Mechanical Regulation of Burn Wound Scarring DISSERTATION

Mechanical Regulation of Burn Wound Scarring

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Jayne Y. Kim, MS

Graduate Program in Biomedical Engineering

The Ohio State University

2015

Dissertation Committee:

Heather M. Powell, PhD, Advisor

Douglas Kniss, PhD

Alan Litsky, MD, ScD

J. Kevin Bailey, MD

William Notz, PhD

Jayne Y. Kim

2015

Abstract

Hypertrophic scars (HTS) are common following cutaneous injuries with up to 91% of burn wounds resulting in HTS and can lead to significant loss of mobility and function for patients. HTS may also be complicated by erythema, hyperpigmentation, pruritus, ulceration, burning and secondary infection, further reducing the quality of life for patients. Although pressure garment therapy (PGT) has been used clinically for many decades, there remain some issues that need to be further investigated: 1) creation of reproducible, uniform, full-thickness burn wounds to reduce variability in observed results that occur with burn depth, 2) controversy remaining regarding efficacy, 3) loss of tension and ability to deliver adequate pressure by garments, 4) validation of appropriate HTS model in large animal and 5) determination of optimal time for pressure application.

To investigate the issue with uniform wound creation, a custom burn device was developed with an electrically heated burn stylus and a temperature control feedback loop via an electronic microstat and compared to a standard burner. The custom burn device was able to continually heat the burn stylus and actively control pressure and temperature, allowing for more rapid and reproducible burn wounds in comparison to the standard burner. ii

With the creation of uniform starting injury with the custom burner, the efficacy of

PGT was then evaluated in a female, Red Duroc pig (FRDP) burn model. PGT was effective in reducing scar contraction and improving biomechanics compared to control scars. However, these improvements were modest. Therefore, we sought to enhance the efficacy of PGT by optimizing the garment fabrication design for better pressure delivery.

In order to design more effective compression garments, a study was performed to investigate the mechanical and structural properties of pressure garment fabrics in order to design garments which deliver the most consistent pressure levels and maintain good durability after repeated wear and laundering. The results suggest

Powernet (a 9:1 nylon-Spandex composite fabric) as being more suitable than moleskin fabric for engineering compression garments.

Additionally, burn and dermatome models for hypertrophic scarring were developed and investigated. The results suggest that burn wounds provide a scar model in

FRDPs that most closely resemble hypertrophic scarring in humans. Using this knowledge, the efficacy of pressure garments fabricated from Powernet fabric was analyzed on a FRDP burn + split-thickness model. We found that early application

(day 7 post-injury) may be more beneficial in the treatment of scars than late application (day 35 post-injury).

iii

Dedication

This thesis is dedicated to my Father God (Phillipians 4:13), my parents Sang and In

Sook, my older brother Elliot, and my sisters Victoria and Sun.

Acknowledgments

First and foremost, I want to thank my advisor Dr. Heather M. Powell. I feel extremely blessed that I had the opportunity to be her PhD student. I am grateful for all of her contributions of time, advice and funding as well as her patience and encouragement during tough times. She has truly helped my PhD experience to be productive and stimulating. For everything you have done for me, I am sincerely thankful. I could not have made it without your guidance and friendship.

I am also grateful for the members of the Powell research group, especially for Dr.

Britani Blackstone and Danielle Dunham, who have been a source of friendship as well as good advice and collaboration. Other past and present group members that I have had the pleasure of working alongside are David Gutschick and Dr. Carol Lee. I would also like to thank my undergraduate assistants, J. Alex Clark and M. Jacob

Kresslein, for aiding in my research.

I would like to thank the members of my doctoral committee for their input, valuable discussion and accessibility: Dr. Douglas Kniss, Dr. Alan Litsky and Dr. J. Kevin

Bailey. In particular, I would like to thank Dr. J. Kevin Bailey, whom I had the

v pleasure of working with on various projects. Thank you for teaching me valuable skills in the operating room as well as providing encouragement and support. I have really enjoyed working with you.

I would like to thank Shriners Hospitals for Children in Cincinnati, Ohio for their collaborative work, especially Dr. Dorothy M. Supp and Kevin McFarland. Thank you for your valuable input and work. I am also grateful to the University Laboratory

Animal Resources staff for providing excellent care of our research animals during and after surgical procedures: Lori Mattox, Crystal Sims, Jeffrey Kim, Alisha Herriott,

Sarah Harp, Richardo Hairston and Alison Gallagher. I’d like to extend my thanks to the clinical veterinarians who provided their expertise: Dr. Valerie Bergdall, Dr.

Dondrae Coble and Dr. Stephanie Lewis.

Although my work with amputee patients was not included in this thesis, I’d like to thank Dr. Chandan Sen, Dr. Cameron Rink, Dr. Surya Gnyawali and Lynn Lambert for their collaborative work. I’d also like to thank Dr. Matthew Wernke, Dr. Jim

Colvin and Ryan Schroeder from Ohio Willow Wood.

Furthermore, I am very thankful that I had the opportunity to work for the Engineering

Education Innovation Center at the Ohio State University as the lead graduate teaching associate and graduate advising associate for the Green Engineering Scholars Program.

Special thanks to Dr. John Merrill, Dr. Phil Schlosser, and Elizabeth Riter. I truly enjoyed working with you.

I would also like to thank all of my friends. There are too many to name, but I’d like to especially thank Seiko Yamashita, Minyet Hua, Rachel Childers and Monica Okon.

You ladies have been my source of comfort during hard times and source of happiness all the time. I feel immensely blessed to have you in my life.

Lastly, I would like to thank my family. For my parents, Sang and In Sook, who raised me with unconditional love and support. For my older brother, Elliot, for being someone I can always look up to. And for my sisters, Victoria and Sun, for always being there to laugh, cry and most importantly, eat with. For my grandmother and grandfather who I lost during graduate school, I wish you could be here. Thank you for always cheering me on. And to all my other family members, thank you for loving me and for all your prayers.

vii

Vita

2003...... Sycamore High School, Cincinnati, Ohio

2006...... B.S. Biological Sciences, The Ohio State University

2013...... M.S. Biomedical Engineering, The Ohio State University

2012 to present ...... Graduate Research Associate, Department of Biomedical Engineering, The Ohio State University

2010 to 2012 ...... Graduate Teaching Associate, Engineering Education Innovation Center, First-Year Engineering Program, The Ohio State University

viii

Publications

Kim, JY, Willard, JJ, Supp, DM, Roy, S, Gordillo, GM, Sen, CK, and Powell, HM. Burn Scar Biomechanics Following Pressure Garment Therapy. Plast Reconstr Surg. May 8, 2015. (Epub Ahead of Print).

Gnyawali, SC, Barki, KG, Dixith, S, Kim, JY, Dickerson, JL, Mathew, SS, Powell, HM, Roy, S, Bergdall, V, and Sen, CK. High-resolution harmonics ultrasound imaging for non-invasive characterization of wound healing in a pre-clinical swine model. PLoS One. 2015;10(3):e0122327.

Kim, JY, Dunham, DM, Supp, DM, Sen, CK, and Powell, HM. Novel Burn Device for Rapid, Reproducible Burn Wound Generation. Burns. 2015. (Accepted)

Rink, CL, Wernke, MM, Gnyawali, SC, Schroeder, RM, Kim, JY, Powell, HM, Colvin, JM, and Sen, CK. Elevated Vacuum Suspension Attenuates Reactive Hyperemia and Improves Residual Limb Perfusion. PLoS One. 2015. (In Revision)

Dent A, Kim JY, Powell HM, Szalacha L, Landers T. Physical and microbiotic properties of nitrile vs. latex in glove juice sampling. (Manuscript in preparation)

Fields of Study

Major Field: Biomedical Engineering

Table of Contents

Abstract ...... ii Dedication ...... iv Acknowledgments...... v Vita ...... viii List of Figures ...... xiv Chapter 1: Introduction ...... 1 1.1 Skin Structure and Function ...... 1 1.2 Wound Healing and Scar Formation ...... 4 1.3 Clinical Problem ...... 6 1.3.1 Burn Injury ...... 6 1.3.1.1 The Body's Response to Burns ...... 7 1.4 Scar Management ...... 10 1.4.1 Surgical Management ...... 10 1.4.2 Non-Surgical Management ...... 12 1.4.2.1 Silicone Dressings ...... 12 1.4.2.2 Pharmacologic Treatments ...... 13 1.4.2.3 Pressure Garment Therapy ...... 13 1.4.2.4 Laser Therapy ...... 15 1.5 In Vitro Models of Abnormal Scarring ...... 18 1.5.1 Female, Red Duroc Pigs ...... 19 1.6 Summary ...... 20 Chapter 2: Novel Burn Device for Rapid, Reproducible Burn Wound Generation ...... 22 2.1 Introduction ...... 22 2.2 Materials and Methods ...... 24 2.2.1 Burn Devices ...... 24 x

2.2.2 Temperature Logging System ...... 25 2.2.3 Burn Wound Creation, Anesthesia, Animal Care ...... 26 2.2.4 Histology ...... 28 2.3 Results ...... 29 2.3.1 Initial Heating Time ...... 29 2.3.2 Stylus and Skin Surface Temperature during Use ...... 30 2.3.3 Elapsed Time ...... 32 2.3.4 Burn Depth as a Function of Time, Pressure and Device ...... 33 2.4 Discussion ...... 37 2.5 Conclusion ...... 40 2.4 Acknowledgements ...... 40 Chapter 3: Hypertrophic Scar Biomechanics Following Pressure Garment Therapy ...... 41 3.1 Introduction ...... 41 3.2 Materials and Methods ...... 44 3.2.1 Burn Wound Generation ...... 44 3.2.2 Pressure Garment Therapy and Pressure Quantification ...... 45 3.2.3 Scar Contraction ...... 45 3.2.4 Laser Doppler ...... 46 3.2.5 Scar Biomechancis ...... 47 3.2.6 Immunohistochemistry ...... 47 3.2.7 Transmission Electron Microscopy ...... 48 3.2.8 Statistics ...... 48 3.3 Results ...... 49 3.3.1 Full-Thickness Wound Generation and Pressure Garment Fabrication ...... 49 3.3.2 Scar Morphology and Contraction ...... 50 3.3.3 Scar Biomechanics ...... 52 3.3.4 Scar Perfusion and Blood Vessel Density ...... 54 3.3.5 Scar Structure and Collagen Organization ...... 55 3.4 Discussion ...... 57 3.5 Conclusions ...... 59 Chapter 4: Structural, Chemical and Mechanical Properties of Pressure Garments as a Function of Use and Laundering ...... 61 4.1 Introduction ...... 61 xi

4.2 Materials and Methods ...... 63 4.2.1 Fabric Material ...... 63 4.2.2 Laundering of Vest and Fabric ...... 64 4.2.3 Pressure Quantification of Custom-made Compression Vests ...... 64 4.2.4 Mechanical Testing of Fabric Samples ...... 66 4.2.4.1 Load Relaxation Testing ...... 66 4.2.4.2 Cyclic Testing ...... 67 4.2.5 Scanning Electron Microscopy (SEM) ...... 68 4.2.6 Attenuated Total Reflectance - Fourier Transform Infrared Spectroscopy (FTIR) ...... 68 4.3 Results ...... 69 4.3.1 Pressure Quantification of Custom-made Compression Vests ...... 69 4.3.2 Mechanical Analysis of Fabric ...... 71 4.3.3 Cyclic Testing ...... 74 4.3.4 Scanning Electron Microscopy (SEM) ...... 76 4.3.5 FTIR ...... 76 4.4 Discussion ...... 78 4.5 Conclusion ...... 80 Chapter 5: Female Red Duroc Pig as an Animal Model for Hypertrophic Scarring ...... 81 5.1 Introduction ...... 81 5.2 Materials and Methods ...... 84 5.2.1 Burn Injury and Animal Care ...... 84 5.2.2 Wound Closure and Transepidermal Water Loss ...... 86 5.2.3 Scar Contraction ...... 87 5.2.4 Scar Morphology and Structure ...... 87 5.2.5 Scar Thickness ...... 87 5.2.6 Scar Biomechanics ...... 88 5.2.7 Statistics ...... 88 5.3 Results ...... 89 5.3.1 Rate of Wound Healing ...... 89 5.3.2 Scar Thickness ...... 90 5.3.3 Scar Contraction ...... 92 5.3.4 Scar Morphology and Structure ...... 93 xii

5.3.5 Scar Biomechanics ...... 96 5.4 Discussion ...... 97 5.4.1 Wound Closure ...... 98 5.4.2 Clinical and Histological Appearance ...... 99 5.4.3 Scar Biomechanics ...... 100 5.5 Conclusions ...... 100 Chapter 6: Early versus Late Application of Pressure Garment Therapy ...... 102 6.1 Introduction ...... 102 6.2 Materials and Methods ...... 103 6.2.1 General Animal Care, Generation and Excision of Burn Wounds ...... 103 6.2.2 Split-Thickness Autograft Harvest, Meshing and Application ...... 105 6.2.3 Bolster and Dressing ...... 105 6.2.4 Percent Engraftment ...... 106 6.2.5 Pressure Garment Therapy and Pressure Quantification ...... 107 6.2.6 Scar Morphology and Contraction ...... 108 6.2.7 Scar Thickness ...... 108 6.2.8 Scar Biomechanics ...... 109 6.3 Results ...... 109 6.3.1 Percent Engraftment ...... 109 6.3.2 Scar Morphology and Contraction ...... 110 6.3.3 Scar Thickness ...... 112 6.3.4 Scar Biomechanics ...... 113 6.4 Discussion ...... 115 6.5 Conclusions ...... 115 Chapter 7: Summary and Future Work ...... 117 7.1 Future Work ...... 118 7.1.1 Tension Off-Loading and Scar Development ...... 118 7.1.2 Optimal Pressure Range (Pressure Garment Therapy) ...... 121 7.1.3 Laser Therapy ...... 121 References ...... 122

xiii

List of Figures

Figure 1.1. From deep to superficial, the epidermis is further divided into four functional layers: stratum basale, statum spinosum, stratum granulosum and stratum corneum. Lower portions project into the dermis with rete ridges ...... 2

Figure 1.2. The dermis can be divided into thin, superficial papillary layer and thick, deep reticular layer. The papillary dermis connects with the epidermis and contains thin loose collagen fibers, while thicker bundles of collagen are present in the deeper reticular layer ...... 3

Figure 1.3. Z-plasty: A) Two triangular flaps with three incisions of equal length are created with varying degrees of angulation (30-75°, with 60° being most common) to the central axis. B) The flaps are then moved into new positions to lengthen the scar. C) Lastly, the flaps are interlocked and the wound is closed...... 12

Figure 1.4. A) Fractional ablative laser therapy for the treatment of scars involves ablating micro-holes or columns into the scar tissue. The damage created by the laser causes regrowth of collagen and a more rapid healing response. B) Patient suffers from scarring as a result of second and third-degree burns. C) A noticeable improvement in the texture and appearance of the scar is seen post-treatment with fractional ablative laser therapy ...... 16

Figure 2.1. Photograph of A) standard burn stylus heated to temperature using a laboratory hotplate (inset: close up of stylus and port for thermocouple) and B) a custom designed burn device with an electrically controlled internal heater and pressure gauge (upper inset: close up of burn stylus with thermocouple in port, lower inset: electronically controlled microstat) ...... 25

Figure 2.2. Time required to bring each burn stylus to temperature (200°C) ...... 29

Figure 2.3. Time-temperature traces for both the inside of the burn stylus (A) and the skin surface (B). Note that the internal temperature of the burner decreases linearly at a high rate in the standard device whereas the custom device held temperature to within 7 degrees throughout contact with the skin surface. The external skin temperature was rapidly raised to 97ºC using the custom burn device and held constant xiv at this temperature until the device was removed. In contrast, the standard device required between 7 and 20 seconds to reach peak temperature and had only a short period of steady surface temperature...... 31

Figure 2.4. Elapsed time between subsequent burns ...... 32

Figure 2.5. Representative H&E stained sections of burns generated using the standard and custom device heated to 200°C and applied to the skin at 3 lbs of pressure for 5, 10, 20 or 40 seconds...... 33

Figure 2.6. Representative H&E stained sections of burns generated using the standard and custom device heated to 200°C and applied to the skin for 40 seconds at either 1 or 3 lbs of pressure ...... 34

Figure 2.7. H&E stained histological section of a burn injury showing methodology for quantifying burn depth. Burn depth quantified by measuring distance from top of tissue section to the depth where distinct collagen fibers can be visualized ...... 35

Figure 2.8. Box and whisker plot of burn injury depth using the standard (A&C) and custom (B&D) device. Burn depth increased as a function of contact time while pressure is held at 3 lbs (A&B). Burn depth increased with increasing pressure during a 40 sec contact time ...... 37

Figure 3.1. A) Photograph of female Red Duroc pig immediately after burn wound generation. After wounds heal for 28 days, compression garments were applied at a reduction in circumference of 10% for the treatment group and 0% for the control group (B). C) Quantification of pressure generated by a compression garment tailored to a 10% reduction in circumference. D) Histological section of burn wound 7 days after initial thermal injury ...... 46

Figure 3.2. Representative photographs of the scars 28-78 days post burn. Compression garments were applied at day 28...... 50

Figure 3.3. Scar contraction, presented as percent of original area, as a function of time and treatment. As compression garments were applied at day 28, all area measurements are normalized to the scar area at this time point. After 28 days of garment application (56 days post burn), compression garment treated scars were significantly less contracted than control scars ...... 51

Figure 3.4. Scar mechanics at day 78 post burn. A) Ultimate tensile strength and B) linear stiffness of compression garment treated wounds were significantly higher than control wounds. C) Torsional ballistometry of control scars, PGT treated scars and normal pig skin showed an increase in probe indentation and elasticity compared to control scars ...... 53 xv

Figure 3.5. A) Laser Doppler imaging of scars 78 days post burn showing no difference in blood flow between control and pressure garment treated groups. B) Immunostaining of endothelial cells (VWF) within burn scars at day 78. Quantification of endothelial cells density showed no significant difference in blood vessel density between control and pressure garment treatment ...... 54

Figure 3.6. H&E stained histological section of control (A&C) and pressure garment treated (PGT) burn scars (B&D) 42 (A&B) and 78 (C&D) days post injury ...... 55

Figure 3.7. Transmission electron micrographs of collagen fibers in control and pressure garment treated burns scars 78 days after injury. Scale bar = 100 nm. Control scars contain large, loosely packed collagen fibers whereas PGT scars are more densely packed with smaller diameter collagen fibers. B) Histogram of collagen fiber diameter distribution...... 56

Figure 4.1. Moleskin fabric (A) and Powernet fabric (B) with finished edge labeled .... 64

Figure 4.2. A) Child size lifelike dummy used to take pressure measurements of compression vests as a function of time. Pressure measurements were obtained at 5 key anatomical locations: 1) upper left torso/chest, 2) lower right torso/abdomen, 3) lower right torso/lateral abdomen, 4) right shoulder, 5) upper central back via Kikuhime sensors. Image shows sensor secured at location 1. The Kikuhime air bladder transmits surface pressures (mmHg) to the data reporting device. B) Dummy wearing powernet compression vest with Kikuhime pressure sensors attached at 5 locations. C) Dummy wearing moleskin compression vest with Kikuhime pressure sensors attached at 5 locations...... 66

Figure 4.3. Hourly pressure measurements of custom-made powernet (A-C) and moleskin (D-F) compression vests worn by child size lifelike dummy (Figure 1) over the course of 23 hours before laundering (A and D), after 1 launder cycle (B and E), and after being launder 5X (C and F). Data suggests pressure loss in both garments with each successive launder cycle, as well as a decrease in pressure over time. After 5 washes, the moleskin compression vest displayed a greater pressure reduction compared to the powernet compression vest ...... 70

Figure 4.4. Powernet fabric generated on average 2 to 4 times the initial load and relaxed at a rapid pace (A-C), while moleskin fabric generated very small load at an initial strain of 10% (D-F)...... 73

Figure 4.5. Average overall energy dissipation (mJ/mm2) of unlaundered, laundered 1X and laundered 5X powernet and moleskin fabric samples cut at (A) 0˚ orientation, (B) 45˚ orientation and (C) 90˚ orientation ...... 75

xvi

Figure 4.6. Gold plated SEM images of Powernet and moleskin fabric at each of the three launder cycles. No fiber structure or orientation change was found between each launder cycle for both fabrics ...... 76

Figure 4.7. FTIR absorbance spectrum for moleskin fabric (A) and Powernet fabric (B). FTIR analysis showed the presence of nylon with Amide I and Amide II absorbance peaks at 1635 cm-1 and 1543 cm-1 in moleskin and the presence of spandex and nylon with absorbance peaks at (urea carbonyl) 1755 cm-1, (Amide I) 1635 cm-1 and (Amide II) 1535 cm-1 in Powernet fabric ...... 78

Figure 5.1. At day 0 post-injury: Burn wounds (A) and dermatome wounds with TDS ~ 0.060 in (B) and TDS > 0.075 in (C). Scale bar = 10mm ...... 86

Figure 5.2. Average transepidermal water loss (TEWL, g/m2/hr) at day 28 post- wounding for normal skin, burn wounds, dermatome wounds with a TDS > 0.075 in and dermatome wounds with a TDS approximately 0.060 in. Average TEWL values of burn wounds were significantly greater than normal skin and both dermatome wounds...... 90

Figure 5.3. Average scar thickness (mm) at day 150 post-wounding for normal skin, burn wounds, dermatome wounds with A TDS > 0.075 in and dermatome wounds with A TDS approximately 0.060 in. The average thicknesses of burn scars were statistically greater than dermatome scars with a TDS > 0.075 and dermatome wounds with a TDS approximately 0.060 ...... 91

Figure 5.4. Average scar contraction (% original area) of wounds created via burner and dermatome. Contraction of burn scars were significantly greater than scars resulting from a dermatome. Wounds created with a dermatome at a TDS of approximately 0.060 in contracted very little ...... 93

Figure 5.5. Representative photographs of wounds and scars resulting from burn injury and dermatome wounding (TDS of ~ 0.06 in and > 0.075 in) 28, 90 and 150 days post wounding. Scale bar = 10mm ...... 95

Figure 5.6. Representative H&E stained histological section of scars resulting from burn injury and dermatome wounding (TDS of ~ 0.06 in and > 0.075 in) 90 and 150 days post wounding...... 96

Figure 5.7. Ultimate tensile strength of burn scars were significantly lower than dermatome scars (TDS > 0.075 in) and dermatome scars (TDS > 0.075 in) at day 150 post-wounding ...... 97

xvii

Figure 6.1. After 1 x 1 inch burn wounds were created with a custom burn device (200 + 10°C, 40 seconds, 3lbs of pressure) (A), burn sites were immediately excised to eschar level (B) and meshed split-thickness autografts were applied (C)...... 104

Figure 6.2. Bolsters were covered with a cotton bolster over a silver-based dressing (A), followed by cotton padding and fiberglass casting material (B) and finally Vetrap and Elastikon to secure the soft circumferential dressing (C)...... 106

Figure 6.3. Example wound sites with 100% engraftment ...... 110

Figure 6.4. Representative photographs of control and pressure treated scars 7-130 post injury. Compression garments were applied at day 7 and day 35. Scale bar = 10mm ...... 111

Figure 6.5. Scar contraction presented as percent of original area, as a function of time and treatment ...... 112

Figure 6.6. Thickness of pressure treated scars are significantly lower than control scars. There is also a significant difference in thickness of scars treated with pressure at day 7 in comparison to day 35 ...... 113

Figure 6.7. Scar biomechanics at day 130 post wounding. A) Linear stiffness and B) ultimate tensile strength of pressure treated scars were slightly greater than control scars. No statistical difference was observed ...... 114

Figure 7.1. Schematic diagram of proposed mechanisms of scar formation within the burn wound. (Modified from Gurtner et al.) ...... 119

Figure 7.2. Off-loading elevated stress in incisional wounds with a stress-shielding device resulted in decreased scar formation, epidermal thickening and rete ridge regeneration (From Gurtner et al.) ...... 120

xviii

Chapter 1: Introduction

1.1 Skin Structure and Function

The skin is the largest organ of the body, accounting for 16% of total body weight1-2.

Skin along with its derivatives (such as sebaceous glands, hair, nails and sweat glands) is responsible for several critical biological functions including providing a physical barrier to prevent infection and excessive water loss1-2. Human skin is comprised of three structural layers: the epidermis, the dermis and subcutaneous layer. Each of these layers is different in composition and organization and thus imparts different properties to the tissue as the whole.

The epidermis is the outermost layer of skin and serves as a physical and chemical barrier to the exterior. It is thinner than the dermis, and the lower portions project into the dermis with undulating structures called rete ridges1-2 (Figure 1.1). From deep to superficial, the epidermis is further divided into four functional layers: stratum basale, stratum spinosum, stratum granulosum and stratum corneum (Figure 1.1). Within the stratum basale, keratinocytes are anchored to the basement membrane by hemidesmosomes1-2. These basal keratinocytes have significant regenerative capacity,

1 continually proliferating over a human’s lifetime. As the basal cells divide and differentiate, they move towards the more superficial layers. In the stratum spinosum, cells produce larger amounts of cytokeratin and are strongly connected to each other via desmosomes1-2. Keratinocytes continue to flatten in the stratum granulosum and lose their nuclei. At the upper region of this layer, the keratinocytes begin to secrete lipid and protein-containing lamellar bodies into the extracellular space which leads to the development of the hydrophobic envelope and subsequent barrier properties1-2. The outermost stratum corneum is comprised of layers of corneocytes, which are non-viable cornified cells that are filled with water-retaining keratin and surrounded by protein. This structure provides a tough barrier for protection and the maintenance of moisture1-2.

Rete ridge

The second layer of the skin, the dermis, can be divided into a thin, superficial papillary layer and a thicker, deep reticular layer1-2 (Figure 1.2). The papillary dermis connects with the epidermis and contains thin loose collagen fibers, while thicker bundles of collagen are present in the deeper reticular layer (Figure 1.2). The dermis contains mast cells, macrophages and fibroblasts, which produce the extracellular matrix of collagen, elastin and proteoglycans. Collagen type I fibers provide structural integrity, while elastin and proteoglycans provide elasticity and maintain hydration, respectively1-2. Hair follicles, sweat glands, nerves and lymphatic vessels are also embedded in the fibrous tissue of the dermal layer. The third and deepest layer is the subcutaneous layer, also known as subcutis. This layer is mainly comprised of fat and loos connective tissue, providing a heat source and protection from injury.

1.2 Wound Healing and Scar Formation

Skin provides numerous protective functions of our body and thus injuries to integument must be repaired rapidly. As described above, the layers of skin provide different function yet are all linked structurally and biologically. Thus, healing cutaneous injuries requires a complex, coordinated response between multiple cell types. Wounds, such as lacerations, surgical incisions and burns, compromise the function of skin and trigger complex and dynamic biological processes in an attempt to restore skin integrity. Normal wound healing can be divided into three classic stages: inflammation, cell proliferation and matrix remodeling. Immediately following injury, platelet degradation and clotting cascades are activated (conversion of fibrinogen into fibrin by the action of thrombin) to form a fibrin clot, which acts as a temporary scaffold for wound repair3-4. Platelet degranulation causes the release of cytokines, such as platelet-derived growth factor

(PDGF), transforming growth factor (TGF-β) and insulin-growth factor (IGF-1), leading to the recruitment of macrophages, neutrophils and mast cells3-4. In addition, the released

TGF-β1 and IGF-1 stimulates collagen production in dermal fibroblasts5. Within two to five days post-injury, the wound healing process progresses into the proliferation phase.

Fibroblasts synthesize granulation tissue, which serves as a framework to bridge the wound and allow vascular ingrowth3,6. This granulation tissue is comprised of elastin, proteoglycans, procollagen and hyaluronic acid and is highly vascularized. Within the granulation tissue, fibroblasts differentiate into actin-containing myofibroblasts that aid in wound contraction3. The proliferation phase may last anywhere from three to six weeks6,

4 dependent on the size and severity of the wound. The final remodeling phase may last several months and involves the degradation of the extracellular matrix (ECM) and the conversion of immature type I collagen to mature type III collagen3,6. During the maturation process, the healed wound flattens and assumes a pigmentation that is similar to normal skin.

In non-fetal humans, the normal outcome to an injury is some level of scarring. When the scar is flat with no change in pigmentation, it is termed normotrophic or normal scarring and this represents the most favorable outcome following an injury7.

Unfortunately, in some individuals, alterations in the wound healing process result in abnormal scar formation7-9. In these individuals, prolonged inflammation and fibroblast proliferation result in excessive collagen deposition, collagen bundling and alignment and a shift in the composition of collagen7-9. This leads the scars to be raised, and in the case of keloid scarring, the scar tissue will extend further than the original boundary of the injury7-9.

Though all forms of scarring may be debilitating, hypertrophic scars (HTS) are common following cutaneous injuries with up to 91% of burn wounds resulting in HTS3 and can lead to significant loss of mobility and function for patients3,10. Hypertrophic scars by definition represent an exaggerated proliferative response to wound healing where contraction and rapid growth of connective tissue are carried to excess10-12. There is excessive deposition of collagen fibers in (increased type III to type I ratio in HTS

5 compared to normal skin), highly aligned in orientation8,13-14. The scar tissue is also densely populated with myofibroblasts, which are responsible for the contractile activity during the remodeling phase10-12. HTS are characterized by poor rete ridge regeneration and thicker, aligned collagen bundles leaving the tissue weaker than normal skin. In addition, throughout the entire scar maturation period, which could be 2 years or more, the scar is actively contracting10,15. Excessive scar contracture greatly reduces skin pliability and results in deformity and loss of function in the affected area, diminishing the quality of life for patients with HTS16-18. A hypertrophic scar may also be complicated by erythema, hyperpigmentation, pruritus, ulceration, burning and secondary infection, further reducing the quality of life for patients19-20. The erythema frequently present in hypertrophic scars can be attributed to residual dilated blood vessels from early wound repair21.

1.3 Clinical Problem

1.3.1 Burn Injury

Burns are not only physically debilitating, but are psychologically devastating for the survivor. The injury represents a diverse challenge to medical staff, as it affects a wide demographic and all ages. Burns account for an estimated 490,000 injuries receiving medical treatment in the United States each year22 and can be the result of flame injuries, scald burns and, to a lesser extent, electrical and chemical burns. 50% of adult burns are

6 due to flame injuries23. They are often associated with inhalational injury and other concomitant trauma23. The types of burn injury are classified into three groups by depth of tissue and damage. Superficial or first-degree burns only affect the epidermis.

Superficial epidermal burns will heal spontaneously and generally do not require medical intervention24-25. When damage extends past the epidermis and into the dermis, it is a partial thickness or second-degree burn. Partial thickness burns can be further divided into superficial dermal and deep dermal. These burns are generally treated with topical antimicrobial agents, which commonly include silver-based ointments and dressings to reduce the incidence of burn wound sepsis, and occlusive dressings made of carboxymethylcellulose-based hydrofiber or arginine-glycine-aspartic acid (RGD) peptide matrix to promote wound healing25. A full thickness or third-degree burn penetrates through all skin layers into the subcutis24-25. Third-degree burns require a more aggressive treatment regimen with early tangential excision of the burn tissue following by wound closure via skin grafts25. Autografts from uninjured skin are generally used; however if there is a lack of adequate skin donor sites, cultured epithelial autografts are utilized25.

1.3.1.1 The Body’s Response to Burns

Thermal injuries result in immediate local responses in the body. Jackson first described three zones of burn injury in 1953. The zone of coagulation represents the area of the injury nearest the heat source and the point of maximum damage23. The tissue loss in this

7 zone is irreversible due to coagulation and denaturation of the constituent proteins23.

Surrounding this area is the zone of stasis, which is comprised of less damaged tissue and decreased tissue perfusion. Tissue in this zone has the potential to recover with proper treatment23. The outermost zone of hyperaemia represents tissue with the least amount of damage and increased tissue perfusion23. Tissue within this zone generally recovers completely unless complicated by infection. These three zones of a burn are three dimensional, and loss of tissue in the zone of stasis will lead to an increase in depth and width of the wound margins23. Tissue necrosis is further induced by the activation of toxic inflammatory mediators such as proteases and oxidants and the loss of fluid and proteins. Thus, it is important to promptly remove the tissue in the zone of coagulation and monitor the zone of stasis for signs of necrosis to prevent further tissue damage and increase in total area of tissue damaged.

Burns greater than 30% of total body surface area (TBSA) will not only result in direct damage to the local area due to the thermal injury but also trigger systemic responses within cardiovascular, respiratory, metabolic and immunologic systems23. The release of histamine and bradykinin immediately following burn injury results in increased capillary permeability within the burn wound, leading to fluid and protein loss into the extravascular compartment26-27. Histamine increases fluid leakage by increasing intracellular junction space in venules26-27. Bradykinins are also known to disrupt normal capillary barriers27. Additionally, myocardial contractility is decreased and splanchnic and peripheral vasoconstriction occurs23. Respiratory changes in the body involve

8 bronchoconstriction, which is caused by inflammatory mediators23. In severe burns, death can occur with the onset of acute respiratory distress syndrome (ARDS). ARDS is believed to occur when neutrophil accumulation in the microcirculation of the lung causes damage to vascular endothelium, leading to decreased lung compliance and difficulty exchanging air28. Two to seventeen percent of burn patients develop ARDS, and the percentages jump to 40-54% in patients requiring mechanical ventilation28.

Metabolic changes in the body include a decrease in glomerular filtration rate (GFR) and an increase in basal metabolic rate, up to three times its original rate23,29-30. Such a severe increase in metabolic rate results in decreases in lean body mass and can have serious consequences30. Ten percent loss of total body mass is associated with immune dysfunction, 20% loss correlates with decreased wound healing, 30% loss leads to increased risk of pneumonia and 40% loss of total body mass could lead to death30.

Severe thermal injuries induce a non-specific down regulation of the immune response that can result in subsequent sepsis and multiple organ failure23,29-30.

The combined local and systemic responses of the body to thermal injuries cause burns to be one of the most challenging injuries to treat. Advances in medicine have significantly increased survival after burn injury, especially in children. Today, virtually all children with a 90% or less total body surface area burn are expected to survive31-32, whereas just

20 years prior the estimated chance of survival would be less than 50%32. Concomitant with advances in survival, there has been a growing appreciation for the necessity to

9 improve the form and function of the healed wound, including the reduction and prevention of scar formation.

1.4 Scar Management

Prevention of a pathological event is commonly more effective than treating the outcome; however, the prevalence of severe scarring in the current population indicates that both solutions to prevent and treat scars are severely needed. Depending on the location and extent of the scarring, a number of treatment modalities may be utilized including pressure garments, silicone sheeting, topical and intralesion steroid, massage and laser therapy. Surgical excision of the scar or incisional release of tension may also be necessary.

1.4.1 Surgical Management

Efforts to prevent scar formation have relied largely on immediate excision of necrotic tissue and replacement with grafts. Skin grafts can be divided into subgroups dependent on the thickness. Split thickness skin grafts include only a portion of the dermis, while full thickness autografts include the entire dermal layer33. The thickness also determines the postoperative degree of wound contraction. Generally, the thicker grafts result in a lower degree of contraction33. Although, this technique has shown clinical benefit in some case studies, split thickness and full thickness autografts have disadvantages that must be considered. Split thickness grafts are more fragile and lack elasticity.

Furthermore, they tend to be abnormally pigmented and contraction tends to be more pronounced33. Although full thickness grafts tend to contract less, there still remains the problem of potential graft loss33. Graft loss is usually associated with poor graft adherence to the underlying debrided wound, nourishment deprivation resulting from low vascularity to the affected site or bacterial infection33.

Surgical techniques to treat hypertrophic scars include reducing scar tension by Z-plasty.

By creating two triangular flaps with three incisions of equal length, a Z-plasty can help lengthen a contracted scar, relax tension lines, interrupt a scar for better camouflage or change the direction of a scar (Figure 1.3)34-36. The angle from the central axis affects how much scar lengthening is achieved. The most common angulation is 60°, which creates a 75% scar length increase35-36. A challenge with this technique is that the gain in directional change or scar lengthening is almost always less than calculated34. This is because mathematical predictions of the angles at which to cut do not translate accurately when dealing with live skin because of factors such as tensile force and pliability that place constraints on skin34.

1.4.2 Non-Surgical Management

1.4.2.1 Silicone Dressings

Topical silicone materials are another option that has been used for the treatment of HTS since the early 1980s37-40. It is recommended for patients to apply silicone gel twice a day or wear silicone gel sheets for at least 12 hours daily for 2 to 12 months after the wound is healed39-41. The mechanism of action of this technique is not well understood, but it is believed to work by increasing hydration of the stratum corneum and protecting the scar tissue from bacterial invasion, thereby reducing collagen production in the treated area37,40-41. In one study where 14 HTS were treated with silicone gel sheeting for

12 to 24 hours daily for at least two months, scar were flattened with a 21% recurrence rate and minimal side effects38. Although other studies report similar benefits, a

Conchrane systematic review concluded that trials evaluating the efficacy of silicone gel sheeting for the prevention and treatment of HTS are highly susceptible to bias and of 12 poor quality42. Low patient compliance due to excessive sweating and skin breakdown, rashes and difficulty in application also attributes to difficulty assessing efficacy37,39.

1.4.2.2 Pharmacologic Treatments

Corticosteroid injections have been a mainstay nonsurgical option in the treatment of hypertrophic scars since the mid-1960s37,39-40,43-44. Triamcinolone acetonide (Kenalog) is injected intralesionally at a dose of 10 to 40 mg per mL depending on the size of the scar40,43-44. It is recommended to begin therapy as soon as the scar is identified, with two to six injections a month apart40. Corticosteroid injections reduce inflammation, fibroblast proliferation and collagen synthesis, while increasing vasoconstriction, thereby flattening and improving symptoms of hypertrophied scars39-40,44. One study reported complete flattening in 64% of Kenalog-treated scars and symptomatic improvement in 46 out of 65 patients45. Another study reported alleviation of itching and complete flattening in the majority of scars injected with Kenalog43. However, recurrence of scars was found in 50% of the cases after 5 years43. In addition to recurrence, common adverse effects include atrophy, hypopigmentation, telangiectasia and pain with injection40,44.

1.4.2.3 Pressure Garment Therapy

Pressure garment therapy (PGT) has been the preferred modality for the prevention and treatment of HTS for the past 45 years46-48. PGT involves the use of compression

13 garments that apply static pressure to the affected area of skin. Although the precise effect of this treatment on scar development and maturation has not been scientifically defined, it is widely believed that compression garments control and reduce collagen synthesis by limiting the supply of blood, oxygen and nutrients to the scar tissue10-11,46-

47,49. Therefore, it has been suggested that PGT can arrest or suppress the production of

HTS tissue46,50 and enhance the natural remodeling process51-52. Although PGT has been used clinically for over 40 years, controversy remains regarding its efficacy53. Prior clinical studies reported outcomes with pressure garment therapy ranging from no evidence-based benefit54 to significant increases in scar pliability55. Furthermore, the application of PGT also places high demands on patients. Not only must the garment be worn continuously for at least 23 hours a day, pressure must be applied until the scar is mature, which could take at least 6 months from wound closure and up to 2 years10,46,56.

According to clinical reports, only 41% of pressure garment wearers were fully compliant with the treatment10,46. During continuous wear, patients have reported restriction of movement and discomfort, such as pain, itching, blistering, ulceration, increased transpiration, or rashes10,46. Emotional distress, such as feeling self-conscious due to poor appearance of the garment has also been reported as contributing factors to low compliance10,46. Another problem is that the optimal pressure required for effective treatment has never been scientifically established46,57-58. Studies measuring actual pressure exerted by compression garments have recorded a wide range of results, leading to contradictory evidence regarding safe and effective pressures46,57.

1.4.2.4 Laser Therapy

To overcome the limitations of PGT and other options, new laser-based treatment modalities have been developed and used for the treatment of HTS for more than 25 years. Laser therapy is based on photothermolysis that can either be ablative or non- ablative, which results in tissue vaporization or coagulation, respectively, stimulating the denaturation of collagen, neocollagenesis and dermal remodeling59-60. Ablative fractional carbon dioxide laser resurfacing to aid in the correction of burn scar thickness, stiffness and abnormal texture and vascular-specific pulsed dye laser21 therapy have been introduced as possible treatment options for hypertrophic scars21,61.

Within the last decade, fractional resurfacing has become an increasingly popular treatment for photodamage and more recently has emerged into a therapeutic option for the aesthetic and functional repair of scar tissue62-63. Fractional lasers are able to ablate tissue at variable depths dependent on treatment settings (Figure 1.4A). Treatment settings generally include pattern size (mm2), fractional coverage, fluence (mJ) and density14. The laser targets intracellular water, resulting in the vaporization of tissue61.

The injury caused by fractional lasers triggers a molecular cascade comprised of heat shock proteins among other factors, leading to extended neocollagenesis with subsequent collagen remodeling and a more rapid healing response62-63. It is therefore believed that the mechanisms of improvement behind ablative fractional laser therapy may include the removal of a portion of fibrotic scar and a relative normalization of collagen structure and

15 composition14,63. In one case study, a single treatment session using an ablative fractional

62 CO2 laser was performed to treat second and third-degree burn scars (Figure 1.4B) .

After two passes (Deep FX mode on the first pass with a 12.5 mJ/cm2 fluence and density setting of 3, and Active FX mode on the second pass with a 80 mJ/cm2 fluence and density setting of 1), improvement in the texture and appearance of the scar could be observed (Figure 1.4C)62.

Figure 1.4. A) Fractional ablative laser therapy for the treatment of scars involves ablating micro-holes or columns into the scar tissue. The damage created by the laser causes regrowth of collagen and a more rapid healing response. (From http://www.abmedispa.com/CO2laser.htm) B) Patient suffers from scarring as a result of second and third-degree burns. C) A noticeable improvement in the texture and appearance of the scar is seen post-treatment with fractional ablative laser therapy. (From Waibel, 200962)

The vascular-specific flashlamp-pumped 585 and 595 nm pulsed dye lasers were first developed several decades ago to treat port wine stains and capillary malformations59 and has been used in the management of hypertrophic scars in the past decade21. The laser targets hemoglobin, coagulating microvasculature in the dermis (up to a depth of 1.2 mm)61. When administered to scars resulting from thermal injury, PDL serves to terminate the hypervascular response. The 585-nm pulsed dye laser has shown a high degree of success in treating hypertrophic scars resulting from burns by improving erythema and pliability without epithelial disruption61,64. The mechanisms behind the reduction of scar height and increase in pliability remain unknown, but many speculate that hemodynamic effects and cytokine production are likely to play a role in the interaction between PDL and scar tissue61.

Despite cases demonstrating the benefits of laser therapy, carbon dioxide laser ablation of hypertrophic scars and keloids has commonly led to scar recurrence and sometimes scar worsening21. One study team described five case studies where patients developed hypertrophic scarring of the neck after treatment with the carbon dioxide fractional system65-66. In another case study, a 30 year old female patient was treated with a single

10,600 nm carbon dioxide fractional laser treatment session66. Difference passes of laser were delivered to scar tissue with contracture and without contracture. Two passes with a laser fluence of 50 mJ (depth of 1500 μm) and a density of 5% coverage was first administered to areas of scar contracture using the Deep FX mode, while one pass with a laser fluence of 35 mJ (depth of 1050 μm) and a density of 10% coverage was delivered

17 to areas without contractur66. No evidence of blistering or secondary infection was seen immediately after treatment; however, after 30 days post-treatment, new hypertrophic nodules that were erythematous to brown in color were observed along the treated area66.

Accordingly, the efficacy of this type of laser therapy for the treatment of HTS remains unclear66.

1.5 In Vivo Models of Abnormal Scarring

An obstacle in researching scarring and developing more efficient therapeutic strategies is the lack of an in vivo model for investigation. Longitudinal studies in the human patient population lack controlled burn depth, size and location along with proper negative controls67-68. Another challenge is that in most cases these abnormal scars are specific to humans8. A major difference between humans and loose-skinned laboratory animals is the panniculus carnosus, which is a fibromuscular layer in animals. This layer enables the skin to slide over underlying fascia and facilitates rapid contraction and an accelerated healing time for thermal wounds8.

Many groups have attempted to generate animal models using various rodents. However, rodent models do not possess the complex mechanical and biological environment observed in human skin and do not naturally form hypertrophic scars3,69. A hypertrophic scar model was reported in the rabbit ear after a group of surgeons noticed that surgical scars in rabbit ears remained elevated months after wounding8,67,70. However, due to the structure of cartilaginous tissue underlying the rabbit ear epithelium, collagen maturation

18 and other histologic parameters found in human HTS cannot be studied in this model67,69.

Another scar model was reported in tight-skin mice (TSM), a mutant mouse strain that was able to shield large amounts of wound contraction that affected their loose-skinned counterparts8. Although hypertrophy of connective tissue and vessels were histologically similar to human HTS and abundant collagen fibers were observed, the scars were unable to sustain these characteristics for long periods of time8,71.

1.5.1 Female, Red Duroc Pigs

More recently, female, red Duroc pigs (FRDPs) have been proposed as a model for hypertrophic scarring5,67-68,72. Studies have confirmed that FRDPs form robust scars following deep cutaneous wounds and that these scars are similar in appearance to human

HTS. Studies have shown that FRDP scars were thick, raised above surrounding uninjured skin, hyperpigmented, firm, contracted and lacked hair46,67,. Additionally, the pattern of inflammatory cytokine expression is similar between humans and FRDPs.

TGF-β1 and IGF-1 are growth factors that are associated with tissue fibrosis5,73-74. In deep FRDP wounds, TGF- β1 mRNA expression and protein production increased at day

10 post-injury and returned back to baseline by day 1505,67. IGF-1 mRNA and protein were also significantly elevated early in these scars5,67. These observations are consistent with reports that TGF- β1 and IGF-1 expression is significantly greater (61% and 78%, respectively) in human HTS compared to normal skin67,73. Decorin is a small proteoglycan that is abundant in normal human dermis and has been found associated

19 with collagen fibril formation5,75. Decorin expression was greatly reduced for 90 days in deep FRDP wounds, which is similar to the long delay in expression seen in human hypertrophic tissue5. The expression of versican, a proteoglycan involved in fibrosis, was seen to increase from days 30 to 150 post-injury in deep FRDP wounds5. The increase in versican expression is also seen in human HTS tissue5,76. These structural and biological similarities between human HTS and the thick scars on FRDPs provide a new platform for the study of scar development and anti-scar therapies.

Despite the challenges associated with wound repair and developing clinically relevant models for investigation, the female red Duroc pig may represent an important model for creating hypertrophic scars, where wound depth and location can be controlled and patient compliance can be tightly monitored. This model allows for the study of current anti-scar therapies and for the development of new strategies with higher clinical benefit.

1.6 Summary

The human integument plays an important physiological role in human health and injury to this organ can have devastating consequences. As significantly more patients survive massive injuries than just 20 years ago, the focus in treatment must shift towards improving the quality of life for these individuals. Scarring, particularly HTS, is a common outcome following burn injury that significantly reduces quality of life. Many therapies have shown promise to treat and reduce HTS but none fully prevent or eradicate

20 the scarring. Therefore, investigation is needed. The development of a new animal model with high levels of homology to human skin and scar are required to represent an environment where the efficacy of current therapies can be rigorously tested and where new therapies can be developed.

Chapter 2: Novel Burn Device for Rapid, Reproducible Burn Wound Generation

2.1 Introduction

Severe scarring is estimated to affect as many as 70% of burn patients10 and results in both cosmetic and functional deformities that negatively impact on the quality of life for those affected67,77. As a result, significant research has been conducted to develop therapies to prevent and treat scarring3,9,11,46,67-68,78-79. Despite the volume of scar research, a highly effective treatment has not yet been developed. A complication to scar research is the availability of an in vivo testing environment. Longitudinal studies in the human patient population lack controlled burn depth, size and location along with proper negative controls67-68. Rodent models do not possess the complex mechanical and biological environment observed in human skin and do not naturally form hypertrophic scars69-70,80-82. More recently, female, red Duroc pigs (FRDPs) have been proposed as a model for scarring5,67-68,72. Studies have confirmed that FRDPs form robust scars following deep cutaneous wounds and that these scars are similar in appearance to human hypertrophic scars. FRDP scars are thick, raised above surrounding skin, lack hair and contain elevated populations of myofibroblasts and mast cell5,67,83. Additionally, the pattern of inflammatory cytokine expression is similar between humans and female red

Duroc pigs5,67,73,83. These structural and biological similarities between human hypertrophic and the thick scars on FRDPs provide a new platform for the study of scar development and anti-scar therapies.

One of the main obstacles in studying burn wounds in animal models is the difficulty in producing burns with uniform depth84. For consistent wound generation, it has been proposed that the initial temperature of burner, contact time, thermal capacity of burner material, and the amount of pressure applied to the skin all need to be tightly controlled84.

A number of different burning devices have been created to satisfy this set of needs.

Heating a metal or glass stylus in a water bath has been used in prior studies to generate cutaneous burn wounds84-86. One study created contact burns placing water heated to

92°C in a plastic bottle on the skin surface for 15s87. Although the operator and procedure were kept consistent, variability in injury depth and subsequently differences in healing time of local areas within each wound were reported87. To improve reproducibility, the addition of a pressure monitor to the standard burn device has been employed84 resulting in more uniform wounds.

The goal of this study was to develop a new burn device that could provide real time pressure monitoring, in addition to real time control and monitoring of device temperature, for rapid generation of reproducible burn wounds. This device was then compared to a standard burn stylus comprised of a heated block. Time required to bring the device to temperature, ability to raise and maintain skin surface temperature

23 throughout the burn experiment, and elapsed time between generation of consecutive burn wounds was examined. In addition, burn depth was assessed as a function of burn device, time, pressure and user.

2.2 Materials and Methods

2.2.1 Burn Devices

To create full-thickness wounds, 1 x 1 in stainless steel blocks were placed onto pig skin to create a thermal injury. Two different burn devices were utilized. The first was a standard burn stylus consisting of a stainless steel block (1in.3) connected to a metal rod and handle (Figure 2.1A), which can be heated using a hot water bath or using a hot plate.

A hot plate (Corning PC-420D, Tewksbury, MA) was used in this study to heat the block to 200°C (Figure 2.1A). The second device was a custom burner fabricated in house that consisted of a stainless steel block (2.5 cm3) connected to a metal stylus, an electronic microstat (JUMO GmbH and Co. KG, Fulda, Germany) and electronic scale (Guangzhou

WeiHeng Electronics Co. Ltd., Guangdong, China) (Figure 2.1B). The custom burn device was electronically heated internally, and the electronic scale allowed for precise measurements of pressure applied to the skin surface.

2.2.2 Temperature Logging System

Internal, mid-block temperature of the standard burner was logged in real time during initial heating, burning, and post-burn using a thermocouple (J-type, OMEGA, Stamford,

CT) inserted in a centrally-located hole in the stainless steel block (Figure 2.1A). The 25 thermocouple was connected to a Hi-Speed USB Carrier (National Instruments, Austin,

TX) and a computer for data acquisition. LabVIEW™ SignalExpress 2010 (National

Instruments, Austin, TX) recorded temperature of the thermocouple at 0.5 second intervals. To obtain mid-block temperature of the custom burner during heating, burning, and post-burn, a thermocouple connected to an electronic microstat controller was inserted in the stainless steel block (Figure 2.1B). Video footage was taken of the microstat controller during the burning procedure and was used to log temperature changes every 1 second. To quantify the surface temperature of the pig skin throughout the entire burn procedure, an additional thermocouple, connected to the same USB carrier, was placed directly in between the skin surface and burner and temperature recorded each second. Time-temperature plots were utilized to quantify time to reach initial temperature, maximum skin surface temperature and time required to bring the burn device back to temperature after each burn. Average elapsed time data was plotted + standard error and statistical differences assessed via student’s t-test.

2.2.3 Burn Wound Creation, Anesthesia, Animal Care

All experiments and data collection were performed following The Ohio State University

Institutional Laboratory Animal Care and Use Committee (ILACUC) approved protocols.

Four female, Red Duroc pigs were utilized for the study: two pigs for analysis of heating time and temperature using a fixed pressure and contact time, and two pigs for analysis of time, pressure, and intra-user variability on burn depth. Pigs were anesthetized with

Telazol followed by isoflurane and the dorsal trunk shaved and surgically prepared with alternating chlorohexidine2% and alcohol 70% scrubs (Butler Schein, Columbus, OH).

On the first two pigs, full-thickness wounds were generated using the standard burner heated via hot plate and the custom designed burner (n = 8 per group). Thermal injury was induced by heating each stylus to 200 + 5°C and applying the stylus to the surface of the skin for a contact time of 40 seconds. During the burn wound generation, time– temperature data was logged within the block and on the pig skin surface continuously.

Three pounds of pressure was applied with the custom-made burner. To deliver a roughly equivalent of pressure with the standard burner, users practiced by pressing the commercial device onto a scale to achieve the desired pressure, prior to wounding the pigs. Sixteen total wounds were generated on each pig, eight with the standard burner and eight with the custom-made burner. Photographs of the wounds were taken immediately after injury and day 0 biospies taken from half of the wounds. Wounds were then covered with non-stick gauze pads (Curad) and Elastikon™ (3M). Taffeta dressings were placed over wounds and secured with Elastikon. A fentanyl patch (NOVAPLUS patch, Watson Laboratories Inc, 100 mcg) was placed in the pig ear pinna and removed three days post wounding. The pigs were maintained on standard chow ad libitum, fasted overnight before the procedures, housed individually and euthanized 7 days post-burn.

The second two pigs were used to assess the ability of the devices to control burn depth via contact time or applied pressure, and to assess intra-user variability. In the first experiment, contact time was examined (5, 10, 20 and 40 sec) while holding contact

27 pressure at 3 lbs. Three independent users created three wounds per contact time. In the second experiment, pressure was examined (1 or 3 lbs) using a constant contact time of

40 seconds (n = 3 per user and a total of 9 per group). Following wound generation, pigs were euthanized and biopsies collected.

2.2.4 Histology

Pig skin was excised from each wound in 1cm x 2cm strips to collect the wound margin and the midpoint of each wound. These biopsies were fixed in formalin for 6 hours prior to processing and paraffin embedding. Sections were stained with hematoxylin and eosin

(H&E) and imaged with light microscopy (Nikon Eclipse 90i with NIS-Elements AR3.1 software) at 4X. Individual images were digitally stitched together (Adobe Photoshop

Elements 6.0) and representative composite images shown for each group. Wound depth

(at the wound center and at the margins) was quantified using ImageJ. Box-and-Whisker plots were generated from each burn depth data set using SigmaPlot software, with the upper boundary of the box representing the upper quartile of burn depths for each group, the lower boundary showing the lower quartile, the end of the upper whisker identifying the maximum value and the lower end of the whisker identifying the minimum value.

The solid line within the box represents the mean value of the data. Statistical analyses were performed using SigmaPlot software (One-way ANOVA with post-hoc test of

Tukey).

2.3 Results

2.3.1 Initial Heating Time

Internal burn stylus temperature as a function of time post heating was collected to compare the time required to bring each burner device from room temperature (20+5°C) to the desired stylus temperature for thermal injury (200 + 5°C). The standard burn stylus heated on standard hotplate (Figure 2.1A) took an average of 13.4 + 2.1 minutes to reach target temperature (Figure 2.2). The custom designed burn device with an electrically controlled internal heater and pressure gauge (Figure 2.1B) took an average of 16.9+ 2.9 minutes to reach desired temperature (p > 0.05, Figure 2.2).

Figure 2.2. Time required to bring each burn stylus to temperature (200ºC).

2.3.2 Stylus and Skin Surface Temperature during Use

In order to examine the ability of each burner device to raise and maintain skin surface temperature throughout the entire burn procedure, mid-block burn stylus temperature and skin surface temperature measurements were simultaneously collected on each burn.

Temperature data acquisition was repeated for a total of 16 burns per pig (4 sequential burns per device in each of the two pigs, before switching to other device). The internal temperature of the standard device decreased linearly at an average rate of 0.4°C/sec while in contact with the porcine skin (Figure 2.3A). In contrast, the custom device maintained the internal burn stylus temperature within 7°C of target temperature throughout contact with the skin surface. It took the standard burner between 7 and 20 seconds to raise the external skin temperature to 97°C (Figure 2.3B). In addition, there was significantly greater variability in the skin surface temperature during device contact when using the standard device versus the custom device (Figure 2.3B). The custom burn device was able to quickly raise the skin surface temperature to 97°C (Figure 2.3B), and consistent skin surface temperature was maintained during the full contact time of 40 seconds (Figure 2.3B).

Figure 2.3. Time-temperature traces for both the inside of the burn stylus (A) and the skin surface (B). Note that the internal temperature of the burner decreases linearly at a high rate in the standard device whereas the custom device held temperature to within 7 degrees throughout contact with the skin surface. The external skin temperature was rapidly raised to 97ºC using the custom burn device and held constant at this temperature until the device was removed. In contrast, the standard device required between 7 and 20 seconds to reach peak temperature and had only a short period of steady surface temperature. 31

2.3.3 Elapsed Time

The elapsed time between each of the four consecutive burns showed how quickly each burner was able to return internal stylus temperature to 200 + 5ºC after each burn wound.

The average elapsed time between burns for the custom design device was 2.8 minutes

(Figure 2.4). The standard device took approximately 4.5 minutes longer (p < 0.005), an average of 7.3 minutes between each burn, than the custom design device to heat up to target temperature in between burns (Figure 2.4). This resulted in the standard burner taking more than 2.5X as long as the custom design burner to generate 4 subsequent burn wounds.

Figure 2.4. Elapsed time between subsequent burns. 32

2.3.4 Burn Depth as a Function of Time, Pressure and Device

Histological analysis of wounds at day 0 showed necrotic tissue at increasing depths as a function of time in both groups (Figure 2.5). At 5 seconds, the standard device damaged only the upper epidermis whereas the custom device destroyed the full epidermis. At each contact time point, the thickness of tissue damage appeared to be greater in the custom device group compared to the standard group. In addition, wound depth appeared to be more sensitive to time in the custom group compared to the standard group where the difference in observed burn depth was less drastic between 5 and 20 seconds (Figure 2.5).

Figure 2.5. Representative H&E stained sections of burns generated using the standard and custom device heated to 200°C and applied to the skin at 3 lbs of pressure for 5, 10, 20 or 40 seconds. 33

A similar trend was observed with increasing pressure and constant contact time. Intact collagen fibrils and other indications of non-damaged tissue were observed higher in the dermis in the standard device group compared to the custom device (Figure 2.6).

Figure 2.6. Representative H&E stained sections of burns generated using the standard and custom device heated to 200°C and applied to the skin for 40 seconds at either 1 or 3 lbs of pressure.

To quantitatively assess burn depth, images were processed using ImageJ with burn depth calculated from the top of the tissue section to the point at which clear collagen fibrils

34 could be observed (Figure 2.7). No significant differences were observed among different users when either time or pressure was varied.

Therefore, data from all users within a group (time-pressure-device combination) was pooled and box-and-whisker plots of burn depth as a function of time or pressure and device were constructed. A trend of increasing mean burn depth with increasing time was observed in both the standard and custom device group (Figure 2.8A&B). With the standard device, a wider variance in intragroup burn depths was observed compared to the custom group and this intragroup variance increased with contact time (Figure 2.8A).

No significant difference in burn depth was found in the standard group between 5 and 20 seconds of contact time (p = 0.529). Burn depth at 40 seconds was significantly greater than 5, 10 or 20 seconds of contact time (p < 0.005). Using the custom burn device, intragroup variance was low at all contact time points (Figure 2.8B). Burn depth also increased significantly with contact time (p < 0.01). Similar trends were observed when device pressure was examined. With both devices, burn depth increased with increasing pressure. Mean burn depth was similar with both devices (standard = 140.4 + 82.9 µm, custom = 190.3 + 11.2 µm) at 1 lb of force. Intragroup variance was substantially greater with the standard device compared to the custom device (Figure 2.8C&D).

Figure 2.8. Box and whisker plot of burn injury depth using the standard (A&C) and custom (B&D) device. Burn depth increased as a function of contact time while pressure is held at 3lbs (A&B). Burn depth increased with increasing pressure during a 40 sec contact time (C&D).

2.4 Discussion

One of the most significant differences in the performance of the custom burn device compared to the standard device was its ability to generate reproducible, user- independent, controlled-thickness burn wounds at a rapid pace. The custom burner’s

37 ability to continuously be heated internally not only significantly shortened the overall amount of time required to create the wounds, but also made the custom burn device more convenient to use. Whereas the standard burn device needed to be transported to and from the hot plate to be heated between burns, a delay that can allow temperature to drop before contact with the pig’s skin, this was not required for the custom burner.

When studying burn wound healing, it is important to maintain consistency within the burns that are generated. This not only includes burn depth but also the initial time point of injury because the healing cascade begins instantaneously following injury88.

Therefore, it is ideal to use a burner that is capable of generating quick, highly reproducible burn wounds. Recently, this custom burner design was utilized successfully to create large (2” x 2”) full-thickness burn wounds in a Yorkshire pig model to facilitate the study of biofilm infected wounds89.

Initial temperature of burner, contact time, thermal capacity of burner material, and applied pressure all play a role in burn depth84. Burn depth scaled positively with both contact time and contact pressure using either unit; however the custom burn device generated burns with much greater reproducibility. Both the standard and custom burners were heated to the same initial temperatures, and both could produce tightly controlled burn depths by altering contact times. Thus, it is most likely the ability to control applied pressure and maintain burner temperature while in contact with the skin that improves the precision of the custom burn device. Both burners utilize a stainless steel stylus, therefore are presented with the same low thermal conductivity of 24 watts per meter

Kelvin90 and heat transfer capabilities. The difference lies within the custom burner’s ability to be heated continuously and internally during burning. This constant heat source provided continuous heat transfer from the stylus to the skin, resulting in the maintenance of a more uniform skin surface temperature throughout burning. This contributed to greater damage and consistency in wound depth, both between and within wounds, compared to the wounds created by the standard burner. The standard burner immediately started losing heat once detached from the hot plate and continued to cool down at a rapid rate while in contact with the skin. The amount of pressure applied to the skin during thermal injury has been suggested to be a key indicator of extent of damage with greater tissue damage resulting from increased pressure. Compression of the underlying tissue is believed to decrease the rate of heat dissipation84. Though each of the three independent users practiced controlling applied pressure with the standard device by pressing the device into a scale, it is highly likely that the magnitude of applied pressure varied from burn to burn. The ability to see the digital pressure gauge in the custom device allowed the user to monitor the applied pressure initially and maintain this pressure throughout the duration of contact. This level of control is not achievable with the standard device and provides a distinct advantage in generating reproducible burn injuries of tailored depths.

2.5 Conclusion

When creating burn wounds to study wound healing and scarring in an animal model, it is ideal to have a burn device that can quickly create consistent, consecutive burn wounds of uniform depth and damage. In this study, a new custom burn device was developed that allowed control and monitoring of real time pressure, internal burner temperature, and elapsed time throughout burn wound generation. In comparison to the standard burner, the custom burner not only was quicker and more convenient to use, but also created burn wounds of tailorable depth and high reproducibility. The burner design used for these studies can be further customized to produce burns of different sizes or shapes for studies in other animal models.

2.6 Acknowledgements

This project was supported by the Shriners Hospital Research Foundations Grant #85100 to HMP. Supported in part by NIH grant GM077185, GM069589, NR013898 and DOD

W81XWH-11-2-0142 to CKS. The authors would like to thank the Physics Machine

Shop at The Ohio State University for their assistance with machining and assembly of parts for the custom burner.

Chapter 3: Burn Scar Biomechanics Following Pressure Garment Therapy

3.1 Introduction

Burns account for roughly 1.25 million injuries in the United States annually91,92.

Scarring is the most common form of morbidity for burn survivors. This exaggerated proliferative response to wound healing results in rapid growth of connective tissue and excessive contraction11-12 greatly reducing skin pliability and quality of life for patients.

The current standard of care for the prevention and treatment of scarring following a burn injury is the use of pressure garments. These garments exert static compression on skin55-

57,93-96 and this pressure is hypothesized to limit blood flow, nutrient and oxygen supply to the scar tissue, reducing collagen synthesis50,56,97-100.

With little modification from the original, custom-fitted Jobst garments, pressure garment therapy (PGT) remains the preferred treatment modality for the prevention and treatment of burn scars47-48,101-102. Although this therapy has been used clinically for over 40 years, controversy remains regarding its efficacy53. Prior clinical studies reported outcomes with pressure garment therapy ranging from no evidenced-based benefit54 to significant increases in scar pliability55. A systematic review of four randomized controlled trials

41 showed a trend towards decreased scar height in PGT-treated scars53. Significant reductions in scar redness and thickness were also observed in scars receiving PGT103.

Because pressure garments are often used in conjunction with other forms of therapy, and due to the high incidence of patient non-compliance, the efficacy of PGT has been difficult to evaluate and neither the efficacy nor the optimal protocol for delivery have been scientifically established,57-58,104.

Complicating the study of pressure garment therapy is the inherent variation of burn depth and location found in human clinical trials and the variation in actual levels of pressure delivered by garments based on manufacturer and anatomical site. One clinical study used a low flow transducer to directly measure the cutaneous pressures generated by a pressure garment on various body parts58. The results showed an increase in subdermal pressures in the range of 9-90 mm Hg depending on the anatomical site58.

Garments over soft sites, such as the medial and posterior mid-calf, exerted pressures ranging from 9 to 33 mm Hg with a mean of 21 mm Hg. Subdermal readings taken over bony prominences showed an increase, ranging from 47 to 90 mm Hg58. In one study, the clinical effectiveness of compressive bandages was observed on 210 separate anatomic burn sites, which included sites in the head/neck, trunk, upper extremities, hands, lower extremities, and feet105. The study found that a critical factor in determining the effectiveness of pressure therapy was the anatomic area of the scar. Scars present on flat areas, such as the foot, exhibited the best improvement in contracture while the trunk region showed intermediate results and treatment of the hand and neck were

42 unsuccessful105. As outcomes are so closely linked with anatomical location and initial degree of injury, an animal model is needed where burn site and depth can be controlled.

More recently, female, red Duroc pigs (FRDPs) have been proposed as a model for studying excessive scarring5,67,72. Studies have confirmed that scars on FRDPs from deep wounds are red, thick, firm, and lack hair,5,67. In addition, the timescale for wound healing and scar development in FRDP correlates with that of deep dermal wounds in humans5,67 and the expression of transforming growth factor β1 (TGF-β1), insulin-like growth factor 1 (IGF-1), decorin, and versican in FRDP displayed similar changes following deep wounds in humans5,67,73,106-108. The similarities in skin structure, wound healing and scar development between female Red Duroc pigs and humans provide an ideal model for the study of PGT efficacy.

The current study examines the efficacy of pressure garment therapy on scar prevention in a Red Duroc pig full-thickness burn wound model. Burn wounds were created on the dorsum of each pig and treated with PGT or treated with garments without compression.

Scar contraction, biomechanics, blood flow, vascularization and extracellular matrix production and structure were evaluated at multiple time points over the 78 day study to investigate PGT’s ability to maintain skin pliability and to identify possible mechanisms of action.

3.2 Materials and Methods

3.2.1 Burn Wound Generation

All experiments and data collection were performed following The Ohio State University

Institutional Laboratory Animal Care and Use Committee (ILACUC) approved protocols.

Female Red Duroc pigs (n=8, 60 lb; Isler Genetics Inc., Prospect, OH) were anesthetized with Telazol followed by isoflurane and the dorsal trunk shaved and surgically prepared with alternating chlorhexidine 2% and alcohol 70% scrubs (Butler Schein, Columbus,

OH). Full-thickness wounds were induced by heating a 1 x 1 inch stainless steel stylus to

200 + 6 °C and applying to the skin for 30 seconds. Eight total wounds were generated per pig, 4 control and 4 receiving PGT (Supplementary Figure 1A-B, total of 32 PGT treated burns and 32 control burns). Wounds were covered with non-stick gauze pads

TM (Curad) and Elastikon (3M). A fentanyl patch (NOVAPLUS patch, Watson

Laboratories Inc, 100 mcg) was placed in the pig ear pinna and removed three days post wounding. Animals received a single intramuscular injection of Buprenorphine

(Buprenex, Reckitt Benckiser Healthcare, 0.3 mg/ml) during recovery from anesthesia.

Animals were maintained on standard chow ad libitum, fasted overnight before the procedures, and were housed individually. Animals were euthanized following the completion of experiments.

3.2.2 Pressure Garment Therapy and Pressure Quantification

Compression vests (The Marena Group Inc., Lawrenceville, GA) were modified for pig use. Two sets of adjustable, wrap-style garments with Velcro closures were fabricated for each pig to accommodate body circumferences ranging from 20-35 inches (Figure 3.1B).

At 28 days post-burn, pig circumference was measured and compression garments applied at a 10% reduction in circumference (Figure 3.1B). Vests with no reduction in circumference were placed over control burns (Figure 3.1B). To evaluate the magnitude of pressure exerted by the compression garments, compression garment fabric was wrapped around a load cell of a mechanical tester (TestResources; Shakopee, MN) at

10% reduction in circumference and pressure was tracked continuously for 12 hours. A representative plot of pressure versus time was reported (Figure 3.1C).

3.2.3 Scar Contraction

Photographs of the scars were taken at days 0, 7, 28, 42, 56 and 78. Each photograph of the scars (n=16 per group days 0-56, n = 14 per group at day 78) was taken with a scale in the field of view for scar area quantification. Scar contraction was quantified using computer planimetry (Image J89,109-111) and defined as total scar area, including hyperpigmented border, at a specific time point divided by original scar area (at day 28, time of therapy application) x 100. Data are presented as percent original area (mean + standard error of the mean (SEM)).

3.2.4 Laser Doppler

The MoorLDI-Mark 2 laser Doppler blood perfusion imager (Moor Instruments Ltd.,

UK) used a visible red laser beam (633 nm) to map tissue blood flow using in control and

PGT scars immediately after garment removal at days 28, 56 and 78 (n=16, day 78 n=14). 46

3.2.5 Scar Biomechanics

Hardness and elasticity of the upper regions of the skin (to a depth of approximately 0.75 mm) was measured using torsional ballistometry (Torsional Ballisometer, Dia-Stron

Limited, Broomall, PA) at post-burn days 28, 42, 56 and 78 (n=16 per group, n =14 at day 78). Hardness was reported as average indentation (mm) + SEM and elasticity reported as elasticity coefficient + SEM. Elasticity coefficient, α, is inversely proportional to the elasticity of the tissue. Additionally, failure biomechanics were assessed at day 78. Strips of tissue were removed from the pig parallel to the circumference of the pig. Dogbone shaped samples were cut from the tissue with the scar centered within the length of the sample112. Skin thickness was measured and strained at

2 mm/sec until failure (TestResources, Shakopee, MN). Ultimate tensile strength and linear stiffness were reported as mean + SEM (n=12 per group).

3.2.6 Immunohistochemistry

Biopsies were taken from control and PGT scars at post-burn days 7, 28, 42, 56 and 78 for histology (n=4 per group for days 7, 28, 42 and 56, n = 6 for day 78, no scars were biopsied more than once). The biopsies were embedded in OCT resin, frozen, and stored at -80oC until sectioning. Sections were stained with hematoxylin and eosin or immunostained to visualize general anatomy. To visualize blood vessels, sections were immunostained with von Willebrand factor protein (VWF, Santa Cruz Biotechnology)

47 and DAPI. The stained sections were imaged using an Olympus FV1000 Multi-Photon confocal microscope. Quantitative analysis for VWF positive cells in the dermis was performed by calculating the percent area of the total field of view that was positive for

VWF (n = 6 per group).

3.2.7 Transmission Electron Microscopy

On day 78, biopsies (n = 6 per group) were fixed in 2.5% glutaraldehyde in 0.1M phosphate buffer (pH=7.4) overnight at 4°C. Skin was post-fixed with a 1% osmium tetroxide. En block staining was performed using 2% uranyl acetate in 10% ethanol followed by dehydration in a graded ethanol series and embedding in Eponate 12 epoxy resin (Ted Pella, Redding, CA). Ultrathin sections were cut (Leica Microsystems), collected on copper grids and stained with lead citrate and uranyl acetate. Images were acquired with an FEI Tecnai G2 Spirit (FEI; Hillsboro, OR) transmission electron microscope. A minimum of 50 collagen fiber diameters were measured (Image J) from each sample (n = 6 per group) and plotted as a histogram for each sample type.

3.2.8 Statistics

Statistical analyses were performed using SigmaStat (Systat Software, Inc., San Jose,

CA). Statistically significant differences were detected using either student’s t-test or a

One Way ANOVA with a posthoc test of Tukey. Statistical significance was considered at p<0.05.

3.3 Results

3.3.1 Full-thickness Wound Generation and Pressure Garment Fabrication

Application of a 200 ºC burn stylus to the skin for 30 seconds resulted in a full-thickness burn injury as evidenced by the complete damage of the epidermis and dermis of the skin

(Figure 3.1D). Wounds were allowed to heal naturally for 28 days until the majority of wounds had fully re-epithelialized. The PGT group’s garments were manufactured to be

10% reduction in circumference (Figure 3.1B) resulting in an initial pressure of 10.1 mm

Hg (Figure 3.1C) which slowly decreased to 9.5 mm Hg after 12 hours. Garments were repositioned on the pigs daily to maintain an average pressure of 10 mm Hg.

Figure 3.2. Representative photographs of the scars 28-78 days post burn. Compression garments were applied at day 28.

3.3.2 Scar Morphology and Contraction

Immediately after thermal injury, the affected area appeared dry with linear wound margins (Figure 3.1A). Scars were hairless and hypopigmented in the center with a thin line of hyperpigmentation outlining the scar. Small areas of epidermal damage near the center of the scars were observed at day 28 in 6 of the 64 scars. This was not observed past day 28 (Figure 3.2). Scars contracted with time, becoming more star-shaped (Figure 50

1). Quantitative analysis of scar area showed a significant decrease in scar area from 42 days to 56 days post injury (Figure 3.3). At day 56, PGT scars were 88.1 + 3.5% original area as compared to the control, which were 73.8 + 4.8% original area (Figure 3.3). By day 78, scar contraction plateaued in the PGT group with no statistical difference between average scar area at days 56 and 78 (Figure 3.3). In contrast, control scars contracted an additional 8.5% from day 56 to 78 (Figure 3.3).

3.3.3 Scar Biomechanics

Non-destructive mechanical analysis of scars revealed that hardness of PGT treated skin was lower than that of the controls at day 78 with an average indentation (lower depth of indentation = higher hardness) of 0.27 + 0.08 mm in control scars and 0.42 + 0.06 mm in the PGT group. Additionally, elasticity coefficient, α, which is inversely related to elasticity, was significantly lower in the PGT group (α = 0.033 + 0.0071) than the controls (α = 0.052 + 0.009) and approached normal pig skin values (α = 0.026 + 0.0056) by day 78 (Figure 3.4C). Tensile testing of excised scar tissue showed that the ultimate tensile strength of the scars with PGT was significantly stronger than that of control scars

(Figure 3.4A). Additionally, PGT increased linear stiffness of scars compared to control scars (Figure 3.4B).

3.3.4 Scar Perfusion and Blood Vessel Density

Laser Doppler imaging of control and PGT treated scars showed no differences in perfusion at any time point (Figure 3.5A). Immunohistochemistry showed no observable change in blood vessel density between the control and PGT group at day 78. A quantitative analysis of blood vessel density within the dermis of PGT and control scars confirmed this observation (Figure 3.5B).

3.3.5 Scar Structure and Collagen Organization

At day 42, both control and PGT scars had completely epithelialized and a uniform epidermis was visible (Figure 3.6A&B). The junction between the epidermis and dermis in the control burns contained few rete ridges whereas a greater number of rete ridges were seen in the PGT group. Dense, cellular infiltration was apparent in both groups

(Figure 3.6A&B). By day 78, thick bands of collagen formed in the dermis of both the control and PGT groups with no gross difference in organization (Figure 3.6C&D).

Depth and number of rete ridges was greater in the PGT group compared to control.

Figure 3.6. H&E stained histological section of control (A&C) and pressure garment treated (PGT) burn scars (B&D) 42 (A&B) and 78 (C&D) days post injury. 55

Collagen organization was examined in greater detail using transmission electron microscopy. Collagen fibrils in control burns were larger and less tightly-packed than in

PGT treated scars, with approximately 27% less free space between fibers in the PGT group compared to controls. Collagen fibril diameter distribution ranged slightly larger in the control scars with an average of 17 nm greater fibril diameter than in PGT treated scars (Figure 3.7).

3.4 Discussion

In this study, the female red Duroc pig model was used to study the efficacy of PGT following burn injury. Advantages of this model over human clinical trials, such as uniformity of burn depth and body site, and ability to collect tissue biopsies for analysis, enabled a detailed evaluation of the effects of PGT on scar formation. Though control scars were thick and raised at day 78, their thickness and excess erythema was not believed to be significant enough to categorized the scars as hypertrophic in the current injury model. Significant differences in scar contraction were observed between scars receiving PGT and control burns that received no pressure. Pressure garments exert compressive forces normal to the scar and also parallel to the surface of scar. These forces act to oppose the direction of contracture113. It has been recently proposed that wound tension acts upon integrins by stretching them, which leads to increased phosphorylation of focal adhesion kinase (FAK) and downstream upregulation of smooth muscle actin (SMA) and collagen production114. When a compressive force was applied to incisional wounds in an opposite direction to the wound tension, it was shown that scars did not form115. These data suggest that the mechanical forces applied to the scar can assist in reducing differentiation of fibroblasts to myofibroblasts, ultimately decreasing scar contraction and collagen deposition. It is likely that the reduced scar contraction observed in the current study was, in part, a result of reducing the strain state within the scar, which subsequently abates myofibroblast differentiation and excessive collagen deposition.

Scar strength was improved with PGT compared to controls with a 34% increase in ultimate tensile strength. In addition to improvements in strength, PGT altered collagen deposition in the dermis with PGT scars comprised of smaller, more densely packed collagen fibers. Correlations between compression, tissue mechanics and collagen structure have previously been reported. Human scars treated with pressure via elastic bandages resulted in thinner reticular collagen fibers that resemble that of normal skin44.

A direct relationship between collagen fiber diameter and tensile strength was previously observed, with small diameter collagen fibers found in low strength wounds in the proliferative phase of healing, and large diameter fibers comprising high strength wounds in the remodeling phase of healing116. In the current study, control collagen fiber diameter was ~1.2-fold greater than collagen in PGT treated scars, thus it is unlikely that collagen fiber diameter was a dominant factor in controlling tissue mechanics. The difference in collagen fiber density was more dramatic. The decrease in interfiber free space likely inhibited fiber-fiber motion during deformation and led to increases in tissue strength and stiffness.

It is widely believed that pressure exerted by compression garments limits blood, nutrient and oxygen supply to the scar tissue limiting collagen synthesis50,56,93,97,99. In the current study, no differences in scar perfusion or blood vessel density were observed between the

PGT group and the controls (Figure 3.5). A possible reason for equivalent levels of scar perfusion and blood vessel density may have been the magnitude of pressure applied.

Garments manufactured to a 10% reduction in circumference resulted in approximately

10 mm Hg on the scar. This level of pressure is considered to be in the low range, and has been shown previously to result in increased redness and vascularity in scar tissue compared to high pressure (20-25 mm Hg)103. Prior clinical studies have also shown no difference in scar vascularity between pressure garment treatment groups and controls53.

Limitations to the current study include maintenance of pressure magnitude and duration of garment wear. Though garments were repositioned daily, the pigs were able to shift the position of the garments, relieving some pressure. Shifting of garment position by the pig was observed zero to two times per week, effectively reducing the total duration of compression. The maintenance of pressure magnitude for 23 hours of daily wear, which is currently the standard of care for burn patients, is challenging and a problem for all garment materials. Measurement of pressure exerted by the garments over a 12 hour period showed that pressure was reduced from 10 to 9.45 mm Hg in this time frame, and likely reduced further in the following 11 hours. As a result, the total effect of pressure garment therapy may have been moderately suppressed by challenges with garment position and pressure maintenance.

3.5 Conclusions

The Red Duroc burn scar model provides the ability to probe the efficacy of a variety of treatments with internal controls and greater ability to control patient compliance.

Pressure garment therapy at 10 mm Hg was found to be effective at reducing scar contraction. Modest improvements to scar biomechanics and structure also resulted from

PGT use. While the current study indicated the efficacy of pressure garments, improvements to the therapy to provide greater benefit to skin biomechanics are needed.

Chapter 4: Structural, Chemical and Mechanical Properties of Pressure Garments as a Function of Use and Laundering

4.1 Introduction

Each year over 490,000 burn injuries require medical treatment in the United States22.

Restoration of skin function to pre-injury status for the burn victim is difficult due to high incidence of excessive scarring following thermal injury18,117. These burn scars are commonly characterized by pruritus, pain, active contracture, and their erythematic appearance3,10-11,117. As a result, these scars result in functional and cosmetic deformities that can dramatically diminish a patient’s quality of life3,18,77,117.

Pressure garment therapy (PGT) is the primary management option for the prevention and treatment of burn scars46-47,57,117. In PGT, patients wear custom made compression garments that apply static pressure to the scar. Some studies suggest that lower pressures

(15-24 mm Hg) can be effective46,53,56, while a large number propose that applied pressure must exceed capillary pressure (~25 mm Hg)4,56. To deliver these levels of pressure to the scars, two main methods of garment fabrication are used: the reduction factor method and the Laplace’s law method. The Reduction Factor method involves reducing the patient’s circumferential measurements by a certain percentage, regardless 61 of the fabric properties57,80,117-118. After the patient is carefully measured, the measurements are decreased by a standard reduction factor of 10, 15, or 20%57,117-118.

Generally, patient’s measurements are reduced approximately 10% for the first set of pressure garments and 15% or 20% for all following garments47,57. In contrast, the

Laplace’s Law method takes into account the fabric’s tension profile when reducing measurements of the patient for construction117-118. Although this method provides more accurate delivery of the desired pressure magnitudes, no equations or design tools have been developed for ease of use117-118.

Compression garments are cut and constructed from elastic fabrics, often Powernet or knitted fabrics comprised of interwoven nylon and Spandex yarn80,117-119. Garments can either be constructed in house by occupational therapists or commercially by a number of companies47,57. In commercial construction, measurements are taken using the manufacturer’s measuring charts and tapes and sent to the factory where specialized computer aided design (CAD) systems are used in the pattern construction procedure47,57.

A common challenge in PGT is the loss of tension within the fabric and the inability of the garment to deliver the same amount of pressure with time and use. In only 1 month, garments lose approximately 50% of their compression56,117 due to daily wear and laundering10,53,56=57. Therefore, it is necessary to replace garments at least every 2-3 months in order to maintain sufficient pressure117. However, patients are also instructed to wear the garments until the scar is fully mature, which can take up to 2 years10,46. The

62 high cost of garments combined with the need to replace them multiple times throughout treatment place high demands on patients. Therefore, it is necessary to examine the mechanical, chemical and structural properties of these fabrics in order to construct the most effective garment for patients, both in terms of delivering consistent pressure levels and in maintaining good durability after repeated wear and laundering.

The current study examines two compression garment fabrics; Powernet and moleskin.

Change in applied pressure of custom-fitted compression vests made with these fabrics was measured on 5 key anatomical sites of a life-like dummy over a period of 23 hours.

Measurements were repeated after each garment was washed once and again after 5 washes. Load relaxation and energy dissipation was evaluated for each fabric at 3 different orientation angles, with no launder, after 1 wash, and after 5 washes. Surface topography was also examined to evaluate changes in fiber texture and structure as a result of applied tension and laundering.

4.2 Materials and Methods

4.2.1 Fabric Material

Moleskin fabric is comprised of 80% Nylon and 20% Spandex (Figure 4.1A) (Spandex

House Inc., New York, NY). Powernet fabric is composed of 90% Nylon and 10%

Spandex (Figure 4.1B) (Darlington Fabrics, Westerly, RI).

A B

Figure 4.1. Moleskin fabric (A) and Powernet fabric (B) with finished edge labeled

4.2.2 Laundering of Vest and Fabric

The vests and fabric were laundered in a household washing machine with regular detergent (Tide, Proctor and Gamble, Columbus, OH). No bleach or fabric softener was added to or present in the detergent. A normal launder cycle was used with warm water.

Fabric and vests were line dried to avoid heat degradation from the dryer.

4.2.3 Pressure Quantification of Custom-made Compression Vests

A child size lifelike dummy (Dummies Unlimited Inc., Pomona, CA) was used to observe changes in pressure measurements over time of 2 compression vests (Figure 4.2).

Dimensions of the dummy were measured using a tape measure, and 2 custom-fit compression vests (Shriners Hospitals for Children, Cincinnati, OH) were fabricated at a 64

10% reduction in dummy body size. The first compression vest was fabricated out of

Powernet fabric (Figure 4.2B), and the second compression vest was made out of moleskin fabric (Figure 4.2C). Pressure measurements were obtained at 5 key anatomical locations on the dummy: 1) upper left torso/chest, 2) lower right torso/abdomen, 3) lower right torso/lateral abdomen, 4) right shoulder, and 5) upper central back via Kikuhime pressure sensors (Figure 1A; Advancis Medical USA, Plainvew, NY). The Kikuhime pressure sensors consist of an air bladder that transmits surface pressure (mmHg) to a digital data reporting device. The air bladders were secured with double sided tape to the surface of each anatomical location before calibrating the reporting device to 0 mmHg.

The dummy was then dressed with a compression vest and each sensor was turned on to obtain initial pressure measurements. Hourly pressure measurements were obtained from each sensor for 23 consecutive hours. This procedure was repeated for both Powernet and moleskin vests. To observe the effects of machine laundering on the compression vests, the same pressure tracking procedure was repeated after laundering once and after

5 launder cycles for a total of 3 23-hour runs per vest. Pressure (mmHg) versus time

(hours) plots are reported for all 3 runs for each vest.

4.2.4 Mechanical Testing of Fabric Samples

4.2.4.1 Load Relaxation Testing

The effect of constant strain on the resultant load of Powernet and moleskin fabric over time was quantified with load relaxation testing. Fabric samples were cut 2.5 inches x

0.5 inches at 0°, 45° and 90° with respect to the finished edge of the fabric. Samples 66 were cut at 0º, 45º, and 90º from larger squares of fabric after laundering once and after 5 total launder cycles. Samples were mounted into the grips of a uniaxial tensile tester

(TestResources, Shakopee, MN, USA), strained to 10% at a rate of 2 mm/sec and held at maximum strain for 12 hours while load was continuously recorded by MTestWr Version

1.3.6 software (TestResources, Shakopee, MN). Load relaxation testing was performed on three different samples per fabric type and laundering condition. The raw force-time relaxation data were fit to

F(t) = At-n (1)

where A is the magnitude of force at t=0 and n is rate of relaxation. Average relaxation rate + standard deviation were reported. Average peak force during relaxation + standard deviation was also reported.

4.2.4.2 Cyclic Testing

To assess the fatigue properties of each fabric, cyclic testing was performed. Powernet and moleskin fabric that was not laundered, was laundered once, and was laundered a total of 5 times were cut into 2.5 inches x 0.5 inches samples at fabric orientations of 0º,

45º, and 90º with respect to the finished edge. Samples were mounted into the grips of a uniaxial tensile tester (TestResources, Shakopee, MN), where cyclically strained with amplitude of 10% at a frequency of 0.1 Hz for 3 hours (n = 3 per fabric/laundering

67 condition). Load (N) and position (mm) was continuously recorded by MTestWr Version

1.3.6 software (TestResources, Shakopee, MN). To quantify overall energy dissipation of each sample, the area underneath the resultant hysteresis curve was calculated using the trapezoidal numerical integration function in MATLAB® (The MathWorks, Inc.,

Natick, MA).

4.2.5 Scanning Electron Microscopy (SEM)

Scanning Electron Microscopy (Sirion; FEI, Hillsboro, OR) was used to characterize the effects of laundering on the structure of powernet and moleskin fabric. Samples (0.1 inches x 0.2 inches) of powernet and moleskin fabric that had been laundered 0, 1 and 5 times were attached to aluminum (Al) stubs (Ted Pella, Reading, CA) using adhesive carbon tape and sputter-coated with gold palladium. Samples were imaged in secondary electron mode at 5 kV accelerating voltage.

4.2.6 Attenuated Total Reflectance – Fourier Transform Infrared Spectroscopy (FTIR)

All samples were characterized with the attenuated total reflectance (ATR) mode of a

Thermo Nicolet Nexus670 FTIR spectrometer, using the Continuum microIR accessory and liquid nitrogen-cooled MCT detector (Thermo Scientific, Waltham, MA). A germanium crystal was placed in contact with each fabric sample and 25-40 scans were collected at a 8 cm-1 resolution. Peak analysis included determination of peak height for

Amide I (1650 cm-1), Amide II (1540 cm-1) and urea carbonyl (1735 cm-1) peaks without normalization or baseline.

4.3 Results

4.3.1 Pressure Quantification of Custom-made Compression Vests

Custom fit compression vests made of powernet and moleskin fabric were worn by a life like dummy to quantify the magnitude of pressure below each fabric at 5 key anatomical locations as a function of time and as a result of laundering (Figure 4.2). As anticipated, the applied pressure was strongly influenced by anatomic location. Below the fabric in the upper central back, which is slightly concave, the maximum applied pressure measured was 1 mm Hg with the majority of measurements at 0 mm Hg (Figure

4.3A&D). Maximum initial applied pressures observed with the powernet fabric were at positions 1, 2, and 3 or the chest, abdomen and lateral abdomen respectively (Figure

4.3A). The moleskin vest applied greatest pressure at area 3 (lateral abdomen) and area 1

(chest) (Figure 4.3D). Hourly pressure measurements over 23 consecutive hours showed a decrease in pressure over time at areas 1-4 for both compression vests. After one day, pressure applied by the powernet vest at areas 1-4 was reduced by 22-35% from pressure measurements at t = 0 hours (Figure 4.3A), while there was a 12-43% pressure reduction in the moleskin vest (Figure 4.3D).

Both vests were laundered once and pressure measurements repeated at the same anatomical areas on the dummy. Small reductions (average -2.5%) in applied pressure was observed with the powernet vest at areas 1, 3, 4 and 5 (Figure 4.3B). In contrast, applied pressure at site 3 was 1/3 of the unlaundered condition (Figure 4.3B). After laundering the moleskin vest once, applied pressure was reduced an average of 15% in areas 1, 2, 4, and 5 with an 86% reduction at area 3 (Figure 4.3D&E). As with the no launder condition, applied pressure decreased over time with each vest at all areas (Figure

4.3B&E). After 5 laundering cycles, the moleskin fabric displayed a substantial reduction in applied pressure with garments applying only one third of the original pressure (Figure 4.3F). In contrast, no additional change in applied pressure was measured at areas 3, 4, and 5 with the powernet vest after 5 laundering cycles compared to 1 laundering cycle (Figure 4.3B&C). With respect to applied pressure in the no laundering condition, powernet fabric decreased on average 27% for all sites after 5 laundering cycles (Figure 4.3A&C).

4.3.2 Mechanical Analysis of Fabric

Relaxation tests of powernet and moleskin fabric cut at orientations of 0°, 45°, and 90° showed that powernet fabric exhibited larger peak loads than moleskin, when strained to

10% for all launder conditions (Figure 4.4). The average peak loads for unlaundered powernet fabric were 0.63 N + 0.00 for 0° orientation, 0.39 N + 0.07 for 45°, and 0.61 N

+ 0.03 for 90°. After laundering once, the peak loads for powernet fabric remained higher than moleskin at 0.58 N + 0.25 for 0°, 0.65 N + 0.04 for 45°, and 0.75 N + 0.02 for 90° orientation. Average peak loads for powernet fabric laundered 5 times were 0.50

N + 0.03, 0.40 N + 0.10, 0.70 N + 0.01 for 0°, 45°, and 90° orientations, respectively.

The average peak loads were 0.28N + 0.07 for 0° orientation, 0.21 N + 0.02 for 45°, and

0.16 N + 0.02 for unlaundered moleskin. After 1 laundering cycle, average peak loads for moleskin were 0.17 N + 0.07 for 0˚ orientation, 0.22 N + 0.01 for 45˚ orientation, and

0.37 N + 0.02 for 90˚. Average peak load for moleskin after 5 washes was 0.18 N + 0.02,

0.30 N + 0.02, and 0.48 N + 0.02 for 0˚, 45˚, and 90˚ orientation, respectively.

For each laundering group and fabric orientation, moleskin fabric generated very small load at an initial strain of 10% (Figure 4.4D-F). This load relaxed rapidly, within 100 sec, and continued to relax slowly for the remainder of the experiment. In contrast, the

Powernet fabric generated on average 2-4X the initial load and relaxed at a rapid pace within the first 200 seconds (Figure 4.4A-C). Load continued to decrease over the course of the experiment at a slow rate. Laundering slightly reduced the average maximum load at 10% but did not significantly alter the steady state relaxation rate for either fabric.

After 5 cycles of laundering, the Powernet fabric generated more load than that of the moleskin fabric at all conditions.

A D

B E

A A

C F

A A

Figure 4.4. Powernet fabric generated on average 2 to 4 times the initial load and relaxed at a rapid pace (A-C), while moleskin fabric generated very small load at an initial strain of 10% (D-F).

4.3.3 Cyclic Testing

Cyclic testing with amplitude of 10% at a frequency of 0.1 Hz of unlaundered powernet samples of 0° orientation revealed a lower overall energy loss than unlaundered Powernet samples of 45° and 90° orientation (Figures 4.5). In this orientation, no change in energy dissipation with laundering was observed with the moleskin fabric and after 5 laundering cycles with the Powernet fabric (Figure 4.5A). When fabric was oriented at 45° and 90° to the finished edge, the moleskin fabric has a significant increase in energy dissipation or fatigue with each laundering cycle (Figure 4.5B&C). In contrast, no significant difference was observed as a function of laundering in the Powernet fabric at 45° and 90°

(Figure 4.5B&C).

4.3.4 Scanning Electron Microscopy (SEM)

SEM analysis of Powernet and moleskin fabric samples that have been unlaundered, laundered once, and laundered 5 times revealed no change in fiber structure or orientation as a result of laundering (Figure 4.6).

Powernet

Moleskin

Figure 4.6. Gold plated SEM images of Powernet and moleskin fabric at each of the three launder cycles. No fiber structure or orientation change was found between each launder cycle for both fabrics.

4.3.5 FTIR

FTIR analysis of moleskin fabric showed the presence of nylon with Amide I and Amide

II absorbance peaks at 1635 cm-1and 1543 cm-1, respectively, in samples that were

76 unlaundered, laundered 1X and laundered 5X (Figure 4.7A). FTIR analysis of Powernet fabric showed the presence of Spandex in addition to nylon with absorbance peaks at

1755 cm-1, 1635 cm-1 and 1535 cm-1, respectively, in all three laundering conditions

(Figure 4.7B). The absorbance of Amide I decreased with laundering (Figure 4.7A).

Unlaundered moleskin fabric showed an Amide I absorbance of 0.113, which decreased to 0.072 and 0.063 after one and five laundering cycles, respectively (Figure 4.7A). The absorbance of Amide II (1543 cm-1) also decreased, from 0.098 when unlaundered to

0.045 after five washes. In contrast, the absorbance of Amide I and II from Powernet fabric increased after one laundering cycle and decreased back to similar values as unlaundered samples after five washes. Powernet fabric displayed an Amide I absorbance of 0.077, which increased to 0.081 after one wash and decreased to 0.071 after five washes. Similarly, unlaundered Powernet displayed an Amide II absorbance of

0.097, which increased to 0.10 after one laundering cycle and decreased to 0.089 after being laundered five times. Powernet showed the presence of Spandex (1755 cm-1) in addition to nylon, which was not detected in the moleskin fabric samples. Unlaundered

Powernet displayed an absorbance of 0.022, which remained the same after one wash and decreased to 0.011 after five laundering cycles.

A B

A A

4.4 Discussion

As anticipated, magnitude of pressure delivered was a function of anatomical location. In areas where the body contour is concave low to no pressure was generated as a result of little to no contact between the garment and the tissue. Over more solid, bony protuberances, higher pressures resulted as seen in the shoulder and chest in this study.

Throughout a day’s use, the magnitude of pressure delivered decreased as the fabric material relaxed. This relaxation can occur at the macroscopic level with the distances between thread bundles within the fabric rearranging to accommodate the strain and at the microscopic level where polymer chains within the thread fibers realigning to reduce 78 the force on each chain. We did not observe any large scale alteration between macroscopic fiber spacing and orientations within the fabric during the course of 23 hours thus believing this relaxation is occurring at the molecular level within each fiber.

Magnitude of initial applied force and fatigue was dependent on fabric type and cycles of laundering. However, these observations were most evident in the moleskin fabric. This fabric applied the least amount of pressure to the pediatric dummy and within the tensile testing apparatus. In all mechanical tests, the moleskin fabric was the most susceptible to fatigue and to deterioration in properties following laundering. This may be due to how the two materials are blended in each fabric. Though the moleskin contained 20%

Spandex, 10% more than the Powernet fabric, no Spandex peaks were observed in ATR-

FTIR analysis. This suggests that the two materials (nylon and Spandex) may be in a layered composite architecture rather than a true blend. This may lead to layers within the fiber that exhibit less elasticity and resistance to fatigue. As the nylon is the majority component its properties may mask the fatigue properties of the Spandex.

Additionally a strong dependence between properties and fabric orientation was observed in each fabric. As both fabrics are woven, anisotropy in properties was expected; however it was not anticipated that there would be a combinatorial effect of orientation and laundering. While the Powernet did not display significant changes in fatigue with laundering, the moleskin fabric, on average, fatigued more after laundering and this increase in fatigue was most apparent when the fabric was oriented 90° to the finished

79 edge. As the moleskin fabric has a very strong weave orientation with braids of fabric oriented parallel to the finished edge, microscopic pilling of the fabric could have occurred during laundering leading to greater fatigue.

4.5 Conclusion

Fabrics, Powernet and moleskin, commonly used for pressure garment fabrication were assessed for their ability to apply load at a specific reduction factor, their fatigue properties, they structure and chemistry and how these properties change with care/laundering. These data suggest that to provide garments with the least amount of fatigue both during use and as a result of laundering, Powernet fabrics should be used in the construction of pressure garments with the leading edge of the fabric oriented parallel to the Langer’s lines within the patient’s skin.

Chapter 5: Female Red Duroc Pig as an Animal Model for Hypertrophic Scarring

5.1 Introduction

Hypertrophic scars (HTS) are a common form of morbidity resulting from burn injury and surgical procedures, with incidence rates varying from 30% to 91% following burns and 40 to 94% following surgery117. Hypertrophic scars develop as a result of a heavily amplified proliferative response during wound healing10-12. In most cases, patients suffering from HTS report a severe impairment of quality of life120. Throughout the entire scar maturation period of 2 years or more, the scar is actively contracting10.

Contracture leads to deformity, restriction of mobility and ultimately loss of function in the affected areas10,15,117. In addition, hypertrophic scars exhibit significant erythema, pruritus, pain, burning and stiffness10,117. A large number of therapies, including corticosteroid injections37,40,43-44, silicone gels37-41, pressure garment therapy10-11,46-49,51-

52,121 and laser therapy40,44,59,61-64,66, have been developed in attempts to mitigate the problem of HTS. To date, none of these therapies can completely prevent scarring and reported outcomes using these therapies vary greatly from highly effective38-39,45,53,55,61-

62,103 to no benefit and deleterious outcomes some cases37,39,42-43,54,64.

The vast number of treatment options is a reflection of the significant need to treat hypertrophic scarring and the difficulty in conducting tightly controlled scar studies using human clinical trials. As a result, developing a ubiquitously effective treatment option remains elusive40,44,117. With unclear efficacies of therapies and unknown mechanisms of action, hypertrophic scar therapy remains challenging and controversial. Therefore, further HTS research is critical to gain a better understanding of the governing mechanisms and to develop more efficient therapeutic strategies for the prevention and treatment of this devastating condition. Unfortunately, scar studies have a common challenge: the lack of an in vivo model for investigation. Longitudinal studies in the human patient population cannot tightly control for burn depth, size and location and, in most cases, cannot include proper negative controls67-68. Additionally, these abnormal scars are specific to humans68.

Many past studies have attempted to use rodents to create valid animal scar models.

However, rodent skin does not naturally form HTS and does not have the complex biological and mechanical environment that human skin possesses3,69. A rabbit ear model for hypertrophic scarring was reported after surgeons noticed elevated scar tissue developing after surgical wounding8,70. However, structural parameters observed in human hypertrophic scarring cannot be investigated in this model due to histological differences in rabbit ear skin including thickness and the healing process over a cartilaginous base69,122. Another rodent scar model was described in tight-skin mice

(TSM), a mutant mouse strain characterized by firmly attached skin and lack of skin folds

8,71. Although the scars developed in this model were characterized by histological similarities as human HTS, these similarities did not last for more than a couple months8,71.

More recently, a model for excessive scarring has been proposed in female red Duroc pigs (FRDPs)5,67,71,83. To date, the most validated model of HTS is reported by Zhu et al.

In their first study, tangential wounds (approximately 8 cm x 8 cm) were created on the dorsum of 6 female red Duroc pigs with a Padgett dermatome67. They reported that deeper wounds, classified as wounds created with a total dermatome setting (TDS) of

0.060 in and > 0.075 in, were not healed by week 3 post-wounding67. Scars caused by wounds with a TDS of 0.060 in had an average thickness of 6.9 mm, while deeper wounds with a TDS > 0.075 in had an average thickness of 9.3 mm67. When inspecting the clinical appearance of the wounds at 5 months, the group reported that deeper wounds demonstrated hair loss, contraction and hyperpigmentation67. Further similarities between HTS in red Duroc pigs and humans with comparable expression patterns of

Decorin, Versican, insulin-like growth factor (IGF-1) and transforming growth factor

(TGF-β1) were reported following injury and scar formation.

Although this dermatome model for hypertrophic scarring provides a platform for the study of scar development in FRDPs, some limitations remain. Histological observations of deep FRDP wounds showed disorganized collagen fibers formed into whorls and nodules5,67. However, HTS in humans have collagen bundles that are aligned in the same

83 plane as the epidermis123-124. In addition, the shape of FRDP scars differs from that of human HTS. Only 1-2 mm of FRDP scars extend above the surrounding normal skin125.

However, human hypertrophic scars are more raised, extending up to 1 cm above normal skin with abrupt edges67,125. Furthermore, the precise Padgett dermatome settings do not translate into accurate wound depths. Multiple tangential excisions must be performed to achieve deeper wounds, introducing additional error and variable depth67,69. Due to these limitations, it is necessary to investigate new models for HTS.

The goal of this study was to develop and investigate new models for hypertrophic scarring in FRDPs. A full-thickness burn model was generated and compared to the dermatome model recreated following the methods established by Zhu et al. Scar morphology and structure were assessed at multiple time points over the 150 day study.

Further, scar biomechanics and contraction were analyzed in each model.

5.2 Materials and Methods

5.2.1 Burn Injury and Animal Care

All experiments and data collection were performed following The Ohio State University

Institutional Laboratory Animal Care and Use Committee (ILACUC) approved protocols.

Four FRDPs were used for the study for the development and assessment of a dermatome model and burn model. Pigs were anesthetized with Telazol followed by isoflurane. The

84 dorsal trunk was shaved and sterilized with two alternating chlorohexidine 2% and 70% ethanol scrubs (Butler Schein, Columbus, OH). On each of four pigs, two burn wounds and two dermatome wounds were created on the dorsum (n = 8 per group). Thermal injury was induced by heating a stainless steel stylus (2 in x 2 in) to 200 + 10°C and pressing the stylus to the surface of the skin for a contact time of 40 seconds at a pressure of three pounds (Figure 5.1A). A Zimmer® Air Dermatome (Zimmer Inc., Warsaw, IN) with a 2 inch width plate was used to create tangential excisional wounds (Figure 5.1B and C). Two wounds (approximately 2 in x 2 in) were created on the back of each of four pigs with a TDS of > 0.075 or ~0.060 in. To ensure accuracy in TDS, the tissue that was tangentially excised with the dermatome was measured in between glass microscope slides using a digital dial caliper. The wounds were photographed immediately after wounding and then covered with non-adhesive gauze pads (Curad) and Tegaderm™ (3M,

St. Paul, MN). Vetrap™ (3M, St. Paul, MN) bandage tape was wrapped around dorsum and secured with Elastikon® (Johnson & Johnson). Bandages were removed 7 days post- wounding. A fentanyl patch (NOVAPLUS path, Watson Laboratories Inc., 100 mcg) was placed in the ear pinna of each pig and removed three days post wounding. The pigs were maintained on standard chow ad libitum, fasted overnight before the procedures, housed individually and euthanized 150 days post-wounding.

A B C

Figure 5.1. At day 0 post-injury: Burn wounds (A) and dermatome wounds with TDS ~ 0.060 in (B) and TDS > 0.075 in (C). Scale bar = 10mm

5.2.2 Wound Closure and Transepidermal Water Loss

The appearance of each wound was inspected at day 28 post injury to assess if the wounds were healed. Portions of the wound that were pink and tender or covered by a tight scab were considered unhealed. A DermaLab® Combo unit (Cortex Technology

ApS, Denmark) was used to assess epidermal barrier function by quantifying transepidermal water loss (TEWL). With the probe placed over the center of the wounds, relative humidity of the air directly above the wound is measured, compared to the relative humidity in the ambient conditions of the environment. Total water loss from the region of the skin is calculated as grams of water lost per square meter of skin per hour.

A large amount of transepidermal water loss is an indication of a non-healed or open wound.

5.2.3 Scar Contraction

Photographs of scars for the analysis of contraction were taken at 28, 90 and 150 days post wounding. Each scar photograph (n = 8 per group) was taken with a scale in the field view for scar area quantification. Scar contraction was quantified using computer planimetry (Image J121) and defined as total scar area a specific time point divided by the original scar area (at day 28) x 100. Data are presented as percent original area (mean + standard error of the mean (SEM)).

5.2.4 Scar Morphology and Structure

Six mm diameter punch biopsies were taken from each wound at days 10, 28, 90 and 150 days post-wounding. The biopsies were embedded in OCT resin, frozen and stored at -

80°C until sectioning. Sections were stained with hematoxylin and eosin to visualize general tissue anatomy.

5.2.5 Scar Thickness

At day 150 post-wounding, multiple 4 mm x 50 mm tissue biopsies were excised from each wound. Photographs with a scale in the field view were taken of each sample. To quantitatively assess scar thickness, images were processed using ImageJ with scar

87 thickness calculated from top of the epidermis to bottom of the scar tissue in the dermis and reported as average + standard error of the mean (SEM).

5.2.6 Scar Biomechanics

The strength of each scar was assessed at day 150 by analyzing failure biomechanics.

Strips of tissue (approximately 4 mm x 50 mm) were cut from excised biopsies from each wound parallel to the circumference of each pig. The wound sites were positioned centrally within the skin biopsies and mounted into the grips of a TestResources mechanical tester (TestResources, Shakopee, MN) and strained at 2 mm/sec until failure.

Ultimate tensile strength was reported as mean + SEM (n > 50 per group).

5.2.7 Statistics

Statistical analyses were performed using SigmaPlot (Systat Software, Inc., San Jose,

CA). Differences that were statistically different were detected using either a student’s t- test or a One Way ANOVA with a posthoc test of Tukey. Statistical significance was considered at p < 0.05.

5.3 Results

5.3.1 Rate of Wound Healing

At 4 weeks days post-wounding, 0 out of 8 burn wounds were healed. All wounds were red with portions covered in desiccated blood and wound exudate (Figure 5.5). Average

TEWL value for burn wounds was 114.5 + 8.1 g/m2/hr, which was significantly greater

(p < 0.02) than average TEWL for normal skin (7.0 + 0.9 g/m2/hr) (Figure 5.2). For dermatome wounds with a TDS > 0.075 in, 4 out of 4 wounds were partially healed. A small portion of the wound centers were covered in dark, tight scabbing; however, the edges appeared to have been re-epithelialized (Figure 5.5). Average TEWL for these wounds was 64.2 + 10.9 g/m2/hr, which was also significantly greater than normal skin (p

< 0.02) (Figure 5.2). For dermatome wounds with a TDS of ~0.060 in, 4 out of 4 wounds were visually healed at week 3 post-wounding. No scabbing or red tender tissue was present (Figure 5.5). Average TEWL for these wounds was 13.8 + 3.5 g/m2/hr, which was not significantly different from normal skin (Figure 5.2). Average TEWL values for burn scars were also significantly greater than dermatome scars created with a TDS >

0.075 in and a TDS of approximately 0.060 in (p < 0.020).

Figure 5.2. Average transepidermal water loss (TEWL, g/m2/hr) at day 28 post-wounding for normal skin, burn wounds, dermatome wounds with a TDS > 0.075 in and dermatome wounds with a TDS approximately 0.060 in. Average TEWL values of burn wounds were significantly greater than normal skin and both dermatome wounds.

5.3.2 Scar Thickness

The average scar thickness of the burn group (5.73 + 0.06 mm) was statistically greater (p

< 0.02) than scars resulting from dermatome wounds at a TDS of 0.075 in (4.23 + 0.07 mm) and approximately 0.060 in (3.52 + 0.05 mm) (Figure 5.3). The average thickness of normal pig skin was 3.47 + 0.07 mm (Figure 5.3). Scar thicknesses in the dermatome

TDS > 0.075 in and the burn wound group (Figure 5.3) were statistically greater than normal pig skin (p < 0.02).

5.3.3 Scar Contraction

Quantitative analysis of wound area showed contraction from 0 days to 28 days post injury for all groups (Figure 5.4). At day 28, burn wounds contracted to 59.2 + 1.8% original wound area, while dermatome wounds at a TDS > 0.075 in contracted to 59.3 +

5.7% original area and dermatome wounds at a TDS ~ 0.060 in were 66.2 + 6.5% original area (Figure 5.4). There was no significant difference in contraction between all wounds.

Burn scars and dermatome scars (TDS > 0.075 in) continued to contract after day 28, while contraction plateaued in dermatome scars from a TDS ~ 0.060 in (Figure 5.4). At day 150, burn scars were 39.0 + 3.3 % original area at day 28 as compared to dermatome scars at a TDS = 0.075 in and 0.060 in, which were 52.5 + 2.4 % and 65.2 + 0.8 %, respectively (Figure 5.4). Contraction of burn scars were significantly greater (p < 0.001) than scars resulting from dermatome wounding.

5.3.4 Scar Morphology and Structure

Burn scars at day 28 post-injury do not appear healed, with the presence of red moist tissue and dark tight scabbing (Figure 5.5). Dermatome wounds with a TDS > 0.075 in appeared to be healed everywhere except for a small portion at the centers of the wounds that were covered in a dark scab (Figure 5.5). Dermatome wounds with a TDS of

93 approximately 0.060 in appeared to be completely healed. No visible dark scabbing was observed (Figure 5.5). By day 90, burn scars demonstrated contraction and deformation in the initial square shape (Figure 5.4 and 5.5). Burn scars were hairless and hypopigmented in the center with a line hyperpigmentation outlining the scar (Figure

5.5). The central tissue of the burn scar was also raised above the surrounding normal skin (5.5). Dermatome wounds with a TDS of 0.060 in demonstrated minute contraction, re-growth of hair and hyperpigmentation at day 150 post-wounding (Figures 5.4 and 5.5).

There was no portion of the scar that was raised above the surrounding uninjured tissue

(Figure 5.5). At the same time point, dermatome wounds with a TDS > 0.075 in demonstrated slight contraction, hair loss and hyperpigmentation (Figures 5.4 and 5.5).

Portions of the scar tissue were slightly raised above the adjacent normal skin (Figure

5.5). At day 150, all scars were morphologically similar to the scars from day 90 (Figure

5.5).

Figure 5.5. Representative photographs of wounds and scars resulting from burn injury and dermatome wounding (TDS of ~ 0.06 in and > 0.075 in) 28, 90 and 150 days post wounding. Scale bar = 10mm

At day 90 post-injury, a uniform epidermis is visible on all scars resulting from burn and dermatome scars (Figure 5.6). The junction between the epidermis and dermis in burn scars and dermatome scars (TDS > 0.075 in) contained few to no rete ridges (Figure 5.6) whereas a greater number of rete ridges were visible in dermatome scars (TDS of approximately 0.060 in) (Figure 5.6). Rete ridges formed at the epidermis and dermis 95 boundary by day 150 for burn scars and dermatome scars from a TDS > 0.075 in (Figure

5.6). There was no gross difference in collagen organization between scars and between days 90 and 150, with thick bands of collagen formed in the dermis (Figure 5.6). In addition, dense cellular infiltration was visible in all scar tissue.

Figure 5.6. Representative H&E stained histological section of scars resulting from burn injury and dermatome wounding (TDS of ~ 0.06 in and > 0.075 in) 90 and 150 days post wounding.

5.3.5 Scar Biomechanics

Tensile testing of excised scar tissue at day 150 post-injury showed that the ultimate tensile strength of burn scars (0.61 + 0.13 MPa) and dermatome scars from a TDS >

0.075 in (0.96 + 0.34 MPa) were significantly weaker than normal skin (2.22 + 0.09

MPa) (p < 0.05) (Figure 5.7). Burn scars were also significantly weaker than dermatome scars from both total dermatome settings (Figure 5.7).

Figure 5.7. Ultimate tensile strength of burn scars were significantly lower than dermatome scars (TDS > 0.075 in) and dermatome scars (TDS > 0.075 in) at day 150 post-wounding.

5.4 Discussion

In this study, the female red Duroc pig was used to develop and investigate new models for hypertrophic scarring. A burn model was developed and compared to the dermatome model by Zhu et al by assessing re-epithelialization and barrier function at 4 weeks post- 97 wounding, scar thickness and clinical appearance at 5 months post-wounding, scar anatomy and collagen organization,. Biomechanics and scar contracture was also assessed from scar tissue of each model.

5.4.1 Wound Closure

It has been previously reported that in humans HTS has been reported to be more prevalent in wounds that require greater than 3 weeks to heal (Zhu 2003). Zhu et al. reported that deeper dermatome wounds, classified as wounds created with a TDS of

0.060 in and > 0.075 in, were not healed by week 3 post-wounding. However, in this current study, all dermatome wounds with a TDS of approximately 0.060 in (0.066,

0.055, 0.066 and 0.060 in) were completely healed and dermatome wounds with a TDS >

0.075 in (0.075, 0.075, 0.078 and 0.079 in) were largely healed visually at 4 weeks post- injury. Burn scars remained open using visual inspection and TEWL measurements. The increased time to heal in the burn wounds was expected as the initial injury is more significant, not only destroying the tissue immediately beneath the burn stylus but also triggering a cascade of events in the surrounding tissue that can lead to damage in the area surrounding the contact region.

5.4.2 Clinical and Histological Appearance

Human HTS is characterized by abrupt edges, hair loss, hypopigmentation and contraction3. Dermatome wounds with a TDS of ~0.060 in. were hyperpigmented, contained hair and had gradual increase in thickness from the scar edge towards the center. Additionally, these scars contracted very little over the 150 day experiment.

When dermatome wounds were generated at greater depths, scar contraction increased slightly and no hair was present within the scars. The scars formed using the burn model were hairless, contracted and hypopigmented. However, like the dermatome scars, none demonstrated abrupt edges; the boundaries between scarring and uninjured skin were more gradual. These findings are in contrast with prior studies reporting on the dermatome model which reported hairless, thick scars in both TDS = 0.06 in and TDS >

0.075 in groups. A possible mechanism for these observed differences is a difference in initial depth. TDS was validated in the current study using a digital caliper. It is possible that the depths of the wounds in the prior studies were larger than the dermatome setting.

When compared to the dermatome model, the burn model demonstrated complete hair loss in all of the wounds and hypopigmentation, macroscopic properties more analogous to what is observed clinically.

In addition, burn scars were also thicker than dermatome scars, and the average thicknesses of burn scars and dermatome scars with a TDS > 0.075 in were significantly greater than normal skin. This increased thickness is likely a result of the greater initial

99 tissue damage, longer period for wound closure and the resultant increase in the inflammatory and proliferative phases of wound healing.

Furthermore, there is a lack of elongated rete ridges seen in the H&E stained histological sections at day 90 of burn scars and dermatome scars from a TDS > 0.075 in, which coincides with the poor rete ridge regeneration and epidermis flattening that is seen in human HTS52,126.

5.4.3 Scar Biomechanics

At day 150 post-wounding, the ultimate tensile strength of scars from burn and dermatome injury were lower than normal skin, with a statistical difference between normal skin and burn scar (p < 0.001) and normal skin and dermatome scar created with a

TDS > 0.075 in (p < 0.005). The burn scar had the lowest ultimate tensile strength in comparison to dermatome scars of both total dermatome settings, suggesting a greater scar formation with thermal injury.

5.5 Conclusions

The burn scar model presented in this study resulted in hairless, hypopigmented scars that were thicker and weaker than scars formed using a dermatome model. These scars demonstrated more similarities to human HTS compared to dermatome wounds. As a result, this model may provide an improved platform for studying the pathophysiology of

100 hypertrophic scarring and for investigating current anti-scar therapies/development of new strategies with higher clinical benefit.

101

Chapter 6: Early versus Late Application of Pressure Garment Therapy

6.1 Introduction

Up to 91% of severe burn injuries result in abnormal scarring, which is characterized by pain, active contracture, redness and pruritus3,10-11,18,117. Standard care for the prevention and treatment of these scars includes pressure garment therapy (PGT), which involves the use of elastic garments to apply pressure to the scar tissue46-47,57,117. It is recommended that patients begin PGT as soon as the burn wound has re-epithelialized and can tolerate the pressure3,56=57. However, others deem it beneficial to use PGT during earlier stages of wound healing56,127. Therefore, optimal time of application has not been scientifically defined. It is known that scarring can be correlated with a prolonged inflammatory phase during healing128,130. One experimental study reported that when inflammation was reduced in porcine wounds via wound chambers, scar formation also was decreased129.

Similarly, early applications of pressure may affect inflammation and aid in the prevention of scar formation.

Furthermore, other forms of anti-scar therapy have been shown to be beneficial with early application to the wound. In a study investigating mechanical modulation of scar tissue,

102 a stress-shielding polymer device was used to off-load tension in incisional wounds131.

Early application of the device, at approximately one week post-surgery, resulted in significant improvements in scar appearance and reduction in scar hypertrophy131.

Another study evaluating the timing of silicone gel sheeting to facial scars revealed that early initiation of treatment resulted in improvements in Modified Vancouver Scar Scale

(MVSS)132.

The goal of this study was to examine if earlier application of pressure garment therapy also resulted in greater benefits than the recommended application after complete wound closure. A female Red Duroc pig burn + autograft scar model was used to assess and compare resultant scar morphology, thickness, contraction and biomechanics between control scars, scars treated with pressure one week post-surgery and scars treated with pressure approximately one month post-surgery.

6.2 Materials and Methods

6.2.1 General Animal Care, Generation and Excision of Burn Wounds

The Ohio State University Institutional Laboratory Animal Care and Use Committee

(ILACUC) approved protocols were used for all animal procedures and data collection.

Female Red Duroc pigs (n=6, 60 lbs; Isler Genetics Inc., Prospect, OH) were anesthetized with Telazol followed by Isoflurane and the dorsal trunk shaved and surgically prepared

103 with alternating scrubs of chlorhexidine 2% and alcohol 70% (Butler Schein, Columbus,

OH). Six total full-thickness wounds were created using a custom designed, electrically heated stainless steel burn stylus (1 inch x 1inch) per pig. Burn wounds were created by heating the stylus to 200 + 10°C and pressing the stylus on the skin at approximately 3 lbs of pressure for a contact time of 40 seconds (Figure 6.1A). Immediately after thermal injury, the pigs were draped in sterile fashion and burn sites were excised (Figure 6.1B).

Scalpels were used to incise along the border of each burn, between the zone of stasis and normal skin. Once the burns were outlined, the sites were tangentially excised to eschar level. A fentanyl patch (NOVAPLUS patch, Watson Laboratories Inc., 100 mcg) was placed in the pig ear pinna and removed three days post wounding. Animals received a single intramuscular injection of Buprenorphine (Buprenex, Reckitt Benckiser

Heatlhcare, 0.3 mg/ml) during recovery from anesthesia. Animals were fasted overnight prior to surgery, were housed in individual runs and fed with standard chow ad libitum.

Animals were euthanized following the completion of experiments on day 130 post wounding.

A B C

6.2.2 Split-Thickness Autograft Harvest, Meshing and Application

Two split-thicknes autografts were harvested from the dorsum of each pig (on either sides of the spinal cord, approximately 2 x 8 inch in dimension) using a Zimmer® Air

Dermatome (Zimmer Inc., Warsaw, IN) with a 2 inch width plate. Immediately after harvesting, the grafts were meshed using a Zimmer® Skin Graft Mesher with a cutter offering a 1-1/2 graft expansion ratio. Once the autografts were meshed, they were cut to fit dimensions of each wound and placed inside each excised area with slight overlap over normal skin on each edge (Figure 6.1C).

6.2.3 Bolsters and Dressing

The autografts were sutured to normal skin using Sofsilk™ Wax Coated Braided Silk sutures (Covidien llc, Mansfield, MA) with two knots per wound edge. Sutures were kept long (approximately 100 mm). Restore® Calcium Alginate Silver dressing

(Hollister Wound Care, Libertyville, IL) and sterile moist cotton bolsters were placed on top of each graft. The suture ends affixing the graft to normal skin were then used to secure the bolsters in place (Figure 6.2A). Bacitracin antibiotic ointment was applied to graft donor sites and covered with Restore dressing. To minimize mechanical trauma to the wounds, the bolsters were covered in a soft circumferential dressing: cotton sterile cast padding (BSN Medical Inc., Rutherford College, NC) and Scotchcast™ Plus casting

105 tape (3M, St. Paul, MN) (Figure 6.2B). Vetrap™ (3M, St. Paul, MN) and Elastikon®

(Johnson & Johnson) were used to secure the cast in place (Figure 6.2C).

A B C

6.2.4 Percent Engraftment

After bandages and bolsters were removed on day 7 post injury, photographs of each wound were taken with a scale in the field of view for wound area and graft take quantification. Percent graft take was quantified using computer planimetry (ImageJ) and defined as are of engraftment divided by total wound area at day 7 x 100. Data are presented as percent engraftment (mean + standard error of the mean (SEM)).

106

6.2.5 Pressure Garment Therapy and Pressure Quantification

Compression garments were designed and fabricated in house for pig use. Two sets of adjustable, wrap-style garments with Velcro closures were fabricated with a double layer of powernet fabric that was cut so that the long weave ran parallel to body circumference.

The Powernet used was a 9:1 nylon-spandex composite fabric (Shriners Hospitals for

Children, Cincinnati, OH). Boning with polyester covering (Joann Fabric and Craft

Stores, Columbus, OH) was added towards the center of the garment to prevent gathering and ensure comfort for the pigs. At 7 days post injury after bolster removal, the first set of compression garments were applied to two wounds out of six wounds for each pig at

15 + 2 mmHg of pressure. The location of garment application was randomized. The remaining four wounds were left untreated, and all wounds were covered with Vetrap and

Elastikon. At 35 days post injury, the second set of compression garments were applied to another two wounds per pig at 15 + 2 mmHg of pressure. The two wounds per pig receiving no treatment were left as controls. To evaluate the magnitude of pressure exerted by the compression garments on the wounds, a Kikuhime pressure sensor, which consists of an air bladder that transmits surface pressure (mmHg) to a digital data reporting device, was used (Advancis Medical USA, Plainview, NY). The pigs were visited every other day to check on garment position, pressure magnitudes and whether garments were soiled. The garments were replaced at least twice a week and applied at magnitudes of 15 + 2 mmHg, with more replacements as needed. The compression

107 garments were laundered by hand with regular detergent (no bleach or fabric softener added) and cold water and line dried to avoid heat degradation from the dryer.

6.2.6 Scar Morphology and Contraction

Photographs of the scars were taken at days 7, 35, 72, 100 and 130 post wounding. Each photograph of the scars (n=10 per group) was taken with a scale in the field view for scar area quantification. Computer planimetry using ImageJ software was used to quantify the contraction of the wounds. Wound contraction was defined as wound area at a specific time point divided by the original wound area (at day 7, time of first pressure garment therapy application) x 100. Data are presented as percent original area (mean + standard error of the mean (SEM)). General scar morphology was assessed at each time point as well.

6.2.7 Scar Thickness

At day 130, strips of tissue (approximately 15 x 60 mm) were excised from each scar parallel to the circumference of the pig. The scar sites were positioned centrally within the skin biopsies and specimens were cut to approximately 3 x 30 mm. Photographs with a scale in the field view were taken of each cut sample. The quantitatively assess scar thickness, images were processed using ImageJ with scar thickness calculated from top of

108 the epidermis to depth of scar tissue in the dermis and reported as average + standard error of mean (SEM).

6.2.8 Scar Biomechanics

Failure biomechanics were assessed at day 130 to assess the final strength of each scar.

The strips of tissue used to measure scar thickness were mounted into the grips of a

TestResources mechanical tester model 1000R12 (TestResources, Shakopee, MN). The skin samples were tested to failure at a strain rate of 2 mm/s using a 100lbf load transducer. Ultimate tensile strength and linear stiffness were reported as mean + SEM.

6.3 Results

6.3.1 Percent Engraftment

Total mean percent graft take was 89.41% + 4.66. One hundred percent engraftment was observed in animals with no damage to their dressing (Figure 6.3).

109

Figure 6.3. Example wound sites with 100% engraftment

6.3.2 Scar Morphology and Contraction

Engraftment of meshed split-thickness autografts can be seen in wound sites by day 7 post injury when the first set of compression garments are applied to two wounds per pig

(Figure 6.4). By day 35, it can be observed that wounds have fully healed and compression garments applied at day 7 have not interfered with healing of wounds

(Figure 6.4). Scars have contracted with time; untreated scars have maintained rectangular shapes, while scars treated at day 7 with pressure garments exhibit flattening in the vertical direction and have changed into a rhombus-like shape (Figure 6.4). By day

72, scars treated at day 7 with pressure maintain their rhombus-like shape and scars treated at day 35 begin to slant into a rhombus-like shape as well (Figure 6.4). The flattening observed in scars treated at day 7 is not seen in scars treated at day 35. Control wounds at day 72 have contracted significantly in the horizontal direction. By day 130, control scars were raised, hyperpigmented along the edges and are firm to the touch.

Contrastingly, scars that were treated at day 7 were almost level with surrounding

110 uninjured skin, similar in pigmentation to the surrounding tissue and not firm to the touch.

Figure 6.4. Representative photographs of control and pressure treated scars 7-130 post injury. Compression garments were applied at day 7 and day 35. Scale bar = 10mm.

Quantitative analysis of wound and scar area showed a significant decrease in wound area from 7 days to 35 days post injury (Figure 6.5). At day 35, scars that received compression garments 28 days prior were 82.0 + 3.8% original area as compared to scars not receiving any pressure, which were 77.0 + 2.6% original area (Figure 6.5). The size of untreated and treated scars increased in size throughout the study, due to the natural growth of the animal (day 35 – day 100, Figure 6.5). By day 130, scar contraction plateaued in both pressure treated groups with no statistical difference between average scar area at days 72 and 130 (Figure 6.5). In contrast, additional contraction could be 111 observed in control scars between day 100 and day 130. There was a significant difference in contraction between control scars and pressure treated scars (p < 0.02) at day 130.

Figure 6.5. Scar contraction presented as percent of original area, as a function of time and treatment.

6.3.3 Scar Thickness

Thicknesses of pressure treated scars are significantly lower than control scars, which were 4.44 + 0.07 mm thick. There was also a significant difference in thickness of scars

112 treated with pressure at day 7 (2.73 + 0.05 mm) in comparison to day 35 (3.37 + 0.06 mm) (Figure 6.6).

6.3.4 Scar Biomechanics

Although pressure garment therapy increased the linear stiffness (N/mm2) of scars compared to control scars, the difference was not significant (Figure 6.7B). Similarly, tensile testing of excised scar tissue revealed that the ultimate tensile strength (UTS) of the scars treated with PGT was stronger than that of control scars (Figure 6.7A). The

113

UTS of control scars was 5.19 MPa + 0.34, which was slightly lower than that of scars treated at day 7 (5.41 + 0.50) and day 35 (5.99 + 0.35) with pressure.

114

6.4 Discussion

The female red Duroc pig model was used to the study the efficacy of pressure garment therapy with an early application at day 7 and a later application at day 35 following a burn injury, excision and split-thickness autograft application. Significant decreases in in scar contraction were observed between pressure-treated scars and untreated controls. By day 130, most of the scars treated at day 7 and 35 are similar in size as the original wound, if not slightly bigger. This shows that these scars were unaffected by contraction and were able to grow with the pig.

Scar tissue treated with pressure at day 7 post wounding was significantly thinner than both untreated scars and scars treated at day 35. This could be due to the effects of compression on collagen structure, as previously reported. When human scars were treated with elastic bandages, the reticular collagen fibers began to resemble normal skin and were thinner52. Furthermore, it is widely believed that PGT may work by restricting nutrient and blood supply to the scar tissue, thereby decreasing collagen deposition, which could also explain the thinner pressure-treated scar tissue10-12.

6.5 Conclusions

Pressure application of 15 mmHg at an earlier time point (day 7) was found to be more effective at reducing scar thickness and improving clinical appearance than when applied

115 at a later time point (day 35). Furthermore, pressure application at both time points was found to be effective in reducing scar contraction. This study shows that pressure applied at an earlier time during wound healing might be beneficial and needs further investigation, especially for mechanisms of action behind these benefits.

116

Chapter 7: Summary and Future Work

A custom burn device with the ability to continually heat the burn stylus and actively control pressure and temperature was successfully developed to allow for more rapid and reproducible burn wounds. With uniform starting injury, the efficacy of pressure garment therapy was analyzed in a female Red Duroc pig burn model. The efficacy of pressure garments was confirmed with reduction in scar contraction and improvement in biomechanics compared to control scars. These improvements were modest, thus we sought to enhance the efficacy of the therapy by optimizing the garment fabrication design to deliver higher levels of pressure with less fatigue during use and as a result of care (i.e. laundering). In order to fabricate more effective pressure garments, two different fabrics were analyzed for magnitude of pressure delivered and fatigue during cyclic stretch and following laundering. The results suggest Powernet (an interwoven fabric comprised of 90% nylon and 10% spandex) as being more suitable than moleskin fabric for engineering compression garments as they were able to deliver more pressure with the same level of strain and fatigued less during use and after laundering.

Additionally, different injury models for hypertrophic scarring were developed and investigated. The results suggest that burn injury provide a scar model in female red

Duroc pigs that most closely resemble hypertrophic scarring in humans. Using this

117 knowledge, the efficacy of pressure garments fabricated from Powernet fabric was analyzed on a FRDP burn + split-thickness model. The results suggest that the early application (day 7 post-injury) was more effective than late application (day 35 post- injury). With this knowledge, future work will include investigating the molecular mechanisms governing pressure therapy and determining optimum pressure ranges for pressure garment therapy.

7.1 Future Work

7.1.1 Tension Off-Loading and Scar Development

Studies have linked wound tension to excessive fibroproliferation, which may be a result of FAK activation131,133-135. A study investigating incisions on fibroblast-specific FAK knockout mice reported significantly less inflammation and scar formation in the FAK knockout mice in comparison to control mice131. The results suggested increased FAK activation potentiated by tension triggers an inflammatory pathway (FAK-ERK-MCP-1), which induces fibrosis131,133. Future work will further investigate the link between FAK and scar formation by determining if tension and the structure of the extracellular matrix act on integrins to increase phosphorylation of focal adhesion kinase (FAK) in a FRDP burn scar model (Figure 7.1). One study reported that mechanical stimuli and actin alignment increased FAK phosphorylation in high glucose endothelial cells136. Collagen fibrils become highly aligned in the direction of tension in wounds. Therefore, the

118 increased alignment of ECM and mechanical activation of FAK may be a possible mechanism for increased collagen production and myofibroblasts differentiation found in exuberant fibrosis.

Figure 7.1. Schematic diagram of proposed mechanisms of scar formation within the burn wound. (Modified from Gurtner et al.131).

Future work will investigate alterations in stress state within the wound using compression garments at different magnitudes of delivered pressure. Gurtner et al. showed that incisional wounds in red Duroc pigs showed an elevated stress state and resulted in scarring131,135. However, when the stress was off-loaded with a stress- shielding device during wound healing, scar hypertrophy was significantly reduced in stress-shielded wounds in comparison to scars under high tension (Figure 7.2)131,135.

119

Furthermore, stress-shielded scars also exhibited epithelial thickening and restoration of rete ridges (Figure 7.2)131,135. Although his findings are significant, more work must be done to apply it to burns, as burn wounds have a more complex stress state in comparison to excisional wounds. Burn wounds have a different shape, larger area and complex strain magnitudes that need to be accounted for. Thus, the design of the compression garments will be key to this experiment.

7.1.2 Optimal Pressure Range (Pressure Garment Therapy)

In addition, future work will include investigating the optimum pressure range for effective treatment via compression garments. Many studies have reported on conflicting pressures. Some studies have reported benefits with very low pressure of 5-15 mmHg117, while several authors recommend pressures of 15-24 mmHg46,52,56. Unspecified higher pressures exceeding capillary pressure (~25 mmHg) have also been proposed4,56.

However, there is evidence that 30-40 mmHg of pressure cause discomfort for patients and can be potentially harmful117,137.

7.1.3 Laser Therapy

Current and on-going collaboration with Dr. J. Kevin Bailey has involved validating another model for burn scars using excision and immediate application of split-thickness autografts. Future work will include comparing this model to our current FRDP burn scar model. This collaborative team will also investigate the efficacy of laser therapy in the treatment of burn scars. Preliminary data has been collected on the efficacy of pulsed dye laser and ablative fractional carbon dioxide laser on the treatment of scarring in the

FRDP burn + autograft scar model. Further analysis must be performed. Future work will include determining optimum laser fluence, number of passes and frequency of treatment.

121

References

1. Gawkrodger DJ. Dermatology, An Illustrated Colour Text. 3rd ed. Edinburgh: Churchill Livingstone; 2002. 2. Zaidi Z, Lanigan S. Skin: structure and function. Dermatology Clin Pract. 2010:1- 15. 3. Gauglitz GG, Korting HC, Pavicic T, Ruzicka T, Jeschke MG. Hypertrophic scarring and keloids: pathomechanisms and current and emerging treatment strategies. Mol Med. 2011;17(1-2):113-125. 4. Tredget EE, Nedelec B, Scott PG, Ghahary A. Hypertrophic scars, keloids, and contractures: the cellular and molecular basis for therapy. Surg Clin North Am. 1997;77:701-30. 5. Zhu KQ, Engrav LH, Tamura RN, Cole JA, Muangman P, Carrougher GJ, Gibran NS. Further similarities between cutaneous scarring in the female, red Duroc pig and human hypertrophic scarring. Burns. 2004;30:518-530. 6. Slemp AE, Kirschner RE. Keloids and scars: a review of keloids and scars, their pathogenesis, risk factors, and management. Curr Opin Pediatr. 2006;18:396-402. 7. Gangemi EN, Gregori D, Berchialla P, Zingarelli E, Cairo M, Bollero D, Ganem J, Capocelli R, Cuccuru F, Cassano P, Risso D, Stella M. Epidemiology and Risk Factors for Pathologic Scarring After Burn Wounds. Arch Facial Plast Surg. 2008;10(2):93-102. 8. Seo BF, Lee JY, Jung SN. Models of Abnormal Scarring. Biomed Res Int. 2013;1-8. 9. Juckett G, Hartman-Adams H. Management of keloids and hypertrophic scars. Am Fam Physician. 2009;80(3):253-260. 10. Ripper S, Renneberg B, Landmann C, Weigel G, Germann G. Adherence to pressure garment therapy in adult burn patients. Burns. 2009;35:657-664. 11. Zurada JM, Kriegel D, Davis IC. Topical treatments for hypertrophic scars. J Am Acad Dermatol. 2006;55:1024-31. 12. English RS, Shenefelt PD. Keloid and hypertrophic scars. Dermatol Surg. 1999;25:631-8. 13. Oliveira GV, Hawkins HK, Chinkes D, Burke A, Tavares AL, Ramos-e-Silva M, Albrecht TB, Kitten GT, Herndon DN. Hypertrophic versus non hypertrophic scars compared by immunohistochemistry and laser confocal microscopy: type I and III collagens. Int Wound J. 2009;6(6):445-52. 14. Qu L, Liu A, Zhou L, He C, Grossman PH, Moy RL, Mi QS, Ozog D. Clinical 122

and Molecular Effects on Mature Burn Scars After Treatment With a Fractional CO2 Laser. Lasers Surg Med. 2012;44:517-524. 15. Candy LH, Cecilia LT, Ping ZY. Effect of different pressure magnitudes on hypertrophic scar in a Chinese population. Burns. 2010;36:1234-41. 16. Lee SJ, Yeo IK, Kang JM, Chung WS, Kim YK, Kim BJ, Park KY. Treatment of Hypertrophic Burn Scars by Combination Laser-Cision and Pinhole Method Using a Carbon Dioxide laser. Lasers Surg Med. 2014;46:380-384. 17. Vrijman C, van Drooge AM, Limpens J, Bos JD, van der Veen JP, Spuls PI, Wolkerstorfer A. Laser and intense pulsed light therapy for the treatment of hypertrophic scars: A systemic review. Br J Dermatol. 2011;165:934-942. 18. Goel A, Shrivastava P. Post-burn scars and scar contractures. Indian J Plast Surg. 2010;43(Suppl):S63-S71. 19. Clark RA. Biology of dermal wound repair. Dermatol Clin. 1993;11:647-66. 20. Alster TS. Laser treatment of hypertrophic scars, keloids, and striae. Dermatol Clin. 1997;15:419-29. 21. Alster TS, Nanni CA. Pulsed dye laser treatment of hypertrophic burn scars. Plast Reconstr Surg. 1998;102(6):190-5. 22. American Burn Association National Burn Repository. 2014. Burn incidence and treatment in the US: 2014 fact sheet. http://www.ameriburn.org. 23. Hettiaratchy S, Dziewulski P. ABC of burns: pathophysiology and types of burns. BMJ. 2004;328(7453):1427-1429. 24. Hettiaratchy S, Papini R. Initial management of a major burn: II--assessment and resuscitation. BMJ. 2004;329(7457):101-103. 25. Atiyeh BS, Gunn W, Hayek SN. State of the Art in Burn Treatment. World J Surg. 2005;29:131-148. 26. Bergmann U, Scheffer J, Köller M, Schönfeld W, Erbs G, Müller FE, König W. Induction of inflammatory mediators (histamine and leukotrienes) from rat peritoneal mast cells and human granulocytes by Pseudomonas aeruginosa strains from burn patients. Infect Immun. 1989;57(7):2187-95. 27. Rocha E, Silva M, Antonio A. Release of bradykinin and the mechanisms of production of thermic edema (45C) in the rat’s paw. Med Exp. 1960;3:371-378. 28. Belenkiy SM, Buel AR, Cannon JW, Sine CR, Aden JK, Henderson JL, Liu NT, Lundy NT, Lundy JB, Renz EM, Batchnisky AI, Cancio LC, Chung KK. Acute respiratory distress syndrome in wartime military burns: application of the Berlin criteria. J Trauma Acute Care Surg. 2014;76(3):821-7. 29. Çakir B, Yeǧen BÇ. Systemic responses to burn injury. Turkish J Med Sci. 2004;34(4):215-226. 30. Williams FN, Herndon DN, Jeschke MG. The hypermetabolic response to burn injury and interventions to modify this response. Clin Plast Surg. 2009;36(4):583- 596. 31. Tompkins RG. Survival of children with burn injuries. Lancet. 2012;379(9820):983-984.

123

32. Sheridan RL, Remensnyder JP, Schnitzer JJ, Schulz JT, Ryan CM, Tompkins RG. Current expectations for survival in pediatric burns. Arch Pediatr Adolesc Med. 2000;154(3):245-9. 33. Kamolz LP, Horch RE, Jeschke MG, Kuntscher M, Brychta P. eds. Handbook of Burns. Vienna: Springer, 2012. Print. 34. Aasi SZ. Z-plasty made simple. Dermatol Res Pract. 2010;2010:982623. 35. Kaur G, Malik S, Grover D, Sharma A. Surgical management of aberrant labial frenum using modified z-frenuloplasty: a case report. IJSS Case Reports & Reviews. 2015;1(10):5-8. 36. Lalwani AK. Current Diagnosis and Treatment in Otolaryngology Head and Neck Surgery. 3rd ed. New York: McGraw Hill Medical, 2012. 37. Wilhelmi BJ, Weiner LJ, Polk HC, Molnar JA. Widened and hypertrophic scar healing treatment & management. http://emedicine.medscape.com/article/1298541-overview. 38. Katz BE. Silicone gel sheeting in scar therapy. Cutis. 1995;56(1):65-7. 39. Rabello FB, Souza CD, Júnior JAF. Update on hypertrophic scar treatment. Clinics. 2014;69(8):565-573. 40. Juckett G, Hartman-Adams H. Management of keloids and hypertrophic scars. Am Fam Physician. 2009;80(3):253-260. 41. Puri N, Talwar A. The efficacy of silicone gel for the treatment of hypertrophic scars and keloids. J Cutan Aesthet Surg. 2009;2(2):104-106. 42. O’Brien L, Pandit A. Silicon gel sheeting for preventing and treating hypertrophic and keloid scars. Cochrane Database Syst Rev. 2006;(1):CD003826. 43. Kiil J. Keloids treated with topical injections of triamcinolone acetonide (kenalog). Immediate and long-term results. Scad J PLast Reconstr Surg. 1977;11(2):169-72. 44. Leventhal D, Furr M, Reiter D. Treatment of Keloids and Hypertrophic Scars: A Meta-analysis and Review of the Literature. Arch Facial Plast Surg. 2006;8(6):362-368. 45. Darzi MA, Chowdri NA, Kaul SK, Khan M. Evaluation of various methods of treating keloids and hypertrophic scars: a 10-year follow-up study. Br J Plast Surg. 1992;45(5):374-9. 46. Cheng J, Evans J, Leung K, Clark J, Choy T, Leung P. Pressure therapy in the treatment of post-burn hypertrophic scar – a critical look into its usefulness and fallacies by pressure monitoring. Burns 1984;10:154-63. 47. Puzey G. The use of pressure garments on hypertrophic scars. J Tissue Viability. 2002;12(1):11-5. 48. Perkins K, Davey Rb, Wallis K. Current materials and techniques used in burn scar management programme. Burns 1987;13:406-10. 49. Spilker G, Toklsdorf-Kremmer A, Kuppers S. Kompressionstherapie [Compression Therapy]. In: Bruck, Muller, Steen (Hrsg.). Handbuch der Verbrennungstherapie. Landsberg; ecomed; 2002. 50. Kischer W, Shetlar M, Shetlar C. Alteration of hypertrophic scars induced by mechanical pressure. Arch Dermatol. 1975;111:60-4. 124

51. Renò F, Sabbatini M, Stella M, Magliacani G, Cannas M. Effect of in vitro mechanical compression on Epilysin (matrix metalloproteinase-28) expression in hypertrophic scars. Wound Repair Regen. 2005;13(3):255-261. 52. Costa AM, Peyrol S, Porto LC, Comparin JP, Foyatier JL, Desmouliere A. Mechanical forces induce scar remodeling. Study in non-pressure-treated versus pressure-treated hypertrophic scars. Am J Pathol. 1999;155:1671-9. 53. Anzarut A, Olson J, Singh P, Rowe BH, Tredget EE. The effectiveness of pressure garment therapy for the prevention of abnormal scarring after burn injury: a meta-analysis. J Plast Reconstr Aes. 2009;62:77-84. 54. Mustoe TA, Cooter RD, Gold MH, Hobbs FDR, Ramelet AA, Shakespeare PG. International clinical recommendations for scar management. Plast Reconstr Surg. 2002;110:56-571. 55. Larson DL, Abston S, Evans EB, Dobrkovsky M, Linares HA. Techniques for decreasing scar formation and contractures in the burned patient. J Trauma. 1971;11:807-23. 56. Staley MJ, Richard RL. Use of pressure to treat hypertrophic burn scars. Adv Wound Care. 1997;10:44-46. 57. Macintyre I, Baird M. Pressure garment for the use in the treatment of hypertrophic scars – a review of problems associated with their use. Burns 2006;32:10-15. 58. Giele H, Liddiard K, Currie K, Wood F. Direct measurement of cutaneous pressure generated by pressure garments. Burns. 1997;23(2):137-141. 59. Hantash BM, Mahmood MB. Fractional photothermolysis: A novel aesthetic laser surgery modality. Dermatol Surg. 2007;33:525-34. 60. Manstein D, Herron GS, Sink RK, Tanner H, Anderson RR. Fractional photothermolysis: a new concept for cutaneous remodeling using microscopic patterns of thermal injury. Lasers Surg Med. 2004;34:426-38. 61. Hultman CS, Edkins RE, Lee CN, Calvert CT, Cairns BA. Shine on: Review of Laser- and Light-Based Therapies for the Treatment of Burn Scars. Dermatol Res Pract. 2012:1-9. 62. Waibel JS, Beer K, Narurkar V, Alster T. Preliminary observations on fractional ablative resurfacing devices: Clinical impressions. J Drugs Dermatol. 2009;8:481- 485. 63. Waibel JS, Wulkan AJ, Shumaker PR. Treatment of Hypertrophic Scars Using Laser and Laser Assisted Corticosteroid Delivery. Laser Surg Med. 2013;45:135- 140. 64. Alster TS. Improvement of erythematous and hypertrophic scars by the 585-nm flashlamp-pumped pulsed dye laser. Ann Plast Surg. 1994;32:186. 65. Avram MM, Tope WD, Yu T, Szachowicz E, Nelson JS. Hypertrophic scarring of the neck following ablative fractional carbon dioxide laser resurfacing. Lasers Surg Med. 2009;41(3):185-8. 66. Lee SJ, Kim JH, Lee SE, Chung WS, Oh SH, Cho SB. Hypertrophic Scarring After Burn Scar Treatment with a 10,600-nm Carbon Dioxide Fractional Laser. Dermatol Surg. 2011;37:1168-1172. 125

67. Zhu KQ, Engrav LH, Gibran NS, Cole JK, Matsumura H, Piepkorn M, Isik FF, Carrougher GF, Muangman PM, Yunusov MY, Yang TM. The female, red Duroc pig as an animal model of hypertrophic scarring and the potential role of the cones of skin. Burns. 2003;29:649-664. 68. Liang Z, Engrav LH, Muangman P, Muffly LA, Zhu KQ, Carrougher GJ, Underwood RA, Gibran NS. Nerve quantification in female red Duroc pig (FRDP) scar compared to human hypertrophic scar. Burns 2004;30:57-64. 69. Gallant-Behm CL and Hart DA. Genetic analysis of skin wound healing and scarring in a porcine model. Wound Rep Regen 2006;14(1):46-54. 70. Morris DE, Wu L, Zhao LL, Bolton L, Roth SI, Ladin DA, Mustoe TA. Acute and chronic animal models for excessive dermal scarring: quantitative studies. Plast Reconstr Surg 1997;100:674-81 71. Erlich HP, Needle AL. Wound healing in tight-skin mice: delayed closure of excised wounds. Plast Reconstr Surg. 1983;72(2):190-198. 72. Harunari N, Zhu KQ, Armendariz RT, Deubner H, Muangman P, Carrougher GJ, Isik FF, Gibran NS, Engrav LH. Histology of the thick scar on the female, red Duroc pig: Final similarities to human hypetrophic scar. Burns. 2006;32(6):669- 677. 73. Ghahary A, Shen YJ, Scott PG, Tredget EE. Immunolocalization of TGF-beta 1 in human hypertrophic scar and normal dermal tissues. Cytokin. 1995;7:184-90. 74. Krein PM, Winston BW. Roles for insulin-like growth factor I and transforming growth factor-beta in fibrotic lung disease. Chest. 2002;122:289S-93S. 75. Weber IT, Harrison RW, Iozzo RV. Model structure of Decorin and implications for collagen fibrillogenesis. J Biol Chem. 1996;271:31767-70. 76. Scott PG, Dodd CM, Tredget EE, Ghahary A, Rahemtulla F. Immunohistochemcial localization of the proteoglycans Decorin, biglycan and Versican and transforming growth factor-beta in human post-burn hypertrophic and mature scars. Histopathology. 1995;36:423-31. 77. Engrav LH, Heimbach DM, Reus JL, Harnar TJ, Marvin JA. Early excision and grafting vs. nonoperative treatment of burns of indeterminant depth: a randomized prospective study. J Trauma 1983;23:1001–4. 78. Askoy B, Atakan N, Askoy HM, Tezel GG, Renda N, Özkara HA, Önder E. Effectiveness of topical zinc oxide application on hypertrophic scar development in rabbits. Burns 2010;36:1027-1035. 79. Berman B, Perez OA, Konda S, et al. A review of the biologic effects, clinical efficacy, and safety of silicone elastomer sheeting for hypertrophic and keloid scar treatment and management. Dermatol Surg 2007;33(11):1291-1303. 80. Wang JF, Ding J, Jiao H, Honardoust D, Momtazi M, SHankowsky HA, Tredget EE. Human hypertrophic scar-like nude mouse model: Characterization of the molecular and cellular biology of the scar process. Wound Rep Regen 2011;19:274-285. 81. Aksoy HM, Vargel I, Canter HI, Erk YE, Sargon M, Pinar A, Tezel GG. A New Experimental Hypetrophic Scar Model in Guinea Pigs. Aesth Plasat Surg 2002;26:388-396. 126

82. Engrav LH, Garner WL, Tredget EE. Hypertrophic scars. Journal of Burn Care & Research 2007;28(4):1-5. 83. Zhu KQ, Carrougher GJ, Gibran NS, Isik FF, Engrav LH. Review of the female Duroc/Yorkshire pig model of human fibroproliferative scarring. Wound Rep Reg 2007;15:S32-S39. 84. Singer AJ, Taira BR, Anderson R, McClain SA, Rosenberg L. Does pressure matter in creating burns in a porcine model? J of Burn Care and Res 2010;31(4):646-651. 85. Papp A, Kiraly K, Harma M, Lahtinen T, Uusaro A, Alhava E. The progression of burn depth in experimental burns: a histological and methodological study. Burns 2004;30:684-690 86. Gaines C, Poranki D, Du W, Clark RA, Van Dyke M. Development of a porcine deep partial thickness burn model. Burns 2013;39(2):311-9. 87. Kempf M, Cuttle L, Liu PY, Wang XQ, Kimble RM. Important improvements to porcine skin burn models, in search of the perfect burn. Burns 2009;35:454-455. 88. Diegelmann RF, Evans MC. Wound Healilng: An Overview of Acute, Fibrotic and Delayed Healing. Frontiers of Bioscience 2004;9:283-289. 89. Roy S, Elgharably H, Sinha M, Ganesh K, Chaney S, Mann E, Miller, C., Khanna, S., Bergdall, V.K., Powell HM, Cool CH, Gordillo, GM, Wozniak DJ, Sen CK. Mixed-species Biofilm Compromises Burn Wound Healing by Disrupting Epidermal Barrier Function Journal of Pathology In Press 2014 Apr 25 90. Baumbach PL. Temperature Measuring Device. Patent WO 2002066946 A2. 29 Aug. 2002. Print. 91. American Burn Association. 2000. Burn incidence and treatment in the US: 2000 fact sheet. http://www.ameriburn.org. 92. Church D, Elsayed S, Reid O, Winston B, Lindsay R. Burn Wound Infections. Clinical Microbiology Reviews 2006: 403-434. 93. Reid WH, Evans JH, Naismith RS, Tully AE, Sherwin S. Hypertrophic scarring and pressure therapy. Burns 1987;13:S29-32. 94. Van den Kerckhove E, Stappaerts K, Fieuws S, Laperre J, Massage P, Flour M, Boeck W. The assessment of erythema and thickness on burn related scars during pressure garment therapy as a preventive measure for hypertrophic scarring. Burns 2005;31:696-702. 95. Fournier R, Pierard GE. Skin tensile strength modulation by compressive garments in burn patients. A pilot study. J Med Eng Technol 2000;24:277-280. 96. Garciavelasco M, Ley R, Mutch D, Surkes N, Williams HB. Compression treatment of hypertrophic scars in burned children. Can J Surg 1978;21:450-452. 97. Kloti J, Ponchon J. Conservative treatment using compression suits for second and third degree burns in children. Burns 1982;8:180-187. 98. Baur P, Larson D, Stacey T, Barratt G Dobrkovsky M. Ultrastructural analysis of pressure treated human hypertrophic scars. J Trauma 1976;16:958-967. 99. Rose M, Deitch E. The effective use of a tubular compression bandage, tubigrip, for burn scar therapy in a growing child. J Burn Care Rehabil 1983;4:197-201. 100. Kealey G, Jensen K, Laubenthal K, Lewis R. Prospective randomized comparison 127

of two types of pressure therapy garments. J Burn Care Rehabil 1990;11:334-336. 101. Linares HA, Larson DL, Willis-Galstaun BA. Historical notes on the use of pressure in the treatment of hypertrophic scars or keloids. Burns 1993;19(1):17- 21. 102. Linares HA, Larson DL, Baur PS. Influence of mechanical forces on burn scar contracture and hypertrophy. Symposium on Basic Science in Plastic Surgery 1975;101. 103. Yan LH, Ping LTW, Ping ZY. Effect of different pressure magnitudes on hypertrophic scar in a Chinese population. Burns. 2010;36(8):1234-1241. 104. Mann R. Do custom-fitted pressure garments provide adequate pressure? J of Burn Care and Rehabilitation. 1997;18:247. 105. Rose MP and Deitch EA. The clinical use of a tubular compression bandage, Tubigrip, for burn-scar therapy: a critical analysis. Burns. 1985;12:58-64. 106. Ghahary A, Shen YJ, Nedelec B, Scott PG, Tredget EE. Enhanced expression of mRNA for insulin-like growth factor-1 in post-burn hypertrophic scar tissue and its fibrogenic role by dermal fibroblasts. Mol Cell Biochem. 1995;148:25-32. 107. Sayani K, Dodd CM, Nedelec B, Shen YJ, Ghahary A, Tredgett EE, et al. Delayed appearance of decorin in healing burn scars. Histopathology. 2000;36:262-72. 108. Scott PG, Dodd CM, Tredget EE, Ghahary A, Rahemtulla F. Chemical characterization and quantification of proteoglycans in human post-burn hypertrophic and mature scars. Clin Sci. 1996;90:417-425. 109. Abramoff MD, Magalhaes PPJ, Ram SJ. Image Processing with ImageJ. Biophotonics International, volume 11, issue 7, pp. 36-42, 2004. 110. Powell, HM, Supp DM, Boyce ST. Influence of electrospun collagen on wound contraction of engineered skin substitutes. Biomaterials 2008,29(7);834-43. 111. Powell HM, Boyce ST. Wound Closure with EDC Cross-linked Skin Substitutes Grafted to Athymic Mice. Biomaterials 2007;28(6):1084-92. 112. Christoforidis JB, Wang J, Jiang A, Willard J, Pratt C, Abdel-Rasoul M, Roy S, Powell HM. The effect of intravitreal bevacizumab and ranibizumab on cutaneous tensile strength during wound healing. Clinical Opthamology 2013;7:185-191. 113. Wong VW, Rustad KC, Akaishi S, Sorkin M, Glotzbach JP, Januszyk M, Nelson ER, Levi K, Paterno J, Vial IN, Kuang AA, Longaker MT, Gurtner GC. Focal adhesion kinase links mechanical force to skin fibrosis via inflammatory signaling. Nature Medicine 2012;18(1):148-153. 114. Gurtner GC, Dauskardt RH, Wong VW, Wu K, Vial IN, Padois K, et al. Improving Cutaneous Scar Formation by Controlling the Mechanical Environment Annals of Surgery 2011;254:217-225. 115. Serpooshan V, Muja N, Marelli B, Nazhat SN. Fibroblast contractility and growth in plastic compressed collagen gel scaffolds with microstructures correlated with hydraulic permeability. J Biomed Mater Res A 2011;94(6):609- 20. 116. Doillon CJ, Dunn MG, Bender E, Silver FH. Collagen fiber formation in repair tissue: development of strength and toughness. Collagen and Related Research 128

1985;5(6):481-492. 117. Atiyeh BS, El Khatib AM, Dibo SA. Pressure garment therapy (PGT) of burn scars: evidence-based efficacy. Ann Burns Fire Disasters 2013;26(4):205-212. 118. Macintyre L, Ferguson R. Pressure garment design tool to monitor exerted pressures. Burns.2013;39:1073–82. 119. "JOBST® Medical Legwear Opaque."Compression Stockings. BSN Medical, 2013. Web. 16 June 2014 120. Bock O, Schmid-Ott G, Malewski P, et al. Quality of life of patients with keloid and hypertrophic scarring. Arch Dermatol Res. 2006;297:433-438. 121. Kim JY, Willard JJ, Supp DM, Roy S, Gordillo GM, Sen CK, and Powell HM. Hypertrophic Scar Biomechanics Following Pressure Garment Therapy. Plast Reconstr Surg. 2015. 122. Davidson JM, Yu F, Opalenik SR. Splinting Strategies to Overcome Confounding Wound Contraction in Experimental Animal Models. Adv Wound Care. 2013;2(4):142-148. 123. Tuan TL, Nichter LS. The molecular basis of keloid and hypertrophic scar formation. Mol Med Today. 1998:19-24. 124. van Zuijlen PP, Ruurda JJ, van Veen HA, van Marle J, van Trier AJ, Groenevelt F, Kreis RW, Middlekoop E. Collagen morphology in human skin and scar tissue: no adaptations in response to mechanical loading at joints. Burns. 2003;29(5):423-31. 125. Zhu KQ, Carrougher GJ, Gibran MS, Isik FF, Engrav LH. Review of the female Duroc/Yorkshire pig model of human fibroproliferative scarring. Wound Repair Regen. 2007;15(Suppl 1):S32-S39. 126. Moshref SS, Mufti ST. Keloid and Hypertrophic Scars: Comparative Histopathological and Immunohistochemical Study. JKAU: Med Sci. 2010;17(3):3-22. 127. Sherris D, Larrabee W, Murakami C. Management of scar contractures, hypertrophic scars, and keloids. Otolaryngol Clin North Am. 1995;28(5):1057-68. 128. Robson MC, Barbul A. Guidelines for the best care of chronic wounds. Wound Repair Regen. 2006;14(6):647-8. 129. Reish RG, Zuhaili B, Bergmann J, Aflaki P, Koyama T, Hackl F, Waisbren E, Canseco JA, Verna KD, Eriksson E, Yao F. Modulation of scarring in a liquid environment in the Yorkshire pig. Wound Repair Regen. 2009;17(6):806-16. 130. Junker JPE, Kamel RA, Caterson EJ, Eriksson E. Clinical impact upon wound healing and inflammation in moist, wet, and dry environments. Adv Wound Care. 2013;2(7):348-356. 131. Wong VW, Beasley B, Zepeda J, Reinhold HD, Yock PG, Longaker MT, Gurtner GC. A Mechanomodulatory Device to Minimize Incisional Scar Formation. Adv Wound Care. 2013;2(4):185-194. 132. Parry I, Sen S, Palmieri T, Greenhalgh D. Nonsurgical Scar Management of the Face: Does Early Versus Late Intervention Affect Outcome? J Burn Care Res. 2013;34(5):569-75. 133. Wong VW, Rustad KC, Akaishi S, Sorkin M, Glotzbach JP, Januszyk M, Nelson 129

ER, Levi K, Paterno J, Vial IN, Kuang AA, Longaker MT, Gurtner GC. Focal adhesion kinase links mechanical force to skin fibrosis via inflammatory signaling. Nature Med. 2012;18(1):148-153. 134. Januszyk M, Wong VW, Bhatt KA, Vial IN, Paterno J, Longaker MT, Gurtner GC. Mechanical offloading of incisional wounds is associated with transcriptional downregulation of inflammatory pathways in a large animal model. Organogenesis. 2014;10(2):180-193. 135. Gurtner GC, Dauskardt RH, Wong VW, Bhatt KA, Wu K, Vial IN, Padois K, Korman JM, Longaker MT. Improving Cutaneous Scar by Controlling the Mechanical Environment: Large Animal and Phase I Studies. Ann Surg. 2011;0(0):1-9. 136. Kemeny SF, Figueros DS, Clyne AM. Hypo- and Hyperglycemia Impair Endothelial Cell Actin Alignment and Nitric Oxide Synthase Activation in Response to Shear Stress. PLoS One. 2013;8(6):e66176. 137. Chan AP, Fan J. Effect of clothing pressure on the tightness sensation of girdles. Int J Cloth Sci Technol. 2002;14(2):100-10.

130