also attempted on a small sample group with some success. This would be the preferred method for simulating in vivo delivery o f the cross-linker. Successive studies conducted in parallel showed that color changes across the annulus fibrosis and the nucleus were not as dramatic. Instead, small blotches o f purple and light purple coloration were seen throughout the disc (Hua 2002). Prior studies by Adams et al. (Adams 1986), where discs were injected with contrast media for discogram analysis, had shown some degree o f success for fluid injection method. However, the author noted that high pressure was required to inject fluid. The injected media often coalesced in pools o f contrast, rather than 12 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. covering the entire disc, which may be indicative of the condition o f the disc itself. Further refinement of the injection technique is required to develop a successful in vivo method of delivery. It must also be considered that when performing in vitro injection tests, the disc should be appropriately loaded in compression to precisely represent in vivo conditions. The final protocol for specimen treatment required a 36 ± 1 hour soak in a 0.5 L ml of 0.33% Genipin-PBS (phosphate buffer solution) solution at room temperature. Sample containers were sealed with a plastic film to prevent water loss. Through mechanical testing o f treated calf specimen, it was determined that a formulation containing 0.33% Genipin would ensure that sufficient number of cross links are formed between the type I collagen o f the disc matrix (Gray and Hedman 2001). The standard method for maintaining the hydration o f the control specimens was refined over the course of the non-destructive pilot study. Initially a 1 M NaCl soak was proposed to reduce swelling o f the disc tissue. Swelling refers to the osmotic influx of fluid into the collagenous tissue. It was noted that the 1 M NaCl soaking was changing the material properties o f the disc (Gray 2001). This was proven by measuring the material properties before and after the application of the control treatment. Meanwhile, the samples with cross-linking treatment showed litde change in their material properties at pre and post treatment. Based on these observations, an investigation was undertaken using various saline concentrations to determine a soaking method that both brought about no change in material properties and matched the hydration levels of the cross-linker treated discs. 13 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Hydration plays a significant role in IVD material properties. It was determined that a 0.15M NaCl (normal physiological saline) soaked control was equivalent in hydration to the Genipin treated samples. This control soaking method was used in all subsequent studies, which include the human non-destructive study of this chapter. The hydration study will be discussed in detail in chapter 4. Specimen Preparation Specimens were harvested from calf lumbar spines. The harvested specimens consisted o f individual lumbar motion segments (Figure 2-3). After removing the muscles, tendons, and miscellaneous soft tissue, the lumbar spine was sawed midway across two adjacent vertebral bodies to obtain a motion segment. Further preparation work was required to gain access to the posterior annulus o f the disc. This included cutting the posterior processes and removing the spinal cord. The PLL (posterior longitudinal ligament) was left intact. A similar method of preparation was used for the human cadaveric specimens. Testing o f the specimens on the MTS mechanical testing machine required potting of specimens in polyurethane blocks. This involved placing screws in the vertebral bodies, such that the screws projected outward from the cortical shell. Potting was accomplished using two-part liquid polyurethane (Figure 2-3). 14 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 2-3 Motion segment. A prepared specimen consists of a motion segment with posterior processes removed (left). The segment above shows the location of screws installed in the vertebrae for fixation (gray screws). Six additional screws (black screws) were required for human samples due to porosity in the spongy bone of the vertebrae. Significant increase in fixation was achieved by taking advantage of the stronger cortical shell of the vertebrae as shown in the treated human sample on the right. Degeneration simulated through non-traumatic fatigue cycling It was identified that uniaxial compressive loads at the center o f the disc alone can not produce injury to the disc. W hen low compressive forces are combined with large flexion bending moments (Figure 2-4), sufficiently high tensile forces can be generated at the posterior annulus. This results in tears or severe injury to the posterior annulus tissue (Hedman and Femie 1997). The effect of tensile forces applied in this manner to the intervertebral disc has been confirmed empirically in prior studies by Adams (Adams and Hutton 1982) and Hedman (Hedman and Femie 1997). Therefore, the prescribed method for producing degradation in a specimen disc is by means o f a flexion fatigue test. This test involves repetitively applying a compressive force to an IVD segment at a specified distance anterior to the center o f the disc (Figure 2-4). A non-traumatic load o f 15 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 200 N was chosen to simulate repetitive loads a human spine may encounter during the day. This load was applied 40 mm from the center o f the disc to produce a moment o f 8000 N-mm. The total number o f fatigue cycles to produce disc degradation were limited to 6000 cycles, however, significant changes in material properties can be observed as early as 2000 cycles (Gray 2001). 200 N 0 mm- Posterior A B e n d in g ? Anterior " T M a « U i « 4 t Figure 2-4: Schematic of loads applied to disc specim ens. The schematic shows an intervertebral disc potted in polyurethane blocks with specified load applied. A 200 N load applied 40 mm from the point of rotation A, causes a bending moment. The reaction of the moment is shown at point A. This would be considered a light load that a human disc would encounter in the course of a day. 16 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Non-destructive Test Methods Developed The aim o f this study was to quantify disc degradation over a number o f fatigue cycles. This would require the specimen to remain intact for the duration of the test. Therefore, non-destructive techniques were applied, as oppose to destructive. Destructive tests require the specimen to be destroyed in order to obtain any material property information. Destructive testing techniques, such as the tensile test, were employed to human cadaver discs to determine material properties and will be discussed in detail in chapter 5. Before proceeding with the non-destructive test methods, a thorough understanding of what is being measured is important. Material properties need to be measured to quantify disc degradation, since they generally describe how a material behaves when subjected to various forces such as tension, compression, or shear. For centuries scientists have studied these forces on materials such as metals and ceramics, also known as elastic solids. These materials can often be modeled as simple mechanical devices, such as springs and dashpots in various configurations. Scientists have also studied material properties o f fluids such as oil, water, and molten metal by describing viscosity, density and other properties of fluids. This second class of materials falls under the category o f viscousfluids (Strong 1996). Elastic solids, when subjected to a force, tend to recover their shape and restore the energy imparted on them. Viscous fluids, however, have the capability to flow when a shear force is imposed, yet as the force is removed they stop flowing and the energy that was applied is not returned (Strong 1996). In short, there is no 17 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. recovery. Within the last century a new class of materials were discovered that could not be classified under either o f the two classes. At room temperature when force is imparted on them they can flow like viscous fluid. Upon removal o f that force some shape may be recovered similar to elastic solids. This class o f materials is called viscoelastic. Polymers make up a major segment of viscoelastic materials. In addition, biological tissue are also classified as viscoelastic materials. W hat makes these viscoelastic materials unique is the rate dependence o f their mechanical properties (Figure 2-5). For example, if a found piece of silly putty were thrown on to the floor it would bounce back and return to its original shape. However, if that same silly putty were left on a table over night, it would flow into a puddle shape. Hence, at high rates o f deformation (throwing the putty) where times are short due to impact, the putty behaves like an elastic solid. When times are long and the rate of deformation is low, the putty behaves like a viscous fluid (Strong 1996). 18 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Stress vs. Strain for Viscoelastic Materials 20 - Figure 2-5: Stress vs. Strain Graph for a viscoelastic material. The graph shows rate dependence of viscoelastic materials. As the same material is loaded at a higher rate, the resulting tensile stress will appear to increase. The component tissue of the intervertebral disc behaves like viscoelastic materials from a bulk material, macroscopic, point o f view. Its material properties such as tensile strength, elastic modulus are rate dependent. Although, the mechanism o f its viscoelastic behavior is considerably different than that of traditional viscoelastic materials like polymers. When a polymer is deformed, the mechanism o f deformation is by means o f breaking weak secondary bonds and allowing polymer chains to move and entangle. If the rate o f deformation is increased, causing higher resistance in the entangled polymer chain, a greater force will be required to overcome this resistance. This is easily viewed in the figure 19 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. above, where the graph shows a rise in strain rate causes a rise in stress for the same polymer being tested. Biological materials respond to strain rate changes in a similar manner. More specifically, IVDs respond to deformation through fluid movement across their collagenous matrix. A component o f the matrix, proteogylcan, is responsible for the fine pore si2e o f the matrix, which provides high resistance to fluid flow under increased rate of deformation (Urban et al. 1981). Depending on the magnitude of deformation, the change in shape o f the disc, and subsequent movement o f fluid across the matrix typically is recoverable. However, a point can be reached where large deformation, or prolonged loading, may cause damage to the matrix itself, resulting in plastic deformation. Therefore, intervertebral disc material properties can include an elastic component and a plastic component. The proposed method for quantifying degeneration o f the posterior annulus of an IVD was to measure its material properties by means o f an indentation test (Hedman et al. 2001). Indentation testing is a non-destructive test method that is commonly used in material science to characterize the hardness o f metals. Hardness measurements are based on plastic deformation, indicating that a metal has reached its yield point. Typically, the depth of penetration and diameter o f the indentation are used to quantitatively assign a hardness number. The relative yield strength of materials can be determined by comparing hardness numbers. Indentation testing thus provides a non-destructive means o f measuring elastic-plastic properties. 20 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. In recent tim e, this m eth o d o f indentation testing has been applied to softer materials like polymers. Due to the viscoelastic nature of polymers, which respond to a force by deformation and flow, hardness measurement becomes difficult to interpret, since it is a measure o f a material’s ability to resist time-dependent and static penetration. Many scientists have applied indentation techniques to polymers by distinguishing between viscoelasticity and viscoplasticity. Viscoelasticity is the measure o f a material’s time dependent material properties, where the relationship between deformation and time is critical. Viscoplasticity is the measure o f elastic- plastic properties o f a material, and deformation over time is insignificant. That is, the viscoelastic effects have been removed. Intervertebral disc tissue displaces when subjected to indentation testing. One component o f this displacement is due to elastic deformation, while the other component is due to plastic deformation (Gray 2001). In order to apply indentation hardness test to a calf IVD specimen, Gray distinguished the plastic component from the elastic component. The plastic component o f displacement was achieved by repeated indentation to a predetermined load, thereby removing most o f the elastic displacement in the tissue. An empirical equation related to the indentation force and plastic displacement formed a hardness number (Gray 2001). indenter penetration (2.1) Where: indenter penetration — applied indenter force = penetration depth 21 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Equation 2.1 can be used to determine the hardness index, H i n ( U x , which allows comparison between specimens of their relative yield strength. This indentation technique o f measuring hardness preserves the integrity o f the IVD sample, while allowing for repeated testing on the same sample within a well defined location on the posterior annulus o f the disc (Figure 2-6). Indenter Posterior Annulus Lumbar Motion Segment Figure 2-6: Indentation test sites. The schematic shows the site of indentation used for the calf pilot study as well as the human pilot study. The center of the posterior annulus would be used for evaluating material properties prior to fatigue cycling. The left and right posteriolateral sites would be used for indentation after 3000, and 6000 cycles. Thus far, the focus o f indentation testing has mainly concerned elastic-plastic properties, which are time independent. Since the intervertebral disc is a viscoelastic material, it is essential to characterize its time and rate dependent behavior. The 22 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. elastic-plastic properties o f biological tissues can be determined by a pseudo hardness measurement technique that was proposed by Gray. However, viscoelastic material properties of biological tissues must be determined by measuring the elastic deformation over time. This is performed by applying a constant force over a specified amount o f time and observing the time dependent deformation known as creep. The loads applied in this manner are usually well below the yield point. The first half o f the figure below shows a typical response of a viscoelastic material to constant force over time. £ o LL Creep-Retaxation of Viscoplastic Materials Time (s) 12 3.5 10 8 2.5 6 4 2 0.5 0 40 60 0 20 80 100 E E ® E o o « Q . m Q Figure 2-7: Theoretical viscoelastic material response to load over time. The dotted curve shows a creep response to a constant stress. The material is loaded to a specified stress level. The deformation over time can be observed until the material is unloaded. The solid curve shows the material loaded to a specified deformation and held constant over time. During this time the material stress relaxation can be observed. This is a typical response o f a viscoelastic material. 23 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. When a constant load is applied to a viscoelastic material, it begins to flow immediately. If the force applied remained constant, the viscoelastic material would continue to displace as long as the force is applied. Although, once the force is removed, the viscoelastic material would attempt to recover. This recovery takes place over time, and the material may not fully recover depending on the load applied and the duration o f time. It is important to note that the application o f load discussed above is for a finite amount o f time. At some moment, most biological tissues reach a steady state creep rate, where the tissue will deform until it reaches the point o f failure. Similar to the method o f creep testing where a load is applied for duration of time, a deformation can also be applied and held constant over time. A material deformed axially up to a specified force and subsequently held constant at that displacement is said to undergo relaxation. The curve for relaxation looks similar to the second half of the curve shown above (Figure 2-7). For the calf pilot study, indentation testing was used to determine viscoelastic properties o f the posterior annulus by employing creep deformation and stress relaxation methods. The complete hardness measurement protocol, as well as creep and stress relaxation protocol as used in the calf specimen pilot study, and subsequent cadaver pilot study, will be summarized below. 24 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. C alf Pilot Study Non-Destructive Testing Protocol Three types o f non-destructive indentation tests were established in this study to determine the viscoelastic and elastic plastic material properties of the IVD. Stress relaxation and static creep were used to determine the viscoelastic properties, while the hardness test was used to determine the elastic-plastic properties. Although these methods have been applied by scientists on many materials, including biological tissue, the application to the IVD was initiated by Hedman. The following describes Gray’s optimized test protocol used for the calf pilot study. An MTS servo-hydraulic test frame with an externally connected precision load cell formed the mainstay of the indentation testing program. A 2.5 mm diameter rod with a hemispherical tip was incorporated in the load cell to form the indenter. Measurements o f displacement were digitally acquisitioned from the movement o f the crosshead to an accuracy o f ±0.01 mm. Strain rate was controlled by the MTS controller using a force feedback from the external load cell. Stress relaxation experiments were conducted by loading the surface o f the posterior annulus tissue with a 2.5 mm indenter to a predetermined force o f 10 N. The applied stress was chosen to be substantially lower than the yield stress of the bulk tissue. The indenter was programmed to reach the targeted load in 30 seconds. Once the targeted load was reached, the indenter was held at that displacement for 60 seconds while the MTS data acquisition system recorded the force (Figure 2-8). Experiments conducted in the lab by Gray revealed that to obtain viscoelastic properties, the targeted site must not have been previously tested. Experiments 25 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. proved that repeated indentation on the same site decreased viscoelasticity (Gray 2001). Therefore, it was decided that stress relaxation would be the initial test performed followed by static creep and then hardness. 0.8 z 0.2 10 0 20 30 40 50 60 70 90 80 Time (s) Stress Relaxation Initial 5% «-* - Final 5 % Displacement | Figure 2-8: Stress relaxation data sample. The graphs show force and displacement data recorded during stress relaxation testing. Linear regression analysis of the initial and final points o f the stress relaxation curve is also shown graphically. Although data from a human cadaver sample is used, this is atypical of calf samples Similarly, static creep experiments were conducted by loading the surface of the posterior annulus tissue with a 2.5 mm indenter to the same predetermined load of 10N. The sample was loaded to 10 N over a 30 second period. To seamlessly initiate a creep test, the control mode was locked into load control upon reaching the targeted load. The load was then held constant while displacement was recorded by the MTS data acquisition system (Figure 2-9). It is again important to consider the 26 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. load history o f the site o f indentation. Since a 10 N load was applied only momentarily during a stress relaxation test, it was possible to retest this site for static creep. 12 i z o u. 0 60 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 E £. * - » c o E v o ra o. < o 5 Time (s) Figure 2-9: Static creep data sample. The graphs show force and displacement data recorded during static creep testing. Notice the initial force overshoot before stabilization occurs at 20 seconds into the 60-second creep test. The hardness test consisted of applying a load of 10 N for 10 consecutive cycles to remove most of the viscoelastic behavior of the disc. A final 10 N indentation was then made and measurements o f force and depth were recorded by the MTS data acquisition system. Equation 2-1 proposed by Gray was used to calculate the hardness index. Fatigue cycling o f the specimen followed initial posterior annulus indentation testing. During fatigue cycling, all samples were wrapped in saline soaked paper 27 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. towels to prevent the loss of hydration from the disc. During other parts o f the testing, the samples were periodically sprayed with normal saline to keep the tissue moist. A second set o f fatigue cycling followed after indentation testing at post 3000 cycles. A final indentation test was performed at post 6000 fatigue cycles (Figure 2- 6). Testing Sequence The sequence of the three indentation tests, stress relaxation, static creep, and hardness are a critical part o f the test protocol. Previous discussion o f viscoelastic material properties had lead to the conclusion that there are two types of material properties o f interest, viscoelastic and elastic-plastic. Viscoelastic properties are determined by measuring force and deformation over time, and hence are sensitive to the “force and deformation history” o f a sample. Fresh tissue site is essential for accurate stress relaxation and static creep measurements. Therefore, these two tests were conducted before the hardness test. In addition to being sensitive to deformation and loading history, viscoelastic material properties o f the disc are also sensitive to the hydration level of the disc. The dependency o f stress relaxation and static creep on hydration will be discussed in greater detail in chapter 4. In contrast, elastic-plastic properties are tim e in d ep en d en t and are not sensitive to the deformation history o f a specimen. Through repeated viscoelastic measurement at a fixed site using indentation techniques, it was shown by Gray that 28 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. there is a gradual reduction in viscoelastic effects due to repeated indentation (Gray 2001). The result o f this experiment was that repeatable measurement o f elastic- plastic property could be made after a site is repeatedly indented for more than 10 cycles (Gray 2001). For reasons indicated above, the hardness test was last out of the sequence of three indentation tests. As previously mentioned, indentation testing was conducted at three intervals. Pre-fatigue indentation testing allowed recording o f non-fatigued intervertebral disc material properties. This would include initial stress relaxation, static creep, and hardness prior to fatigue cycling the specimen. Subsequent testing intervals at post 3000 and 6000 cycles would have to be performed on fresh posterior annulus tissue and not on a site where pre-fatigue data was obtained. Recall that once a site has been repeatedly indented, it will not be possible to obtain accurate stress relaxation data since fresh tissue is required to measure viscoelastic properties. Therefore a regional variation experiment was conduct by Gray to show that within ± 4mm o f the center of the posterior annulus, no significant differences in the means for creep and stress relaxation measurements existed for calf lumbar disc specimens (Gray 2001). The final test protocol allowed for indentation testing at center, left and right o f the posterior annulus, as long as the indentation sites were within the specified ± 4mm (Figure 2-6). Normally, most o f pre-fatigue indentation data was taken at the posterior center, while the data for post 3000 cycles and 6000 cycles were taken at posteriolateral sites. Table 2-1 below shows the order o f testing. This exact same order was also used for the cadaver human disc study, the results of which will follow after a brief presentation o f the results of calf pilot study. 29 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Indentation Measurement Sequence Measurement Indentation Location 1 Stress Relaxation Center of the Posterior Annulus 2 Creep Displacement Center of the Posterior Annulus 3 Hardness Center of the Posterior Annulus 3000 Compression/Flexion Fatigue Cycles 4 Stress Relaxation 4 mm Lateral to Center 5 Creep Displacement 4 mm Lateral to Center 6 Hardness Center of Posterior Annulus 6000 Compression/Flexion Fatigue Cycles 7 Stress Relaxation 4 mm Lateral to Center (Opposite Side) 8 Creep Displacement 4 mm Lateral to Center (Opposite Side) 9 Hardness Center of Posterior Annulus Table 2-1: Calf pilot study test protocol (Gray, 2001). Results and Conclusions Drawn from the Calf Pilot Study The calf pilot study was designed to investigate the effect of the cross-linking agent on the posterior annulus tissue o f the disc in response to repetitive loading. Comparison between the cross-linked IVDs and control specimens were made. Regression analysis was applied to interpolate noisy force signals. The goodness o f the regression analysis was determined by evaluating the coefficient o f determination, the R2 term. Two principal effects were examined, that is, the results o f indentation tests were used to evaluate the impact o f the biochemical treatment prior to fatigue loading, and changes measured as a result o f repetitive loading. Stress relaxation analyses o f the calf discs were performed by examining the differences in force over the 60-second relaxation period (Figure 2-8). Linear regression was used at the initial and final 5% o f the stress relaxation force curve, to 30 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. determine this difference precisely. The results indicated an 18.4% (p=0.026) downward shift in viscoelasticity due to the biochemical treatment (Gray 2001). The results also showed that the rate o f change in stress relaxation with repetitive loading decreased for treated specimens. These results were found to be statistically significant, and indicate a change in reduced viscoelasticity for treated specimens. Creep deformation o f the calf discs was performed by examining the differences in indenter depth over the 60-second static load period (Figure 2-9). The static load o f 10 N was applied and was maintained by the MTS controller, through a feedback loop from the external load cell. Additional analysis was performed on the data by observing the rate of creep deformation at the end o f 60 seconds. Regression analysis was used to determine the slope of the creep curve. All o f the creep data was normalized to avoid errors resulting from drift o f the targeted static force. The results indicated a downward shift in creep deformation due to cross-linking treatment, however, the results failed to satisfy the significance level of 0.05 (Gray 2001). Similarly, an 18% reduction in creep behavior was observed for treated samples in response to fatigue loading. Also, these results were found not to be statistically significant (Gray 2001). Hardness data was evaluated using equation 2-1 and the force deformation curve obtained from indentation testing. Regression analysis was used to determine the endpoints of the force curve. Hardness index was evaluated with a minimum indenter load of 0.25 N to avoid re-zeroing errors observed on many samples. An apparent increase in hardness index was evident as a result of biochemical treatment 31 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. prior to fatigue cycling. However, this increase over the control specimens was not maintained following fatigue cycling. As with the creep analysis, these results were not statistically significant (Gray 2001). Application of Test Methods to Human Intervertebral Disc The challenge o f the present preliminary human study was to apply non destructive test methods developed in the calf pilot study to human IVDs. The initial focus was to use the procedures developed in the calf study, along with quantifying degeneration of the human IVDs. Unlike the young IVD specimens harvested from calves, where all o f the specimens were assumed to have an undamaged homogeneous disc matrix, human cadaver specimens tend to have various degrees o f degeneration. Since the human disc matrix is non-homogenous (variation from sample to sample), comparison o f material properties can only be made between specimen samples of similar grades. For the preliminary human study, the results of which will follow, four lumbar IVDs were tested. The age o f the cadavers ranged between 30 and 40 years. Following the third and final indentation test, the discs were cut across the mid-plane using a scalpel. The appearance o f the discs was examined and recorded using high- resolution digital photographs o f each specimen. This information was used to grade discs for degeneration o n a 4-p oin t scale m eth od p rop osed by R olander (Rolander 1966). Indentation test methods have already been introduced in this chapter from the calf study. The following section will present the method used for 32 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. degeneration scoring and results o f viscoelastic and elastic plastic material property testing o f human cadaver IVD. The post indentation tested sliced disc samples were also used to determine the hydration level. The results o f which will be discussed in chapter 4. Grading strategy o f disc degeneration The disc matrix o f a human cadaver can have varying amounts o f damage depending on that person’s genetics, loading history and health history. Persons involved in spinal injury, or working in a job requiring heavy labor will tend to have greater damage to their discs. This occurs due to the high, more frequent or more sustained loading o f their spine. Environmental factors are not the sole cause; genetics can play a role as in the case o f disc degeneration disease. This disease affects individuals as early as in their late twenties. When comparing material properties o f cadaver specimens, these factors must be kept in mind. Material properties of healthy disc tissue are very different from degenerated disc tissue. Previous sections in this chapter have already discussed to some degree the profound effects hydration levels can have on viscoelastic properties. Therefore, degenerated discs, which have tears and discontinuities in the matrix, will show markedly different material property values in comparison to healthy discs. To make comparisons between sample populations, researchers have devised a scale consisting o f four to five grade levels to quantify the observed degeneration. This allows comparison between specimens o f similar degeneration scores. 33 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Several researchers have proposed methods of quantifying degeneration through qualitative methods. These include visually looking for signs o f wear on the discs. In most cases the disc is sliced in half for observation. While others have used a non-destructive method were samples are measured for disc height and observed for bulge and osteophytes on the exterior to indicate the level of degeneration (Rolander 1966). Radiographic images can also be taken in an effort to preserve the disc’s integrity (Adams and Hutton 1986). The most common method for in vitro testing involves slicing the specimen in half at the end o f an experiment to make observations. These observations are given a score, the higher the score, the greater the amount o f observed degeneration. A method proposed by Rolander (Table 2-2) shows a established standards for scoring cadaver disc (Rolander 1966). Degenera tio n Score Description o f Disc 0 Macroscopically normal discs without signs of ruptures or other structural changes. Both the annulus fibrosus and the nucleus pulposus are shiny white. 1 Disc with normal appearance in general but with a somewhat more fiberous structure in the nucleus. A distinct boundary between the nucleus and annulus. Yellowish discoloration. 2 Clear deterioration of the central structure o f the nucleus, which is definitely drier than normal and usually discolored. There may be isolated fissures in the annulus. 3 Marked degenerative changes in the nucleus as well as he annulus fibrosus with ruptures and sequestra in the nucleus or the annulus and/or scarring o f the nucleus. Table 2-2: Degeneration grade for human cadaver discs (Rolander 1966). Quantifying the level o f degeneration in a human IVD is a subjective process. For this study a scale o f 0 to 3 was used (Table 2-2). An IVD with a score o f 0 would be given to a specimen with virtually no degeneration and visibly 34 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. gelatinous nucleus, while a score o f 3 would be given to a specimen with a high degree o f degeneration and dehydrated nucleus. Following examples illustrate these differences. Ht612 (Lumbar L1/L2 Disc) Male 32 yrs. Grade: 0 (Genipin Treatment) Ht623 (Lumbar L2/L3 Disc) Male 32 yrs. Grade: 0 (Control Treatment) Ht712 (Lumbar L1/L2 Disc) Male 39 yrs. Grade: 1 (Control Treatment) Ht745 (Lumbar L2/L3 Disc) Male 39 yrs. Grade: 3 (Genipin Treatment) Figure 2-10: Human specimen cross-section. Cross section of all 4 samples used in human cadaver disc study. 35 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Results ofpreliminary human cadaver disc study The preliminary human study contained samples with degeneration scores varying from 0 to 3, despite the fact that all the samples were relatively young (Table 2-3). Most samples received a score of 0 or 1. Since only 4 samples (2 control and 2 biochemicaly treated) were used for the preliminary study, statistically significant results cannot be obtained. Normally, n=10 samples are required to establish a statistical significance o f observed phenomenon, provided the variance o f the measurements are low. With a sample group o f 4, we can only comment on observed trends. Subject no. Age /Sex Original Condition Treatment Degeneration Score Ht612 32/M Frozen Fresh Genipin Treatment 0 Ht623 32/M Frozen Fresh Control Treatment 0 Ht712 39/M Frozen Fresh Control Treatment 1 Ht745 39/M Frozen Fresh Genipin Treatment 3 Table 2-3: Specimen information. Stress Relaxation Analysis of Human Cadaver IV D Samples Stress relaxation data was obtained by using the indentation procedures outlined above for the calf pilot study. This involved indentation testing at 3 different sites in a well defined area on the posterior annulus. From this point forward, this method will be known as multiple point indentation testing. All cadaveric samples were first indented at the center o f the posterior annulus to determine pre fatigue viscoelastic properties. The second set o f indentation measurements were 36 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. obtained at post 3000 cycles o f flexion-compression fatigue on the posteriolateral annulus tissue. Third and final indentation measurements were taken at the opposite posteriolateral side at post 6000 cycles o f fatigue. Stress relaxation results for cadaver specimens were determined in the same manner as for the calf specimens. Regression analysis was used to determine the end points o f the relaxation curve. As with the calf study by Gray, two important trends were evaluated. The impact of fatigue cycling on annulus stress relaxation, and the impact o f the biochemical treatment on stress relaxation prior to and following fatigue cycling. Samples are normally divided across degeneration scores in order to make reasonable comparisons, however, a small sample group (n=2) for each treatment required them to be pooled together. An upward shift in stress relaxation was seen due to the biochemical treatment. The mean o f pre-fatigue stress relaxation for the treated specimens was 3.35 N (st. dev.: 0.48 N). This is a 249% increase relative to the 1.34 N (st. dev.: 0.50 N) recorded for control samples. The biochemical treatment had an opposite effect on the human cadaveric discs in comparison to the calf pilot study. The calf study revealed an 18.4% reduction in viscoelasticity due to treatment. Although these results are interesting, they are not statistically significant when compared to the calf pilot study. It must be noted that the average standard deviation for both control and treated human specimens was approximately half that of the calf study. The following graph shows stress relaxation results for all four samples plotted against fatigue cycles. Specimen Ht612 and Ht623 are a pair, that is, one 37 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. sample, Ht623, is a saline soaked control while Ht612 (from the same spine) is treated with Genipin. Ht612 shows that stress relaxation is decreasing as fatigue cycling is increasing. Similarly Ht712 and Ht745 are control and treated samples, showing the same trend. The effect o f fatigue cycling on stress relaxation for human cadaveric specimens appeared to be opposite that o f calf. This is in fact an incorrect response, and runs counter to what is intuitive. Greater flexion-compression fatigue on the disc posterior annulus should cause further damage to the tissue, which would result in increased viscoelastic property. Only Ht623 showed an increase in stress relaxation as the sample was fatigued. Pre-Fatigue Post 3k Fatigue Cycles (100 Post 6k Fatigue Cycles (200 min.) min.) Cycles (Time) — Ht612-G----------- Ht623-C * * Ht712-C - - - Ht745-G Figure 2-11: Human cadaver study stress relaxation for all four specimens. 38 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Since erroneous trends were observed for post fatigue data, mean differences o f relxation from 0 to 6000 cycles and standard deviations were not calculated. Instead, an explanation will be presented as to why the actual data is in fact the opposite o f what was anticipated. It is interesting to note that the specimen with the highest degeneration score was observed to be the most distorted data set. Sample Ht745 with a degeneration score o f 3 has the greatest decrease in stress relaxation. Figure 2-10 shows a cross sectional photograph o f Ht745, where an arrow points to the center (pre-fatigue) indentation site. It has been observed on several cadaveric discs (n=12) with high degeneration scores that the center posterior annulus, as well as the nucleus, is substantially damaged. Dehydrated tissue and tears in the fiber are commonly observed. Sample Ht745 exhibits these characteristics specifically in the center. Figure 2-12 shows another sample with a degeneration score o f 3. The photograph shows the nucleus has almost disintegrated and an area in the center posterior annuls where the density o f fibrous tissue has been severely diminished. 39 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Ht834 (Lumbar L3/L4 Disc) Male 44yrs. Grade: 3 (Control Treatment) Figure 2-12: A severely degenerated sample. The sample above was sliced after indentation testing. A large gap filled with fluid and tissue was observed (shown by the tweezers) in the center of the posterior annulus, which may be additional proof that large relaxation values were observed due to the severely degenerated center of the posterior annulus. By examining the center posterior annulus cross section of specimens, and with the knowledge that the first indentation on all experiments takes place at this site, we can explain the reverse trend in viscoelastic properties. Substantial decrease in pre-fatigue viscoelastic material properties was observed at the center posterior annulus primarily due to a vast amount of tissue degradation. Posteriolateral sides, where post 3000 and 6000 cycle indentation measurements are made, were observed to have a denser annulus structure with conceivably lower deterioration o f the tissue, resulting in lower viscoelasticity. Therefore, most samples, specifically the ones with higher degeneration score, seemed to show changes in material properties due to location and not fatigue. 40 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. These observations contradict the assumption made in the calf pilot study, where the general area o f indentation (± 4 mm from center posterior annulus) was considered to be homogeneous. That is, insignificant variation in material properties was observed from one site to another within the specified area. It is therefore reasonable to conclude that this assumption is not true for moderately degenerated human specimens. With greater degeneration o f disc there is the likelihood of greater non-homogeneity. Therefore, the starting level o f viscoelasticity will vary from one location to another. Creep Analysis of Human Cadaver IV D Samples Creep data was obtained by using procedures outlined above for the calf pilot study, which used multiple point indentation technique. Erroneous results for creep data were expected for reasons outlined in the human disc stress relaxation analysis. The results were further compromised due to overshoot o f indenter force by the MTS controller. It required 15 to 20 seconds for the force to recover to its static value o f 10 N, leaving less than 45 seconds of creep displacement measurement. In some cases, the overshoot caused the slope o f the creep curve to reverse momentarily before proceeding in the correct direction. As with the calf study by Gray, creep data was analyzed using more than one calculation to obtain m eaningful results. C reep displacem ent w as calculated to obtain material behavior at an earlier creep stage. While additional insight into equilibrium creep behavior was gained by evaluating the rate o f creep deformation at 41 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. the end of the 60 seconds test period (Gray 2001). Regression techniques were used to evaluate the rate of creep behavior. The creep deformation data was fitted to a natural logarithmic function {y — a * ln(x) + b), however, a power law function proved (y = be ") to be an equally good fit for the data. Coefficient o f determination, R2 , was evaluated to determine the goodness o f the least squares fit curve. Analogous to the stress relaxation data, the impact o f biochemical treatment on creep deformation was evaluated pre and post fatigue cycling. The number of samples was limited to two for each treatment; therefore, the results that follow will not be statistically significant Table 2-4 shows creep deformation results prior to fatigue cycling. There was an upward shift in creep measurements due to the biochemical treatment. Gray observed the opposite in the calf pilot study. The 60- second creep deformation mean for treated specimens was 0.113 mm (st. dev.: 0.054 mm) prior to repetitive loading. This is substantially more than the lone control sample with a creep deformation value o f 0.0363 mm. The value for control sample Ht623 was excluded due to force overshoot. For the creep experiments, the MTS uses a feedback loop from the load cell to stabilize the force at 10N. The overshot and reaction time to recover often depend on the material being tested. With such large variation in human discs due to degeneration, the quality o f the creep experiments can vary from one specimen to another. Further testing will be required to show whether this trend in increased viscoelasticity was due to the biochemical treatment. 42 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Specimen ID Ht612 Ht623 Ht712 Ht745 (Treated) (Control) (Control) (Treated) Scaled Creep (mm) 0.0742 -0.0225 0.0363 0.1511 Table 2-4 : Creep deformation results prior to fatigue cycling. A reverse trend in viscoelastic creep deformation was observed from pre fatigue to post-6000 cycle measurements. All pre-fatigue indentation measurements were performed at the center of the posterior annulus, while post fatigue measurements were performed on the less degenerated posterolateral annulus. Therefore, initial creep measurements were much greater than post-6000 cycle creep deformation measurements, leading to inconclusive results. As with the stress relaxation data, the observed changes in material properties were due to location and not fatigue. N o relevant trends can be stated regarding creep deformation o f human IVD. Hardness Analysis o f Human Cadaver IV T ) Samples In this study we have been unable to observe any trends in the data due to the dependency of viscoelastic properties on disc degeneration and the location of indentation. Elastic-plastic properties, however, express the strength o f the proteoglycan matrix. Since elastic-plastic properties of IVD do not depend on the interstitial fluid movement, as do viscoelastic properties, it is anticipated that degeneration will have a lesser affect on hardness property o f IVD. 43 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Hardness measurements of human cadaveric discs were obtained using procedures from the calf pilot study. The 30-second test period, where the indenter is ramp loaded at constant strain, was used to calculate the hardness index. Recall that the hardness index is the indenter force divided by the depth o f indentation over a specified time period (2.1). As with the analysis o f stress relaxation and creep, two principal effects o f biochemical treatment will be evaluated. An impact on the measured hardness prior to and following fatigue cycling. The results will not be statistically significant because o f the limited sample size for each treatment group. A representative analysis o f hardness data over the fatigue cycle period is shown (Figure 2-13). Indentation force was plotted against penetration depth, where the hardness index is the slope o f the curve. Since the sample has gone through repeated indentations as part of the hardness protocol to remove viscoelastic behavior, the plotted curve should display an approximate linear relationship. A possible source o f error in computing the hardness index is determining at which two points to compute the slope. The “slack” in the beginning of the curve may be due to residual viscoelastic effects or in some cases improper re-zeroing during the hardness test. If the slope o f the entire curve is taken, this initial effect is computed in the hardness index, which can add variability to the results. However, there is a greater linear relationship at the last 20% o f the curve, which may serve as a better measure o f the posterior annulus hardness. Unlike the calf pilot study, where hardness measurements were taken at forces above 0.25 N, the hardness for the human cadaver disc study will be determined using the last 20% o f the data. In all 44 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. four samples, the strain rate was kept constant at 0.1 m m /sec, which allowed comparison between specimens. 12 10 8 6 4 2 0 0 2 1 Displacement (mm) Figure 2-13 : Force vs. deform ation curves sam ple. The sample above shows three graphs for calculating hardness index at pre-fatigue, post 3000 and 6000 cycle fatigue. There was an increase in hardness due to cross-linking treatment, prior to fatigue loading. The hardness index for treated specimens was 21.08 N /m m (st. dev.: 5.89 N /m m ). This is a 28.7% increase relative to the control specimens’ mean of 16.38 N /m m (st. dev.: 7.03). Due to the small sample population, the standard deviation for the human cadaver specimens was observed to be twice as high as for the calf pilot study specimens. The increase in the standard deviation may also be attributed to specimen degeneration. Both control samples, which received a low 45 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. degeneration score o f 1 (scale 0-3), showed a very high standard deviation o f 43% of the mean. Similar to the trend observed in the calf study, pre-fatigue hardness was not maintained following fatigue cycling. Post 3000 load cycle data revealed that the mean hardness index o f the treated specimens was 13.58 N /m m , and 13.75 N /m m for the control specimens. This is a 35.6 % and 16.1% decrease from pre-fatigue values for treated and control specimens, respectively. Similarly, a minor downward trend in hardness was observed at post 6000 load cycles. Between 3000 and 6000 cycles, most samples recorded a decrease of less than 12%. One important observation regarding the plot o f force vs. deformation shown in figure 2-13 was the evident shift in the curvature of the graphs due to fatigue loading. The curvature becomes larger especially at the initial half o f the graph. This change in curvature, which was consistent for all the samples, may suggest that there is deterioration o f the posterior annulus due to fatigue loading. However, it would be difficult to decipher the effect o f the biochemical treatment by measuring the curvature differences. Although these results are not statistically significant due to the small sample size, it is interesting to note that similar results were obtained for the calf pilot study where the sample size was 12 per treatment. We conclude that no qualitative observations can be drawn from this data. Perhaps this may not be the best way to observe elastic-plastic material property trends for viscoelastic materials. Destructive test methods, such as a tensile test, yield far more precise results. The yield point and slope from a simple tensile test can be used to effectively gauge a material’s 46 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. strength. Chapter 5 will determine the effectiveness o f the biochemical treatment by comparing yield stress and strain values, as well as other material properties, of control and treated posterior annulus IVD samples. Concluding Remarks Regarding M ulti Point Non-Destructive Testing of Cadaveric IV D The analysis o f a small population o f human cadaver discs did not reveal any qualitative results. However, through this study, insight into the indentation testing of human cadaveric discs was gained. More specifically, there are two principal findings that prevent us from obtaining successful results through multiple point indentation testing: 1. The posterior annulus tissue of a human cadaveric specimen is a “non- homogeneous” structure, which prevents comparison between two adjacent tissue sites. 2. The non-homogenous nature o f each specimen will result in high variance within a sample group. Recall that 3 out o f the 4 human cadaveric discs tested showed a reverse trend in viscoelastic behavior due to fatigue cycling, which was not physically possible. In other words, a decrease in viscoelastic behavior was observed with fatigue loading o f the posterior annulus. This discovery was made because all pre fatigue data was obtained at the center posterior annulus. Upon dissection, it was revealed to have a greater amount of degeneration when compared to the posterolateral annulus. As a result, the assumption of local homogeneity is not true 47 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. for moderately degenerated human cadaveric specimens. Consequently, the multiple point method o f indentation testing used for the calf specimens proved to be unsuccessful. Strategy for Improving Non-destructive Test Methods The multiple point indentation testing and results generated in the calf pilot study are valid for calf specimens, because they have non-degenerated homogenous disc matrix. The controls in that study proved that the results obtained were due to fatigue and not to location on the calf posterior annulus (Hedman 2002). For the human cadaver specimens, where the location on the annulus does affect results, a new technique o f single point indentation is proposed. The basis for this technique is to eliminate the variability o f location on the posterior annulus by indentation testing repeatedly on a single site. This may allow comparison of viscoelastic measurements with repetitive loading. There are several issues regarding single point indentation testing. Recall that validation o f stress relaxation and creep methods by Gray required a fresh unaltered annulus surface. An indentation site, if previously loaded, would compromise stress relaxation results. Therefore, once a site is indented, a recovery process would be required to remove the affects of a previous indentation. This recovery process would allow repeated indentation at one site following each fatigue cycle. 48 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Additionally, it was noted over the course o f this study that several mildly degenerated samples (score < 1) displayed much greater viscoelastic behavior at the center posterior annuls than posteriolateraly. This information may be critical when comparing data from single point indentation method, as repeated indentation at the center o f posterior annulus will generate much larger values o f stress relaxation and creep than posterolateral annulus. Therefore, if validation o f single point repeated indentation testing proves to be successful, we should compare results, not only with respect to degeneration scores, but with respect to center or lateral locations o f the posterior annulus as well. The next chapter will introduce the recovery technique along with experiments conducted using single point indentation testing. It is hoped that this technique will not only give adequate results for human test samples, but it should verify the result o f the calf validation study that used multiple point indentation technique. 49 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Chapter 3 — A New Approach to Non-destructive Testing New methodology for non-destructive testing of human IV D s An alternate method of non-destructive testing o f human cadaveric discs was revealed in chapter two, known as single point indentation testing. This method was introduced in order to overcome the challenges in determining the changes in viscoelastic properties with repetitive loading. Recall that multiple point indentation method, as conducted in previous fatigue resistance pilot studies using calf IV D s, assumes that the posterior annulus tissue of the IVD is homogeneous. While this allows comparison of indentation force from one tissue site to another in a calf specimen, the assumption of homogeneity does not hold true for the adult human discs due to its non-homogeneous tissue structure. Based on results presented in chapter 2, it was determined that repeated indentation measurements at a fixed site may be the only means of obtaining viscoelastic material property from a moderate to severely degenerated human disc (degeneration score > 1). Therefore, the intent of the study presented in this chapter is to develop the techniques for measuring viscoelastic properties utilizing single point indentation. A consequence of this approach is that the tissue site must be restored from prior deformations. This is critical in order to compare material property, such as creep and stress relaxation, between successive indentations. Recall that viscoelastic effect decreases asymptotically with repeated indentation, hence the choice of 50 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. indenting at multiple sites over a well defined path on the posterior annulus, in previous studies (Gray 2001). With the fixed site indentation technique, we wish to observe changes in material property due to fatigue and not due to the deformation history o f the tissue. Therefore, a tissue recovery process in between indentations is essential for removing deformation after each indent. A verification study was undertaken in our laboratory to prove the feasibility of the single point indentation method. The following sections will summarize a verification study that was undertaken in our laboratory to prove the feasibility o f the single point indentation method. Comparisons o f material property will be made with previously established methods, namely multiple point indentation. A preliminary non-destructive test protocol will be presented later, utilizing single point indentation method, specifically developed for moderately degenerated human IVDs. Finally, early test results on a cadaver disc will illustrate what promise this method holds. Requirementsforfeasibility of Single Point Indentation Method O ne of the main challenges to implementing single point indentation method is to produce a repeatable recovery (from prior indentation) process. Since viscoelastic properties of a disc is in part dependent on fluid flow characteristics through the proteoglycan matrix (Hedman 2001), restoration of the tissue was attempted by re-hydrating the disc in a normal saline (0.9% NaCl) bath. The effectiveness of this process can be verified by measuring material property at a fixed 51 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. site on the posterior annulus, re-hydrating the tissue, and then comparing it with a successive measurement at that site. Additionally, another key factor for repeatable viscoelastic measurements is that the indenter must be placed precisely on the same location. For the verification study, 1 mm grid marks positioned on the test fixture were used to replace the potted specimen after each re-hydration cycle. Although repositioning will have a minimal effect on viscoelastic measurements obtained from calf specimens, its impact on human cadaveric specimens is much greater, due to the non-homogeneity in the disc tissue. Recovery techniquesfor removal of tissue deformation caused by repeated indentation An approach to recovering deformation from previously indented disc tissue was proposed by Hedman (Hedman 2002). This deformation recovery step consisted o f hydrating the disc tissue for a specified amount o f time in normal saline. This simulates the diurnal cycling that human discs experience on a daily basis. High loads during the day cause the discs to deform and loose their height. Disc height is restored over night during recumbent sleep through absorption o f surrounding fluid, when they are no longer under excessive compressive load. Concentration of proteoglycan within the nucleus imparts a large swelling potential, causing the disc to im bibe fluid and swell. 52 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Scientists in the area o f spine research use a damp environment during testing, and are generally concerned with the level of hydration in an IVD. Adams and Hutton have used a bath o f Ringer’s solution during their experiments, a solution containing 8.6 gm of NaCl, 0.3 gm potassium chloride, and 0.33 gm calcium chloride per liter. This solution simulates the body’s fluid surrounding the spine. Similarly, Hedman et al have used normal saline bath, a 0.15 M NaCl solution, in their experiments involving compressive loading o f human cadaveric lumbar spine. Researchers, when performing these types of experiments, have used loads of 100N to as much as 450 N to simulate axial compressive preloads on the IVDs. A load of 100 N was used by Hedman to simulate preload and to limit the imbibition of fluid into the discs. Between experiments, this compressive load was reduced by 50 N for a period of 30 minutes to 1 hour to allow the discs to recover their height by reabsorbing the fluid lost during previous experiments (Hedman et al. 1997). Comparable loads and duration of re-absorption were used by Pflaster and Krag et al. who demonstrated that a disc under a preload of 450 N can re-hydrate in 30 to 60 minutes (Pflaster and Krag et al. 1997). The recovery procedure used in this study is similar to the one used by Hedman. Discs were re-hydrated for 30 minutes in normal saline in between indentation experiments. Unlike previous studies, the discs were not preloaded while in the saline bath. In this manner, the recovery time could be minimized by incurring a large swelling potential to radially expand the disc and remove deformations caused by the indenter. The choice o f an unloaded recovery was 53 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. reasonable by the fact that prior treatment practices involved 36 hours of soaking in a solution, without a load. However, the effect o f loaded fluid recovery was evaluated in a small study, the results o f which will be presented later in this chapter. Fundamentally, the effect of unloaded treatment o f discs should be reconsidered. Particularly, since in vivo environment dictates a preload on the discs. Further refinement o f the treatment process will be discussed in greater detail in chapter 6. Recall from prior fatigue resistance studies, three indentation measurements were desired (2 indentations are the required minimum) to observe degradation of viscoelastic material property with repetitive loading. Therefore, for a single non destructive test the tissue site must recover a minimum of two times in order to obtain three force indentations, at pre fatigue, post 3000 cycles, and post 6000 cycles. Since there are three tests to be conducted, stress relaxation, creep, and hardness, up to 8 recovery cycles may be required. Verification Protocolfor Single Point Indentation Testing Using Calf Specimens Verification of the single point method was carried out by means o f a test protocol comparing stress relaxation obtained utilizing the proposed method against the standard method used in prior studies (Table 3-1). The two methods were performed on each sample to avoid the variability between samples from influencing the statistical analysis. Consequently, the data may be looked at qualitatively by 54 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. comparing methods for each sample, or may be combined to give statistically significant results. Notice, that there are two types of tissue variations. O ne type of variation is from sample to sample, while the other is a regional variation within a specimen. The latter is o f greater interest in this comparison. The regional variation within the posterior annulus o f the calf disc was shown by Gray to be negligible within a 8mm section o f the central_posterior annulus. The purpose of this comparison is to show that variability between repeated indentations over a fixed site will be equivalent to this regional variations observed in prior studies. Measurement Sequence Multiple Point Indentation Single Point Indentation 1 Indent at Center, Left and Right (Record Stress Relaxation Data) Soak for 30 Minutes in Normal Saline 2 Repeat Indent at Center, Left, and Right (Record Stress Relaxation Data) Soak for 30 Minutes in Normal Saline 3 Repeat Indent at Center, Left, and Right (Record Stress Relaxation Data) Soak for 30 Minutes in Normal Saline 4 Repeat Indent at Center, Left, and Right (Record Stress Relaxation Data) Table 3-1 Multiple vs. single point indentation test protocol. Eight calf IVD samples were tested using the protocol outlined above (table 3-1. Data was evaluated by comparing the difference in stress relaxation between indentations. For the multiple point test method, stress relaxation differences were determined between the center and left indent, center and right indent, and left and right indent. For the single point method, stress relaxation differences were 55 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. determined between the first indentation and each successive indentation. That is, the difference between 1s t and 2n d , 1s t and 3rd , and 1s t and 4th indentation. For each test method, the mean o f the stress relaxation differences between indentations and the standard deviation were calculated. This allowed comparison to be made qualitatively between methods for each sample. Additionally, the data for all samples was combined, and segregated by method. A 2-way analysis of variance (ANOVA) was applied to determine the general effect on viscoelastic measurement due to the applied indentation method. Results of the Verification Testing Viscoelastic measurements indicate strongly that there is an equivalency between the two test methods. The results of individual specimens show qualitatively that the difference in stress relaxation between repeated indentations is frequently lower for single point versus multiple point indentation (Table 3-2). More importantly, similar conclusion can be made with regards to the standard deviation of the differences, where specimens utilizing single point method exhibited a 43% decrease (0.176 N vs. 0.307 N). This alone suggests that greater repeatability of viscoelastic measurements is gained through indentation testing at a fixed site. 56 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Sample # Differences Between Stress Relaxation Values (multiple point indentation) Differences Between Stress Relaxation Values (single point indentation) Mean of Differences (N) Standard Deviation of Differences (N) Mean of Difference (N) Standard Deviation of Differences (N) CT156 0.489 0.406 0.445 0.451 CT323 0.258 0.215 0.150 0.100 CT245 0.549 0.416 0.263 0.160 Ct312 0.664 0.292 1.175 0.188 Ct334 0.522 0.239 0.380 0.144 Ct234 0.707 0.511 1.324 0.148 Ct356 0.170 0.094 0.086 0.093 Ct456 0.499 0.281 0.275 0.123 Table 3-2: Comparison of two indentation methods performed on the same specimen. The mean for each sample is calculated from three relaxation difference values. The data can be observed with greater significance by combining the individual stress relaxation differences from all samples. For the combined sample group, 24 data points were obtained for each method. Differences obtained for single point indentation was recorded at 0.51 N (st. dev.: 0.48 N). This is a 6.3% increase relative to 0.48 N (st. dev.: 0.33 N) obtained through multiple point indentation. However, this is not a significant increase, when the data is observed in light of the standard deviation in mind. This may in part be due to the outliers in the data (sample Ct234 and Ct312). These outliers will be discussed later in this chapter. A two way ANOVA applied to the stress relaxation difference measurements for both methods strongly indicates that the two methods are equivalent (2-way ANOVA p=0.77). If the desired difference to be detected is 20% of the control measurement (i.e. multiple point indentation) with a significance level o f 0.05, the experiment design provides a statistical power of 0.9. In other words, there is a 5% probability of finding a false positive, and a 10% probability o f finding a false 57 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. negative. In summary, single point indentation appears to be a viable technique for testing non-homogeneous IVD tissue. An additional note regarding this methodology is that it is primarily feasible for the stress relaxation test. At an early stage in the testing, it was determined that this technique would not be suitable for measuring creep. Creep measurement test deforms the tissue site significantly, making complete recovery of the site difficult. This prevents accurate viscoelastic material property measurement from successive indentations. Although non-destructive hardness measurement may be carried out using this method, it will require measurement at a different site. The hardness test deforms the tissue significantly, which prevents complete recovery o f the site for successive stress relaxation measurement. Therefore, hardness may be measured at an adjacent location within the central posterior annulus. Comparison o f Weighted and Non-Weighted Recovery Techniques Using Single Point Indentation Method Determination o f the effect o f load used during the hydration recovery process is illustrated by the bar chart below (Figure 3-1), which shows two variations on the recovery technique performed on calf discs. One involves limiting the imbibition o f hydration simply by placing the specimen in a normal saline bath and allowing osmotic equilibrium to be reached. While the other involves additionally placing a 10 lb. load over the specimen to further limit the imbibition of fluid in to the disc. 58 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Comparison of Recovery Techniques : 0 lb vs. 10 lb Load J I I Ct234 Ct334 Ct312 Ct245 0 ib. 0 lb. 101b. 101b Ct645 Ct456 Ct356 Calf IV D Sample 11st Indent - 2nd Indent ■ 1st Indent - 3rd Indent □ 1st Indent - 4th Indent Figure 3-1: Comparison of recovery techniques. Percent difference in stress relaxation between initial and subsequent indentation is shown. Each calf sample was repeat indentation tested 3 times at a fixed site. Viscoelastic property data from the verification study is represented as a percent difference of the initial stress relaxation value (Figure 3-1). Percent Difference = S.R; — 3’Rx / 3R; *100 (3.1) Where: S R f — stress relaxation in N o f first indentation x = 2n d , 3rd , or 4th indentation We are interested in detecting trends exhibiting increasing differences (positive differences) between repeated indentations. This suggests that the deformation in the tissue is not recovering to its initial point, making the re- 59 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. hydration recovery technique unreliable. However, a decreasing (negative differences) trend is also possible, suggesting errors in recorded viscoelasticity due to poor instrumentation or sensitivity to indenter replacement as a result o f large variation within the tissue site. In either case, it is imperative to minimize the percentage difference between indentations since our goal is to measure changes in viscoelastic property due to repetitive loading not to repeated indentation. Consequently, large standard deviation between repeated indentation measurements will make it difficult to observe changes in viscoelastic property between pre and post fatigue data. Prior studies conducted by Gray have established a 20% increase as a significant change in viscoelastic properties due to fatigue. Most samples (Figure 3-1) were within ± 12% of the initial stress relaxation, which is well below the significant 20% value. However, two samples exceeded this value, at 27.7% and 21.8 %, Ct312 and Ct234, respectively. These increases are not due to the applied load during recovery, as they are seen both in 0 lb. and 10 lb. recovery data (Figure 3-1). They are as a result o f high initial stress relaxation values, which could not be matched after the 1st recovery. There are several explanations for large positive stress relaxation differences. It is possible that due to a 36 hour unloaded control treatment in a saline bath, some specimen may be imbibing large quantity of fluid, resulting in super-hydrated specimens with high initial stress relaxation measurement when compared with subsequent indentations, which are given 0.5 hrs to recover. Conversely, it is also possible that a small subset o f specimens may imbibe fluid at a slower rate, and 60 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. longer recovery time in the saline bath is required. Also, since the level o f hydration is close to equilibrium (due to 36 hour saline treatment) it may take longer for substantial pressure to build up to radially expand and remove the deformation created by the indenter. A repeated measure ANOVA applied to the stress relaxation values obtained for this data weakly suggests that the two recovery methods are similar (ANOVA p= 0.09). That is, the weight placed on the sample to prevent the disc from imbibing excessive fluid, has not significantly influenced the measured viscoelastic property. Perhaps a 10 lb weight, acting for a reasonably short time, may be insufficient to affect the fluid transport across the IVD matrix. This is supported by experiments conducted by Pflaster et al. where load in access o f 100 lb. was required to significantly affect the hydration level in a cadaveric IVD placed in a saline bath (Pflaster and Krag 1997). Also, adding a 10 lb. weight will limit increases in disc height, but may enhance radial expansion by placing the nucleus under compression, and redistributing the fluid content within. This may induce radial expansion o f the outer annulus as well, and possibly assist in removing deformation caused by the indenter. By making the assertion that the initial indentation value is not representative of the tissue’s material property, improvement in repeatability can be achieved. Percent difference in stress relaxation between indentations is shown with the initial indentation data removed (Figure 3-2). Percent differences were calculated between 2n d and 3rd , and 2n d and 4th indent. Removing the initial relaxation value gave better 61 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. repeatability, even after 6 indentations. Differences in stress relaxation for all samples were observed to be less than 9%, a 3% improvement from prior interpretation o f the results (10% and 15% improvement for Ct234 and Ct312, respectively). ■ sg o c 13 £ 8 iE « Q ® E * m U ) £3 Comparison of Recovery Techniques : 0 lb vs. 10 lb Load (1st indentation point removed) 27 24 21 ~ 18 15 12 9 6 3 0 -3 -6 -9 -12 E - Q Ct456 0 lb Gt356 Olb. ie : Ct234 Olb. E Ct312 10 i b n Ct323 10 to Ct334 01b. I I Ct245 10 ,b- ■ * 5 L-J ™ 101b. Calf IV D Sample 12nd Indent - 3rd Indent Q2nd Indent - 4th Indent Figure 3-2: Comparison of recovery techniques. Percent difference in stress relaxation between 2n d and subsequent indentation is shown. A vast improvement in consistency of obtaining viscoelastic property was made. A Proposed Protocol fo r Obtaining Changes in Viscoelastic Property Utilising the Single Point Indentation Method Once a recovery method was established, a study was designed to determine the tissue’s response to repetitive loading. A test paradigm similar to pervious fatigue resistance studies was employed to demonstrate the effects o f fatigue cycling 62 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. on viscoelastic property by utilizing single point indentation method to measure the changes in stress relaxation. For this study, calf specimens were subdivided into two groups. One group of specimens was subjected to sub-traumatic, cyclical fatigue loading. Meanwhile a control group was allowed to stand at room temperature for the duration o f time, to match the exposure to ambient conditions experienced by the fatigued group. All samples were wrapped in saline soaked towels to limit the loss of hydration. Recovery test results presented earlier indicated a reduction in error of recorded material property when the first indentation is removed. Based on this premise, specimens were subjected to pre-indention. That is, following an initial indentation test, the specimen was placed in a saline bath for 30 minutes to recover. The specimen was then indentation tested once again to determine its pre-fatigue viscoelastic property (stress relaxation) value. Following this 2n d indentation, the entire cycle o f hydration recovery followed by indentation was repeated once again. After completing a pre and initial indentation, the fatigued sample were loaded at 200 N for 3000 cycles at 0.5 Hz. Visocelastic measurements of fatigued sample group were obtained at post 3000, and 6000 cycles, meanwhile the control sample were concurrently indentation tested at intervals of 0,100, and 200 minutes. A total o f 5 samples were used, where 3 samples were fatigued, and 2 samples were non-fatigued controls. To increase the number of data sets (a set consists o f viscoelastic material property at 3 intervals), samples were tested at several sites within the central posterior annulus. A total of 7 data sets were obtained from the cycled specimens, while 6 data sets were obtained from the two non- 63 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. fatigued controls. The recorded data was evaluated in the same manner as described in chapter 2, utilizing linear regression to determine the endpoints o f the stress relaxation curve. The results o f this experiment indicated an increase in viscoelastic material property due to fatigue cycling (Figure 3-3). The stress relaxation o f fatigued samples increased by 2.451 N (st. dev.: 0.395 N) following 6000 cycles o f repetitive loading. This is 63.3 % higher relative to the 3.872 N (st. dev.: 0.438 N) obtained for the non-fatigued control specimen following a corresponding time interval o f 200 minutes. A 2-way ANOVA applied to post fatigue stress relaxation results strongly indicates an increase in viscoelastic properties due to fatigue (p=0.0004). Effect of Fatigue Cycling on Viscoelastic Property 8 7 6 5 4 3 2 1 0 o % 3 & ( A (0 £ 55 Pre-Fatigue Post 3k Fatigue Cycles Post 6k Fatigue Cycles (100 min.) (200 min.) ■ Fatigue Cycled Non-Cycled Figure 3-3: Effect of fatigue cycling on viscoelastic material property. 64 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The results presented thus far have confirmed the feasibility of using single point indentation technique as a non-destructive means of determining changes in viscoelastic property o f calf IVD with repetitive loading. Based on these results, a non-destructive test protocol is proposed (Table 3-3), utilizing the aforementioned method with hydration recovery. Indentation Measurement Sequence Measurement Indentation Location 1 Stress Relaxation Single Site Within + /- 8mm of Central Posterior Annulus 2 Hardness Single Site Within + /- 8mm of Central Posterior Annulus Recovery: 30 minute Soak in Normal Saline 10 Compression/Flexion Cycles (100 N load) 3 Stress Relaxation Repeated at the Exacdy the Same Location 4 Hardness Repeated at the Exacdy the Same Location Recovery: 30 minute Soak in Normal Saline 3000 Compression/Flexion Fatigue Cycles 5 Stress Relaxation Repeated at the Exacdy the Same Location 6 Hardness Repeated at the Exacdy the Same Location Recovery: 30 minute Soak in Normal Saline 6000 Compression/Flexion Fatigue Cycles 7 Stress Relaxation Repeated at the Exacdy the Same Location 8 Hardness Repeated at the Exacdy the Same Location Table 3-3: Alternate test protocol using single point indentation method. O ur motivation is to use this protocol to measure changes in viscoelastic property o f human cadaveric discs. The following section will discuss the results o f a limited evaluation o f cadaveric samples using this protocol to determine viscoelastic property (stress relaxation). Additionally, experiments analogous to the verification study were performed to determine the repeatability of single point indentation method using human cadaveric discs. 65 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Single Point Indentation Method Applied to Human IVHs It has been shown above that single point indentation method can be used to determine changes in viscoelastic property, namely stress relaxation, o f calf IVD. However, our intent was to employ this method on human cadaveric discs. Having said that, it is important to realize the impact of non-homogeneous cadaver tissue on repeatability o f this technique. Therefore, prior to employing the test protocol (Table 3-3), verification experiments were carried out to determine the constancy of repeated indentation measurements. As with the prior verification study, test procedures presented in table 3-1 were used to determine the constancy o f repeated indentation. Prior to testing, specimens were exposed to a saline bath for 36 hours without load. Single point indentation testing was carried out at several locations within an 8 mm section o f the central posterior annulus. Stress relaxation measurements were repeated 4 times at precisely the same fixed locations on each specimen. Grid lines were used to facilitate the placement o f specimens. The recovery technique was employed between each successive indentation to remove prior deformation created by the indenter. Three cadaver specimens, each with a degeneration score of 2, were indented at 3 different locations, resulting in a total o f 9 data sets. Analysis, similar to the verification study discussed earlier was carried out to obtain percent difference between stress relaxation data. Stress relaxation differences between 1s t and 2n d , 1s t and 3rd , and 1s t and 4th indentations are presented below (Figure 3-4). 66 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Although these results are limited by the number o f samples tested, the percent difference (positive and negative) between successive stress relaxation measurement are much higher than that obtained for the calf samples (Figure 3-4). 50 Ht1634 0 lb. Ht1523 0 lb. Human IVD Sample | B 1st Indent - 2nd Indent B 1st Indent - 3rd Indent □ 1st Indent - 4th Indent F igure 3-4: P relim inary rep eated stress relaxation results. Relative to the calf data, large percent differences were observed for human cadaveric specimens, -31.8% (st.dev.: 50.0%) vs. 8.7% (st.dev.: 7.9%). However, most human cadaveric specimens exhibited a dramatic decrease in percent difference with repeated indentation, indicating increasing stress relaxation obtained following the initial indentation (Figure 3-1 vs. Figure 3-4). Recall from prior studies that a 20% increase in viscoelastic property was considered a significant change due to fatigue. Therefore, a standard deviation of 67 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. percent difference o f less than this value is vital, in order to allow changes in viscoelastic properties to be observe with fatigue loading. The non-fatigued loaded samples o f this study, however, have exhibited a standard deviation o f 50.0%, making it difficult to observe changes in material property with fatigue loading, reliably. Earlier, improvements in percent stress relaxation differences were observed for the calf samples when initial indentation data was removed. With the assertion that the initial stress relaxation data was not representative of the sample’s material property, substantial decreases (8.7% vs. 4.7%) in the standard deviation o f the percent difference was obtained. A similar paradigm was used to improve the constancy o f repeated indentation measurements o f human IVDs. Results indicated that the standard deviation decreased from 50.0 % to 21.9%, after removing the initial indentation measurement from the data sets (Figure 3-5). However, in this case the results fail to satisfy our desired mean percent difference and standard deviation o f less than 20%. 68 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Ht1534 0 lb. Ht1634 0 lb. Ht1523 0 lb. fc -140 Q . T W t534 Ht1«34 0 lb. 0 lb. Ht1523 Olb. HHS34 Olb. r Ht1634 Olb. Human IV D Sample ■ 2nd Indent - 3rd Indent 02nd Indent - 4th Indent Figure 3-5: Preliminary repeated stress relaxation results. Initial indentation is removed. Summary of single point indentation method respect to multiple point indentation method, by determining the percent difference between left, right, and center stress relaxation measurement obtained from the initial indentation o f each sample (Figure 3-4). Although comparison o f percentages depends on which two differences are taken, it is interesting to note that viscoelastic properties of non-homogeneous human disc analyzed using the aforementioned method resulted in substantially lower percent difference values (Figure 3-2 vs. Figure 3-4). Calf specimen results presented earlier (Figure 3-2) can also be viewed with 69 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Multiple Point Indentation Stress Relaxation Measurement 50 Ht1634 Ht1523 Ht1534 -130 -140 J- Human IV D Sample ■ Center - Right ■ Center - Left □ Right - Left Figure 3-6: Preliminary repeated stress relaxation results. The failure to show efficacy o f single point indentation testing may be attributed to the small sample size used in the present study, as well as several possible sources of error. 1. Changes in material property during experimentation 2. Errors in precisely replacing the sample at the fixed location. 3. Errors generated due to small range o f measurements. 4. Errors generated by the MTS test system. Insufficient recovery from previous indentation and super-hydration of the indented tissue result in decreased viscoelasticity with repeated testing (Gray 2001, Syed 2002). A measure increase in viscoelasticity is expected when the tissue undergoes some degradation. An apparent decrease in viscoelasticity occurred in 70 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. only 1 of 3 cadaver samples, whereas, it was observed in 6 o f 8 calf samples. Earlier it was hypothesized that 36 hours o f free swelling in a saline bath resulted in high initial stress relaxation measurement (1st indentation) with respect to subsequent indentations. However, this does not hold true for 2 of 3 cadaver samples, which exhibited high stress relaxation measurements following the 1st indentation. It is hypothesized that the observed increase in the viscoelastic behavior from the first to the subsequent indentations may be due to changes in tissue material property, i.e. tissue degradation, caused by the initial indent. Further experimentation will be required to prove this assertion. It is also believed that large standard deviation between successive stress relaxation measurements may be attributed to errors in precisely locating the specimen’s fixed indentation sites. A grid pattern placed on the MTS fixture was used to replace the sample. The precision o f this grid pattern is limited to + 0.5 mm. Larger standard deviation may result if viscoelastic property o f degenerated tissue is found to be highly sensitive to location. Determination of the relative sensitivity of location o f viscoelastic property is beyond the scope o f this study. It is important to realize that these hypotheses presented thus far are made without the benefit o f a large sample size. Additionally, it must be noted that the range o f measurement o f stress relaxation for cadaveric specimens was 1.37 N to 4.57 N, versus 3.37 N to 7.07 N for calf specimens. A small error in measurement will translate into large changes in percent difference o f stress relaxation. 71 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Measurement errors may have been introduced in a subset of the data presented above by changes in rate o f loading the indenter. The ramp loading is controlled by the MTS machine, which uses the force from the load cell as its feedback. Although the target rate o f loading is set at 0.33 N /sec, speeds as high as 2 N /sec were observed. Faster indenter speeds will result in greater stress relaxation force measurement. In summary, single point indentation method has shown promise as a feasible non-destructive means of obtaining viscoelastic property data. Results from the verification study using calf specimens strongly indicate that the proposed method is equivalent to the established method of indentation. However, these results could not be duplicated on a limited number o f human cadaveric samples. Considering that multiple point indentation method is not a choice, additional feasibility testing o f single point method using human cadaveric discs may be the best alternative for developing a non-destructive method for testing degenerated tissue. In light o f the fact that these methods may be required to show efficacy of a pharmaceutical product to FDA’s (Food and Drug Administration) satisfaction, this alone merits further research into developing non-destructive test methods specifically for human IVDs. 72 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Chapter 4 — Effects of Disc Hydration on Material Properties Hydration Content of Intervertebral Discs Several experimental factors that influence viscoelastic material properties of intervertebral discs, such as load and strain rate, have been investigated systematically in the past. However, the impact o f IVD water content on viscoelastic material properties has not been adequately understood. This is of vital importance, since the amount of water content direcdy influences the viscoelastic behavior of the disc, which depends on the interstitial fluid movement through the micro-porous collagen matrix. The unique inner and outer structures of the disc, the nucleus pulposus and the annulus, are largely made up o f water, type I and type II collagen. The nucleus pulpous o f the disc is rich in proteoglycans, which aid in binding o f water. Consequendy, water makes up 85% o f the mass o f the nucleus, while the remaining 15% consists o f PG (proteoglycan) and type II collagen. In contrast, the concentration o f PG is lower in the annulus, thereby resulting in a water mass of 65% of the total, while the remainder consists of crisscross arrangement of the coarse collagen fiber bundles (Lindh 1989). The PG solution also, indirectly, plays two critical roles in determining the viscoelastic behavior of the disc. It has a high osmotic pressure and tends to inflate the collagen fibers by imbibing surrounding fluids (Urban and Maroudas 1981). This allows the disc to hydrostatically resist high compressive loads. In addition, PGs are 73 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. responsible for the fine effective pore size o f the collagen matrix, thereby providing high resistance to fluid flow (Urban and Maroudas 1981). With this in mind, it is particularly important to determine the initial water content in an IVD specimen and to understand the mechanisms of dehydration and re-hydration during testing. This allows indentation measurements to be interpreted as representative o f the tissue’s material property. Therefore, a study was designed to understand several fundamental factors that can influence the results of indentation measurements over a fatigue-loading regimen. The aim was to determine 1. If there exists a dehydration mechanism as a result o f cyclical compression loading. 2. A suitable control treatment to match the initial hydration level of the treated. The indentation studies summarized previously in chapters 2 and 3 have had the benefit o f the results of the disc hydration study outlined in this chapter. Consequently, the method for treating controls was determined to be a 36 hours soak in a saline bath, as oppose to a no-soak treatment used in earlier pilot studies. Prior Research on Quantification of Disc Hydration A s a result, experim enters have taken precautions to prevent dehydration o f the disc during testing. These include wrapping the disc with saline soaked towels, 74 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. immersing the disc in saline with an appropriate pre-load, or working in a high humidity chamber (Pflaster and Krag 1997). Although researchers often do not measure the effect of hydration on physical properties of the disc, they have quantified the process of disc hydration, and the resulting physical changes. Johnson et al. measured physical changes caused by hydrating porcine tail IVDs in normal saline (0.9% saline). The specimens were allowed to swell freely. The results concluded that there is a greater percent change in vertical height than in radial expansion (Johnson et al. 2001). The maximum radial swelling of 2% was reached within 40 minutes of hydrating the discs. The IVDs then reduced in size to their nominal radial dimension after 100 minutes. Vertical dimension also increased by 16% at an average linear rate of 0.064% per minute, until a plateau was reached. Similar experiments conducted by Pfaster and Krag focused on the effects of test environment on intervertebral disc hydration. Several methods for preventing dehydration of specimens during in vitro testing were examined by measuring the change in water content. These methods included, humidified air, wrapping with saline moistened towels, repeated spraying, and wrapping with plastic film. They concluded that specimen exposure to saline spray, plastic film wrap, or moistened towels did not result in a change in hydration (Pfaster and Krag, 1997). However, unloaded exposure to normal saline bath resulted in substantial swelling, which increased the water content by 24% over a 7 hours period. Approximately 44% of this increase occurred within the first 0.5 hour, which corroborates with prior research. 75 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The effects o f compressive load on swelling o f discs in a normal saline bath were studied as well. A load of 100 lbs was chosen as a nominal compressive force to simulate in vivo preload on intervertebral discs. The results indicated an 8% increase in hydration level for specimens exposed to a saline bath with a 100 lb preload (Pfaster and Krag, 1997). Preparation and Measurement of Hydration Level o f Intervertebral Discs Samples were prepared by slicing the disc in half along the transverse plane, and subdividing both halves (Figure 4-1). Tissue was removed from the posterior half o f the disc, since indentation viscoelastic measurements are taken along the outer annulus o f this region. Subdivision o f the disc halves was accomplished by determining the boundaries between the nucleus, inner annulus, and the outer annulus. The boundary between the annulus and the nucleus o f healthy disc tissue can be discerned by examining the change in structure from fibrous to gelatinous. The nucleus pulposus was then removed from both halves by cutting along this boundary and removing the tissue off the endplates. The remaining annulus was then subdivided into two equal parts to form the boundary between the inner and outer annulus (Figure 4-1). The outer annulus was removed by cutting around the posterior perimeter o f the endplates. The inner annulus was then removed by scraping the remaining tissue from the posterior halves of both endplates. 76 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Nucleus Anterior Inner Annulus Posterior O uter Annulus Figure 4-1: Sectioning of the disc. The regions of the disc dissected for evaluation o f water content are shown in cross hatch patterns. While the nucleus and the annulus have distinct boundaries, the inner annulus and outer annulus where prepared by dividing the annulus into two equal section. The cutting process required to subdivide the disc must be completed rapidly to prevent loss o f water through evaporation prior to weighing the samples. Measurement of hydration level was performed by determining the percentage o f water lost following oven drying o f tissue samples. The water content in a disc was determined by taking the difference between the mass o f the tissue before and after drying, and dividing it by the initial mass according to the following equation. Percent Water Content = Mass (wet)— Mass (drvi x 100 (4.1) Mass (wet) Samples were dried in an air oven at 90 degrees Celsius for a specified amount of time based on sample weight A minimum of 2 hrs drying time was used for all samples, while samples weighing greater than 1 gm were dried for additional hour for each additional 0.5 gm o f weight Generally, calf specimens were dried for 2 hours, while much larger human cadaver IVD samples harvested from the outer annulus region required longer drying times. 77 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Finally, it must be indicated that due to the lack of precision by the weighing scale (± 0.01 gm) to measure such small differences in tissue weight, and the fact that water evaporates during sample preparation at a rate of 0.07 gm per hour, measurement errors can add up to 1.5-2%. Therefore results are much more meaningful when water content changes greater than 2% are observed. A Study Designed to Determine the Existence o f a Dehydration Mechanism in Response to Fatigue Loading A pilot study was designed to investigate if fatigue loading o f IVD specimens can result in dehydration o f the disc tissue through exudation o f fluid. If there exits such a mechanism, it may have a significant impact on the accuracy of measured viscoelasitc material property at post fatigue cycling. Calf IVD specimens, which are rich in proteoglycans, were employed for this study due to their homogeneity between samples. Four types o f treatments were tested for fatigue-induced disc dehydration. These included a 0.33 Mol % Genipin in phosphate buffer solution (1 x PBS), a cross-linker used previously to prepare treated samples for indentation testing, along with three experimental salt solutions used to prepare the controls. Note that it is hypothesized that the Genipin treatment may have a positive effect on disc hydration due to the effect o f increased cross-linking on the micro-porous structure. The strategy for the control treatments was to vary the salt content in the 0.5 L bath o f distilled water used to soak calf IVDs. A primary reasons for examining 78 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. several treatments was to find a hydration matched control to the Genipin treated samples. Since water will flow from the side of low solute concentration to the side of high solute concentration, increasing the solute concentration of the bath will result in decreased water content in the disc. Therefore, a solution of 0.15 M NaCl (normal saline), simulating the fluid surrounding the spine, was prepared, along with a higher concentration 0.30 M NaCl solution (Table 4-1). In both cases, the discs were soaked for a period of 36 hours (±1 hour) to obtain equilibrium between the water content in the disc and the bath. Treatment Description Number of Specimen Loading Condition at the Time of Dehydration Testing Genipin (Tl) 0.33 Mol% Genipin Treatment 4 Pre-Fatigue 0.33 Mol% Genipin Treatment 4 Post 3000 Fatigue Cycles 0.33 Mol% Genipin Treatment 3 Post 6000 Fatigue Cycles Control 1 (Cl) As Harvested (Fresh Frozen) 3 Pre-Fatigue As Harvested (Fresh Frozen) 3 Post 3000 Fatigue Cycles As Harvested (Fresh Frozen) 3 Post 6000 Fatigue Cycles Control 2 (C2) 1.00 M NaCl Solution 5 Pre-Fatigue Control 3 (C3) 0.30 M NaCl Solution 3 Pre-Fatigue 0.30 M NaCl Solution 3 Post 3000 Fatigue Cycles 0.30 M NaCl Solution 3 Post 6000 Fatigue Cycles Control 4 (C4) 0.15 M NaCl Solution (Saline) 3 Pre-Fatigue 0.15 M NaCl Solution (Saline) 3 Post 3000 Fatigue Cycles 0.15 M NaCl Solution (Saline) 4 Post 6000 Fatigue Cycles Table 4-1: Treatment and loading condition of specimens. The third control treatment consisted of non-soaked discs, which represented hydration level of fresh frozen samples (Table 4-1). These samples were allowed to thaw while wrapped in saline soaked towels to minimize dehydration. 79 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Additionally, a 1M NaCl salt solution, which was previously investigated in a pilot study, was also examined. However, high concentration o f salt dramatically depleted the water content in the nucleus o f the disc, and therefore it was removed from the fatigue testing. Each IVD was subjected to cyclical compressive loading in a manner similar to the protocol established for indentation testing, as described in chapter 2. However, specimens were subjected to one o f two compression-flexion fatigue cycling regimen, either 3000 or 6000 cycles (Table 4-1). A set o f specimens from each treatment group were not cycled in order to observe changes in hydration content from 0 to 6000 fatigue cycles. Analysis of hydration content with respect to fatigue hading Statistical analysis to determine the effect o f fatigue loading on hydration content o f the disc was carried out independently for the nucleus, and inner and outer annulus o f the disc. Spearmen correlation was used to determine if there exists a significant change in the water content in these regions due to repeated loading. The hydration results o f the outer annulus for all treatment groups indicate that there is no significant change (p>0.05) in the water content due to fatigue cycling (Table 4-2). A primary reason for not observing a decrease in hydration with fatigue cycling may be due to the fact that all samples were wrapped in saline soaked towels during testing. This prevented samples from drying out, thereby allowing them to maintain their initial hydration level through out the fatigue cycling test. 80 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The average percent hydration for Genipin treated samples, and 0.15 M and 0.30 M NaCl solution samples was similar at 70.5% (± 1.5%), while that o f non soaked samples was 66.3% (± 1.5%). It is interesting to note that the saline soaked towels, despite being wrapped around samples for up to 3 hrs during fatigue cycling (6000 cycles @0.5 Hz), had no effect on the percent hydration level o f the non soaked group. This indicates that the rate o f water absorption from the soaked towel is approximately the same as the hydration loss during fatigue loading. A submersion, and perhaps increased hydrostatic pressure, may be required to force the fluid into the annulus. Genipin (%) Non-soaked (control 1) (%) 0.30 M NaCl (control 3) (%) 0.15 M NaCl (control 4) (%) Pre-Fatigue Water Content 71.31 65.38 70.34 71.23 Post 3000 Cycles Water Content 70.63 66.24 71.29 70.25 Post 6000 Cycles Water Content 70.38 67.24 68.60 70.42 Spearman Correlation Results (p) 0.326 0.493 0.329 0.485 Table 4-2: Effect of fatigue loading on water content of disc outer annulus. Results of percent water content with fatigue loading for Genipin treatment and three proposed control treatments. The Spearman correlation results of p>0.05 support the hypothesis that fatigue loading does not influence disc hydration content (bovine specimen). In contrast to the results observed for the outer annulus, the nucleus exhibited significant change (p<0.05) in water content due to fatigue cycling for at least one treatment group (Table 4-3). Controls 3 and 4 exhibited high pre-fatigue water content similar to the Genipin treated specimens. However, a significant 81 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. change in hydration content of control 3 (p— 0.001), and near significant level for control 4 (p=0.076), was observed with respect to fatigue cycling. This may indicates that the cyclical compressive loading has forced a small amount o f fluid out o f the nucleus, and into the annulus or adjacent endplates. Moreover, all control samples asymptotically approached a range o f 78% to 79%. Perhaps a larger compressive load may induce greater amount o f fluid movement out o f the nucleus. Note that this trend was not observed in the Genipin treated samples. Perhaps this can be explained by the fact that the effect o f increased cross-linking has reduced the effective pore size o f the collagen structure, thereby reducing the ability o f the disc to exude fluid. Although a percent hydration content o f 79.48% was observed at 3000 cycles, that may be indicative of the amount o f water absorbed by this sample group, rather than a result o f the repeated loading. Genipin (%) Non-soaked (control 1) (%) 0.30 M NaCl (control 3) (%) 0.15 M NaCl (control 4) (%) Pre-Fatigue Water Content 81.02 79.81 81.27 82.32 Post 3000 Cycles Water Content 79.48 79.16 79.76 79.29 Post 6000 Cycles Water Content 80.21 79.36 78.42 79.73 Spearman Correlation Results ,< * > ) ..................... 0.216 0.493 0.001 0.076 Table 4-3: Effect of fatigue loading on water content o f disc nucleus. Results of average water content with fatigue loading for Genipin treatment and three proposed control treatments. The S p e a r m a n c o r r e la tio n r e s u lts o f p > 0 . 0 5 s u p p o r t th e h y p o t h e s is th a t fa tig u e lo a d in g d o e s n o t in f lu e n c e disc hydration content of like specimen (bovine specimen). 82 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The data for the inner annulus must be examined cautiously, as it is the boundary between the annulus and the nucleus, and can be comprised o f mixed results. Statistical analysis indicated that there is no significant change in water content due to fatigue loading. However, a near significant trend (p=0.08) was observed for the Genipin treatment group, indicating a decrease in percent hydration with fatigue loading. This may suggest that there may be a fluid driving force from the outer and inner annulus to the nucleus. In addition, the average percent hydration o f the non-soaked group was equivalent to the saline soaked group, 74.64% versus 74.74%, respectively, further indicating poor reliability o f the data in this region to demonstrate any significant trend. Genipin (%) Non-soaked (control 1) (%) 0.30 M NaCl (control 3) (%) 0.15 M NaCl (control 4) (%) Pre-Fatigue Water Content 76.81 75.45 75.16 75.80 Post 3000 Cycles Water Content 75.09 74.87 76.25 73.56 Post 6000 Cycles Water Content 74.06 73.61 75.31 74.85 Spearman Correlation Results t o ............................. 0.08 0.407 0.767 0.645 Table 4-4: Effect of fatigue loading on water content of disc inner annulus. Results of average water content with fatigue loading for Genipin treatment and four proposed control treatments. The Spearman correlation results of p>0.05 support the hypothesis that fatigue loading does not influence disc hydration content (bovine specimen). Results from a Study Designed to Determine a Hydration Matched Control Results o f the hydration study compared the percent water content o f three controls against Genipin treated samples in each region o f the disc, in order to 83 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. determine the best hydration matched control (Table 4-5). Mann & Whitney test was utilized to determine statistical significance. The null hypothesis states that there is no difference in the hydration level of the specified control and treated samples. Since it was not conclusively shown that fatigue cycling has no effect on hydration content, the results were further segregated, not only by region, but also in terms of the amount o f fatigue a set of samples received. Nucleus Inner Annulus Outer Annulus Cl C2 C3 C4 C l C2 C3 C4 Cl C2 C3 C4 Mann & Whitney Test 0.48 0.03 1.00 0.16 0.29 0.05 0.29 0.29 0.03 0.62 0.16 0.72 Table 4-5: Results of comparison of hydration level of four control treatments with Genipin treatment. The Mann & Whitney Test was utilized to test the hypothesis that there is no change in disc hydration for like specimens due to Genipin treatment. Results shown for pre-fatigue samples. Comparisons of pre-fatigued controls with the treated samples are presented in the table above. Results of the Mann & Whitney test concluded that there is no significant difference (p>0.05) in the amount of water content between Genipin treated samples and 0.30 M and 0.15 M NaCl soaked controls (C3 & C4 in Table 4- 5). These results were observed in all three regions of the disc. However, pre-fatigue hydration results are varied for non-soaked and 1 molar NaCl controls. The hydration level of the nucleus and inner annulus of the non- soaked samples (Cl in Table 4-5) were not significantly different than the Genipin treated samples (p> 0.05). However, the outer annulus of the non-soaked control resulted in a significantly lower hydration level when compared with its Genipin treated counterpart (p= 0.03). This was expected, as non-soaked samples are subject 84 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. to drying out following extraction o f the spine until it is frozen. The outer annulus, being exposed to dry ambient environment, will tend to dry out faster than the more protected nucleus, which is not directly exposed. The opposite effect is seen for the 1 M NaCl sample group (C2 in Table 4-5), where the pre-fatigue hydration of the nucleus and inner annulus was significandy different than the hydration level of the treated samples (p< 0.05). Note that a 36 hour soak in a 1 M NaCl solution has the effect o f equilibrating the water from the side o f low solute concentration (i.e. nucleus) to the side of high solute concentration (i.e. 1 M NaCl bath). Therefore, it is expected that the nucleus, which has an abundance of water, would loose its water content, while no net effect would be observed on the outer annulus since it has relatively low initial water content. Consequendy, the pre-fatigue hydration content o f the outer annulus was not significandy different than that of the treated samples (p >0.05). Similar results were obtained from post 3000 and 6000 cycle hydration content data. However, the effect o f water exuding out o f the nucleus due to fatigue loading is evidenced from the low statistical significance achieved for saline soaked samples (C3 in Table 4-7). 0.30 M NaCl soaked samples dropped below hydration levels o f 79% in the nucleus at post 6000 fatigue cycles, which resulted in hydration content being significandy lower than that o f Genipin treated samples (p=0.05). 85 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Nucleus Inner Annulus Outer Annulus Cl C2 C3 C4 Cl C2 C3 C4 Cl C2 C3 C4 Mann & Whitney Test 0.72 N /A 0.72 0.72 0.72 N /A 0.29 0.48 0.08 N /A 0.72 1.00 Table 4-6: Results of comparison of hydration level of four control treatments with Genipin treatment. The Mann & Whitney Test was utilized to test the hypothesis that there is no change in disc hydration for like specimens due to Genipin treatment. Results shown for post 3k-fatigue cycle samples. Nucleus Inner Annulus Outer Annulus C l C2 C3 C4 Cl C2 C3 C4 Cl C2 C3 C4 Mann & Whitney Test 0.28 N /A 0.05 1.00 0.83 N /A 0.28 0.29 0.05 N /A 0.28 0.48 Table 4-7: Results of comparison of hydration level of four control treatments with Genipin treatment. The Mann & Whitney Test was utilized to test the hypothesis that there is no change in disc hydration for like specimens due to Genipin treatment. Results shown for post 6k-fatigue cycle samples. When determining a hydration matched control, pre-fatigue data should be used in order to exclude the influence of fatigue loading on the choice for a control. That is, the control should match the starting point. Statistically, both 0.15 M NaCl (0.9 % saline) and 0.30 M NaCl (1.8% saline) make an excellent choice for a control treatment. The hydration levels match the Genipin samples well for either solution (Table 4-8). However, since the bodies own fluid property surrounding the disc in vivo is similar in nature to normal saline, the 0.15 M NaCl solution was chosen as the appropriate control. 86 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Nucleus (% hydration) Inner Annulus (% hydration) Outer Annulus (% hydration) Genipin 81.02 76.81 71.31 Non-soaked (control 1) 79.81 75.45 65.38 1 M NaCl (control 2) 75.96 73.64 70.98 0.30 M NaCl (control 3) 81.27 75.16 70.34 0.15 M NaCl (control 4) 82.32 75.80 71.23 Table 4-8: Average hydration content for each treatment. Results indicate the distribution of water content within the three main regions of the disc, nucleus, inner annulus, and outer annulus. Table represents pre-fatigue values. Effect o f Hydration Content on Viscoelasticity The hydration study results indicate that the choice o f control treatment dictates the amount o f water content in the disc. Further more, a weak correlation was discovered between fatigue cycling at 200 N load level and water content, which suggested a decrease in water content of the nucleus over the course of 6000 cycles. This change in hydration may pose a significant impact on measured indentation creep and stress relaxation. For this reason, Gray designed a study to determine the impact o f control treatments, more importantly the hydration content o f the disc, on measured viscoelastic properties (Gray 2001). A small sample size o f calf disc were divided into two treatment groups, normal saline soaked (0.15M NaCl) and non-soaked. Samples were subjected non destructive indentation testing to assess viscoelastic mechanical properties. Stress relaxation and creep was measured on the posterior outer annulus o f the disc. 87 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Additionally, water content of the discs was measured to verify differences between the two treatments. The results o f this experiment indicated that stress relaxation generally does not change with hydration content, 96.8% o f the maximum value (Figure 4-2). Creep behavior o f annular tissue, however, does decrease with increase in water content, 74.4% o f the maximum value (Figure 4-2). These results suggest that quantifying hydration level o f test specimens is vital for proper interpretation of material property data. Mechanical Properties as a Percentage of Maximum Value Stress Relaxation Creep 0 0 .9 % Saline Soaked B N on-Soaked Figure 4-2: Effect of hydration on viscoelastic properties. Comparison of stress relaxation property and creep property of soaked and non-soaked samples of bovine intervertebral discs (Gray and Hedman 2001). Generally, higher hydration level greatly improves resistance to creep deformation. The difference in hydration dependence between creep and relaxation could be explained by the existence o f ongoing deformation, and thus displacement 88 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. of fluid, during a static creep test while there is no change in deformation (and consequently little fluid flow) during a stress relaxation experiment. Changes in Hydration o flV D Observed for Human Cadaver Specimens Hydration content o f human cadaver discs was recorded during nondestructive studies, presented earlier in chapters 2 and 3. In contrast to the calf specimens, human cadaver discs comprise morphological changes due to degeneration. Scientists have demonstrated that these alterations parallel the changes in the biochemical characteristics o f the disc. As a consequence, the water content in a human disc is highest at birth, and decreases with age in both regions, approaching a value of 70%. Therefore, hydration results o f human cadaver specimens must be viewed with respect to the level o f degeneration. Normal saline soaked cadaver IVD specimens were segregated by regions at post 6000 fatigue cycles, and then weighed and dried in accordance with the method described for the calf specimen hydration study. A 4-point grading scheme based on the method of Rolander, described in chapter 2, was used to determine the level o f degeneration. The results were then grouped by degeneration score and treatment and then plotted (Figure 4-3, Figure 4-4, and Figure 4-5). Nucleus o f a human disc contains 88% water mass at birth, and the annulus can contains up to 78% w ater m ass. T h e specim ens for this study range from as young as 32 years, to as old as 90, and will contain noticeably lower levels of water content. The results o f hydration content of saline treated discs are plotted below 89 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. (Figure 4-3). The results indicate the hydration level o f the disc progressively decreases with degeneration. More importantly, the hydration levels o f all regions converge to the same value, beyond a degeneration score o f 1. This is evidenced by the fact that the disc loses proteoglycans as it ages, thus the nucleus is unable create the osmotic pressure required to retain a high level o f hydration. 85 80 75 70 '5 65 •3 X 60 55 3 0 1 2 Degeneration Score (0-3) Nucleus — ■ —Inner Annulus — £ —Outer Annulus Figure 4-3: Hydration content of saline treated cadaver discs. Results show a change in hydration with morphological changes in the disc. The results o f the Genipin treated human cadaver discs indicated that the hydration content o f the inner annulus decreased dramatically for samples with a degeneration score o f 3. This may have been due to the fact that only one lumbar spine segment was sampled for both treatments, therefore, the impact o f these results is limited by the sample size. 90 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 75 - ■ a 70 0 2 1 3 Degeneration Score (scale 0-3) — Nucleus — ■—Inner Annulus — A—Outer Annulus Figure 4-4: Hydration content of Genipin treated cadaver discs. Results show a change in hydration with morphological changes in the disc. Qualitatively, the hydration content o f the nucleus o f saline treated samples appears to be greater then the Genipin treated samples. However, the opposite was shown to be true for calf specimens. This further indicates that the level of degeneration in human cadaver specimens can greatly affect results. The combined hydration content data for both treatments are presented in terms o f specimen age (Figure 4-5). The hydration level for all three regions converges as discs age, possibly indicating a biochemical change with age. 91 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 80 <J 70 - ■ 60 61-75 76-90 30-45 46-60 Range of Cadaver Age (years) ♦ 1 Nucleus — ■— Inner Annulus — A— Outer Annulus j Figure 4-5: Change in hydration content of the disc with age. Summary of Hydration Content of Disc Study The series of experiments presented above suggests two general guidelines for viscoelastic material property testing o f intervertebral discs. First, hydration content of controls must match that o f treated samples, so that measurements reflect differences in tissue material properties. Second, just as with degeneration score, hydration content must be reported along with material property data, in order to clarify understanding of the overall results. 92 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Chapter 5 — Destructive Testing of Human Intervertebral Discs Material Characterisation o flV D M atrix by Means of a Destructive Test Tensile forces exceeding the physiological limit of the disc as a consequence of direct trauma, or age related degradation o f the tissue, have been reported to cause failure o f the annulus fibrosis (Osti et al. 1990). Further more, rim lesions and circumferential tears in the annulus have been shown to be a prerequisite to prolapse of the nucleus, which leads to accelerated degradation of the disc (Osti et al. 1990). Therefore, the study o f load deformation response o f the annulus fibrosis by means of a destructive test is particularly interesting in understanding the elastic-plastic behavior of the disc and the effect mechanically induced degradation has on these properties. A Rationale fo r Studying the Destructive Material Properties of the Posterior Annulus Fibrosis Scientists generally take one o f two approaches in their pursuit to identify the pathogenesis o f disc degeneration. One such approach is a qualitative one, which relies on relating observable changes in disc morphology to a known, or predetermined, condition. A s an exam ple, Friberg and H irsch attem pted to link disc degeneration to ruptures observed in the posterior annulus (Friberg and Hirsch, 1950). They postulated that degeneration observed as a result o f ruptures were separate from 93 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. notmal physiological aging. Additionally, they concluded that annulus ruptures were observed more frequendy in the posterior of the two lower lumbar discs. Hilton and Ball analyzed the frequency, distribution, and histological characteristics of rim lesions from a large population o f spines. They concluded that the rim lesions in a population over and under the age of 20 were equally common, and may be the result o f direct trauma to the disc (Hilton and Ball 1984). Based on these findings, Osti et al used animal models to test the hypothesis that discrete peripheral tears within the annulus lead to secondary degenerative changes in other components o f the disc (Osti et al. 1990). In their experiment sheep were preconditioned with 5 mm deep lesions located circumferentially around the annulus of the lower lumbar discs. They concluded that although collagen can be synthesized to repair rim lesions, it does litde to prevent the accelerated biochemical degradation o f the nucleus and inner annulus as a result of repetitive loading and protrusion through the lesions. Other researchers have taken a more quantitative approach by measuring the material properties of the disc in response to a specific applied loading condition. However, the types of loading condition and test environment are often controversial. Galante tested the tensile elongation o f the annulus fibrosis in the fiber directions as well as alternate angles to the fiber direction. Strips o f 2x1 mm cross- sections o f annulus fibrosis with a gage length o f 10 mm were loaded in 0° (i.e. horizontal direction), 15°, 30°, 50°, 70°, and 90 degrees. He concluded that the 94 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. annulus is most extensible in the vertical direction, and least extensible in the fiber direction. Based on that evidence, it was postulated that the disc annulus is subjected primarily to tensile stresses acting between 0 and ± 30 degrees (i.e. the fiber direction of the lamella) (Galante 1967). In a similar study, Ebara et al. reported that tensile behavior of multiple layer non-degenerated annulus samples varied more with radial position in the disc than from anterior to posterior regions (Ebara et al. 1996). In addition, the study found that fiber orientation, and the number o f layers could affect the tensile properties of the annulus fibrosis. While both studies focused on quantification of material properties of the disc, they did not relate any specific condition o f the disc, natural or mechanically induced, to the changes observed in material properties. In addition, variation in material properties was attributed to age, or structural composition o f the disc. In contrast to the studies cited above, we plan to condition the disc annulus samples by cyclically loading them to sub-traumatic, pseudo physiological limits while evaluating the change in their destructive material properties. More specifically, only the posterior annulus will be subjected to this type of loading. Our rational has been formed by prior studies that have indicated the prevalence of degradation in this region of the annulus (Friberg and Hirsch, 1950), (Osti et al. 1990). The loading of the samples will also differ gready from prior studies. Excised samples from the posterior o f the disc, consisting o f upper and lower 95 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. vertebral bone and a section of the annulus, will be loaded in the longitudinal direction. This is in stark contrast to the loading regimen proposed by Galante and Ebara, who loaded single and multiple layer o f the annulus in the tangential direction (Galante 1966). The objective o f their investigation was largely to obtain destructive material properties o f minimally degenerated, annulus fibrosis tissue that makeup the disc. Their rational for tensile testing the annulus fibrosis layers in the tangential direction was that in young healthy discs vertical pressure on the nucleus is transformed into tangential stresses estimated to be three to Jive times the applied stress. In contrast to the tangential stresses, the vertical (longitudinal) stresses on the annulus were thought to be low for substantially high compressive loads on the nucleus (Nachemson 1963). However, in degenerated discs, where the nucleus has a reduced capacity to transform compressive stresses to tangential stresses, most o f the load is found to be carried by the vertical forces on the annulus (Nachemson 1966). Additionally, leaning forward while lifting a weight can lead to excessive tensile forces on the posterior annulus o f the disc. Since our primary goal is to measure the material properties of moderately degenerated disc tissue subjected to this type o f cyclical loading, a rational for tensile testing the posterior annulus tissue in the longitudinal direction rather than the tangential direction may be valid. The objective o f this chapter is to develop the protocol by which posterior annulus tissue can be tensile tested in the longitudinal direction. In addition, a preliminary study designed to test these methods will be carried out on human cadaver disc. 96 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Earlier non-destructive material properties were measured to determine the ability o f Genipin on improving the fatigue resistance of moderately degenerated human cadaver discs by cross-linking the collagen matrix. Therefore, another objective o f this study is to continue the evaluation of the Genipin treatment on disc tissue by evaluating the change destructive material properties against a control. The Tensile Test The tensile test provides the most basic description of a material. In this simple test, the load required to produce a given elongation is monitored as the specimen is pulled in tension at a constant rate. The load deformation data is then normalized for geometry, resulting in a stress vs. strain curve (Figure 5-1). Soft Tissue Tensile Test Plot 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 Ultimate Tensile Strength Y ield Poinl Young's Modulus Toughness 0.2 0.4 0 0.05 0.1 0.15 0.25 0.3 0.35 0.45 Strain (mm/mm) Figure 5-1: Typical Stress vs. Strain Plot. 97 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Several interesting material characteristics can be obtained from this plot Notice that the stress-strain curve is divided into two distinct regions, elastic and plastic, which are defined by the deformation o f the material. Within the elastic region there is a linear change in stress with strain, allowing the deformed material to recover upon removal o f load. Following elastic deformation, the relationship between stress and strain is no longer linear and the material is said to be deforming plastically. Plastic deformation on the other hand is not recoverable when the load is removed, although a small elastic component is recovered (Shackelford 1988). The transition point from linear elastic to plastic deformation is often difficult to determine. The usual convention is to define a yield point, which is the intersection o f the deformation curve with a straight line parallel to the elastic portion and offset by 0.2% on the strain axis (Figure 5-1). The slope of this line is defined as the modulus o f elasticity, which is a measure of the materials stiffness. Other significant material properties are the ultimate tensile strength, toughness, and modulus o f resiliency. Ultimate tensile strength (UTS) is defined as the point o f maximum tensile stress on the stress-strain curve. The area under the entire stress-strain curve, defined as toughness, is a measure o f the amount o f energy required to break the sample. Similarly, the area under the elastic strain portion of the curve is termed modulus of resiliency, and is a measure o f the material's ability to absorb energy without deforming permanently. 98 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Challenges Presented in Tensile Testing o flV D There are many challenges in obtaining reliable tensile test data o f soft tissue. Unlike solid materials, where cross sectional area can be determined using conventional tools with relatively good precision, soft tissues require non-contact methods to obtain reasonably accurate area measurements. The desired mode o f failure in a tensile test is the failure o f the material along its gage section, defined as the smallest cross-section on a test sample. Failures at arbitrary location near the clamping point may not be representative o f the sample’s materials properties. Normally, a gage section can be machined or formed to ASTM requirements in solid materials, such as metals and most plastics, in order to prepare the sample for tensile testing. This precisely defines the length and cross sectional area o f the section with the highest stress. A similar approach must be taken with disc tissue, where a gage section must be formed by hand on each sample to define a point of failure. However, given the dimensional constraints o f the annulus section o f the disc, this task becomes increasingly difficult, resulting in variability between samples. Further more, since the disc is essentially a composite material formed by its alternating lamella structure, its material property may be dependent upon the number o f layers and the overall ratio o f width to depth o f the formed cross-section. T h e dim ensions o f the tissue sam ple’s gage section, along w ith a suitable test fixture, determines the outcome of a successful failure. When working with biological tissue, failures are often seen at the jaw, or sometimes samples slip out due 99 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. to poor ability o f the fixtures to hold the sample in place during testing. The viscoelastic nature o f soft tissue, adds yet another challenge in designing suitable holding fixtures. For the purpose o f testing longitudinal sections o f the IVD annulus, vertebral bone and polyurethane will be utilized to hold the disc tissue in place. The following sections detail the methodology involved in determining the solution to these challenges. The A r t Behind the Science of Perfect Failures Biological tissues pose a special challenge during tensile testing, in that, a controlled failure point is often difficult to obtain. Ideally, the point o f failure should coincide with the point where minimum cross-section area measurement was taken. Often, these two points are not the same due to flaws within the tissue or high stress concentration at the clamping point, which nucleate a premature tare in the tissue. In this study we are faced with having to hold the intervertebral disc segment at the weaker vertebral bone while pulling apart the disc annulus tissue. Therefore, fabrication o f a narrow gage section in the disc annulus tissue is required to obtain an acceptable failure mode. A method for preparing IVDs into longitudinal sections consisting of bone- annulus-bone was developed in our lab to ensure that the tensile failures are across the disc tissue gage section. T o prevent failures at the endplate, the ratio betw een the largest and smallest width along the annulus was kept at a ratio o f two to one (Figure 5-2). 100 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. endplate m axim um tissue width Figure 5-2: Schematic of a disc segment. Posterior annulus is shown along with the gage section. Formation o f a gage section required working with the disc tissue in its frozen state. This allowed precise formation o f an “hour glass” shaped gage section by drilling away the excess tissue as illustrated by the figures below. A treated frozen disc was initially divided into four segment o f equal width by scribing the top o f the cut vertebral bone. The scribed lines would continue across the posterior annulus to mark the three points where a hole would be drilled to form the “hour glass” gage pattern. An appropriate drill is then chosen based on the ratio described above to drill out the excess tissue. Once the excess tissue is removed, the disc is replaced back into the freezer for 30 minutes to 1 hour to refreeze the tissue, which has softened from the heat created by the drilling operation. The disc is then sectioned into four segments along the markings scored earlier by using a thin blade hacksaw (Figure 5-3). Segments are then replaced into the freezer for 30 minutes, or until frozen, depending on the thickness. 101 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 5-3: Sectioning the disc into segments. The last drilling step involves forming the depth o f the annulus section. Individual segments are measured to a desired 5 mm outer annulus depth plus the radius o f the drill. After marking this position, the tissue is then drilled transverse to the sagital plane (Figure 5-4). The remaining inner posterior annulus, nucleus, and anterior annulus are removed by slicing along the endplates with a scalpel (Figure 5- 6). 102 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 5-4: Marking the depth of the posterior annulus. Figure 5-5: Slicing the rem aining tissue. Disc segments should be trimmed, using a scalpel or a small scissor, to prepare for cross-sectional area measurement using the laser interference system. This requires carefully removing loose tissue that may interfere with the laser, or 103 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. unattached ligamentous tissue that do not add to the strength o f the disc annulus (Figure 5-6). The segments may be refrozen for this operation if necessary. Notice, that it is not desirable for the sample to undergo repeated freeze-thaw cycles, therefore, at each step samples should be refrozen if the slightest amount of softening o f the tissue is evident. Figure 5-6: Prepared posterior annulus segm ent. Medial segment is shown. Non-Contact A n a Measunment of Soft Tissues Assessment of mechanical properties o f soft tissues requires accurate measurement of the cross sectional area. Several techniques have been developed from weighted micrometer measurement with error correction to image reconstruction technique based on measurements from collimated laser beams (Lee and Woo 1988). 104 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The accuracy o f most techniques is limited by the geometry of the cross section. For example, mechanical micrometers, although highly accurate, can only measure thickness and require circular or square cross sectional shapes in order to accurately determine the area. Additionally, since a small amount o f pressure is applied to the part during measurement, soft tissues become inherently difficult to measure accurately. Collimated laser measurement systems, however, do not require contact with the specimen, and achieve accuracy o f 0.0001 in. or greater when thickness of solids or soft tissue are measured. Many sophisticated laser micrometer systems employed for measurement of cross-section area o f soft tissues use image reconstruction techniques to determine the area. The cross-sectional shape is obtained by rotating the specimen in the path of the laser beam and collecting several hundred width measurements. Once the cross-sectional shape is reconstructed, the area can be calculated by integration (Woo 1990). Chan proposed an alternative method o f measuring cross-section, where the specimen remains stationary while a laser reflectance system rotates around the specimen and measures the distance to the surface (Chan et al. 1996). Since the laser is fixed about an axis o f rotation, the area o f its circular path is known. The area around the specimen can be determined by obtaining the laser transducer to specimen surface distance measurements and integrating using the trapezoidal rule. The greater the number o f points obtained by the laser, the higher the accuracy will be. The area o f the specimen can be determined by subtracting the total area o f the laser’s path by the area around the specimen. 105 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. A similar system was built in our lab to measure the cross-sectional area of prepared posterior annulus disc segments (Figure 5-7). Specimens were placed under light tension using upper and lower jaws attached at the vertebral halves. The MTS material testing software was utilized to sample the laser’s output at a rate o f 0.05 seconds while the laser rotated at a speed of 10° per second. This resulted in approximately 670 data points for one full revolution around the specimen. A minimum of three trials were recorded for each disc segment and the area was determined by analyzing the data and smoothing the errors from the raw data files (drawing o f laser reflectance system located in appendix). 106 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. R 77.722 mm 65 mm ■ 75 mm. 65 mm - 95 mm Lb ** icSQ C O O mm **icdy Laser itSO.DOO nun exactly Radius o f rotation o f thelasechead * Figure 5-7: Schematic of laser reflectance system: Two instantaneous positions of the laser module shown moving in the clockwise direction. The specimen must be within the 65-95 mm range of the laser. Notice the trapezoidal area piece being formed between the two instantaneous measurements. 107 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Figure 5-8: Area measurement of the posterior annulus disc segment using a laser reflectance system. Y'ixtuning o f Specimens In a quest to force tissue samples to fail at predefined failure points, researchers have devised clever methods. Often a wire reinforced potting method is used to fix the ends o f a specimen. A less common approach has been to develop clamps that strengthen the ends o f a specimen with respect to the gage (mid) section by freezing the tissue. Fixturing o f IVD specimens in preparation for the tensile test was done in the usual manner by reinforcing the specimens with screws and rigidly fixing them in polyurethane blocks. In addition, a swivel clamp was constructed to hold the upper potting block to the MTS tensile testing machine’s pull rod. This swivel joint allows the specimen to be loaded axially with minimal application of a moment, normally caused by misalignment in a rigidly fixed system. 108 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The cortical bone was utilized over the weaker spongy bone for ridged fixation of the specimen during tensile testing. Screws were driven through the inferior and superior cortical segments to increase the hold with the potting material (Figure 5-9). ' Sifbey wixe f o i p o ttin g Figure 5-9: Completely restrained specimen ready for potting. Mid posterior annulus specimen shown. Notice the screws in the cortical bone. Safety wire prevents disc from being loaded during handling, and is cut just prior to testing the specimen. During the initial course o f experimentation with different methods of fixation, it was determined that the interface between the endplate and the disc tissue segment was the weakest link, specifically with degenerated discs. To improve the strength o f this interface, several steps were taken. The interface was reinforced with a string to prevent premature failure (Figure 5-9). Secondly, carefiil attention was paid to keep a minimum o f 1 mm distance between the disc tissue and potting level, to avoid denaturing o f the tissue from the heat produced by the exothermic reaction 109 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. of the 2 part polyurethane solution. By following these steps, failure on the gage section was observed in more than 75% o f the samples. A Proposed Method for Pensile Testing A protocol for tensile testing of the posterior annulus segments could then be established based on preliminary work completed to improve the method for holding and measuring the cross sectional area of the disc segment. Recalling that the objective o f the non-destructive testing was to relate elastic and viscoelastic material property changes on the posterior annulus of the disc to repetitive loading o f the IVD complex. The current study aimed to quantify degradation o f the posterior annulus by relating material property changes measured using destructive testing methods to repetitive loading of the disc segment. We further examine the relationship between collagen cross-linking in the IVD matrix and degradation o f disc tissue by studying the effects of a cross-linking reagent treatment. Human cadaver lumbosacral motion segments consisting of a section o f the superior and inferior vertebral bodies and the disc were entered into the study. When ever possible, attempts were made to select a spine from each age category, one spine from 30-40, one from 40-55, and another from 55-70 (Hedman 2001). All pre p rocessin g o f the sam ple w ill be d on e in the frozen state, in order to allow precise form ation o f the disc gage section using a drilling m ethod. In addition, 110 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. repeated freezing and thawing will be avoided by reffeezing the sample after each preparation step or after 5 minutes o f elapsed time, which ever comes first. Prior studies by Hoshaw have measured a change in strength and stiffness of canine femoral bone due to fatigue damage (Hoshaw 1997). In this study the control femur o f each pair was loaded to failure, and then the contralateral femur was cycled at 50% o f the ultimate load of the control femur. After completing the cycling regimen, the femur was tensile tested to failure, to determine the change in material properties due to fatigue. Using this testing strategy, Hedman has shown that a significant amount o f degradation of disc tissue can occur at cycling loads equivalent to 33% o f theyield stress (Hedman 2001). A test paradigm similar to Hoshaw’s will be used to measure changes in material properties due to fatigue loading of matched pairs of posterior annulus disc segments. Frozen discs will be sectioned along the sagital plane into left and right lateral and medial pairs. Each pair of lateral or medial segments will be prepared and measured for cross-sectional area according to procedures outlined earlier. One segment from each pair will be used as a non-fatigued control. The non-fatigued control will be wrapped in a saline soaked towel and set aside for 3.33 hours to replicate the duration of time that the fatigued specimen would undergo cycling. After the time has elapsed, the non-fatigued control will be tensile tested to failure, and the yield stress will be calculated from the resulting load deformation data. The yield stress of the non-fatigued control will then be used to set the cycling load limits of the contralateral segment at 33% of the yield stress. I l l R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The MTS model 870 servo-hydraulic axial testing machine was utilized for fatigue cycling as well as tensile testing o f the sample. Once the cycling was complete, samples were tensile tested to obtain load deformation data. From this data the modulus of elasticity, yield point, and the ultimate tensile strength was calculated. During the entire preparation and testing process it is critical to keep the sample hydrated as loss of water content can change the tissue’s material properties. Since non-fatigued samples remain stationary for the duration prior to cycling the fatigued samples, preparation steps were staggered to prevent spoilage o f the samples. During idle time, samples were kept in a 2 ± 1 0 C refrigerator. Samples will be treated with a cross-linking solution, Genipin, and a control saline solution to examine the effect o f the cross linker on fatigue resistance o f the disc annulus. Initially a treatment method consisting of soaking individual pairs of disc segments in their respective solution for 36 ± 1 hours was chosen. However, upon further evaluation, it was determined that both the control and treatment solutions were swelling the disc segments. For that reason, an alternate method was chosen where the intact intervertebral disc complex was soaked in solution for 36+1 hours, and then frozen to prepare for cutting individual segments. Preliminary Destructive Material Property Results of Posterior Annulus Disc Tissue The test results presented here are limited by the small samples size (n=4), and therefore are primarily qualitative. However, valuable information can be 112 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. learned o f the material properties of the posterior annulus tissue, as well as the tensile testing procedure itself. The results from the preliminary study suggest that there is a difference in material property with location on the disc annulus. Therefore, the following plots of material property results will indicate the location of the disc annulus that was tested, medial or lateral. The mean material properties for moderately degenerated posterior annulus are presented for both medial and lateral locations o f the disc (Table 5-1). Mechanical Properties Lateral Control Samples (Mean Value) Medial Control Samples (Mean Value) Modulus of Elasticity (Mpa) 11.77 22.05 Modulus of Resiliancy (kj / cm2 ) 0.25 0.27 Yield Strength (MPa) 1.89 3.14 Ultimate Tensile Strength (Mpa) 2.31 4.08 Toughness (kj/cm 2 ) 1.42 1.32 Table 5-1: Mean destructive material properties of moderately degenerated human cadaver discs. Typical properties of the posterior annulus are shown (n=5). The modulus o f elasticity is measured in the elastic region o f the stress-strain curve, and therefore is a reliable material property to compare differences between samples. Typically with soft tissues, other material properties such as toughness, UTS, and yield point are far less reliable. This is mainly due to the fact that these properties are measured in the elastic-plastic region of the curve, and are also dependent on the consistency of the failure. Therefore, comparisons between groups will be made by evaluating their respective modulus o f elasticity, while remaining material property comparisons will be included in the appendix section. 113 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The modulus o f elasticity fot non-fatigued lateral control samples was 47.8% less than that of medial control samples (11.76 MPa vs. 22.05 MPa, n=4). These results are in agreement with prior study by Galante, which concluded a significant difference in material properties with circumferential location at which the sample was harvested in the annulus. The change in material properties with location, however, was not limited to control samples. Similar increase in material properties was evident in treated samples from both medial and lateral locations. The mean modulus o f elasticity for non-fatigued lateral treated samples was 46.6% less than that o f medial treated samples (14.13 MPa vs. 26.32 MPa, n=4) (Figure 5-10). Similar trends were observed in other material properties with location on the posterior annulus and are presented in the appendix section. 40 Control (Saline) Treatment (Genipin) Group ■ Laterals ■Medials j Figure 5-10: Change in modulus of elasticity with location on the posterior annulus of pre^ fatigued specimens. 114 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. The preliminary study test results also indicated that there is an apparent change in material properties due to treatment (Figure 5-11). Similar to the results observed during non-destructive testing, a general increase in tissue strength due to cross-linking was observed following Genipin treatment. The mean modulus of elasticity of lateral treated samples increased by 20.0% over the control samples (14.13 MPa vs. 11.77 MPa, n=4), while medial treated samples showed an increase of 19.4% over the medial control samples (26.32 vs. 22.05). O ther material properties displayed similar changes due to Genipin cross-linking treatment (Refer to their plots in the appendix). | 15 "o 10 Laterals Medials Location ■ Control (Saline) ■Treatment (Genipin)J Figure 5-11: Test results comparing the change in tissue property with type of treatment for pre-fadgued specimens. Modulus of elasticity o f non-fatigue disc segments were compared with their contralateral fatigue segments for both control and treatment groups. The aim was to 115 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. determine if there was any difference in the fatigue resistance of control versus treatment group. Preliminary results indicate that there was a 3.4% decrease in stiffness due to fatigue (13.38 MPa vs. 12.92 MPa, n=3) o f medial control samples (Figure 5-12). While, medial Genipin treatment group exhibited a 3.6% increase (26.3 MPa vs. 27.3 MPa, n=4). When looking at the variance, essentially no change in strength due to fatigue was observed with either control or treatment groups. 45 40 - - - Control (Saline) Treatment (Genipin) Group B Non-Farigued B Fatigued j Figure 5-12: Change in modulus of elasticity with fatigue cycling of posteromedial annulus samples. Similarly, inconclusive results are obtained from the lateral treated and control disc. Results indicate that there is a 77.1% increase in stiffness due to fatigue (11.77 MPa vs. 20.84 MPa, n — 4) of lateral control samples (Figure 5-13). While, lateral Genipin treatment group exhibited a 7.8% increase (12.83. MPa vs. 13.83 MPa, 116 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. n— 3). When looking at the variance, essentially no change in strength due to fatigue was observed with either control or treatment groups. 35 30 Control (Saline) Treatment (Genipin) Group B Non-Fatigued B Fatigued | Figure 5-13: Change in modulus of elasticity with fatigue cycling of posteriolateral annulus samples. Discussion of results According to prior research, material properties o f single layer disc tissue vary circumferentially as well as radially. A positive increase in material property was observed when tissue is sampled medially versus laterally. The results o f this preliminary study showed a similar conclusion, where medial samples were substantially stronger than lateral samples. We are interested in showing a 20% or greater decrease in material property strength with fatigue cycling for the control group, while ideally showing no change 117 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. in properties with fatigue for the treatment group. However, an increase in material properties with fatigue cycling was observed for the lateral control group, while the medial control group showed essentially no change. This may be attributed to fatigue cycling limits that did not degrade the tissue of the contralateral sample during cycling. However, primary limitation is that a larger sample size o f 16-24 discs is required to observe trends. One reason for this may be errors in measurement o f the area due to unattached soft tissue surrounding some o f the prepared specimens. These include pieces of PLL (posterolateral ligament) loosely attached, fatty ligmentous tissue surrounding the lateral section o f the disc, and the disconnected fibers on the periphery o f the gage section that were cut during the sample shaping process. Consequently, the area may be overestimated, resulting in reported pre-fatigue mechanical properties that are lower than actual values. In addition, the 33% yield stress criteria used to set the cycling load limits on the contralateral specimen may be lower due to lower calculated yield stress. As a result, samples may have been fatigued far below the desired yield stress limit of 33% to induce degradation of the tissue. Applying additional steps in removing tissues that are not connected from endplate-to-endplate can increase the accuracy of area measurement, and will most likely result in higher yield stress calculations. Another possible reason for not observing the decrease in strength o f control samples with fatigue loading may be due to the criteria placed for cycling load limit itself. A 33% o f the yield stress limit may not be high enough to induce tissue 118 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. degradation during fatigue cycling. Review of this criteria is beyond the scope o f this research. It must also be noted that the sample size for each group was small (n=4), and therefore clear trends cannot be deciphered. In addition, other sources o f errors such as the lack o f homogeneity o f the disc material contribute to increase in the variance of destructive material properties. Therefore, determining the effect o f a cross-linking solution in increasing the fatigue resistance o f human cadaver annulus tissue may require a large sample size. It may be interesting to repeat this test using calf discs, which are more circumferentially homogeneous in their structure. Although earlier tensile testing revealed that the growth plates on calf disc specimens were as yet un-fused in many o f the vertebra, resulting in bone failure rather than tissue failure, repeat o f those experiments using the latest techniques developed for the human specimens will be beneficial. This will allow us to view the material properties independent o f the disc tissue quality, which will give better insight into the test procedure and the cycling load criteria. It is important to recognize that the main purpose o f this chapter is to put into practice the techniques for tensile testing posterior annulus IVD segments in the longitudinal direction. In addition, a larger sample size will be required to determine statistically significant results due to inherently large variances observed with human cadaver disc specimens. Further refinement o f the techniques is discussed in chapter 6. 119 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Chapter 6 — Recommendations with Regards to In Vitro Testing of Human Cadaver Discs. Lessons learned from non-destructive and destructive testing of human cadaver lumber discs. The goal o f these studies has been to establish and refine the non-destructive and destructive testing techniques required to show the effectiveness of the Genipin cross-linking reagent in improving the fatigue resistance of the intervertebral disc by measuring changes in material properties to quantifying the degradation due to fatigue cycling. However, one key point that has reshaped the course o f this research is that viscoelastic mechanical properties o f human cadaver disc tissue are sensitive to morphological condition o f the specimen. That is, the material properties of the disc are intrinsically related to its level of degradation. However, this poses two problems which must be overcome in order to be able to compare the effect of fatigue on control versus treated human cadaver disc specimens. There is an increased variance in the initial pre-fatigue material property value due to varying level o f degradation between samples. Although samples were segregated by level o f degeneration based on a four level qualitative grading method proposed by Rolander (Rolander 1966), it is clear that this method o f determining the mechano-biological condition o f the specimen correlated poorly with measured mechanical property. In other word an improved method of screening the quality of the specimens is required to reduce the variance between samples in initial mechanical properties. 120 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Prior studies by Gray using calf IVD specimens have shown significant changes in material properties between controls and treateds. This was primarily due to the homogeneity in the morphology o f the calf samples. Therefore, human cadaver specimens with a low degeneration score will be desirable in order to reduce the pre-fatigue variance between samples. However, implementation o f such a method may not be practical, as it may be cosdy in terms o f obtaining young human cadaver specimens. Another difficulty that exists when comparing pre and post fatigue material properties as the lack o f homogeneity within the specimen as result o f random lesions and tears within the disc matrix and variability in the hydration content o f the nucleus. However, this was investigated in chapter 3, were single point indentation technique was introduced as an alternate method to multiple point indentation technique used previously. This allowed indentation measurements to be representative of tissue’s material properties rather than location on the disc. Screening for highest quality disc specimens and implementing single point indentation technique to determine non-destructive material properties can achieve a significant reduction in variance o f stress relaxation material property. While the evidence strongly indicates sensitivity o f viscoelastic material properties with degradation o f the disc tissue, the extent to which degradation effects destructive material properties will require further investigation. Additional inconsistency with regards to accurate measurement o f the effective cross sectional area (i.e. tissue that is actually loaded) o f specimens requires a larger sample size to delineate the effect o f tissue degeneration on destructive mechanical properties. 121 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Further consideration with regards to single point indentation method and control treatment Although the adaptation o f single point indentation testing method has made it possible to compare material properties at pre and post fatigue cycles, a tool must accompany this method that ensures the specimen are repeatedly indented at the same location after each fatigue cycle. Repeatability o f viscoelastic measurement can be improved by fastening the specimen block on an x-y table with fine adjustments (Figure 6-1). Currently, the specimen block is placed by hand on the MTS testing platform by aligning the block on the table’s grid pattern. The precision o f this method in relocating the indenter is ± 0.5 mm at best. By improving the repeatability of indentation measurements at single site, differences observed at post-fatigue can then be attributed to fatigue cycling rather than change in location. Figure 6-1: X-Y table with precise linear motion. (Excel-Seiki, incorporated) 122 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Additional improvements to single point indentation can be made by improving the method use to segregate specimens by determining the degeneration index locally. As described earlier, many researchers quantify degeneration as an average appearance o f the disc. That may be satisfactory for application where the applied nominal load is on the entire disc, as in axial compression loading. However, indentation testing is highly localized, therefore determination o f the level of degradation should be limited to the area o f indentation. Through visual analysis of transversely sliced disc specimens under the microscope, the level o f degradation o f a selected area may be qualitatively evaluated. In addition, a new set o f grading scheme must be established to group samples for appropriate statistical comparison. As a last note, it will be virtually difficult to compare treated vs. control samples, unless they are adjacent levels from the same spine. O r possible half o f a disc that is injected with the control treatment while the other half is not. This is actually an important consideration, as it represents a more realistic situation. Consider injecting one side of the disc with treatment, while the other side remains as is. Place the disc in the saline solution, as per the control treatment protocol, and test each respective side. One must also consider weather this solution will washout as well. Further improvements with regards to destructive testing Despite achieving high levels o f precision in area measurement, additional accuracy may be gained by upgrading the laser reflectance module. Since the tissue structure, 123 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. specifically at the edges, is some what translucent, replacing the current laser with one that measures translucent objects may reduce the number of missing data points. Although an algorithm was used to replace the missing data points, files containing more than 100 missing points tend to create area measurement errors in access of 15-20%. Although this occurred infrequently, the number of missing points can be dramatically reduced by replacing with a laser that reflects well off o f translucent objects. Another improvement on the laser reflectance system may be the addition of an x-y linear motion plate to facilitate the adjustment of the laser beam on smaller samples, such as a rat tail disc. Currently the sample is placed by hand in the path of the laser, which requires additional time to setup the specimen. Treatment and Delivery Issues Thus far the treatment method o f specimens has been a 36 hour soak in an appropriate solution. However, practically speaking, the cross linking reagent needs to be placed into the disc, not surrounding the disc, preferably using a percutaneus approach. One o f the possibilities is to injected the cross linking solution in to the disc by means of a catheter-pump system and a small diameter needle. Therefore, a concern may be the effectiveness of the injection method in cross-linking the collagenous matrix, and h o w it com pares w ith the soak m ethod. M ore im portandy, it is unknown if the coverage o f the reagent using this method will be evenly 124 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. distributed. Theoretically, the soak method may be superior to the injection method, which has been proven through rigorous testing using calf IVD specimens. However, it is not practical to continue the studies using the soak method, since at some point it will be required to show equivalency between that method and the injection through quantification o f cross links and mechanical testing. Since cadaver specimens are difficult to obtain, it may be wise to reevaluate the using of soaking method for treatment prior to additional human cadaver studies. Summary Ultimately if augmenting cross linking is to be a viable treatment for slowing the progression of disc degradation through strengthening of the collagen matrix in the annulus, efficacy must be proven both in vitro as well as in vivo. More importantly, proving its effectiveness in vitro on human cadaver disc is desirable. Therefore, a case can be made for further refinement of non-destructive and destructive test procedures to adapt to human cadaver disc specimens. Implementation of the refinements outlined in this and prior chapters will result in protocols that may get us close to that goal. 125 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Bibliography Adams MA, Hutton WC. Prolapsed intervertebral disc: a hyperflexion injury. Spine. 7:184-191,1982. Adams MA, Hutton WC. The effect of fatigue on the lumbar intervertebral disc. The Journal of Tone and Joint Surgery. 65-B(2): 199-203, 1983. Adams MA, Dolan P, Hutton WC. The stages of disc degeneration as revealed by discograms. The Journal of Bone and Joint Surgery. 68-B(l):36-41, 1986. Adams MA, Green TP, Dolan P. The strength in anterior bending of lumbar intervertebral discs. Spine. 19(19):2197-2203, 1994. Chan SS, Livesay GA, Woo SL-Y. A new system to accurately determine the cross- sectional shape and area o f soft tissues. Paper presented at the 42n d Annual Meeting of Orthopaedic Research Society, Atlanta, Georgia, 1996. Clemente CD. Gray’s anatomy, 13th ed. Philadelphia: Lea and Febiger, 1985, 197- 230. Deyo RA, Weinstein JN. Low back pain. New England Journal o f Medicine. 344(5) :363- 369, 2001. Ebara S, Iatridis JC, Setton LA, Foster RJ, Mow VC, Weidenbaum M. Tensile properties of nondegenerate human lumbar annulus fibrosus. Spine. 21(4):452-461, 1996. Friberg S, Hirch C. Anatomical and clinical studies on lumbar disc degeneration. Acta Orthopaedica Scandinavica Supplementum. 19:222-242, 1950. Frymoyer JW, Pope MH, Clements JH, Wilder D G, MacPherson B, Ashikaga T. Risk factors in low-back pain. The Journal of Bone and Joint Surgery. 65A(2):213-218, 1983. 126 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Galante JO. Tensile properties of the human lumbar annulus fibrosus. Acta Ortbopaedica Scandinavica Supplementum. 100(1):1 — 91, 1967. Gray D. Conversation with author. 2001. Gray, D. A preliminary investigation to determine the effects o f a crosslinikng reagent on the fatigue resistance of the posterior annulus of the intervertebral disc. University o f Southern California, 2002. Hedman TP, Femie GR. Mechanical response of the lumbar spine to seated postural loads. Spine. 22(7):734-743, 1997 Hedman TP. Whitaker Foundation grant application: Modification o f intervertebral disc. Whitaker Foundation. 2001. Hedman TP. Conversation with author. 2002. Hilton RC, Ball J. Vertebral rim lesions in the dorsolumbar spine. Annals of the Rheumatic Diseases. 43:302-307,1984. Hoshaw SJ, Cody D D, Saad AM, Fyhrie DP. Decrease in canine proximal femoral ultimate strength and stiffness due to fatigue damage. Jounal of Biomechanics. 30(4):323-329, 1997. Hua M. Conversation with author. 2002. Johnson S, Jones MG, Halliwell M, McNally D. Swelling o f the intervertebral disc: Observations of external and internal morphological change. Paper presented at the 47th Annual Meeting of Orthopaedic Research Society, San Francisco, California, 2001. Kelsey JL. An epidemiological study of acute herniated lumbar discs. Rheumatology and Rehabilitation. 14:144-159, 1975. Lee TQ, Woo SL-Y. A new method for determining cross-sectional shape and area of soft tissues. Transactions of theA SM E . 110(5):110-114,1988. 127 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Lieber RL. Statistical significance and statistical power in hypothesis testing. Journal of Orthopaedic Research. 8:304-309,1990. Lindh, M. Biomechanics of the lumbar spine. In Basic Biomechanics of the Musculoskeletal System 2n d ed.., ed. Nordin M, Victor FH. Philadelphia: Lippincott Williams and Wilkins, 1989, 183-207. Magora A. Investigation of the relation between low back pain and occupation: sitting, standing and weight lifting. Industrial Medicine. 41:5-9,1972. Markolf KL, Morris JM. The structural components o f the intervertebral disc. The Journal of Bone and Joint Surgery. 56-A (4):675-687, 1974. McMillan DW, Garbutt G, Adams MA. Effect o f sustained loading on the water content o f intervertebral discs: implications for disc metabolism. Annals of the Rheumatic Diseases. 55(12):880-887,1996. Nachemson A. Lumbar intradiscal pressure: Experimental studies on post-mortem material. Acta Orthopaedica Scandinavica Supplementum.. 43,1960. Nachemson A. The influence of spinal movements on the lumbar intradiscal pressure and on the tensile stress in the annulus fibrosus. Acta Orthopaedica Scandinavica Supplementum.. 33,1963. Nachemson A. The load on lumbar disc in different positions of the body. Clin. Orthopaedica. 45:107, 1966. Osti OL, Vemon-Roberts B, Fraser RD. Annular tears and intervertebral disc degeneration: An experimental study using an animal model. Spine. 15(8): 762-767, 1990. Osti OL, Vemon-Roberts B, Moore R, Fraser RD. Annular tears and disc degeneration in the lumbar spine. A post-mortem study o f 135 discs. The Journal of Bone and Joint Surgery. 74-B:678-682, 1992. Panjabi MM, Krag MG, Chung TQ. Effects of disc injury on mechanical behavior of the human spine. Spine. 9(7): 707-713, 1984. 128 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Pflaster DS, Krag MH, Johnson CC, Haugh LD, Pope MH. Effect of test environment on intervertebral disc hydration. Spine. 22(2): 133-139, 1997. Pope MH, Wilder D G , Goel VK.. Biomechanics of the lumbar spine. In The A dult Spine: Principles and Practice, 2 “ l ed., ed. Frymoyer JW. Philadelphia: Lippincott-Raven, 1997. Rolander SD. Motion o f the lumbar spine with special reference to the stabilizing effect of posterior fusion: an experimental study on autopsy specimens. Acta Orthopaedica Scandinavica Supplementum.. 90(1):64— 67,1966. Shackelford JF. Introduction to material science for engineers, 2n d ed. New York: Macmillan, 1988,157-179. Strong AB. 1996. Plastics: Materials and processing, 2n d ed. New Jersy: Prentice Hall, 1996, 38-59. Thompson JP, Pearce RH, Schechter MT, Adams ME, Tsang MB, Bishop PB. Preliminary evaluation o f a scheme for grading the gross morphology of the human intervertebral disc. Spine. 15(5):411-415, 1990. Urban JPG, Maroudas A. Swelling of the intervertebral disc in vitro. Connective Tissue Research. 9(1):1-10,1981. Videman T, Nurminen M, Troup JDG . Lumbar spinal pathology in cadaveric material in relation to history o f back pain, occupation, and physical loading. Spine. 15(8):728-738,1990. Woo Sl-Y, Danto MI, Ohland KJ. The use of a laser micrometer system to determine the cross-sectional shape and area o f ligaments: A comparative study with two existing methods. Transactions of the A S M E . 112(11): 426-431,1990. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Appendix 80 -40 J ------------- -------------------- -------- Group Type ■ Hardness Control ■ Hardness Treated j Appendix A-l: Pre-fatigue hardness of human cadaver discs. Hardness measurement using multiple point indentation method (n=2). Recording Device (to computer) Rotating Motor with Controler Appendix A-2: Drawing of laser reflectance system. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 0.45 CM Saline Genipin ■ Laterals ■Medials Appendix A-3: Change in modulus of resiliency due to location. Saline Genipin ■ Laterals ■ Medials j Appendix A-4: Change in yield point due to location. 131 R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. 5 Saline Genipin ■ Laterals ■ Medials Appendix A-5: Change in ultimate strength due to location. Laterals Medials I Saline ■ Genipin Appendix A-6: Change in modulus of resiliency due to treatment. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission. Laterals Location M edials ■ Saline ■Genipin Appendix A-7: Change in yield point due to treatment. Laterals Medials (■Saline ■ Genipin Appendix A-8: Change in ultimate strength due to treatment. R eproduced with perm ission of the copyright owner. Further reproduction prohibited without perm ission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A preliminary investigation to determine the effects of a crosslinking reagent on the fatigue resistance of the posterior annulus of the intervertebral disc

PDF

Biological materials investigation by atomic force microscope (AFM)

PDF

A fluorescence microscopy study of quantum dots as fluorescent probes for brain tumor diagnosis

PDF

Computation and validation of circulating blood volume with the indocyanine green dilution technique

PDF

Development of ceramic-to-metal package for BION microstimulator

PDF

Cellular kinetic models of the antiviral agent (R)-9-(2-phosphonylmethoxypropyl)adenine (PMPA)

PDF

Effects of prenatal cocaine exposure in quantitative sleep measures in infants

PDF

Comparing signal processing methods for spectral bio-imaging

PDF

A user interface for the ADAPT II pharmacokinetic/pharmacodynamic systems analysis software under Windows 2000

PDF

Dynamics of the newly formed neuromuscular synapse

PDF

Characteristic acoustics of transmyocardial laser revascularization

PDF

Evaluation of R.F. transmitters for optimized operation of muscle stimulating implants

PDF

Comparisons of deconvolution algorithms in pharmacokinetic analysis

PDF

Assessment of minimal model applicability to longitudinal studies

PDF

A model of upper airway dynamics in obstructive sleep apnea syndrome

PDF

Comparison of evacuation and compression for cough assist

PDF

Propofol Effects On Eeg And Levels Of Sedation

PDF

Bayesian estimation using Markov chain Monte Carlo methods in pharmacokinetic system analysis

PDF

Head injury biomechanics: Quantification of head injury measures in rear-end motor vehicle collisions

PDF

Cardiopulmonary coupling in the control of breathing during exercise

Asset Metadata

Creator Syed, Baber R. (author)

Core Title Destructive and non-destructive approaches for quantifying the effects of a collagen cross-linking reagent on the fatigue resistance of human intervertebral disc

School School of Engineering

Degree Master of Science

Degree Program Biomedical Engineering

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

Tag engineering, biomedical,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Hedman, Tom (committee chair), D'Argenio, David (committee member), Yamashiro, Stanley (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-323673

Unique identifier UC11337101

Identifier 1424233.pdf (filename),usctheses-c16-323673 (legacy record id)

Legacy Identifier 1424233.pdf

Dmrecord 323673

Document Type Thesis

Rights Syed, Baber R.

Type texts

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

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA