Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Spinal cord tissue changes due to residual compression in a novel rat contusion model Sjovold, Simon Gerhard 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-0307.pdf [ 15.23MB ]
Metadata
JSON: 831-1.0080765.json
JSON-LD: 831-1.0080765-ld.json
RDF/XML (Pretty): 831-1.0080765-rdf.xml
RDF/JSON: 831-1.0080765-rdf.json
Turtle: 831-1.0080765-turtle.txt
N-Triples: 831-1.0080765-rdf-ntriples.txt
Original Record: 831-1.0080765-source.json
Full Text
831-1.0080765-fulltext.txt
Citation
831-1.0080765.ris

Full Text

SPINAL CORD TISSUE C H A N G E S D U E C O M P R E S S I O N IN  A  NOVEL  RAT  TO  RESIDUAL  CONTUSION  MODEL  by SIMON GERHARD SJOVOLD B.A.Sc, The University of British Columbia, 2001  A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF MASTER OF APPLIED SCIENCE  in THE FACULTY OF GRADUATE STUDIES  (MECHANICAL ENGINEERING)  THE UNIVERSITY OF BRITISH COLUMBIA April 2006 © Simon Gerhard Sjovold, 2006  Abstract Residual compression can potentially lead to the exacerbation of the initial spinal cord trauma and is currently corrected by performing surgical decompression.  Decompression is one of  the few treatment options currently available to clinicians, but until direct beneficial effects are proven, its use will remain controversial. Many investigators have developed experimental models to evaluate the effects of maintained compression; however replication of the clinical injury incorporating a primary contusion followed by residual compression of the spinal cord has not been a primary focus. The objectives of this thesis were to i) establish injury protocols of graded residual compression following a reproducible contusion injury, ii) develop a clamp to rigidly support the thoracic vertebrae, iii) develop a method to monitor the long-term microvascular blood flow of the spinal cord, and iv) determine the influence of 40% and 90% residual spinal cord compression on the extent and progression of neurological damage following a moderate contusive spinal cord injury. The initial contusion produced considerable disruption of the localized tissue, preventing it from supporting further compression. As a result, the load relaxation of the spinal cord was rapid for both levels of residual compression. Contusion did not adversely affect the microvascular blood flow, but residual compression significantly increased the blood flow one level caudal to the'injury epicentre. Total haemorrhage volume was similar for contusion and both levels of residual compression although the extent was greater following 90% residual compression. High levels (90%) of residual compression resulted in an extension of the gray and white matter damage beyond that of the initial contusion injury. Low levels (40%) of residual compression did not appear to increase the cellular damage in the white and gray matter, at least in the acute stage (initial 3 hours). Although this thesis did not evaluate the blood flow in all regions of the spinal cord and only compared two levels of residual compression for a contusion injury of a single magnitude, we have produced compelling evidence that residual compression does not entirely restrict the blood  ii  Abstract  flow through the cord, distant to the injury. As well, the relative magnitude of the residual compression, in relation to the initial contusion injury, is an important factor in determining the resulting level of neurological injury.  iii  Table of Contents Abstract  ii  Table of Contents  iv  List of Tables  vii  List of Figures  viii xi  Acknowledgements  xn  Dedication Chapter 1. Introduction 1.1 Clinical Motivation 1.2 Epidemiology 1.3 The Spinal Cord - Anatomical Perspective 1.3.1 Osseous Structure 1.3.2 Soft Tissue Structure & Organization 1.3.3 Vascular Structure 1.3.4 Cellular Structure 1.3.5 Biomechanical Properties 1.4 Clinical Spinal Cord Injury 1.4.1 Parameters of Injury 1.4.2 Surgical Decompression 1.5 Experimental Models of Spinal Cord Injury 1.5.1 Maintained Compression 1.5.2 Dynamic Contusion 1.5.3 Residual Decompression 1.6 Blood Flow Measurement 1.6.1 Laser Doppler Flowmetry 1.6.2 Normal Spinal Cord Blood Flow 1.7 Pathology of Spinal Cord Injury 1.7.1 Progression of Injury 1.7.2 Vascular Injury 1.8 Summary 1.9 Objectives k Scope 1.10 References  1 2 3 6 7 8 10 13 15 16 16 18 20 20 21 22 23 24 25 26 26 27 28 29 32  Chapter 2. M o d e l of Residual Compression 2.1 Introduction 2.2 Methods 2.2.1 Surgical Preparation & Fixation 2.2.2 Injury Procedure & Assessment  43 43 43 44 44  iv  Table of Contents  2.3  2.4  2.5  2.2.3 Statistical Analysis Results 2.3.1 Mechanical Parameters 2.3.2 Laser Doppler & Pulse Oximeter Discussion 2.4.1 Injury Mechanism 2.4.2 Mechanical Parameters of Injury 2.4.3 Load Relaxation 2.4.4 Blood Flow References  47 48 48 51 55 57 58 60 61 67  Chapter 3. Histopathological Changes Following Residual Compression 3.1 Introduction 3.2 Methods 3.2.1 Injury Procedure 3.2.2 Tissue Harvesting & Preparation 3.2.3 Histology & Immunohistochemistry 3.2.4 Image Analysis & Quantification 3.2.5 Variation in Techniques 3.2.6 Statistical Analysis 3.3 Results 3.3.1 Haemorrhage 3.3.2 Gray Matter Injury 3.3.3 White Matter Injury 3.4 Discussion 3.4.1 Haemorrhage 3.4.2 Gray Matter Damage 3.4.3 White Matter Damage 3.5 References  72 72 72 73 73 74 75 78 80 80 80 83 86 94 95 96 99 102  Chapter 4. General Discussion and Conclusions 4.1 General Limitations 4.1.1 Anaesthesia 4.1.2 Animals as Models of Human Injury 4.1.3 Rigid Support of Vertebral Column 4.1.4 Bias & Variability in Analysis Methods 4.2 Summary & Conclusions 4.3 Future Directions 4.4 References  105 105 106 107 108 110 110 114 116  Appendix A . Specimens & Experimental Apparatus  121  A.l A.2 A.3 A.4 A.5  Specimens Contusion Apparatus Temperature Measurements and Control Vitals Support Frame  v  121 121 123 124 124  Table of Contents  A.6 A. 7  Laser Doppler Flowmetry System Artificial Cerebral Spinal Fluid  126 127  Appendix B. Injury Protocols B. l PID Settings B.2 Pure Contusion B.3 Contusion & 40% Residual Compression B. 4 Contusion & 90% Residual Compression  128 128 129 130 131  Appendix C. Statistical Results Cl Mechanical Parameters C. 2 H & E C.3 NeuN & F J C.4 /3-APP  132 132 135 135 137  vi  List of Tables 1.1 1.2 1.3  Functional goals associated with specific levels of complete cervical injury Etiology, Vertebral Level, Type of Bony Injury and Age of SCI patients Average yearly health care and living expenses and the estimated lifetime costs that are directly attributable to SCI  3 4 5  2.1  Injury parameters of the primary contusion injury for each group  4.1  Dosage of urethane anaesthesia used in this study was not statistically different between groups (p=0.6)  107  B.l B.2 B.3 B.4 B.5 B.6 B. 7  PID settings for linear electromagnetic actuator Waveform settings for pure contusion injury Data acquisition settings for pure contusion injury Waveform settings for contusion & 40% residual compression injury Data acquisition settings for contusion Sz 40% residual compression injury Waveform settings for contusion & 90% residual compression injury Data acquisition settings for contusion & 90% residual compression injury  128 129 129 130 130 131 131  Cl C. 2 C.3 C.4 C.5 C.6 C.7 C.8 C.9 C.10 C.ll C.12 C.13 C.14 C.15  Univariate Tests of Significance - Preload Force Univariate Tests of Significance - Contusion Injury Parameters Newman-Keuls Multiple Comparison Test - Peak Displacement Kruskal-Wallis ANOVA by Ranks - Load Relaxation Multiple Comparisons p values (2-tailed) - Load Relaxation Repeated Measures Analysis of Variance - H&E Newman-Keuls Multiple Comparison Test - H&E Injury Groups Repeated Measures Analysis of Variance - NeuN Newman-Keuls Multiple Comparison Test - NeuN Injury Groups Repeated Measures Analysis of Variance - Fluoro-Jade Newman-Keuls Multiple Comparison Test - Fluoro-Jade Injury Groups Repeated Measures Analysis of Variance - 8-APP Newman-Keuls Multiple Comparison Test - 3-APP Injury Group Newman-Keuls Multiple Comparison Test - 3-APP Region Newman-Keuls Multiple Comparison Test - /3-APP Extent from Epicentre  132 132 132 133 134 135 135 135 136 136 136 137 137 137 138  vii  50  List of Figures 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 2.1 2.2 2.3 2.4  2.5 2.6 2.6 2.6 2.7 2.8 2.9 2.10 3.1 3.2 3.3 3.4  Typical cervical, thoracic and lumbar vertebrae illustrating key anatomical componenets Cross-section of the spinal cord, showing bilateral symmetry of gray and white matter across ventral median fissure and dorsal median sulcus Cross-section of spinal cord, showing axonal tracts of the white matter Blood supply and major arteries of the spinal cord Comparison of arterial supply and venous drainage of spinal cord Primary components of a neuron Reduction of the backscatter spectral power distribution following the accumulation of blood below the laser doppler (LD) probe Progression of SCI from initial focal primary injury to encompass surrounding tissues through secondary mechanisms of degeneration The vertebral body clamp secures the vertebrae on the edge of the pedicle just between the lateral process and the posterior surfaces of the rib Experimental set-up for the creation of SCIs Typical displacement, force and velocity time curves for each injury group Characterization of the load relaxation for 40% and 90% residual compression by plotting median times to reach 75%, 50%, 40%, 30%, 20%, 10% and 5% of the peak force Laser Doppler (LD) microperfusion measured from the dorsolateral surface of the exposed spinal cord Comparison of laser doppler microperfusion at 2, 30, 58, 62 and 90 minutes post-contusion Comparison of laser doppler microperfusion at 2, 30, 58, 62 and 90 minutes post-contusion - continued Comparison of laser doppler microperfusion at 2, 30, 58, 62 and 90 minutes post-contusion - continued Median heart rate and haemoglobin saturation values measured from the base of the tail with a pulse oximeter Comparison of force and displacement from the current study to the results of other studies employing a contusion injury of the rat thoracic spinal cord Frame stiffness is 29.8 N/mm, corresponding to flexion of less than 6% of the peak contusion displacement Tests of injury protocols using surrogate silicone tubing Methods for selecting the maximum rostral and caudal extent of gray matter damage for NeuN and F J labelled spinal cords Analysis technique for quantifying the density of 3-APP in regions of the white matter Comparison of haemorrhage near the injury epicentre (midsagittal plane) of the different injury groups Total volume of haemorrhage for each injury group  viii  8 9 10 11 13 14 25 26 45 46 49  50 51 52 53 54 55 59 61 62 77 78 81 82  List of Figures  3.5 3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 3.15 3.16 3.17 3.18 3.19  Rostral and caudal extent of haemorrhage for pure contusion, 40% and 90% residual compression Extent of gray matter damage within the medial gray matter by assessing the diminished immunofluorescence of NeuN and presence of F J positive neurons... Extent of gray matter damage within the ventral gray matter by assessing the diminished immunofluorescence of NeuN and presence of F J positive neurons... Extent of gray matter damage within the dorsal gray matter by assessing the diminished immunofluorescence of NeuN and presence of F J positive neurons... Variation in NeuN rostral and caudal extent of gray matter damage at the midsagittal, ventrolateral and dorsolateral plane Variation in F J rostral and caudal extent of gray matter damage at the midsagittal, ventrolateral and dorsolateral plane Mean rostral and caudal extent of /3-APP density in dorsal white matter tracts (error bars not plotted) Mean rostral and caudal extent of /5-APP density in ventral white matter tracts (error bars not plotted) Mean rostral and caudal extent of B-APP density in ventrolateral white matter tracts (error bars not plotted) Mean rostral and caudal extent of /3-APP density in lateral white matter tracts (error bars not plotted) Main effect comparing the overall mean densities of /?-APP for the different injury groups Main effect comparing the distribution of /3-APP The distribution of 3-APP for pure contusion, 40% and 90% residual compression. The distribution of /3-APP in the dorsal, ventral, ventrolateral k, lateral white matter tracts Three way interaction between the rostral-caudal distribution of /3-APP for each injury in the dorsal, ventral, ventrolateral & lateral white matter tracts. ..  82 83 84 84 85 85 86 87 87 88 89 90 91 92 93  4.1  Comparison of the major axonal tracts of the white matter of the (a) human [32] and (b) rat [47] 108  A.l  Linear electromagnetic actuator and support frame used for the creation of the spinal cord injuries Impact tip fabricated from a 2 mm stainless-steel dowel pin and aluminum housing that allowed the overall length to be adjusted Heating pad and rectal probe for maintenance of specimen body temperature at 37°C Heating coils used to maintain the temperature of the CSF fluid at 37°C Support structures for rigid fixation of specimen during tests The vertebral body clamp secures the vertebrae on the edge of the pedicle just between the lateral process and the posterior surfaces of the rib Primary components of clamp: support bar, rostral/caudal swing-arms and rostral/caudal teeth Distance between teeth can be finely adjusted by changing angular orientation of swing-arms Installation tool for placing laser doppler probe in probe holder  A.2 A.3 A.4 A.5 A.6 A.7 A.8 A.9  ix  122 122 123 124 125 126 126 126 127  List of Figures  A. 10 Probe holder and vertical alignment device for positioning laser doppler probe.. 127 A. 11 IV burette used for control of CSF flow rate 127  x  Acknowledgements This thesis is dedicated with much love to my beautiful wife Leah and funny man Caleb. Their support and patience over the past few years has been appreciated beyond words and I hope this document is worthy of the many nights and weekends I spent without them. The drawings they both contributed, especially Caleb's figure depicting the primary components of the neuron are so much better than anything I produced. Our parents, Rasmus and Joan Sjovold, Randy and Shelley Malone and Brett Chapman have provided unconditional love and support, kept our refrigerator and freezer stocked with food and helped in a million other ways. Sarah and Stephanie (plus family) have been the sisters and sister-in-laws we needed, providing an ear and encouragement when needed. This journey would have been much harder without you all, thanks from the bottom of my heart. I would like to thank my committee for their input into the development of this thesis. A sincere thanks to Dr. Oxland, "The Boss", for letting me join and manage his laboratory five long years ago and picking me to be one of his graduate students a little over two years ago. These opportunities have provided me with the environment and guidance to grow and expand my talents and introduced me to the world of biomechanics. I would also like to thank Dr. Tetzlaff, "Wolf", for welcoming another engineer into his pack and providing such an encouraging and creative environment for research. Thanks to Jie for all your surgical expertise and Clarrie for leading my way through the world of biological sciences. Hopefully I did not ask too many questions. Thank you to Anthony for developing the test system that made my work possible and to Liz (Down Under) for coming to UBC and making sure I was not the only engineer that felt like a lab rat. To all my many friends and colleagues for being you. Dave L., Dave W., Chris and Polly, Victor and family, Peter, Hanspeter, Qingan, Juay-Seng, Chad, Derek, Amy, Tim, JD, Saija, Carolyne, Teresa, Sarah, Emily, Mr. P and everyone else that obviously did not fit on the page.  xi  Dedication  To my beautiful wife Leah and wonderful son Caleb.  xii  Chapter 1  Introduction  Considerable research has been invested in an absolute cure for spinal cord injury (SCI), but the fundamental understanding of how SCI progresses and what components of the primary injury influence the final outcome are still largely unknown. This thesis was proposed for the development of a model of SCI that effectively modelled both the initial spinal cord contusion and subsequent residual compression. Residual compression can potentially lead to the exacerbation of the injury following the initial trauma and is currently treatable within the clinical setting. However, conclusive evidence of the effects of residual compression on human SCI trauma are still relatively unknown. Detrimental effects of (residual) compression have been demonstrated using a variety of experimental models that employ persistent compression of the spinal cord, but very few of these studies modelled the initial rapid contusion of clinical SCI that creates significant trauma to the cord. It is well established that contusion causes immediate disruption to the local tissue but it is not known what incremental effects residual compression may have. The remainder of the chapter will discuss relevant literature that is pertinent to the understanding of residual compression injuries, followed by the objectives of the research that were undertaken and the overall scope of the work. Chapter 2 outlines the experimental model of contusion and residual compression that was developed for this research, with a presentation of the mechanical parameters and blood flow measurements that were obtained. Discussion of the results with relation to the pre-existing literature is also included. 1  Chapter 1.  Introduction  Chapter 3 is a continuation of the experiments conducted in Chapter 2, with emphasis on the pathohistological findings. Details of the histological and immunohistological results are presented with a discussion of their importance to the understanding of residual compression. Chapter 4 reviews the limitations of the current study as well as a general discussion of the overall project. Final conclusions will be presented as well as recommendations for future work.  1.1  Clinical Motivation  In the past century, considerable advancements have been made within the medical field, resulting in cures for previously fatal inflictions and promising outcomes for many types of injuries. Improved surgical techniques and rehabilitation practices for SCI have resulted in fewer fatalities and a reduction in the average hospital stay [7, 58, 88, 95, 98, 107, 114]. However, relative to triumphs achieved in other areas of modern medicine, the treatment of SCI remains relatively unsolved, resulting in lifelong disabilities and many secondary complications for the majority of patients [7, 98]. SCI as the result of burst fractures and fracture-dislocation are frequently accompanied by persistent residual compression of the cord from fragments of the surrounding tissue [44]. The initial contusion is a dynamic event that produces a significant high force injury, while residual compression is a static event during which the tissue remains compressed. It has been established that contusion injuries result in significant damage to the spinal cord [6, 11, 17, 126, 152], however the influence of residual compression remains controversial [7, 28, 48, 49, 141, 145]. Despite all the research and progress in clinical management, there is currently no cure for SCI. The emergence of a critical breakthrough, as seen in many other areas of medical care, is yet to be found and potential treatments such as methylprednisolone and surgical decompression are plagued by controversy. Still it is important to determine whether surgical techniques, such as decompression, can improve the overall neurological outcome. Animal studies have shown that significant function below the level of the spinal cord lesion can be realized when as little as 5-10% of the axonal tracts remain intact [15]. This is encouraging as even minor increases  2  Chapter 1.  Introduction  in the lowest level of neurological function can have profound effects on the livelihood of SCI patients and decrease their dependence upon others [9, 141] (Table 1.1). Table 1.1: Functional goals associated with specific levels of complete cervical injury. Figure modified from Gray [56].  Functional Goals  Level  Abilities  C l to C3  Head & Neck  Ventilator dependent Talking difficult/impossible  C4  Diaphragm  C5  Deltoids Biceps  C6  Wrist extensors  C7  Triceps  C8  1 lands  Talk normally Use adaptive controls Powered wheelchair Manage own health care Driving possible Perform daily task Manual wheelchair use Require few adaptive aids Daily manual wheelchair use Live independently without adaptive aids  1.2 Epidemiology The epidemiology of SCI varies considerably throughout the world, yet similar trends can be found within countries of comparable economic status. The mean age of SCI patients tends to be higher in developed countries, possibly due to the increased life-expectancy of those populations [3]. Higher male-to-female ratios are often found in developing countries, a finding that is thought to be the result of a larger manual labour force (the majority of which are male) and risk taking among young males [3]. The cause, level, type of bony injury and age of SCI patients are summarized in Table 1,2 The incident rate of SCI is estimated to be between 14.5 to 57.8 cases per million people annually [3, 98, 120]. Canada (52.5) and Portugal (57.8) have the highest rates worldwide and are thought to more accurately predict the actual rate of SCI, as they are the only countries to include individuals that died before admission to hospital [3]. The US National Spinal Cord  3  Chapter 1.  Introduction  Table 1.2: Etiology, Vertebral Level, Type of Bony Injury and Age of SCI patients. Compiled from the records of Toronto General and Sunnybrook Hospitals from 1948-1973. Adapted from Tator [132, 134]  Cause of Injury  Traffic accidents Work Sports and recreation Falls Violence  Level of Injury  Cervical (CI to C7-T1) Thoracic (Tl-11) Thoracolumbar (Tll-12 to L l - 5 )  Type of Bony Injury  Minor fracture Fracture dislocation Dislocation only Burst fracture Other  Age of SCI Patients  Birth - 10 11 - 20 21 - 30 31 - 40 40 - 60+  40% - 50% 10% - 25% 10% - 25% 20% 10% - 25% 56.5% 19.7% 23.8% 10% 40% 5% 30% 15% 10% 20% 25% 15% 10% per decade  Injury Statistical Center reports that 11,000 new SCIs occur in the US each year, and there are approximately 250,000 citizens currently living with some level of SCI [98]. Most SCIs occur as a result of motor vehicle accidents, falls, violence or sports related injuries [3, 98, 120, 134]. Worldwide, motor vehicle accidents account for the majority of injuries, whereas falls in the elderly account for most injuries in developed countries, and violence is the leading cause of injuries in lower-middle income countries such as Brazil and South Africa [3, 134]. In the US, the prevalence of SCI resulting from falls (22.9%) is steadily increasing, whereas injuries resulting from violence have dropped from a peak of 24.8% in 1999 to 13.8% in 2000 [98]. SCI is primarily an affliction of young adults, with the majority of injuries in the US between 1973 and 1979 occurring between the ages of 16 and 30 (mean age of 28.7 years) and since 2000 the average age at injury is 37.6 years [98, 134]. Worldwide, similar rates are found in almost all countries (mean age between 20 & 40 years); however, as the mean age of developed countries  Chapter 1.  Introduction  has increased, so too has the mean age of SCI patients [3]. In the US the percentage of SCI patients over 60 years of age has increased from 4.7% in 1980 to 10.9% in 2000 [98]. The ratio of males-to-females within the US has conversely decreased from 81.8% in 1980 to 79.6% in 2000 [98]. Other developed countries have a similar ratio of 3 to 1, whereas the male-to-female ratio is reported to be greater than 7.5 to 1 for countries such as Zimbabwe and Bangladesh, where manual labour and civil unrest is more prevalent [3]. The economic and social costs of SCI are staggering. SCI is the second most expensive condition to treat in the US after respiratory syndromes in children [3] and requires over $4 billion annually [120]. Annual and lifetime costs vary substantially based upon education, pre-injury employment history and neurological level of injury (Table 1.3). As well as having enormous financial costs, SCIs result in many social burdens. Due to their young age, 51.5% of SCI patients are single when injured and have a lower chance of becoming married or maintaining an existing marriage, compared to the general populace [98]. Prior to injury 64.1% of SCI patients are employed, whereas only 32.8% of paraplegics and 24.7% of tetraplegics are employed 10 years after injury [98]. The majority of SCIs occur from rapid cord compression, resulting from osseous injuries such as fracture-dislocations (40%) or burst fractures (30%) that often result in subsequent residual  Table 1.3: Average yearly health care and living expenses and the estimated lifetime costs that are directly attributable to SCI. Values vary substantially based upon education, severity of injury and pre-injury employment history. Adapted from the National Spinal Cord Injury Statistical Center [98]. Severity of Injury  Average Yearly Expenses  Estimated Lifetime Costs by Age at Injury  First Year  Subsequent Years  25 years old  50 years old  High Tetraplegia (C1-C4)  $710,275  $127,227  $2,801,642  $1,649,342  Low Tetraplegia (C5-C8)  $458,666  $52,114  $1,584,132  $1,003,192  Paraplegia  $259,531  $26,410  $936,088  $638,472  Incomplete (Any Level)  $209,324  $14,670  $624,441  $452,554  5  Chapter 1.  Introduction  compression of the spinal cord [120]. Anterior dislocation injuries that result from extremely high energy insults to the spinal column are more likely to result in complete neurological disruption of the spinal cord, compared to posterior dislocations or burst fractures [134]. Injuries at the level of the cervical vertebrae account for more than half (41-76%) [3, 120, 132, 134] of all injuries and are approximately three times more frequent than injuries in the thoracic or thoracolumbar regions (Table 1.2). Although cervical injuries account for the majority of SCIs, only 40% of chronic cases have cervical lesions due to the high mortality rates [120]. The ratio of complete to incomplete spinal cord lesions has decreased from almost 66% in 1960 to approximately 45% in recent years [134]. Although variations in the epidemiology of SCI can be found throughout the world, many similarities exist for the majority of cases. SCIs primarily result from very high energy events such as vehicle accidents, falls and violence producing significant trauma. These injuries often occur in young individudals at the level of the cervical spine by means of burst fractures or fracture dislocations with persisting residual compression of the spinal cord, culminating in social encumbrances for the patient and large economic losses over the patient's lifetime. However, the potential for significant improvements in the neurological function at the cervical level is promising, as even minor increases in the lowest level of neurological function can have profound effects on the livelihood of SCI patients and decrease their dependence upon others.  1.3  The Spinal Cord - Anatomical Perspective  The central nervous system, consisting of the brain, spinal cord and retinae, forms the control centre of the body. The brain processes information received from the sensory nerves of the body and allows us to interact with our environment by sending motor impulses to our muscles. The spinal cord forms the pathway through which these signals are transmitted to and from the various levels of the body. The localized organization of the brain and spinal cord allows the body to regulate the environment of these specialized cells, thereby ensuring that the high metabolic requirements of the nervous tissue are met [77]. As beneficial as this organization is for meeting the functional requirements of the cells, it can result in devastating consequences if the  6  Chapter 1.  Introduction  spinal cord is traumatically injured. Insult to the spinal cord not only disrupts the transmission of nervous signals at the level of injury, but also often results in the loss of conduction caudal to the injury site. For brevity the following subsections will focus on the anatomy as seen in the human body. Information comparing the similarities and differences between humans and specifically the rat can be found in Chapter 4.1.2. 1.3.1  Osseous S t r u c t u r e  The spine can be considered one of the most important and complex anatomical structures of the human body. The series of vertebrae that extend from the cranium to the coccyx provide a strong, structural framework to support the weight of the body, while providing flexibility to perform the many complex motions of daily activity. The spinal column also provides structural protection for the spinal cord and nerve roots which transfer the motor and sensory impulses allowing us to interact with our local environment. The vertebral column consists of 33 vertebrae divided into 5 regions: 7 cervical, 12 thoracic, 5 lumbar, 5 sacral and 4 coccygeal, of which the sacrum and coccyx are normally fused within the adult [96]. The vertebrae vary in size (increasing interiorly) and functional characteristics with level, but typically consist of a vertebral body, vertebral arch and multiple processes (Figure 1.1). The vertebral arch consists of a pair of pedicles and thin plate-like laminae, which form the walls of the vertebral foramen along with the posterior surface of the vertebral body. The successive stacking of vertebral foramina forms the spinal canal, which contains the spinal cord, meninges and spinal nerve roots. Intervertebral foramina are created by adjacent vertebral notches located above and below the pedicles. These gaps in the spinal column give passage to the nerve roots and blood vessels of the spine. The first two vertebrae of the cervical spine, C l (atlas) and C2 (axis), are unique in geometry, facilitating the extensive bending and rotational motions of the head. The five lower cervical vertebrae, C3-C7, are characterized by large vertebral foramen to account for the enlarged cervical spinal cord that innervates the upper limbs. The thoracic vertebrae are characterized by the articulation of the ribs with the vertebral bodies and transverse processes (T1-T10)  7  Chapter 1.  Introduction  CERVICAL  THORACIC  LUMBAR  Figure 1.1: Typical cervical, thoracic and lumbar vertebrae illustrating key anatomical componenets: A - vertebral body, B - processes, C - pedicle, D - vertebral foramina, E - lamina, F - vertebral arch. Figures modified from Gray [56]. imparting greater strength and stiffness to this region of the spinal column. The lumbar spine is the lowest portion of the articulating spine, containing massive bodies and thick intervertebral discs.  1.3.2  Soft Tissue S t r u c t u r e & O r g a n i z a t i o n  The spinal cord, located within the spinal canal, is protected by epidural fat, cerebrospinal fluid and three layers of spinal meninges: dura, arachnoid and pia mater. It is a cylindrical structure, with two enlargements in the cervical (C4-T1) and lumbosacral (L2-S3) regions for innervation of the limbs. Similar to the vertebral column, the spinal cord is divided into 31 segments, giving rise to 31 pairs of spinal nerves [125]. The nerves in turn are formed by ventral (sensory) and dorsal (motor) nerve roots that emerge from the dorsolateral and ventrolateral surfaces of the cord. In the adult, the spinal cord does not run the whole length of the vertebral column, but terminates at the medullary cone (T12-L3) [96]. The length of the roots increase caudally down the cord, and each segment is numbered according to the vertebral level at which the corresponding nerves exit the canal.  Chapter 1.  Introduction  The cross-section of the spinal cord differs at each level but is characterized by an inner core of gray matter and surrounding white matter (Figure 1.2). The gray matter consists primarily of nerve cell bodies while the white matter is composed of myelinated axons. The ventral median tissue and dorsal median sulcus divide the cord into two bilaterally symmetric halves [125]. Each half of the gray matter is further divided into the dorsal horn, comprised of cells that receive and process sensory information, the ventral horn, where the motor neurons reside, and an intermediate zone. A small lateral horn is also present in the thoracic and upper lumbar segments. A similar organization exists within the white matter, which is anatomically divided into the dorsal, ventral and lateral funiculi (columns) (Figure 1.2).  Figure 1.2: Cross-section of the spinal cord, showing bilateral symmetry of gray and white matter across ventral median fissure and dorsal median sulcus. Gray matter is divided into an intermediate zone separating the dorsal and ventral horns. White matter is divided into dorsal, ventral and lateral funiculi. Figure based on images from Kiernan and Barr [77].  Collections of axons with similar origins and terminations are termed tracts. The tracts of most clinical significance are the corticospinal, spinothalamic and fasiculus gracilis and cuneatus (Figure 1.3) [77]. The corticospinal tracts are the primary descending pathways for the conduction of voluntary motor signals and are located in the lateral (most important) and medial ventral funiculi. The spinothalamic tract relays the ascending signals of pain, temperature and light touch. It is located along the periphery of the gray and white matter in both the lateral and ventral funiculi. The fasiculus gracilis and fasiculus cuneatus (present above T6) reside in the 9  Chapter 1.  Introduction  dorsal funiculi and contain ascending sensory projections for discriminative touch, pressure and proprioception. Ascending Tracts:  Descending Tracts: Fasiculus Gracilus Lateral Corticospinal  Fasiculus Cuncatus Dorsal Spinocerebellar  Ventral Spinocerebellar  Rubrospinal  Vestibulospinal  Ventral  Spinothalamic  Corticospinal  Cross-section of spinal cord, showing axonal tracts of the white matter. Descending (motor) tracts shown in red on left side. Ascending (sensory) tracts shown in blue on right side. Figure based on images from Kiernan and Barr [77]. Figure  1.3.3  1.3:  Vascular Structure  The spinal cord is richly supplied with blood, providing multiple anastomosing channels to prevent the impairment of circulation to any level. It is interesting to note that the neonatal spinal cord blood supply is greater than at maturity. While the cord volume increases almost thirteen-fold in mature adults, the vascular supply to the cord only increases by one-half [151]. The blood supply of the spinal cord originates from branches of the vertebral arteries and multiple medullary (radicular) arteries derived from segmental arteries of the aorta [151]. Three arterial channels run longitudinally throughout the length of the spinal cord. The anterior spinal artery originates in a y-shaped configuration from the left and right vertebral arteries below the medulla and descends along the anterior median fissure. In the cervical region, the anterior spinal artery is often found to be duplicated but continues as a single vessel along the thoracic and lumbar segments of the cord [37]. Paired posterior spinal arteries, also derived from the vertebral arteries, descend on the posterior surface of the spinal cord, medial to the dorsal root entry zones. There are limited connections between the anterior and posterior spinal arteries 10  Chapter 1.  Introduction  until they anastomose at the medullary cone, but they receive variable contributions throughout their length from medullary arteries [37]. ANTERIOR  POSTERIOR  (a)  (b)  Figure 1.4- Blood supply and major arteries of the spinal cord: (a) arterial supply to anterior and posterior spinal arteries; (b) major arteries of the spinal cord. Figures modified from Drake et al. [41]  Radicular and medullary arteries pass through the intervetebral foraminae and supply blood to the nerve roots and spinal cord. Whereas radicular arteries are present at each level of the spinal column and provide blood only to the spinal nerve roots and meninges, medullary arteries are found at varying levels of the spinal column and anastomose directly with the spinal arteries of the cord [77, 142]. Medullary arteries enter the spinal canal through the intervertebral foramina and divide into anterior and/or posterior branches. Approximately 610 anterior medullary arteries contribute to the anterior spinal artery. The cervical segments receive 0-6 arteries, the thoracic receives 2-4, while the lumbar has only 1 or 2. The artery of Adamkiewicz (lumbar enlargement) is the largest and considered by many to be the most important of the anterior radicular arteries, as it supplies the majority of the blood to the lower two-thirds of the spinal cord. The posterior medullary arteries anastomose with the paired posterior spinal arteries and are more numerous than the anterior arteries, ranging in numbers  11  Chapter 1.  Introduction  from 10-23 but are of smaller calibre. Sulcal arteries arise from the anterior spinal artery and pass posteriorly into the anterior median fissure branching to one side or the other [142]. The anterior two-thirds of the spinal cord derive blood from the sulcal arteries that are more numerous in the cervical and lumbar regions of the spinal cord. Sulcal arteries are also of a larger calibre than those in the thoracic segments [136, 142]. The periphery white matter and posterior one-third of the spinal cord derives its blood from the posterior spinal arteries. Within the gray matter terminal branches of the sulcal arteries form an extensive interlocking capillary network to ensure that the high metabolic demands of the neuronal cell bodies are met [136, 142]. In the cervical region anastomotic pathways are numerous, and the obstruction of an artery to the cervical cord can be bypassed in most cases without neurological damage. However, in the thoracic region interconnections between the medullary, anterior and posterior spinal arteries are significantly less, resulting in the possibility of serious vascular compromise if a vessel becomes occluded [77]. The veins of the spinal cord have a distribution similar to that of the spinal arteries. Radicular veins anastomose between the spinal veins and the internal venous plexus of the spinal canal. There are normally three anterior and three posterior spinal veins. Primary drainage is provided by the anteromedian and posteromedian veins through sulcal veins within the spinal cord. The lateral veins form a plexus communicating freely with each other and the anteromedian and posteromedian vessels as well as providing drainage for the lateral portions of the cord. In the arterial system, the sulcal arteries provide the majority of blood supply to the cord and originate from the anterior spinal artery. However, in the venous system, the sulcal veins drain to both the anterior and posterior vessels (Figure 1.5). The vasculature of the central nervous system is distinct from other regions of the body due to the presence of the blood-brain barrier. The blood-brain barrier is formed by tight junctions between successive capillary endothelial cells, limiting the diffusion of large molecules such as proteins from the blood supply to the interstitial fluid of the central nervous system. The  12  Chapter 1.  Introduction A R T E R I A L SUPPLY  VENOUS DRAINAGE  ANTERIOR  Figure  1.5:  Comparison of arterial supply and venous drainage of spinal cord.  presence of the blood-brain barrier does present significant challenges for the treatment of spinal cord disease and injury, as many potentially beneficial pharmaceuticals are unable to penetrate this barrier. 1.3.4  Cellular Structure  The spinal cord consists of two principal cells, glia and neurons. Glia play vital roles in controlling the environment and providing structural support and protection for nerve cells. There are three types of glial cells, which outnumber neurons by 10 to 1 in the spinal cord: astrocytes, microglia and oligodendrocytes [125]. Neurons act as the functional component of the spinal cord, integrating information and sending signals across considerable distances. The primary components of the neuron are the cell body, dendrites, synapses and axon (Figure 1.6). The cell body is the nutritional power-house, containing many ribosomes and ribonucleic acids for the synthesis of proteins essential for the proper function of the neuron [125]. Each neuron has multiple branching dendrites which conduct the electrical stimulation from other cells to the cell body. Synapses are specialized junctions that allow the transmission of signals between one cell and another by the release of chemical neurotransmitters from the synaptic terminal. Axons are the pathways of the neuron that allow the propagation of the nervous signal over very large distances. Each neuron contains only one axon that is uniform in diameter, unlike the distally tapering dendrites. Axons propagate the nervous signal through ionic shifts called action potentials that generate a voltage change across the membrane of the axon. In large  13  Chapter 1.  Introduction  myelinated axons, such as those found in the corticospinal tract, the action potential jumps down the axon by a process known as saltatory conduction, thereby increasing the conduction speed of the signal.  Figure  1.6:  Primary components of a neuron.  The cytoskeleton of the neuron is composed of microfilaments, neurofilaments and microtubules [125]. Microfilaments are the thinnest component of the cytoskeleton, measuring 4-5 nm in diameter and are prominent in the growth cones of axons and dendrites. Neurofilaments, which are 10 nm in diameter, are the primary structural component of the cell. They are made of three proteins of different molecular weights and function. The neurofilament-L (70kd) forms the central core, and the neurofllament-M (140kd) and neurofilament-H (220kd) form the sidearms that interact between neighbouring neurofilaments and microtubules [71]. Microtubules are hollow cylinders 25 nm in diameter, formed by 13 protofilament subunits that form the perimeter of the tube [14]. Each protofilament is a polymer of alternating a and /3-tubulin proteins. Microtubules play an important role in the growth and maintenance of the axon. They provide the rapid phase of intracellular transport, moving proteins and vesicular elements  14  Chapter 1.  Introduction  along the axon and ensuring continuous replenishment of the cytoskeleton [14]. 1.3.5  Biomechanical Properties  The spinal cord is a soft tissue structure with a central core of gray matter (cell bodies) surrounded by longitudinally aligned white matter tracts (axons). Like all soft tissues, the spinal cord exhibits a rate dependent viscoelastic behaviour [18]. Viscoelasticity is characterized by the following behaviours: creep (constant applied stress results in the continued deformation of the tissue); load relaxation (constant applied strain results in the reduction of the internal stresses over time); and hysteresis (cyclic loading results in different stress-strain behaviours during the loading and unloading processes) [54]. In vitro uniaxial tension tests of fresh cervical spinal cord sections have demonstrated the viscoelastic behaviour of the spinal cord [18]. Stress relaxation experiments loaded the segments at moderate strain rates to a peak strain within the physiological range of normal motion. The characteristic nonlinear relationship between stress and strain was observed in which the stiffness increased with applied strain rate. Considerable load relaxation (20-30%) was observed within the initial 30 seconds of the experiment. A similar stress-strain rate dependence has been shown in axial tension tests of swine brain tissue in vitro [93]. Both of these investigations utilized strain rates that were significantly slower than would be seen during a burst fracture injury, and due to their in vitro nature, physiological effects such as perfusion pressure could not be evaluated. It is believed that differences in the material properties of the gray and white matter may account for the preferential destruction of the gray matter following trauma. Traditionally it was believed that the gray matter damage occurred due to its softer consistency and the presence of greater vasculature. A recent comparison of bovine gray and white matter demonstrated that the gray matter is more rigid and fragile than the white matter [69]. Both white and gray matter had a stress-strain curve with an initial nonlinear region followed by a linear region. In the nonlinear region (<10%) the stresses were not statistically different but were higher in the gray matter than white matter in the linear region. The gray matter ruptured at lower strains than  15  Chapter 1.  Introduction  the white matter [69].  1.4  Clinical Spinal Cord Injury  Managed treatment for SCI begins at the scene of the injury with the arrival of an emergency response team, followed by various stages of clinical care, ending when the patient is discharged from a rehabilitation centre. Throughout each stage, the goals of treatment remain the same: [7, 58] 1. maximize neurological recovery; 2. restore normal alignment and correct deformity; 3. promote spinal stability/fusion; 4. minimize acute/chronic pain; 5. facilitate early mobilization and rehabilitation; 6. minimize hospitalization and cost; 7. detect secondary complications early and treat promptly. Over the past several decades, the treatment strategies for SCIs have improved considerably. The development of rapid emergency transport systems, with trained technicians aware of the importance of early immobilisation, has resulted in a reduction of the number of SCI patients presented with complete lesions [7]. Imaging modalities, such as computed tomography (CT) and magnetic resonance imaging (MRI), have enabled unprecedented visualization of both the osseous and soft tissue injuries [112]. Improved routines of daily care and rehabilitation have reduced the incidence of secondary complications, mortality rate (46% reduction) and average length of hospitalization (50%) [58, 98, 138, 149]. However, direct clinical treatment of the spinal cord lesion has seen few successes thus far.  1.4.1  Parameters of Injury  The detection of acute SCIs remains challenging, as 9.1% of SCIs are missed during the initial assessment [110], regardless of advancements in imaging [39, 58]. This has been attributed to a 16  Chapter 1.  Introduction  lack of consistent imaging protocols for visualization of the extent of spinal cord compression. Most SCI patients receive a CT scan of their injury, but CT alone often underestimates the extent of spinal cord compression. Excellent visualization of the soft tissue components can be achieved through M M , but due to the limited availability and relatively high costs, only 54% of SCI patients in North America receive a MRI at some time in their treatment regime [139]. However, even when thorough imaging of both the osseous and soft tissue is performed, the magnitude of the initial trauma remains difficult, if not impossible, to ascertain. Many attempts have been made to find an association between the geometry of the bony encroachment within the spinal canal and the level of neurological compromise. No correlations have been found between the absolute size of the skeletal injury (retropulsion distance, interpedicular distance, anterior/posterior body heights) [21, 75, 89, 118, 144, 145], but the shape, as determined by the sagittal-to-transverse diameter ratio, and reduction in cross-sectional area have been shown to be associated with the extent of neurological damage [61, 144]. Burst fractures and fracture-dislocation are the primary mechanisms of spinal cord contusion (Chapter 1.2) resulting in residual compression following the primary injury [44]. Burst fractures are associated with the greatest residual canal occlusion, and fracture-dislocations produce the highest percentage of complete injuries [75]. A complete injury is considered to be the absence of sensory or motor function caudal to the level of injury, 48 hours post-injury [7, 9]. The assessment is made at this time because the probability of neurological recovery following complete injury is rare; < 3% if no improvement by 24 hours, and almost no chance if no recovery by 48 hours [7, 38]. The distinction between complete and incomplete injuries is important because it impacts the amount of time that is spent on rehabilitation attempts to restore bowel and bladder function [7]. Regardless of surgery or intervention, most incomplete SCIs experience some spontaneous neurological recovery that eventually plateaus. Once this threshold is met, the prognosis for further neurological recovery is limited [141]. However, numerous investigators have reported further significant neurological recovery following delayed surgical decompression, months to years following the plateau [7, 22, 49]. This suggests that compression of the spinal cord is an important 17  Chapter 1.  Introduction  contributing factor of neurological function and that earlier decompression may increase the final functional level that is achieved. 1.4.2  Surgical Decompression  Decompression is the process by which a surgeon alleviates pressure on the spinal cord and/or nerve roots through surgical or non-surgical intervention, with the aim of improving the condition of the remaining intact neurological elements [9]. Surgical decompression includes operative reduction of a dislocation or removal of impinging bone or disc fragments. Non-surgical treatments can be administered by the reduction of a dislocation through traction or the reduction of cord swelling via pharmacological agents. The immediate benefit of decompression is the elimination of distortion of the cord thereby lessening compression of the neural tissue and vasculature. Although this may not help the site of primary lesion, it may benefit adjacent levels that derive their blood supply from sulcal arteries that pass through the injury site [2]. There exists considerable controversy on the merit of surgical decompression. Wagner and Chehrazi claimed that the strong correlation that they observed between the initial neurological score at admission with that obtained one year post-injury supported the importance of the initial trauma as the primary determinant of injury severity and concluded that decompression in any form is unwarranted [145]. It has also been shown that incomplete injuries, regardless of treatment, do not demonstrate significant motor recovery following treatment and decompressed facet dislocations have poorer outcomes than uncompressed dislocations [147]. Furthermore, surgical decompression of thoracic injuries do not favour anterior decompression, as these patients do not recover appreciably to note changes in their functional activities of daily living [108, 141]. Recent advances in the safety and efficacy of surgical decompression have begun to change earlier perceptions that no difference existed between operative and non-operative treatment in lengths of stay and neurological recovery [137] as well as reports that some patients became worst neurologically following surgery [141]. In a nine-month retrospective review of 36 SCI treatment centres in North America (1994-1995), Tator et al [139] found that surgery was performed in  18  Chapter 1.  Introduction  66% of SCI patients. Decompressive surgery was performed in 68.2% of the operations, and fusion was performed in 85.7%. Of the 173 cervical injuries reported, non-surgical traction was performed in 47% of these cases, but persisting deformation of the spinal cord remained in 43%. Furthermore, 8.1% of patients were reported to experience neurological deterioration during traction. Operative treatment has also been associated with a lower mortality rate (6.1%) compared to conservative treatment (15.2%) [137], decreased length of hospitalization, earlier movement and rehabilitation, and improved bladder function and pain patterns [21, 86, 97, 114, 141]. Positive effects associated with spinal cord decompression following injury have been reported in a few studies [4, 86, 148]. Aebi et al. reported that 23 of the 30 patients (100 total) that demonstrated neurological improvement following cervical SCI had undergone decompressive surgery within six hours of the initial trauma [4]. Weinshel et al. also found that 71% (53 of 74 patients) of patients that underwent decompressive procedures showed neurological improvement in comparison to 49% (7 of 16) that only received fusion without decompression [148]. Krengel et al. also demonstrated significant improvement in neurological function in 13 patients that underwent early reduction, stabilization or decompression [86]. However, control groups were not included in these retrospective reviews, so the improved neurological function that they observed cannot conclusively be associated with decompression. Following an extensive review of published literature concerned with the role of decompression in SCI between 1966 and 2001, Fehlings et al [49] reported than no conclusive data exists showing the benefit of surgical vs. conservative treatment. Spinal cord decompression is one of the few treatment options that clinicians can currently use to treat patients with SCIs, until direct beneficial effects are proven its use will remain controversial. When the clinician chooses to perform decompression they are also faced with the decision of using a conservative (non-surgical) versus a surgical technique. The additional benefits of surgical treatment that have been shown in recent years suggests that surgery may be the more viable option. A miraculous cure for spinal cord trauma does not exist and until that time treatment options such as decompression that have the potential to influence the spread and severity of SCIs will remain topics of considerable interest for clinicians and SCI researchers. 19  Chapter 1.  1.5  Introduction  Experimental Models of Spinal Cord Injury  The validation of SCI models, as relevant representations of clinical injuries, is hindered by the complex, arbitrary and ill-defined nature of clinical trauma [20]. The vague understanding of clinical SCI is one of the primary reasons that various models have been developed and refined over the past century. While clinical studies can take years and easily become convoluted, the objectives of experimental models can be focused to provide valuable insight into the pathology of SCI within a relatively short period of time. It has been established that the pathology of SCI is a two-stage process consisting of primary and secondary damage [8, 17, 27, 44, 120, 135] (See Section 1.7), but the contribution of different primary mechanisms of injury, such as contusion, distraction, shear and residual compression, has not been fully elucidated. Models of SCI have been conducted in various species, including cats [26, 51, 67, 73, 99, 122, 140, 146], dogs [5, 28, 30-32, 38, 43, 83, 129, 130], ferrets [10, 11, 74], mice [70, 121, 126], monkeys [24, 78-80, 82], and rats [39, 46, 50, 59, 60, 65, 76, 92, 94, 100, 104, 111, 128]. The primary modes of injury that have been established are partial/complete cord transection, maintained compression and rapid dynamic contusion. Transection models have been used extensively in regeneration studies, as they allow easy assessment of axonal re-growth across the lesion. Their use in evaluating the influence of injury mechanisms on the pathology of SCI is not as well suited. Complete transection of the spinal cord is relatively rare in humans which primarily arises from non-penetrating contusive injuries. 1.5.1  Maintained Compression  Maintained compression has been produced by balloon catheters [80-82, 104, 129, 130, 140], static weights [5, 35], trans-laminar screws [60], hydraulic pistons [29, 30, 32], and modified aneurysm clips [13, 46, 76, 113]. While these methods do allow stable, long-term compression of the spinal cord, the rate of initial compression is relatively slow. The clinical significance of slow and maintained compression is somewhat distinct from rapid contusion techniques. It has previously been shown that SCI is rate sensitive [74, 124] and can generally tolerate slow compression [80] and stretching, as deformation by one-third does not produce significant damage  20  Chapter 1.  Introduction  if applied slowly [152]. Although the modified aneurysm clip can be closed at a significantly faster rate, this model applies a constant force to the spinal cord tissue that does not allow dissipation of the load due to viscoelastic relaxation. 1.5.2  Dynamic Contusion  The first readily reproducible model of spinal cord contusion, developed by Allen in 1911, was based upon the potential energy of a dropped mass. Due to its relatively simple functionality, the weight-drop technique has become the primary method by which experimental contusion injuries are produced. The New York University (NYU) [15] weight-drop system was developed with strict experimental parameters to address shortcomings in the original model. Variations in both weight and height for constant g-cm products, presence and size of an impact plate and occurrence of secondary impacts have been shown to influence the severity of injury [10, 11, 76, 83]. Recently, computer-controlled models of contusion have been produced that allow the direct measurement of the physical parameters of injury. The quantification of impact parameters (displacement, rate and force) is necessary if correlations are to be made with the histological and functional outcome severities. The Ohio State University (OSU) [101, 123, 127] controlled contusion model utilizes an electromagnetic shaker to produce a rapid impact to the exposed spinal cord. The impact is generated in displacement control referenced to a baseline level of 3 kdynes (0.03 N) that produces a slight dimpling of the spinal cord. The current, commercially available version of the OSU uses an impact plate to control the displacement of the contusion tip as opposed to a direct feedback loop. This set-up produces considerable vibration at the peak of the injury stroke, complicating the interpretation of the peak injury parameters and preventing the measurement of visoelastic load relaxation. A force-defined feedback controlled device utilizing a commercially available Infinite Horizons (IH) [117] spinal cord injury system has also been developed. This system does not require the establishment of a baseline level to ensure a highly reproducible injury, but the use of force as the feed-back mechanism prevents the controlled retraction of the impact tip to a specific percentage of the peak displacement.  21  Chapter 1.  1.5.3  Introduction  R e s i d u a l Decompression  Strong correlations between the degree of injury have been shown to exist for the displacement and duration of compression, as well as the force and velocity of the injury, using the previously discussed injury models [12, 20, 25, 74, 101, 113, 127]. The degree of injury has also been found to be affected by the type and concentration of anaesthesia, size and species of animal, rigidity of spinal column fixation and velocity of the injury [12, 51, 66, 74, 105, 115, 116]. The neurological deficit following experimental studies investigating the effects of residual compression of the spinal cord has been shown to be determined by the magnitude and duration of the compression [29-31, 38, 40, 57, 60, 82, 113, 129-131]. These findings associated with residual compression of the spinal cord have primarily been derived from experiments utilizing slow, maintained compression. Only one published study to date has evaluated the effects of spinal cord compression following an initial contusion injury of the cord [39]. They found that compression alone (20% or 30% of spinal canal area) did not produce neurological injury until canal narrowing reached 50%, whereas trauma that included an initial contusion followed by the placement of spacers that occupied 20% or 35% of the canal area produced greater injury than a purely contusive injury. A contusion injury initiates a cascade of cellular damage [17], increasing the sensitivity of the spinal cord to compression [39], and disrupts the cellular network, possibly reducing the ability of the spinal cord to support a load. While these models have produced compelling evidence that persistent compression progressively increases the severity of SCI, the relationship between the primary contusion injury and resulting residual compression of the spinal cord, as observed in the clinical setting, has not been fully investigated. Cadaveric studies of spinal canal occlusion following a burst fracture have demonstrated that the post-injury level (18%-66%) of occlusion is less than the maximum transient level [33, 34, 106, 150]. Therefore, to create a clinically representative model of residual spinal cord compression, it is necessary to produce a dynamic primary contusion, followed by maintained compression of the spinal cord, that is equal to or less than the peak displacement of the initial contusion injury.  22  Chapter 1.  1.6  Introduction  Blood Flow Measurement  The blood provides the means for the cells of the body to acquire the nutrients necessary for their survival. When the local flow of blood becomes arrested, the neighbouring cells quickly become starved of oxygen, and waste byproducts resulting from their metabolism begin to accumulate. If the blood flow remains low for an extended period of time, the cells may become damaged, leading to their eventual death. This is especially true for the neural cells of the spinal cord. The brain and spinal cord are two of the most metabolically active organs of the body [44, 77, 103], yet the spinal cord only receives 30-45% (based on weight) of the blood flow to the brain, and due to limited anastomotic pathways, it has a minimal capacity to adapt to blood flow changes [105]. It is therefore not surprising that valuable information can be ascertained through local blood flow measurements of the traumatically injured spinal cord, as vascular injury plays a key role in both primary and secondary damage to the cord [136]. Techniques that have previously been utilized in the measurement of blood flow through the spinal cord include: hydrogen clearance; radioactive or fluorescent labelled microspheres; and laser doppler (LD)flowmetry.The hydrogen clearance method is based upon the rate at which inhaled hydrogen is cleared from the surrounding tissue, producing a change in the potential of a platinum electrode [37]. This technique can be used for repetitive measurements in specific regions of white or gray matter. The main disadvantages of the hydrogen clearance method are the relatively small volume of measurement and the need for invasive insertion of the probe [37]. Microsphere techniques depend upon the uniform injection of millions of distinctly labelled lbum spheres into the circulatory system [90]. It is assumed that the microspheres become completely entrapped within the first passage though the capillary network of the body and that they do not disrupt the circulatory or physiological status of the tissue [52]. This technique allows the determination of blood flow throughout the circulatory system while providing excellent spatial resolution. However, only a limited number of measurements can be taken due to the finite number of distinct radioactive and fluorescent labels commercially available.  23  Chapter 1.  1.6.1  Introduction  Laser D o p p l e r F l o w m e t r y  L D flowmetry is an optical technique for estimation of the microcirculation based upon the doppler principle [19, 90]. It is applied by directing a continuous laser beam via a fibre optic cable to the surface of the tissue. The monochromatic light undergoes a shift in its frequency proportional to the velocity of any moving particles, while backscatter from the underlying stationary tissues remains unaltered [19]. The measurement volume of L D flowmetry is considered to be within a hemisphere of 1 mm radius but varies according to the optical properties of the tissue [87]. A spatial resolution of this magnitude is relatively high compared to many other blood flow measurement techniques [52, 90]. However, due to variability in the measurement volume, it is not possible to attain absolute blood flow values [90], resulting in relative L D flowmetry signals recorded in perfusion units, a product of the local velocity and concentration of blood cells [87]. L D flowmetry is simple to use, requiring the placement of the fibre optic probe near vascularized tissue for the creation of a perfusion signal and the establishment of a baseline level from the initial recordings. However, consideration must also be given to the set-up and operational environment in which L D flowmetry is used. Ambiguous signals can easily be created due to multiple doppler shifts, motion artefact noise or displacement of the probe from the baseline position. Multiple doppler shifts result from scattering of the laser light by more than one blood cell, resulting in an accumulation of the respective velocities and a non-linear response to changes in the flow [87]. For this reason the surface of the cord must be devoid of visible vessels, as the theoretical justification (linear relationship between changes in L D signal and blood flow) is only supported at the microcirculation level [90]. To overcome motion artefacts and unwanted displacement of the L D probe, meticulous stability and isolation is necessary [90]. Motion artefact noise arises from unwanted relative motion between the probe and tissue sample. Due to the non-homogeneous distribution of vessels within the spinal cord, movement of the L D probe across the surface will result in substantial shifts in the perfusion signal, requiring the establishment of a new baseline level for relative measurements [52, 87]. For internal measurements requiring the surgical exposure of the sampling tissue, complete hemostasis is  24  Chapter 1.  Introduction  required. The accumulation of blood below the probe can cause aberration of the backscatter light, producing variations in the spectral power distribution of the detected light. This results in an apparent shift from the baseline level and corresponding change in the perfusion signal [90] (Figure 1.7). Aberrated Backscatter Signal  Unaltered Backscatter Signal  Wavelength  Wavelength  (a) Spectral power distribution of  (b) Spectral power distribution of  aberrated back scatter signal.  unaltered back scatter signal.  Figure 1.7: Reduction of the backscatter spectral power distribution following the accumulation of blood below the laser doppler (LD) probe. Perceived as a reduction in the measured blood flow and corresponding perfusion signal.  1.6.2  N o r m a l Spinal C o r d B l o o d Flow  The blood flow through the spinal cord is three times higher in the gray matter (60 ml/100 gm/min) than the white matter (20 ml/100 gm/min) and is relatively homogeneous across different regions (ventral, lateral and dorsal) and segments (cervical, thoracic and lumbar) [62, 63]. Normal spinal cord blood flow is independent of the mean systolic blood pressure over a wide range due to the process of autoregulation [42, 105]. Autoregulation maintains the spinal cord blood flow by regulating the calibre of the small arteries, contracting their muscular walls if the pressure inside rises and relaxing if the pressure falls [77]. The spinal cord blood flow will decrease in a linear fashion if the blood pressure drops below 50-60 mmHg and increases linearly as the blood pressure rises above 130-135 mmHg [64, 78].  25  Chapter 1.  1.7  Introduction  Pathology of Spinal Cord Injury  Trauma to the spinal cord results in immediate primary damage that is followed by a cascade of cellular processes (secondary injury) leading to an extension of the injury into the surrounding tissue (Figure 1.8). The primary damage is a direct result of the mechanical insult to the local tissues, leading to the disruption of the cellular and microvascular networks. The understanding of the concept of secondary injury has evolved in recent years to include a multitude of effects such as: systemic effects (changes in heart rate and blood pressure), vascular effects (haemorrhage, loss of autoregulation and microcirculation), electrolyte changes (increased intracellular ion concentrations), biochemical changes (neurotransmitter accumulation, free radical production), and apoptosis [47].  Primary  Gray Matter  Region of  Axons  Progression of SCI from initial focal primary injury to encompass surrounding tissues through secondary mechanisms of degeneration. Based on figure from Beattie [16]. Figure  1.8:  1.7.1  Progression of Injury  Following acute trauma of the spinal cord, the first observable indications of tissue trauma are the presence of disrupted axons and cell bodies within the local region of injury. Within a few minutes haemorrhage and edema spread throughout the central gray matter due to increased vascular permeability [72]. Within 48 hours the spinal cord becomes soft and liquified due to the conversion of myelin to fat, and by 5 days activated glial cells have formed fibrous tissue  26  Chapter 1.  Introduction  within the lesion [72]. Within a few weeks the lesion has developed into a spindle shaped cavity aligned along the long axis of the spinal cord [36]. Three to six months following the SCI Wallerian degeneration of the ascending and descending axons is visible as well as atrophy of the adjacent nerve roots [72]. The lesion cavity is normally restricted to the central gray matter and the immediately adjacent white matter. 62% of injuries demonstrate continuity across the lesion site in the lateral, anterior and posterior white matter [36, 72]. Ventral motor tracts are more likely to be damaged than those of the sensory systems due to their large calibre and dependence upon the myelin sheath for effective signal conduction [133]. Neurological recovery may occur in the nerve roots at the level of injury or within the spinal cord itself. Nerve roots are the structures that have the greatest tendency to recover following SCI, followed by local gray matter and local white matter [133]. 1.7.2  V a s c u l a r Injury  Vascular insult plays a key role in primary and secondary injury, as cells of the central nervous system have very high metabolic rates and are therefore vulnerable to reductions in blood flow and eventual ischemia [97, 136]. Ischemia exacerbates the injury, leading to the death of the affected cells due to the creation of free radicals and other biproducts of cellular metabolism [44]. The conversion from aerobic to anaerobic metabolism leads to the buildup of lactate that cannot be removed due to poor perfusion [105]. These detrimental effects lead to the death of the cells by necrotic and apoptotic pathways. Acute trauma results in the disruption of the anterior sulcal arteries and capillary network of the gray matter, leading to haemorrhage in the gray matter followed by diminished perfusion [53, 136]. However, the main arteries such as the anterior and posterior spinal arteries and veins usually remain intact [44, 72, 85]. The surrounding white matter that derives its blood from the sulcal arteries also experiences varying levels of damage. Moderate to severe injuries result in the loss of autoregulation and changes in the regional blood flow patterns [37, 44, 102, 135]. Systemic effects include hypotension and diminished 27  Chapter 1.  Introduction  cardiac output. The general consensus is that SCI results in the diminished perfusion of the gray matter [1, 44, 53, 85, 128] and high variability is seen in the white matter (hyperaemia or diminished perfusion) [37, 64, 84].  1.8  Summary  Clinical treatment involves diagnosis of the mechanism of injury, surgical realignment and stabilization of the spinal column as well as efforts to minimize the cellular-level damage. Basic and applied research has investigated the effects of various surgical techniques (hypothermia, decompression, etc.) and pharmacological agents (growth factors, antioxidants, channel blockers, antibodies, corticosteroids, etc.) [37, 45]. Yet, the only therapeutic window established in humans for minimizing the extent of cellular damage to the spinal cord is the administration of methylprednisolone within 8 hours of injury [23, 120, 139, 153]. Methylprednisolone often results in only modest improvements in neurological recovery [49], and some clinical studies and retrospective reviews have found no effect of high-dose corticosteroid administration on neurological function [55, 68, 109]. Another equally controversial form of intervention is surgical decompression to alleviate residual compression of the spinal cord [7]. The controversy arises from the fact that there exists no consensus for the necessity of surgical decompression or timing when surgery does occur. A disconnection exists between the conclusions that arise from experimental studies and clinical reviews of spinal cord compression. Numerous investigators have documented that the degree and duration of experimental compression determine the extent of neurological impairment [31, 38-40, 57, 60, 113, 129, 130]. These findings have resulted in many individuals recommending the immediate clinical decompression of the persistently compressed cord. However, the clinical evidence is not as compelling. While retrospective reviews have suggested that early decompression of the spinal cord may improve the neurological outcome [86, 95, 107, 114, 119], others have not [91, 109, 137, 143, 145]. Clinical reviews have even suggested that surgical decompression may further impair the neurological status of the patient and create unnecessary secondary complications. Regardless, there is no conclusive evidence that clinical decompression  28  Chapter 1.  Introduction  is beneficial to SCI patients presented with residual compression of the spinal cord [49]. The majority of experimental models that have been employed in the investigation of residual compression of the spinal cord have utilized slow maintained compression. These models are incapable of replicating the initial high-rate contusion injury typical of clinical SCIs. It has previously been shown that SCI is rate sensitive [74, 124], and can generally tolerate slow compression [80] and stretching, as deformation by one-third does not produce significant damage if applied slowly [152]. Therefore, to create a clinically representative model of residual spinal cord compression, it is necessary to produce a dynamic primary contusion, followed by maintained compression of the spinal cord, that is equal or less than the peak displacement of the initial contusion injury. Such a model may help to elucidate the interaction of the primary contusion injury and subsequent residual compression and their contributions to the overall spinal cord trauma.  1.9  Objectives & Scope  The primary objective of this thesis was to determine the influence of 40% (0.4 mm) and 90% (0.9 mm) of residual spinal cord compression on the extent and progression of neurological damage within the rat, following a moderate (1 mm, 700 mm/s) contusive SCI. This was accomplished by monitoring the microvascular blood flow following injury and comparing the damage to the gray and white matter of the spinal cord postmortem. It was hypothesized that residual compression would produce ischemia at the injury epicentre and diminished blood flow in the adjacent tissue resulting in an extension of injury to the spinal cord that was dependent upon the magnitude of the compression. To facilitate the completion of the primary objective, a number of secondary tasks were performed: 1. Establish a protocol for the development of graded residual compression of the rat spinal cord following a contusion injury of reproducible magnitude;  29  Chapter 1.  Introduction  2. Establish a method to rigidly support the spinal column during the contusion injury and the subsequent period of residual compression; 3. Establish a method to continuously monitor the microvascular blood flow with a LD flowmetry system through the dorsolateral spinal cord; 4. Complete a series of animal tests for comparison of the gray and white matter injury following pure contusion or contusion plus subsequent residual compression of the spinal cord; 5. Analyze the injuries using histological and immunohistochemical procedures. The scope of this thesis was to determine if the injury to the spinal cord, relative to a pure contusion injury, was affected by residual compression and whether the relative magnitude of the residual compression was an important factor in the progression of SCI pathology. Previously developed models of experimental SCI do not allow the independent control of the contusion and residual compression magnitudes. Therefore a major focus of this thesis was the integration of these two mechanisms of injury into a simple and reproducible method. Although the model developed for this thesis allowed essentially infinite variation in the mechanical factors of the contusion (velocity, displacement...)  and residual compression (magnitude, duration...), a  minimal number were chosen. The contusion injury was designed to be similar for all injury groups, producing a 1 mm injury at 700 mm/s.  Two levels of residual compression were  chosen: 0.4 mm and 0.9 mm, or 40% and 90% of the peak contusion depth. The duration of compression was 1 hour for both residual groups, and all animals were euthanized 3 hours post-contusion (controls: following baseline touch). These levels and durations were chosen due to the preliminary nature of the investigation. The levels (90% and 40%) correspond to realistic limits of residual compression, and the durations (1 hour and 3 hours) were chosen to limit the compounding effects of secondary damage and allow multiple tests to be performed on a single day (See Chapter 2.4.2 for rational). To evaluate the effects of residual compression on SCI, the blood flow one level distal to the injury epicentre was measured with a LD flowmetry system, and histopathology was performed to quantify the damage to the white and gray matter. Hematoxylin and Eosin (H&E) was used for quantification of the overall haemorrhage. 30  Chapter 1.  Introduction  The rostral-caudal extent of neuronal gray matter damage was quantified using neuronal nuclei (NeuN) and Fluoro-Jade (FJ), and the extent of axonal disruption in the white matter was quantified using /3-amyloid precursor protein (/3-APP). It is anticipated that the data collected from this study will be used in part, with future investigations of residual compression, to elucidate the interaction between varying levels of contusive and compressive trauma and the resultant levels of neurological and functional impairment.  31  Chapter 1.  1.10  Introduction  References  [1] Blood pressure management after acute spinal cord injury. S58-62, 2002.  Neurosurgery,  50(Suppl 3):  [2] Management of acute central cervical spinal cord injuries. S166-72, 2002.  Neurosurgery,  50(Suppl 3):  [3] Ackery A., Tator C , and Krassioukov A. A global perspective on spinal cord injury epidemiology. J Neurotrauma, 21(10):1355-70, October 2004. [4] Aebi M., Mohler J., Zch G. A., and Morscher E. Indication, surgical technique, and results of 100 surgically-treated fractures and fracture-dislocations of the cervical spine. Clin Orthop Relat Res, (203):244-257, February 1986. [5] Aki T. and Toya S. Experimental study on changes of the spinal-evoked potential and circulatory dynamics following spinal cord compression and decompression. Spine, 9(8): 800-9, November-December 1984. [6] Allen A. R. Remarks on the Histopathological changes in the Spinal Cord due to Impact. An Experimental Study. J. Nerv. Ment. Dis., 41:141-147, 1914. [7] Amar A. P. and Levy M . L. Surgical controversies in the management of spinal cord injury. J Am Coll Surg, 188(5):550-66, May 1999. [8] Amar A. P. and Levy M . L. Pathogenesis and pharmacological strategies for mitigating secondary damage in acute spinal cord injury. Neurosurgery, 44(5): 1027-39, May 1999. [9] Anderson P. A. and Bohlman H. H. Anterior decompression and arthrodesis of the cervical spine: long-term motor improvement. Part II-Improvement in complete traumatic quadriplegia. J Bone Joint Surg Am, 74(5):683-92, June 1992. [10] Anderson T. E. A controlled pneumatic technique for experimental spinal cord contusion. J Neurosci Methods, 6(4):327-33, November 1982. [11] Anderson T. E. Spinal cord contusion injury: experimental dissociation of hemorrhagic necrosis and subacute loss of axonal conduction. J Neurosurg, 62(l):115-9, January 1985. [12] Anderson T. E. and Stokes B. T. Experimental models for spinal cord injury research: physical and physiological considerations. J Neurotrauma, 9(Suppl l):S135-42, March 1992. [13] Anthes D. L., Theriault E., and Tator C. H. Characterization of axonal ultrastructural pathology following experimental spinal cord compression injury. Brain Res, 702(1-2): 1-16, 1995. [14] Baas P. W. Elaboration of the Axonal Microtubule Array During Development and Regeneration. In Kalb R. G. and Strittmatter S. M., editors, Neurobiology of Spinal Cord Injury, Contemporary Neuroscience, pages 155-168. Humana Press, Totowa, N.J., 2000. [15] Basso D. M., Beattie M. S., and Bresnahan J. C. Graded histological and locomotor outcomes after spinal cord contusion using the N Y U weight-drop device versus transection. Exp Neurol, 139(2):244-56, June 1996. 32  Chapter 1.  Introduction  [16] Beattie M. S., Li Q., and Bresnahan J. C. Cell death and plasticity after experimental spinal cord injury. Prog Brain Res, 128:9-21, 2000. [17] Beattie M . S., Hermann G. E., Rogers R. C , and Bresnahan J. C. Cell death in models of spinal cord injury. Prog Brain Res, 137:37-47, 2002. [18] Bilston L. E. and Thibault L. E. The mechanical properties of the human cervical spinal cord in vitro. Ann Biomed Eng, 24(l):67-74, January-February 1996. [19] Bircher A., de Boer E. M., Agner T., Wahlberg J. E., and Serup J. Guidelines for measurement of cutaneous blood flow by laser Doppler flowmetry. A report from the Standardization Group of the European Society of Contact Dermatitis. Contact Dermatitis, 30(2):65-72, February 1994. [20] Blight A. R. Experimental Spinal Cord Injury Models. In Narayan R. K . , Wilberger J. E., and Povlishock J. T., editors, Neurotrauma, chapter 101, pages 1367-79. McGraw Hill, Health Professions Disivion, 1996. [21] Boerger T. O., Limb D., and Dickson R. A. Does 'canal clearance' affect neurological outcome after thoracolumbar burst fractures? J Bone Joint Surg Br, 82(5):629-35, July 2000. [22] Bohlman H. H. and Anderson P. A. Anterior decompression and arthrodesis of the cervical spine: long-term motor improvement. Part I-Improvement in incomplete traumatic quadriparesis. J Bone Joint Surg Am, 74(5):671-82, June 1992. [23] Bracken M . B., Shepard M . J., Collins W. F., Holford T. R., Young W., Baskin D. S., Eisenberg H. M . , Flamm E., Leo-Summers L., and Maroon J. A randomized, controlled trial of methylprednisolone or naloxone in the treatment of acute spinal-cord injury. Results of the Second National Acute Spinal Cord Injury Study. N Engl J Med, 322(20): 1405-1411, May 1990. [24] Bresnahan J. C. An electron-microscopic analysis of axonal alterations following blunt contusion of the spinal cord of the rhesus monkey (Macaca mulatta). J Neurol Sci, 37 (l-2):59-82, June 1978. [25] Bresnahan J. C , Beattie M . S., Todd r., F. D., and Noyes D. H. A behavioral and anatomical analysis of spinal cord injury produced by a feedback-controlled impaction device. Exp Neurol, 95(3):548-70, March 1987. [26] Brodkey J. S., Richards D. E., Blasingame J. P., and Nulsen F. E. Reversible spinal cord trauma in cats. Additive effects of direct pressure and ischemia. J Neurosurg, 37(5): 591-3, November 1972. [27] Carlson G. D. and Gorden C. Current developments in spinal cord injury research. J, 2(2):116-28, March-April 2002.  Spine  [28] Carlson G. D., Minato Y . , Okada A., Gorden C. D., Warden K . E., Barbeau J. M . , Biro C. L., Bahnuik E., Bohlman H. H., and Lamanna J. C. Early time-dependent decompression for spinal cord injury: vascular mechanisms of recovery. J Neurotrauma, 14(12):951-62, December 1997. 33  Chapter 1.  Introduction  [29] Carlson G. D., Warden K . E., Barbeau J. M . , Bahniuk E., Kutina-Nelson K . L., Biro C. L., Bohlman H . H., and LaManna J. C. Viscoelastic relaxation and regional blood flow response to spinal cord compression and decompression. Spine, 22(12) :1285—91, June 1997. [30] Carlson G. D., Gorden C. D., Nakazowa S., Wada E., Warden K., and LaManna J. C. Perfusion-limited recovery of evoked potential function after spinal cord injury. Spine, 25 (10):1218-26, May 2000. [31] Carlson G. D., Gorden C. D., Nakazawa S., Wada E., Smith J. S., and LaManna J. C. Sustained spinal cord compression: part II: effect of methylprednisolone on regional blood flow and recovery of somatosensory evoked potentials. J Bone Joint Surg Am, 85(A1): 95-101, January 2003. [32] Carlson G. D., Gorden C. D., OliffH. S., Pillai J. J., and LaManna J. C. Sustained spinal cord compression: part I: time-dependent effect on long-term pathophysiology. J Bone Joint Surg Am, 85(Al):86-94, January 2003. [33] Carter J. W., Mirza S. K., Tencer A. F., and Ching R. P. Canal geometry changes associated with axial compressive cervical spine fracture. Spine, 25(l):46-54, January 2000. [34] Chang D. G., Tencer A. F., Ching R. P., Treece B., Senft D., and Anderson P. A. Geometric changes in the cervical spinal canal during impact. Spine, 19(8):973-980, April 1994. [35] Croft T. J., Brodkey J. S., and Nulsen F. E. Reversible spinal cord trauma: a model for electrical monitoring of spinal cord function. J Neurosurg, 36(4):402-406, April 1972. [36] Croul S. E. and Flanders A. E. Neuropathology of human spinal cord injury. Adv Neurol, 72:317-23, 1997. [37] de la Torre J. C. Spinal cord injury. Review of basic and applied research. Spine, 6(4): 315-35, July-August 1981. [38] Delamarter R. B., Sherman J., and Carr J. B. Pathophysiology of spinal cord injury. Recovery after immediate and delayed decompression. J Bone Joint Surg Am, 77(7): 1042-9, July 1995. [39] Dimar n., J. R., Glassman S. D., Raque G. H., Zhang Y . P., and Shields C. B. The influence of spinal canal narrowing and timing of decompression on neurologic recovery after spinal cord contusion in a rat model. Spine, 24(16):1623-33, August 1999. [40] Dolan E. J., Tator C. H., and Endrenyi L. The value of decompression for acute experimental spinal cord compression injury. J Neurosurg, 53(6):749-55, December 1980. [41] Drake R. L., Vogl W., Mitchell A. W. M . , and Gray H. Gray's anatomy for students. Elsevier/Churchill Livingstone, Philadelphia, 2005. [42] Ducker T. B. and Perot J., P. L. Spinal cord blood flow compartments. Trans Am Neurol Assoc, 96:229-31, 1971. 34  Chapter 1.  Introduction  [43] Ducker T. B. and Perot J., P. L. Spinal cord oxygen and blood flow in trauma. Surg Forum, 22:413-5, 1971. [44] Dumont R. J., Okonkwo D. O., Verma S., Hurlbert R. J., Boulos P. T., Ellegala D. B., and Dumont A. S. Acute spinal cord injury, part I: pathophysiologic mechanisms. Clin Neuropharmacol, 24(5):254-64, September-October 2001. [45] Dumont R. J., Verma S., Okonkwo D. O., Hurlbert R. J., Boulos P. T., Ellegala D. B., and Dumont A. S. Acute spinal cord injury, part II: contemporary pharmacotherapy. Clin Neuropharmacol, 24(5):265-79, September-October 2001. [46] Fehlings M . G. and Nashmi R. A new model of acute compressive spinal cord injury in vitro. J Neurosci Methods, 71(2):215-24, March 1997. [47] Fehlings M. G. and Sekhonm L. H. Cellular, Ionic, and Biomolecular Mechanisms of the Injury Process. In Contemporary Management of Spinal Cord Injury: from Impact to Rehabilitation, Neurosurgical Topics, chapter 5, pages 33-50. American Association of Neurological Surgeons, Park Ridge, 111., 2000. [48] Fehlings M . G. and Tator C. H. An evidence-based review of decompressive surgery in acute spinal cord injury: rationale, indications, and timing based on experimental and clinical studies. J Neuro surg, 91(1 Suppl):l-ll, July 1999. [49] Fehlings M . G., Sekhon L. H., and Tator C. The role and timing of decompression in acute spinal cord injury: what do we know? What should we do? Spine, 26(24 Suppl): S101-10, December 2001. [50] Fiford R. J., Bilston L. E., Waite P., and Lu J. A vertebral dislocation model of spinal cord injury in "rats. J Neurotrauma, 21(4):451-8, April 2004. [51] Ford R. W. A reproducible spinal cord injury model in the cat. J Neurosurg, 59(2): 268-75, August 1983. [52] Frerichs K . U . and Feuerstein G. Z. Laser-Doppler flowmetry. A review of its application for measuring cerebral and spinal cord blood flow. Mol Chem Neuropathol, 12(l):55-70, January 1990. [53] Fried L. C. and Goodkin R. Microangiographic observations of the experimentally traumatized spinal cord. J Neurosurg, 35(6):709-14, December 1971. [54] Fung Y . C. Biomechanics: Mechanical Properties of Living Tissues. Springer-Verlag, New York, 2nd edition, 1993. [55] George E. R., Scholten D. J., Buechler C. M., Jordan-Tibbs J., Mattice C , and Albrecht R. M . Failure of methylprednisolone to improve the outcome of spinal cord injuries. Am Surg, 61(8):659-63, August 1995. [56] Gray H. and Lewis W. H . Anatomy of the human body. Lea k, Febiger, Philadelphia and New York,, 20th edition, 1918. URL http://www.bartleby.com/107/.  35  Chapter 1.  Introduction  [57] Guha A., Tator C. FL, Endrenyi L., and Piper I. Decompression of the spinal cord improves recovery after acute experimental spinal cord compression injury. Paraplegia, 25(4):324-39, August 1987. [58] Gunnarsson T. and Fehlings M . G. Acute neurosurgical management of traumatic brain injury and spinal cord injury. Curr Opin Neurol, 16(6):717-23, December 2003. [59] Gupta R., Rowshan K., Chao T., Mozaffar T., and Steward O. Chronic nerve compression induces local demyelination and remyelination in a rat model of carpal tunnel syndrome. Exp Neurol, 187(2):500-8, July 2004. [60] Hashimoto T. and Fukuda N . New spinal cord injury model produced by spinal cord compression in the rat. J Pharmacol Methods, 23(3):203-12, May 1990. [61] Hashimoto T., Kaneda K., and Abumi K . Relationship between traumatic spinal canal stenosis and neurologic deficits in thoracolumbar burst fractures. Spine, 13(11):1268-72, November 1988. [62] Hayashi N., Green B. A., Gonzalez-Carvajal M., Mora J., and Veraa R. P. Local blood flow, oxygen tension, and oxygen consumption in the rat spinal cord. Part 2: Relation to segmental level. J Neurosurg, 58(4):526-30, April 1983. [63] Hayashi N., Green B. A., Gonzalez-Carvajal M., Mora J., and Veraa R. P. Local blood flow, oxygen tension, and oxygen consumption in the rat spinal cord. Part 1: Oxygen metabolism and neuronal function. J Neurosurg, 58(4):516-25, April 1983. [64] Hickey R., Albin M. S., Bunegin L., and Gelineau J. Autoregulation of spinal cord blood flow: is the cord a microcosm of the brain? Stroke, 17(6):1183-9, November-December 1986. [65] Hiruma S., Otsuka K., Satou T., and Hashimoto S. Simple and reproducible model of rat spinal cord injury induced by a controlled cortical impact device. Neurol Res, 21(3): 313-23, April 1999. [66] Holobotovskyy V. V., Arnolda L. F., and McKitrick D. J. Effect of anaesthetic and rat strain on heart rate responses to simulated haemorrhage. Acta Physiol Scand, 180(1): 29-38, January 2004. [67] Hung T. K., Lin H. S., Bunegin L., and Albin M . S. Mechanical and neurological response of cat spinal cord under static loading. Surg Neurol, 17(3):213-7, March 1982. [68] Hurlbert R. J. The role of steroids in acute spinal cord injury: an evidence-based analysis. Spine, 26(24 Suppl):S39-S46, December 2001. [69] Ichihara K., Taguchi T., Shimada Y., Sakuramoto I., Kawano S., and Kawai S. Gray matter of the bovine cervical spinal cord is mechanically more rigid and fragile than the white matter. J Neurotrauma, 18(3):361-367, March 2001. [70] Jakeman L. B., Guan Z., Wei P., Ponnappan R., Dzwonczyk R., Popovich P. G., and Stokes B . T. Traumatic spinal cord injury produced by controlled contusion in mouse. J Neurotrauma, 17(4):299-319, April 2000. 36  Chapter 1. Introduction  [71] Julien J. P. and Mushynski W. E. Neurofilaments in health and disease. Prog Nucleic Acid Res Mol Biol, 61:1-23, 1998. [72] Kakulas B. A. Pathology of spinal injuries. Cent Nerv Syst Trauma, l(2):117-29, Winter 1984. [73] Katsuta M . Influence of spinal cord compression on the C3-C4 propriospinal neurons in cats. J Orthop Sci, 8(3):367-73, 2003. [74] Kearney P. A., Ridella S. A., Viano D. C , and Anderson T. E. Interaction of contact velocity and cord compression in determining the severity of spinal cord injury. J Neurotrauma, 5(3):187-208, 1988. [75] Keene J. S., Fischer S. P., Vanderby J., R., Drummond D. S., and Turski P. A. Significance of acute posttraumatic bony encroachment of the neural canal. Spine, 14(8):799-802, August 1989. [76] Khan M . and Griebel R. Acute spinal cord injury in the rat: comparison of three experimental techniques. Can J Neurol Sci, 10(3):161-5, August 1983. [77] Kiernan J. A. and Barr M . L. Barr's the human nervous system: an anatomical viewpoint. Lippincott-Raven, Philadelphia, 7th edition, 1998. [78] Kobrine A. I., Doyle T. F., and Martins A. N . Autoregulation of spinal cord blood flow. Clin Neurosurg, 22:573-81, 1975. [79] Kobrine A. I., Doyle T. F., Newby N., and Rizzoli H. V. Preserved autoregulation in the rhesus spinal cord after high cervical cord section. J Neurosurg, 44(4):425-8, April 1976. [80] Kobrine A. I., Evans D. E., and Rizzoli H . Correlation of spinal cord blood flow and function in experimental compression. Surg Neurol, 10(l):54-9, July 1978. [81] Kobrine A. I., Evans D. E., and Rizzoli H. V. Correlation of spinal cord blood flow, sensory evoked response, and spinal cord function in subacute experimental spinal cord compression. Adv Neurol, 20:389-94, 1978. [82] Kobrine A. I., Evans D. E., and Rizzoli H . V. Experimental acute balloon compression of the spinal cord. Factors affecting disappearance and return of the spinal evoked response. J Neurosurg, 51(6):841-5, December 1979. [83] Koozekanani S. H., Vise W. M . , Hashemi R. M . , and McGhee R. B. Possible mechanisms for observed pathophysiological variability in experimental spinal cord injury by the method of Allen. J Neurosurg, 44(4):429-34, April 1976. [84] Koyanagi I. and Tator C. H. Effect of a single huge dose of methylprednisolone on blood flow, evoked potentials, and histology after acute spinal cord injury in the rat. Neurol Res, 19(3):289-99, June 1997. [85] Koyanagi I., Tator C. H., and Lea P. J. Three-dimensional analysis of the vascular system in the rat spinal cord with scanning electron microscopy of vascular corrosion casts. Part 2: Acute spinal cord injury. Neurosurgery, 33(2):285-91, August 1993. 37  Chapter 1.  Introduction  [86] Krengel I., W. F., Anderson P. A., and Henley M . B. Early stabilization and decompression for incomplete paraplegia due to a thoracic-level spinal cord injury. Spine, 18(14): 2080-7, October 1993. [87] Leahy M . J., de Mul F. F., Nilsson G. E., and Maniewski R. Principles and practice of the laser-doppler perfusion technique. Technol Health Care, 7(2-3):143-62, 1999. [88] Lifshutz J. and Colohan A. A brief history of therapy for traumatic spinal cord injury. Neurosurg Focus, 16(l):l-8, January 2004. [89] Limb D., Shaw D. L., and Dickson R. A. Neurological injury in thoracolumbar burst fractures. J Bone Joint Surg Br, 77(5):774-7, September 1995. [90] Lindsberg P. J., O'Neill J. T., Paakkari I. A., Hallenbeck J. M., and Feuerstein G. Validation of laser-Doppler flowmetry in measurement of spinal cord blood flow. Am J Physiol, 257(2 Pt 2):H674-80, August 1989. [91] Marshall L. F., Knowlton S., Garfin S. R., Klauber M . R., Eisenberg H. M., Kopaniky D., Miner M . E., Tabbador K., and Clifton G. L. Deterioration following spinal cord injury. A multicenter study. J Neurosurg, 66(3):400-4, March 1987. [92] Metz G. A., Curt A., van de Meent H., Klusman I., Schwab M . E., and Dietz V. Validation of the weight-drop contusion model in rats: a comparative study of human spinal cord injury. J Neurotrauma, 17(1):1—17, January 2000. [93] Miller K. and Chinzei K. Mechanical properties of brain tissue in tension. J Biomech, 35 (4):483-490, April 2002. [94] Mills C. D., Hains B. C , Johnson K . M . , and Hulsebosch C. E. Strain and model differences in behavioral outcomes after spinal cord injury in rat. J Neurotrauma, 18(8): 743-56, August 2001. [95] Mirza S. K . , Krengel r., W. F., Chapman J. R., Anderson P. A., Bailey J. C , Grady M . S., and Yuan H. A. Early versus delayed surgery for acute cervical spinal cord injury. Clin Orthop, 1(359):104-14, February 1999. [96] Moore K. L. and Dalley A. F. Clinically oriented anatomy. Lippincott Williams & Wilkins, Philadelphia, 4th edition, 1999. [97] National Institute of Neurological Disorders and Stroke [NINDS] . Spinal Cord Injury: Emerging Concepts. Technical report, National Institutes of Health, October 1 1996. URL http://www.ninds.nih.gov/news_and_events/proceedings/sci_report.htm. [98] National Spinal Cord Injury Statistical Center [NSCISC] . Spinal Cord Injury - Facts and Figures at a Glance. Technical report, National Spinal Cord Injury Statistical Center, 2005. URL http://www.spinalcord.uab.edu/show.asp?durki=21446. [99] Nelson E., Gertz S. D., Rennels M. L., Ducker T. B., and Blaumanis O. R. Spinal cord injury. The role of vascular damage in the pathogenesis of central hemorrhagic necrosis. Arch Neurol, 34(6):332-3, June 1977.  38  Chapter 1.  Introduction  [100] Noble L. J. and Wrathall J. R. Spinal cord contusion in the rat: morphometric analyses of alterations in the spinal cord. Exp Neurol, 88(l):135-49, April 1985. [101] Noyes D. H. Electromechanical impactor for producing experimental spinal cord injury in animals. Med Biol Eng Comput, 25(3):335-40, May 1987. [102] Ohashi T., Morimoto T., Kawata K., Yamada T., and Sakaki T. Correlation between spinal cord blood flow and arterial diameter following acute spinal cord injury in rats. Acta Neurochir (Wien), 138(3) :322-9, 1996. [103] Okonkwo D. O., Pettus E. H., Moroi J., and Povlishock J. T. Alteration of the neurofilament sidearm and its relation to neurofilament compaction occurring with traumatic axonal injury. Brain Res, 784(1-2): 1-6, February 1998. [104] Oro J. J., Gibbs S. R., and Haghighi S. S. Balloon device for experimental graded spinal cord compression in the rat. J Spinal Disord, 12(3):257-61, June 1999. [105] Osterholm J. L. The pathophysiological response to spinal cord injury. The current status of related research. J Neurosurg, 40(l):5-33, January 1974. [106] Panjabi M . M., Kifune M., Wen L., Arand M., Oxland T. R., Lin R. M., Yoon W. S., and Vasavada A. Dynamic canal encroachment during thoracolumbar burst fractures. J Spinal Disord, 8(l):39-48, February 1995. [107] Papadopoulos S. M., Selden N. R., Quint D. J., Patel N., Gillespie B., and Grube S. Immediate spinal cord decompression for cervical spinal cord injury: feasibility and outcome. J Trauma, 52(2):323-32, February 2002. [108] Petitjean M . E., Mousselard H., Pointillart V., Lassie P., Senegas J., and Dabadie P. Thoracic spinal trauma and associated injuries: should early spinal decompression be considered? J Trauma, 39(2):368-72, August 1995. [109] Pollard M . E. and Apple D. F. Factors associated with improved neurologic outcomes in patients with incomplete tetraplegia. Spine, 28(l):33-9, January 2003. [110] Poonnoose P. M . , Ravichandran G., and McClelland M . R. Missed and mismanaged injuries of the spinal cord. J Trauma, 53(2):314-320, August 2002. [Ill] Raines A., Dretchen K . L., Marx K., and Wrathall J. R. Spinal cord contusion in the rat: somatosensory evoked potentials as a function of graded injury. J Neurotrauma, 5 (2):151-60, 1988. [112] Rao K. C. V. G. MRI and CT of the spine. Williams & Wilkins, Baltimore, Md., U.S.A., 1994. [113] Rivlin A. S. and Tator C. H. Effect of duration of acute spinal cord compression in a new acute cord injury model in the rat. Surg Neurol, 10(l):38-43, July 1978. [114] Rosenfeld J. F., Vaccaro A. R., Albert T. J., Klein G. R., and Cotler J. M . The benefits of early decompression in cervical spinal cord injury. Am J Orthop, 27(l):23-8, January 1998. 39  Chapter 1.  Introduction  Salzman S. K . , Mendez A. A., Sabato S., Lee W. A., Ingersoll E. B., Choi I. FX., Fonseca A. S., Agresta C. A., and Freeman G. M . Anesthesia influences the outcome from experimental spinal cord injury. Brain Res, 521(1-2):33-9, June 1990. Salzman S. K . , Lee W. A., Sabato S., Mendez A. A., Agresta C. A., and Kelly G. Halothane anesthesia is neuroprotective in experimental spinal cord injury: early hemodynamic mechanisms of action. Res Commun Chem Pathol Pharmacol, 80(1):59-81, April 1993. Scheff S. W., Rabchevsky A. G., Fugaccia I., Main J. A., and Lumpp J., J. E. Experimental modeling of spinal cord injury: characterization of a force-defined injury device. J Neurotrauma, 20(2):179-93, February 2003. Scher A. T. Is the pattern of neurological damage of diagnostic value in the radiological assessment of acute cervical spine injury? Paraplegia, 19(4):248-52, 1981. Schlegel J., Bayley J., Yuan H., and Fredricksen B. Timing of surgical decompression and fixation of acute spinal fractures. J Orthop Trauma, 10(5):323-30, 1996. Sekhon L. H. and Fehlings M . G. Epidemiology, demographics, and pathophysiology of acute spinal cord injury. Spine, 26(24 Suppl):S2-12, December 2001. Seki T., Hida K., Tada M., Koyanagi I., and Iwasaki Y . Graded contusion model of the mouse spinal cord using a pneumatic impact device. Neurosurgery, 50(5):1075-81, May 2002. Senter H. J. and Venes J. L. Loss of autoregulation and posttraumatic ischemia following experimental spinal cord trauma. J Neurosurg, 50(2):198-206, February 1979. Somerson S. K . and Stokes B. T. Functional analysis of an electromechanical spinal cord injury device. Exp Neurol, 96(l):82-96, April 1987. Sparrey C. The Effect of Impact Velocity on Acute Spinal Cord Injury. Master's thesis, University of British Columbia, Vancouver, 2004. Steward O. Functional Neuroscience. Springer, New York, 2000. Stokes B. T. and Jakeman L. B. Experimental modelling of human spinal cord injury: a model that crosses the species barrier and mimics the spectrum of human cytopathology. Spinal Cord, 40(3):101-9, March 2002. Stokes B. T., Noyes D. FL, and Behrmann D. L. An electromechanical spinal injury technique with dynamic sensitivity. J Neurotrauma, 9(3): 187-95, Fall 1992. Taoka Y. and Okajima K . Spinal cord injury in the rat. Prog Neurobiol, 56(3):341-58, October 1998. Tarlov I. M . Spinal cord compression studies. III. Time limits for recovery after gradual compression in dogs. AMA Arch Neurol Psychiatry, 71(5):588-97, May 1954. Tarlov I. M . and Klinger H. Spinal cord compression studies. II. Time limits for recovery after acute compression in dogs. AMA Arch Neurol Psychiatry, 71(3):271-90, March 1954. 40  Chapter 1. Introduction  [131] Tarlov I. M., Klinger H., and Vitale S. Spinal cord compression studies. I. Experimental techniques to produce acute and gradual compression. AMA Arch Neurol Psychiatry, 70 (6):813-819, December 1953. [132] Tator C. H . Spine-spinal cord relationships in spinal cord trauma. Clin Neurosurg, 30: 479-94, 1983. [133] Tator C. H. Biology of neurological recovery and functional restoration after spinal cord injury. Neurosurgery, 42(4):696-707; discussion 707-8, April 1998. [134] Tator C. H. Epidemiology and General Characteristics of the Spinal Cord-Injured Patient. In Contemporary Management of Spinal Cord Injury: from Impact to Rehabilitation, Neurosurgical Topics, chapter 3, pages 15-19. American Association of Neurological Surgeons, Park Ridge, 111., 2000. [135] Tator C. H . and Fehlings M . G. Review of the secondary injury theory of acute spinal cord trauma with emphasis on vascular mechanisms. J Neurosurg, 75(l):15-26, July 1991. [136] Tator C. H. and Koyanagi I. Vascular mechanisms in the pathophysiology of human spinal cord injury. J Neurosurg, 86(3):483-92, March 1997. [137] Tator C. H., Duncan E. G., Edmonds V. E., Lapczak L. I., and Andrews D. F. Comparison of surgical and conservative management in 208 patients with acute spinal cord injury. Can J Neurol Sci, 14(l):60-9, February 1987. [138] Tator C. H., Duncan E. G., Edmonds V. E., Lapczak L. I., and Andrews D. F. Neurological recovery, mortality and length of stay after acute spinal cord injury associated with changes in management. Paraplegia, 33(5):254-262, May 1995. [139] Tator C. H., Fehlings M . G., Thorpe K., and Taylor W. Current use and timing of spinal surgery for management of acute spinal surgery for management of acute spinal cord injury in North America: results of a retrospective multicenter study. J Neurosurg, 91(1 Suppl):12-8, July 1999. [140] Thienprasit P., Bantli H., Bloedel J. R., and Chou S. N . Effect of delayed local cooling on experimental spinal cord injury. J Neurosurg, 42(2):150-4, February 1975. [141] Transfeldt E. E., White D., Bradford D. S., and Roche B. Delayed anterior decompression in patients with spinal cord and cauda equina injuries of the thoracolumbar spine. Spine, 15(9):953-7, September 1990. [142] Turnbull I. M . Chapter 5. Blood supply of the spinal cord: normal and pathological considerations. Clinical Neurosurgery, 20:56-84, 1973. [143] Vaccaro A. R., Daugherty R. J., Sheehan T. P., Dante S. J., Cotler J. M., Balderston R. A., Herbison G. J., and Northrup B. E. Neurologic outcome of early versus late surgery for cervical spinal cord injury. Spine, 22(22):2609-13, November 1997. [144] Vaccaro A. R., Nachwalter R. S., Klein G. R., Sewards J. M., Albert T. J., and Garfin S. R. The significance of thoracolumbar spinal canal size in spinal cord injury patients. Spine, 26(4):371-6, February 2001. 41  Chapter 1.  Introduction  [145] Wagner J., F. C. and Chehrazi B. Early decompression and neurological outcome in acute cervical spinal cord injuries. J Neurosurg, 56(5):699-705, May 1982. [146] Wagner J., F. C , VanGilder J. C , and Dohrmann G. J. Pathological changes from acute to chronic in experimental spinal cord trauma. J Neurosurg, 48(l):92-8, January 1978. [147] Waters R. L., Adkins R. FL, Yakura J. S., and Sie I. Effect of surgery on motor recovery following traumatic spinal cord injury. Spinal Cord, 34(4):188-92, April 1996. [148] Weinshel S. S., Maiman D. J., Baek P., and Scales L. Neurologic recovery in quadriplegia following operative treatment. J Spinal Disord, 3(3):244-249, September 1990. [149] Wilberger J. E. Diagnosis and management of spinal cord trauma. J Neurotrauma, 8 Suppl LS21-8; discussion S29-30, July 1991. [150] Wilcox R. K., Boerger T. O., Hall R. M., Barton D. C., Limb D., and Dickson R. A. Measurement of canal occlusion during the thoracolumbar burst fracture process. J Biomech, 35(3).381-384, March 2002. [151] Windle W. Blood Supply of the Spinal Cord. In Windle W., editor, The Spinal Cord and its Reaction to Traumatic Injury, volume 18 of Modern Pharmacology-Toxicology, chapter 4, page 384. Marcel Dekker Inc, New York, 1980. [152] Young W. Spinal cord contusion models. Prog Brain Res, 137:231-55, 2002. [153] Young W. and Bracken M . B. The Second National Acute Spinal Cord Injury Study. J Neurotrauma, 9 Suppl LS397-S405, March 1992.  42  Chapter 2 Model of Residual Compression  2.1  Introduction  This chapter provides a detailed description of the experimental setup and methodology that was used in the creation of the contusion and residual compression SCIs. The mechanical parameters of the contusion injuries for each injury group as well as the blood flow measurements that were obtained are presented in the Results (Section 2.3).  2.2  Methods  Thirty-two male Wistar rats with a mean weight of 314.0 g (sd = 8.6 g) were randomly assigned to one of four groups. Group one (n = 6) served as a sham surgical control. These animals were surgically exposed and mounted in the testing frame, and the spinal cord was touched to the baseline level. The control animals did not receive a SCI but were monitored in a similar manner to the injury groups. Group two (n = 8) received a primary contusion injury at 700 mm/s to a depth of 1 mm to the exposed spinal cord at T9. Group three (n = 7) also received a primary contusion injury and subsequent residual compression of 90% of the contusion depth (0.9 mm). The final.injury group (n = 8) was residual compressed to 40% of the contusion depth (0.4 mm) after receiving a contusion injury. Three additional animals were tested but excluded from the results after confirmation of additional non-specific injury to the spinal cord. Detailed description of the specimens and experimental apparatus can be found in Appendix A.  43  Chapter 2. Model of Residual Compression  2.2.1  Surgical Preparation & Fixation  The animal was weighed and prepared for surgery by administering three successive intraperitoneal injections of urethane (U2500, Sigma-Aldrich, St. Louis, MO) (2 mg/kg total) at 10 minute intervals. This protocol was adopted as a single large dose had been found to cause premature death in some animals. The back and base of the tail were shaved and the animal was placed in a prone position within a stereotactic frame. Traction was applied to the base of the tail. A mid-sagittal incision was made along the back of the animal and the muscle was separated from the spinal column (T9-T11). The spinal process and caudal regions of the T9 and T10 laminae were removed to create two separate exposures of the spinal cord approximately 5 mm apart. The T9 laminectomy was centred in the midline of the spinal cord and sized so that a 2 mm diameter stainless steel dowel pin could freely touch the exposed cord. The smaller T10 laminectomy was elliptical in shape (enlarged laterally) to minimize the distance between the injury and microperfusion measurement location and to ensure that a region of the spinal cord devoid of visible blood vessels was accessible. The vertebral column (T9-T11) was rigidly clamped using a custom built vertebral body clamp to minimize the movement of the animal and to isolate the effect of impact to the spinal cord alone. The design was based on a similar technique previously developed in our laboratory for the creation of cervical SCIs [20]. The clamp secured the vertebrae on the edge of the pedicle just between the lateral process and the posterior surfaces of the rib (Figure 2.1). The laminectomy sites were covered in Gelfoam (Johnson & Johnson Inc, New Brunswick, NJ), and the exposed tissue was packed with cerebrospinal fluid moistened gauze to absorb haemorrhage. The animal was transferred to a temperature controlled container for 30 minutes to allow the bleeding to diminish. All surgeries were performed by the ICORD micro-surgeon, Dr. Jie Liu. 2.2.2  Injury Procedure & Assessment  To facilitate the creation of a spinal cord injury, the animal was transferred to a modified stereotactic frame, and the vertebral body clamp was secured. The skin around the surgical  44  Chapter 2. Model of Residual Compression  Figure 2.1: The vertebral body clamp secures the vertebrae on the edge of the pedicle just between the lateral process and the posterior surfaces of the rib. exposure was elevated by sutures at the four corners of the clamp to form a small trough. Artificial cerebrospinal fluid was administered through a 26 gauge biopsy needle into the confines of the cavity to prevent the accumulation of blood. The temperature of the cerebrospinal fluid was maintained at 37°C by a heating coil, as measured by a digital multimeter (Fluke 179, Fluke Electronics Canada, Mississauga, ON) and thermocouple submersed in the fluid. Excess fluid was removed by suction from a vacuum pump. Core body temperature was regulated at 37°C (±1°C) by means of a heating plate and rectal probe connected to a temperature control unit (TCAT-2LV, Physitemp Instruments, Clifton, NJ). Heart rate and oxygen saturation were continuously monitored from the base of the tail using a veterinary pulse oximeter (8600V, Nonin Medical, Plymouth, MN). Blood flow microperfusion was measured with a single channel LD system (PF 3, Perimed AB, Stockholm, Sweden). The LD probe (Probe 306, Perimed AB, Stockholm, Sweden) was positioned in the T10 laminectomy over a region of the spinal cord devoid of visible blood vessels. A custom probe holder connected to a horizontally aligned electrode manipulator (Model 960, David Kopf Instruments, Turjunga, California) allowed precise placement of the probe just above the surface of the spinal cord. Baseline blood flow measurements were taken for a 10 minute period pre-injury, and the probe was left in place during the injury sequence. Continuous blood flow microperfusion measurements were recorded while the animal remained in the test frame.  45  Chapter 2. Model of Residual Compression  Figure 2.2: Experimental set-up for the creation of SCIs. The spinal cord injury was induced using an electromagnetic linear actuator (TestBench ELF, EnduraTec, Minnetonka, MN) mounted to a custom built support frame [19, 20]. The vertical height of the actuator was controlled using a linear ball screw actuator (SuperSlide 2DB, Danaher Motion, Radford, VA) for establishment of the baseline level. An xy-table was used to make fine adjustments in the position of the animal. Force measurements were made using a 22.2 N precision miniature load cell (Model 31, Honeywell Sensotec, Columbus, OH). Acceleration was measured using a 490 m/s (50 g) (Model 355B03, PCB Piezotronics, Depew, NY) 2  accelerometer. Displacement was measured using the linear variable differential transformer (LVDT) integrated into the actuator (MHR250 10FT, Measurement Specialties Inc, Hampton, VA). A 2 mm diameter stainless steel impact tip was incrementally lowered in 0.05 mm (0.002 in) steps until the surface of the cord was contacted as measured by a slight increase in the force trace with a 100 Hz digital filter. The injury baseline level was established by slightly compressing the cord with two additional increments of the impactor tip (0.1 mm). The 100 Hz filter was removed from the force trace.  16  Chapter 2. Model of Residual Compression  The injury sequence was initiated by raising the impact tip 2 mm above the baseline level for 10 seconds. Raising the actuator from the baseline level was also the starting point for the three hour analysis period of the surgical control animals. For the injury groups, raising the impact tip allowed the generation of a higher velocity injury and allowed the spinal cord to fully relax from the minor baseline compression. The contusion injury was produced at 700 mm/s to a depth of 1 mm past the baseline level, followed by immediate retraction of the impactor to a level 4 mm above the surface of the spinal cord. Subsequent residual compression for the 40% and 90% injury groups were created by controlling the retraction of the impact tip to the desired level. Typical displacement-time, velocity-time and force-time curves for each injury group can be found in Section 2.3.1. Residual compression was maintained for 60 minutes following the contusion injury, at which point the actuator was raised 4 mm above the spinal cord. Half of the animals in each group were removed from the test frame after 90 minutes following the contusion injury. The surgical exposure was carefully covered in cerebrospinal fluid soaked gauze, and the animal was placed in a temperature controlled container. The remaining animals in each group were monitored for the entire 180 minutes of the experiment in the testing frame. All animals were euthanized 3 hours following the primary contusion injury (baseline touch for controls) by intracardial perfusion with phosphate buffered saline followed by 4% paraformaldehyde. Detailed histological procedures can be found in Chapter 3. The use of animals for this experiment was approved by the University of British Columbia Animal Care Committee, as adhering to the regulations of the Canadian Council on Animal Care (CCAC). 2.2.3  Statistical Analysis  The baseline preload force and mechanical parameters of the contusion injury (displacement, force and velocity) were assessed using an analysis of variance (ANOVA) to evaluate if significant differences existed across the test groups. Comparisons between the specific groups were made using the Student-Newman-Keuls (SNK) post-hoc test. A nonparametric ANOVA (KruskalWallis) was used to assess differences between the injury groups for the load relaxation force, LD microperfusion, pulse oximeter heart rate and haemoglobin oxygen saturation due to the 47  Chapter 2. Model of Residual Compression  considerable variability of these measurements and the small, unequal specimen numbers in each group. The change in LD microperfusion following decompression for the 40% and 90% residual compression groups was analyzed using a Wilcoxon matched pairs test, by comparing the perfusion signal at 58 and 62 minutes. For all tests, a p-value less than 0.05 was considered to be significant. The statistical package Statistica 7.1 (Statsoft Inc., Tulsa, OK) was used for all statistical analyzes. Key tables of the statistical results can be found in Appendix C.  2.3 Results 2.3.1  Mechanical Parameters  Prior to injury, the force (<0.01 N) on the spinal cord following the establishment of the 0.1 mm pre-injury baseline level was similar for all test groups and control (p>0.5). Upon initial dimpling of the cord surface, the load was observed to be approximately 0.015N but rapidly dissipated to the recorded preload level. The loading pattern for each injury group was designed so that identical primary contusion injuries would be generated with variation in the subsequent level of residual compression (Figures 2.3(a)-(c)). The peak displacement, force and velocity of the contusion in the 40% residual and 90% residual injury groups are summarized in Table 2.1. Peak displacement for the 90% residual group was statistically greater than the pure contusion and the 40% residual groups (p<0.002). However, the practical importance of this finding is inconsequential, as the difference only amounts to 2% (0.02 mm) of the overall displacement and arises from the extreme reproducibility of the testing system (Coefficient of variation <0.5% for all groups). The peak velocity (p=0.57) of all groups was reached at the point of spinal cord impact and was not different between the groups (denoted by red line on Figures 2.3(a)-(c)). The peak force (p=0.2) was also similar for all groups. The load relaxation following the primary contusion injury for 40% and 90% residual compression was characterized by plotting the median times to reach 75%, 50%, 40%, 30%, 20%, 10% and 5% of the peak force (Figure 2.4). The force curve for 40% residual compression resulted in a residual force of less than 5%, 28 msec after the primary contusion injury and was not 48  8  t  CD •1  IN-  E E  -  E E  1  ~ 1  10  10.005  10.01  10.015 Time (s)  n t> E  \\ i  Cord Surface  7.  ?E  \ •  \  1  Cord Surfa«e  10.02  10.025  10.03  •i  10.005  10.01  10.015 Time (s)  10.02  10.025  \\ :^  o  ^  V  o  • 1, 10  '  10.03  10  Cord Surface  K " : \ iV  10.005  ' :  10.01  ]  •-  50V. RC  10.015 Time (s)  10.02  10.025  *  9  -0.5  I"  1  -1.5  io to  10.005  10.01  10.015 Time (s)  10.02  10.005  10.01  10.015 Time (s)  10.02  10.025  10.03  10  10.005  10.01  10.015 Time (s)  10.02  10.025  10.03  10  10.005  10.01  10.015 Time (s)  10.02  10.025  10.03  10.025  10.03  250 0 ;-250 -500  10.01  10.015 Time (s)  10.02  10.025  10.03  -750 10  • Contact Cord Surface - Peak Displacement  (a) C o n t u s i o n Injury  Contact Cord Surface PeakDsplacement  (b) C o n t u s i o n & 40% R e s i d u a l Compression Injury  Contact Cord Surface Peak Displacement  (c) C o n t u s i o n & 90% R e s i d u a l Compression Injury  Figure 2.3: Typical displacement, force and velocity time curves for each injury group.  c  10.03  0  3 J  Si  Chapter  2.  Model of Residual  Compression  Injury parameters of the primary contusion injury for each group [Mean (st. dev.)]. Superscript * denotes statistical difference from the other groups (p<0.002). Table 2.1:  Group  Peak Displacement (mm)  Peak Force (N)  Peak Velocity (mm/s)  Contusion  -1.01 (0.004)  -1.59 (0.26)  -708.3 (6.3)  40% Residual  -1.01 (0.005)  -1.60 (0.22)  -708.3 (7.7)  90% Residual  -1.03 (0.002)*  -1.41 (0.19)  -711.9 (8.13)  statistically different from the pure contusion injury. The relaxation curve for 90% residual compression was statistically different from pure contusion and 40% residual compression for values less than <75% of the peak force (p<0.006). The load relaxation occurred rapidly in the 90% residual compression group, reaching less than 5% of the peak force within 11.8 sec (range: 7.4-114). Load Relaxation Post-Contusion Injury: [Median (25th-75th)J  100 Time (s)  2.4- Characterization of the load relaxation for 40% and 90% residual compression by plotting median times to reach 75%, 50%, 40%, 30%, 20%, 10% and 5% of the peak force. Unloading curve of pure contusion is included for relative comparison. Figure  Chapter 2. Model of Residual Compression  2.3.2  Laser Doppler &; Pulse Oximeter  LD flowmetry is incapable of measuring in absolute units. Therefore, the blood flow microperfusion measurements were evaluated relative to the median baseline levels recorded (minimum 10 minute interval) prior to the the primary contusion injury. For statistical comparison the median relative perfusion signals for each group were calculated over 2 minute intervals and evaluated at 2, 30, 58, 62 and 90 minutes (Figure 2.5). Laser Doppler Microperfusion:  2.6  n  [Median] * Control - © - Contusion 40% Residual - B - 90% Residual  80  100  180  Time (min)  Figure 2.5: Laser Doppler (LD) microperfusion measured from the dorsolateral surface of the exposed spinal cord [Median values]. The median relative perfusion values were calculated over two minute intervals. Statistical comparisons were made at 2, 30, 58, 62, and 90 minutes post-injury (vertical magenta lines).  Pure contusion resulted in an initial reduction of the dorsolateral blood flow that increased throughout the experiment but was not statistically different from the surgical control (p=1.0). 40% and 90% residual compression resulted in immediately elevated blood flow that steadily increased over 60 minutes, as compared to pure contusion and the surgical control, but varied considerably across individual animals. 40% residual compression was statistically greater than  51  Chapter 2. Model of Residual Compression  control at 2 (p=0.046) and 58 minutes (p=0.021), marginally greater at 30 minutes (p<0.053) and statistically greater than pure contusion at all three time points (p<0.021). 90% residual compression was not statistically greater than the surgical control (p>0.1) at any analysis points prior to decompression or pure contusion (p>0.075). Decompression at 60 minutes resulted in an immediate reduction of the blood flow within 2 minutes for both 40% (p=0.01i) and 90% (p=0.028) residual compression (compared to blood flow at 58 minutes post-contusion) to levels similar to contusion and the surgical control (p>0.26). LD microperfusion was monitored in a few specimens from each group (n<4) for the entire three hour duration of the investigation to evaluate the trend of the blood flow for an extended period of time. However, due to the initial variability and small number of subjects, a statistical comparison was not performed. Boxplots comparing each group at 2, 30, 58, 62 and 90 minutes can be found in Figures 2.6(a)-(f). Comparison of Laser Doppler Microperfusion: 2 Minutes  Control  • Median J • 25%-75% I Min-Max  40% Residual Contusion  90% Residual  (a) Laser doppler microperfusion at 2 minutes. Figure 2.6: Comparison of laser doppler microperfusion at 2, 30, 58, 62 and 90 minutes postcontusion. • and if - denotes statistical difference between common markers p<0.05. • denotes marginal difference between common markers p=0.053.  52  Chapter 2. Model of Residual  Compression  Comparison of Laser Doppler Microperfusion: 30 Minutes  Control  • Median • 25%-75% I Min-Max  40% Residual Contusion  90% Residual  (c) Laser doppler microperfusion at 30 minutes. Comparison of Laser Doppler Microperfusion: 58 Minutes  Control  • Median • 25%-75% I Min-Max  10% Residual Contusion  90% Residual  (d) Laser doppler microperfusion at 58 minutes.  Figure 2.6: Comparison of laser doppler microperfusion at 2, 30, 58, 62 and 90 minutes postcontusion - continued. • and -A" - denotes statistical difference between common markers p<0.05. © - denotes marginal difference between common markers p=0.053.  53  Chapter 2. Model of Residual Compression  Comparison of Laser.Doppler Microperfusion: 62 Minutes 3.0 2 8 2.6  2.4 CO  2.2 2.0  1.6 1.4 1.2 1.0  T  0.8  T • Median • 25%-75% X Min-Max  0.6 Control  40% Residual Contusion  90% Residual  (e) Laser doppler microperfusion at 62 minutes. Comparison of Laser Doppler Microperfusion: 80 Minutes  Control  • Median • 25%-75% X Min-Max  40% Residual Contusion  80% Residual  (f) Laser doppler microperfusion at 90 minutes (ie. 30 minutes after decompression).  Figure 2.6: Comparison of laser doppler microperfusion at 2, 30, 58, 62 and 90 minutes postcontusion - continued. • and + - denotes statistical difference between common markers p<0.05. • - denotes marginal difference between common markers p=0.053.  54  Chapter 2.  Model of Residual Compression  Heart rate and percent haemoglobin saturation were measured from the base of the tail to assess the stability of the animals throughout the LD measurement period (Figure 2.7. The median heart rates (two minute intervals)(p>0.1) as well as the oxygen saturation (p>0.5) were similar for all groups throughout the evaluation period. PulseOx Heart Rate & Haemoglobin Saturation: [Median] 500  180  80 100 Time (min) 100r cO 2 3  to  90s 80^ 70-  O CO O  60-  GJ  50 0  £ I  L  I  20  40  60  Control - G - Contusion - 0 - 40% Residual - B - 90% Residual  1  80 100 Time (min)  120  140  160  180  Figure 2.7: Median heart rate and haemoglobin saturation values measured from the base of the tail with a pulse oximeter. The median relative perfusion values were calculated over two minute intervals. Statistical differences were calculated at 2, 30, 58, 62, and 90 minutes post-injury (vertical magenta lines).  2.4  Discussion  A somewhat controversial clinical practice is the surgical decompression of the persistently compressed spinal cord. While operative treatment has been associated with a reduced mortality rate, decreased length of hospitalization, earlier movement and rehabilitation and improved bladder and pain patterns [8, 45, 49, 54, 62, 64], there is no conclusive evidence that clinical decompression is beneficial to the neurological outcome of SCI patients presented with resid-  55  Chapter 2. Model of Residual Compression  ual compression of the spinal cord [30]. For this reason many experimental models have been developed to evaluate the effects of maintained compression of the spinal cord. Tarlov et al. conducted a number of classical experiments evaluating the compression of the canine spinal cord with an inflatable balloon with the hopes of determining the time limits for recovery following compression [59-61]. They concluded that the extent of functional recovery was dependent upon the rate, duration and magnitude of the compression. Since this initial work, many other investigators have explored the effects of spinal cord compression using balloon catheters [41-43, 52, 59, 60, 63], static weights [1, 22], trans-laminar screws [37], hydraulic pistons [12-14], modified aneurysm clips [4, 27, 40, 53], and canal spacers [24]. Their conclusions have been consistently similar: the neurological deficit is determined by the magnitude and duration of the compression. The findings associated with residual compression of the spinal cord have primarily been derived from experiments utilizing slow maintained compression that does not replicate the initial high rate contusion injury. Only one published study to date has evaluated the effects of spinal cord compression following an initial contusion injury of the cord [24]. Dimar et al. effectively replicated both contusion and compression components of the injury. A moderate contusion injury was created using the NYU impactor, and subsequent compression was achieved by the placement of epidural spacers within the canal. They also compared the effects of spacer placement without an initial contusion injury and found that compression alone resulted in no neurological injury until canal narrowing reached 50%, whereas trauma that included an initial contusion followed by the placement of spacers that occupied 20% or 35% of the canal area produced greater trauma than a purely contusive injury. However, the contusion and spacer placement occurred at different levels of the spinal cord, and the manipulation required for the placement of the spacer could have resulted in unwanted aggravation of the disrupted tissue following the initial injury.  56  Chapter 2. Model of Residual Compression  2.4.1  Injury Mechanism  A range of devices have been developed for the creation of reproducible contusion injuries that demonstrate graded histopathological and function loss in relation to the injury severity. Many of these models have utilized a weight drop mechanism based on the first such descriptions by Allen [2, 3]. The most common weight-drop model is the NYU impactor [5, 34] that was standardized to minimize the variability in the resulting injuries [31, 40, 44]. Graded lesions are created by dropping a 10 g rod from incremental heights of 6.25, 12.5, 25 and 50 mm, and the impact velocity and cord compression are measured. The NYU model is not ideal for the creation of a residual compression injury, since the impact rod is not directly controlled and thereby precludes the direct creation of subsequent spinal cord compression. Controlled mechanisms for the creation of spinal cord contusion injuries are the OSU [50, 57, 58] and more recently developed IH [55] impactors that allow the direct measurement of the impact force and displacement. The OSU model utilizes an electromagnetic shaker to create a contusion to a predefined distance below the baseline reference level. The baseline level is established by compressing the cord until a load of 1.5-3 kdynes (ie. 0.03-0.015 N) is achieved. The currently available commercial OSU model does not allow direct feedback control of the electromagnetic shaker, thus residual compression levels less than the maximum contusion depth cannot be created. Unlike the OSU model, the IH impactor uses direct force feedback to control the impact. While this mechanism does benefit from the fact that an initial baseline does not have to be established, the use of force control prevents the establishment of fixed displacement residual compression following the initial injury. In theory a constant force could be used to define the level of residual compression in a similar manner to the clip compression model [25, 26, 28, 29, 53]. Applying a constant force to the spinal cord does not allow the reduction of the stress within the tissue due to viscoelastic relaxation. This will most likely produce significantly more damage than a fixed displacement, as the structurally intact tissue will have to support increasingly more stress as the surrounding cells become damaged and incapable of supporting the applied load. In the current study a custom-mounted, feedback-controlled linear actuator, recently developed 57  Chapter 2. Model of Residual Compression  by our laboratory [19-21], was used to create a primary contusion injury and immediate residual compression to a predefined level of the maximum contusion depth. This device operates in a similar fashion to the OSU model and incorporates feedback control similar to the IH machine, thereby allowing the reproducible creation of both the contusion and subsequent residual compression injuries. Additionally, once the parameters of the injury sequence (ie. PID settings, waveform) are determined the actuator can reproduce the desired injury to an extremely high degree without the need for recalibration on subsequent test days. The reproducibility was demonstrated in this study by the low coefficient of variation for the peak displacements of the contusion injuries (Coefficient of variation <0.5% for all groups).  2.4.2 Mechanical Parameters of Injury The creation of reproducible SCIs was based on the establishment of a consistent baseline reference level. Unlike the OSU mechanism that utilizes an initial small force (1.5-3 kdynes) [6, 9, 10, 57], the current study used a fixed displacement. Initial contact with the spinal cord was established using a combination of visual observation through a stereoscopic microscope and a shift in the force trace while lowering the actuator. From our earlier experience, a force-defined baseline is dependent upon the loading rate of the spinal cord, as load relaxation occurs during the compression of the spinal cord. This may result in a far more severe injury than intended if the load is allowed to relax multiple times during the baseline procedure. For this reason we switched to a combination of force and displacement to establish a reproducible baseline level. The displacement was fixed to 0.1 mm past the initial contact with the cord, and load relaxation was observed between the establishment of the baseline level and initiation of the injury protocol (5-10 sec). The mechanical parameters of the contusion injuries from the current study were compared to the values obtained by others during thoracic SCIs in the rat (Figure 2.8). The peak displacement used for the current injuries was chosen to create a moderate SCI based on the results of previous investigations employing functional evaluations [6, 10, 55]. The peak force resulting from the current 1 mm injuries was similar to the values obtained by Scheff et al. for their moderate injury group but were smaller than the values obtained using the OSU device. The peak 58  Chapter 2. Model of Residual Compression  velocity of our injuries was significantly greater (700 mm/s vs. 430 mm/s [9], 170 mm/s [6] or 122 mm/s [55]) than those obtained using electromagnetic devices, although greater velocities have been achieved using other devices (pneumatic actuator [39]). Our parameters were also comparable to those obtained using the NYU impactor for the creation of a moderate injury (12.5 g: 1.6 mm, 484 mm/s), although the force was not measured during the weight-drop injury. While the parameters of the current study were not identical to previous results, it is most likely that a moderate injury would be achieved, as severe injuries require a greater displacement (NYU>1.5 mm, OSU>1.5 mm and IH>1.2 mm) of the spinal cord and result in the generation of higher loads (OSU and IH>2 N).  -  O * * x + • *  !  Behrmann, DL 1992 Bresnahan, J C 1987 Bresnahan, J C 1991 Scheff, S W 2003 Somerson, S K 1987 Sparrey, C J 2004 Current Study  i  + o;  +  c  2.5  •  * x  •  •  * •  0.5  o11 0  ii  iI  I  0.2  0.4  0.6  iI i I 0.8 1 1.2 Displacement (mm)  I  iI  I  I  1.4  1.6  1.8  2  Comparison of force and displacement from the current study to the results of other studies employing a contusion injury of the rat thoracic spinal cord. Figure 2.8:  The two levels of residual compression where chosen to represent the maximum and minimum depths that would most likely be of clinical significance for a contusion injury of 1 mm. The  59  Chapter 2. Model of Residual Compression  lower level of compression, 40% (0.4 mm), was chosen as this corresponds to 16% of the dorsalventral depth and 12% of the total area of the rat spinal canal [T9 canal elliptical area: 6.9 mm , 2  2.5 mm (D-V) x 3.5 mm (L-R); impact tip rectangular area: 0.8 mm , 0.4 mm (depth) x 2 mm 2  (width)]. The maximum occlusion depth of 90% (0.9 mm) was chosen as a value closer to 1 mm is difficult to control (overshoot, oscillation of actuator) when the contusion velocity is so high, and during the clinical injury, natural vibration and recoil of the occluding particles will occur. While these values are arbitrary in nature, there is no consensus to the recommended minimum threshold to perform clinical decompression (25%, 50%, complete injury, any compression) [8]. The rat thoracic spinal cord encompasses a proportionally greater volume of the spinal canal than the human, so a value smaller than the minimum 25% reported in literature was chosen. Cadaveric studies of burst fracture occlusion have demonstrated that the final occlusion is less than the transient level [16, 18, 66]. For this reason, the maximum residual compression was restricted to a level less than or equal to the peak contusion depth.  2.4.3 Load Relaxation Following the contusion injury, the load relaxation of the spinal cord was monitored for the residual compression groups. -40% residual compression was almost indiscernible from the pure contusion injury, and 90% residual compression resulted in relaxation to <5% of the peak load within 11.8 sec. The monitoring of the load was complicated by the high velocity of the injury creating vibration in the impact tip and by the fact that the spinal cord was not completely constrained within the rigidly supported vertebral column. To confirm that the observed relaxation was not the result of flexion in the support frame, a stiffness test and confirmation of the injury protocols using surrogate silicone tubing were performed. The stiffness of the support frame was 29.8 N/mm resulting in a deflection of less than 0.06 mm (6%) for loads equivalent to the peak contusion (Figure 2.9). Tests of the three injury protocols using surrogate silicone tubing loosely supported in the test frame resulted in lower peak contusion forces than the actual SCIs but the subsequent residual compression of the spinal cord produced considerably more force (Figure 2.10). Interface pressures following dynamic cord compression have also been monitored by Carlson 60  Chapter 2. Model of Residual Compression Frame Stiffness  I -5  i  Frame Stiffness = 29.8 N/mm :  A  ;  z  I Loading R e g i on|  o li.  -2  i  -1  0  -0.02  -0.04  -0.06  -0.08  -0.1  -0.12  -0.14  -0.16  -0.18  Displacement (mm) |~« L o a d C y c l e # 1  * L o a d C y c l e # 2 — F r a m e Stiffness]  Figure 2.9: Frame  stiffness is 29.8 N/mm, corresponding to flexion of less than 6% of the peak contusion displacement. et al. with a hydraulic piston at 0.17 mm/min to an average depth of 2 mm [11, 12]. They observed that the maximum interface pressure (30.5 kPa; 1.17 N for 7 mm diameter piston) occurred at the end of the dynamic compression, dropped by 50% within 4 minutes and reached a constant level within 90 minutes (<12%). Their observations are in stark contrast to our own, in which the load dissipated within a few seconds. We believe this observation highlights the importance of producing the initial contusion at a high velocity that will ultimately result in acute disruption of the cellular structure and the inability of the spinal cord to support further load. While our peak load was of a similar magnitude (1.6 N vs 1.17 N), the interface pressure calculated for the 2 mm diameter impact tip was 16 times greater (509.3 kPa vs 30.5 kPa). 2.4.4  Blood Flow  The blood flow through the spinal cord was monitored using a LDflowmetrysystem. The great appeal of LD flowmetry for the purpose of blood flow measurements is its simple operation. Following the connection of the fibre optic probe to the LD control unit, measurements can  61  Chapter 2. Model of Residual Compression  Evaluation of injury Protocols with Surroqale Silicone Tubing  0.2  0  -02  _-o< z  I 6  -0.8  * -12 10  10  10  90% Residua! 10  Time (s)  Figure 2.10: Tests of injury protocols using surrogate silicone tubing. Peak surrogate contusion force was less than average contusion force for all injury groups (1.54 N). Residual load was considerably more for 40% and 90% residual compression.  be performed by placing the probe-end over any vascularized tissue. Although the operation is fairly straightforward, the interpretation of the measured values can be rather ambiguous if consideration is not given to the set-up and operational environment. LD Flowmetry functions by comparing the relative motion between the probe and any reflective substance within its measurement volume. For vascularized tissue this results in an estimation of the red blood cell perfusion. However, if undesired movement between the LD probe and tissue sample occurs, motion artefact noise will be created [46, 47]. For this reason considerable effort was extended to the rigid stabilization of the spinal column and LD probe (See Section 4.1.3 for discussion of the vertebral clamp used for stabilization of the spinal column). During the contusion injury, vibration of the testing system (actuator & stereotactic frame) inevitably occurred. The vibration was minimized by a damping table upon which the entire testing system was located, and the rigid designs of the vertebral clamp and LD probe holder prevented relative motion between them. Due to the non-homogeneous distribution of vessels within the spinal cord, movement of the LD probe across the surface may result in substantial shifts in the perfusion signal [32, 46]. Although the spinal cord was displaced below the impact  62  Chapter 2. Model of Residual Compression  tip, the location of the LD probe 5 mm caudal prevented substantial motion between the probe and the baseline location. A custom probe holder and horizontally aligned electrode manipulator were used for precise placement of the LD probe over the exposed spinal cord. In this study the probe was positioned just above the surface of the cord in an area devoid of visible vessels. The small distance between the probe and tissue surface prevented accidental compression of the vasculature without affecting the recorded signal. The theoretical justification (linear relationship between changes in LD signal and blood flow) of LDflowmetryis only supported at the microcirculation level [47]. LD measurements including larger vessels are more likely to result in multiple doppler shifts of the back-scattered light, resulting in an accumulation of the respective velocities and a non-linear response to changes in the flow [46]. During setup, multiple acceptable measurement locations were compared in each specimen, and the location furthest from visible vessels producing a relatively low perfusion signal was chosen. This process was performed to ensure that large vessels were unlikely to be located below the surface of the spinal cord within the measurement volume (1 mm ) of the LD probe. Prior to testing the LD system was calibrated using the 3  routines specified by the manufacturer to facilitate a comparison of the baseline levels across specimens. This ensured that the selected measurement location produced a perfusion signal that was typical of the microvascular level, as large underlying vessels may be present at multiple locations. The accumulation of blood below the LD probe can cause aberration of the backscatter light [47]. This would adversely affect the measurement results by being perceived as a reduction in the perfusion signal (similar to the perception of the eyes after the application of sunglasses). Since the measurements were taken within a surgical exposed area over 1.5-3 hours, the accumulation of blood below the LD probe was most likely to occur. This was prevented by flooding the surgical cavity with artificial cerebrospinal fluid. The placement of the LD probe occurred prior to the release of the cerebrospinal fluid, and no observable shift in the perfusion signal was detected as the cavity was filled. Pilot tests also demonstrated that mixing a small amount of blood with the cerebrospinal fluid did not produce an artificial perfusion signal when the  63  Chapter 2. Model of Residual Compression  probe was immersed. The use of cerebrospinal fluid complicated the tests due to the fact that the specific heat of water is four times that of air. The cerebrospinal fluid pool could result in significant cooling of the surrounding tissue and inadvertently have neuroprotective effects. Hypothermia has been suggested as a clinically feasible neuroprotective method for many years, and directly applied cooling of the rat spinal cord has been shown to be beneficial in preventing injury secondary to ischemic cellular damage [23] even at modest systemic hypothermia levels (32-33°C) [67]. The blood flow reduction resulting from systemic hypothermia has also been shown to exceed the response following moderate spinal cord compression [65], highlighting the importance of maintaining normal body temperature throughout the experiment. Small heating coils were used to maintain the temperature of the cerebrospinal fluid at 37°C, as monitored by a thermocouple located on the periphery of the T10 laminectomy. Care was taken to prevent overheating of the fluid, as hyperthermia has conversely been shown to reduce functional outcome and increase the percentage of tissue damage following SCI [68]. The measurement volume of LD flowmetry varies with the optical properties of the tissue but is generally considered-to be 1 mm -[32, -46,- 47]. Due to uncertainty in the measurement 3  volume, absolute blood flow measurements cannot be achieved, but relative linear changes with blood flow have been demonstrated [32]. LD has been strongly correlated with microsphere injection [32, 47] and simultaneous hydrogen clearance recordings [65] in the brain and spinal cord. However, the direct comparison of LD flowmetry with other techniques is often difficult and questionable due to differences in the measurement volume [46]. The perfusion values we obtained following residual compression were unexpected. Prior to the study we had postulated that residual compression would result in ischemia at the injury epicentre and diminished blood flow in the adjacent tissue. Our results showed the opposite effect, with increasing perfusion distant to the injury epicentre until the compression was relieved. Interspecimen variability was very high for both residual compression groups, thus the need for a nonparametric statistical comparison. While many of the residual compression specimens demonstrated drastically increased perfusion, there were some in each group that were not 64  Chapter 2. Model of Residual Compression  vastly different from the initial baseline levels (40% residual compression: n = 2; 90% residual compression: n = 3). However, following decompression the large reduction in perfusion signals for both the 40% and 90% residual compression groups suggests that the compression caused the increase in blood flow and it did not arise from an unmeasured source. Heart rate and haemoglobin blood saturation, measured from the base of the tail, were not different for any group and there was no observable effect on these measurements following contusion, residual compression or decompression. We observed no statistical difference between pure contusion and the surgical control group during the measurement of the dorsal-lateral blood flow. A few of the contusion specimens experienced an initial hypoperfusion that was not observed in the control, but values returned to normal within 20 minutes. Prior blood flow measurements following compression of the spinal cord demonstrated a different response than our own observations. Using the hydrogen clearance technique, the spinal cord blood flow following clip compression was shown to decrease at the injury epicentre and 1 cm caudal with increasing severity of the injury [29, 35]. A similar decrease was shown 7 mm rostral and caudal to the injury epicentre using LD flowmetry following epidural clip compression [51]. Carlson et al. [12, 13, 15] also found the spinal cord blood flow to be significantly less following compression with a hydraulic actuator within a 1 cm section. The blood flow increased as the pressure decreased due to viscoelastic relaxation of the tissue and returned to normal within 5 minutes of decompression. Although these measurements are in stark contrast to our own, we believe there are numerous differences that may account for the discrepancy including: severity of injury; choice of anaesthesia; load relaxation influenced by method of injury; and region of measurement. While the exact severities of our injuries are unknown, we believe that the pure contusion injury would result in a moderate functional recovery. Many of the blood flow measurement studies have utilized what is considered to be severe SCIs that most likely resulted in a more significant initial disruption of the local vascular structure. Urethane anaesthesia has also been shown to have considerably less effect on the systemic blood pressure and arterial dilation than halothane, which was used in the studies by Carlson et al. and Ohashi et al., and pentobarbital [38, 48].  65  Chapter 2. Model of Residual Compression  As stated earlier, the load relaxation we observed was extremely fast for both 40 and 90% residual compression. Without a significant force compressing the spinal cord it is unlikely that the blood flow through the main arteries was completely arrested, although the compression could have been great enough to have restricted the local flow. Therefore, to compensate for the restricted flow at the injury epicentre the blood flow rate was increased (through autoregulatory processes that are known to be present within the uninjured spinal cord) as we observed one level caudal to the injury. Comparison of blood flows is also complicated by the fact that different measurement techniques were used in different regions and volumes of spinal cord tissue. By using a LD system positioned over the dorsolateral spinal cord it is most likely that we observed the blood flow primarily in the white matter. Regional differences in flow have been reported by numerous observers with stark contrasts been the gray and white matter following contusion. The blood flow in the gray matter has consistently been shown to be significantly reduced following trauma [7, 17, 33, 36] while the white matter blood flow is only slightly decreased [33] or even higher than the uninjured control [7, 36]. Following severe contusion injury of the cat spinal cord, the dorsolateral blood was not different from pre-injury levels up to Ihr following trauma [17, 56]. The general consensus seems to be that SCI is accompanied by a dramatic decrease in the blood flow to the gray matter (leading to the development of an ischemic environment) and considerable variability in the blood flow to the white matter. Our results suggest that a contusion injury generated at 700 mm/s to a depth of 1 mm in the rat does not significantly influence the dorsal white matter blood flow distant to the injury epicentre, while subsequent residual compression may result in considerably elevated flow rates. The exact cause of the elevated flow following residual compression is unknown, as well as why some specimens do not demonstrate this behaviour. The variation in the vascular structure of the spinal cord most likely had an effect; similar results may have been obtained if simultaneous measurements were taken 5 mm rostral to the epicentre. Please see Chapter 4 for conclusions and recommended future directions relevant to this portion of the study.  66  Chapter 2. Model of Residual Compression  2.5  References  [1] Aki T. and Toya S. Experimental study on changes of the spinal-evoked potential and circulatory dynamics following spinal cord compression and decompression. Spine, 9(8): 800-9, November-December 1984. [2] Allen A. R. Surgery of experimental lesion of spinal cord equivalent to crush injury of fracture dislocation of spinal column. A preliminary report. J. Am. Med. Assoc., 57:878880, 1911. [3] Allen A. R. Remarks on the Histopathological changes in the Spinal Cord due to Impact. An Experimental Study. J. Nerv. Ment. Dis., 41:141-147, 1914. [4] Anthes D. L., Theriault E., and Tator C. H. Characterization of axonal ultrastructural pathology following experimental spinal cord compression injury. Brain Res, 702(1-2):1—16, 1995. [5] Basso D. M . , Beattie M . S., and Bresnahan J. C. Graded histological and locomotor outcomes after spinal cord contusion using the NYU weight-drop device versus transection. Exp Neurol, 139(2):244-56, June 1996. [6] Behrmann D. L., Bresnahan J. C , Beattie M. S., and Shah B. R. Spinal cord injury produced by consistent mechanical displacement of the cord in rats: behavioral and histologic analysis. J Neurotrauma, 9(3):197-217, Fall 1992. [7] Bingham W. G., Goldman H., Friedman S. J., Murphy S., Yashon D., and Hunt W. E. Blood flow in normal and injured monkey spinal cord. J Neurosurg, 43(2):162-171, August 1975. [8] Boerger T. 0., Limb D., and Dickson R. A. Does 'canal clearance' affect neurological outcome after thoracolumbar burst fractures? J Bone Joint Surg Br, 82(5):629-35, July 2000. [9] Bresnahan J. C , Beattie M . S., Todd r., F. D., and Noyes D. H. A behavioral and anatomical analysis of spinal cord injury produced by a feedback-controlled impaction device. Exp Neurol, 95(3):548-70, March 1987. [10] Bresnahan J. C , Beattie M . S., Stokes B. T., and Conway K. M . Three-dimensional computer-assisted analysis of graded contusion lesions in the spinal cord of the rat. J Neurotrauma, 8(2):91-101, Summer 1991. [11] Carlson G. D., Minato Y., Okada A., Gorden C. D., Warden K. E., Barbeau J. M., Biro C. L., Bahnuik E., Bohlman H. H., and Lamanna J. C. Early time-dependent decompression for spinal cord injury: vascular mechanisms of recovery. J Neurotrauma, 14(12): 951-62, December 1997. [12] Carlson G. D., Warden K. E., Barbeau J. M., Bahniuk E., Kutina-Nelson K. L., Biro C. L., Bohlman H. H., and LaManna J. C. Viscoelastic relaxation and regional blood flow response to spinal cord compression and decompression. Spine, 22(12):1285-91, June 1997.  67  Chapter 2. Model of Residual Compression  [13] Carlson G. D., Gorden C. D., Nakazowa S., Wada E., Warden K., and LaManna J. C. Perfusion-limited recovery of evoked potential function after spinal cord injury. Spine, 25 (10):1218-26, May 2000. [14] Carlson G. D., Gorden C. D., Nakazawa S., Wada E., Smith J. S., and LaManna J. C. Sustained spinal cord compression: part II: effect of methylprednisolone on regional blood flow and recovery of somatosensory evoked potentials. J Bone Joint Surg Am, 85(A1): 95-101, January 2003. [15] Carlson G. D., Gorden C. D., Oliff H. S., Pillai J. J., and LaManna J. C. Sustained spinal cord compression: part I: time-dependent effect on long-term pathophysiology. J Bone Joint Surg Am, 85(Al):86-94, January 2003. [16] Carter J. W., Mirza S. K., Tencer A. F., and Ching R. P. Canal geometry changes associated with axial compressive cervical spine fracture. Spine, 25(l):46-54, January 2000. [17] Cawthon D. F., Senter H. J., and Stewart W. B. Comparison of hydrogen clearance and 14C-antipyrine autoradiography in the measurement of spinal cord blood flow after severe impact injury. J Neurosurg, 52(6):801-807, June 1980. [18] Chang D. G., Tencer A. F., Ching R. P., Treece B., Senft D., and Anderson P. A. Geometric changes in the cervical spinal canal during impact. Spine, 19(8):973-980, April 1994. [19] Choo A., Liu J., Lam C., Dvorak M., Tetzlaff W., and Oxland T. Comparing Mechanisms of Spinal Cord Injury - Contusion, Dislocation & Distraction. San Diego, 2004. Presented at the 34th Annual Meeting of the Society for Neuroscience and 22nd Annual National Neurotrauma Society Symposium. [20] Choo A., Liu J., Lam C , Dvorak M., Tetzlaff W., and Oxland T. New Device for Producing Different Mechanisms of Spinal Cord Injury. The American Society of Mechanical Engineers, June 2005. Summer Bioengineering Conference. [21] Choo A., Liu J., Lam C , Dvorak M., Tetzlaff W., and Oxland T. Primary and Secondary Damage in Three Mechanisms of Spinal Cord Injury: Contusion, Dislocation & Distraction. Presented at the 23rd Annual National Neurotrauma Society Symposium, 2005. Poster: 306. [22] Croft T. J., Brodkey J. S., and Nulsen F. E. Reversible spinal cord trauma: a model for electrical monitoring of spinal cord function. J Neurosurg, 36(4):402-406, April 1972. [23] Dimar J. R., Shields C. B., Zhang Y. P., Burke D. A., Raque G. H., and Glassman S. D. The role of directly applied hypothermia in spinal cord injury. Spine, 25(18):2294-2302, September 2000. [24] Dimar n., J. R., Glassman S. D., Raque G. H., Zhang Y . P., and Shields C. B. The influence of spinal canal narrowing and timing of decompression on neurologic recovery after spinal cord contusion in a rat model. Spine, 24(16):1623-33, August 1999. [25] Dolan E. J. and Tator C. H. A new method for testing the force of clips for aneurysms or experimental spinal cord compression. J Neurosurg, 51(2):229-233, August 1979. 68  Chapter 2. Model of Residual Compression  [26] Dolan E. J., Tator C. H., and Endrenyi L. The value of decompression for acute experimental spinal cord compression injury. J Neurosurg, 53(6):749-55, December 1980. [27] Fehlings M. G. and Nashmi R. A new model of acute compressive spinal cord injury in vitro. J N euro sci Methods, 71(2):215-24, March 1997. [28] Fehlings M. G. and Tator C. H. The relationships among the severity of spinal cord injury, residual neurological function, axon counts, and counts of retrogradely labeled neurons after experimental spinal cord injury. Exp Neurol, 132(2):220-8, April 1995. [29] Fehlings M. G., Tator C. H., and Linden R. D. The relationships among the severity of spinal cord injury, motor and somatosensory evoked potentials and spinal cord blood flow. Electroencephalogr Clin Neurophysiol, 74(4):241-59, July-August 1989. [30] Fehlings M. G., Sekhon L. H., and Tator C. The role and timing of decompression in acute spinal cord injury: what do we know? What should we do? Spine, 26(24 Suppl):S101-10, December 2001. [31] Ford R. W. A reproducible spinal cord injury model in the cat. J Neurosurg, 59(2):268-75, August 1983. [32] Frerichs K. U. and Feuerstein G. Z. Laser-Doppler flowmetry. A review of its application for measuring cerebral and spinal cord blood flow. Mol Chem Neuropathol, 12(l):55-70, January 1990. [33] Griffiths I. R. Spinal cord blood flow after acute experimental cord injury in dogs. J Neurol Sci, 27(2):247-259, February 1976. [34] Gruner J. A. A monitored contusion model of spinal cord injury in the rat. J 9(2):123-6, 1992.  Neurotrauma,  [35] Guha A., Tator C. H., and Rochon J. Spinal cord blood flow and systemic blood pressure after experimental spinal cord injury in rats. Stroke, 20(3):372-7, March 1989. [36] Hall E. D. and Wolf D. L. Post-traumatic spinal cord ischemia: relationship to injury severity and physiological parameters. Cent Nerv Syst Trauma, 4(l):15-25, 1987.  [37] Hashimoto T. and Fukuda N . New spinal cord injury model produced by spinal cord compression in the rat. J Pharmacol Methods, 23(3):203-12, May 1990. [38] Holobotovskyy V. V., Arnolda L. F., and McKitrick D. J. Effect of anaesthetic and rat strain on heart rate responses to simulated haemorrhage. Acta Physiol Scand, 180(1): 29-38, January 2004. [39] Kearney P. A., Ridella S. A., Viano D. C., and Anderson T. E. Interaction of contact velocity and cord compression in determining the severity of spinal cord injury. J Neurotrauma, 5(3):187-208, 1988. [40] Khan M. and Griebel R. Acute spinal cord injury in the rat: comparison of three experimental techniques. Can J Neurol Sci, 10(3):161-5, August 1983.  69  Chapter 2. Model of Residual Compression  [41] Kobrine A. I., Evans D. E., and Rizzoli H. Correlation of spinal cord blood flow and function in experimental compression. Surg Neurol, 10(l):54-9, July 1978. [42] Kobrine A. I., Evans D. E., and Rizzoli Ft. V. Correlation of spinal cord blood flow, sensory evoked response, and spinal cord function in subacute experimental spinal cord compression. Adv Neurol, 20:389-94, 1978. [43] Kobrine A. I., Evans D. E., and Rizzoli Ft. V. Experimental acute balloon compression of the spinal cord. Factors affecting disappearance and return of the spinal evoked response. J Neurosurg, 51(6):841-5, December 1979. [44] Koozekanani S. H., Vise W. M., Hashemi R. M., and McGhee R. B. Possible mechanisms for observed pathophysiological variability in experimental spinal cord injury by the method of Allen. J Neurosurg, 44(4):429-34, April 1976. [45] Krengel I., W. F., Anderson P. A., and Henley M . B. Early stabilization and decompression for incomplete paraplegia due to a thoracic-level spinal cord injury. Spine, 18(14):2080-7, October 1993. [46] Leahy M . J., de Mul F. F., Nilsson G. E., and Maniewski R. Principles and practice of the laser-doppler perfusion technique. Technol Health Care, 7(2-3):143-62, 1999. [47] Lindsberg P. J., O'Neill J. T., Paakkari I. A., Hallenbeck J. M., and Feuerstein G. Validation of laser-Doppler flowmetry in measurement of spinal cord blood flow. Am J Physiol, 257(2 Pt 2):H674-80, August 1989. [48] Longnecker D. E. and Harris P. D. Microcirculatory actions of general anesthetics. Fed Proc, 39(5):1580-1583, April 1980. [49] National Institute of Neurological Disorders and Stroke [NINDS] . Spinal Cord Injury: Emerging Concepts. Technical report, National Institutes of Health, October 1 1996. URL http: //www. ninds . nih. gov/news_and_events/proceedings/sci_report. htm. [50] Noyes D. H. Electromechanical impactor for producing experimental spinal cord injury in animals. Med Biol Eng Comput, 25(3):335-40, May 1987.  [51] Ohashi T., Morimoto T., Kawata K., Yamada T., and Sakaki T. Correlation between spinal cord blood flow and arterial diameter following acute spinal cord injury in rats. Acta Neurochir (Wien), 138(3):322-9, 1996. [52] Oro J. J., Gibbs S. R., and Haghighi S. S. Balloon device for experimental graded spinal cord compression in the rat. J Spinal Disord, 12(3):257-61, June 1999. [53] Rivlin A. S. and Tator C. H. Effect of duration of acute spinal cord compression in a new acute cord injury model in the rat. Surg Neurol, 10(l):38-43, July 1978. [54] Rosenfeld J. F., Vaccaro A. R., Albert T. J., Klein G. R., and Cotler J. M . The benefits of early decompression in cervical spinal cord injury. Am J Orthop, 27(l):23-8, January 1998.  70  Chapter 2. Model of Residual Compression  [55] Scheff S. W., Rabchevsky A. G., Fugaccia I., Main J. A., and Lumpp J., J. E. Experimental modeling of spinal cord injury: characterization of a force-defined injury device. J Neurotrauma, 20(2):179-93, February 2003. [56] Senter H. J. and Venes J. L. Loss of autoregulation and posttraumatic ischemia following experimental spinal cord trauma. J Neurosurg, 50(2):198-206, February 1979. [57] Somerson S. K. and Stokes B. T. Functional analysis of an electromechanical spinal cord injury device. Exp Neurol, 96(l):82-96, April 1987. [58] Stokes B. T., Noyes D. H., and Behrmann D. L. An electromechanical spinal injury technique with dynamic sensitivity. J Neurotrauma, 9(3):187-95, Fall 1992. [59] Tarlov I. M. Spinal cord compression studies. III. Time limits for recovery after gradual compression in dogs. AMA Arch Neurol Psychiatry, 71(5):588-97, May 1954. [60] Tarlov I. M. and Klinger H. Spinal cord compression studies. II. Time limits for recovery after acute compression in dogs. AMA Arch Neurol Psychiatry, 71(3):271-90, March 1954. [61] Tarlov I. M., Klinger H., and Vitale S. Spinal cord compression studies. I. Experimental techniques to produce acute and gradual compression. AMA Arch Neurol Psychiatry, 70 (6):813-819, December 1953. [62] Tator C. H., Duncan E. G., Edmonds V. E., Lapczak L. I., and Andrews D. F. Comparison of surgical and conservative management in 208 patients with acute spinal cord injury. Can J Neurol Sci, 14(l):60-9, February 1987. [63] Thienprasit P., Bantli H., Bloedel J. R., and Chou S. N. Effect of delayed local cooling on experimental spinal cord injury. J Neurosurg, 42(2):150-4, February 1975. [64] Transfeldt E. E., White D., Bradford D. S., and Roche B. Delayed anterior decompression in patients with spinal cord and cauda equina injuries of the thoracolumbar spine. Spine, 15(9):953-7, September 1990. [65] Westergren H., Farooque M., Olsson Y., and Holtz A. Spinal cord blood flow changes following systemic hypothermia and spinal cord compression injury: an experimental study in the rat using Laser-Dopplerflowmetry.Spinal Cord, 39(2):74-84, February 2001. [66] Wilcox R. K., Boerger T. O., Hall R. M., Barton D. C., Limb D., and Dickson R. A. Measurement of canal occlusion during the thoracolumbar burst fracture process. J Biomech, 35(3):381-384, March 2002. [67] Yu C. G., Jimenez O., Marcillo A. E., Weider B., Bangerter K., Dietrich W. D., Castro S., and Yezierski R. P. Beneficial effects of modest systemic hypothermia on locomotor function and histopathological damage following contusion-induced spinal cord injury in rats. J Neurosurg, 93(1 Suppl):85-93, July 2000. [68] Yu C. G., Jagid J., Ruenes G., Dietrich W. D., Marcillo A. E., and Yezierski R. P. Detrimental effects of systemic hyperthermia on locomotor function and histopathological outcome after traumatic spinal cord injury in the rat. Neurosurgery, 49(l):152-8, July 2001.  71  Chapter 3 Histopathological Changes Following Residual Compression  3.1  Introduction  This chapter is a continuation of the experimental study outlined in Chapter 2, focusing on the histopathological analysis that was performed to evaluate the extent of the spinal cord lesion following the contusion and residual compression injuries. The influence of 40% and 90% residual compression on the magnitude of the spinal cord lesion was evaluated by comparing the volume and extent of haemorrhage. The extent of gray and white matter damage was compared using immunohistochemical techniques. Details of the histopathological results are presented with a discussion of their importance to the understanding of residual compression injuries.  3.2  Methods  Thirty-two male Wistar rats with a mean weight of 314.0g (sd = 8.6g) were randomly assigned to one of three injury groups or sham surgical control. The injury groups received a primary contusion injury (n = 8), contusion plus 40% residual compression (n = 8), or contusion plus 90% residual compression (n = 7). The residual compression was decompressed 60 minutes after the contusion injury. The spinal cord of the control animals (n = 6) was touched (baseline level = 0.05 mm) and monitored in a similar manner to the injury groups. A brief description of the method to produce the spinal cord injuries will ensue; detailed descriptions can be found in Chapter 2.  72  Chapter 3. Histopathological Changes Following Residual Compression  3.2.1  Injury Procedure  Three successive intraperitoneal injections of urethane (2 mg/kg total) (2500, Sigma-Aldrich, St. Louis, MO) at 10 minute intervals was given to each animal. Laminectomies were performed at T9 and T10 to expose the spinal cord. The vertebral column (T9-T11) was rigidly clamped between the lateral processes and the posterior surface of the ribs with a custom vertebral body clamp. The animal was secured in a modified stereotactic frame, and a contusion injury (1 mm, 700 mm/s) was performed using an electromagnetic linear actuator (TestBench ELF, EnduraTec, Minnetonka, MN). Residual compression of the spinal cord was created by stopping the retraction of the impounder tip at the desired level (0.9 mm or 0.4 mm below baseline). Decompression was performed 60 minutes following the contusion injury by raising the impact tip. A heating plate was used to regulate the core body temperature at 37°C (±1°C), and heart rate and haemoglobin oxygen saturation were continuously monitored. Artificial cerebrospinal fluid was released into the surgical cavity and maintained at 37°C by a heating coil. Laser Doppler microperfusion was continuously monitored from the T10 laminectomy site. 3.2.2  Tissue Harvesting & Preparation  All animals were euthanized 3 hours following the primary contusion injury (baseline touch for controls) by intracardial perfusion with 150ml of phosphate buffered saline (PBS) followed by 300ml of 4% paraformaldehyde (PF) in PBS. Following perfusion, the spinal cord was harvested, and the dura mater and nerve roots were removed. The spinal cord was post-fixed in 4% PF for 24 hours and sequentially transferred to 12%, 18% and 24% sucrose solutions for 24 hours at 4°C. Eighteen millimetre sections of the cord, centered at the injury epicentre were rolled in optimal cutting temperature medium (OCT Tissue-Tek, Sakura Finetek, Torrance, CA), mounted on blotting paper and flash frozen on dry ice. The tissue was stored at -80°C until sectioned. Each spinal cord was cut using a cryostat along the parasagittal plane (Leica CM3050, Leica Microsystems (Canada) Inc, Richmond Hill, On) into  20mTi  thick sections. The sections were  mounted in parallel onto six slides (Superfrost Plus, Fisher Scientific, Houston, TX) such that  73  Chapter 3. Histopathological Changes Following Residual Compression  each slide had sections 12G>m apart, representing each region of the cord. The slides were stored at -80°C until processed for histology or immunohistochemitry. 3.2.3  Histology &; Immunohistochemistry  Tissue sections were histologically processed using Hematoxylin and Eosin (H&E) staining performed by Wax-it Histology Services (Wax-it Histology Services, Vancouver, BC) for the evaluation of haemorrhage. Immunohistochemistry was performed for the assessment of the following proteins (primary antibodies): /3-amyloid precursor protein (/3-APP), a protein carried by the fast anterograde axoplasmic transport system (rabbit polyclonal, CT-695, Zymed-Invitrogen Canada Inc, Burlington, ON); and neuronal nuclei (NeuN), a neuron-specific nuclear protein found within the neurons of the spinal cord (mouse monoclonal, MAB-377, Chemicon International Inc, Temecula, CA). Fluoro-Jade (FJ), an anionicfluorochromethat has been shown to selectively stain degenerating neurons (FJ, Histo-Chem Inc, Jefferson, AR), was also used in conjunction with NeuN. For visualization of the previously listed primary antibodies, the followingfluorescentsecondary antibodies were used: Cy3-conjugated donkey anti-mouse (Jackson ImmunoResearch Laboratories Inc., West Grove, PA); and Alexa Fluoro 488 goat anti-rabbit (Molecular Probe-Invitrogen Canada Inc, Burlington, ON). The general immunohistochemical procedures used for visualization of the primary antibodies is outlined below. Slides were defrosted for 15 minutes on a 37°C heating pad and rehydrated for 5 minutes in 0.01M PBS. A 10% block of either normal donkey serum or normal goat serum diluted in PBS plus 0.1% Triton X-100 (PBS+Tx) was applied to prevent non-specific antibody binding. After 30 minutes at room temperature, excess serum was removed, and the slides were incubated with primary antibodies for B-APP (1:200) or NeuN (1:100) diluted in PBS+Tx overnight in a humidity chamber. The following day, the slides were washed (3x5 minutes) in PBS and then incubated in a 1:200 PBS-TX dilution of the corresponding secondary antibody for 90 minutes. Following the final wash ( 3 x 5 minutes) in PBS, the 3-APP slides were coverslipped using mounting medium (Gel Mount, Sigma-Aldrich, St. Louis, MO) and sealed with  74  Chapter 3. Histopathological Changes Following Residual Compression  nail varnish, and the NeuN slides were prepared for FJ staining. For incorporation of FJ with the NeuN slides, the following modified procedure from the recommendations of Tom Hallam [14] was used. A 0.01% FJ stock solution (FJ-s) was prepared by mixing lOmg of the FJ dye powder in 100ml of distilled water (dH^O) and stored at 4°C until needed. The 0.0001% FJ working solution (FJ-W) was created immediately before use by combining 1000^1 of FJ-S in 99ml of 0.1% acetic acid diluted with dH^O. Slides were rinsed for 5 minutes in dH^O following the previously described molecular labelling procedure and immersed in a coplin jar containing the FJ-W for 60 minutes at 4°C. The slides were dried for 20 minutes on a 37°C heating pad after being washed (3x1 minute) in dH^O. The sections were cleared in xylene ( 2 x 2 minutes) cover-slipped using Entellan (Electron Microscopy Sciences, Hatfield, PA) and allowed to dry overnight. To protect the FJ from photo-bleaching during the staining procedure, each step was carefully performed with aluminum foil covering to minimize the exposure to ambient light. 3.2.4  Image Analysis & Quantification  Digital images were obtained using a Zeiss Axioplan II microscope (Carl Zeiss Canada Ltd, North York, ON) fitted with a motorized microscope stage (Scan 100x100 - 1 mm, Mdrzhduser Wetzlar, Wetzlar-Steindorf, Germany) and Retiga Exi digital ccd camera (RET-EXi-F-M012, Q-Imaging, Burnaby, BC). Using a customized script for the Northern Eclipse software (Northern Eclipse 6.0, Empix Imaging Inc, Mississauga, ON), consecutive views were automatically captured and merged into a complete image of the tissue sections. This procedure enabled the tissue sections to be analyzed as a single image, preventing the duplicate counting of positively stained areas and ensuring the use of uniform settings. All images were analyzed using ImagePro Plus software (Version 4.5, Media Cybernetics Inc, Silver Spring, MD) and a LCD drawing tablet (Cintiq 21 UX, Wacom Technology Corp, Vancouver, WA) for improved efficiency. H&E stained tissue was quantified for the macroscopic spread of haemorrhage from the injury epicentre and total volume within the spinal cord. 24-bit, red/blue/green (RBG) images were captured using a five times objective lens (50x total magnification) and RGB filter (RGB slider,  75  Chapter 3. Histopathological Changes Following Residual Compression  Q-Imaging, Burnaby, BC). The area of haemorrhage was determined by thresholding an area of interest selected with the LCD tablet for the exclusion of normal tissue structures. Area measurements from seven sections (240/um apart) centred about the midsagittal plane were combined using the frustum of a cone equation (Equation 3.1) to calculate a total volume for the entire spinal cord. The maximum rostral and caudal extent of haemorrhage from the injury epicentre was measured from a section within ±360//m of the midsagittal plane. V=±h[B  1  + B + (Bi*B )*} 2  (3.1)  2  Where: V Bi B2 h  = = = =  volume of haemorrhage between two segments [mm ] area of haemorrhage of section [mm ] area of haemorrhage of next section [mm ] distance between the two sections [mm] 3  2  2  NeuN and FJ labelled slides were used for quantification of the rostral-caudal extent of gray matter damage from the injury epicentre (Figure 3.1(a)). Images were captured in 18-bit greyscale using a 5 times objective (50x total magnification) and FITC (FJ) or Cy3 (NeuN) fluorescent filter. Neuronal damage was regarded as the occurrence of shrunken, diminished NeuN immunofluorescence or FJ positively labelled neurons. The extent of neuronal damage was measured to the steepest transition between normal appearing and diminished NeuN staining (Figure 3.1(b)) and the furthest occurrence of positively labelled FJ neurons (Figure 3.1(c)). Single measurements were taken from sections best representing the midsagittal plane, ventrolateral and dorsolateral gray matter. The maximum extent of gray matter damage was independently selected for both the NeuN and FJ labelled tissue. The injury epicentre for matching sections was selected concurrently to ensure that similar reference points were used. /3-APP was used for quantification of the injury to the white matter tracts of the spinal cord. Grey-scale, 18-bit images were captured using a 5 times objective and FITC fluorescent filter. Rostral and caudal regions of the white matter were selected using the LCD tablet and ImagePro Plus (Figure 3.2(a)). From the areas of interest, the /3-APP positive tissue was thresholded out and a subsequent image identifying the surrounding background (grey), normal white matter 76  (a) Merged fluorescent image showing combined NeuN (red cell bodies) and F J  assessed by the steepest transition  as assessed by the presence of F J  (bright green cell bodies) labelling for  between normal and diminished NeuN  positively labelled neurons.  neuronal damage assessment.  immunofluorescence.  Figure 3.1:  cords.  (b) Location of maximal extent of damage as (c) Location of maximal extent of damage  Methods for selecting the maximum rostral and caudal extent of gray matter damage for NeuN and FJ labelled spinal  Chapter 3. Histopathological Changes Following Residual Compression  (black) and 0-APP (white) was produced (Figure 3.2). The density of 0-APP (white) within the total white matter (black + white) over 0.5 mm rostral-caudal windows (254 pixels) was performed using a custom image processing script in Matlab (Version 6.1, The MathWorks Inc, Natick, MA). The average density of d-APP accumulation was calculated from left and right sections representing the ventral, dorsal, ventrolateral and lateral white matter (Figure 3.2(c)).  (a)  (c)  (b)  Figure 3.2: Analysis technique for quantifying the density of 0-APP in regions of the white matter: (a) area of interest (red outline) selected with LCD tablet; (b) 0.5 mm analysis windows (green outline), density was determined from the ratio of white {0-APP) vs black (normal tissue) + white pixels; (c) analysis regions of the white matter.  3.2.5  Variation in Techniques  Care was taken to ensure that every specimen was processed in an identical manner, however some variation occurred for the initial six specimens. This arose from the accidental mounting of  78  Chapter 3. Histopathological Changes Following Residual Compression  the tissue sections onto uncharged slides (Superfrost, Fisher Scientific, Houston, TX), resulting in the separation of the tissue from the slides during the initial staining attempt. The uncharged slides used for H&E staining were coated in celloidin prior to staining. Celloidin prevented the sections from lifting off of the slides but did not affect the staining of the tissue as the membrane is permeable to small molecules such as H&E. To prevent the loss of tissue, the sections used for immunohistochemistry were transferred to charged slides using the following routine: 1. Defrost slides for 5 minutes at 37°C; 2. Pipette  dH20  onto uncharged slides until tissue sections are covered by a thin layer;  3. Wait 5 minutes to allow tissue to loosen on uncharged slides; 4. Drain excess  dH20  using the edge of a paper towel;  5. Carefully lower a charged slide (Superfrost Plus) over the tissue sections, ensuring the slides remain parallel throughout; 6. Rotate the slides 180° so that the uncharged slide is on top and slide it in the direction of the long-axis of the sections until the slides are separated; 7. Pipette 100 mL d H 2 0 onto charged slide containing transferred sections and correct placement of tissue using a fine tipped brush; 8. Drain excess  dH20  and dry slides for 30 minutes at 37°C;  9. Follow previously outlined immunohistochemistry techniques. The six specimens included two controls, two pure contusion and two 40% residual compression specimen. These techniques had minor effects on the image analysis and quantification of the results. Due to the transfer process the immunohistochemistry sections had their dorsalventral orientation reversed and the celloidin coating on the H&E slides required a slightly longer exposure during the image capture process. Only the results of the /3-APP analysis were affected, as the tissue on these slides also underwent an additional 30 minute wash in 4% PF following the 30 minute drying stage. The 4% PF wash was used to ensure that the sections  79  Chapter 3. Histopathological Changes Following Residual Compression  were bonded to the slides, but it resulted in denaturing of the targeted proteins and no visible staining of the tissue. Subsequent slides did not undergo this additional step after the above process was shown to be adequate to bond the tissue to the slides. Additional slides were not stained for /3-APP due to loss of tissue during the initial staining attempts on uncharged slides. However 6 specimen for each of the affected injury groups (contusion and 40% residual compression) were still available for image analysis and statistical comparison. 3.2.6  Statistical Analysis  Statistical analysis was performed using Statistica software (Version 7.1, Statsoft Inc., Tulsa, OK). An analysis of variance (ANOVA) was used to evaluate the total haemorrhage volume between injury groups. A two-way ANOVA with the rostral and caudal distances as a repeated measure was used for analysis of the rostral and caudal extents of haemorrhage. For comparison of the gray and white matter damage at multiple regions of the spinal cord, a three-way ANOVA was used with two factors (injury group and anatomical region) and one repeated measure (gray matter : rostral and caudal distance from the injury epicentre; white matter: rostral and caudal distances from the injury epicentre ± 1 , ±2, ±3, ±4 and ±5 mm). All post-hoc analyzes were made using the Student-Newman-Keuls (SNK) test. A p-value less than 0.05 was considered to be significant. Key tables of the statistical results can be found in Appendix C.  3.3 3.3.1  Results Haemorrhage  Residual compression had a varying effect upon the total haemorrhage volume (Figure 3.3. 90% residual compression produced relatively little haemorrhage in some animals (n=2) (Figure 3.3(d)), while resulting in extensive haemorrhage in others that extended beyond the 18 mm sections of harvested spinal cord (Figures 3.3(c)). Due to the variability the total haemorrhage volume was not found to be statistically different between injury groups (p=0.28) (Figure 3.4). Overall, the extent of haemorrhage between the different injury groups was statistically different (p=0.016). The extent of haemorrhage following 90% residual compression was greater than  80  Chapter 3. Histopathological Changes Following Residual Compression  (a) Typical contusion.  (b) Typical 40% residual compression.  (c) Maximum haemorrhage observed for 90% residual compression (total visible haemorrhage length before sectioning intact cord 25 mm).  (d) Minimum haemorrhage observed for 90% residual compression of the spinal cord. Figure 3.3: Comparison of haemorrhage near the injury epicentre (midsagittal plane) of the different injury groups. Haemorrhage was similar for pure contusion (a) and 40% residual compression (B). 90% residual compression resulted in considerable variability in the extent and volume of haemorrhage (c) & (d). pure contusion (p=0.02) and 40% residual compression (p=0.014). 40% residual compression resulted in a similar extent of haemorrhage compared to pure contusion (p=0.78) (Figure 3.5). There was no difference between the direction of the (rostral or caudal) extent of haemorrhage for all groups (p=0.53) (main effect) or between injury groups (p=0.99) (interaction).  81  Chapter 3. Histopathological Changes Following Residual Compression  Total Haemorrhage Volumes  € 1.5  h-  40% Residual  90% Residual  Figure 3.4'- Total volume of haemorrhage for each injury group. Rostral-Caudal Extent of Haemorrhage from Injury Epicentre 10 8  • ^ •  H & E - Rostral H&E - Caudal Mean (st. dev.)  2 6  w o *  4  1° _ CO  o  -4 _ K D  -8  -10  \  Contusion  40% Residual  90% Residual  Figure 3.5: Rostral and caudal extent of haemorrhage for pure contusion, 40% and 90% residual compression. The overall extent (rostral -f caudal) of haemorrhage was greater for 90% residual compression (p<0.02). However, the individual rostral and caudal extent was not statistically different across all injury groups. 82  Chapter 3. Histopathological Changes Following Residual Compression  3.3.2  G r a y M a t t e r Injury  NeuN and FJ were used for quantification of the rostral-caudal extent of gray matter damage in the medial, ventrolateral and dorsolateral gray matter (Figures 3.6, 3.7 & 3.8). The main effect comparing the different injury groups was statistically different for both fluorescent labels (p<0.001). The extent, regardless of direction (rostral or caudal) or anatomical region (medial, ventrolateral or dorsolateral), of diminished NeuN immunofluorescence was greater for 90% residual compression than pure contusion (p=0.0001) or 40% residual compression (p=0.0001). The overall extent of positively labelled FJ neurons was also greater for 90% residual compression than either contusion (p=0.0001) or 40% residual compression (p=0.0001). NeuN & Fluoro-Jade: Medial Gray Matter [Mean (st. dev.)] 8 -  6  tn  NeuN - Rostral NeuN - Caudal FJ - Rostral FJ-Caudal  o  % 2  1° -2{ --A T3  TO.«  Contusion  40% Residual  90% Residual  Figure 3.6: Extent of gray matter damage within the medial gray matter by assessing the diminished immunofluorescence of NeuN and presence of FJ positive neurons.  The main effects of the gray matter damage were not statistically different between the three regions (NeuN: p=0.07, FJ: p=0.12) or in the rostral/caudal directions (NeuN: p=0.49, FJ: p=0.09) for NeuN or FJ. However, the midsagittal plane tended to extend further (rostral + caudal) for all injury groups (Figures 3.9 & 3.10). No statistical differences existed for any of the interactions (injury + region, injury + direction, etc.) (p>0.2).  83  Chapter 3. Histopathological Changes Following Residual Compression  NeuN & Fluoro-Jade: Ventrolateral Gray Matter [Mean (st. dev.)] 8  S6  NeuN-Rostral I NeuN - Caudal FJ - Rostral FJ-Caudal  00 o  2 1 °  «T  4  TJ ZS  O  D  -8  Contusion  40% Residual  90% Residual  Figure 3.7: Extent of gray matter damage within the ventral gray matter by assessing the diminished immunofluorescence of NeuN and presence of FJ positive neurons.  NeuN & Fluoro-Jade: Dorsolateral Gray Matter [Mean (st. dev.)] 8, NeuN - Rostral  2 6 00 o  NeuN - Caudal FJ - Rostral FJ-Caudal  * 4 2  1° -2 "O  O  °  Contusion  40% Residual  90% Residual  Figure 3.8: Extent of gray matter damage within the dorsal gray matter by assessing the diminished immunofluorescence of NeuN and presence of FJ positive neurons.  Si  Chapter 3. Histopathological Changes Following Residual Compression  Gray Matter D a m a g e a s s e s s e d by Diminished N e u N Immunofluorescence Mean ± 95% confidence intervals R o s t a l Extent  D.5 0.0  Caudal Extent  i  '  •  Contusion  '  •  1  90% R e s i d u a l  I  ;  f  !  '  i  '  Contusion  40% Residual Midsagittal  90% R e s i d u a l  40% Residual Ventrolateral  Dorsolateral  Figure 3.9: Variation in NeuN rostral and caudal extent of gray matter damage at the midsagittal, ventrolateral and dorsolateral plane. Gray Matter D a m a g e a s s e s s e d by Positively Labelled F J Neurons [Mean ± 95% confidence intervals] Rostral Extent  Contusion  C a u d a l Extent  90% R e s i d u a l  Contusion  40% R e s i d u a l Midsagittal  90% R e s i d u a l 40% Residual  J2£ Ventrolateral  3ii  Dorsolateral  Figure 3.10: Variation in FJ rostral and caudal extent of gray matter damage at the midsagittal, ventrolateral and dorsolateral plane. 85  Chapter 3. Histopathological Changes Following Residual Compression  3.3.3  White Matter Injury  Disruption of the axonal tracts of the spinal cord were quantified by comparing the density of /3-APP every 1 mm (rostral and caudal: 0.5 mm windows) from the injury epicentre in the dorsal, ventral, ventrolateral and lateral regions of the white matter (Figures 3.11-3.14). P-APP Density: Dorsal White Matter 0.25  Caudal  Rostral  Mean rostral and caudal extent of /3-APP density in dorsal white matter tracts (error bars not plotted). Density calculated as percentage of white matter within 0.5 mm rostral-caudal windows. Group means were statistically compared every 1 mm. Figure 3.11:  All main effects were significantly different: type of injury (p=0.004), region of white matter (p<0.001), and distance from injury epicentre (p<0.001). The overall density of /3-APP was similar for pure contusion and 90% residual compression (p=0.5) but less following 40% residual compression (p<0.009) (Figure 3.15). The overall density was less than 3% for contusion and 90% residual compression and only present within 1.5% of the 40% residual compression tissue. Within the different regions of the spinal cord the overall density was higher in the lateral (p<0.001) and ventrolateral (p<0.045) tracts than either the ventral or dorsal tracts (Figures 3.11-3.14). Comparing the accumulation of /3-APP in the rostral/caudal directions from the injury epicentre the overall density was greatest near the periphery (±1 mm) of the impact tip (p<0.0001), accumulating in a bell shaped distribution primarily within 2 mm caudal 86  Chapter 3. Histopathological Changes Following Residual Compression  p-APP Density: Ventral White Matter - O - Contusion 40% Residual - B - 90% Residual 03  ^  0.2  °0.15 O O X  Q_0.05 D_ < CO.  Caudal  (mm)  Rostral  Mean rostral and caudal extent of /?-APP density in ventral white matter tracts (error bars not plotted). Density calculated as percentage of white matter within 0.5 mm rostral-caudal windows. Group means were statistically compared every 1 mm. Figure 3.12:  Figure 3.13: Mean rostral and caudal extent of /3-APP density in ventrolateral white matter tracts (error bars not plotted). Density calculated as percentage of white matter within 0.5 mm rostral-caudal windows. Group means were statistically compared every 1 mm.  87  Chapter 3. Histopathological Changes Following Residual Compression  Mean rostral and caudal extent of /3-APP density in lateral white matter tracts (error bars not plotted). Density calculated as percentage of white matter within 0.5 mm rostral-caudal windows. Group means were statistically compared every 1 mm. Figure 3.14'-  and 3 mm rostral of the injury epicentre (Figure 3.16). The distribution was skewed rostrally, with a greater density accumulating 2 mm (p=0.002) and 3 mm (p=0.02) rostrally than at the equivalent caudal distances. There was no significant interaction between the region within the white matter where the /3-APP accumulated and the type of SCI injury (p=0.4). However, the density tended to be the greatest in the lateral white matter, followed by the ventrolateral tract, dorsal tract and lowest in the ventral tract for all groups. The rostral-caudal distance from the injury epicentre was affected by the type of injury (p<0.001) (Figure 3.17). Pure contusion resulted in a dense accumulation of /3-APP around the injury epicentre within ±1 mm, with mean densities more than double those of the residual compression injuries. The contusion values at the impact tip periphery (±1 mm), were statistically greater than any other location for all groups (p<0.013). For rostral-caudal extents greater than ±1 mm, pure contusion was similar to 40% and 90% residual compression at all corresponding locations (p>0.35) and both residual groups were similar at all corresponding locations 88  Chapter 3. Histopathological Changes Following Residual Compression  0.040  0.035  J  0.030  E & 1  0.025  2  0,020  S  0.015  0010  0.005 Contusion  4 0 % Residual  9 0 % Residual  Figure 3.15: Main effect comparing the overall mean densities of /3-APP for the different injury groups. • and -fc - denotes statistical difference between common markers p< 0.009.  (p>0.61). Comparing similar distances in the rostral and caudal directions, the distributions about the epicentre line were skewed towards the rostral end but symmetrical for all levels of 40% residual compression (p>0.54) and pure contusion (p=0.082). 90% residual compression was also statistically symmetrical in the rostral and caudal directions but appeared slightly more skewed towards the rostral end, with the greatest difference at ± 3 mm (p=0.099). The extent (comparing densities in the rostral-caudal distances for a specific injury group) of the /3-APP densities were narrower for contusion and 40% residual compression than for 90% residual compression (Figure 3.17). The contusion densities within 1 mm caudal and 2 mm rostral were statistically greater than those outside ±3 mm (p<0.027). 40% residual compression also had a narrow band of densities within -1-2 mm that were greater than those beyond ±4 mm (p<0.0036) (Except -1 mm and 4 mm; p=0.061). The extent of axonal disruption was greater following 90% residual compression, with densities within -2-3 mm statistically greater than those outside -4 and 5 mm (p<0.014). The rostral-caudal distance from the injury epicentre also varied within the different regions of the spinal cord (p<0.001) (Figure 3.18). Two distinct profiles were seen in the density distributions of the white matter tracts. Similar dense accumulations of /3-APP around the 89  Chapter 3.  Histopathological Changes Following Residual Compression  Mean p - A P P Densities as a Function of Distance from the Injury Epicentre [Mean ± 95% confidence intervals]  -  5  -  1  -  3  -2 Caudal  -  1 (mm)  1  2  3  4  5  Rostral  Main effect comparing the distribution of /3-APP. The distribution was skewed rostrally, with a greater density accumulating 2 mm (p=0.002) and 3 mm (p=0.02) rostrally than at the equivalent caudal distances, if and X - denotes statistical difference between Figure 3.16:  all unmarked points p<0.0001 and p<0.02.  • and  ~k - denotes statistical difference between  common markers p<0.02 and p<0.002.  injury epicentre (±1 mm) occurred in the lateral and ventrolateral tracts (p>0.62) that were statistically greater than the dorsal and ventral tracts (p<0.012). The overall profile was similar for the dorsal and ventral tracts (p>0.84). A significant three-way interaction between the distance from the epicentre, type of injury and region within the white matter was also present (p<0.001) (Figure 3.19). Contusion was statistically greater at ±1 mm in the lateral tracts than in any other region for all groups (p<0.015). 90% residual compression was also statistically greater at 1 mm and 3 mm in the lateral region than at the majority of greater distances for each group in all regions. All other values were similar.  90  Figure  3.17:  The distribution of /3-APP for pure contusion, 40% and 90% residual compression. X -  greater than all similarily - denotes statistical greater than •  unmarked points p<0.013.  • - denotes statistical difference greater than •  difference greater than • marked points p<0.0036,  marked points p<0.014-  except with superscript *.  •  denotes statistical  difference  marked points p<0.027.  - denotes statistical  •  difference  Figure 3.18: The distribution of /3-APP in the dorsal, difference between all unmarked points p<0.013.  ventral, ventrolateral & lateral white matter tracts. X - denotes statistical  Figure 3.19: Three way interaction between the rostral-caudal distribution of 0-APP for each injury in the dorsal, ventral, ventrolateral & lateral white matter tracts. Contusion was statistically greater at ±1 mm in the lateral tracts than in any other region for all groups. In the lateral tract 90% residual compression was also greater at 1 mm and 3 mm than at many other distances for each group in all regions of the cord.  Chapter 3. Histopathological Changes Following Residual Compression  3.4  Discussion  The development of experimental models of SCI allows the investigation of the injury pathology that results from different injury mechanisms. Trauma to the spinal cord results in immediate primary damage which is followed by a cascade of secondary processes leading to an extension of the injury into the surrounding tissue. The primary damage is a direct result of the mechanical insult leading to the disruption of the cellular and microvascular networks. However, the amount of chronic damage that arises directly from the initial injury as opposed to secondary processes is not known [6]. To understand how different injury mechanisms such as contusion and residual compression influence the injury, it is important to evaluate these changes and their progression from the acute to chronic lesion. The histopathological analysis of this study was limited to the initial few hours following the SCI. This was done to limit the progression of the secondary injury so that the trauma resulting directly from the contusion and subsequent levels of residual compression could be easily ascertained. Regardless of injury severity the pattern of damage following acute SCI is the development of central haemorrhagic necrosis followed by the progression of the injury in a centrifugal and longitudinal direction [4, 5, 7-9, 11, 16, 32]. Lesion volume, length and area of gray and white matter damage at the injury epicentre have all been correlated with the severity of a contusion injury [20]. However, the total volume and overall lesion length are thought to be better predictors of the eventual functional deficit [18, 21], so the current study focused on the effects of 40% and 90% residual compression on the rostral-caudal extent of the damage. The assessment techniques used in this study have not typically been used in other investigations. Many reports of the histopathology of experimental SCI only report qualitative observations as unbiased quantitative values require considerably more time and effort. The current techniques gave simple reproducible values which allowed the overall extent of haemorrhage, gray and white matter damage to be statistically compared between the injury groups. These techniques do not indulge much information about the progression and differences of the injuries but provide a method of determining if one injury results in a smaller or larger lesion than another. 94  Chapter 3. Histopathological Changes Following Residual Compression  3.4.1  Haemorrhage  H&E has been used in the evaluation of haemorrhage and tissue necrosis in the spinal cord in numerous studies [2, 15, 17, 20, 25]. Haemorrhage is one of the earliest indicators of spinal cord trauma and both the area and volume have been strongly correlated with the resulting functional impairment [7]. Other studies have not always found a similar correlation between the area of haemorrhage and the severity of the injury [3, 20] but the amount of haemorrhage has been shown to increase with the magnitude of the trauma [31]. The central cavity that forms in the later stages of SCI pathology occurs in a similar area to where the haemorrhage was located following a weight-drop injury [31]. Haemorrhage within the spinal cord can also be visualized in the human spinal cord using magnetic resonance imaging (MRI). Patients without visible spinal cord haemorrhage had a significantly higher motor scale score (better function) at discharge than those with haemorrhage at early time points [13]. In this study H&E was used to determine the maximum extent and overall volume of haemorrhage within the spinal cord for each injury group. The haemorrhage areas were calculated using a combination of area tracing and colour thresholding. This combined technique was used to ensure that tissue areas obviously devoid of haemorrhage were not inadvertently included in the measurements, and the presence of rather diffuse haemorrhage at the injury epicentre of the residual compression groups necessitated the use of colour thresholding. Otherwise, the selection of the areas containing diffuse haemorrhage would have been extremely tedious or easily overestimated. By limiting the quantification to the overall volume and greatest rostral and caudal extents a simple comparison of the different injuries could be made. However, overall measurements do not take into consideration local variations such as differences in the density (solid versus diffuse accumulation) of the red blood cells at the injury epicentre. The volume and individual rostral and caudal extents were statistically similar for all groups, but the overall extent (rostral + caudal) was greater following 90% residual compression. While these results suggest that residual compression does not considerably increase the damage to the tissue, the diffuse haemorrhage qualitatively observed at the epicentre of the residual compression groups is difficult to capture in a quantitative measurement and most likely had an 95  Chapter 3. Histopathological Changes Following Residual Compression  impact on the results. Following pure contusion the haemorrhage was typically contained near the injury epicentre in dense clumps in the obviously necrotic tissue. The residual compression groups also displayed red blood cells at the epicentre but the overall density appeared to be more dispersed and visible necrotic tissue was sometimes devoid of any observable haemorrhage. This observation is apparent when comparing the volume and extent of the pure contusion and 40% residual compression groups. While both results were not statistically different the volume tended to be less following 40% residual compression but the extent tended to be slightly more. It would be expected that if the extent was greater the overall volume would be greater as well. The 90% residual compression specimens also displayed considerable variability in the amount of visible haemorrhage. While some specimens had similar volumes and extents to the contusion and 40% residual compression animals in others the haemorrhage was considerably greater or considerably less (Figures 3.4 and 3.5). Haemorrhage is one of the earliest observable indicators of damage to the spinal cord and has been correlated with injury severity in a number of studies. Using a similar correlation this study would suggest that the long-term severity of the three injuries would be quite similar. However, the presence of diffuse haemorrhage at the injury epicentre of the residual compression groups most likely led to the underestimation of the associated cellular damage and staining for Nissl substance (cresyl echt violet stain) to indicate the volume of necrotic tissue would possibly give a better indication of the group differences. 3.4.2  Gray Matter Damage  Injury to the gray matter was assessed using the neuronal specific immunofluorescent markers NeuN and FJ. NeuN stains the neuronal cells of the spinal cord and the majority of the regions of the adult brain [19]. Its immunoreactivity is concentrated in the soma and extends a short distance into the dendritic processes. FJ is an anionic tribasicfluorescein(crystalline dye with a bright yellow-greenfluorescence[26]) derivative that has been shown to selectively stain degenerative cerebral neurons [23]. The exact mechanism by which FJ targets neurons is unknown, but the degenerating molecule that it binds with is normally absent in the cell as increased membrane permeability without cellular trauma does not result in the appearance of 96  Chapter 3. Histopathological Changes Following Residual Compression  positively labelled cells [23]. The reduction of NeuN immunoreactivity was used as an indication of neuronal damage and has previously been used as evidence of neuronal loss [29]. The loss of NeuN immunoreactivity may not necessarily indicate the disintegration and disappearance of the targeted neurons as Western blot analysis 24 hours following cerebral trauma does not show a decrease in the intensity of the NeuN protein [30]. It was suggested that the loss of NeuN reactivity resulted from the loss of antigenicity but the region of dramatically reduced staining correlated with increased TUNEL labelling. For this reason we limited our analysis of NeuN to indicate the extent of neuronal damage and used FJ as a positive counterstain. FJ produces bright green positive neurons and may also result in strong background staining unless a potassium permanganate pre-treatment is used [23]. However, the permanganate treatment may eliminate or degrade fluorescent counterstains such as NeuN. Two new versions of FJ (FJ-B and FJ-C) have been developed that exhibit greater signal to background ratios and require significantly reduced permanganate pretreatments [12, 22, 24]. Although, FJ-B stains damaged neurons in the brain in a similar manner to its predecessor a recent study found no evidence of FJ-B staining at numerous time points in the traumatically injured spinal cord [1]. In fact, FJ-B was found to stain astrocytes in the normal and injured spinal cord but not the brain [1]. FJ-C is reported to have the highest affinity and resolution of all the FJ stains, but its use in the spinal cord has not previously been reported. Our attempts to use FJ-C did not successfully label any positive cells in the injured spinal cords although brain test sections that had known ischemic damage produce positive cells for both FJ and FJ-C. The original FJ stain was used in this investigation with a modified staining technique to facilitate the use of NeuN as an immunofluorescent counterstain. The permanganate pretreatment was omitted and a reduced dye concentration and treatment temperature were used. Only 90% residual compression was found to cause an extension of the lesion in comparison to a pure contusion injury; 40% residual compression tended to produce a slightly longer lesion but was not statistically different. No difference was detected between the midsagittal, ventrolateral and dorsolateral sections although the use of parasagittal sections made the localization of the 97  Chapter 3. Histopathological Changes Following Residual Compression  analysis to a specific region of the gray matter along the rostral-caudal length of the cord more complicated. Slight variations in the orientation or the spinal cord during the freezing and cutting processes can result in similar regions of the spinal cord occurring on different slides from the rostral and caudal ends. The overall maximum extent was chosen in each section so it is likely that the lesion extended further in the dorsal region of the ventrolateral and the ventral region of the dorsolateral sections. However, the dorsolateral sections were always located more laterally than the ventrolateral sections, so these measurements give a rough overview of the lesion size throughout the gray matter. The different regions of the spinal cord were not found to be affected by the type of injury, suggesting that distinct populations of neurons are not more likely to be sensitive to residual compression. However, the fairly nonspecific nature of the gray matter measurements used in this study may obscure this type of analysis, and more localized measurements in the ventral and dorsal horns may show differences. As expected, no difference was observed in the rostral or caudal extents of the gray matter damage for either marker. FJ and NeuN are neuron specific markers that target the stationary gray matter cells and do not require the transfer of a protein from an external source to produce a positive result. Therefore, FJ and NeuN are unlikely to show a directional preference as the neurons rostral and caudal to the injury epicentre are both as likely to be affected by the SCI and subsequent residual compression. This also suggests that the blood supply to the gray matter is similar rostral and caudal to the injury epicentre, as a variation between the two directions would most likely result in greater cell death at the side of reduced blood flow. In this study similar rostral-caudal extents were seen for both NeuN and FJ in each group. This observation was expected since the two markers were being used as indicators of the extent of gray matter damage. The fact that similar results we obtained from two independent yet related measurements imparts confidence that we were successfully able to measure the extent of gray matter damage for each injury group.  98  Chapter 3. Histopathological Changes Following Residual Compression  3.4.3  White Matter Damage  APP is a protein found within many organs of the body including the nervous system, but its normal function is presently unknown. APP is transported by the fast mode of the anterograde axonal transport system mediated by the microtubules in the axons [10]. Following axonal trauma this protein quickly pools at the sites of focally impaired transport [27] and has become the most sensitive method of demonstrating axonal injury in the brain [10]. APP accumulation has also been used in the detection of white matter damage of the spinal cord. 0-APP immunoreactivity has been shown to be confined to the swollen portions of traumatically injured axons without targeting the surrounding uninjured white matter [28]. This study utilized an antibody that targets the C-terminus of 0-APP located on the external surface of the transport vesicle, producing a greater affinity of the antibody with reduced background staining in comparison to other APP antibodies [27]. In this study 0-APP was used as an indicator of white matter disruption due to the rapid accumulation of the protein following a SCI and is a less subjective analysis in comparison to other axonal markers. Background staining for the detection of 0-APP using CT-695 is very low and positive staining is very bright in comparison. These properties reduce the chances of including nonspecific tissue during the image analysis and thereby adversely affecting the results. However, in order for the majority of axons to properly conduct electrical signals, the myelin sheath provided by oligodendrocytes present in the white matter must also be uninjured. Therefore, axons that appear to be structurally intact do not function properly if the associated oligodendrocyte is destroyed and the overall functional loss will be more severe for an injury group than what will be determined by only evaluating the axonal disruption. The overall density of the 3-APP measurements were significantly lower for the 40% residual compression injuries in comparison to the other injury groups. This finding is somewhat surprising as it would be expected that the accumulation would be equal or greater for the residual compression injuries since the initial contusion should produce a similar level of axonal disruption to the pure contusion injuries. If the subsequent compression had no effect the accumulation should be equivalent, but if it aggravated the injury then the accumulation should 99  Chapter 3. Histopathological Changes Following Residual Compression  increase. This result is possibly accounted for by the noticeable presence of APP accumulating in the gray matter of the residual compression groups. If the axon is disrupted within a short distance of the soma, the accumulation will appear in the gray matter and not the white matter. However, /3-APP was used as an indicator of white matter injury so the presence in the gray matter was not accounted for in the measurement results. Regional differences were observed in the white matter with the accumulation of greater densities in the more lateral tracts (ventrolateral and lateral). An inverse relation was expected in which the more medial tracts would display a greater overall accumulation due to their location at the centre of the injury. However, the selection of white matter tissue is much easier in lateral regions as the boundary between gray and white matter is not always distinct in the medial regions of the cord where gray and white matter tissue can appear disorganized due to the close proximity to the injury epicentre, and in the lateral region only white matter is present so all tissue is selected. Reduced /3-APP accumulation in the dorsal and ventral regions was also affected by the degree of injury to the axons in those regions. Qualitatively the white matter in the lateral regions appeared fairly intact, with obvious visible axonal tracts. In comparison, the dorsal and ventral white matter appeared very disrupted. For these reasons it is likely that the trauma to the white matter in the central regions was greater than the lateral tracts; a result that is converse to the trauma indicated by the densities of /3-APP. The overall rostral-caudal distribution of /3-APP was greatest at the periphery of the injury epicentre and skewed rostrally. The greatest accumulation should be present at the epicentre periphery near the primary site of axonal disruption, and a rostrally skewed distribution is expected due to the anterograde transport of the protein and primary presence of descending axons in the regions of analysis. The distribution of /3-APP was affected by the type of injury. Pure contusion resulted in a dense accumulation of /3-APP around the perimeter of the impact tip in comparison to the residual compression groups. Beyond ±1 mm the density for pure contusion was similar to both of the residual compression injuries. 90% residual compression resulted in a greater rostral-caudal extension of the white matter injury. This effect was observed by comparing the maximum rostral and caudal distances at which the injuries were statistically  100  Chapter 3. Histopathological Changes Following Residual Compression  different from more extreme distances from the injury epicentre. For pure contusion and 40% residual compression the densities within -1-2 mm were statistically greater than those beyond ±4 mm, but following 90% residual compression the densities were still greater at -2 mm and 3 mm than beyond 5 mm. This result suggests that 40% residual compression did not result in a greater extension of the white matter injury in comparison to pure contusion, but 90% residual compression does influence the overall injury, extending it in the rostral and caudal directions. Please see Chapter 4 for conclusions and recommended future directions relevant to this portion of the study.  101  Chapter 3. Histopathological Changes Following Residual Compression  3.5  References  [1] Anderson K. J., Pugaccia I., and Scheff S. W. Fluoro-jade B stains quiescent and reactive astrocytes in the rodent spinal cord. J Neurotrauma, 20(11):1223-31, November 2003. [2] Anderson T. E. Spinal cord contusion injury: experimental dissociation of hemorrhagic necrosis and subacute loss of axonal conduction. J Neurosurg, 62(l):115-9, January 1985. [3] Anderson T. E. and Stokes B. T. Experimental models for spinal cord injury research: physical and physiological considerations. J Neurotrauma, 9(Suppl l):S135-42, March 1992. [4] Basso D. M., Beattie M . S., and Bresnahan J. C. Graded histological and locomotor outcomes after spinal cord contusion using the NYU weight-drop device versus transection. Exp Neurol, 139(2):244-56, June 1996. [5] Behrmann D. L., Bresnahan J. C., Beattie M. S., and Shah B. R. Spinal cord injury produced by consistent mechanical displacement of the cord in rats: behavioral and histologic analysis. J Neurotrauma, 9(3):197-217, Fall 1992. [6] Blight A. R. Experimental Spinal Cord Injury Models. In Narayan R. K., Wilberger J. E., and Povlishock J. T., editors, Neurotrauma, chapter 101, pages 1367-79. McGraw Hill, Health Professions Disivion, 1996. [7] Bresnahan J. C., Beattie M . S., Todd r., F. D., and Noyes D. H. A behavioral and anatomical analysis of spinal cord injury produced by a feedback-controlled impaction device. Exp Neurol, 95(3):548-70, March 1987. [8] Bresnahan J. C., Beattie M . S., Stokes B. T., and Conway K. M . Three-dimensional computer-assisted analysis of graded contusion lesions in the spinal cord of the rat. J Neurotrauma, 8(2):91-101, Summer 1991. [9] Carlson G. D., Gorden C. D., Nakazawa S., Wada E., Smith J. S., and LaManna J. C. Sustained spinal cord compression: part II: effect of methylprednisolone on regional blood flow and recovery of somatosensory evoked potentials. J Bone Joint Surg Am, 85(A1): 95-101, January 2003. [10] Cornish R., Blumbergs P. C , Manavis J., Scott G., Jones N. R., and Reilly P. L. Topography and severity of axonal injury in human spinal cord,trauma using amyloid precursor protein as a marker of axonal injury. Spine, 25(10):1227-33, June 2000. [11] Dimar n., J. R., Glassman S. D., Raque G. H., Zhang Y. P., and Shields C. B. The influence of spinal canal narrowing and timing of decompression on neurologic recovery after spinal cord contusion in a rat model. Spine, 24(16):1623-33, August 1999. [12] Fernandes A. M., Maurer-Morelli C. V., Campos C. B., Mello M. L., Castilho R. F., and Langone F. Fluoro-Jade, but not Fluoro-Jade B, stains non-degenerating cells in brain and retina of embryonic and neonatal rats. Brain Res, 1029(l):24-33, December 2004. [13] Flanders A. E., Spettell C. M., Friedman D. P., Marino R. J., and Herbison G. J. The relationship between the functional abilities of patients with cervical spinal cord injury and 102  Chapter 3. Histopathological Changes Following Residual Compression  the severity of damage revealed by MR imaging. AJNR Am J Neuroradiol, 20(5):926-934, May 1999. [14] Hallam T. Protocol for Fluoro-Jade Histofluorescent Labeling of Degenerating Neurons, 1999. URL http://pappone.ucdavis.edu/tom/protocol/FJ/. [15] Hiruma S., Otsuka K., Satou T., and Hashimoto S. Simple and reproducible model of rat spinal cord injury induced by a controlled cortical impact device. Neurol Res, 21(3): 313-23, April 1999. [16] Holtz A., Nystrom B., Gerdin B., and Olsson Y. Neuropathological changes and neurological function after spinal cord compression in the rat. J Neurotrauma, 7(3):155-67, Fall 1990. [17] Kearney P. A., Ridella S. A., Viano D. C., and Anderson T. E. Interaction of contact velocity and cord compression in determining the severity of spinal cord injury. J Neurotrauma, 5(3):187-208, 1988. [18] Metz G. A., Curt A., van de Meent H., Klusman I., Schwab M. E., and Dietz V. Validation of the weight-drop contusion model in rats: a comparative study of human spinal cord injury. J Neurotrauma, 17(1):1-17, January 2000. [19] Mullen R. J., Buck C. R., and Smith A. M. NeuN, a neuronal specific nuclear protein in vertebrates. Development, 116(1):201-11, September 1992. [20] Noble L. J. and Wrathall J. R. Spinal cord contusion in the rat: morphometric analyses of alterations in the spinal cord. Exp Neurol, 88(1): 135-49, April 1985. [21] Scheff S. W., Rabchevsky A. G., Fugaccia I., Main J. A., and Lumpp J., J. E. Experimental modeling of spinal cord injury: characterization of a force-defined injury device. J Neurotrauma, 20(2):179-93, February 2003. [22] Schmued L. C. and Hopkins K. J. Fluoro-Jade B: a high affinity fluorescent marker for the localization of neuronal degeneration. Brain Res, 874(2):123-30, August 2000. [23] Schmued L. C., Albertson C , and Slikker J., W. Fluoro-Jade: a novel fluorochrome for the sensitive and reliable histochemical localization of neuronal degeneration. Brain Res, 751(l):37-46, March 1997. [24] Schmued L. C , Stowers C. C , Scallet A. C , and Xu L. Fluoro-Jade C results in ultra high resolution and contrast labeling of degenerating neurons. Brain Res, 1035(1):24-31, February 2005. [25] Sparrey C. The Effect of Impact Velocity on Acute Spinal Cord Injury. Master's thesis, University of British Columbia, Vancouver, 2004. [26] Stedman T. L. Stedman's concise medical dictionary for the health professions. Lippincott Williams & Wilkins, Philadelphia, illustrated 4th edition, 2001. [27] Stone J. R., Singleton R. H., and Povlishock J. T. Antibodies to the C-terminus of the beta-amyloid precursor protein (APP): a site specific marker for the detection of traumatic axonal injury. Brain Res, 871(2):288-302, July 2000. 103  Chapter 3. Histopathological Changes Following Residual Compression  [28] Stone J. R., Okonkwo D. O., Dialo A. 0., Rubin D. G., Mutlu L. K., Povlishock J. T., and Helm G. A. Impaired axonal transport and altered axolemmal permeability occur in distinct populations of damaged axons following traumatic brain injury. Exp Neurol, 190 (l):59-69, November 2004. [29] Sugawara T., Lewn A., Noshita N., Gasche Y., and Chan P. H. Effects of global ischemia duration on neuronal, astroglial, oligodendroglial, and microglial reactions in the vulnerable hippocampal CAI subregion in rats. J Neurotrauma, 19(l):85-98, January 2002. [30] Unal-Cevik I., Kilinc M., Gursoy-Ozdemir Y., Gurer G., and Dalkara T. Loss of NeuNimmunoreactivity after cerebral ischemia does not indicate neuronal cell loss: a cautionary note. Brain Res, 1015(l-2):169-74, July 2004. [31] Wagner J., F. C , VanGilder J. C , and Dohrmann G. J. Pathological changes from acute to chronic in experimental spinal cord trauma. J Neurosurg, 48(l):92-8, January 1978. [32] Young W. Spinal cord contusion models. Prog Brain Res, 137:231-55, 2002.  104  Chapter 4 General Discussion and Conclusions  Spinal cord decompression is one of the few treatment options that clinicians can currently use to treat patients with SCI, however until direct beneficial effects are proven, its use will remain controversial. Many investigators have developed experimental models to evaluate the effects of maintained compression, however replication of the clinical injury incorporating a primary contusion followed by residual compression of the spinal cord has not been a primary focus. With the utilization of a novel electromagnetic system for the creation of precisely controlled experimental SCI [13? ], we have been able to accurately replicate the dynamic nature of the clinical contusion injury as it is currently understood. Furthermore, we have demonstrated the ability to produce different levels of residual compression subsequent to an identical contusion injury providing a means to effectively evaluate the contribution that various levels of residual compression influence the overall spinal cord trauma.  4.1  General Limitations  Experimental models play a vital role in the research and advancement of clinical care. They provide the investigator with the means to control the focus and variability of a study that would be difficult, if not impossible, to achieve in the clinical setting. However, experimental models are only representations that are inherently limited by the assumptions that were adopted. For these reasons, careful consideration of the limitations of the model must be made before the experimental results are compared with other studies or extrapolated to the clinical stage.  105  Chapter 4. General Discussion and Conclusions  4.1.1  Anaesthesia  Anaesthesia is a required component of in vivo studies of injury for obvious ethical reasons. However, its use can have considerable effect on the outcome of the study if the type and concentration is not careful selected. General anaesthesia with either inhalation (halothane, ethyl ether, nitrous oxide) or nonvolatile (pentobarbital, urethane or ketamine) anaesthetics can result in significant changes in the calibre of the microvasculature and response to haemorrhage [40]. The effects vary with the type and dosage of anaesthesia. This study was conducted using 2 g/kg of urethane administered in three successive intraperitoneal injections at 10 minute intervals (Table 4.1). Urethane has been reported to result in a high mortality rate [19], which we observed when a single large injection was used in preliminary animals. Once the administration routine was modified to three successive injections, no further animals were lost due to overdose. Urethane was chosen due to its long lasting effects [19] and minimal interference with physiological responses [33, 41]. Following a recommended dosage of 1.2-1.5 g/kg [41], urethane does not influence the heart rate or eliminate the haemodynamic response to haemorrhage [26]. However, large doses (>1.5 g/kg) can suppress the cardiac and respiratory function [33]. Our initial attempts to use the recommended dosage of 1.5 g/kg did not result in a condition of surgical anaesthesia, so the dosage was increased to approximately 2 g/kg. However, urethane is a rather volatile compound, and depending upon the storage conditions and experimental preparation, the calculated concentrations may be overestimated [33]. In comparison to other standardly used injectable anaesthetics, urethane produced the least arteriolar dilation and resulted in decreased systemic blood pressure for the initial 30 minutes following administration to a bat [40]. The initial reduction in mean arterial blood pressure is seen in spontaneously hypertensive but not Wistar or Sprague Dawley rats [26]. This study involved the contusion and subsequent residual compression of the rat spinal cord requiring the use of anaesthesia. The microvascular blood flow was monitored for each animal as it was postulated that residual compression might adversely affect the blood flow in the spinal cord differently from pure contusion. For this reason an anaesthetic that provided a stable, long-term level of anaesthesia while not adversely affecting the cardiovascular system 106  Chapter 4. General Discussion and Conclusions  was needed. Urethane was chosen for this study as it provided all of the desired properties and is an injectable anaesthetic that does not require the additional hardware associated with inhalants such as isoflurane that are often used for studies requiring long-term, controlled levels of anaesthesia. Table 4-T. Dosage of urethane anaesthesia used in this study was not statistically different between groups (p=0.6) Control  Contusion  Mean  2.08  2.04  1.93  2.07  Std Dev  0.26  0.22  0.25  0.20  Group Anaesthesia (g/kg)  4.1.2  40% Residual 90% Residual  A n i m a l s as M o d e l s of H u m a n Injury  Models of SCI have been conducted in various species, including cats [8, 22, 27, 29, 44, 50, 58, 59], dogs [1, 9-12, 14, 17, 38, 56, 57], ferrets [2, 3, 30], mice [28, 49, 54], monkeys [6, 3437], and rats [15, 18, 21, 23-25, 31, 42, 43, 45, 46, 48, 55]. Animal models have proven to be invaluable for the development of experimental therapies [39], allowing research into a multitude of potentially beneficial treatments in a timely manner that would otherwise be impractical and ethically questionable in humans. In recent years the rat and mouse have become the most popular, primarily due to their cost and accessibility. Although mice have the added advantage of their transgenic potential, their histopathological response following injury is distinct from that seen in rats and humans [53]. The general composition of the spinal cord is similar in both rats and humans. The central gray matter is divided into two symmetrical halves of cell bodies (sensory neurons in dorsal horn; motor neurons in ventral horn) surrounded by tracts of myelinated white matter [32, 60] and the arterial supply of the rat is almost identical to that of humans [47]. Following uniaxial tension the spinal cords of both humans and rats have been shown to exhibit a similar "J" shaped non-linear stress-strain response [4, 20]. The main differences in anatomy relate to the overall dimensions of the cords in humans and rats [16, 47], although the size of the individual cellular components are similar [5].  107  Chapter 4. General Discussion and Conclusions  Another factor that must be taken into consideration when using rats as models of human SCI, is the variation in the location of similar functional groupings of white matter tracts (Figure 4.1). In the rat the rubrospinal tract plays an important role is controlling the large motor functions and is prominently located in the lateral funiculus [47]. In humans the rubrospinal tract is rudimentary and terminates at the second cervical level [32]. Its function is superceded by the corticospinal tract in humans which is primarily located in the lateral funiculus as opposed to the dorsal funiculus in the rat [32, 47]. Descending I c i c n .  (a)  Ascending Tmcls:  (b)  Figure 4-1" Comparison of the major axonal tracts of the white matter of the (a) human [32] and (b) rat [47]. Descending (motor) tracts shown in red on left side. Ascending (sensory) tracts shown in blue on right side. Figure based on images from Kiernan and Barr [32].  4.1.3  Rigid  Support of Vertebral Column  The current experiments involved the extended compression of the spinal cord following an initial high rate contusion and the continuous monitoring of the dorsolateral blood flow with a LD flowmetry system. Rigid stabilization of the vertebral column was a vital requirement for the generation of a consistent SCI and to ensure that the LD probe monitored the blood flow from the same location throughout the experiments. The vertebral column was stabilized using a technique previously developed in our laboratory for the creation of cervical SCIs [13]. The clamp secures the vertebrae on the edge of the pedicle just between the lateral process and the posterior surface of the rib. Although this technique requires a more extensive exposure of the posterior column than the application of allice clamps to the spinous processes, the support 108  Chapter 4. General Discussion and Conclusions  that is confers it considerably more rigid and reliable over an extended period of time. When using allice clamps extreme care must be taken throughout the set-up process to ensure that the spinous processes are not fractured, because once a fracture occurs rigid support of the animal will not be possible and use of the animal in the experiment will not be possible, fn comparison the animal can easily be supported solely by the vertebral body clamp without worry of fracturing the spinal column or loosening of the clamp. Allice clamps have been used extensively for supporting the vertebral column during the creation of contusion injuries. However, previous work in our laboratory using modified allice clamps required the exclusion of 10 to 20% of the animals due to slippage during the injury or fracturing of the spinal processes during the clamping process [52]. A review of vertebral column support techniques also demonstrated that when rostral and caudal clamps are used to secure the spine, the motion of the spinal column accounts for 26% of the overall motion recorded during the contusion injury [51]. In the current set of experiments no animals were excluded due to slippage or adverse motion of the spinal column. The vertebral column clamp rigidly supported all of the animals throughout the measurement period. However, proper use of the clamps requires extra attention during the creation of the laminectomy for the exposure of the spinal cord. The clamp produces a rigid connection with the vertebrae by grasping the sides of the pedicles and squeezing them medially, ff too much of the laminar bone is removed, the force of the clamp flexes the pedicles inward, thereby reducing the exposure cross section to a diameter that is too small for the passage of the impactor. Further enlargement of the laminectomy often results in fracturing of the pedicles, making a rigid connection very difficult. To overcome these difficulties the clamps were designed in a modular fashion, with each side consisting of two separate teeth. The rostral tooth clamped the T9 vertebrae and the caudal tooth spanned T10 and T i l . Each tooth was also supported on an adjustable spring-loaded arm, allowing the distance between opposing teeth to be easily varied. These features allowed the clamp to be adjusted for variations in anatomy and the clamping force to be reduced when the laminar bone was unable to support the full force of the clamp.  109  Chapter 4.  4.1.4  General Discussion and Conclusions  Bias & Variability in Analysis Methods  This thesis evaluated the effects of subsequent 40% and 90% residual compression in comparison to a pure contusion injury on the microvascular blood flow and acute histopathological changes in the gray and white matter. Every attempt was made to ensure that the measurements were unbiased, however since all of the measurements and locations of analysis were made by the primary author this is unlikely. As the objectives of the study were fairly basic the preliminary nature of the investigation did not warrant the additional involvement of another observer. Only twenty-nine specimens were included in the final analysis of this study and large variations were observed in a number of the measured parameters. These factors impacted the results of the experiment, causing some comparisons which appeared to be quite different to not reach statistical significance. The inclusion of additional animals may have increased the power of the results, although a formal power analysis was not performed. The number of specimens was limited for each group due to the in vivo nature and preliminary scope of the experiment, and the use of additional animals is likely unwarranted. Selection of the appropriate tissue sections, analysis regions and thresholding values were subjective. To minimize the variability associated with these measurements, the majority of the analysis routines were automated to ensure consistency between specimen. Some steps required a manual process due to natural variability in the anatomy of the specimen, variation in the alignment of the parasagittal sections and strength of the tissue staining. These effects were minimized by using basic analyses to quantify the overall spinal cord trauma, however even these type of measurements were often difficult to properly assess and did not capture the entire extent of the damage. The amount of tissue disruption was rather extensive in the central regions of the spinal cord making the analysis difficult.  4.2 Summary &; Conclusions This thesis evaluated the effects of 40% and 90% residual compression on the microvascular blood flow and acute histopathological changes in the gray and white matter subsequent to a  110  Chapter 4.  General Discussion and Conclusions  contusion injury to a depth of 1 mm at 700 mm/s. ft involved the establishment of injury protocols which included the consistent production of residual spinal cord compression subsequent to a contusion injury of a reproducible magnitude. Methods to rigidly support the thoracic vertebral column and monitor the microvascular blood flow of the spinal cord for extended periods of time were also developed. This was followed by a series of animal tests conducted using the established protocols. The harvested spinal cords were assessed using H&E to measure the volume and extend of haemorrhage, NeuN and F J to assess the extent of gray matter damage and /3-APP to assess the extent of white matter disruption. The load relaxation of the spinal cord is rapid for any level of residual compression following a moderate contusion injury of the spinal cord. The initial contusion results in considerable disruption of the localized tissue such that it cannot support any further compression. The local vascular structure is likely to be compressed resulting in restricted blood flow to the tissue undergoing residual compression. The dorsolateral, microvascular blood flow, one level (5 mm) caudal to the injury epicentre is not adversely affected by a moderate contusion of the spinal cord. Throughout the analysis period the blood flow following a pure contusion injury was similar to the pre-injury and surgical control levels. Residual compression magnitudes of 40-90% can result in considerable increases in the blood flow distant to the injury epicentre. This may be an effect of restricted blood flow to the injury epicentre resulting from the residual compression. It has been shown elsewhere that the uninjured spinal cord possesses an autoregulatory process to maintain the blood flow at a consistent level. If the blood flow is restricted at the compression site, the surrounding vasculature may become dilated in an effort to increase the restricted flow. Such a regulated process is likely to have occurred in this study, as the blood flow immediately dropped to within normal levels once the spinal cord was decompressed. Therefore it is likely that the residual compression following the contusion injury resulted in a localized reduction of the blood flow to the tissue at the injury epicentre. In an effort to correct the reduction in blood flow, the surrounding tissue increased its blood flow by utilizing an autoregulatory process. Once the spinal cord was fully decompressed and the flow in the local tissue was restored, the  111  Chapter 4.  General Discussion and Conclusions  autoregulatory process returned the blood flow in the surrounding tissue to a normal level. Haemorrhage volume was similar for pure contusion and both levels of residual compression, though the extent was greater following 90% residual compression. Qualitatively the residual compression injuries appeared to have more diffuse haemorrhage at the injury epicentre and areas of necrotic tissue devoid of red blood cells. This observation suggests that within the injury epicentre the blood flow was diminished, supporting the autoregulatory explanation for the increased distal perfusion as measured by the laser doppler probe. Haemorrhage following 90% residual compression varied considerably, with a few specimens displaying very few red blood cells, while other specimens had haemorrhage which extended beyond the analyzed 18 mm cord section. Previously, haemorrhage has been used to estimate the severity of a spinal cord injury, and the volume and area have been correlated with the long-term functional outcome [7]. In comparison, the findings of this thesis imply that the extent of haemorrhage is a better predictor of the overall injury severity than the total volume. When residual compression is involved, the volume of necrotic tissue may be a better predictor, as some specimens with obvious cellular damage demonstrate very little haemorrhage. High levels (90%) of residual compression result in an extension of the gray matter damage beyond the immediate area disrupted by the initial contusion injury. Low levels (40%) of residual compression do not appear to increase the cellular damage beyond that of the contusion. These assessments were made using two unrelated neuron specific markers: F J and NeuN. Similar lengths of the lesion were calculated for both markers, providing confidence that the overall extent of gray matter damage were correct. The lesion length was similar in both the rostral and caudal directions and throughout the central, ventrolateral and dorsolateral regions. It is likely that our measurements of gray matter injury overestimated the extent of injury, as viable cells were evident around and closer to the injury epicentre. However, these measurements were taken at a fairly early time point (3 hrs) following the initial injury and further cell death is going to occur due to processes such as apoptosis. The results clearly show that high levels of residual compression adversely increase the lesion throughout the gray matter.  112  Chapter 4.  General Discussion and Conclusions  Following a pure contusion injury the disruption of the white matter appears to be focused within the immediate periphery of the injury epicentre. Low levels of residual compression (40%) do not appear to increase the extension of the injury to the white matter, although the overall accumulation of /3-APP was less for these animals. The cause of this effect is unknown. There may be an association between the immediate gray and white matter injury as suggested by the qualitative observation of /3-APP accumulation in the gray matter. However, high levels of residual compression (90%) result in further disruption of the white matter from the injury epicentre than pure contusion. The lateral and ventrolateral regions had higher densities than either the ventral or dorsal regions. This was especially true for pure contusion at the periphery of the epicentre (±1 mm). Qualitatively, the white matter in the central regions appeared very disrupted in comparison to the lateral tracts. This probably accounts for the lower accumulations of /3-APP in these regions, as the protein cannot be detected because it has not been transferred into these areas. Although /3-APP is an excellent marker for axonal disruption, in areas of severe damage it may underestimate the cellular trauma because the protein will not be transferred into this region. Therefore an additional marker of white matter damage should be used in combination with /3-APP to determine the exact extent and severity of the injury. Moderate contusion of the spinal cord results in disruption of the localized cellular structure producing injury throughout the gray and surrounding white matter that is focused within the rostral-caudal periphery of the injury epicentre. A moderate contusion, alone, does not result in changes in the microvascular blood flow distant to the injury. When the contusion is followed by subsequent residual compression, the blood flow increases for at least the initial hour of compression, but returns to normal levels immediately following decompression. This suggests that the autoregulatory process of the spinal cord remains intact following moderate injury and increases the blood flow distant to the injury epicentre in an attempt to corrected the restricted flow created by the compression of the cord. The effects of residual compression are dependent upon its relative magnitude in comparison to the initial contusion. One hour of 40% (0.4 mm) residual compression does not increase the injury to the gray and white matter in comparison to a purely contusive injury within the initial three hours following injury. In contrast, a high 113  Chapter 4.  General Discussion and Conclusions  level (90%) of subsequent residual compression adversely increases the extents of both the gray and white matter injury. Therefore in the presence of high levels of residual compression it is advisable to perform decompression as soon as possible following the initial injury to prevent exacerbation of the injury. However, further research is required to fully determine the effects of residual compression on the local blood flow and injury to the gray and white matter and how different magnitudes of contusion and relative residual compression interact to create the overall spinal cord trauma.  4.3 Future Directions This thesis probably spawned more questions than it answered. However, the scope of this thesis was not to determine the exact mechanisms and progressions of spinal cord pathology when residual compression is involved. Its primary objective was to compare two distinct levels of residual compression subsequent to a contusive trauma with a pure contusion injury to determine if residual compression influences the extent of spinal cord trauma. As simple as this objective may seem no model previously existed that allowed the combined and independent control of a contusion and residual compression injury. Therefore considerable effort was expanded in the creation of such a model. However, there are a number of unknowns and observations from this basic study that warrant further investigation. Functional evaluations are the current gold standard of evaluating experimental SCI research. Therefore each new model must undergo a long-term functional evaluation to determine the severity of the resulting spinal cord trauma. While the magnitudes of the contusion injuries used in this study were similar to previous investigations, the greater velocity and more rigid clamping method may significantly alter the severity of the injury. Further evaluations on the effects of residual compression on spinal cord blood flow are also warranted. Techniques that allow more regionalized measurements, such as fluorescent microspheres may be advantageous to determine the individual effects on the gray and white matter.  114  Chapter 4.  General Discussion and Conclusions  The vertebral body clamp was not designed with the intention of performing long-term function assessments, but its modular design may allow the integration of additional components that have a smaller profile and are capable of maintaining the residual compression of the spinal cord after the animal is removed from the actuator. The high level of precision obtained in the current experiments may be difficult to obtain if the compression on the cord is to be maintained while the animal is disconnected from the actuator. However, the ability to do so will be advantageous, as it will allow experiments to be conducted more rapidly. Using the current protocol only one animal can be injured/compressed at one time, therefore longer compression times will result in considerably longer experiment times. Still, disconnecting the animal from the actuator may not be entirely necessary as clinically relevant times for decompression may only be in the initial few hours following injury, as suggested by earlier maintained compression studies. The load relaxation was also observed to occur extremely rapidly in the current study so it may be interesting to evaluate the load relaxation following a more gradual initial contusion (compression) and subsequent residual compression. For lower velocities the disruption to the spinal cord may be less, enabling it to support a residual load and thereby affecting the local blood flow and cellular damage. It is also anticipated that future investigations of residual compression will be undertaken utilizing multiple levels of the initial contusion and subsequent residual compression. Hopefully such experiments in combination with this one, will fully elucidate the interaction between varying levels of contusive and compressive traumas and finally determine if residual compression is a significant factor contributing to the overall severity of SCIs. Maybe such data will provide conclusive evidence for the beneficial effects of decompression of the persistently compressed spinal cord and eliminate the controversy surrounding its use.  115  Chapter 4.  4.4  General Discussion and  Conclusions  References  [1] Aki T. and Toya S. Experimental study on changes of the spinal-evoked potential and circulatory dynamics following spinal cord compression and decompression. Spine, 9(8): 800-9, November-December 1984. [2] Anderson T. E. A controlled pneumatic technique for experimental spinal cord contusion. J Neurosci Methods, 6(4):327-33, November 1982. [3] Anderson T. E. Spinal cord contusion injury: experimental dissociation of hemorrhagic necrosis and subacute loss of axonal conduction. J Neurosurg, 62(l):115-9, January 1985. [4] Bilston L . E. and Thibault L. E. The mechanical properties of the human cervical spinal cord in vitro. Ann Biomed Eng, 24(l):67-74, January-February 1996. [5] Blight A. R. Experimental Spinal Cord Injury Models. In Narayan R. K., Wilberger J. E., and Povlishock J. T., editors, Neurotrauma, chapter 101, pages 1367-79. McGraw Hill, Health Professions Disivion, 1996. [6] Bresnahan J. C. An electron-microscopic analysis of axonal alterations following blunt contusion of the spinal cord of the rhesus monkey (Macaca mulatta). J Neurol Sci, 37 (l-2):59-82, June 1978. [7] Bresnahan J. C , Beattie M . S., Todd r., F. D., and Noyes D. H. A behavioral and anatomical analysis of spinal cord injury produced by a feedback-controlled impaction device. Exp Neurol, 95(3):548-70, March 1987. [8] Brodkey J. S., Richards D. E., Blasingame J. P., and Nulsen F. E. Reversible spinal cord trauma in cats. Additive effects of direct pressure and ischemia. J Neurosurg, 37(5):591-3, November 1972. [9] Carlson G. D., Minato Y., Okada A., Gorden C. D., Warden K . E., Barbeau J. M . , Biro C. L., Bahnuik E., Bohlman H. H., and Lamanna J. C. Early time-dependent decompression for spinal cord injury: vascular mechanisms of recovery. J Neurotrauma, 14(12): 951-62, December 1997. [10] Carlson G. D., Gorden C. D., Nakazowa S., Wada E., Warden K., and LaManna J. C. Perfusion-limited recovery of evoked potential function after spinal cord injury. Spine, 25 (10):1218-26, May 2000. [11] Carlson G. D., Gorden C. D., Nakazawa S., Wada E., Smith J. S., and LaManna J. C. Sustained spinal cord compression: part II: effect of methylprednisolone on regional blood flow and recovery of somatosensory evoked potentials. J Bone Joint Surg Am, 85(A1): 95-101, January 2003. [12] Carlson G. D., Gorden C. D., Oliff H. S., Pillai J. J., and LaManna J. C. Sustained spinal cord compression: part I: time-dependent effect on long-term pathophysiology. J Bone Joint Surg Am, 85(Al):86-94, January 2003. [13] Choo A., Liu J., Lam C , Dvorak M . , Tetzlaff W., and Oxland T. New Device for Producing Different Mechanisms of Spinal Cord Injury. The American Society of Mechanical Engineers, June 2005. Summer Bioengineering Conference. 116  Chapter 4.  General Discussion and  Conclusions  [14] Delamarter R. B., Sherman J., and Carr J. B. Pathophysiology of spinal cord injury. Recovery after immediate and delayed decompression. J Bone Joint Surg Am, 77(7):10429, July 1995. [15] Dimar n., J. R., Glassman S. D., Raque G. H., Zhang Y. P., and Shields C. B. The influence of spinal canal narrowing and timing of decompression on neurologic recovery after spinal cord contusion in a rat model. Spine, 24(16): 1623-33, August 1999. [16] Dobkin B. H. and Havton L. A. Basic advances and new avenues in therapy of spinal cord injury. Annu Rev Med, 55:255-282, 2004. [17] Ducker T. B. and Perot J., P. L. Spinal cord oxygen and blood flow in trauma. Surg Forum, 22:413-5, 1971. [18] Fehlings M . G. and Nashmi R. A new model of acute compressive spinal cord injury in vitro. J Neurosci Methods, 71(2):215-24, March 1997. [19] Field K . J., White W. J., and Lang C. M . Anaesthetic effects of chloral hydrate, pentobarbitone and urethane in adult male rats. Lab Anim, 27(3):258-69, July 1993. [20] Fiford R. J. and Bilston L. E. The mechanical properties of rat spinal cord in vitro. J Biomech, 38(7):1509-1515, July 2005. [21] Fiford R. J., Bilston L. E., Waite P., and Lu J. A vertebral dislocation model of spinal cord injury in rats. J Neurotrauma, 21(4):451-8, April 2004. [22] Ford R. W. A reproducible spinal cord injury model in the cat. J Neurosurg, 59(2):268-75, August 1983. [23] Gupta R., Rowshan K., Chao T., Mozaffar T., and Steward O. Chronic nerve compression induces local demyelination and remyelination in a rat model of carpal tunnel syndrome, Exp Neurol, 187(2) :500-8, July 2004. [24] Hashimoto T. and Fukuda N . New spinal cord injury model produced by spinal cord compression in the rat. J Pharmacol Methods, 23(3):203-12, May 1990. [25] Hiruma S., Otsuka K., Satou T., and Hashimoto S. Simple and reproducible model of rat spinal cord injury induced by a controlled cortical impact device. Neurol Res, 21(3): 313-23, April 1999. [26] Holobotovskyy V. V., Arnolda L. F., and McKitrick D. J. Effect of anaesthetic and rat strain on heart rate responses to simulated haemorrhage. Acta Physiol Scand, 180(1): 29-38, January 2004. [27] Hung T. K., Lin H. S., Bunegin L., and Albin M . S. Mechanical and neurological response of cat spinal cord under static loading. Surg Neurol, 17(3):213-7, March 1982. [28] Jakeman L. B., Guan Z., Wei P., Ponnappan R., Dzwonczyk R., Popovich P. G., and Stokes B. T. Traumatic spinal cord injury produced by controlled contusion in mouse. J Neurotrauma, 17(4):299-319, April 2000.  117  Chapter 4.  General Discussion and Conclusions  Katsuta M . Influence of spinal cord compression on the C3-C4 propriospinal neurons in cats. J Orthop Sci, 8(3):367-73, 2003. Kearney P. A., Ridella S. A., Viano D. C , and Anderson T. E. Interaction of contact velocity and cord compression in determining the severity of spinal cord injury. J Neurotrauma, 5(3):187-208, 1988. Khan M . and Griebel R. Acute spinal cord injury in the rat: comparison of three experimental techniques. Can J Neurol Sci, 10(3):161-5, August 1983. Kiernan J. A. and Barr M . L. Barr's the human nervous system: an anatomical viewpoint. Lippincott-Raven, Philadelphia, 7th edition, 1998. Koblin D. D. Urethane: help or hindrance? Anesth Analg, 94(2):241-2, February 2002. Kobrine A. I., Doyle T. F., and Martins A. N . Autoregulation of spinal cord blood flow. Clin Neurosurg, 22:573-81, 1975. Kobrine A. I., Doyle T. F., Newby N., and Rizzoli H . V. Preserved autoregulation in the rhesus spinal cord after high cervical cord section. J Neurosurg, 44(4):425-8, April 1976. Kobrine A. I., Evans D. E., and Rizzoli H. Correlation of spinal cord blood flow and function in experimental compression. Surg Neurol, 10(l):54-9, July 1978. Kobrine A. I., Evans D. E., and Rizzoli H. V . Experimental acute balloon compression of the spinal cord. Factors affecting disappearance and return of the spinal evoked response. J Neurosurg, 51(6):841-5, December 1979. Koozekanani S. H., Vise W. M . , Hashemi R. M . , and McGhee R. B. Possible mechanisms for observed pathophysiological variability in experimental spinal cord injury by the method of Allen. J Neurosurg, 44(4):429-34, April 1976. Kwon B. K., Oxland T. R., and Tetzlaff W. Animal models used in spinal cord regeneration research. Spine, 27(14):1504-1510, July 2002. Longnecker D. E. and Harris P. D. Microcirculatory actions of general anesthetics. Fed Proc, 39(5):1580-1583, April 1980. Maggi C. A. and Meli A. Suitability of urethane anesthesia for physiopharmacological investigations in various systems. Part 2: Cardiovascular system. Experientia, 42(3):2927, March 1986. Metz G. A., Curt A., van de Meent H., Klusman I., Schwab M . E., and Dietz V. Validation of the weight-drop contusion model in rats: a comparative study of human spinal cord injury. J Neurotrauma, 17(1):1—17, January 2000. Mills C. D., Hains B. C , Johnson K . M . , and Hulsebosch C. E. Strain and model differences in behavioral outcomes after spinal cord injury in rat. J Neurotrauma, 18(8):743-56, August 2001. Nelson E., Gertz S. D., Rennels M . L., Ducker T. B., and Blaumanis O. R. Spinal cord injury. The role of vascular damage in the pathogenesis of central hemorrhagic necrosis. Arch Neurol, 34(6):332-3, June 1977. 118  Chapter 4.  General Discussion and Conclusions  Noble L. J. and Wrathall J. R. Spinal cord contusion in the rat: morphometric analyses of alterations in the spinal cord. Exp Neurol, 88(l):135-49, April 1985. Oro J. J., Gibbs S. R., and Haghighi S. S. Balloon device for experimental graded spinal cord compression in the rat. J Spinal Disord, 12(3):257-61, June 1999. Paxinos G. The Rat nervous system. Academic Press, San Diego, 2nd edition, 1995. Raines A., Dretchen K . L., Marx K., and Wrathall J. R. Spinal cord contusion in the rat: somatosensory evoked potentials as a function of graded injury. J Neurotrauma, 5(2): 151-60, 1988. Seki T., Hida K., Tada M., Koyanagi I., and Iwasaki Y . Graded contusion model of the mouse spinal cord using a pneumatic impact device. Neurosurgery, 50(5): 1075-81, May 2002. Senter H. J. and Venes J. L. Loss of autoregulation and posttraumatic ischemia following experimental spinal cord trauma. J Neurosurg, 50(2):198-206, February 1979. Somerson S. K . Analysis of an electro-mechanical spinal cord injury device. Matster of science, Ohio State University, Columbus, 1986. Sparrey C. The Effect of Impact Velocity on Acute Spinal Cord Injury. Master's thesis, University of British Columbia, Vancouver, 2004. Sroga J. M., Jones T. B., Kigerl K . A., McGaughy V. M., and Popovich P. G. Rats and mice exhibit distinct inflammatory reactions after spinal cord injury. J Comp Neurol, 462 (2):223-40, July 2003. Stokes B. T. and Jakeman L. B. Experimental modelling of human spinal cord injury: a model that crosses the species barrier and mimics the spectrum of human cytopathology. Spinal Cord, 40(3):101-9, March 2002. Taoka Y . and Okajima K . Spinal cord injury in the rat. Prog Neurobiol, 56(3):341-58, October 1998. Tarlov I. M . Spinal cord compression studies. III. Time limits for recovery after gradual compression in dogs. AMA Arch Neurol Psychiatry, 71(5):588-97, May 1954. Tarlov I. M . and Klinger H. Spinal cord compression studies. II. Time limits for recovery after acute compression in dogs. AMA Arch Neurol Psychiatry, 71(3):271-90, March 1954. Thienprasit P., Bantli H., Bloedel J. R., and Chou S. N . Effect of delayed local cooling on experimental spinal cord injury. J Neurosurg, 42(2):150-4, February 1975. Wagner J., F. C , VanGilder J. C , and Dohrmann G. J. Pathological changes from acute to chronic in experimental spinal cord trauma. J Neurosurg, 48(l):92-8, January 1978. Zeman W., Craigie E. FL, and Innes J. R. M . Craigie's neuroanatomy of the rat, revised and expanded. Academic Press, New York, 1963.  119  Appendices  120  Appendix A Specimens &; Experimental Apparatus  A.l  Specimens  Forty-six male Wistar rats were used in the entirety of this study, of which 29 were evaluated in the final study design. Thirteen animals were used in a preliminary study, where adjustments were made to the testing apparatus and surgical procedure. The survival time following injury was one-half (1.5 hours) of the time used in the final investigation. Four additional animals were also removed from the final analysis. Two specimens were excluded after the exposed spinal cord was severely damaged during set-up, and two animals suspected of additional non-specific injury were excluded after confirmation following tissue harvesting. All specimens were acquired from the UBC Animal Care Centre (South Campus Animal Facility, University of British Columbia, Vancouver, BC) or Charles River Laboratories (Charles River Laboratories International, Inc, Wilmington, MA). Preoperatively, the animals were housed in the ICORD animal house with food and water supplied freely. The mean weight of the 29 rats used for evaluation was 314.0 g (sd = 8.6 g).  A.2  Contusion Apparatus  All contusion injuries were performed using an electromagnetic linear actuator (TestBench ELF, EnduraTec, Minnetonka, MN) mounted to a custom built support frame (Figure A.l). Adjustments in the positioning of the actuator were made by a system of connections that allowed rotation and radial translation in the horizontal plane. The vertical height was controlled using a linear ball screw actuator (SuperSlide 2DB, Danaher Motion, Radford, VA) that allowed continuous rapid translation or incremental stepping of 0.002 inches. An x-y table allowed fine adjustment of the specimen for alignment under the actuator. The actuator was controlled from a PC using the Enduratec supplied software, Wintest Digital Control System. Force measurements were made using a 22.2 N precision miniature load cell (Model 31, Honeywell Sensotec, Columbus, OH) with a full-scale non-linearity of ±0.15% and output of 2 mV/V (Input voltage: 5 V). Acceleration was measured using a 490 m/s (50 g) (Model 355B03, PCB Piezotronics, Depew, NY) accelerometer with a non-linearity of <1% full-scale, resolution of 0.0009 m/s and sensitivity of 10.82 mV/m/s . Displacement was measured using the linear variable differential transducer (LVDT) integrated into the actuator (MHR250 10FT, Measure2  2  2  121  Appendix  A.  Specimens <fe Experimental  Apparatus  Figure A.l: Linear electromagnetic actuator and support frame used for the creation of the spinal cord injuries. ment Specialties Inc, Hampton, VA) which has a full-scale non-linearity of 0.23% and sensitivity of 86 mV/V/mm. The contusion implement was fabricated from a 2 mm stainless-steel dowel pin (91585A024, McMaster-Carr, Atlanta, GA) protruding from an aluminium shank and held in place by 0-80 machine screws (Figure A.2). This arrangement allowed the overall length to be varied while keeping the mass low. One end of the dowel pin was polished to form a smooth surface, and the edges were slightly rounded.  Figure A.2: Impact tip fabricated from a 2 mm stainless-steel dowel pin and aluminum housing that allowed the overall length to be adjusted.  122  Appendix A.  A.3  Specimens &c Experimental Apparatus  Temperature Measurements and C o n t r o l  Body temperature was controlled using a heating plate (HP1M, Physitemp Instruments, Clifton, NJ) connected to a temperature controller (TCAT-2LV, Physitemp Instruments, Clifton, NJ) (Figure A.3). The temperature of the heating plate was regulated to ensure that its temperature never surpassed 40°C, thereby prevented burns to the animal. The internal temperature of the animal was monitored using a rectal probe (RET-2, Physitemp Instruments, Clifton, NJ) that supplied feedback to the controller.  Figure A.3:  Heating pad and rectal probe for maintenance of specimen body temperature at  37°C. Custom heating coils created from 40 gauge magnet wire were used to increase the temperature of the artificial CSF (Figure A.4(a)). A secondary coil enclosed within a section of nylon tubing was used to preheat the CSF fluid before it was released into the surgical cavity (Figure A.4(b)). A primary heating < "il was secured to one side of the vertebral body clamp and was used to regulate the temperature of the CSF in the surgical exposure. The coils were wired in series so that they could be controlled simultaneously. The temperature of the heating coils was manual controlled by regulating the current with a DC power supply (PSA-2530D, RP Electronics, Vancouver, BC). An aquarium electronic thermometer was used to monitor the temperature of the fluid prior to it being released into the surgical exposure. A type-k thermocouple (80PK1, Fluke Electronics Canada, Mississauga, ON) connected to a digital multimeter (Fluke 179, Fluke Electronics Canada, Mississauga, ON) measured the temperature of the fluid in the surgical exposure. -  123  Appendix A.  Specimens & Experimental  Apparatus  (a)  (b)  Figure A.4'- Heating coils used to maintain the temperature of the CSF fluid at 37°C: (a) placed within the surgical cavity for direct maintenance of the fluid temperature; (b) nylon tubing containing secondary coil and temperature probe that were used to preheat the CSF fluid before being released within the surgical cavity.  A.4  Vitals  Heart rate and haemoglobin oxygen saturation were monitored using a veterinary pulse oximeter (8600V, Nonin Medical, Plymouth, MN) and clip-on sensor (2000SL, Nonin Medical, Plymouth, MN) positioned at the base of the tail. Heart rate and percent oxygen saturation were recorded over a serial port at 1 Hz using custom software (Matlab 6.1, The Mathworks, Natick, MA).  A.5  Support Frame  The specimens were supported in a custom modified stereotactic frame (Figure A.5(a)). Two horizontal fixtures were connected to a crossbar supported between two vertical bars on either side of the frame. The crossbar was fabricated with a slot that allowed adjustment in the distance between the horizontalfixtures(Figure A.5(b)). The vertebral column clamp was held in place by two pairs of holes at the ends of the horizontal fixtures (Figure A.5(c)). Thumb screws provided rigid fixation between the clamp and fixtures. The vertebral column clamps rigidly held the vertebral column from T9-T11 on the edge of the pedicle just between the lateral process and the posterior surfaces of the rib (Figure A.6). The clamps were fabricated from two matching sides consisting offiveprimary components: support bar, rostral/caudal swing-arms and rostral/caudal teeth (Figure A.7). The support bars were separated by spacers and bolted together with thumb screws. Knurled locking nuts prevented the thumb screws from loosening and also provided an additional connection between the clamp and the horizontal fixtures. The swing-arms were connected to either end of the support bar, such that they could rotate around the axis of their connection. A second machine screw held fourteen 31 N (7 lb) belleville disc springs (9713K51, McMaster-Carr, Atlanta, GA), stacked serially in seven pairs between the support bar and each swing-arm. This arrangement allowed minor adjustments in the angular orientation of the swing-arms to account for variations in anatomical size (Figure A.8). An angled tooth was bolted to the base of the swing-arms and formed the direct connection to the vertebral column. The larger caudal tooth supported the 124  Appendix A.  Specimens & Experimental  Apparatus  (b)  (c)  Support structures for rigid fixation of specimen during tests; (a) modified stereotactic frame with vertebral column clamp in place; (b) crossbar that provided adjustments for the height of the clamp and distance between the horizontal fixtures; (c) horizontal fixtures that secured clamp in place. Figure A.5:  125  Appendix A.  Specimens & Experimental  Apparatus  T10-11 vertebrae and the rostral tooth supported T9. All custom parts were fabricated from 303-stainless steel.  Figure A.6: The vertebral body clamp secures the vertebrae on the edge of the pedicle just between the lateral process and the posterior surfaces of the rib.  Figure A.7: Primary components of clamp: support bar, rostral/caudal swing-arms and rostral/caudal teeth  Figure A.8: Distance between teeth can be finely adjusted by changing angular orientation of swing-arms.  A.6  Laser D o p p l e r F l o w m e t r y System  Blood flow microperfusion was measured using a 1mm diameter laser doppler probe (Probe 306, Perimed AB, Stockholm, Sweden) connected to a single channel Periflux controller (PF 3, 126  Appendix A.  Specimens <fc Experimental  Apparatus  Perimed AB, Stockholm, Sweden). The probe was supported by a customized holder mounted to a horizontally aligned electrode manipulator (Model 960, David Kopf Instruments, Turjunga, California). The holder consisted of a 4mm diameter tube that held the probe from the rubber ring at the base of the metal probe-tip. An installation tool was used to insert the laser doppler probe into the tube of the holder, as the probe had to be installed after the specimen was positioned in the stereotactic frame (Figure A.9). A long shaft connected the probe holder to a vertical alignment device that raised/lowered the probe by rotating a thumb-screw (Figure A. 10). The electrode manipulator was used to make changes in alignment of the horizontal plane. Data was collected using the Perimed supplied software, Perisoft for Windows Version 2.10.  Figure  Figure A.  A.7  A.9: Installation tool for placing laser doppler probe in probe holder.  10: Probe holder and vertical alignment device for positioning laser doppler probe.  Artificial Cerebral Spinal Fluid  Artificial CSF was used to maintain a blood free environment for the purpose of blood flow measurement. Fluid flow was manually controlled from a modified IV burette elevated above the level of the support frame (Figure A. 11). The end of the IV line was attached to a section of nylon tubing housing the secondary heating coil and connected to a 26 gauge biopsy needle. The needle was used to control the location of where the CSF was released into the surgically exposed area.  Figure A. 11:  IV burette used for control of CSF flow rate. 127  Appendix B Injury Protocols  B.l  P I D Settings Table B.T. PID settings for linear electromagnetic actuator. PID Settings (P)roportional 0.85 (I)ntegral 0.006 (D)erivative -2.27 (M)ass (C)ompensation 0.159  128  Appendix B.  Injury Protocols  B.2 Pure Contusion Table B.2:  Step: 1 2 3 4 5 6  Description: Dwell Ramp Dwell Ramp Ramp Dwell  Table  Step: 1 2 3 4 5 6  Waveform settings for pure contusion injury. Waveform Setting: 0.5 s 2 mm/s to 2 mm 8.5 s 750 mm/s to -1.73 mm 250 mm/s to 4 mm 3.472 s  Total Elapsed Time: 0.5 1.5 10 10.005 10.028 13.5  B.3: Data acquisition settings for pure contusion injury.  Acquisition Sample Time: Sample Points: Total Samples: 0\02 160 675 Continue Continue Continue Continue Continue  129  Time Between Samples: 0T02  Appendix B.  B.3  Injury Protocols  Contusion &; 40% Residual Compression Table B.4'- Waveform settings for contusion & 40% residual compression injury.  Step: 1 2 3 4 5 6 7 8 9  Description: Dwell Ramp Dwell Ramp Ramp Dwell Dwell Dwell Ramp  Waveform Setting: 0.5 s 2 mm/s to 2 mm 8.5 s 750 mm/s to -1.73 mm 250 mm/s to -0.4 mm 3.49 s 586.5 3000 250 mm/s to 4 mm  Data acquisition settings for contusion & 40% residual compression injury. Acquisition Sample Time: Sample Points: Total Samples: Time Between Samples: 0.02 160 675 0.02 Continue Continue Continue Continue Continue 0.5 50 1173 0.5 20 20 150 20 0.5 50 1 0.5  Table B.5:  Step: 1 2 3 4 5 6 7 8 9  Total Elapsed Time: 0.5 1.5 10 10.005 10.01 13.5 600 3600 3600.02  130  Appendix B.  B.4  Injury Protocols  Contusion &; 90% Residual Compression Table B.6: Waveform settings for contusion & 90% residual compression injury.  Step: 1 2 3 4 5 6 7 8 9  Description: Dwell Ramp Dwell Ramp Ramp Dwell Dwell Dwell Ramp  Waveform Setting: 0.5 s 2 mm/s to 2 mm 8.5 s 750 mm/s to -1.73 mm 250 mm/s to -0.9 mm 3.492 s 586.5 3000 250 mm/s to 4 mm  Data acquisition settings for contusion & 90% residual compression injury. Acquisition Sample Time: Sample Points: Total Samples: Time Between Samples: 0.02 160 675 0.02 Continue Continue Continue Continue Continue 0.5 50 1173 0.5 20 20 150 20 0.5 0.5 50 1  Table B.7:  Step: 1 2 3 4 5 6 7 8 9  Total Elapsed Time: 0.5 1.5 10 10.005 10.008 13.5 600 3600 3600.02  131  Appendix C Statistical Results  C.l  Mechanical Parameters Table C.l: Univariate Tests of Significance - Preload Force Intercept Group Error  SS  Degr. Freedom  MS  F  0.001678 0.000039 0.000418  1 3 25  0.001678 0.000013 0.000017  100.3439 0.7681  P 0.000000 0.522687  Table C.2: Univariate Tests of Significance - Contusion Injury Parameters Parameter Peak Displacement Peak Force  Peak Velocity  SS  Degr. Freedom  MS  F  23.58491 0.00153 0.00029 53.75273 0.16611 1.01400 11532987 62 1091  1 2 20 1 2 20 1 2 20  23.58491 0.00076 0.00001 53.75273 0.08306 0.05070 11532987 31 55  1601886 52  P 0.000000 0.000000  1060.208 1.638  0.000000 0.219353  211367.7 0.6  0.000000 0.574898  Table C.3: Newman-Keuls Multiple Comparison Test - Peak Displacement Group  Contusion  40% R e s i d u a l  90% R e s i d u a l  Contusion 0.531775 0.000145 40% Residual 0.531775 0.000151 90% Residual 0.000145 0.000151 Approximate Probabilities for Post Hoc Tests Error: Between MS = .00001, df = 20.000  132  Appendix C. Statistical Results  Table C.4- Kruskal-Wallis A N O V A by Ranks - Load Relaxation % Peak Force Peak Force  75%  50%  40%  30%  20%  10%  5%  Group Contusion 40% Residual 90% Residual Contusion 40% Residual 90% Residual Contusion 40% Residual 90% Residual Contusion 40% Residual 90% Residual Contusion 40% Residual 90% Residual Contusion 40% Residual 90% Residual Contusion 40% Residual 90% Residual Contusion 40% Residual 90% Residual  Code 102 103 104 102 103 104 102 103 104 102 103 104 102 103 104 102 103 104 102 103 104 102 103 104  N 8 8 7 8 8 7 8 8 7 8 8 7 8 8 7 8 8 7 8 8 7 8 8 7  133  Test Statistic  Sum Ranks 96.00000 96.00000 84.00000 86.0000 87.5000 102.5000 72.0000 64.5000 139.5000 73.5000 62.5000 140.0000 69.5000 66.5000 140.0000 68.5000 67.5000 140.0000 41.0000 97.0000 138.0000 ' 36.0000 100.0000 140.0000  H  H  H  H  H  H  H  H  Kruskal-Wallis test: ( 2, N = 23) =0.000000 p =1.000 Kruskal-Wallis test: ( 2, N = 23) =1.570533 p =.4560 Kruskal-Wallis test: ( 2, N = 23) =14.00047 p =.0009 Kruskal-Wallis test: ( 2, N = 23) =14.34872 p =.0008 Kruskal-Wallis test: ( 2, N = 23) =14.12388 p =.0009 Kruskal-Wallis test: ( 2, N = 23) =14.21201 p =.0008 Kruskal-Wallis test: ( 2, N = 23) =17.35590 p =.0002 Kruskal-Wallis test: ( 2, N = 23) =19.64286 p =.0001  Appendix C. Statistical Results  Table C.5: Multiple Comparisons p values (2-tailed) - Load Relaxation % Peak Force 50%  40%  30%  20%  10%  5%  Group Contusion 40% Relaxation 90% Residual Contusion 40% Relaxation 90% Residual Contusion 40% Relaxation 90% Residual Contusion 40% Relaxation 90% Residual Contusion 40% Relaxation 90% Residual Contusion 40% Relaxation 90% Residual  Contusion 1.000000 0.005549 1.000000 0.006203 1.000000 0.003809 1.000000 0.003362 0.117000 0.000097 0.054962 0.000030  134  40% R e s i d u a l  90% R e s i d u a l  1.000000  0.005549 0.002171  0.002171 1.000000 0.001550 1.000000 0.002609 1.000000 0.002963 0.117000 0.091838 0.054962 0.097883  0.006203 0.001550 0.003809 0.002609 0.003362 0.002963 0.000097 0.091838 0.000030 0.097883  Appendix C. Statistical Results  C.2  H&E Repeated Measures Analysis of Variance - H&E SS Degr. Freedom MS F P 142.2920 0.000000 900.4301 900.4301 1 2 32.6957 5.1668 0.015526 65.3914 20 6.3280 126.5609 0.9706 1 0.9706 0.4012 0.533657 0.991705 0.0202 0.0083 0.0403 2 20 2.4195 48.3891  Table C. 6:  Intercept Group: Error RI Rl*Group: Error  Table C.T. Newman-Keuls Multiple Comparison Test - H&E Injury Groups Group Contusion 40% Residual 90% Residual Contusion 0.014239 0,775284 40% Residual 0.014239 0.019510 90% Residual 0.775284 0.019510 Approximate Probabilities for Post Hoc Tests Error: Between MS = 6.3280, df = 20.000  C.3  NeuN & F J •-Table-C-.8:  Intercept Group: Location Group:*Location Error Direction Direction*Group: Direction*Location Direction*Group:*Location Error  Repeated Measures-Analysis of- Variance- NeuN SS Degr. Freedom 1079.714 1 2 102.393 2 9.275 4 6.515 99.981 60 . 0.288 1 2 1.832 2 1.046 4 1.473 35.457 60  135  MS 1079.714 51.197 4.637 1.629 1.666 0.288 0.916 0.523 0.368 0.591  F 647.9532 30.7238 2.7830 0.9774  P 0.000000 0.000000 0.069850 0.426707  0.4873 1.5499 0.8848 0.6230  0.487844 0.220644 0.418104 0.647896  Appendix C. Statistical Results  Newman-Keuls Multiple Comparison Test - NeuN Injury Groups Group Contusion 40% Residual 90% Residual Contusion 0.000109 0.313976 40% Residual 0.000109 0.000117 90% Residual 0.313976 0.000117 Approximate Probabilities for Post Hoc Tests Error: Between MS = 1.6663, df = 60.000 Table C.9:  Repeated Measures Analysis of Variance - Fluoro-Jade Degr. of Freedom MS F SS P 1459.674 754.8975 0.000000 Intercept 1459.674 1 2 39.560 20.4592 0.000000 Group: 79.120 4.304 2.2259 0.116811 Location 8.608 2 1.326 0.6859 0.604506 Group:*Location 5.305 4 Error 116.016 60 1.934 Direction 3.491 1 3.491 2.8992 0.093800 0.3739 0.689609 0.901 2 0.450 Direction*Group: 0.217 0.1806 0.835258 Direction*Location 0.435 2 0.204 0.1692 0.953269 Direction*Group:*Location 0.815 4 1.204 72.256 60 Error Table CAO:  Table CAT. Newman-Keuls Multiple Comparison Test - FluoroJade Injury Groups Group Contusion 40% Residual 90% Residual Contusion 0.000118 0.165132 40% Residual 0.000118 0.000117 90% Residual 0.165132 0.000117 Approximate Probabilities for Post Hoc Tests Error: Between MS = 1.9336, df = 60.000  136  Appendix C. Statistical Results  CA  (3-APP Table CA2: Repeated Measures Analysis of Variance - /3-APP Intercept Group Location Group*Location Error RI Rl*Group Rl*Location Rl*Group*Location Error  SS  Degr. Freedom  MS  F  0.394537 0.026210 0.075076 0.012651 0.136861 0.389811 0.152964 0.179255 0.132957 0.572101  1 2 3 6 64 9 18 27 54 576  0.394537 0.013105 0.025025 0.002109 0.002138 0.043312 0.008498 0.006639 0.002462 0.000993  184.4963 6.1282 11.7025 0.9860  P 0.000000 0.003672 0.000003 0.442256  43.6075 8.5559 6.6843 2.4790  0.000000 0.000000 0.000000 0.000000  Table CA3: Newman-Keuls Multiple Comparison Test - /3-APP Injury Group Group C o n t u s i o n 40% R e s i d u a l 90% R e s i d u a l Contusion 0.003702 0.503639 40% Residual 0.003702 0.009038 90% Residual 0.503639 0.009038 Approximate Probabilities for Post Hoc Tests Error: Between MS = 1.6663, df = 60.000 Table C.14'- Newman-Keuls Multiple Comparison Test - /3-APP Region N e w m a n - K e u l s M u l t i p l e C o m p a r i s o n Test - b - A P P R e g i o n Group Dorsal V e n t r a l Ventrolateral Lateral Dorsal 0.094334 0.045243 0.000787 Ventral 0.094334 0.001226 0.000155 Ventrolateral 0.045243 0.001226 0.068645 Lateral 0.000787 0.000155 0.068645 Approximate Probabilities for Post Hoc Tests Error: Between MS = 1.6663, df = 60.000  137  Newman-Keuls Multiple Comparison Test - /3-APP Extent -1 1 2 -4 -3 -2 0.829259 0.429631 0.000055 0.000010 0.000012 0.000032 0.442803 0.000083 0.000032 0.000010 0.000026 0.005600 0.000017 0.000020 0.000008 0.442803 0.000022 0.000008 0.001860 0.000083 0.005600 0,497838 0.000525 0.000032 0.000017 0.000022 0.000117 0.000010 0.000020 0.000008 0.497838 0.000026 0.000008 0.001860 0.000525 0.000117 0.001426 0.021524 0.426169 0.000008 0.000017 0.000290 0.624798 0.570907 0.001447 0.000020 0.000026 0.000017 0.883381 0.370514 0.000103 0.000026 0.000032 0.000020 Approximate Probabilities for Post Hoc Tests Error: Within MS = .00099, df = 576.00  Table C.15:  Distance -5 -4 -3 -2 -1 1 2 3 4 5  -5 0.829259 0.429631 0.000055 0.000010 0.000012 0.000032 0.000885 0.664619 0.930182  from Epicentre 3 4 0.000885 0.664619 0.001426 0.624798 0.021524 0.570907 0.426169 0.001447 0.000008 0.000020 0.000017 0.000026 0.000290 0.000017 0.011603 0.011603 0.001549 0.436717  5 0.930182 0.883381 0.370514 0.000103 0.000026 0.000032 0.000020 0.001549 0.436717  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0080765/manifest

Comment

Related Items