THE UNIVERSITY OF NEW SOUTH WALES
Faculty of Engineering
Laser-Activated Biomaterials for Tissue Repair
Antonio Lauto
Thesis Submitted for the Degree of Doctor of Philosophy, 2005
Supervisor: A/Professor Albert Avolio
Co-Supervisor: Dr John Foster
1 Declaration
I hereby declare that this thesis submission is my own work and that to the best of my knowledge, it contains no material previously published or written by another person nor material which to substantial extent has been accepted for the award of any other degree or diploma of a university or other institute of higher learning, except where due acknowledgment is made in the text.
Antonio Lauto
2 Acknowledgments
I would like to thank my supervisors Albert Avolio and John Foster for their precious help and suggestions throughout my investigations.
I am also indebt to Drs Stoodley, Liao, Esposito, Mingin, Sarris and McKenzie for their collaborative effort in the surgical procedures and tissue histology preparation.
A special thank to Michael Doran and Lynn Ferris for their help during the cytotoxic tests of the solders and adhesives. I am grateful to Fernando Camacho, Nick Zweneveld,
Katie Levick and Jenny Norman for revising part of the manuscript and processing samples for AFM and SEM analysis. I would like also to thank Dr Poole-Warren for the initial collaboration and supervision of the project. Finally, a cheerful thanks to Damia for her patience and verve, Elda for her striking style, Dr James Hook for his friendship and Dr Bruce Lachter for his invaluable advises.
This work was partially funded by the ARC discovery grant # DPO345899 and partially by an Engineering UNSW Faculty Grant (2004).
3 To Jeroslav Seifert,
…. waiting for the greatest miracle: criminals, politicians, capitalists, soldiers and priests turned into poets!
If You Call Poetry... If you call poetry a song - and people often do - then I've sung all my life. And I marched with those who had nothing, who lived from hand to mouth, I was one of them. I sang of their sufferings, their faith, their hopes, and I lived with them through whatever they had to live through. Through their anguish, weakness and fear and courage and poverty's grief. And their blood, whenever it flowed, spattered me. Always it flowed in plenty in this land of sweet rivers, grass and butterflies and passionate women. Of women, too, I sang. Blinded by love I staggered through my life, tripping over dropped blossoms or a cathedral step. Jeroslav Seifert
4 Abstract
Background. Laser tissue repair usually relies on haemoderivate solders, based on serum albumin. These solders have intrinsic limitations that impair their widespread use, such as limited repair strength, solubility and brittleness. Furthermore, the solder activation temperature (65-70 0C) can induce significant damage to tissue.
In this study, new laser-activated biomaterials for tissue repair were developed to overcome some of the shortcomings of traditional solders.
Materials and Methods. Solder strips were welded onto sheep intestine using a diode laser. The laser delivered continuously a power of 170 ± 10 mW at l = 808 nm, through a multimode optical fiber (core size = 200 mm) to achieve a dose of 10.8 ± 0.5 J/mg.
The solder thickness and surface area were 0.15 ± 0.01mm and 12.6 ± 1.0 mm2 respectively. The solder contained albumin, indocyanine green, water and a natural crosslinker for amino groups: genipin.
Flexible and insoluble strips of chitosan adhesive (surface area ~34 mm2, thickness ~20
µm) were also developed and bonded on sheep intestine with a laser fluence and irradiance of 52 ± 2 J/cm2 and ~15 W/cm2 respectively. The temperature between tissue and adhesive was measured using small thermocouples. The strength of repaired tissue was tested by a calibrated tensiometer. The adhesive was also bonded in vivo to the sciatic nerve of rats to assess the thermal damage induced by the laser (fluence = 65 ±
11 J/cm2, irradiance = 15 W/cm2) four days post-operatively. Finally, fibroblasts were cultured in extracted media from chitosan adhesive to assess cytotoxicity.
5 Results. The repair strength of the genipin-albumin solder was double that of traditional albumin solders (0.21 ± 0.04 vs. 0.11 ± 0.04 N, n=30). Chitosan adhesives successfully repaired intestine tissue, achieving a repair strength of 0.50 ± 0.15 N (shear stress = 14.7
± 4.7 KPa, n=30) at a temperature of 60-65 0C. The laser caused demyelination of axons at the operated site; nevertheless, the myelinated axons retained their normal morphology proximally and distally. Media extracted from chitosan adhesive showed negligible toxicity to fibroblasts.
Conclusion. A novel chitosan-based adhesive has been developed, which is insoluble, flexible and adheres firmly to tissue upon infrared laser activation. Further research is needed to reduce the thermal damage to the tissue.
6 Table of Contents
Abbreviations 10 List of Figures 11 List of Tables 14
Chapter 1. Background 15 1.1 Introduction 15 1.2 Fibrin Sealant 17 1.3 Cyanoacrylate Glues 21 1.4 Laser-Activated Glues 23 1.5 PEG Glues 37 1.6 Chitosan Glues 38 1.7 Genepin 48 1.8 Objectives 49
Chapter 2. Sutureless Laser-Soldering Technique for Reversal Vasectomy 50
2.1 Introduction 50 2.2 Materials and Methods 54 2.2.1 The Protein Solder 54 2.2.2 The Laser System 55 2.2.3 Chitosan Film 55 2.2.4 Thermogravimetric Analysis (TGA) 55 2.2.5 Stent Preparation 56 2.2.6 Elasticity Test 57 2.2.7 Stent Self-Expansion 58 2.2.8 Laser Tissue Soldering 59 2.2.9 Statistical Analysis 61 2.3 Results 62 2.3.1 Thermogravimetric Analysis 62 2.3.2 Stent Preparation 62 2.3.3 Elasticity Test 64 2.3.4 Stent Expansion 64
7 2.3.5 Laser Tissue Soldering 65 2.4 Discussion 67
Chapter 3. Albumin-Genipin Solder for Laser Tissue Repair 72 3.1 Introduction 72 3.2 Materials and Methods 76 3.2.1 The Laser System 76 3.2.2 Solder Preparation 76 3.2.3 Tissue Soldering 77 3.2.4 Tensile Strength 79 3.2.5 Solder Attenuation 79 3.2.6 Cytotoxic Assay 80 3.3 Results 82 3.3.1 Tensile Strength 82 3.3.2 Solder Attenuation 83 3.3.3 Cytotoxicity Assay 84 3.4 Discussion 85
Chapter 4. Chitosan Adhesive for Laser Tissue Repair 91 4.1 Introduction 91 4.2 Materials and Methods 93 4.2.1 Chitosan Adhesive Films 93 4.2.2 Adhesive Attenuation 95 4.2.3 Laser Tissue Repair (LTR) 95 4.2.4 Tensiometer Measurements 97 4.2.5 13C-NMR 98 4.2.6 Thermogravimetric Analysis (TGA) 99 4.2.7 Differential Scanning Calorimetry (DSC) 99 4.2.8 Contact Angle 100 4.2.9 Atomic Force Microscopy (AFM) 100 4.2.10 Young’s Modulus 101 4.2.11 Temperature Measurements 101 4.2.12 Ex-Vivo Histology and Scanning Electron Microscopy (SEM) 102 4.2.13 Cytotoxic Assay 103
8 4.2.14 In Vivo Thermal Damage 104 4.2.15 Statistical Analysis 106 4.3 Results 106 4.3.1 Adhesive Attenuation 106 4.3.2 Tensiometer Measurements 109 4.3.3 13C-NMR 111 4.3.4 Thermogravimetric Analysis and Contact Angle 113 4.3.5 Differential Scanning Calorimeter 114 4.3.6 Atomic Force Microscopy 115 4.3.7 Young’s Modulus 116 4.3.8 Temperature Measurements 117 4.3.9 Ex-Vivo Histology and SEM 118 4.3.10 Cytotoxicity Assay 122 4.3.11 In Vivo Thermal Damage 125 4.4 Discussion 128
Chapter 5. Conclusions 134
References 138 Publications, Patents and Presentations Arising from This Thesis 167
Appendix 168
9 Abbreviations
AFM Atomic Force Microscopy
BSA Bovine Serum Albumin
CB Carbon black
CMAP Compound muscle action potential
CW Continuous wave
DSC Differential Scanning Calorimetry
H&E Hematoxylin and Eosin
IBC Isobutyl cyanoacrylate
IG Indocyanine green
KS Kologorov-Smirnov
LTS Laser tissue soldering
LTW Laser tissue welding
MPC Methoxypropyl cyanoacrylates
PBS Phosphate buffer solution
PEG Polyethylene glycol
PLGA Poly(lactic-co-glycolic acid)
SEM Scanning electron microscopy
TGA Thermogravimetric Analysis
10 List of Figures
Chapter 1
Figure 1. Different approaches for wound closure and sealing in existence today 16
Figure 2. Sutures, staples and clips in today clinical practice 17
Figure 3. Diagram of the two-component fibrin glue and its clot formation 18
Figure 4. Histological characterization of thermally damaged bladder 27
Figure 5. Albumin inside carotid artery 28
Figure 6. Nerve repair with solid albumin solder 30
Figure 7. Distal longitudinal neurorraphy immediately after laser-solder repair 30
Figure 8. Transverse section of solder strips bonded to the serosa layer 33
Figure 9. A solder strip is laser bonded to the intimal layer of a rat aorta 34
Figure 10. Two-layer strip after laser welding with the serosa layer 37
Chapter 2
Figure 1. Diagram of reanastomosing the vasa with sutures 50
Figure 2. Intraluminal sutures are applied to align the vasa 52
Figure 3. Schematic drawing of the stent preparation 56
Figure 4. Diagram of the surgical procedure 60
Figure 5. The weight loss of chitosan films given as a function of temperature 62
Figure 6. SEM view of a 3 mm long chitosan stent 63
Figure 7. Strain-stress diagram of a chitosan strip 65
Figure 8. A vas stenosis in the proximity of the anastomotic site 66
Figure 9. Longitudinal section of a sperm granuloma at the anastomotic site 68
11 Figure 10. Longitudinal section of the proximal site of a sutured vas deferens 69
Chapter 3
Figure 1. Molecular structure of genipin 74
Figure 2. Proposed mechanism of the blue pigment formation 75
Figure 3. Schematic of laser tissue-soldering 79
Figure 4. Histogram of acute tensile strength of laser repaired tissue with
BSA solder and BSA-genipin solder 83
Figure 5. Percent inhibition growth of fibroblasts in media extracted from
BSA solder 86
Figure 6. Percent inhibition growth of fibroblasts in media extracted from
BSA-genipin solder 87
Figure 7. Cell growth inhibition percentages of fibroblasts of ethanol controls and genipin solutions in medium 88
Chapter 4
Figure 1. Chemical structure of chitosan that is derived from deacetylated chitin 92
Figure 2. Schematic top-view of laser tissue repairing 96
Figure 3. Photograph showing a strip of chitosan adhesive bonded to the serosa layer of intestine 101
Figure 4. The adhesive is placed around the sciatic nerve like a collar 106
Figure 5. Typical attenuation spectrum of chitosan adhesive in the visible-NIR region 107
Figure 6. Histograms of acute shear stress of chitosan adhesive bonded to tissue 110
Figure 7. 13C-NMR spectra of chitosan shells and adhesive 112
12 Figure 8. The weight loss of chitosan adhesive is given as a function of temperature 113
Figure 9. Contact angles of a water drop on chitosan adhesive 114
Figure 10. The DSC scans of the chitosan adhesive 114
Figure 11. Topography of the chitosan adhesive obtained with AFM 115
Figure 12. Graph illustrating the stress and strain relationship for wet chitosan 117
Figure 13. Graph illustrating the temperature of the thermocouple, under the chitosan adhesive at tissue interface, as a function of time 118
Figure 14. Longitudinal sections of the adhesive laser-bonded to the intestine serosa 119
Figure 15. SEM cross sections of a chitosan strip bonded to sheep intestine 120
Figure 16. Histogram illustrating the number of cells recovered as a function of media treatment 123
Figure 17. Histogram of viable cells recovered from the extracted media 123
Figure 18. Flowcytometer plots of cell fluorescence and forward/side scatter 124
Figure 19. Cross sections of an operated nerve with chitosan adhesive four days post-operatively. The histology of a non operated nerve is also provided 125
Figure 20. Cross section of the sciatic nerve and chitosan adhesive surrounded by neutrophils four days post-operatively 128
Appendix
Figure 1. Laser beam profile 173
13 List of Tables
Chapter 2
Table 1. Parameters of the elastic test 58
Table 2. The laser soldering parameters 61
Chapter 3
Table 1. The laser parameters and solder characteristics 83
Table 2. Attenuation data of the solder at a wavelength of 808 and 580 nm 85
Chapter 4
Table 1. The laser parameters and adhesive characteristics (first part) 97
Table 2. The laser parameters and adhesive characteristics (second part) 98
Table 3. Attenuation data of the adhesive at a wavelength of 808 and 608nm 109
14 Chapter I
Background
1.1 Introduction
Suture and staple are still considered the gold standard for surgical closure; these devices are similar in their function to devices employed thousands of years ago for wound closure. Use of sutures is essential and undisputedly the optimal solution for closure of a wide variety of simple wounds. However, in many instances the suture is either unable to effect repair or the repair that occurs interferes with the functional rehabilitation of the site. Examples of this include closure of lung tissue following resection (air leakage), nerve repair (scarring and lack of conduction), abdominal repairs
(leading to adhesions) and traumatic injuries (agglomeration of many small lesions) [1-
4]. In these cases, various surgical “glues” have often been used. These include fibrin glues, albumin (protein) solders and synthetic polymer glues like cyanoacrylates or degradable polyethylene glycol (PEG). All of these approaches have problems ranging from lack of strength, associated tissue damage from crosslinking systems (UV light) and thermal damage from laser assisted protein solders. A flowchart has been constructed to illustrate the different approaches to wound closure and sealing in existence today (Figure 1).
The conventional approach to closure of wounds involves approximation of tissue segments with sutures, clips, or staples (figure 2). A common side effect of these techniques is leakage of fluids (i.e. blood) or gases (i.e. air from the lungs). Surgical
“adhesives” like fibrin glue have been used more recently as an adjunct method to achieve sealing of such closures [5,6].
15 TISSUE REPAIR
SUTURES
FOREIN BODY REACTION ADHESIVE SEALANT NO SEALING PROPERTIES Strenght>12 KPa burst pressure >5 KPa
THERMAL DAMAGE VIRAL INFECTION LASER + TISSUE ALBUMIN SOLDERS FIBRIN GLUES ALBUMIN SOLDERS
LOW TENSILE STRENGTH LOW BURST PRESSURE NON UNIFORM APPLICATION
DNA DAMAGE
CHITOSAN GEL + UV LIGHT PEG GEL + VISIBLE LIGHT CHITOSAN GEL + UV LIGHT
Figure 1. Different approaches for wound closure and sealing in existence today.
16
Figure 2. Sutures, staples and clips in today clinical practice.
1.2 Fibrin Sealant
Fibrin sealants essentially mimic the final stages of the coagulation cascade in the human body (figure 3). They usually comprise two components, fibrinogen and thrombin, which are applied to the tissue site. In the presence of calcium ions, thrombin cleaves the fibrinogen chains to form fibrin monomers, which can then polymerize to produce a physiological fibrin clot, independent of the patient’s coagulation pathways.
Fibrinogen provides the substrate for thrombin to cleave into fibrin monomers and it also activates factors XIII, which stabilizes the fibrin clot by polymerizing the fibrin monomers. The fibrin clot induced by the sealant degrades by physiologic fibrinolysis.
Beginning a few hours after tissue adhesion, there is a proliferation of fibroblasts,
17 infiltration of granulocytes, formation of granulation tissue and subsequent proliferation of collagen fibers.
FactorXIII Ca 2+ + Thrombin
Factor 3 (4-10 IU/ml) XIIIa
Ca 2+
Soluble Fibrin Fibrin Fibrin Fibrinogen Monomer Clot
(40-110 mg/ml)
(<103 kIU/ml) Aprotinin
Figure 3. Diagram of the two-component fibrin glue and its clot formation.
After 2 weeks, the clot organization is complete with collagen rich granulation tissue and a markedly reduced cellular infiltration. The tensile strength of the clot is mainly a function of the fibrinogen concentration of the glue [7,8]. Thrombin concentration may also influence the mechanical strength. The speed and completeness of clotting is determined by the thrombin concentration. Increasing the thrombin concentration in vivo from 20 to 1000 IU/ml was reported to speed up the time of hemostasis from 30 to
18 a few seconds clotting [9]. Adding aprotinin in the fibrin glue, which inhibits human trypsin and plasmin, may also increase the stability of the clot against lysis [10]. Fibrin glues are used widely for many surgical procedures, although they have been few randomized and controlled clinical trials. The key indications for these glues are hemostasis and tissue sealing. Clearly, they can provide these two functions together in many procedures such as intracranial and spinal surgery, duraplasty, tumour resection, aneurysm repair and nerve anastomosis [11-13]. In cardiovascular surgery, fibrin sealants have been used successfully during bypass surgery, vascular graft, valve replacements and repair of septal defects [14,15]. Optimal application requires a dry operative field, which is often difficult to achieve and thus fibrin sealants are usually used prophylactically while the suture line is dry to prevent possible haemorrhage.
Fibrin sealants have also been applied to endoscopic procedures for the treatment of peptic ulcers [16].
Complications in the clinical use of fibrin sealant can arise because of potential viral infection, patient sensitivity to bovine thrombin and rudimental delivery devices, which affect the final outcome [17,18]. Sierra et al investigated the effect of mixing the two components of fibrin sealants upon their mechanical properties [17]. The degree of the component mixing was assessed by quantitative stereological techniques. Fibrin sealant
(fibrinogen concentration ~ 70 mg/ml) was applied with 4 different clinical applicators on split-thickness dermal grafts, which were compressed by a 100g weight for 10 minutes (surface area ~6.2 mm2). The tensile strength increased when the fibrin components were better mixed (~28 kPa). However, no significant difference was measure between the different applicator types on the closure strength of in vivo dermal incision (~0.268 ±0.19 N/mm), 3 h after the sealant application. In a rabbit spleen
19 incision model, a more thoroughly mixed sealant corresponded with a decrease in time
(27 s vs ³50s) to obtain complete hemostasis as well as less sealant used (60 ml vs 90 ml). It appeared that the increased strength of the better mixed sealant had no impact in the acute repair strength.
The relative high tensile strength reported by Sierra et al in vitro may be due to particularly dry conditions of the experimental set up. These results contrast sharply with the well-known weakness of fibrin clots, bonded to tissue, when they are stressed by pulling forces. When wounds are closed in vivo by fibrin sealant, a few sutures are usually implemented to reinforce the repair and avoid mend failure. Petratos et al, for example, closed full thickness skin wounds in pigs with fibrin sealant after approximating the tissue edges with a single 5-0 nylon suture [19]. Likewise, Menovski et al approximated transacted edges of rat sciatic nerves with two 10-0 absorbable sutures and subsequently glued the nerves using fibrin sealant [20].
To date, investigations into basic mechanisms of tissue sealant attachment and failure at the tissue interface have not been reported. Sierra et al investigated the macroscopic adhesive properties and bulk characteristics of fibrin sealants [21]. Two test methods were used: uniaxial monotonic tensile testing of the bulk material and blister testing.
The latter test was performed on fresh porcine skin graft as the adherend, which had a hole (25 mmx2 mm) sealed with 2 cc of fibrin (curing time=5 min). Two fibrin concentrations, 30 mg/ml (LFC) and 70 mg/ml (HFC) of fibrin sealant were investigated. The HFC (tensile strength = 28.0 ±11.6 kPa, E = 27.5 ±19.3 kPa) showed higher overall averaged values than LFC (tensile strength = 9.9 ±3.4 kPa, E = 14.6
±4.1kPa) for ultimate tensile strength at 2.5, 25, 250 mm/min strain rates. Visual
20 observation of both sealants showed that at lower strain rates, failure occurred predominantly by strands of the specimen peeling away from the outermost region extending towards the inside of the gage crossection during the course of strain. As the strain rate increased to 250 mm/min a “cup and cone” fracture surface resulted.
Syneresis (expression of the fluid phase of the gel) was observed in all cases during the tests. HFC demonstrated overall greater averaged burst strength (13.37 ± 9.28 kPa) than
LFC (3.17 ± 2.30 kPa) at all strain rates. Failure of fibrin gels likely occurred by percolation of the pressurized water, which displaced the entrapped liquid phase of the gel in regions of relatively low moduli and strength, leading to fracture of the matrix.
The so-called adhesives in use today tend to act mainly as sealants and lack sufficient strength to be used without the support of sutures or staples [22]. Some tissue adhesives like cyanoacrylate (Super Glue) can yield a strong bond, however they are difficult to use and have associated toxicity [23].
1.3 Cyanoacrylate Glues
Cyanoacrylates are esters (alkyl side chains) of cyanoacrylic acid [(C5H5NO2)
CH2=C(CN)COOCH3]. The oxygen double bond with the carbon plays a key role in the polymerization (hardening) of this glue. The cyanoacrylate monomers polymerize through contact with water or a weak base, such as cell membranes and tissue.
Hydroxylation occurs through the exclusion of oxygen from the substances being bonded. The alkyl side chain can be modified to produce cyanoacrylates with different bonding properties. As an ester chain increase from one carbon to higher numbers (10
C, for example) the compound becomes more biocompatible. Short chain esters (<4 C) are toxic either directly or through breakdown products. Early derivates of cyanoacrylates had short side chains and degraded rapidly into cyanoacetate and
21 formaldehyde. The degradation products accumulated in tissues and produced significant histotoxicity characterized by both acute and chronic inflammation. The longer alkyl chains of currently available glues (e.g. N-butyl-2 cyanoacrylate) slow degradation significantly, limiting accumulation of products to amounts that can be effectively eliminated by tissues. Histotoxicity, however, depends on the vascularity of tissues, being grater in well vascularized soft tissues [24].
Cyanoacrylate glues have been successfully applied in ophthalmology to repair corneal perforations. A small skin patch is usually glued to the cornea, which has suffered perforation (typically less than 2 mm). This procedure gives improved visual outcomes without re-epitheliazation into the zone of damaged and naked stroma [24].
Cyanoacrylate glues can also prevent the development of the critical setting for collagenase production that leads to stromal melting, and has a significant bacteriostatic activity against gram-positive organisms. The major concern is toxicity of cyanoacrylates through direct contact with the corneal endothelium and lens. Fibrin glues are generally less toxic, however, they are less strong, less available and more expensive. Linden et al modified the cyanoacrylate glue to repair soft and hard tissue.
Absorbable polymeric oxylates were incorporated in methoxypropyl cyanoacrylates
(MPC). The new adhesive was introduced into a goat skin incision for 3 h and the specimen tested for tensile strength. Isobutyl cyanoacrylate (IBC) and MPC were also used in the same skin model for comparison. The modified cyanoacrylate adhesive had the greatest tensile strength (296 ± 48 KPa) as the other glues reached only 220 ± 34
KPa. On the other hand IBC was the strongest adhesive (10.3 ± 2.0 MPa) when the cortical portion of femoral bovine bones was glued. The other glues performed poorly with strength less than 2.068 MPa. The stiffer IBC appeared to be better suited for
22 adhesion to hard tissue as its modulus matched with the bone substrate modulus [25].
Marcovich et al compared cyanoacrylate glues and fibrin in vivo [26]. Cystostomy was performed in a pig model and the closure was achieved by using 2-octyl cyanoacrylate adhesive (OCA) or fibrin glue (FG, 75%-115 mg/ml fibrinogen) with stay sutures. The closures were tested on postoperative day 2 at less than 200 mm for leakage and 4 out of
6 fibrin glued pigs leaked, while no OCA pig leaked. It was concluded that OCA provided enough strength to hold together a large bladder wound, while fibrin glue did not ensure an adequate closure.
Although the use of sutures is generally reliable, there are a number of problems that limit their use, such as scar formation and foreign body reaction [27,28,4]. Further, the suturing technique is very difficult to perform during endoscopic surgery; the limitation in free movements and vision imposed on surgeons by the endoscopy set up are a serious impairment during suturing. These side effects have prompted a search for alternative methods and devices for wound closure such as photoactivated adhesives and sealants.
1.4 Laser-Activated Glues
Lasers are increasingly used for tissue repair in experimental and clinical procedures
[29,30]. In laser tissue welding (LTW) of tubular structures (anastomosis), proteins within the target tissue are coagulated to form a bond joining the two edges [31]. This technique offers advantages over conventional suturing in that it involves less suture or needle trauma and decreases foreign body reaction [32]. It is also simpler and faster to perform in most instances. Example are LTW of blood vessels, nerves and intestine, which offer the potential for faster operations and improvements in the tissue healing
23 when compared with suturing. However, direct laser welding is often unsatisfactory because it may inflict intolerable thermal damage to cells and has a high failure rate
[33]. Some researchers have used temporary or permanent stay sutures to improve the success rate of laser anastomoses, as the welds tend to be weakest in the first few days postoperatively [34, 35]. This does not eliminate the sutures and their consequent problems and also complicates the operative procedure.
A better alternative is to add extra protein in the form of “solder” to the tissue to supplement and enhance the bond strength. During the laser procedure, the solder is applied across the severed tissue ends and the laser heats and denatures the proteins at
65-70 °C. As a result, the solder bonds to tissue and repairs it. Protein solders are proposed to intertwine and link sterically with collagen fibers following laser irradiation
[36].
Protein solutions such as albumin or fibrin have been used successfully in conjunction with a CO2 laser to heat the fluid solder [37]. The first albumin solder was successfully developed and used to seal tissue against fluid leakage in urethral reconstruction [38].
To date, infrared lasers, such as Tm, Ho: YAG, and CO2 at wavelengths of 2.0 µm and
10.6 µm, respectively, have been primarily used for LTW [32, 39, 40] and for laser tissue soldering (LTS) [30, 37, 41]. The choice of such lasers is based on absorption of the laser energy by water, thereby heating proteins within the target tissue or the applied solder, resulting in coagulation at 65-70°C. The resulting denaturation of protein is not specific to the target tissue or solder, and all irradiated tissues are heated. This is a problem when repairing blood vessels, peripheral nerves and tissue in general. Some workers have successfully used dyes at the weld site [42] to avoid most of the
24 ‘‘collateral’’ thermal damage. The dye is usually applied directly to the tissue, or mixed with the applied protein solder [43, 44]. The laser wavelength is chosen to be strongly absorbed by the dye, but poorly absorbed by water and bodily tissues. Usually an 810 nm diode laser is used to denature the dye-protein solder applied to the tissue, with energy transferal due to absorption in the dye, indocyanine green (IG) or carbon black
(CB) [43-46]. The diode light is delivered by optical fiber in a handheld fiber chuck, so it easily may be incorporated into the surgical procedure.
Protein laser-activated solders are absorbable within days and, therefore, are less likely to cause scar formation. Also, laser-soldering techniques are usually faster and easier than conventional closure procedures [47, 48]. Kirsch et al performed hypospadias repair on 54 boys (mean age 15 months) using 50% human albumin solder doped with
2.5 mg/ml indocyanine green dye and an 808 nm diode laser with an irradiance of
~16W/cm2 [48]. The pulse duration was 0.5 sec with an interval of 0.1 sec between pulses. Eighty-four children were operated with the standard suturing technique. In the laser group, a few sutures were used for tissue alignment only. The mean operative time was five times less for laser tissue soldering hypospadias repair compared to controls.
After12 months, the complication rate was 4.7% in the laser group and 10.7% in controls with fistula in 2 of 54 cases, and fistula and meatal stenosis in 7 and 2 of 84, respectively. These results indicated that laser tissue soldering for hypospadias repair might be performed in almost sutureless fashion and more rapidly than conventional suturing.
Of particular interest was the study conducted by Wadia et al, who evaluated laser soldering by using liquid albumin for welding liver injuries in pigs [49]. Eight
25 laceration (6 x 2 cm) and eight non-anatomic resection injuries (raw surface, 6 x 2 cm) were treated with an 805 nm laser (Irradiance ~51 W/cm2). The pulse duration was 0.1 sec with an interval of 0.1 sec between pulses. Liquid albumin solder (50% w/w) and indocyanine green (12 mM) was used to seal and repair the liver surface, reinforced with a free autologous omental scaffold. All 16 laser mediated liver repairs had minimal blood loss as compared with the suture controls. No dehiscence, hemorrhage, or bile leakage was seen in any of the laser repairs after 3 hours.
Despite important advantages, most of the current albumin solders are still not as strong as sutures and they induce collateral thermal damage. Furthermore, they are often in a fluid state and require a dry operating field along with extreme care during their application to tissue in order to minimize thickness variability. These drawbacks may result in an impairment of the reliability and reproducibility of the laser repair.
Albumin solders, in their fluid state, have inherent problems to solve. The required surface area and thickness of the applied fluid glue cannot be produced repeatedly; the reproducibility is very important because the laser energy necessary to coagulate the solder depends on the solder surface area and thickness. The fluid solder is difficult to apply to tissue (although it is easier and sometimes faster than suturing) and it is not possible to control the uniformity of the applied layer, which may vary considerably
(figure 4) [50]. A significant thermal damage may derive from the variable solder thickness because the tissue under a thin layer can absorb more radiation or heat than the tissue under a thicker layer. When a temperature control system with feed-back is employed to keep the solder temperature at 65-70 0C during operation, the bonding strength between solder and tissue may be weaker where the solder is thicker. In fact,
26 the system measures and controls only the superficial temperature of the solder (~10 mm), which is usually higher than the temperature at the tissue-solder interface.
S
Figure 4. Histological characterization of thermally damaged bladder seen under 34 magnification with Masson’s trichome stain. Albumin solder (S) is visible on tissue with variable thickness. Damage is more evident under the thinner solder.
Furthermore, the fluid solder may penetrate at the anastomotic site of a vessel, for example, and pose a risk of thrombosis for the patient (figure 5) [51]. To overcome these problems, a solid solder with fixed thickness was developed by increasing the albumin concentration to ³56% (w/w). The solid solder had higher cohesive strength than its fluid counterpart and required 1.2 ± 0.5 J/g of heat to denature, while the fluid solder needed 3.2 ± 0.2 J/g [52]. The increased heat measured during protein denaturation might be ascribed to the higher water content of the liquid solder and it
27 might indicate that the solid solder required less heat to bond to tissue than the liquid solder.
C A
Figure 5. Albumin is found at both the luminal and adventitial surfaces of a dog carotid artery (A). At 2 hours after repair, arteries were fixed in formalin, embedded in paraffin, and sectioned. Trichrome-stained sections reveal the presence of albumin (layers of albumin denoted by arrows) inside the artery. Coagulated tissue is visible beneath the solder (C).
McNally et al studied the effect of temperature on the optical and thermal properties of fluid and solid albumin solders, doped with IG (0.25%). These properties were studied as a function of wavelength, between 25 °C and 100 °C. The reduced scattering coefficient of fluid and solid solders at 805 nm increased rapidly with temperature from being a highly nonscattering medium at room temperature to a highly scattering medium at temperatures close to 70 °C (151 cm-1, 191 cm-1 for liquid and solid solders,
28 respectively). The thermal conductivity, thermal diffusivity, and heat capacity increased by up to 30%, 15%, and 10%, respectively, for both solid and liquid solders. The
-1 absorption coefficient (ma) of the fluid and solid solders was ~300 cm , corresponding to a penetration depth (1/ma) of ~33 mm [53].
Solid solder strips were first applied to repair tibial nerves in Wistar rats [54, 55]. A total of 18 Wistar rats had left tibial nerve repaired by either the laser-solder technique or a more conventional microsuture technique. Four protein strips (length ~3mm, thickness ~0.15 mm) were doped with IG (0.2%) and applied across the nerve anastomosis (figures 6, 7). A diode laser bonded them to tissue using a power of 90 mW, irradiance of ~200 W/cm2 and a radiation dose of 16 J/mg (energy per mg of solder). Three months postoperatively electrophysiology showed that the average compound muscle action potential (CMAP) of the laser repair group was not significantly different from the CMAP of the sutured nerves (~2.5 mV). Light microscopy confirmed regeneration of myelinated axons in both groups of animals [55].
The solid solder is rigid and partially flexible if maintained in humid environment, otherwise becomes brittle after a few minutes upon dehydration. The lack of flexibility and brittleness limits therefore the solid solder applications. A significant example of the limited capability of solid solders is given by vessel anastomosis.
29 Figure 6. The operation site immediately after the laser-solder repair. Four protein bands hold together the nerve stumps after the laser irradiation (×30).
S
A P
Figure 7. Distal longitudinal neurorraphy immediately after laser-solder repair. The laser-induced heat has denatured the BSA band (S), which is bonded to the perineurium (P). Nevertheless, the axons (A) underneath appear undamaged (×100 Masson’s Trichrome).
Maitz et al used tubes of solid solder (length ~2 mm, diameter ~1.3 mm) doped with IG
(0.27%), to successfully anastomose the abdominal aorta of rats [56]. The albumin tubes
30 were heated for 1 s in hot water to make them flexible and overcome their brittleness.
The denatured solder decreased its solubility, which is desirable, but unfortunately weaken also its tensile strength. Nevertheless, the surgical technique compensated for the diminished tensile strength of the repaired aorta and the burst pressure was more than satisfactory after laser soldering. In detail, 90 rats were divided into two groups randomly. In the control group, the anastomoses were performed by using conventional microsuturing technique, whereas in the experimental group, the anastomoses were performed by using a diode laser (l=808 nm) with an irradiance of ~286 W/cm2. The mean clamp time of the anastomoses performed with conventional suturing was 20.6 minutes compared with 7.2 minutes for the laser-activated welded anastomoses (p <
0.001). Histologic evaluations revealed proliferation of myofibroblasts, some fibrotic reaction on the adventitia and a near complete resorption of the solder after 6 weeks.
Xie et al performed 21 anastomoses in the carotid arteries of seven pigs with an 800-nm laser (irradiance ~254 W/cm2) and an albumin stent plus solder doped with IG (0.1 mM)
[57]. There were five artery-to-artery and 16 elastin heterograft to native artery anastomoses. The albumin stent played an important role in strength of the anastomosis and out of 21 anastomoses, 20 were patent at 3 h. The laser energy used was 212 Joules in artery-to-artery anastomosis and 273 Joules in elastin heterograft to native artery.
Histology shows coagulative necrosis of the adventitia and no change in elastin heterografts.
Solid and liquid albumin solders are soluble in physiological fluids before laser irradiation and therefore a dry operating field is indispensable to avoid solder wash out or mechanical alterations that weakens the repair strength. The solubility obstacle has
31 been partially overcome by increasing the protein concentration of solders. In a previous study, gauged protein strips of solder (thickness ~0.17 mm, weight ~1.4 mg) were immersed into 0.5 ml saline solution for fixed intervals of time [58]. The strips contained four Bovine Serum Albumin (BSA) concentrations: 56%, 66%, 70%, and
75% (by weight). A Bradford protein assay measured the BSA solubility of the solders in saline solution. The strips with 56% BSA content dissolved > 37% of their weight in
45 s, the 66% solder dissolved 18% in 110 s, and the 70% strips dissolved 10% in 299 s.
The 75% solder dissolved 6% in 50 min. Sections of dog intestine were also repaired with 66% and 75% solid strips of solder (~4.3 x 0.6 x 0.15 mm, weight ~0.4 mg) using a CW diode laser (l=808nm) at a power level and radiation dose of 0.14 W and 14
J/mg, respectively [52]. The 66% solder formed stronger repairs than the 75% solder
(0.23 Vs 0.06 N) because it was more soluble. The lower concentrated strips in fact liquefied a few microns at the tissue interface allowing the solder to better adhere to the tissue and thus enhance the bond strength upon irradiation (figure 8a). The less soluble solder adhered to the tissue only partially resulting in weaker repairs (figure 8b).
It should be noted that soluble solders entangled with tissue might solidify and therefore bond to tissue at temperatures below the BSA denaturation temperature, as long as they are insoluble in water. Solders entrapped in tissue become insoluble when heated around or above the denaturation temperature (65-70 0C) causing the weld to be stable and strong, as previously reported [59]. The non-denatured solder, entrapped and heated at lower temperatures in the tissue, can liquefy again in moist environments and weaken the tissue bond [59, 60].
32 A
B
Figure 8. Transverse section of solder strips bonded to the serosa layer. The soluble strip (B) is well adherent to the serosa layer while the poorly soluble strip (A) badly matched with the tissue edge. (Masson trichrome stain; x100).
The fluid state of the solder (prior to irradiation) in conjunction with the irregular geometry of the tissue surface seemed to play a prominent role in enhancing the bond strength, because of the sterical locking between albumin and tissue collagen [36]
(figure 9). The nature of the molecular bonds between tissue and solder remains unclear; non-covalent forces between BSA and collagen are possible such as van der Waals,
33 S
Figure 9. A solder strip (S) is laser bonded to the intimal layer of a rat aorta. The soluble solder is penetrating the tissue irregularities enhancing the bonding strength. (Masson trichrome; x100). electrostatic interactions and hydrogen bonds [31]. Nevertheless, their specific contribution in strengthening the repair is still uncertain. It seems less likely that covalent crosslinks can occur, as they should not be affected by water unlike the strength of soldered tissue.
The solid solder behaves substantially like a dense fluid solder at the tissue interface.
Both these solders are fluid on tissue with high BSA concentration (£ 56% w/w) and have similar denaturation temperatures (70 ±1 0C) [52]. Hence, the strength of the laser repair should be comparable at tissue interface using solid soluble solders and high concentrated fluid solders. Other investigators found that the albumin concentration should be increased to enhance the repair strength of fluid solders, which contained only
34 albumin and water [61,62]. The reduction in strength of repaired tissue due to less soluble solid solders sets an intrinsic limit on the previous assumption: the concentration of protein cannot be increased above ~65% otherwise the weld strength drops [52]. BSA powder and dehydrated solid solders also showed not to have a denaturation peak temperature. These results raised the question if albumin solders would still denature and bond at very high protein concentration (~95% w/w). Solders containing water and albumin appear therefore to have a limited bonding strength at tissue interface.
As mentioned before, one of the major drawbacks of laser welding and soldering is the thermal damage inflicted on tissue, which can induce necrosis and apoptosis of cells
[34]. Computerised feedback systems have been coupled to lasers to control the solder temperature and minimize thermal injury [63]. Shenfeld et al used a silver halide fiber for IR radiometric measurement of tissue temperature during laser welding of urinary bladder in rats. The investigators were able to close successfully puncture wounds by heating tissue with a CO2 laser for 5 sec at constant temperatures with a deviation of ±
2.5 0C [64]. Shumalinsky et al carried out successfully laparoscopic pyeloplasty in a porcine model using a C02 laser, albumin solder (47% w/w) and a temperature control system, similar to the one mentioned above [65]. Laparoscopic laser soldering provided watertight bonding at 65 ± 3 0C and was found to be faster than suturing.
More recently, an analogous IR radiometric system was also used to laser solder full- thick incisions in the dorsal area of New Zealand rabbits [66]. The laser irradiance was
7 W/cm2 on liquid albumin solder (47% w/w) and the temperature was set between 50 to 90 0C. The optimal temperature was found to be 65 ± 3.6 0C and resulted in repairs
35 with tensile strength higher than sutured skin (1.81 vs 1.08 MPa) 4 weeks post operatively. Histoacryl and Dermabond were also used to close wounds and to compare the immediate tensile strength of the laser technique. Laser and chemical methodologies achieved a tensile strength of ~0.05 MPa.
Although IR radiometric systems with computerized feedback can stabilize the solder external temperature during welding, they may not effectively control the temperature of the solder-tissue interface, which is required to limit thermal damage. Fluid solders vary in thickness unpredictably over the tissue and therefore prevent any reliable correlation between external and tissue interface temperature. In an attempt to overcome this problem, a solid protein solder with two layers has been developed with a fixed thickness to minimize the difference in temperature across the solder depth
[46,67]. In the two-layer solder, the dyed layer in contact to tissue absorbs the laser and heat diffuses almost symmetrically to tissue interface and to the external side of the solder, minimizing its temperature gradient (figure 10). The feedback system can therefore stabilize at once the external solder temperature and the tissue interface temperature. McNally et al used a similar two-layer solder design and was able to achieve a gradient of 3 0C across 0.15 mm solder thickness [68]. The first versions of the two-layer solder lacked flexibility as both strata were made of albumin films [46].
Subsequently, a resorbable poly(lactic-co-glycolic acid) (PLGA) membrane doped with fluid albumin solder made a more flexible and versatile two-layer adhesive for surgical applications [68,69].
Despite feedback systems and the more complex adhesive designs developed to control the temperature, protein denaturation and heat shock still occur in cells during laser
36 repair. The effort of researchers to limit the thermal damage is often obstructed by technical difficulties related to the sophisticated systems, and by the more fundamental reason that tissue can be irreversibly damaged at 65-70 °C.
Figure 10. Transverse section of a two-layer strip after laser welding with the serosa layer (bottom of picture). The lower half of the solder contains scattered granules of carbon black (black dots), while the top half does not (whitish appearance). The air bubbles (generated by the laser heat) are localized mostly in the middle of the strip (black middle plane), and they are scattered on both layers of the solder. The strip substantially preserves its integrity. Tissue manipulation during slide preparation detached the strip from the serosa layer (H&E, ×100).
1.5 PEG Glues
PEG gels activated using visible light have been recently approved for lung sealing by
FDA (Focalsealâ). The primary active component of the sealant is a polyethylene glycol group [(C2H4O)nH2O] linked to trimethylene carbonate and acrylate moieties. A liquid primer is usually applied to provide a template to which the sealant bonds. The viscous sealant is then applied and activated by photopolymerization with visible blue- green light (450-550 nm). This glue is able to produce good sealing of the lung, however it has the disadvantages of being applied as a two-step process and it degrades very slowly, persisting for up to 20 months following implantation (personal 37 communication from FDA panel meeting). This latter point may be of concern for the longer term biocompatibility, particularly the carcinogenicity of the material.
Sweeney et al performed small bowel anastomosis on New Zealand rabbits (n=12) with
4 interrupted sutures and Focalsealâ, which was applied over the suture line to seal the anastomosis [70]. Eight sutures were applied to complete the bowel anastomosis in the control group (n=12). Bursting pressure and other wound healing parameters were assessed after three days, one week and three weeks. There was no evidence of anastomotic dehiscence in 23 of the 24 animals. Furthermore, there was no significant difference in adhesion formation, stenosis or bursting pressure (P ~13 KPa after three days) between the groups. Collagen and blood vessel formation were significantly increased in the sealant group.
1.6 Chitosan Glues
More recently, photochemically activated chitosan gels have been shown to induce no thermal damage to cells and to produce better sealing for air leaks than fibrin [22]. For example, Ono et al introduced lactose and azide moieties in the chitosan structure (Az-
CH-LA), which was more soluble in water than chitosan at neutral pH. Az-CH-LA aqueous solution was gelatinized by UV irradiation presumably because azide groups (-
N3) released N2 upon UV irradiation and were converted into highly reactive nitrene groups. Nitrene groups were supposed to react quickly with each other or with amino groups of the chitosan to generate azo groups (-N=N-) thus causing gelation.
Application of UV light to Az-CH-LA glue produced an insoluble hydrogel in 60 s.
This hydrogel adhered two peaces of sliced ham with each other (4.2 KPa). The binding
38 strength of the chitosan hydrogel prepared from 30-50 mg/ml of Az-CH-LA was similar to that of fibrin glue (3.9 KPa). Compared to the fibrin glue, the Az-CH-LA glue sealed more effectively air leakage from pinholes in isolated small intestine (8.1 vs. 6.8 KPa) and aorta (26.7 vs 10.7 KPa) and from incisions on isolated trachea (10.1 vs 5.6 KPa).
Neither Az-CH-LA nor Az-CH-LA hydrogel showed any cytotoxicity in cell culture tests with human skin fibroblasts, coronary endothelial cells and smooth muscle cells. In vivo studies showed that 6 mice survived over 1 month after implantation of chitosan crosslinked gel (200ml) and other 3 mice survived intraperitoneal administration of up to 1 ml of 30 mg/ml of Az-CH-LA solution. The chitosan hydrogel developed by Ono et al was also tested in full thickness incisions made on the back of mice. About 100 ml of 20 mg/ml of Az-CH-LA aqueous solution was put in each mice incision and irradiated by UV light for 90 s. The wound areas was measured for 12 days postoperatively and the chitosan wounds contracted significantly after 2 days while control incisions were almost double in size respect to the chitosan wounds. These were nearly epithelialized and almost all necrotic tissue was replaced by new granulation tissue in contrast with the control repairs. After 10 days all wounds healed similarly
[71]. A drawback of this technology is the use of ultraviolet (UV) radiation to crosslink the gel to the tissue, which can pose a risk for DNA mutations in cells exposed to UV light. Nakayama et al prepared gelatines as tissue adhesives, which were activated not only with UV radiation but also with visible light [72]. The gelatines from bovine bones were partially derivatized with UV and visible photoreactive groups (benzophenone, fluorescein sodium salts, eosin Y and rose bengal respectively). These photoreactive gelatins were combined with Poly(ethylene glycol) Diacrylate and a saline solution with or without ascorbic acid (reducing agent) to obtain photocurable glues. The adhesive glues were coated on collagen films and irradiated with a mercury lamp (t=60,
39 P=5mW/cm2) or an excimer laser (l=248 nm, tp=15 ns, 70 pulses/s). The visible activated glues were irradiated by a Xenon lamp (400< l<500 nm) for 60 s. The benzophenone glues achieved adhesion strength of 12 KPa. Fibrin glue and cyanoacrylate adhesive were used for comparison and they achieved a strength of 1.5 and 68 KPa, respectively. It is likely the dry conditions, under which the tests occurred, increased the tensile strength of fibrin and gelatine adhesives.
In rats whose livers were injured mechanically, the bleeding spots were coated with the photocurable glues and irradiated through an optical fiber during endoscopic surgery.
The glue was converted into a gel, which firmly attached to tissue presumably by interpenetration, and hemostasis was completed. Abdominal and thoracic aortas of dogs were also incised with a knife and repaired with UV and visible glues. These treatments resulted in minimal bleeding under pulsatile conditions. Histological examination showed that the light-activated glues on the rat livers were gradually degraded in vivo with time with infiltration of inflammatory cells and connective tissue without necrotic signs in surrounding tissue. Ahn et al synthesized poly acrylic acid/chitosan (PAA/CH) complexes by template polymerization of acrylic acid in the presence of chitosan [73].
One percent of 2,2-dimethoxy-2-phenyl-acetophenone (DMPAP) as photoinitiator was added to the solution. FT-IR analysis indicated that a polymer complex was formed between PAA and chitosan through hydrogen bonding. The solution mixture was then poured into a glass petridish and irradiated for 10 min using a low intensity UV lamp
(300 40 Recently a biological glue based on chitosan has been proposed and developed because of the high biocompatibility of this polysaccharide. Yamada et al investigated the possibility that tyrosinase-catalyzed reactions of dopamine could confer water resistance adhesive properties to semidiluted solutions of chitosan [74]. The schematic reaction is as follows: Dopamine + Tyrosinase = 0-quinone 0-quinone + Chitosan = “quinone tanned” Chitosan The evidence for the “quinone-tanning” chitosan was obtained by UV-visible spectrum analysis, which showed a shift towards the absorption peaks of dopaminochrome (280 nm and 475 nm) over time. The observed adhesive properties appear to be related to the increased viscosity associated with the modified chitosan. Samples from this high viscosity modified chitosan were spread onto dry glass slides, the slides were lapped and clipped together after being submerged in water, and the bound slides were held under water for 24 h. The samples were then removed from water, unclipped and tested to determine the shear strength required to separate the glass slides. Adhesive shear strength of over 400 kPa was observed for these samples while control chitosan solution conferred no adhesive strength. Chitosan is a linear polymer of glucosamine with a variable frequency of N-acetyl-D- glucosamine units and it has been reported to have antimicrobial, haemostatic and wound healing properties, making it an excellent biomaterial [75-78]. Rao et al investigated the hemostatic potential of chitosan films [77]. Hemostasis was studied in vitro by using 2% chitosan solution in acetic acid with whole blood. In vivo hemostasis 41 tests were undertaken by applying chitosan to punctured bleeding capillaries. Chitosan formed a coagulum in contact with whole blood in vitro and in vivo, the clotting time was also reduced by 40% in respect to the normal required time. The effect of chitin and chitosan on the proliferation of human dermal fibroblasts and keratinocytes were examined in vitro by Howling et al [79]. Chitosan (50 mg/ml) with high degree of deacetylation (89%) strongly stimulated fibroblasts proliferation after 3 day treatment, as assessed by the 3H thymidine assay. Alternatively, samples with lower degrees of deacetylation (37% and 58%) promoted less cell activity. The stimulatory effect on fibroblast proliferation required the presence of serum in the culture medium, suggesting that the chitosan may interact with growth factors and enhance their effect. The mechanism responsible for fibroblast proliferation is unknown, although it has been postulated that chitosan may function in a similar way to hyaluronan [80]. Chitosan may form polyelectrolyte complexes with serum components such as heparin [81] or potentiating growth factors such as platelet derived growth factors [82]. In contrast to the stimulatory effect on fibroblasts, chitosan (89%) inhibited human keratinocyte mitogenesis (40% inhibition) while showing no effect on keratinocytes proliferation. These results are in agreement with Denuziere et al [83] who examined cytocompatibility of films of chitosan with glycosaminoglycans with human keratinocytes. In their experiments cell inhibition was ~ 40% when compared to the controls. These results indicated that highly deacetylated chitosan is biologically more active than chitin and may possibly have more potential as a wound healing agent or dressing material. Risbud et al demonstrated the biocompatibility of chitosan hydrogel and epithelial cells [84]. The authors synthesized chitosan-gelatine hydrogel and used it as growth supportive substrata for respiratory epithelial cells. Macrophages were 42 employed to test the immunocompatibility of the gel. The gel did not exert a cytotoxic effect on macrophages, as assessed by tetrazolium reduction (MTT) and neutral red uptake assay, although they show round morphology and did not attach well on the gel. The epithelial cells cultured on the gel showed a proper attachment, normal morphology and growth. The hydrogel appeared to be a suitable choice for coating tracheal prostheses. Hidaka et al made membranes of 65, 70, 80, 94 and 100 deacetylated chitin, which were implanted subperiosterally over the calvaria of 100 rats [85]. Reactions were studied at 1, 2, 4 and 8 weeks after implantation. Membranes prepared with 65, 70 and 80% deacetylated chitin initially elicited marked inflammatory reaction that subsided in time with granulation tissue formation and osteogenesis. Membranes made of 94% chitosan showed mild inflammation and minimal osteogenesis. The authors postulated that the negative charge conferred upon chitin by the acetyl groups plays a role in the inflammatory response. Mori et al reported that chitin and chitosan elicited release of interleukin 8 from rat dermal fibroblasts (in vitro) and suggested that migration of neutrophil towards these materials might be related of this chemokine, which is a potent chemoattractant and activator of neutrophils [86]. Chatelet et al investigated the biological properties of chitosan films in respect to the degree of acetylation [87]. In particular, cell adhesion and proliferation on chitosan films was correlated to the degree of acetylation (DA), which was ranging from 2.5 to 47%. The films were cytocompatible at all DA. Higher DA resulted in the lower cell adhesion on the films and human fibroblasts appeared to adhere twice as much as keratinocytes, after 1 h incubation in culture medium. Human keratinocytes 43 proliferation increased when DA of chitosan films decreased, as assessed by the MTT test after 5 days incubation, thus DA influenced the cell growth in the same way as cell adhesion. Fibroblasts remain alive but do not proliferate on chitosan film. This behaviour is related to an extremely high adhesion on the films, which inhibits cell growth. It was also noticed that a smoother surface topology led to a better cell adhesion. Yan et al prepared and characterized chitosan/alginate films of different molecular weight. Chitosan solution (0.25% w/v in a 1:1 v/v solvent mixture of 2% acetic acid and acetone) and alginate solution (0.25% w/v in distilled water) were vigorously mixed for 20 minutes to result in a suspension, which was cast into a polyethylene petri dish overnight at room temperature [88]. The properties of the dried films obtained from the above treatment were studied in relation to the chitosan molecular weight. Films prepared with low molecular weight (1.30 x105D) were twice as thin and 55% less permeable to water vapour, compared to films prepared with higher molecular weight (10x105 D). In vitro cytotoxicity of the films incubated with human fibroblasts was determined using the MTT assay and it showed the cells retained more than 91% viability over a four-day period of exposure. Gaserod et al proved that a chitosan-alginate membrane was strengthened by the addition of calcium ions [89], while Vandevord et al evaluated the biocompatibility of chitosan scaffolds in mice [90]. Porous chitosan scaffolds were implanted in the ventral (n=20) or dorsal (n=20) skin of mice. The animals were sacrificed after 1,2,4,8 or 12 weeks. Macroscopic inspection of the implantation site revealed no pathological inflammatory responses. Histological assessment indicated marked neutrophil accumulation within the implant, which resolved with increasing implantation time. These cells are normally present during the acute phase of inflammation and it was unusual to observe neutrophils responding specifically to chitosan in the absence of the cardinal signs of inflammatory 44 characteristics, such as erythema and edema. Other authors have also reported a chemotactic effect of chitosan on neutrophils [91]. Gram staining and limulus effect assays showed no evidence of infection or endotoxin. Collagen was observed within the chitosan pore spaces, indicating that connective tissue matrix was deposited within the implant. Angiogenic activity associated with the external implant surface was also observed. Cellular immunoresponses were determined by lymphocyte proliferation assay and antibody responses were measured using ELISA technique. These assays indicated a very low incidence of chitosan specific reactions. Similarly, other studies have recently reported that chitosan and its derivatives are potentially favourable materials as substrates for the growth of endothelial cells and chondrocytes [92, 93]. Mi et al demonstrated the use of a chitosan wound dressing device for sustainable antibiotic delivery [94]. A bilayer chitosan membrane was prepared by a combined wet/dry phase inversion method. The membrane consisted of a dense upper layer (skin layer) and a sponge-like lower layer (sublayer). The membrane had a very good oxygen permeability and promoted water uptake capability. The bilayer chitosan was doped with silver sulfadiazine (AgSD) and the release kinetic was studied in a vertical permeation cell at 35 0C by atomic absorbance and spectrophotometrically. The release of sulfadiazine occurred as a burst release on the first day and than tapered off to a much slower release. In contrast, the release of silver was slow and sustained. The inconsistent release of Ag and SD may be attributed to the Ag binding to the amino groups in the chitosan membrane, while the SD was free to diffuse through the bilayer chitosan [95]. The chitosan membrane was also used as a wound dressing device on rat skin incisions, which were previously infected with Staphylococcus Aureus and 45 Pseudomonas aeruginosa (2.5x2.5 cm2). The chitosan dressing was able to significantly reduce the colony-forming units in the wounds. Degradability is an essential characteristic of chitosan and several studies have been reported on this subject. Tomihata et al studied the degradation of chitin and chitosan films both in vitro and in vitro [96]. In this study, Chitin was deacetylated to various percentages with NaOH to obtain chitosan. The specimens used in this study were deacetylated by 0% (Chitin), 68.8%, 73.3%, 84.0%, 90.1% and 100% (Chitosan). Films with a thickness of 150 mm were prepared from these specimens by solution casting. The equilibrated water contents of the films were 52.4 wt% (chitin), 73.8 wt% (68.8 mol%), 64.2 wt% (73.3 mol%), 61.8 wt% (84.0 mol%), 57.8 wt% (90.1 mol%) and 49.7 wt% (chitosan). The tensile strength of the water-swollen films was 2.39 kPa (chitin), 1.93 kPa (68.8 mol%), 2.27 kPa (73.3 mol%), 3.13 kPa (84.0 mol%), 2.87 kPa (90.1 mol%) and 4.24 kPa (chitosan). The maximum water content and the minimum tensile strength observed for specimens deacetylated between 0 and 68.8 mol% might be ascribed to the lowered crystallinity by deacetylation of chitin, since both chitin and chitosan are crystalline polymers. Unlike their physical properties, in vitro and in vivo degradation of these films occurred less rapidly without passing a maximum or minimum, as their degree of deacetylation became higher. The in vitro degradation was carried out by immersing the films in buffered aqueous solution of pH 7 containing lysozyme at 37 0C, while the in vivo degradation was studied by subcutaneous implantation of the films in the back of rats. It was found that the rate of in vivo biodegradation was very high for chitin and 68.8 mole % chitosan, compared with that of for the 73.3 mol% film. The films with higher degree of deacetylation showed slower degradation. The tissue reaction towards highly deacetylated chitin was very mild 46 although they had cationic primary amines in the molecules. This apparent contradiction may be ascribed to the almost null zeta potential of chitosan films. This implies that the surface of the chitosan films may be so bioinert that it does not provide any stimulus to the surrounding tissue for degradation. Lee at al also investigated the biodegradability of chitosan [97]. N-acetylated chitosan films were incubated with Lysozyme (51100 units per mg) in a neutral buffer solution. The higher lysozyme susceptibility was achieved for chitosan having a higher degree of N-acetylation. For example, N-propionyl (41.7 % acetylation) and N-butyryl (40.4 % acetylation) degraded more than 40% (by weight) in less than 8 hours. This result suggests that the segments consisting of glucosamine residues are not accessible to the lysozyme active site [98]. However, others suggested that N-acetyl glucosamine segments, which lysozyme could access, might consist of more than 3 N-acetyl glusosamine monomers and that lysozyme could not act on glucosamine segments where relatively small fractions of N-acetyl glucosamine residues were randomly distributed [99]. Hirano and Hagi reported that lysozyme did not hydrolyzed N-butyryl chitosan [100]. Some authors also reported that the lysozyme from chicken egg white did not hydrolyzed N-propinyl chitosan [101]. The inconsistency of these reports may be due to the origin of the enzyme, the different degree of substitution for N-acryl groups and the physical form of the samples. Rao et al treated the surface of chitosan films with glutaraldehyde, carbodiimide and plasma glow discharge to elicit the effect of enzyme degradation [77]. Of the seven enzymes used, leucine amino peptidase caused maximum degradation (62% weight retained) after 30 days of enzyme incubation at pH 47 8.7. The films were sterilized by autoclaving, which caused the least decrease in tensile strength. These researchers also performed an intramuscular implantation of chitosan films in rabbits for toxicity tests. After 3 and 7 days, the retrieved films did reveal a pyrogen free status. The biocompatible nature of chitosan has stimulated scientists to investigate possible mechanisms of crosslinking this polysaccharide without using toxic substances. Conventional chitosan crosslinking reactions have involved a reaction of chitosan with formaldehyde, glutaraldehyde and epoxy compounds. However, all these crosslinking agents are chemically synthesized and are physiologically toxic [102-104]. 1.7 Genepin Genepin has been used by Sung et al as a crosslinking agent to fix biological tissues as an alternative to the conventional toxic compounds [104]. Genepin is an iridoid glucoside abundantly present in gardenia fruits. A chitosan based polymeric network has been recently crosslinked with genepin [105]. A possible mechanism for this involves a heterocyclic linking of genepin with chitosan by a nucleophilic attack and the formation of amide linkages leading to crosslinking of chitosan and oligomer creation in a crosslinked network. These chitosan networks exhibit pH dependant swelling characteristics. The biocompatibility and degradability of genepin was investigated by Mi et al, who compared it with glutaraldehyde [106]. Genepin and Glutaraldehyde were used to crosslink chitosan microspheres. The spheres were injected intramuscularly in rats. At 1 week postoperatively, the degree of inflammatory reaction for tissue implanted with 48 fresh microspheres declined significantly. However, the inflammatory cells (lymphocytes) surrounding the tissue implanted with glutaraldehyde and genepin spheres were still abundant. After 12 weeks, the inflammatory cells had almost disappeared for the genepin and fresh chitosan microspheres at the injection site, while the inflammation was still significant around the glutaraldehyde microspheres. The degradation rate of the Glutaraldehyde microspheres appeared to be faster than the genepin ones 20 weeks post implantation. 1.8 Objectives Brittleness, lack of flexibility, water solubility, low bonding strength, tissue thermal damage and risk of viral infections are among the most significant faults of laser- activated protein solders. The following investigations aim to develop and characterize new biomaterials to overcome some of the limitations of traditional solders and tissue repair techniques. Genipin and chitosan appear to be promising compounds for the development of novel tissue adhesives. Genipin is indeed a non-toxic crosslinker of amino groups that are abundant in proteins; chitosan is a linear polysaccharide, suitable for the fabrication of films with good mechanical properties and also capable of improving healing in wounds. 49 Chapter II Sutureless Laser-Soldering Technique for Reversal Vasectomy 2.1 Introduction Vasovasostomy is the surgical operation of reanastomosing the vas deferens after previous vasectomy to restore fertility; it is a challenging and laborious microsurgical procedure that requires great skill and precision during intraluminal and adventitial suturing (figure 1). A B Figure 1. (a) Diagram of reanastomosing the vasa with sutures. (b) Vas deferens anatomy. (http://www.malereproduction.com). Current suturing techniques for vasovasostomy have a patency rate of 80%-90% but they require several intraluminal sutures to align the lumen and other adventitial sutures to ensure the vas repair (figure 2) [107]. Operative time is usually in the region of several hours and particular skill and precision is required to successfully complete this procedure. Laser tissue soldering is conversely a sutureless technique that offers the 50 possibility of simplifying this complicated procedure. For example, solid albumin solder proved to decrease the duration time of microsurgical procedures such as peripheral nerve, small vessel and vas deferens anastomosis [54, 56, 108,109]. Figure 1. (c ) Histological cross-section (H&E) of a human vas deferens, which is easily identified by its star-shaped lumen and large amount of surrounding smooth muscle. There is an inner longitudinal layer of smooth muscle, surrounded by middle circular layer, and an outer longitudinal layer attached to connective tissue (adventitia). (http://www.lab.anhb.uwa.edu.au). Seaman et al completed rat vasovasostomy (n=10) with three intraluminal sutures and two adventitial sutures; albumin solder (25% w/v) with IG (0.25% w/v) was applied around the vas and bonded to tissue with a diode laser (l=808 nm, Irradiance ~12.9 W/cm2, pulse duration 0.2 s). Control rats were operated with six intraluminal sutures and eight advential sutures. Patency was determined by microscopic observation of sperm in the vas fluid and unobstructed passage of a dye through the lumen at the anastomotic site. One month after the operation, both techniques achieved a patency rate of ~ 80% but the mean operative time was significantly shorter for the laser group. Eight granulomas developed in the laser group and five developed in the suture group [107]. 51 A B Figure 2. (a) Intraluminal sutures are applied to align the vasa. (b) Adventitial sutures complete the vas anastomosis. (http://www.vasectomymedical.com). Mingin and Ditrolio performed vasovasostomy on five adult men with two intraluminal sutures and 3-5 drops of albumisol that were bonded to tissue by an argon laser (l=488 nm, P=0.5 W, irradiation time = 6s). Sperm count and motility were satisfactory after four months and three man achieved paternity within a year after the operation [110]. Trickett et al attempted a sutureless vas anastomosis in rats [108]. Liquid albumin solder (45% w/v) with IG (0.3% w/v) was applied around the vas deferens and irradiated with a diode laser (l=808 nm, Irradiance ~28 W/cm2, pulse duration 2 s). A stent was placed to align the lumen and removed before the anastomosis. The control group was operated with four intraluminal sutures and four adventitial sutures. Eight 52 weeks after the surgical procedure, the granuloma formation (G) and patency rate (P) for the conventional suture technique (G, 14%; P, 93%) were significantly better than observed for the laser soldering technique (G, 52%; P, 50%). In this thesis, a collar made of solid albumin solder was wrapped around the anastomotic site and activated by a laser to repair the vas deferens. The lumen alignment was achieved by a biocompatible chitosan stent that self-expanded upon insertion in the vas lumen. A stent made of albumin could not be manufactured as the solder brittleness and rigidity prevented the preparation of a narrow stent (diameter~0.5mm). Also, the solder solubility impaired the stent insertion in the vas in vivo. The limited mechanical properties of albumin solders along with their water solubility has prompted the search for new biomaterials in the design and preparation of a self-expandable stent. A large variety of stents are currently used in clinical practice to relieve pathological obstruction of tubular structures in vascular, urologic and gastroenterologic systems. Stents can be manufactured with metals, synthetic polymers or biomaterials; they can also have a rigid or self-expandable structure [111-117]. Temporary stents, which are made of synthetic materials or biopolymers, usually degrade over time once inside the body. Therefore, they typically do not require surgical removal after their implantation in patients [116,118]. As compared to metals and some synthetic materials, biopolymers are more biocompatible, less traumatic to surrounding tissue and possibly beneficial to wound healing [116,119]. In recent years, self-expandable stents have been introduced because of their unique property of enlargement following tissue insertion. Such stents are easier to implant than rigid stents [111,113,120]. Detrimental 53 complications secondary to stent insertion have been clinically described, such as migration, tissue trauma, fragmentation and encrustation [113,121-124]. Improvements in both quality and performance of stents are necessary to reduce clinical complications and failure. 2.2 Materials and Methods 2.2.1 The Protein Solder Lyophilised BSA (Sigma, MO) was placed in a small eppendorf tube with Carbon Black (0.25 ± 0.06 %, w/w) and distilled water to a final concentration of 65% ± 1 %. A metallic rod with a flat end (compressor) mixed and compressed the component to ~2/3 of their initial volume. When albumin exceeds a concentration of ~56%, its viscosity increases and the solder becomes a malleable solid paste [62]. The solder paste can be shaped in different forms, such as hollow cylinders or rectangular bands, to tailor the repair site. For this study, a parallel plate vice pressed the solder and its thickness was reduced to 0.13 ± 0.02 mm. The protein sheet was then cut with a scalpel into rectangular bands (surface area dimensions ~ 3 mm x 6 mm), and accurately rolled around a teflon cylindrical rod to form open collars of ~2.5 mm diameter. The attenuation length (1/e light attenuation) of the solder with carbon black was estimated to be ~0.03 mm at the wavelength of 808 nm, as previously measured [46]. The solder collars were placed in eppendorf tubes and then stored in the refrigerator at 5°C for a week, before the surgical procedure. The solder was composed of sterile compounds, which were handled and mixed using sterile instruments. No further sterilisation was carried out on the solder. 54 2.2.2 The Laser System A GaAlAs diode laser (Qphotonics, L.L.C., VA, USA), coupled with a multimode optical fiber via an FC connector, was used in this study. The fiberoptic cable (core diameter 400 mm, numerical aperture 0.22) was inserted in a hand-held probe to provide easy and precise beam delivery by the operator. The laser emission wavelength was 808 nm, with an output power of 0.25 ± 0.02 W and beam spot size on the solder of approximately 600 mm, corresponding to an irradiance of ~88 W cm-2. The laser emitted light in continuous mode over the time interval required to achieve the prescribed laser exposure dose (energy/solder weight). Because the laser is not eye safe (Class IV), safety glasses were worn by all staff in the operating theatre. 2.2.3 Chitosan Film Deacetylated chitosan from crab shells (Sigma, St. Louis, MO) was dissolved to a concentration of 4.3% (w/v) in a water solution containing 2% (v/v) acetic acid. The resulting chitosan gel was spread evenly (thickness ~2-4 mm) over a plastic plate and allowed to dry for 12 hours at room temperature [125]. The resulting chitosan film was carefully detached from the plate and immersed in a 10 N NaOH solution for 15 minutes, making the film water insoluble. The basic film was then rinsed several times with phosphate buffer solution (PBS) until its pH was neutralized. A digital caliper was used to measure the thickness of the films (0.3-0.6 mm) once dry. 2.2.4 Thermogravimetric Analysis (TGA) Chitosan adhesive films were analysed using a Perkin Elmer Pyris 1 Thermogravimetric Analyser to evaluate the water content and degradation temperature of the films in air 55 (weight 10-15 mg). The temperature was increased from 20 to 600 0C at a rate of 40 0C/min. The mass of the films was continuously recorded as a function of temperature and the first derivative calculated to assess the adhesive degradation (derivative minimum peak). 2.2.5 Stent Preparation The chitosan film was cut into rectangular strips with approximate surface dimensions of ~0.3 x 5 cm and a thickness of 0.55 ± 0.05 mm. These strips were immersed in PBS and wound around a metallic cylindrical rod (diameter = 0.5 mm) in a helical fashion. The strips were elongated about 30% while wound around the rod (figure 3). Metallic Rod Chitosan Strip Elongation Force Helix Pitch Strip Overlap Figure 3. Schematic drawing of the stent preparation. The stent is manufactured by pulling and winding the chitosan strip around the metallic rod. There is a partial superimposition of the strip at each turn to close the helical structure. 56 The film was partially superimposed at each turn of the helix to ensure its closure. A “circular helix” tube (stent) was finally obtained and was coated externally with a thin layer of chitosan solution to prevent its expansion and opening. The stent was allowed to dry, removed from the rod, shortened to the desired length (3-5 mm) by a sterile surgical blade and stored in a sterile plastic container at room temperature. The structure of the stents was analyzed in detail by scanning electron microscopy (SEM). Specimens were sputter coated with AuPd before being viewed at 20 KV in the microscope (Jeol 100cx-II with ASID attachment for SEM). 2.2.6 Elasticity Test A calibrated tensiometer (Instron Mini 55, Instron, Boston, Ma), controlled by a personal computer, measured the Young’s modulus of chitosan films. Merlin IX software recorded and analyzed the data. The tensiometer consisted of a single column system with pneumatic grips and a static load cell (maximum load 50 N). The films were cut into rectangular strips, which were immersed in PBS and clamped to the system grips. The tensiometer stressed each strip as the upper grip moved away from the other grip at a constant speed of 25 mm/min. The test stopped when the strip broke in two. The test parameters are summarized in table 1. 57 Table 1. Parameters of the elastic test. N, number of tested strips. Max Stress, maximum stress. Max strain, Maximum strain. E, Young’s modulus. Max load, maximum load applied on the specimen. Thickness, width, length, dimensions of the specimen. Std. error, standard error. All the test parameters passed the normality test but the specimen length, since the distance between the grips was intentionally kept constant. Parameters N Max Max Strain E Max Load Thickness Width Length Stress (%) (MPa) (N) (mm) (mm) (mm) (MPa) Mean 33 0.5922 105.3 0.7655 0.9255 0.034 4.55 7.61 Std. Error 33 0.0292 6.4 0.0288 0.0572 0.001 0.12 0.30 KS Distance 33 0.1509 0.09040 0.1497 0.09841 0.1335 0.09229 0.3787 P value 33 >0.10 >0.10 >0.10 >0.10 >0.10 >0.10 0.0002 Passed 33 YES YES YES YES YES YES NO Normality Test 2.2.7 Stent Self-Expansion Twenty vas deferens were freshly harvested from adult male wistar rats, which were sacrificed for other experiments. A chitosan stent was inserted in each vas (diameter ~0.5 mm) to observe and test its expansion. The vas was first enlarged by metallic dilators of larger diameters (~0.6-0.8 mm). The stent was placed on the tip of a thin teflon rod (diameter ~0.3 mm) and then inserted into a vas deferens. The procedure was performed with microforceps to precisely handle stent and tissue, under an operative microscope (x20). 58 2.2.8 Laser Tissue Soldering Sutureless laser repair of the vas deferens was attempted using the solder collars and the chitosan stents in a vasovasostomy model. The stability of the stent self-expansion was also tested in vivo as the vas deferens has strong muscular walls capable of contracting and exerting inward pressure. Fourteen adult male wistar rats (weight ~270 g) were anaesthetized with 0.37 cc of sodium pentobarbital based upon the rodents’ weight. The vas deferens of the animals was obstructed bilaterally by a titanium clamp under an operative microscope, as described previously [126]. The vas was re-exposed after two weeks to simulate the clinical scenario of reversal vasectomy. No sperm granuloma was developed in sixteen obstructions. Therefore, the clipped sections were resected and the vasa stumps were approximated end to end with a surgical clamp. Sperm were retrieved from the proximal end of the vas and motility was observed in all cases under an optical microscope. A chitosan stent was partially inserted in the proximal vasal end, following the procedure described in the previous section. The distal vasal end was also enlarged by dilators and gently pushed against the stent until complete insertion of the stent was achieved (figure 4a). 59 A Chitosan Stent Vas Deferens Lumen Solder Collar B Anastomotic Laser Fiber Site Figure 4. Diagram of the surgical procedure. (a) The chitosan stent is inserted in both vasal ends after vasectomy. (b) The solder collar is bonded to the tissue by the laser to complete the vas anastomosis. Sperm were often observed in the stents because of capillary effect. The vas anastomosis was completed without suture, but rather with a laser welding technique [46, 55, 62]. A thin film of solid albumin solder (collar) was wrapped around the contiguous ends of the vas and the surgeon irradiated the adhesive by moving continuously the beam across its surface at a speed of ~1 mm/s (figure 4b). Moderate tissue shrinkage under the adhesive was observed during laser irradiation. The laser was absorbed by the carbon black and the generated heat welded the protein collar around the vas anastomosis. All tissue repairs were performed using a fixed laser power, solder surface area and thickness as summarized in table 2. 60 Table 2. The laser soldering parameters are given as mean and standard deviation. N, number of vasal anastomosis. Collar Surface, dimensions of the solder collar surface. Spot Size, laser spot size on the tissue. Power, laser power. Time, laser irradiation time. Weight, collar weight. Dose, laser energy per mg of solder. N Collar Collar Power Spot Size Time Weight Dose Surface Thickness (mW) (mm) (s) (mg) (J/mg) (mm) (mm) 13 »3 x 6 0.13 ± 0.02 250 ± 20 ~0.6 171 ± 36 4.1 ± 0.9 10.5 ± 1.9 Upon completing the laser anastomoses (n=13), the abdominal muscle layers and skin were closed with a continuous 3-0 chromic catgut suture. Three vasal anastomoses were performed as control, by using the same procedure described above without the chitosan stent. Six laser anastomosis, without stent insertion, were also harvested ex vivo and tested acutely to assess the repair strength of the solder. The other animals were sacrificed 8 weeks later to examine the patency of the anastomosis and chitosan stents inside the vasa. Some specimens were harvested and stained with Masson’s trichrome for histological examination. 2.2.9 Statistical Analysis Merlin IX software generated the strain-stress plot. Data was also analyzed by Prism Software (GraphPad Software, Inc., San Diego, Ca), which calculated the Kologorov- Smirnov (KS) distance to test the normal distribution of the collected measures. The significant level for the p value was 0.05. 61 2.3 Results 2.3.1 Thermogravimetric Analysis The TGA analysis of the film showed a continuous weight loss that sharply increased at 305 ± 4 0C (degradation temperature = derivative minimum peak). The film degraded 42.6% ± 1.7% (n=5) at 305 ± 4 0C (figure 5). It is most likely that the weigh loss was due to water evaporation, before the sharp degradation took place. 100 Degradation Derivative 80 60 40 20 Weight (%) 0 0 50 100 150 200 250 300 350 400 450 500 550 600 650 -20 -40 Temperature (0C) Figure 5. The weight loss of chitosan film is given as a function of temperature. The adhesive burnt at ~ 305 0C, where the weight loss rate achieved its maximum (minimum peak of the first derivative). Water likely evaporated before the sharp degradation occurred. 2.3.2 Stent Preparation A typical stent is shown in figures 6a, 6b and 6c. The stent geometry is a circular helix with constant pitch and inner diameter. Strip overlaps are noticeable at each turn of the helix and the rigid tubular structure appears completely sealed. No appreciable irregularities were detected at the openings of the stents. 62 Figure 6a. SEM view of a 3 mm long chitosan stent (x50). The cylindrical helix shape is revealed in detail. The helix pitch is ~1.4 mm and the strip overlaps at each turn to close the tubular structure. Figure 6b. SEM view of the stent aperture (diameter = 0.5 mm, x100). The aperture profile appears regular without appreciable deformations. 63 Figure 6c. Typical helix turn of the chitosan stent. The tubular structure is sealed along the turn line (x50). 2.3.3 Elasticity Test The wet chitosan strips behaved elastically, as shown by the strain-stress diagram of figure 7. The E value was 0.7655 ± 0.1653 Mpa (mean ± standard deviation); the maximum strain and stress were 105% and 0.5922 Mpa respectively. The test results are shown in table 1. 2.3.4 Stent Expansion After insertion in the vas, the stents were observed to self-expand to about 50% over their initial diameter. This dilation did not regress but remained stable, suggesting that the expansion process was irreversible. The insertion procedure was done safely without crushing or damaging the stent or tissue. 64 Figure 7. Strain-stress diagram of a chitosan strip, tested by the tensiometer. The linear behaviour of the strain proved the elastic nature of the chitoasan film. The strip doubled its initial length before breaking under stress. 2.3.5 Laser Tissue Soldering The maximum load achieved by the laser anastomosis ex vivo was 0.28 ± 0.17 N (n=6); all samples failed under stress because of cohesive rupture. Eight weeks after the operations all the anastomoses were in continuity at the anastomosis site (n=16) and the solder collars were degraded. In the experimental group, sperm granulomas (mean dimensions ~0.9x0.4 mm) developed proximally (n=2) and at the anastomotic site (n=7). The other 4 anastomoses were stenotic (figure 8). The anastomotic site was harvested and sperm retrieved from the vas proximally and distally for microscopic analysis. In all cases, sperm debris were found distally and no motile sperm was observed. Sperm motility was observed proximally in only 4 65 specimens. Twelve chitosan stents were rescued from the vasa and their structural integrity was assessed by inserting a rod inside them (diameter = 0.5 mm). M A Figure 8. A vas stenosis in the proximity of the anastomotic site, eight week after the operation. The thickness of the muscular walls (M) attached to the adventitia (A) appears reduced. (Longitudinal section, Masson’s trichrome, x40). The stents were larger than the rod and they appeared intact under the operative microscope (figure 9). One stent could not be found in a sperm granuloma. Two control animals developed a sperm granuloma (figure 10). The histology showed inflammatory cells such as macrophages, giant cells and lymphocytes, inside the granulomas of the experimental and control groups. Others described similar findings inside rat granulomas after vasectomy [127]. 66 2.4 Discussion Vasovasostomy is a particularly challenging microsurgical technique that requires several intraluminal and adventitial sutures to align the lumen and secure the vas anastomosis. The lumen alignment is crucial to avoid sperm granulation and vas obstruction. The insertion of a stent in conjunction with a solder collar was simpler and easier than suturing, under an operative microscope. Despite the relative simplicity of the laser technique, granuloma formation compromised the overall success of the operations. The development of sperm granulomas was most likely facilitated by the spermicidal properties of chitosan, already reported by Sanford [128]. Unfortunately, we became aware of Sanford’s report only after experiment completion. The laser energy could also induce stenosis secondary to thermal damage and promote granuloma formation [129]. Trickett et al reported a thinning of the muscular walls at the anastomotic site in all vas deferens repaired with the laser soldering technique [110]. The laser irradiance used in that study was ~28 w/cm2, while Seaman et al used only ~13 W/cm2 to successfully perform vasovasostomy [107]. However, both authors neglected to report how much energy and solder they applied on the anastomotic site. The energy dose (energy/solder weight), along with the temperature at the solder-tissue interface, is an important parameter to optimize in order to avoid thermal damage of tissue. We used an irradiance of ~88 W/cm2 and dose of 10.5 J/mg, that ensure patency of the vasa anastomosis eight weeks after the surgical procedure. These laser parameters are also similar to the ones used to repair successfully rat tibial nerves in vivo [55]. 67 S T SG Figure 9. Longitudinal section of a sperm granuloma at the anastomotic site, 8 weeks post surgery. The chitosan stent (T) is open despite the inflammatory response of the granuloma and fragments of solder (S) are not degraded yet. Slide preparation partially disrupted the stent. (Masson’s trichrome, x40). 68 M L Figure 10. Longitudinal section of the proximal site of a sutured vas deferens, 8 weeks post surgery. The muscular walls (M) appear intact and the lumen (L) enlarged. (Masson’s trichrome, x40). Many metallic, synthetic or biodegradable stents are currently used to relieve pathological obstruction of ducts in urology and gastroenterology. Their common side effects are stent migration, fragmentation and encrustation [113,121-124]. The search for a new design with improved performance is therefore necessary. In this study, a temporary chitosan stent was developed and manufactured with a self- expanding mechanism and a helical design. The stents expanded most likely by releasing their elastic energy when in contact with tissue moisture. Upon insertion in rat vasa, they were observed to expand ~ 50% over their initial diameter. Self-expansion occurs when the thin layer of chitosan gel coating the stent is dissolved by tissue moisture. The elastic nature of the chitosan strip and the bending stresses generated during the stent fabrication, are probably responsible for the opening of the helical 69 structure. The water content of the chitosan films may be estimated as 40% (figure 5) and this possibly enhanced the elastic properties of the strips. Chitosan stents can either offer a rigid support or exert a controlled pressure on the tissue walls adjacent to them. It may be possible to estimate the average pressure (p) exerted by the stent upon surrounding tissue, using the Young’s modulus (E). The work (L) done by the tube while opening causes an increase in volume (DV) given by L = pDV, hence p = L/DV. The work is proportional, in a first approximation, to the elastic energy accumulated by the chitosan strip during its elongation to prepare the stent. Therefore, L µ (k/2)Dx2, where Dx is the chitosan strip elongation and k is the elastic constant for the chitosan strip. We also know that k = ES/l0, where S is the strip cross-sectional area and l0 is the initial strip length. Finally, we obtain: 2 p µ (Sl0/DV)(E/2)(Dx/l0) . The pressure exerted by the stent on the tissue may be controlled by varying the relative elongation and cross sectional area of the chitosan strip. Upon substitution of the typical values of our stents: DV = 3.9 mm3 (50% expansion), E = 0.7655 MPa, S = 0.25 2 mm , l0 = 9 mm, Dx = 3 mm, and assuming 50% of the elastic energy was used to expand the vas deferens; we find p ~ 90 mmHg. In these experiment, the stents enlarged the vas until the tissue walls equilibrated with the expansion pressure. The stents were stable for 8 weeks inside the animals and did not shrink back to their initial dimensions, suggesting the self-expanding process was irreversible. The stents remained dilated because tangential frictional forces were probably created between the chitosan strip at the helix turn. Such friction prevented narrowing of the stent diameter. 70 It is remarkable that 7 chitosan stents remained enlarged in sperm granulomas, where the pressure can be higher than the normal physiological level (figure 6). The stents did not migrate in 6 cases, where granuloma did not form at the operation site. No conclusion could be drawn about the stability of the stent position when granulomas were present at the anastomotic site (n=7). For clinical applications, requirements for stent use are variable. Chitosan stents can be prepared with different dimensions. The diameter can vary between 0.3 mm and a few centimeters, and the length can range from a few millimeters to several centimeters. Chitosan was chosen among others biomaterials because of its tissue biocompatibility, elasticity and antimicrobial activity, as well as haemostatic and wound healing properties [119, 76, 130-134]. The stent preparation presented in this study can be extended to other suitable biomaterials with elastic properties. Chitosan stents are not indicated for vasovasostomy because of their spermicidal properties. Nevertheless, they may be implanted in the prostate, bile duct or ureter during open surgery. Problems may arise following endoscopic insertion of chitosan stents if the stents are not implanted quickly, as bodily fluids may prematurely trigger the self-expanding mechanism. 71 Chapter III Albumin-Genipin Solder for Laser Tissue Repair 3.1 Introduction Albumin glues have been developed for LTS, but to date they have not been applied routinely in clinical procedures [29]. One of the main reasons for surgeons not embracing this new laser technology is that suturing still provides the strongest tissue repair. The weak tensile strength of laser glues means that relatively large amounts of solder have to be dispensed on repair sites to ensure proper bonding and prevent tissue dehiscence. Furthermore, albumin glues must be coagulated at 65-70 0C by laser light to achieve satisfactory weld strength and therefore the chance to thermally induce cell death is significant. A stronger solder would reduce the amount of glue required and also may minimise the risk of causing thermal damage. During reversal vasectomy (Chapter 2), the heat produced by the laser inside the solder collar diffused to the tissue interface and probably induced the stenosis of the vas deferens, despite the success of the anastomoses. The first fluid albumin solder was introduced in the late 1980’s and it was shown to successfully repair rat urethra [38]. Subsequently, the protein concentration was increased to obtain solid albumin solders (~62% albumin) and improve their cohesive bond strength after laser irradiation [135, 54]. The solid solder was nevertheless soluble, and lacked flexibility when in dry form, limiting the possibility of tissue manipulation by surgeons during vascular anastomosis. To overcome these limitations, the solid solder was denatured in hot water and transformed into a rubber-like material, which was insoluble and still able to be laser welded. This heat-denatured solder was cylinder- 72 shaped and was used successfully for aortic anastomosis in rats [56]. The major drawback of this technique was the weak tensile strength of the repairs due to the decreased solubility of the denatured solder. Indeed, it has recently been reported by Lauto et al that repair strength is reduced when solder solubility decreases [52]. Another development of solid albumin solders was the design and manufacturing of “2- layer” systems, in which the layer in contact with tissue absorbs the laser and bonds to tissue while the second layer provides cohesive strength and flexibility [46, 68]. The first version of the “2-layer” solder lacked flexibility as both strata were made of albumin films [46]. Subsequently, a resorbable poly(lactic-co-glycolic acid) (PLGA) membrane doped with fluid albumin solder made the 2-layer adhesive flexible and more versatile for surgical applications [68]. Despite advances in design and manufacturing, the crucial issue of increasing bond strength of solders at the tissue interface has not progressed since the early days of LTS, with the basic composition of albumin glues remaining unaltered. To overcome this impasse, it appears necessary to modify either the glue composition or the bonding mechanism. In this investigation, a slow crosslinker (genipin) was added to albumin solid solders to enhance the bonding strength and to exploit molecular crosslinking for albumin glue activation. Genipin (figure 1) is extracted from the fruit of Gardenia jasminoides that provides yellow, red and blue colorants in the food industry [136]. G. jasminoides has been also used in traditional Chinese medicine for diuretic, choleretic and hemostatic purposes [137, 138]. Genipin has been reported to spontaneously react with amino acids or 73 proteins to form blue pigments. Park et al incubated methylamine (2.2 mmol) with genipin (0.22 mmol) at 70 0C for 5 hours to obtain the blue pigments, which Figure 1. Molecular structure of Genipin. were thereafter passed through Bio-Gel P-2 resin yielding five fractions [139]. Four fractions were blue while the last one was colourless with peak absorption at 292 nm. This fraction was very unstable and turned blue if heated, proving to be the true intermediate compound of the blue pigment. 1H-NMR and 13C-NMR showed the fifth fraction was composed of two epimeric isomers, where nitrogen substituted oxygen in the genipin hexagonal ring (figure 2). Paik et al obtained blue pigments by reacting genipin with glycine, lysine, or phenylalanine. Thermal degradation reactions at temperatures of 60-90 0C were carried out at different pH levels within the range 5.0-9.0. The blue pigments remained stable after 10 h at temperatures of 60-90 degrees C, and in some cases, more new pigments formed. The pigments were more stable at alkaline pH than neutral and acidic pH [140]. 74 (G) (M) Figure 2. Proposed mechanism of the blue pigment formation (fifth fraction) due to the interaction of genipin (G) with methylamine (M). Tissue collagen and BSA contains a large amount of aminoacids such as glycine, lysine and proline that spontaneously react with genipin. The toxicity of genipin crosslinked collagen was assessed by Sung et al, who cultured in vitro mouse 3T fibroblasts on the surface of porcine pericardia fixed with genipin or glutaraldehyde [141]. Tissue was fixed at 37 0C for three days, in a buffer solution (pH ~7.4) containing genipin at a concentration of 0.6 % w/v. Most of the collagen amino groups (~98%) were estimated to react with genipin [142]. Light microscopic examination revealed that the cytotoxicity of genipin was significantly lower than that of glutaraldehyde. Moreover, the results obtained by the cytotoxic assay (MTT) implied that genipin was about 10000 times less cytotoxic than glutaraldehyde and the proliferative capacity of cells after exposure to genipin was approximately 5000 times greater than that after exposure to glutaraldehyde. Neocollagen fibrils made by these fibroblasts were also observed on the genipin-fixed tissue while no synthesis of collagen fibrils was observed on glutaraldehyde fixed tissue. 75 Several in vivo studies confirmed the non toxic nature of tissue treated with genipin. In these investigations, crosslinked tissue was implanted in animals and retrieved at different time point to assess the degree of inflammatory reaction [143-146]. Although some inflammatory cells initially infiltrated the genipin crosslinked tissue, they disappeared within a few months. By contrast, glutaraldehyde treated tissue always induced more inflammatory reaction than tissue cured with genipin and the inflammation did not diminish with time. 3.2 Materials and Methods 3.2.1 The Laser System A GaAlAs diode laser (Qphotonics, L.L.C., VA, USA), coupled with a multimode optical fiber via an FC connector, was used in this study. The fiberoptic cable (core diameter 200 mm, numerical aperture 0.22) was inserted in a hand-held probe to provide easy and precise beam delivery by the operator. The laser emission wavelength was 808 nm, with an output power of 0.17 ± 0.01 W and beam spot size on the solder of approximately 300-400 mm, corresponding to an irradiance of 241-135 W cm-2. The laser emitted light in continuous mode over the time interval required to achieve the prescribed laser exposure dose (energy/solder weight). Because the laser is not eye safe (Class IV), safety glasses were worn by all staff in the operating theatre. 3.2.2 Solder Preparation Genipin powder was stirred in deionized water at 5 0C for approximately 24 hours until a uniform suspension was obtained and subsequently heated at 37 0C for 30 minutes to partially dissolve the genipin without inactivating it [142]. The initial genipin 76 concentration could not exceed 1% (w/v), due to the limited solubility of the crosslinker in aqueous solutions. IG was dissolved in the genipin solution as a chromophore for laser absorption, prior to adding BSA (62% ± 2% w/w) [44]. The final genipin and IG concentration were 0.38% ± 0.03% and 0.25% ± 0.02% (w/w) respectively. A typical preparation procedure consisted in placing ~0.2 g of BSA powder in an eppendorf tube (1.5 ml) containing ~0.3 ml of genipin/IG solution. The components were then mixed and compressed with a metallic rod, designed to fit the bottom of the eppendorf tube. The application of this rod step to the procedure allowed a thorough compression and homogenisation of the BSA and genipin solution. The resulting protein mixture (solder) was solid and malleable until dried, at which point became rigid. Before dehydration, the solder was reduced to a fixed thickness film with a vice, the parallel plates of which were coated with plastic (polyethylene). Rectangular parts of the protein film were cut with a surgical blade and weighed by a digital balance (s.d. = ± 0.1 mg). These protein strips were prepared to a fixed weight (2.1 ± 0.2 mg) and thickness (0.15 ± 0.01 mm) to standardize the solder dimensions (table 1). The solder thickness was measured with a digital caliper (s.d. = ± 0.01 mm). Protein strips were stored at 5 0C and used after five days. BSA solder controls in the absence of genipin were similarly prepared. 3.2.3 Tissue Soldering The diode laser was used in conjunction with the protein strips to weld rectangular sections of sheep small intestine (area: ~1 x 0.5 cm, thickness: ~1mm) under an operative microscope (X 20). Fresh intestinal tissue was harvested from sheep immediately after euthanasia and then frozen for ~2 months at –80 0C. Prior to use, tissue was immersed for 15 min in deionized water to defrost and rehydrate at room temperature. 77 The serosa layer of intestine consists of connective tissue that is reach of collagen and therefore suitable for laser tissue repair. Intestine is abundant in the sheep body and a single animal can provide enough tissue for several trials. Two cm long rectangular sections were bisected by a full thickness incision with a #10 blade. The stumps were then approximated and joined with the irradiated solders. In more detail, a solder strip was positioned across the incision on the serosa layer with microforceps, to ensure full contact with the intestine, and it was welded to the tissue. Thereupon, the laser irradiated continuously all over the strip by moving continuously the beam across its surface at a speed of ~1 mm/s with a spot size of 300-400 mm, for the prescribed time interval. Moderate tissue shrinkage under the solder was observed during laser irradiation. Two strips were soldered to complete the intestine repair as shown in figure 3. The intestine was kept moist and excess water was carefully absorbed with gauze prior to laser irradiation to avoid dilution of the solder. All tissue repairs were done using fixed laser power, radiation dose, solder surface area and thickness (table 1). The dose was similar to that previously used to successfully repair peripheral nerves in vivo [55]. 78 LASER FIBER INTESTINE LASER BEAM SECTION TISSUE INCISION SOLDER STRIPS Figure 3. Schematic of laser tissue-soldering. A strip of albumin is applied across the incision and irradiated afterwards by the laser. Two strips are applied to complete the intestine repair. 3.2.4 Tensile Strength The soldered intestine was tested ten minutes after LTR using a calibrated tensiometer (Instron Mini 5543, MA, USA) to assess the tensile strength of the repairs. Tissue was kept in wet gauze after laser irradiation to mimic in vivo conditions and avoid desiccation of the sample. A specimen was clamped to the tensiometer using pneumatic grips, which moved 22 mm/min until the solder weld failed, separating the two intestine segments. The maximum load under which the weld failed was recorded. 3.2.5 Solder Attenuation A spectrophotometer measured the laser attenuation at both 808 nm and 580 nm within the solder films to observe any possible change in the attenuation characteristics of the solder due to the presence or absence of genipin. The first wavelength (808 nm) corresponds to the absorption peak of IG and to the laser radiation used for LTS; the 79 second wavelength (580 nm) is the typical absorption peak of the blue pigment formed by genipin after crosslinking with amino groups [140]. Solder films were fixed inside a quartz cuvette and adhered to the wall facing the light beam. The attenuation length (1/e -A attenuation) was calculated by assuming the validity of Beer’s law: I=I0e , where I0 is the incident beam intensity and A is the film attenuation. The attenuation measurements were performed with baseline subtraction, five days after the solder was prepared to replicate the LTS conditions. 3.2.6 Cytotoxic Assay A cell growth inhibition assay was used to assess the cytotoxic potential of BSA solder with and without genipin and also with and without thermal treatment. Solutions of Genipin were also tested to establish its baseline cytotoxicity levels. Inhibition of cell growth under these samples was compared with unperturbed samples and positive controls containing 4%, 5% and 7.5% ethanol [147,148]. Murine L929 cells (Earle's L Cells - NCTC Clone 929) were established at 105 cells in 2 o 35 mm petri dishes. They were incubated at 37 c for 24 hours in a 2%CO2 humidified atmosphere using Earle’s minimum essential medium supplemented with 10% Foetal Bovine Serum (EMEM). After 24 hours, the sub-confluent cell monolayer was established and the medium was aspirated from the petri dishes. The medium was replaced with a medium extract of the BSA genipin solders and solutions. The cell monolayers were then cultured for a further 48-hour period. During this time any cytotoxic components emanating from the test materials would have disrupted the growth of cells numbers in the culture dish. At the end of the test period, cells were 80 harvested from the dishes, their numbers assessed by a coulter counter and compared with unperturbed cultures (null samples). Differences in cell numbers were expressed as a percentage inhibition in comparison to null samples. The standard deviation of the percentage inhibition was calculated by propagating the error with the formula: 2 2 2 2 2 s (Z) = [¶ N(Z)] *s (N)+ [¶ S(Z)] *s (S) where s=standard deviation, ¶N= derivative respect to N, ¶S= derivative respect to S, Z= 100*(N-S)/N percentage inhibition, S=number of growth cells in the experimental groups, N=number of growth cells in the null group. Three replicas of BSA solder with and without genipin were prepared using non- pyrogenic water with no IG dye included in the solder. All solder samples were g- radiated in sealed sterifilm® at a level of ~2.4 KGy to ensure sterile conditions (0.1 KGy/h for 24 hours) [149]. Heated samples underwent thermal treatment at 70 0C for five minutes in order to denature albumin and crosslink genipin with the amino groups. Pellets of solder £5 mm in diameter and approximately 0.55 g, were extracted in 2.5 ml medium (EMEM) in glass vials for 24h at 37oC. The extract was removed and placed directly on the cell monolayer for 48 hours, as prescribed by the international standards (ISO/DIS 10993-12.2, 1996). A filter sterilised stock solution of 0.1%w/v genipin was prepared using non-pyrogenic water. This genipin stock was diluted in EMEM at concentrations from 0.0125% to 0.00078% w/v to assess the cytotoxic level of genipin when it is not bonded to the solder amino groups. 81 Statistical comparison of means was made using ANOVA one-way and the Student’s t test for unpaired observations at 0.05 level of significance, Excel 2000 generated the histograms. Mean and standard deviation were calculated from three counts, performed on three samples for each group (n=9). 3.3 Results 3.3.1 Tensile Strength The BSA-genipin solder doubled the strength of tissue repair (0.21 ± 0.4 N) when compared with the BSA tissue welds (0.11 ± 0.4 N, p=5.5 x10-15, unpaired t-test, N=30). The welded strips either failed cohesively (broken strips) or at the tissue interface (detached strip); in particular, the BSA-genipin solder suffered 55% of interface failure while the BSA solder failed at the tissue interface 72% of the times. The tensile test results are summarized in figure 4 and table 1. The increased temperature, produced by the laser absorption, most likely accelerated the speed of reaction between genipin and BSA molecules and change in solder colour, from green to blue-brown [140]. The temperature at tissue interface was estimated to be around 65 0C, as measured in a previous study [150]. 82 0.3 55% Interface Failure 72% Interface 0.25 Failure 0.2 * BSA Solder 0.15 BSA-Genipin Solder -15 0.1 * P~10 , N=30 Maximum Load (N) 0.05 0 Solder Type Figure 4. Histogram of acute tensile strength of laser repaired tissue with BSA solder and BSA-genipin solder (N=30). Table 1. The laser parameters and solder characteristics (mean ± SD) are given for the tissue soldering of rat intestine. N, number of laser-solder repairs; Thickness, solder thickness; Area, averaged solder surface area in contact to the intestine during laser welding; Power, laser power during soldering; Time, laser irradiation time; Dose, averaged laser dose used for soldering. Strength, maximum load in Newtons under which the tissue welds failed. Solder Type Thickness Area Power Time Dose Strength (mm) (mm2) (W) (s) (J/mg) (N) BSA 0.15 ± 0.01 12.8 ± 1.1 0.17 ± 0.01 139 ± 10 10.8 ± 0.4 0.11 ± 0.4 (N=30) BSA-Genipin 0.15 ± 0.01 12.6 ± 1.0 0.17 ± 0.01 137 ± 10 10.9 ± 0.8 0.21 ± 0.4 (N=30) 3.3.2 Solder Attenuation The attenuation of the BSA-Genipin solder at 808 nm was not significantly different than the attenuation of BSA solder (0.017 ± 0.006 mm vs. 0.019 ± 0.004 mm, p=0.31, t- test); therefore we can infer that genipin did not change the laser absorption of IG in the 83 solder. Similarly, BSA-genipin solder did not show an increased attenuation at 580 nm when compared with BSA-solder containing no genipin (0.045 ± 0.006 mm vs. 0.047 ± 0.008 mm, p=0.34, t-test). It is likely that genipin reacted with lysin, arginine, glutamine and other amino acids present in albumin even at low temperature (4 0C); nevertheless, the rate of reaction was considerably slowed down and no significant change in the attenuation was detected in the solder five days after its preparation. The solder without genipin and IG attenuated light at 580 and 808 nm in a comparable way. The roughness of the solder surface, due to the plastic coating on the vice plates, and BSA molecules were probably responsible for scattering light and therefore attenuating radiation through the solder. The complete results are summarized in table 2. 3.3.3 Cytotoxicity Assay Medium extracts of BSA solder heat-treated and non-heat treated caused a negligible cell growth inhibition (<10%, n=9, figure 5). The BSA genipin solder that was heat- treated showed also a negligible cytotoxicity level of 10.8% (n=9). The BSA genipin solder that was not heated had a cytotoxic impact with a mean inhibition of 31.5% (n=9), which was comparable to cytotoxic level induced by the 5% ethanol solution (figure 6). When genipin is not bonded to the solder amino groups it shows a clear cytotoxic inhibition of 60.0% at a concentration of ~0.003%w/v (figure 7). 84 Table 2. Attenuation data of the solder at a wavelength (l) of 808 and 580 nm; the mean value ± the standard deviation are given. N, number of solder samples analysed. Attenuation Length, 1/e attenuation of the solder sample, as calculated from Beer’s law. Thickness, thickness of the solder sample. Solder l Solder Thickness Attenuation Length Type (nm) (mm) (mm) BSA+IG+Genipin 808 0.07 ± 0.02 0.017 ± 0.006 (N=5) BSA+IG 808 0.08 ± 0.01 0.019 ± 0.004 (N=5) BSA 808 0.10 ± 0.01 0.223 ± 0.069 (N=5) BSA+IG+Genipin 580 0.07 ± 0.02 0.045 ± 0.006 (N=5) BSA+IG 580 0.08 ± 0.01 0.047 ± 0.009 (N=5) BSA 580 0.10 ± 0.01 0.190 ± 0.050 (N=5) 3.4 Discussion Genipin proved to be effective in increasing the weld strength of tissue repairs. In particular, the modality of the bond failure demonstrated that genipin enhanced the strength at the tissue interface: the soldered strips detached from tissue without breaking in 55% of the cases at a load twice as high as the failure load for the BSA solder. The latter had a failure rate of 72% at the tissue interface (figure 3). The bonding mechanism of the BSA-genipin solder may be the synergistic action of mechanical adhesion with chemical crosslinking by genipin. The intermolecular crosslinking of BSA may have enhanced the strength at the tissue interface by decreasing the solubility of the solder and therefore stabilizing the bonds. The darkening of the solder during laser irradiation supports this hypothesis. The colour change observed in Genipin solders was most 85 likely due to the well known observation that genipin reacts with primary amines (NH2), abundant in proteins, to produce polymeric blue pigments [5]. BSA SOLDER 100 80 60 40 20 CELL INHIBITION (%) 0 -20 4% Ethanol 5% Ethanol 7.5% Ethanol BSA Heated#1BSA Heated#2BSA Heated#3 BSA Non heated#1BSA Non heated#2BSA Non heated#3 Figure 5. Percent inhibition growth of fibroblasts in media extracted from BSA solder and other controls when compared to the null group (Y-axes). Legend. Ethanol, samples of cells incubated in media+ ethanol at various concentrations (positive controls). Heated, samples of cells incubated with media extracted from heated solder. Non-heated, samples of cells incubated with media extracted from non-heated solder. Mean and standard deviation were calculated from three counts, performed on three samples for each group (n=9). An Albumin molecule contains one amino terminal and several amino acid residues in its side chain, with a NH2 group that reacts with genipin. These residues are lysin (n=60), asparagine (n=14), arginine (n=26), and glutamine (n=58). The number of residues per BSA molecule is indicated in brackets. 86 BSA-GENIPIN SOLDER 90 80 70 60 50 40 30 20 CELL INHIBITION (%) 10 0 Heated#1 Heated#2 Heated#3 4% Ethanol 5% Ethanol 7.5% Ethanol Non heated#1Non heated#2Non heated#3 Figure 6. Inhibition percentages of fibroblast growth (Y-axis) for ethanol controls and medium extracts from BSA-genipin solder when compared to the null samples (X-axis). Extract of heated BSA-genipin solder is not cytotoxic with an inhibition of 10.8%. Non- heated BSA-genipin solder had cytotoxic effects on cells with an inhibition of 31.6%. Legend. See figure 4. Genipin may have also crosslinked the solder BSA with the collagen in the tissue through their amino groups, although there was no macroscopic evidence of this. No change in colour was observed in the BSA solder during tissue welding, with the exception of the solder top, which became brown as the IG absorbed the infrared light. Genipin did not significantly change the attenuation characteristics of the solder and therefore the photon absorption of IG, which could have been ascribed to increasing the tensile strength of the welds. It should be stressed that the spectrophotometric measures did not provide an absolute measure of the solder absorption but merely quantified the solder attenuation, which is due to reflection, scattering and absorption of the incident light. We assumed that adding IG or genipin could only change the absorption and not scattering or reflection of the solder. 87 GENIPIN SOLUTION 90 80 70 60 50 40 30 20 10 CELL INHIBITION (%) 0 4% Ethanol5% Ethanol 7.5% Ethanol 0% Genipin 0.0125% 0.0063% Genipin 0.0031% Genipin 0.0016%Genipin 0.0008% Genipin Genipin Figure 7. Cell growth inhibition percentages of fibroblasts (Y-axis) of ethanol controls and genipin solutions in medium (X-axis) when compared to the null. Genipin at 0.0031% w/v gives a 60.0% cytotoxic inhibition. The inhibition levels fall to ~24% at a concentration of 0.0016% and to 9.9% inhibition at 0.0008% genipin. Four percent ethanol inhibits cell growth by 33.4% (positive control). Legend. Ethanol, samples of cells exposed to media+ ethanol at various concentrations (v/v). Genipin, concentrations of genipin solutions w/v in medium. The effect of increasing the weld strength, observed in the solid solder, due to the genipin crosslinking ability, should also occur in fluid albumin solders. It has been previously shown that a solid protein strip dissolves a few microns of its thickness when in contact with tissue moisture and therefore is equivalent to a concentrated fluid solder at the tissue interface [52]. In this report, the solid solder is preferred over its fluid counterpart since it allows control of the solder thickness and surface area with much higher precision. The fluid solder can have a thickness variability larger than 100% and its surface area at the operating site can fluctuate even if a fixed amount of solder is 88 applied for each weld [51]. Both these factors play a key role in the tensile strength outcome of the repaired site and must be controlled for a reliable statistical comparison. No more than 1% of genipin could be diluted in water limiting the final genipin concentration of the solder to 0.38 %. This level of genipin concentration proved to efficiently crosslink tissue in previous studies [142, 151]. Some researchers have reported that genipin reacts slowly in phosphate buffer solution (pH~7.4) with 98% of primary amines after three days, when the genipin concentration ranges between 0.25% and 1% w/v at 25 0C [142]. The slow reactivity of genipin with protein amino groups at neutral pH has the potentiality to increase the bond strength of the solder repairs in the body at 37 0C during the first days after the operation. This would prove beneficial since previous reports have showed that the tensile strength of albumin laser welds decreased post- operatively for the first three days [152]. The hypothesis that genipin may compensate for the drop of acute tensile strength of the solder welds needs to be confirmed with further experiments in animal models. Long-term storage of BSA-genipin compounds appears also problematic, as crosslinking of albumin molecules is a continuing slow process that could compromise the welding properties of solders. The in vitro and in vivo toxicity of crosslinked genipin has been reported to be minimal and significantly lower than the toxic level of more popular crosslinkers, such as glutaraldehyde [153-156]. These reports are consistent with the in vitro cytotoxicity study conducted in this investigation, as media extracted from BSA-genipin solders that underwent thermal treatment did not inhibit significantly the growth of murine 89 fibroblasts. When the BSA-genipin solder was not heated and not crosslinked to albumin, it induced a significant level of cell growth inhibition (31.5 %). This toxic effect is likely due to genipin leaking out of the solder and reacting with cell proteins. This hypothesis was supported by the cytotoxic impact of low concentrations of genipin (~0.003% w/v) in solution that inhibited ~60% of fibroblasts growth. The non-toxic nature of bonded genipin strongly suggests this crosslinker as a suitable enhancing factor for future albumin glues, which may become to surgeons more appealing than other glues currently used for tissue repair. Fibrin glue, for example, is the most popular tissue sealant, mainly due to its non-toxicity and biodegradability although it lacks high tensile strength and burst pressure [157]. 90 Chapter IV Chitosan Adhesive for Laser Tissue Repair 4.1 Introduction Laser-activated solders for tissue repair are traditionally based on three proteins: albumin, fibrinogen and collagen [158-160, 38, 13]. The first two are blood-derived and therefore have an intrinsic, although limited, risk of viral infection for the host organism [161, 162]. Albumin solders are the most investigated and promising candidates for specific procedures in urology, vascular surgery and microsurgery due to their bonding characteristics [48, 56, 55]. Nevertheless, the use of albumin solders, especially in the solid form, is limited by their water solubility, lack of flexibility and brittleness before being irradiated by lasers. These disadvantages, along with the potential thermal damage to tissue caused by the laser, may compromise the safe and reliable application of solders in surgical procedures [163, 33]. It is not surprising therefore, that clinicians and surgeons are reluctant to routinely employ albumin solders for tissue repair and wound closure. In chapter III, the addition of genipin to albumin based solders resulted in enhanced bonding strength. Despite this, the unique albumin/genipin solder remained soluble and brittle [164]. Consequently, alternative biopolymers have investigated in the design of a new tissue adhesive to overcome shortcomings of currently used solders. 91 Chitosan is a polysaccharide derived from deacetylated chitin and can be readily solvent cast in a film that has excellent mechanical properties and low toxicity (figure 1) [119, 131, 77]. Figure 1. Chemical structure of chitosan that is derived from deacetylated chitin. Chitosan may also bind to collagen as demonstrated by previous studies, which reported polyanionic-polycationic and hydrogen bonding interaction between these two biopolymers [165, 166]. Taravel and Domard showed that a pure polyanion-polycation complex is formed between the protein carboxylate group (COO-) and the chitosan + -2 amino group (NH3 ) when a solution of chitosan hydrochloride (5*10 M) is added to a solution of collagen (1.7 g/l) with all its ionic sites fully dissociated (pH ~7.8). Such complex was insoluble in water and the polyanionic-polycationic interaction improved if the collagen was denatured by heat at 60 0C. Further, a second mechanism of interaction occurred between chitosan and collagen when the polysaccharide exceeded the protein amount. FTIR analysis suggested hydrogen bonds formed between collagen 92 molecules and chitosan amino groups. These complexes were also insoluble in water and circular dichroism spectra suggested that collagen was denatured. In an attempt to improve bonding strength, Ono et al. modified chitosan with lactobionic acid and p-azidebenzoic acid that was crosslinked with ultra-violet (UV) light to repair holes in pig aortas [167]. The authors reported a bursting pressure of ~30 KPa, significantly higher than the bursting pressure of specimens sealed with fibrin glue (~11 KPa). Histologic examinations of rabbit carotid arteries showed that 30 days after gel application in vivo, a fraction of the chitosan gel was phagocytosed by macrophages, partially degraded, and induced the formation of fibrous tissues around the gel. This chitosan gel was unfortunately water soluble prior to light crosslinking and therefore subject to fluid dilution like albumin solders. In the present investigation a chitosan-based film has been developed without UV- crosslinker modifications. The novel film is insoluble, flexible and adheres firmly to tissue upon infrared laser activation. 4.2 Materials and Methods 4.2.1 Chitosan Adhesive Films In the first experimental group, deacetylated chitosan (³ 85%) from crab shells (Sigma, St. Louis, MO, USA) was dissolved to a concentration of 2% w/v in a water solution containing acetic acid (2% v/v) and indocyanine green (IG, 0.02% w/v). In the second experimental group, the chitosan solution of group I was prepared without the dye Indocyanine Green (IG). 93 In the third experimental group, an ethanol solution of genipin (10% w/v) was added to the green chitosan gel of the first group to obtain final concentrations of 1 % (w/v) genipin and 0.7% (v/v) ethanol. In the fourth experimental group, a genipin-chitosan solution was prepared as in group III but without IG dye. All gelatinous chitosan solutions (pH ~4.0) were stirred for 6 hours at 4 0C before spreading evenly (thickness ~2 mm, surface area ~12 cm2) over a sterile and dry, perspex plate. The chitosan solutions were then dried for ~6 days under clean conditions and atmospheric pressure at 4 0C. The resulting chitosan films were carefully detached from the plate without damage and were insoluble in water. A digital caliper was used to measure the adhesive thickness, which ranged from 15 to 30 µm. All films were thereafter cut in rectangular strips (~7.5 x 4.5 mm), placed between sterile glass slides to preserve their flat shape and stored in the dark at 4 0C. Genipin was added to the chitosan solution in order to explore a possible enhancement of film adhesion to tissue, as previously demonstrated for albumin solders. The final concentrations of genipin (1%) and ethanol (~0.7%) were suggested by previous studies [143]. Further, 0.7% of ethanol likely induces negligible cytotoxicity, as shown in chapter 3. 4.2.2 Adhesive Attenuation A UV-Visible spectrophotometer was used to measure the laser attenuation at 808 nm within the films and to observe the attenuation characteristics of the adhesive due to the presence of IG and genipin. The wavelength of 808 nm corresponds to the absorption 94 peak of IG and to the laser radiation used for laser tissue repair [44]. Adhesive films were fixed inside a plastic cuvette and placed in the light beam, which scanned the adhesives at the wavelength range of 400-890 nm. The attenuation length (1/e -A attenuation) was calculated by assuming the validity of Beer’s law: I=I0e , where I0 is the incident beam intensity and A is the film attenuation. The attenuation measurements were performed with baseline subtraction. 4.2.3 Laser Tissue Repair (LTR) Tissue repair was investigated by using a GaAlAs diode laser (Qphotonics, L.L.C., VA, USA), coupled with a multimode optical fiber through an FC connector. The fiberoptic cable was inserted in a hand-held probe to provide easy and precise beam delivery by the operator. The laser emitted at 808 nm, with a fiber core diameter of 200 mm, numerical aperture 0.22 and a beam spot size on the adhesive of approximately 1 mm. Because the laser is not eye safe (Class IV), safety goggles were worn by all staff in the operating theatre. The diode laser was used to irradiate the chitosan-based adhesive strips to repair rectangular sections of sheep small intestine (~2x1 cm) under an operative microscope (X 20). Fresh intestinal tissue was harvested from sheep immediately after euthanasia and stored at –80 0C. Prior to use, tissue was immersed in deionized water for 15 minutes to defrost and hydrate at room temperature. The serosa layer of intestine consists of connective tissue that is rich in collagen and therefore suitable for testing the tensile strength of laser adhesive repairs. Intestine is abundant in the sheep body and a single animal can provide enough tissue for several trials. 95 Intestine sections were bisected by a full thickness incision with a #10 blade. The intestine was kept moist using deionized water; excess water was absorbed with sterile gauze or cotton tips prior to tissue repair. The incision stumps were approximated end- to-end and a chitosan strip was positioned across the incision on the serosa layer with microforceps ensuring full contact with the intestine. Thereupon, the operator irradiated the adhesive by moving continuously the beam across its surface at a speed of ~1 mm/s and without charring or ablating the adhesive (figure 2). LASER FIBER LASER BEAM INTESTINE SECTION TISSUE INCISION CHITOSAN STRIP Figure 2. Schematic top-view of laser tissue repairing. A strip of chitosan adhesive is applied across the incision then subsequently irradiated using a laser. A fluence of approximately 52 J/cm2 was required to irradiate three times the adhesive strips at a speed of ~ 1 mm/s and a power level of 120 mW. A similar fluence was also chosen for the other experimental groups. Moderate tissue shrinkage under the adhesive was observed during laser irradiation. Variable power and therefore irradiance was used in the first part of the experiment, to choose suitable parameters for tissue repair (table 1). In the second part of the experiment, the chitosan strips were tested using the 96 selected power and irradiance (table 2). Control tissue repairs were also carried out, applying chitosan strips without the use of laser radiation. Table 1. The laser parameters and adhesive characteristics (mean ± SD) are given for the repairing of sheep intestine. The chitosan strips had a thickness of 21 ± 2 mm. Legend. Power, laser power during tissue repair; Area, averaged adhesive surface area in contact to the intestine during laser repair; Time, laser irradiation time; Shear Stress, maximum load divided by the surface area of the chitosan adhesive. Power Area Time Irradiance Fluence Shear Stress (W) (mm2) (s) (W/cm2) (J/cm2) (KPa) 0.16 ± 0.01 22 ± 2 75 ± 11 ~20 54 ± 4 7.9 ± 2.8 0.12 ± 0.01 23 ± 2 99± 6 ~15 53 ± 1 13.2 ± 3.9 0.08 ± 0.01 22 ± 2 140 ± 11 ~10 51 ± 2 10.5 ± 4.2 0 34 ± 4 0 0 0 1.3 ± 0.7 4.2.4 Tensiometer Measurements The intestine was tested 10 minutes after tissue repair with calibrated tensiometer (Instron Mini 55, MA, USA) to assess the tensile strength of the repaired wound. Tissue was maintained in wet gauze after being repaired to mimic in vivo conditions and avoid sample desiccation. A specimen was clamped to the tensiometer using pneumatic grips, separating at a rate of 22 mm/min until the adhesive failed. The maximum load under which the adhesive failed was recorded by Merlin IX software. 97 Based on the tensiometer results, further physical, chemical and biological characterization of the strongest chitosan adhesive (among the four groups) was carried out as described in the following sections. Table 2. The laser parameters and adhesive characteristics (mean ± SD) are given for the repairing of rat intestine. The chitosan strips had a thickness of 20 ± 5 mm. Adhesive Group Area Power Time Fluence Maximum Shear Stress (mm2) (W) (s) (J/cm2) Load (N) (KPa) (N=30) Group 1+Laser 34 ± 4 0.12 ± 0.01 147 ± 7 52 ± 2 0.50 ± 0.15 14.7 ± 4.3 Group 1 34 ± 4 0 0 0 0.07 ± 0.04 1.9 ± 1.3 Group 3+Laser 34 ± 4 0.12 ± 0.01 146 ± 4 51 ± 1 0.31 ± 0.10 9.1 ± 2.9 Group 3 34 ± 4 0 0 0 0.02 ± 0.01 0.6 ± 0.4 4.2.5 13C-NMR Solid-state 13C-NMR spectra were acquired using a Varian Inova-300 spectrometer operating at 75.45 MHz with Chemagnetics 7.5 mm double air-bearing cross- polarization (CP) probe. Samples of chitosan shells and white chitosan adhesive films (group II) (~200 mg) were packed as strips and pieces into 7.5 mm od rotors made from partially-stabilized zirconia and subjected to “magic-angle spinning” at 2-3 kHz. IG dye was not included in the analysed adhesive to avoid possible signal interference in the spectrum. Spectra were acquired at 294 K using single-contact cross-polarization experiments with high-power 1H decoupling during acquisition. The following parameters were found optimal: pulse width 5.2 µs (90º); contact time, 1 ms; recycle 98 time, 5 s. Free induction decays were acquired and zero-filled to 8K prior to Fourier Transformation. Up to 10,000 scans were collected for sufficient signal/noise. The Hartman-Hahn match was set using hexamethylbenzene, which was also used as a secondary external reference that gives the methyl peak, dC = 17.3 ppm on the tetramethylsilane scale (13C TMS = 0 ppm). 4.2.6 Thermogravimetric Analysis (TGA) Chitosan adhesive films were analysed using a Perkin Elmer Pyris 1 Thermogravimetric Analyser to evaluate the water content and degradation temperature of the films in air (weight 10-15 mg). The temperature was increased from 20 to 600 0C at a rate of 40 0C/min. The mass of the films was continuously recorded as a function of temperature and the first derivative calculated to assess the adhesive degradation (derivative minimum peak) and water content (derivative=0). 4.2.7 Differential Scanning Calorimetry (DSC) Thermograms of chitosan adhesives were obtained and recorded by using a differential scanning calorimeter (DSC 7, Perkin-Elmer, USA). The samples (10–15 mg) were accurately weighed into solid aluminum pans and sealed. Heating rates of 10°C/min, 30 °C/min and temperature ranges of 20–110 0C (n=14) or 25-220 0C (n=4) were selected for scanning under air with a flow rate of 20 mL/min. The scans were repeated twice at a cooling rate of 30 0C/min to thermally stabilize the system and eliminate possible thermogram artifacts. The temperature range was not extended further to avoid heavy degradation of samples and allow the detection of a possible melting transition. The software Pyris analyzed data and calculated the temperature of transition points. 99 4.2.8 Contact Angle The static contact angle of chitosan adhesives was measured by the sessile drop method using the RHI system (model 100-00-230, New Jersey, USA). The films were positioned on a translating stage and a drop of water poured on them. A CCD camera recorded the image of the drop on films provided by a telescope with magnification of 23; the system software measured the contact angles. Six readings from different parts of the film surface were averaged to give the mean contact angle. 4.2.9 Atomic Force Microscopy (AFM) The surface topography of chitosan adhesives was imaged by using a commercial atomic force microscopy (Digital Instruments Dimension 3000, Urbana, Illinois, US). The microscope was operated in tapping mode using silicon cantilevers with a nominal spring constant of 40 N/m and oscillating with average amplitude of 100 nm and a resonance frequency between 200 and 450 kHz. The scanning rate was automatically selected. Surface roughness was calculated as the Z RMS value: 2 ½ Rq = [Si( Zi-Zave) /N] Where Zave is the average Z value within the scanned area, Zi is the current Z value and N is the number of points in the scanned area. Nine measurements of this surface parameter were performed on three separate films; Rq mean and standard deviation were also calculated. 4.2.10 Young’s Modulus Rectangular strips of chitosan adhesives from groups I and II were also tested alone by the tensiometer to measure their Young’s modulus (E) without laser activation. The strips (N=15) were wet with water and fixed to the grips, which moved apart until the 100 repair failed. The software generated a strain-stress plot and calculated E in the strain region between 10% and 20%, assuming the strip thickness remained constant throughout elongation. 4.2.11 Temperature Measurements The temperature of LTR was measured with an insulated K-type thermocouple (diameter = 0.25 mm, response time = 0.1 s) positioned between the intestine and adhesive as illustrated in figure 3. Chitosan Adhesive Thermocouple Serosa Tissue 1 mm Figure 3. Photograph showing a strip of chitosan adhesive bonded to the serosa layer of intestine. A thermocouple was placed between tissue and adhesive to record the temperature during LTR. The thermocouple was connected to a signal-conditioning unit (SCXI-1121) with 4Hz low pass analogue filter. Data were sampled at 10 Hz with a 12-bit data acquisition board (PCI-6024E) controlled by LabView® (National Instruments, Austin, TX, U.S.) The strips were irradiated, as described previously, by moving continuously the beam across the adhesive surface; temperature data were recorded for 20 s while the beam 101 was directed on the thermocouple or in its proximity. After LTR, the adhesive was detached using microforceps to ensure that tissue adhesion occurred. 4.2.12 Ex-Vivo Histology and Scanning Electron Microscopy (SEM) Intestine was harvested immediately after sacrificing sheep and stored in phosphate buffer solution at 4 0C for half an hour. Tissue sections (n=4) were thereupon repaired as explained previously using the same parameters listed in table II. They were stored in 10% buffered formalin and fixed for Hematoxylin and Eosin (H&E) staining. The purpose of this study was to observe any acute thermal damage such as ablation, charring or coagulation. Other repaired tissue was prepared for SEM analysis (n=4). Chitosan repaired intestine was cut with a scalpel to small pieces (~1x2 mm) and fixed in Karnovsky’s solution (2.5% paraformaldehyde and 2% glutaraldehyde in 0.1 M phosphate buffer) after being gently washed in 0.1 M phosphate buffer and dehydrated in alcohol at several dilutions (30%, 50% and 70% v/v). Specimens were further dehydrated to critical point and sputter coated with Cu before being viewed at 10 KV by the microscope in high vacuum mode (FEI Quanta 200, Hillsboro, Oregon, US). 4.2.13 Cytotoxic Assay A cell growth inhibition assay was used to assess the cytotoxic potential of the chitosan adhesive films. Inhibition of cell growth in contact with extracts from these samples was compared with untreated medium and positive controls containing 7.5% ethanol. Murine L929 cells (Earle's L Cells - NCTC Clone 929) were established at ~105 cells per 35 mm diameter petri dishes. Cells were incubated at 37 oC for 24 hours in a 5% 102 CO2 humidified atmosphere using Earle’s minimum essential medium (EMEM) supplemented with 10% Foetal Bovine Serum. After 24 hours, a sub-confluent cell monolayer was established and the medium was aspirated from the petri dishes. The medium was replaced with either a medium extract of the chitosan adhesives (group II) or with the controls. All solder samples were g-radiated in sealed sterifilm® at a level of ~2.4 KGy to ensure sterile conditions (0.1 KGy/h for 24 hours). Chitosan films without IG (group II) and with surface area of ~18 cm2 were placed in 3 ml medium (EMEM) in glass vials for 24h at 37oC. The extract was removed and placed directly on the cell monolayer for 48 hours, as prescribed by the international standards (ISO/DIS 10993-12.2, 1996). During this time any cytotoxic components emanating from the test materials would have disrupted the growth of cells in the culture dish. At the end of the test period, cells were harvested, their numbers assessed through Flow Cytometry (FACS, Becton Dickinson, USA) and compared with untreated cultures (blank and null samples). No IG was added to the chitosan strips to avoid possible interference with the fluorescence signal of the assay. Assessment via Flow Cytometry was facilitated by the addition of Propidium Iodide (10µg/ml), which stains non-viable cells with a disrupted membrane. Cells were harvested into a known volume and a known number of ~10 µm diameter polystyrene beads was added to the cell suspension. The number of cells in suspension was back calculated by comparing the ratio of cells to beads acquired by flow cytometry. Viability staining, forward scatter and side scatter evaluation by Flow Cytometry provided a more sensitive indication of material toxicity than inhibition alone. 103 4.2.14 In Vivo Thermal Damage The thermal damage induced by the laser activation of chitosan adhesive was assessed in vivo by irradiating chitosan strips on rat sciatic nerves. This animal model proved in previous studies to be a reliable test for the safety and efficacy of laser-activated albumin solders, as myelinated axons of peripheral nerves are sensitive to thermal damage [168]. Six adult male Wistar rats, weighing approximately 600 g, were used in this preliminary study. Approval and consent for the study were obtained from our Institution's Research and Animal Ethical Review Committees. Anaesthesia was induced and maintained during surgery with Halothane/O2 mix (4% during induction, 2% thereafter) using a standard anaesthetic machine. The surgical operation was performed using a Zeiss OPMI 7 operating microscope, and the operative procedure was performed under sterile conditions. An oblique skin incision of about 3 cm was made in the dorso-lateral part of the right thigh. The muscles were gently retracted and the sciatic nerves (dimeter ~1 mm) exposed at the sciatic notch. The adventitia of the sciatic nerve was partially peeled and trimmed with straight micro- scissors before applying the adhesive; excess water was absorbed with sterile gauze or cotton tips. A chitosan strip (group I) with a thickness of ~20 mm and surface dimensions of ~ 6 x 5 mm was then positioned underneath the sciatic nerve using microforceps. The chitosan strip adhered fully to the nerve like a collar and assisted with rotation of the nerve, during the procedure (figure 4). The laser beam was passed over the length of the strip three times to activate its adhesion to the epineurium, as described in the previous LTR section. The output power, fluence and beam spot size were 0.12 ± 0.01 W, 65 ± 11 J/cm2 and ~1 mm respectively. During laser irradiation, the nerve perineurium was observed to shrink moderately. Rotation of the nerve was 104 obtained by gently moving the chitosan collar, so that the chitosan strip could be irradiated around the nerve (figure 4). The redundant chitosan was trimmed from the collar before closing the muscles and the skin with five 3-0 sutures. The animals were thereupon placed in individual cages with no restriction of movement. Sections of the operated nerves were harvested at the operated sites, proximally and distally, four days after surgery. They were stored in 10% buffered formalin and fixed with Luxol Fast Blue and H&E staining to assess myelinated axons and tissue thermal damage. Sciatic nerves were also harvested from the left thigh to serve as controls. At the end of the procedure the animals were sacrificed by an intracardiac injection of 2 ml Sodium Pentobarbital. Three sections (~2cm) of sciatic nerves were also used for laser repair and measurement of shear stress. Each nerve section was cut in two equal parts by a microscissor, a chitosan strip was then positioned underneath the severed nerve and the stumps aligned end-to-end with micro-forceps. The chitosan strip adhered fully to the severed nerve like a collar and assisted with rotation of the nerve during the laser repair, as detailed above. The output power, fluence and beam spot size were 0.12 ± 0.01 W, 46 ± 2 J/cm2 and ~1 mm respectively. The shear stress of the adhesive bonded to the nerve was tested by the Instron tensiometer as described before. 4.2.15 Statistical Analysis Statistical comparison of means was made using the two-tails unpaired Student’s t-test, ANOVA one-way and Bonferroni’s multiple comparison test at 0.05 level of significance. Excel 2000 generated the graphs and histograms. 105 Adhesive A B Laser Sciatic Nerve Figure 4. The adhesive is firstly placed underneath the sciatic nerve (a). The chitosan strip adheres to the nerve like a collar before the laser irradiation takes place (b). 4.3 Results 4.3.1 Adhesive Attenuation The attenuation length of the chitosan adhesive with and without IG (groups I and II) at 808 nm were respectively 5 ± 1 mm and 228 ± 72 mm. The presence of genipin in the adhesive (group III) did not change significantly the attenuation length (6 ±1 mm, n=3, p=0.38 t-test). The attenuation length at 608 nm of the bluish chitosan adhesive with genipin (group IV) was 6 ±2 mm; the colour change of the chitosan strips and the peak absorption signalled crosslinking between genipin and chitosan amino groups. The presence of IG in the adhesive (group III) did not change significantly the attenuation length (7±1 mm, n=3, p=0.54 t-test). Assuming insignificant scattering and reflection, we may ascribe to IG the efficient absorption of the laser energy at 808 nm inside the chitosan adhesive, 106 independently to the presence of genipin. In contrast, chitosan adhesive without IG was virtually transparent to the laser. The complete results are reported in table 3 and figure 5. Chitosan + IG (A) 2 1.5 1 0.5 Absorption (OD) 0 390 490 590 690 790 890 Wavelenght (nm) Chitosan+IG+Genipin (B) 3 2.5 2 1.5 1 Absorption (OD) 0.5 0 390 490 590 690 790 890 Wavelenght (nm) 107 Chitosan+Genipin (C) 1.4 1.2 1 0.8 0.6 0.4 Absorption (OD) 0.2 0 390 490 590 690 790 890 Wavelenght (nm) Chitosan (D) 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 390 490 590 690 790 890 Figure 5. Typical attenuation spectrum of chitosan adhesive in the visible-NIR region. A strong attenuation peak is localized at 808 nm, corresponding to the well-known absorption wavelength of IG dye (a). Another peak is located at 608 nm due to the genipin crosslinked amino groups (b, c). No peaks are present in chitosan films without IG and Genipin (d). 108 Table 3. Attenuation data of the adhesive at a wavelength of 808 and 608nm; the mean value ± the standard deviation are given. N, number of adhesive samples analysed. Attenuation Length, 1/e attenuation of the chitosan adhesive, as calculated from Beer’s law. Thickness, thickness of chitosan adhesives. Adhesive Type (N=3) l Thickness Attenuation Length (nm) (mm) (mm) Chitosan+IG (Group 1) 808 30 ± 10 5 ± 1 608 30 ± 10 25 ± 6 Chitosan (Group 2) 808 30 ± 10 228 ± 72 608 30 ± 10 223 ± 68 Chitosan+IG+Genipin 808 40 ± 10 6 ± 1 (Group 1) 608 40 ± 10 7 ± 1 Chitosan+ Genipin 808 20 ± 10 129 ± 16 (Group 4) 608 20 ± 10 6 ± 2 4.3.2 Tensiometer Measurements First Part The chitosan strips failed at the tissue interface in all types of repair procedures (group I). The laser-irradiated strips fully bonded to tissue and adhered more firmly at 80 and 120 mW laser output (Anova one-way, p<0.001, figure 6a). Although there was no statistical difference between the shear stress of the adhesive irradiated at 80 mW and 120 mW (13.2 ± 3.0 KPa and 10.5 ± 4.2 KPa respectively, p>0.05 Bonferroni’s post- test), it appeared that there was a trend for higher shear stress at 120 mW. The shear stress decreased to 8.0 ± 2.8 KPa at 160 mW laser output; the chitosan strips sporadically burned and twisted during irradiation. Thus, a power level of 120 mW appeared the best choice for the second part of the experiment because of the high shear stress and reduction in irradiation time. All results are displayed in figure 6a and table 1. 109 Shear Stress Versus Power 18 *P<0.001, N=10 16 14 * 12 10 8 6 * 4 Shear Stress (KPa) 2 0 0 80 120 160 Power (mW) Figure 6a. Histogram of acute shear stress of chitosan adhesive (group I) versus laser power (mean ± SD). Second Part The adhesive with IG (group I) performed a stronger repair than the adhesive with IG+Genipin (group III) after laser activation (14.7 ± 4.3 KPa and 9.1 ± 2.9 KPa respectively, n=30, p<0.001 t-test). Strips from the same groups resulted in a significantly lower adhesion to intestine without the laser irradiation step (shear stress= 1.9 ± 1.3 KPa and 0.6 ± 0.4 KPa, n=30, p<0.001, unpaired t-test). The strips failed at the tissue interface in all types of repair procedures but for the IG+ genipin adhesive (group III); 30% of the repairs broke in two pieces under the pulling maximum load. The chitosan strips from groups I and II were therefore selected for further characterization. All data are displayed in figure 6b and table 2. 110 Shear Stress Test 20 * *P<0.001, N=30 18 16 14 12 10 8 6 4 Shear Stress (KPa) 2 0 IG IG+Laser IG+Genipin IG+Genipin+Laser Figure 6b. Histogram of acute shear stress of chitosan adhesives from groups I and III, with and without the aid of laser radiation (mean ± SD). 4.3.3 13C-NMR The solid-state 13C-NMR spectra of commercially available chitosan, and the chitosan adhesive (group II) are displayed in Figures 7a and 7b respectively. No significant chemical change occurred during the preparation of chitosan adhesive. In both spectra there are the peaks expected for the glucosamine moiety at ~ 104(C1), 77(C3-5), 57(C2, C6) ppm. In the commercial chitosan sample the peaks for the residual acetate group from the unhydrolysed chitin at ~ 174(C=O) and 24(CH3) ppm could be readily observed, while in that for the chitosan adhesive, the acetate peaks are augmented by the acetic acid used in the preparation of the adhesive [169]. The peaks are relatively broad reflecting the amorphous nature of these materials. 111 A B Figure 7. 13C-NMR spectra of chitosan shells (a) and adhesive (b). The glucosamine moiety at ~104(C1), 77(C3-5), 57(C2, C6) ppm, are present in chitosan shells and adhesive. The residual acetate group from the unhydrolysed chitin, at ~174(C=O) and 24(CH3) ppm are augmented in the adhesive spectrum by acetic acid. No significant chemical changes occurred during the preparation of the chitosan adhesive. 112 4.3.4 Thermogravimetric Analysis and Contact Angle The TGA analysis of the adhesive with IG (group I) showed an initial sample weight loss of 11.9 ± 0.6% (n=6) at 202 ± 15 0C (figure 8). It is most likely that during this phase water evaporated. The adhesive degraded further and its weight diminished to 34.5 ± 3.6% at 304 ± 2 0C, the temperature of maximum rate of weight loss (derivative minimum peak). Similar results were achieved by the adhesive without IG (group II). The adhesive water content and the mass loss were not statistically different respect to the group I adhesive (13.2 ± 0.6% and 37.4 ± 1.5% respectively, n=5, p>0.05 student’s t-test), though the degradation temperature was higher (317 ± 2 0C, p<0.001 student’s t- test). TGA 100 Degradation Derivative 80 60 40 Weight (%) 20 0 0 100 200 300 400 500 600 -20 Temperature (0C) Figure 8. The weight loss of chitosan adhesive is given as a function of temperature. Water evaporated up to 211 0C, reducing the adhesive weight of 13 % (first derivate =0). The adhesive burnt at ~ 317 0C, where the weight loss rate achieved its maximum (minimum peak of the first derivative). The contact angle of the adhesive was 47 ± 4 degrees (n=6), consistent with a moderately hydrophilic nature (figure 9). 113 Figure 9. Contact angles of a water drop on chitosan adhesive. 4.3.5 Differential Scanning Calorimeter The thermograms of the adhesive (group I) showed a broad peak of heat absorption between 150 and 220 0C that disappeared during the second scan (figure 10). Chitosan degradation therefore occurred rather than melting at these temperatures, in agreement with the TGA study. If the adhesive melted, a similar peak should have also appeared in the second scan. No other transition was observed below 150 0C. DSC Analysis 80 70 60 50 First Scan 40 Second Scan 30 20 Heat Flow (mW) 10 0 0 50 100 150 200 250 Temperature (0C) Figure 10. The DSC scans show a broad degradation peak between 150 and 200 0C that disappears in the second scan. No other transition occurred below 1500C. 114 4.3.6 Atomic Force Microscopy The profile of the film surface was relatively smooth (Rq = 6.7 ± 4.0 nm, n=9) as irregular corrugations appeared sporadically over the chitosan film. The surface topography is shown in figure 11. Figure 11a. Topography of the chitosan adhesive obtained with AFM. Irregular peaks appeared sporadically throughout the film surface. 115 Figure 11b. Three-dimensional view of the adhesive surface. 4.3.7 Young’s Modulus Young’s modulus of wet chitosan strips from groups I and II were respectively 7.85 ± 1.5 MPa and 6.81 ± 1.26 MPa (n=15, p>0.05 student’s t-test). The presence of IG appeared not to alter significantly the elastic modulus of the adhesive. When combined, the above results yield to E = 7.3 ± 1.5 MPa (n=30). The adhesive not only showed high tensile strength but also flexibility as it could be flexed to ~160 degrees without suffering macroscopic damage. The tensile test results are summarized in figure 12. 116 Figure 12. Graph illustrating the stress and strain relationship for wet chitosan adhesive. The E modulus (~7.3 MPa) was calculated in the strain region between 10 and 20 %, where the adhesive shows its elastic behaviour. 4.3.8 Temperature Measurements The average temperature for LTR with chitosan adhesive (group I, n=15) was 57 0C (figure 13). The laser output raised sporadically the temperature to a maximum level of ~65 0C and this always resulted in charring of the adhesive. Moderate tissue shrinkage under the adhesive was observed whenever the thermocouple measured temperatures above 60 0C. The efficient absorption of IG prevented IR radiation to directly affect the thermocouple and alter the temperature measures. The irregular profile of the temperature plots in figure 6 is a reflection of the free hand use of the laser and is typical of surgical procedures. 117 65 60 55 50 TEMPERATURE (0C) 45 0 5 10 15 20 25 TIME (S) Figure 13. Graph illustrating the temperature of the thermocouple, under the chitosan adhesive at tissue interface, as a function of time. The mean temperature ± standard deviation (n=15) is also provided to indicate the temperature range during LTR. 4. 3.9 Ex-Vivo Histology and SEM Although the laser heat caused tissue coagulation in the proximity of the chitosan adhesive (~20 mm), no signs of tissue ablation, charring or vacuoles were observed in the histological sections (figures 14). SEM images confirmed the intestine serosa bonded to chitosan strips, even if the adhesion line was sporadically interrupted. Artefacts during tissue manipulation may have caused such detachment (figure15). 118 S A Figure 14a. Longitudinal section of the adhesive (A) laser-bonded to the intestine serosa (S) (H&E, x10). D S A Figure 14b. The serosa suffered mild thermal damage (D) soon after irradiation if compared to the control. Cell nuclei are stained blue (N) (H&E, x20). 119 S Figure 14c. Non laser irradiated intestine; the serosa appears healthy (H&E, x10). A S Figure 15a. SEM cross section of a chitosan strip (A) bonded on sheep intestine (S). 120 A S Figure 15b. SEM cross section of a chitosan strip (A) bonded on sheep intestine (S) at higher magnification. A F Figure 15c. Collagen fibers (F) are stretched and clearly bonded on the adhesive (A). 121 4.3.10 Cytotoxicity Assay The chitosan adhesive appeared to induce negligible cytotoxicity (figure 16). One-way variance analysis indicated a significant difference between the controls and experimental groups (p<0.0001). In particular, there was no statistical difference between the cell number of the null (n=3), blank (n=6) and chitosan extracted samples (n=9, cell number >280*103, p>0.05, Bonferroni’s post-test). The growth of cells treated with ethanol (n=3) was significantly inhibited when compared to the growth of the chitosan extracted samples and negative controls (cell number <55*103, p<0.001, Bonferroni’s post-test). Viability of the cells cultured in the chitosan extract, null and blank were greater than 95%, while cells cultured in the 7.5% ethanol control had a viability of 53% (figure 17). Forward and side scatter provided a qualitative evaluation of cell health; cells cultured in chitosan extract appeared equivalent to cells harvested from the blank or null (figure 18). The addition of 7.5% ethanol to culture media resulted in a qualitative changes to the cell shape, which was easily observed in their forward and side scatter. This detailed evaluation allowed for detection of toxic properties at doses lower than what would be statistically detectable in growth inhibition and confirmed the non cytotoxic properties of the chitosan adhesive. 122 Cells recovered from media treatments 400 350 300 250 200 Cells / ml 3 150 10 100 50 0 chitosan blank Null ethanol 7.5% Treatment Figure 16. Histogram illustrating the number of cells recovered as a function of media treatment (mean ± standard deviation). The chitosan adhesive extracts were not cytotoxic to fibroblasts when compared to the null and blank. Solutions 7.5% ethanol significantly inhibited cell growth (positive controls). Legend. Ethanol, samples of cells exposed to media with ethanol at 7.5% concentrations (v/v). Null, sample of cells exposed to fresh media. Blank, sample of cells exposed to 0 media, which was incubated for 24 hours in 5% CO2 air at 37 C. Chitosan Adhesive, samples of cells exposed to media, which was incubated with 0 chitosan adhesive for 24 hours in 5% CO2 air at 37 C. Cell Viability 100 90 80 70 60 50 40 30 Viable Cells (%) 20 10 0 Chitosan Blank Null 7.5% Ethanol Figure 17. Histogram of viable cells (%) recovered from the extracted media. More than 90% of recovered cells are still available after being treated with chitosan, while only 53% of cells appeared viable after ethanol incubation. 123 A B C Figure 18. Flowcytometer plots of cell fluorescence and forward/side scatter. Cells with their membrane broken fluorescence red light and are visualized in the R2 region while healthy cells are located in the R1 region. Cells extracted from chitosan (a) and untreated media (b) are healthy but many cells have their membrane broken after ethanol incubation (c). 124 4.3.11 In Vivo Thermal Damage The histology of the sciatic nerves showed that part of the axons underneath the irradiated adhesive was demyelinated, signalling potential thermal damage (figure 19a, b). The proximal and distal parts of the operated nerve, about 0.5 cm apart from the irradiated site, appeared less affected by the heat with myelinated axons retaining their normal morphology if compared to controls (figures 19c-e). Neutrophils were also visible around the adhesive, suggesting that an acute inflammation response was triggered (figure 20). The acute shear stress of the adhesive bonded to nerves was 11.1 ± 1.9 KPa (n=3). The chitosan collars failed at tissue interface under stress. A P A M Figure 19a. Cross section of an operated nerve. The adhesive (C) was laser-bonded to the perineurium (P); several axons are demyelinated (A) while others preserve the blue myelin sheet (M) four days post-operatively (Luxol Fast Blue, x20). 125 M A Figure 19b. Particular of figure 19a at higher magnification (Luxol Fast Blue,x40). M A Figure 19c. Cross section of the laser operated nerve at the distal site. Several axons preserve their myelinated sheet (M) while others are demyelinated (A) (Luxol Fast Blue, x20). 126 Figure 19d. Cross section of the proximal site of the laser operated nerve. The majority of axons retained their normal morphology (Luxol Fast Blue, x20). Figure 19e. Cross section of a non operated sciatic nerve. The majority of axons are myelinated (Luxol Fast Blue, x20). 127 4.4 Discussion This chapter described the development of a novel light-activated adhesive that is insoluble in water and has improved mechanical properties compared with existing solders. The chitosan adhesive (group I) bonded to sheep intestine with a relatively high shear stress (~14.7 KPa, maximum load ~0.50 N) and also demonstrated an E modulus of ~7.3 MPa, which resulted in all the adhesive strips investigated detaching from the tissue without breaking. When compared to the conventional albumin-based solders used in chapter III, the chitosan adhesive strips proved to withstand a higher maximum load (~0.19 N) than that of albumin strips (~0.11 N) of the same surface area (~13 mm2). It is problematic however to compare the shear stress of chitosan and albumin strips because the latter failed either cohesively or at the tissue interface under a pulling force. A N Figure 20. Cross section of the sciatic nerve and chitosan adhesive (A) surrounded by neutrophils (N) four days post-operatively (H&E, x20). The laser-activated adhesive was also stronger than the soluble chitosan gel, which was modified by Ono et al with UV crosslinkers (~3.1 KPa) [22]. 128 Chitosan adhesive strips demonstrated an initial adhesion to tissue prior to laser irradiation (~1.9 KPa), which is in agreement with previous reports [170]. Such adhesiveness was greatly enhanced by the laser as the IG dye converted photons into heat that supposedly diffused to the tissue interface, enhancing the bonding between chitosan and tissue. The mechanism responsible for this photo-adhesiveness is not yet clear, although we may hypothesize that polyanionic-polycationic interactions and hydrogen bonding occurred between collagen and chitosan when the laser heat denatured collagen, as previously reported [165, 166]. Such bonding strength appeared to decrease if genipin was added in the composition with or without laser activation (~ 9.1 KPa and 0.6 KPa respectively). Genipin strongly reacted with amino groups of the adhesive as witnessed by the high absorption peak at 608 nm. It might be possible that intermolecular and intramolecular crosslinking impaired the binding capability of collagen and chitosan. The free amino groups of chitosan, for example, might have diminished causing less polyanionic-polycationic interactions and hydrogen bonding with tissue collagen. Also, the chitosan crosslinking weakened the adhesive tensile strength and caused cohesive failure in 30% of the intestine repairs. In vivo histology demonstrated that axons suffered demyelination under the laser- activated adhesive, four days post operatively. The thermal damage though was mostly limited under the adhesive as several axons in the proximal and distal sites were myelinated and preserved their normal morphology. The axon demyelination was likely due to the high fluence (~65 J/cm2) used to activate the adhesive, even if satisfactory bond strength (~11 KPa) was achieved in vitro by using a lower fluence (~46 J/cm2). The surgeon over irradiated the chitosan strips because the adhesive activation lacked a 129 visual end point. A reduction of fluence is therefore needed to avoid or diminish nerve thermal damage; it also may prove beneficial to irradiate the adhesive with pulses rather than in continuous wave, as previously reported [168]. In agreement with the above considerations, the ex vivo histology showed mild tissue coagulation of the serosa beneath the adhesive (~20 mm) but no carbonisation or vacuoles, when the laser fluence was 52 J/cm2. The results of acute histology were most likely due to the low levels of fluence and irradiance (~52 J/cm2 and15 W/cm2 respectively). The laser raised the average temperature at the adhesive-tissue interface from 57 to 65 0C; such temperatures were recorded by the thermocouple and are an estimation of the real temperatures of the tissue interface during LTR. These measures were subjected to errors, mainly due to the thermocouple mass and the limited surface area of the adhesive probed by the thermocouple. The latter error was due to the small diameter of the thermocouple bed; this resulted in a significant temperature drop whenever the laser beam was not in the thermocouple proximity. The error due to the thermocouple mass was estimated to be ~ 2 0C from the manufacturer specifications. The above considerations indicate that (figure 6) and do not liquefy, denature or melt below 70 0C like protein solders. One 130 may assume that only the tissue collagen has to denature at 60-65 0C prior to bonding to chitosan and substantially increase the initial adhesiveness of chitosan films [173-175]. DSC thermograms showed that chitosan adhesives did not undergo a phase transition when the laser activated the adhesion of chitosan strips on tissue. Hence, the fluence characterizes the laser repair of chitosan adhesives better than the radiation dose (J/mg). In contrast, the latter is more appropriate for albumin solders, which undergo a phase transition during denaturation and need a fixed amount of energy per mole to change the protein conformation (1.2 ± 0.5 J/g) [52]. The fluence required to irradiate thoroughly (3 times) the chitosan adhesive was ~52 J/cm2. It is very important that tissue glues, solders and adhesives are not altered or degraded by aqueous solutions such as physiological fluids, which are often used to maintain the moisture of exposed tissues at operated sites. It was noticed during the experiments that the chitosan adhesives swelled, curled and dissolved in water, if not dried thoroughly. Particular attention was therefore necessary to dry the films until they became insoluble. Alternatively, the chitosan films used in chapter II were dried overnight and required to be immersed in a concentrated solution of NaOH before becoming insoluble. This treatment appeared to increase the film water content (~40%) and consequently the E modulus was drastically reduced (~0.77 MPa). Therefore, chitosan treatment with a concentrated solution of NaOH was avoided in this thesis to preserve the high E modulus of chitosan films (~7.3 MPa). In contrast to albumin solders, chitosan adhesives not only have the notable property of being insoluble but also exhibit important mechanical properties such as high E 131 modulus and flexibility. Adhesives with ~20 mm thickness could sustain a shear stress of ~14.7 KPa without breaking, had an E modulus of ~ 7.3 MPa and could be curved or deformed over 160 degrees, returning to their initial shape with no sign of macroscopic damage. This flexibility of the chitosan adhesive permitted the ready manipulation of tissue without fear of breaking or tearing. The adhesive did not fold or breakdown when manipulated with forceps and appeared to be well suited for tissue repair. The application of the adhesive on tissue was also facilitated by its adhesiveness prior to laser activation and hydrophilic properties (contact angle ~47’). For example, the chitosan strip adhered like a collar on the nerve stumps, aligning them end-to-end and allowing nerve rotation during the laser irradiation. Unlike fibrinogen and albumin solders, which are derived from blood-based proteins, the chitosan adhesive is based on a polysaccharide and consequently there are no risks of viral infections associated to this novel adhesive. Furthermore, Chitosan is widely accepted as a non-toxic and biocompatible polysaccharide. Among other applications, chitosan is employed to develop skin grafts, tissue scaffolds and dietary products [176- 180]. Also of consideration in tissue repair is the remarkable antimicrobial properties of chitosan, which reduces potential infection [181, 22]. According to 13C-NMR, the laser-activated adhesive has the same chemical composition of chitosan shells derived from marine crustacean and therefore should retain the above mentioned properties of biocompatibility and non toxicity (figure 4). The results of the cytotoxicity test support this suggestion, with media extracted from the adhesive showing negligible toxicity to fibroblasts [182]. 132 Chitosan adhesives may degrade in the body and also act as a delivery system at the repair site, with the potential of incorporating therapeutic drugs such as growth factors, antibiotics or genes to guide and enhance the wound healing process [183,184]. It has been shown that in vitro and in vivo degradation of chitosan films occurred less rapidly as the film deacetylation became higher [96]. Chitin films were implanted in the back of rats and degraded ~25% in two weeks, while chitosan films (85% deacetylated) degraded 20% in 12 weeks. Furthermore, lysozyme degradation of chitosan has been reported [96,97]. For example, 10% of chitosan films (85% deacetylated) degraded in 48 hours if incubated in buffered aqueous solution (pH~7) with lysozyme (4 mg/ml) at 37 0C [97]. Chitosan adhesives are inexpensive and may be manufactured in a broad spectrum of shapes and dimensions to suit the organ or tissue characteristics where they are applied. 133 Chapter V Conclusions The sutureless laser technique for vasovasostomy, used in chapter II, highlighted the limitations of solid albumin solders. For example, their mechanical properties did not fully match requirements of the surgical procedure as the solder quickly became brittle in air and could not be molded into a stent with small diameter (³1mm). The inability of fabricating a small stent with albumin solders stimulated the search for new biomaterials with better mechanical qualities such as chitosan films. Another limitation of solders is their solubility in aqueous solutions before being laser irradiated; a dry operation field is therefore mandatory for surgeons who are usually reluctant to employ them in a variety of surgical procedures such as the end-to end anastomosis of small blood vessels. The solubility and flexibility problems may be overcome by pre-denaturing the solder in hot water for a short time (100 0C, 1s). A major shortcoming for such procedure is the declining of the repair strength, due to the diminished solubility of the solder. This approach was therefore avoided in the present investigations. Another limitation highlighted by the sutureless vasovasostomy procedure was the thermal damage induced by the laser to the vas deferens, which likely contributed to the stricture and sperm granuloma formation. The repair bonding strength needed to improved as well, though it was sufficient to successfully repair the vasa. Increasing the bonding strength of solders is a necessity for the safety and reliability of laser repairs and also for diminishing the solder amount required during the operation. Less solder 134 implies less surface area of tissue exposed to the laser heat and therefore to thermal damage. The instance of increasing the bonding strength of albumin solders was explored in the third chapter by developing a novel Genipin-BSA solder. Genipin is a natural crosslinker that proved to be capable of doubling the bonding strength of laser repaired tissue, without inducing any significant cytotoxicity. Genipin crosslinked albumin molecules by acting on their amino groups, without photochemical activation. The laser absorbed by the solder increased the temperature and promoted the genipin crosslinking between albumin molecules. The thermal activation of genipin might therefore continue in the body at 37 0C and it has the potential of strengthening the tissue bonds of laser repairs during the first few days post surgery, which is the most critical period for tissue failure. Unfortunately, the shortcoming for the thermal activation of genipin may be the reduced shelf life of the solder, due to protein crosslinking. The Genipin-Albumin solder still has the same limitations related to solubility, brittleness and flexibility as traditional solders, although it increases the strength of tissue bonds. The mechanical and solubility weaknesses were overcome by the development of a novel chitosan adhesive that was insoluble in water, had a Young modulus of ~7.3 MPa and could be bent 1600 without suffering macroscopic damage. The chitosan adhesive was thermally activated by a laser and adhered firmly on tissue. The thermal damage induced on tissue appeared restricted to the adhesive interface, being the tissue temperature 60-65 0C during laser irradiation. Such a temperature range is significantly lower than the one required for albumin solders (65-70 0C). The adhesion strength of chitosan appeared to be higher than that of albumin solders, 135 nevertheless, more studies are needed to validate the superior strength occurs because of tissue interface bonding and not because of the high cohesive strength of chitosan. Additional studies are also required to investigate and optimize the shear stress dependence on the temperature of chitosan adhesion; a temperature control system with feedback is the best option for this task [185]. This adhesive had also the remarkable property to bond to tissue without laser- activation, though the adhesive strength was about seven times lower than the one obtained with the laser. This pre-heating adhesion proved to be crucial in applying easily the adhesive onto tissue and handling it during the laser repair. Chitosan is the second most diffused biomaterial after collagen. It is a polysaccharide and therefore has no risk of viral infections like albumin. Chitosan has remarkable antibacteric and antimicrobial properties that make it particularly suitable for tissue repair. Chitosan is also an excellent drug delivery system and in future developments it may incorporate antibiotics, growth factors and genes to help the wound healing process. Despite these promising qualities, there is still the need of increasing the shear stress of the adhesive on tissue in order to expand its applications not only on a variety of soft tissue such as tendons, dura, peritoneum, fascia and muscles, but also on bones and cartilage. It is of paramount importance to investigate in the future the mechanism of tissue bonding of chitosan adhesives to exploit and increase the adhesion shear stress. Also, the activation temperature needs to be lowered as cells and tissue are irreversibly damaged at 60 0C. Further research should therefore focus on in vivo applications of the 136 chitosan adhesive to confirm the safety and efficacy of this biomaterial, and on the improvement of the chemical formulation to enhance the adhesion strength and lower the activation temperature. In conclusion, the investigations reported hitherto achieved the following results: 1. A new tissue adhesive, based on chitosan, was developed and characterized. The chitosan adhesive overcame typical faults of protein solders such as water solubility, flexibility and risks of viral infection. 2. The addition of genipin in the chemical formulation of traditional albumin solders, doubled the tissue repair strength without inducing significant cytotoxicity. 3. A novel self-expandable chitosan stent was designed and manufactured for surgical applications. 137 References 1. Kjaergard HK. Suture Support:Is It Advantageous? Am J Surg, 2001; 182:15S-20S. 2. Millesi H. Peripheral nerve injuries. Nerve Sutures and Nerve Grafting. Scand J Plast Reconstr Surg Suppl, 1982; 19:25-37. 3. Kramer K, Senninger N, Herbst H, Probst W. Effective prevention of Adhesion with Hyaluronate. Arch Surg, 2002:137:278-282. 4. Pecha RE, Prindiville T, Kotfila R, Ruebner B, Cheung AT, Trudeau W.Gastrointestinal hemorrhage consequent to foreign body reaction to silk sutures: case series and review. Gastrointest Endosc. 1998 Sep;48(3):299-301. 5. Thetter O. Fibrin adhesive and its application in thoracic surgery. Thorac Cardiovasc Surg; 29: 290-292, 1981. 6. Berger A, Tempfer C, Hartmann B, Kornprat P, Rossmann A, Neuwirth G, Tulusan A, Kubista E. Sealing of postoperative axillary leakage after axillary lymphadenectomy using a fibrin glue coated collagen patch: a prospective randomised study. Breast Canver Res Treat; 67: 9-14, 2001. 7. Saltz R, Sierra D, Feldman D. Experimental and Clinical Applications of Fibrin Glue. Plast Reconstr Surg, 1991; 88:1005-1015. 138 8. Kjaergard HD, Weis-Fogh US. Important Factors Influencing the Strength of Autologus Fibrin Glue: the Fibrin Concentration and reaction Time. Eur Surg Res, 1994; 26:273-276. 9. Martinowitz U, Saltz R. Fibrin Sealant. Curr Opin Hematol, 1996; 3:395-402. 10. Spotnitz WD. Commercial Fibrin sealants in Surgical Care. Am J Surg, 2001; 182:8S-14S. 11. Shaffrey CI, Spotnitz WD, Shaffrey ME, Jane JA. Neurosurgical Applications of Fibrin Glue: Augmentation of dural closure in 134 patients. Neurosurgery, 1990; 26:207-210. 12. Bento RF, Miniti A. Anastomosis of the Intratemporal Facial Nerve Using Fibrin Tissue Adhesive. Eear Nose Throat J, 1993; 72:663-667. 13. Polsven B. A New Fibrin seal: Functional Evaluation of Sensory Regeneration Following Primary Repair of Peripheral Nerves. J Hand Surg (Br), 1994; 19:250-254. 14. Von Oppel UO, Zilla P. Tissue Adhesives in Cardiovascular Surgery. J Long- Term Effects Med Implants, 1998; 8:87-101. 15. Von Segesser LK, Fasnacht MS, Vogt PR. Prevention of Ventricular Septal Defects with Fibrin sealants. Ann Thorac Surg, 1995;60:511-516. 139 16. Rutgeerts P, Rauws E, Wara P, Swain P, Hoose A, Solleder E, Halttunen J, dobrilla G, Richter G, Prassler R. Randomized Trials of a Single and Repeated Fibrin Glue compared with Injection of Polidocanol in Treatment of Bleeding peptic Ulcer. Lancet, 1997; 350:692-696. 17. Sierra DH, O`Grady K, Toriumi DM, Foresman PA, Rodeheaver GT, Eberhart A, Feldman DS, lemons JE. Modulation of Mechanical Properties in Multiple- Component Tissue Adhesives. J Biomed Mater Res, 2000; 52:534-542. 18. Jackson RM. New and Potential Uses of Fibrin Sealants as an Adjunct to Surgical Hemostasis. Am J Surg, 2001; 182:, 36S-39S. 19. Petratos PB, Felsen D, Trierweiler G, Pratt B, McPherson JM, Poppas DP. Transforming growth factor-beta2 (TGF-beta2) reverses the inhibitory effects of fibrin sealant on cutaneous wound repair in the pig. Wound Repair Regen. 2002 Jul- Aug;10(4):252-8. 20. Menovsky T, Beek JF. Laser, fibrin glue, or suture repair of peripheral nerves: a comparative functional, histological, and morphometric study in the rat sciatic nerve. J Neurosurg. 2001 Oct;95(4):694-9. 21. Sierra DH, Eberarhart AW, Lemons JE. Failure Characteristics of Multiple- Component Fibrin Based Adhesive. J Biomed Mater Res,2002;59:1-11. 140 22. Ono K, Saito Y, Yura H, Ishikawa K, Kurita A, Akaike T, Ishihara M. Photocrosslinkable chitosan as a biological adhesive. J Biomed Mater Res;49:289- 95,2000. 23. Tseng YC,Hyon SH, Ikada Y. Modification of synthesis and investigation of properties for 2-cyanoacrylates. Biomaterials;11:73-79, 1990. 24. Vote JT, Elder MJ. Cyanoacrylate Glue for Corneal Perforations: A Description of surgical Technique and a Review of the Literature. Clin Experiment Ophthalmol. 2000 Dec;28(6):437-42. 25. Linden CL, Shalaby SW. Performance of Modified Cyanoacrylate Composition as Tissue Adhesives for Soft and hard Tissues. J Biomater Res, 1997;38:348-355. 26. Marcovich R, Williams AL, Rubin MA, Wolf JS Jr. Comparison of 2-octyl cyanoacrylate adhesive, fibrin glue, and suturing for wound closure in the porcine urinary tract. Urology. 2001 Apr;57(4):806-10. 27. Gabrielli F, Potenza C, Puddu P, Sera F, Masini C, Abeni D. Suture materials and other factors associated with tissue reactivity, infection, and wound dehiscence among plastic surgery outpatients. Plast Reconstr Surg; 107:38-45, 2001. 28. Cursio R, Gugenheim J, Mouiel J, Saint-Paul MC, Ragusa L, Bronsino E, Canino V. The biocompatibility of suture materials used in colon surgery. An experimental study in the rat. Minerva Chir; 54:49-55, 1999. 141 29. Kirsch AJ, de Vries GM, Chang DT, Olsson CA, Connor JP, Hensle TW. Hypospadias repair by laser tissue soldering: intraoperative results and follow-up in 30 children. Urology;48:616-23,1996. 30. Poppas DP, Rooke CT, Schlossberg SM. Optimal parameters for CO2 laser reconstruction of urethral tissue using a protein solder. J Urol;148(1):220-4,1992. 31. Schober R, Ulrich F, Sander T, Durselen H, Hessel S. Laser-induced alteration of collagen substructure allows microsurgical tissue welding. Scince;232:1421- 1422,1986. 32. Dubuisson AS, Kline DG. Is laser repair effective for secondary repair of a focal lesion in continuity? Microsurg;14:398-401,1993. 33. Fung LC, Mingin GC, Massicotte M, Felsen D, Poppas DP. Effects of temperature on tissue thermal injury and wound strength after photothermal wound closure. Lasers Surg Med;25:285-290,1999. 34. Reali UM, Gelli R, Giannotti V, Gori F, Pratesi R, Pini R.Experimental diode laser-assisted microvascular anastomosis. J Reconstr Microsurg. 1993 May;9(3):203- 10. 35. Maragh H, Hawn RS, Gould JD, Terzis JK. Is laser nerve repair comparable to microsuture coaptation? J Reconstruct Microsurg;4:189-195,1993. 142 36. Menovsky T, Beek JF, van Gemert MJ.Laser tissue welding of dura mater and peripheral nerves: a scanning electron microscopy study. Lasers Surg Med. 1996;19(2):152-8. 37. Menovsky T, Beek JF, van Gemert MJ. Microsurgery. CO2 laser nerve welding: optimal laser parameters and the use of solders in vitro. 1994;15(1):44-51. 38. Poppas DP, Schlossberg SM, Richmond IL, Gilbert DA, Devine CJ Jr. Laser welding in urethral surgery: improved results with a protein solder. J Urol. 1988 Feb;139(2):415-7. 39. Bass LS, Treat MR, Dzakonski C, Trokel SL. Sutureless microvascular anastomosis using the THC:YAG laser: a preliminary report. Microsurgery. 1989;10(3):189-93. 40. Oz MC, Bass LS, Popp HW, Chuck RS, Johnson JP, Trokel SL, Treat MR. In vitro comparison of thulium-holmium-chromium: YAG and argon ion lasers for welding of biliary tissue. Lasers Surg Med. 1989;9(3):248-53. 41. Korff M, Bent SW, Havig MT, Schwaber MK, Ossoff RH, Zealear DL. An investigation of the potential for laser nerve welding. Otolaryngol Head Neck Surg. 1992 Apr;106(4):345-50. 143 42. Chuck RS, Oz MC, Delohery TM, Johnson JP, Bass LS, Nowygrod R, Treat MR. Dye-enhanced laser tissue welding. Lasers Surg Med. 1989;9(5):471-7. 43. Treat MR, Oz MC, Bass LS. New technologies and future applications of surgical lasers. The right tool for the right job. Surg Clin North Am. 1992 Jun;72(3):705-42. 44. Oz MC, Johnson JP, Parangi S, Chuck RS, Marboe CC, Bass LS, Nowygrod R, Treat MR. Tissue soldering by use of indocyanine green dye-enhanced fibrinogen with the near infrared diode laser. J Vasc Surg. 1990 May;11(5):718-25. 45. Sauda K, Imasaka T, Ishibashi N. Determination of protein in human serum by high-performance liquid chromatography with semiconductor laser fluorometric detection. Anal Chem. 1986 Nov;58(13):2649-53. 46. Lauto A, Kerman I, Ohebshalon M, Felsen D, Poppas DP. Two-layer film as a laser soldering biomaterial. Lasers Surg Med. 1999;25(3):250-6. 47. Cuomo BE, Lauto A, Kirman I, Felsen D, Poppas DP. Assessment of the degradation of denatured albumin solder by human urine. J Urol. 2000 Feb;163(2):634-7. 48. Kirsch AJ, Cooper CS, Gatti J, Scherz HC, Canning DA, Zderic SA, Snyder HM 3rd. Laser tissue soldering for hypospadias repair: results of a controlled prospective clinical trial. J Urol. 2001 Feb;165(2):574-7. 144 49. Wadia Y, Xie H, Kajitani M. Liver repair and hemorrhage control by using laser soldering of liquid albumin in a porcine model. Lasers Surg Med. 2000;27(4):319-28. 50. Bleustein CB, Cuomo B, Mingin GC, Ohebshalom M, Lauto A, Shin SJ, Stewart RB, Felsen D, Soslow RA, Sennett M, Poppas DP. Laser-assisted demucosalized gastrocystoplasty with autoaugmentation in a canine model. Urology. 2000 Mar;55(3):437-42. 51. Kirman I, Lauto A, Phillips A, Hamawy A, Heldman E, Cuomo B, Shin SJ, Soslow R, Felsen D, Poppas DP. Effect of laser welding with human serum albumin on the expression of P-selectin on platelets. Lasers Surg Med. 1999;25(5):438-44. 52. Lauto A, Stewart R, Ohebshalom M, Nikkoi ND, Felsen D, Poppas DP. Impact of solubility on laser tissue-welding with albumin solid solders. Lasers Surg Med; 28:44- 49;2001. 53. McNally KM, Sorg BS, Bhavaraju NC, Ducros MG, Welch AJ, Dawes JM. Optical and Thermal Characterization of Albumin Protein Solders. Applied Optics. 1999; 38 (31): 6661-72. 54. Lauto A, Trickett R, Malik R, Dawes JM, Owen ER. Laser-activated solid protein bands for peripheral nerve repair: an in vivo study. Lasers Surg Med;21:134-141, 1997. 145 55. Lauto A, Dawes JM, Piper JA, Owen ER. Laser nerve repair by solid protein band technique. II: assessment of long-term nerve regeneration.Microsurgery. 1998; 18(1): 60-4. 56. Maitz PK, Trickett RI, Dekker P, Tos P, Dawes JM, Piper JA, Lanzetta M, Owen ER. Sutureless microvascular anastomoses by a biodegradable laser-activated solid protein solder. Plast Reconstr Surg. 1999 Nov;104(6):1726-31. 57. Xie H, Bendre SC, Burke AP, Gregory KW, Furnary AP. Laser-assisted vascular end to end anastomosis of elastin heterograft to carotid artery with an albumin stent: A preliminary in vivo study. Lasers Surg Med. 2004;35(3):201. 58. Lauto A, Poppas DP, Murrell GA. Lasers Surg Med. Solubility study of albumin solders for laser tissue-welding. 1998;23(5):258-62. 59. Bell LN, Hageman MJ, Bauer JM. Impact of moisture on thermally induced denaturation and decomposition of lyophilized bovine somatotropin. Biopolymers 1995; 35:201-209. 60. Hummer G, Garde S, Garcia AE, Paulatis ME, Pratt LR. The pressure dependence of hydrophobic interaction is consistent with the observed pressure denaturation of proteins. Proc Natl Acad Sci USA. 1998; 95:1552-5. 146 61. Poppas DP, Wright EJ, Guthrie PD, Shlahet LT, Retik AB. Human albumin solders for clinical application during laser tissue welding. Lasers Surg Med. 1996;19(1):2-8. 62. Lauto A. Repair strength dependence on solder protein concentration: a study in laser tissue-welding. Lasers Surg Med. 1998;22(2):120-5. 63. Stewart RB, Benbrahim A, LaMuraglia GM, Rosenberg M, L'Italien GJ, Abbott WM, Kung RT. Laser assisted vascular welding with real time temperature control. Lasers Surg Med;19:9-16,1996. 64. Shenfeld O, Ophir E, Goldwasser B, Katzir A. Silver halide fiber optic radiometric temperature measurement and control of CO2 laser-irradiated tissues and application to tissue welding. Lasers Surg Med. 1994;14(4):323-8. 65. Shumalinsky D, Lobik L, Cytron S, Halpern M, Vasilyev T, Ravid A, Katzir A. Laparoscopic laser soldering for repair of ureteropelvic junction obstruction in the porcine model. J Endourol. 2004 Mar;18(2):177-81. 66. Brosh T, Simhon D, Halpern M, Ravid A, Vasilyev T, Kariv N, Nevo Z, Katzir A. Closure of skin incisions in rabbits by laser soldering II: Tensile strength. Lasers Surg Med. 2004;35(1):12-7. 147 67. Lauto A, Stewart RB, Felsen D, Foster J, Poole-Warren L, Poppas D. Low temperature solder for laser tissue welding. Proceedings SPIE Vol. 5287;83-90, Laser Florence 2002. 68. McNally KM, Sorg BS, Welch AJ. Novel solid protein solder designs for laser- assisted tissue repair. Lasers Surg Med. 2000;27(2):147-57. 69. McNally KM, Sorg BS, Hammer DX, Heintzelman DL, Hodges DE, Welch AJ. Improved vascular tissue fusion using new light-activated surgical adhesive on a canine model. J Biomed Opt. 2001 Jan;6(1):68-73. 70. Sweeney T, Rayan S, Warren H, Rattner D. Intestinal anastomoses detected with a photopolymerized hydrogel. Surgery. 2002 Feb;131(2):185-9. 71. Ishihara M, Nakanishi K, Ono K, Sato M, Kikuci M, Saito Y, Yura H, Matsui T, Hattori H, Uenoyama M, Kurita A. Photocrosslinkable Chitosan as a Dressing for Wound Occlusion and Accelerator in Healing Process. Biomaterials, 2002;23:833- 840. 72. Nakayama Y, Matsuda T. Photocurable Surgical Tissue Adhesive Glues Composed of Photoreactive Gelatin and Poly(ethylene glycol) Diacrylate. J Biomed Mater Res, 1999; 48:511-521. 148 73. Ahn JS, Choi HK, Cho CS. A Novel Mucoadhesive Polymer Prepared by Template Polymerization of Acrylic Acid in the Presence of Chitosan. Biomaterials, 2001;22:923-928. 74. Yamada K, Chen T, Kumar G, Vesnovsky O, Topoleski LDT, Payne GF. Chitosan Based Water-Resistant Adhesive. Analogy to Mussel Glue. Biomacromolecules, 2000;1:252-258. 75. Effects of chitosan on experimental abscess with Staphylococcus aureus in dogs. Okamoto Y, Tomita T, Minami S, Matsuhashi A, Kumazawa NH, Tanioka S, Shigemasa Y.J Vet med Sci;57:765-767,1995. 76. Suzuki K, Okawa Y, Hashimoto K, Suzuki S, Suzuki M. Protecting effect of chitin and chitosan on experimentally induced murine candidiasis. Microbiol Immunol;28:903-912,1984. 77. Rao B and Sharma C. Use of Chitosan as a Biomaterial: Studies on Its Safety and Hemostatic Potential. J Biomed Res Mater, 1997;34:21-28. 78. Okamoto Y, Tomita T, Minami S, Matsuhashi A, Kumazawa NH, Tanioka S, Shigemasa Y. Effects of chitosan on experimental abscess with Staphylococcus aureus in dogs. J Vet Med Sci. 1995 Aug;57(4):765-7. 149 79. Howling GI, Dettmar PW, Goddard PA, Hampson FC, Dornish M, Wood AJ. The Effect of Chitn and Chitosan on the Proliferation of Human Skin Fibroblasts and keratinocytes in vitro. Biomaterials, 2001; 22:2959-2966. 80. Mattioli M, Muzzarelli B, Muzzarelli RAA. Chitin and Chitosan in Wound h Healing and Other Biomedical Applications. Carbohyr Europe, 1997;19:30-36. 81. Mori T, Okamura M, Matsuura M, Ueno K, Tokura S, Okamoto Y, Minami S, Fujinaga T. Effect of Chitin and its Derivates on the Proliferation and Cytokine Production of Fibroblasts in vitro. Biomaterials, 1997;18:947-951. 82. Inui H, Tsujikubo M, Hirano S. Low Molecular Weight Chitosan Stimulation of mitogenic Response to Platelet Derived Growth Factor in vascular Smooth Muscle cells. Biosi Biotech Biochem, 1995;59:2111-2114. 83. Denuziere A, Ferrier D, Damour O, Domard A. Chitosan Chondroitin Sulphate and Chitosan Hyaluronate Polyelectrolite Complexes:Biological Properties. Biomaterials, 1998;19:1275-1285. 84. Risbud M, Endres M, Ringe J, Bhonde R, Sittinger M. Biocompatible Hydrogel Supports the Growth of Respiratory Epithelial Cells: Possibility in tracheal Tissue Engineering.J Biomed Mater Res, 2001; 56:120-127. 150 85. Hidaka H, Ito m, Mori K, Yagasaki H, Kafrawy AH. Histopatological and immunohistochemical studies of membranes of deacetylated chitin derivates implanted over rat calvaria. J Biomed Mater res, 1999;46: 418-423. 86. Mori T, Okamura m, Matsuura m, Ueno K, Tokura S, Okamoto Y, Minami S, Fujinaga T. Effects of Chitin and its Derivates on the Proliferation and Cytokine Production of fibroblasts in vitro. Biomaterials, 1997: 18:947-951. 87. Chatelet C, Damour O, Domard A. Influence of the Degree of Acetylation on Some Biological Properties of Chitosan Films. Biomaterials, 2001: 22:261-268. 88. Yan XL, Khor E, Lim LY. Chitosan Alginate Films Prepared with Chitosans of Different Molecular Weights. J Biomed Mater Res, 2001; 58: 358-365. 89. Gaserod O, Smidsord O, Skjak G. Microcapsules of Alginate-Chitosan. Biomater, 1998;19:1815-1825. Chitosan-alginate membrane are strengthened by the addition of calcium ions. 90. VandeVord PJ, Matthew HWT, DeSilva SP, Mayton L, Bin W, Wooley PH. Evaluation of the Biocompatibility of a Chitosan Scaffold in Mice. J Biomed Mater Res, 2002;59:585-590. 91. Usami Y, Okamoto Y, Minami S, Matsuhashi A, Kumazawa NH, Tanioka S, Shigemasa Y. Migration of Canine Neutrophils to Chitin and Chitosan. J Vet Med Sci, 1994;56:1215-1216. 151 92. Mori T, Irie Y, Nishimura SI, Tokura S, Matsuura M, Okumura M, Kadosawa T, Fusjinaga T. Endothelial Cell Responses to Chitin and its Derivates. J Biome Mater Res, 1998; 43:469-472. 93. Sechriest VF, Miao WJ, Niyibizi C, Larson A, Mattew HW, Evans CH, Fu FH, Suh JK. GAG-Augmented Polysaccaride Hydrogel: A novel Biocompatible Material to support chondrogenesis. J Biome Mater Res, 2000; 49:534-541. 94. Mi FL, Wu YB, Shyu SS, Schoung JY, Hang YB, Tsai YH, Hao JY. Control of Wound Infection Using a Bilayer Chitosan Wound Dressing with Sustainable Antibiotic Delivery. J Biomed Mater Res, 2002;59: 438-449. 95. Tsipouras N, Rix CJ, Brady PH. Passage of Silver Ions through Membrane Mimetic Materials and its Relevance to Treatment of Burn Wounds with Silver Sulfadiazine Cream. Clin Chem, 1997;43: 290-301. 96. Tomihata K AndIkada Y. In Vitro and In Vivo Degradation of Films of Chitin and its Deacetylated Derivates. Biomaterials, 1997;18:567-575. 97. Lee KY, Ha WS, Park WH. Blood Compatibility and Biodegradability of partially N-acylated Chitosan Derivates. Biomaterials, 1995;16:1211-1216. 98. Sashiwa H, Saimoto h, Ogawa R, Tokura S. Lysozymic Susceptibility of Partially Deacetylated Chitin. Int J Biol Macromol, 1990;12: 295-296. 152 99. Aiba S. Studies on Chitosan: 4. Chitin. Int J Biol Macromol, 1992; 14: 225-228. 100. Hirano S, Yagi Y. The effect of N-Substitution of Chitosan and The Physical Forms of the products on the Rate of Hydrolysis by Chitinase from Streptomyces Grseus. Carbohydr Res, 1980;83 : 103-108. 101. Hara S, Matsushima Y. Studies on The Substrate Specificity of Egg White Lysozyme. J Biochem, 1967; 62: 118-125. 102. Nishi C, Nakajima N, Ikada Y. In viitro Evaluation of Cytotoxicity of Diepoxy Compounds used for Biomaterial Modification. J Biomed Mater Res, 1995;29:829- 834. 103. Sung HW, Hang RN, Huang LLH, Tsai CC. In vitro Evaluation of cytotoxycity of a natural crosslinking Reagent for Biological Tissue Fixation. J Biomater Sci Polymer Edn, 1999;10:63-78. 104. Sung HW, Chen CN, Hang RN, Hsu JC, Chang WH. In vitro Surface Characterization of a biological Patch Fixed with a Naturally Occurring Crosslinking Agent. Biomaterials, 2000;21:1353-1362. 105. Mi FL, Sung HW, SHYU SS. Synthesis and Characterization of a novel Chitosan-Based Network Prepared Using Natural Occurring Crosslinker. J Polym Sci A: Plym Chem, 2000;28:2804-2814. 153 106. Mi FL, Tan YC, Liang HF, Sung HW. In vivo Biocompatibility and Degradability of a Novel Injectable Chitosan based implant. Biomaterials, 2002;23:181-191. 107. Seaman EK, Kim ED, Kirsch AJ, Pan YC, Lewitton S, Lipshultz LI. Results of laser tissue soldering in vasovasostomy and epididymovasostomy: experience in the rat animal model. J Urol. 1997 Aug;158(2):642-5. 108. Trickett RI, Wang D, Maitz P, Lanzetta M, Owen ER. Laser welding of vas deferens in rodents: initial experience with fluid solders. Microsurgery. 1998;18(7):414-8. 109. Pohl D, Bass LS, Stewart R, Chiu DT. Effect of optical temperature feedback control on patency in laser-soldered microvascular anastomosis. J Reconstr Microsurg. 1998 Jan;14(1):23-9; discussion 29-30. 110. Mingin GC, Ditrolio JV. Vasovasostomy using albumisol solder with an argon laser. Br J Urol. 1998 Apr;81(4):628-9. 111. BH Lee, YS Do, JH Lee, KH Kim, SY Chin. New self-expandable spiral metallic stent: preliminary evaluation in malignant biliary obstruction. J Vasc Interv Radiol 1995; 6:635-40. 154 112. RK Tschada, TO Henkel, KP Junemann, J Rassweiler, p Alken. Spiral reinforced ureteral stent: an alternative to internal urinary diversion. J Endourol 1994; 8: 119-23. 113. Z Braf, J Chen, M Sofer, H Matzkin. Intraprostatic metal stents: more than 6 years’ clinical experience with 110 patients. J Endourol 1996;10:555-8. 114. G Cozzi, MF Colnago, M Bellomi, G Giovannardi, M Salvetti, A Saverini. Protesi metalliche autoespandibili nel trattamento delle stenosi neoplastiche esofagee. Radiol Med 1994;88: 272-6. 115. E van Sonnenberg, HB D’Agostino, R O’Laoide, J Donaldson, RB Sanchez, A Hoyt, CC Pittman. Malignant ureteral obstruction:treatment with metal stents techniques, results, and observations with percutaneous intraluminal US. Radiology 1994; 765-8. 116. Petas A, Talja M, Tammela TLJ, Taari T, Valimaa T, Tormala P. The biodegradable self-reinforced poly-DL-lactic acid spiral stent compared with a suprapubic catheter in the treatment of post-operative urinary retention after visual laser ablation of the prostate. Br J urol 1997;80: 439-443. 117. Petas A, Karkkaainen P, Talja M, Taari K, Laato M, Valimaa T, Tormalaa P. Effects of biodegradable self-reinforced polyglicolic acid, poly-DL-lactic acid and stainless-steel spiral stents on uroepithelium after Nd:YAG laser irradiation on the canine prostate. Br J urol 1997;80:903-7. 155 118. Schlick RW, Planz K. Potentially usefull materials for biodegradable ureteric stents. Br J Urol 1997;80: 908-10. 119. Shigemasa Y, Shibazaki K, Minami S, Matsuhashi A, Tanioka S, Shigemasa Y. Evaluation of chitin and chitosan for biomaterials. Biotech Genet engineer Rev 1996; 13: 383-420. 120. Smits M, Huibregtse K, Tytgat G. Results of the new nitinol self-expandable stents for distal biliary structures. Endoscopy 1995; 27: 505-8. 121. Marsman JW, Hoedmaker HP. Necrotizing fasciitis: fatal complication of migrated biliary stent. Australas Radiol 1996; 40: 80-3. 122. Benoit G, Blanchet P, Eschwege P, Alexandre L, Bensadoun H, Charpentier B. Insertion of a double pigtail ureteral stents for the prevention of uroligical complications in renal transplantation: a prospective randomized study. J Urology 1996; 156: 881-4. 123. Sauter ER, Hoffman JP, Hartz WH, Barber LW, Eisemberg BL. Atraumatic method of intraoperative retrograde transhepatic biliary stent insertion. J Surg Oncol 1996;62: 10-4. 124. Holmes SAV, Miller PD, Crocker PR, Kirby RS. Encrustation of intraprostatic stents: a comparative study. Br J Urol 1992; 69:383-7. 156 125. Costain DJ, Kennedy R, Ciona C, McAlister VC, Lee TDJ. Prevention of postsurgical adhesion with N, O-carboxymethyl Chitosan. Surgery 1997; 121: 314- 19. 126. Young GPH, Li PS, Gardner TA, Goldstein M. Animal models for microsurgical training and research. In: Goldstein M, Surgery of Male Infertility. Saunders Company 1995: 297-320. 127. Caldwell JC, McGadey J, Kerr R, Bennett NK, Mc Donald SW. Cell recruitment to the sperm granuloma, which follow vasectomy in the rat. Clin Anat 1996; 9: 302- 308. 128. Sanford P.A. Chitosan and alginate: new forms of commercial interest. Amer. Chem. Soc. Div. Polym. Chem. 1990;31 (1):628. 129. Trickett RI, Wang D, Maitz P, Lanfetta M, Owep, ER. Laser welding of vas deferens in rodens:initial experience with fluid solders. Microsurgery 1998;18:414- 418. 130. Suzuki K, Okawa Y, Hashimoto K, Suzuki S, Suzuki M. Protecting effect of chitin and chitosan on experimentally induced murin candidiasis. Microbiol immunol 1984; 28: 909-12. 157 131. Muzzarelli R, Tarsi R, Filippini O, Giovanetti E, Biagini G, Varaldo PE. Antimicrobial properties of N-carboxylbutyl chitosan. Antimicro Ag Chemotherap 1990; 34: 2019-23. 132. Chou TC, Fu E, Wu CJ, Yeh JH. Chitosan enhances platelet adhesion and aggregation. Biochem Biophys Res Commun. 2003 Mar 14;302(3):480-3. 133. Rao SB, Shorma CP. Use of chitosan as a biomaterial: studies on its safety and homeostatic potential. J Biomed Mater Res 1997; 34: 21-28. 134. Okamura T, Masui T, Taylor RJ. Evaluation of effect of chitosan in preventing hemorrhagic cystitis in rats induced by cyclophosphamide. Acta Urol Jpn; 41: 289- 96. 135. Lauto A, Cushway T, Dawes J, Piper J, Owem R. Laser nerve repair by solid protein band technique. I: Identification of optimal laser dose, power and solder surface area.”Microsurgery, 1998; 18:55-9. 136. Hendry G, Houghton JD. Natural food colorant, 2nd ed.; Chapman& Hall, 1996:pp 40-79, 310-341. 137. Aburada M, Takeda S, Sakurai M, Harada M. Pharmacological studies of gardenia fruit. Mechanisms of inhibitory effect of genipin on gastric acid secretion and its facilitatory effect on bile secretion in rats. J Pharmacobiodyn. 1980 Aug;3(8):423-33. 158 138. Suzuki Y, Kondo K, Ikeda Y, Umemura K. Antithrombotic effect of geniposide and genipin in the mouse thrombosis model. Planta Med. 2001 Dec;67(9):807-10. 139. Park JE, Lee JY, Kim HG, Hahn TR, Paik YS. Isolation and characterization of water-soluble intermediates of blue pigments transformed from geniposide of Gardenia jasminoides. J Agric Food Chem. 2002 Oct 23;50(22):6511-4. 140. Paik Y, Lee C, Cho M, Hahn T. Physical stability of the blue pigments formed from geniposide of gardenia fruits: effects of pH, temperature, and light. J Agric Food Chem. 2001, Jan; 49(1):430-2. 141. Sung HW, Huang RN, Huang LL, Tsai CC. In vitro evaluation of cytotoxicity of a naturally occurring cross-linking reagent for biological tissue fixation. J Biomater Sci Polym Ed. 1999;10(1):63-78 142. Sung HW, Chang Y, Liang IL, Chang WH, Chen YC. Fixation of biological tissues with a naturally occurring crosslinking agent: fixation rate and effects of pH, temperature, and initial fixative concentration. J Biomed Mater Res. 2000 Oct;52(1):77-87. 143. Sung HW, Huang RN, Huang LL, Tsai CC, Chiu CT. Feasibility study of a natural crosslinking reagent for biological tissue fixation. J Biomed Mater Res. 1998 Dec 15;42(4):560-7. 144. Chang Y, Liang HC, Wei HJ, Chu CP, Sung HW. Tissue regeneration patterns in acellular bovine pericardia implanted in a canine model as a vascular patch. J Biomed Mater Res. 2004 May 1;69A(2):323-33. 159 145. Chang WH, Chang Y, Lai PH, Sung HW. A genipin-crosslinked gelatin membrane as wound-dressing material: in vitro and in vivo studies. J Biomater Sci Polym Ed. 2003;14(5):481-95. 146. Liang HC, Chang WH, Lin KJ, Sung HW. Genipin-crosslinked gelatin microspheres as a drug carrier for intramuscular administration: in vitro and in vivo studies. 147. Fukaya F, Miyazaki M,.Pu H, Katayama B, Inoue Y, Ohashi R, Nakamura C, Namba M. Pyruvate alleviates toxic effect of ethanol on cells in culture. Arch Toxicol. 1997;71:651-654. 148. Tomas M, Lazaro-Dieguez F, Duran J, P Marin, Ranau-Piquers J, Egea G. Protective effect of lysophosphatidic acid (LPA) on chronic ethanol-induced ijuries to the cytosckeleton and on glucose uptake in rat astrocytes. J Neurochem. 2003;87:220-229. 149. Min J, Lee CW, Gu MB. Gamma-radiation dose-rate effects on DNA damage and toxicity in bacterial cells. Radiat Environ Biophys. 2003;42(3):189-92. 150. Lauto A, Stewart RB, Felsen D, Foster J, Poole-Warren L, Poppas DP. Low Temperature Solder for Laser Tissue-Welding. Proceedings SPIE Laser Florence, 2000; vol.5287: 83-90. 160 151. Huang LL, Sung HW, Tsai CC, Huang DM. Biocompatibility study of a biological tissue fixed with a naturally occurring crosslinking reagent. J Biomed Mater Res. 1998 Dec 15;42(4):568-76. 152. Suh DD, Schwartz IP, Canning DA, Snyder HM, Zderic SA, Kirsch AJ. Comparison of dermal and epithelial approaches to laser tissue soldering for skin flap closure. Lasers Surg Med. 1998; 22(5): 268-74. 153. Chang WH, Chang Y, Lai PH, Sung HW. A genipin-crosslinked gelatin membrane as wound-dressing material: in vitro and in vivo studies. J Biomater Sci Polym Ed. 2003;14(5):481-95. 154. Sung HW, Huang DM, Chang WH, Huang LL, Tsai CC, Liang IL. Gelatin- derived bioadhesives for closing skin wounds: an in vivo study. J Biomater Sci Polym Ed. 1999;10(7):751-71. 155. Liu BS, Yao CH, Chen YS, Hsu SH. In vitro evaluation of degradation and cytotoxicity of a novel composite as a bone substitute. J Biomed Mater Res. 2003 Dec 15;67A(4):1163-9. 156. Mi FL, Tan YC, Liang HC, Huang RN, Sung HW. In vitro evaluation of a chitosan membrane cross-linked with genipin. J Biomater Sci Polym Ed. 2001; 12(8):835-50. 161 157. Barrieras D, Reddy PP, McLorie GA, Bagli D, Khoury AE, Farhat W, Lilge L, Merguerian PA. Lessons learned from laser tissue soldering and fibrin glue pyeloplasty in an in vivo porcine model. J Urol. 2000 Sep;164(3 Pt 2):1106-10. 158. Forman SK, Oz MC, Lontz JF, Treat MR, Forman TA, Kiernan HA. Laser- assisted fibrin clot soldering of human menisci. Clin Orthop. 1995 Jan;(310):37-41. 159. Small W 4th, Heredia NJ, Maitland DJ, Da Silva LB, Matthews DL. Dye- enhanced protein solders and patches in laser-assisted tissue welding. J Clin Laser Med Surg. 1997;15(5):205-8. 160. Mueller MP, Kopchok GE, Tabbara MR, Cavaye DM, White RA. Argon laser- welded bovine heterograft anastomoses. J Clin Laser Med Surg. 1993 Feb;11(1):1-5. 161. Hino M, Ishiko O, Honda KI, Yamane T, Ohta K, Takubo T, Tatsumi N. Br J Haematol. Transmission of symptomatic parvovirus B19 infection by fibrin sealant used during surgery. 2000 Jan;108(1):194-5. 162. Kasper CK. Concentrate safety and efficacy. Haemophilia. 2002 May;8(3):161- 5. 163. Simhon D, Ravid A, Halpern M, Cilesiz I, Brosh T, Kariv N, Leviav A, Katzir A. Laser soldering of rat skin, using fiberoptic temperature controlled system. Lasers Surg Med. 2001;29(3):265-73. 162 164. Lauto A, Foster J, Ferris L, Avolio A, Zwaneveld N, Poole-Warren L. Albumin-Genipin Solder for Laser Tissue-Welding. Lasers Surg Med. 2004; 35(2):140-5. 165. Taravel MN, Domard A. Relation between the physicochemical characteristics of collagen and its interactions with chitosan: I. Biomaterials.1993 Oct;14(12):930-8. 166. Taravel MN, Domard A. Collagen and its interaction with chitosan. II. Influence of the physicochemical characteristics of collagen. Biomaterials. 1995 Jul;16(11):865- 71. 167. Ono K, Ishihara M, Ozeki Y, Deguchi H, Sato M, Saito Y, Yura H, Sato M, Kikuchi M, Kurita A, Maehara T. Experimental evaluation of photocrosslinkable chitosan as a biologic adhesive with surgical applications. Surgery. 2001 Nov;130(5):844-50. 168. Menovsky T, Van Den Bergh Weerman M, Beek JF. Effect of CO(2)-Milliwatt laser on peripheral nerves: part II. A histological and functional study. Microsurgery. 2000;20(3):150-5. 169. L. Heux, J. Brugnerotto, J. Desbrieres, M.-F. Versali and M. Rinaudo. ‘Solid State NMR for Determination of Degree of Acetylation Chitin and Chitosans’. Biomolecules, 2000, 1, 746-751. 163 170. Needleman IG, Smales FC, Martin GP. An investigation of bioadhesion for periodontal and oral mucosal drug delivery. J Clin Periodontol.1997, Jun;24(6):394- 400. 171. Naseef GS 3rd, Foster TE, Trauner K, Solhpour S, Anderson RR, Zarins B. The thermal properties of bovine joint capsule. The basic science of laser- and radiofrequency-induced capsular shrinkage. Am J Sports Med. 1997 Sep- Oct;25(5):670-4. 172. Simhon D, Brosh T, Halpern M, Ravid A, Vasilyev T, Kariv N, Katzir A, Nevo Z. Closure of skin incisions in rabbits by laser soldering: I: Wound healing pattern. Lasers Surg Med. 2004, Jul;35(1):1-11. 173. Moran K, Anderson P, Hutcheson J, Flock S. Thermally induced shrinkage of joint capsule. Clin Orthop. 2000 Dec;(381): 248-55. 174. McClain PE, Wiley ER. Differential scanning calorimeter studies of the thermal transitions of collagen. Implications on structure and stability. J Biol Chem. 1972 Feb 10;247(3):692-7. 175. Wall MS, Deng XH, Torzilli PA, Doty SB, O'Brien SJ, Warren RF.Thermal modification of collagen. J Shoulder Elbow Surg. 1999 Jul-Aug;8(4):339-44. 164 176. Azad AK, Sermsintham N, Chandrkrachang S, Stevens WF. Chitosan membrane as a wound-healing dressing: characterization and clinical application. J Biomed Mater Res. 2004 May 15; 6B(2): 216-22. 177. Stone CA, Wright H, Clarke T, Powell R, Devaraj VS. Healing at skin graft donor sites dressed with chitosan. Br J Plast Surg. 2000 Oct;53(7):601-6. 178. Hsu SH, Whu SW, Hsieh SC, Tsai CL, Chen DC, Tan TS. Evaluation of Chitosan-alginate-hyaluronate Complexes Modified by an RGD-containing Protein as Tissue-engineering Scaffolds for Cartilage Regeneration. Artif Organs. 2004 Aug;28(8):693-703. 179. Ylitalo R, Lehtinen S, Wuolijoki E, Ylitalo P, Lehtimaki T. Cholesterol-lowering properties and safety of chitosan. Arzneimittelforschung. 2002;52(1):1-7. 180. Gades MD, Stern JS. Chitosan supplementation and fecal fat excretion in men. Obes Res. 2003 May;11(5):683-8. 181. Rabea EI, Badawy ME, Stevens CV, Smagghe G, Steurbaut W. Chitosan as antimicrobial agent: applications and mode of action. Biomacromolecules. 2003 Nov- Dec;4(6):1457-65. 182. Guggi D, Langoth N, Hoffer MH, Wirth M, Bernkop-Schnurch A. Comparative evaluation of cytotoxicity of a glucosamine-TBA conjugate and a chitosan-TBA conjugate. Int J Pharm. 2004 Jul 8;278(2):353-60. 165 183. Chourasia MK, Jain SK. Design and development of multiparticulate system for targeted drug delivery to colon. Drug Deliv. 2004 May-Jun;11(3):201-7. 184. Bozkir A, Saka OM. Chitosan nanoparticles for plasmid DNA delivery: effect of chitosan molecular structure on formulation and release characteristics. Drug Deliv. 2004 Mar-Apr;11(2):107-12. 185. Nageris BI, Zilker Z, Zilker M, Kariv N, Feinmesser R, Katzir A. Esophageal incisions repair by CO2 laser soldering. Otolaryngol Head Neck Surg. 2004 Dec;131(6):856-9. 166 Publications, Presentations and Patents Arising from this Thesis 1. Lauto A, Hook J, Doran M, Camacho F, Poole-Warren LA, Avolio A, Foster LJ. Chitosan adhesive for laser tissue repair: In vitro characterization. Lasers Surg Med. 2005; 36(3):193-201. 2. Lauto A, Foster J, Ferris L, Avolio A, Zwaneveld N, Poole-Warren LA. Albumin-Genipin Solder for Laser Tissue-Welding. Lasers Surg Med. 2004; 35(2):140-5. 3. Lauto A, J Foster, A Avolio, Poole-Warren LA. Albumin-Genipin Solder for Laser Tissue-Welding. Proceedings of SPIE 5312, 124, 2004. 4. Lauto A, Foster J, Avolio A, Poole-Warren LA. Albumin-Genipin Solder increases tensile strength during laser tissue repair. Seventh World Biomaterials Congress, Sydney, 2004. 5. Provisional Patent # 2004 904 914: Bioadhesive for Tissue Repair. Inventors: Lauto A, Foster J, Poole-Warren LA. 167 Appendix The Laser System A.1 Introduction A diode laser was used in the previously described investigations along with absorbing dyes to activate albumin solders and chitosan adhesives. In this appendix, a brief introduction to diode lasers and optical fibres is given, followed by a characterisation of the diode laser employed in the experiments. A.2 Diode Lasers In a semiconductor, the electron population in the levels of the valence and conduction bands is determined with reference to the Fermi level, which represents the boundary between fully occupied and completely empty levels at T = 0 K. For non-degenerate semiconductors, the Fermi level is situated within the band gap. Electrons, pumped from the valence band to the conduction band by current for example, drop to the lowest levels in that band after very short time (~ 10-13s), thus leaving the top of the valence band full of 'holes'. This creates a population inversion between the valence and conduction bands. The electrons in the conduction band fall back into the valence band, recombining with holes, and emitting photons in the process. The process of stimulated emission due to the radiation from electron-hole recombination produces laser oscillation when the semiconductor is placed in a suitable resonator and the appropriate threshold conditions are fulfilled. 168 The semiconductor diode laser used in this thesis has a double heterojunction structure. In this structure there are two junctions [Al x Ga 1-x As (p) - Ga As and Ga As - Al x Ga 1-x As(n)] between different materials. The active region is a thin layer of Ga As (0.1 - 0.3 µm). Al x Ga 1-x As is a multiple quantum well material, consisting of two layers of AlAs alternating with layers of GaAs in the overall proportions x : 1-x. This structure leads to a modification of the electronic band structure and confinement in the quantum well structure improves electron transport properties. The threshold current density for room temperature operation is reduced by about two orders of magnitude (~ 103 A/cm2) compared with the homo junction device (Ga As - Ga As). This is because the refractive index of Ga As (n ~ 3.6) is significantly larger than that of Al x Ga 1-x As (n ~ 3.4), producing an optically guiding structure. Thus the laser beam is confined to the GaAs layers where the gain exists. Such structures are termed gain guided, and tend to exhibit poorer beam quality than index-guided structures. Since the bandgap in GaAs is less than in Al x Ga 1-x As, the injected holes and electrons are confined within the active layer. As a consequence, the concentration of holes and electrons in the active layer is increased and the gain is also increased. The Ga AlAs lasers emit between 780 - 850 nm and their maximum power is 1 - 3W in CW mode. The output power of semiconductor lasers has increased considerably in recent years due to technological advances in their fabrication. One aspect of this is the cooling system, which is more efficient and allows faster dissipation of the ohmic heat. As a consequence, the diode lifetime is increased and the laser wavelength is more stable. The other major advance is in growth methods. Molecular Beam Epitaxy (MBE) and Metalorganic Chemical Vapor Deposition (MCVD) growth techniques permit fine control of the device layers, avoiding impurities. Such techniques allow 169 fabrication of quantum wells in the laser diode active layer, increasing electrical-to- optical efficiency and lowering threshold current requirements. These characteristics are extremely important for high power lasers. Low threshold and high quantum efficiency allow increased optical power at the lowest possible electrical drive current, and thereby the lowest thermal load. The Diode laser emission wavelength (780 - 850 nm) falls in a region of relatively low attenuation in fused silica optical fibre, and diode laser radiation is often coupled into optical fibre, because of the convenience and efficiency of this approach. This makes such lasers particularly useful for medical or surgical applications, because of the flexibility and easy manipulability of the fiber. A.3 Optical Fibers Optical fibres are circular dielectric waveguides that can transport optical energy and information. They have a central core surrounded by a concentric cladding with slightly lower refractive index. Fibres are usually made of silica with index modifying dopants such as GeO2. One or two layers of cushioning material are used as a protective coating to prevent fibre twisting. The guiding properties of optical fibre may be modelled assuming multiple reflections due to total internal reflection. The sine of the largest angle an incident ray can have for total internal reflectance in the core is defined as the Numerical Aperture (NA). It is a measure of the acceptance light cone for the fibre. 170 The fibre geometry and composition determine the modes, which can propagate in the fibre. Guided modes are reflected in the core and propagate along the fibre. The fibre can simultaneously sustain more than one mode if the core diameter is large enough (multi mode operation). The modes exchange energy between them, travelling along the fibre, and therefore, the distribution of energy among the modes evolves with distance. In particular, energy can be coupled from guided to radiation modes, which are dissipative, by microbending and twisting of the fibre, thereby increasing the attenuation. Light power decays exponentially with propagation length in a fibre because of attenuation due to absorption and scattering losses. The small absorption losses of pure silica in the near infrared and visible regions are due to tails of absorption bands in the far infrared and ultraviolet. In standard optical fibres, the light polarization is not preserved. Laser light can be coupled into an optical fibre by appropriately positioning the fibre core in the focused laser beam. An efficient coupling is achieved for a multimode fibre (core diameter 50 - 200 µm) if the focused spot is comparable to the core size and the incident cone angle does not exceed the arcsine of the NA of the fibre. A.4 The Diode Laser Used in this Thesis The laser diode used in the LTR experiments is a model QSDL – 1500- 808FC (Qphotonics, Chesapeake, VA, USA), directly coupled with a multimode optical fibre. This diode is a double-heterojunction GaAs\ Al x Ga 1-x As structure and is gain guided with an aperture for the laser beam < 1 mm. The operating laser wavelength 171 ranged between 806 and 810 nm at 25 0C and the bandwidth was ~ 2 nm or 1000 GHz. Within this 1000 GHz band, there were about 10 longitudinal modes operating. The laser emission wavelength varies with the temperature, due to temperature related changes in the bandgap energy and refractive index. The wavelength decreases 0.23 to 0.30 nm for every 1 0C temperature decrease. The current driver of the diode was incorporated in the laser system that was air cooled to automatically stabilize the diode temperature and wavelength of operation. The QSDL – 1500- 808FC is a fibre coupled laser diode, with a permanently attached fibre pigtail. The multimode fibre, coupled with the diode, acts as an effective mode mixer. The output is a spatially incoherent, but temporally coherent beam, with a diameter governed by the fibre core diameter. The beam profile is shown in figures 1; it has a typical multimode structure with partial symmetry. A CCD camera recorded the image of the beam that was focused by a lens (f=3.5 cm), positioned 40 ± 0.2 cm apart from the fiber. The beam was attenuated by two neutral density filters before reaching the CCD camera, located 56 ± 0.2 cm from the lens. The multimode fibre has a core diameter of 200 µm and a NA of 0.22. The pigtail fiber was also coupled with another similar fibre via an FC connector to provide more free movement to the operator. The total length of the optical fibre used is ~1.5 m, which introduces negligible attenuation apart from the FC connector. The wavelength required for the LTR experiments was 808 nm, in conformity with the absorption characteristics of the dyes. 172 Figure 1. The multimode beam profile of the diode laser (l=808 nm) emerging from a 200 mm core optical fibre, ~1.5 m long, at a power level of 120 mW. The intensity scale is in arbitrary units. 173