<%BANNER%>

Optimized Sterilization and Decellularization of Small-Caliber Vascular Allografts while Preserving Matrix Integrity


PAGE 1

OPTIMIZED STERILIZATION AND DECELLULARIZATION OF SMALLCALIBER VASCULAR ALLOGRAFTS WHILE PRESERVING MATRIX INTEGRITY By DONNA M. K. SQUILLACE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by DONNA M. K. SQUILLACE

PAGE 3

iii ACKNOWLEDGEMENTS I would like to thank my family for all of their support and patience. My mother and stepfather, and my father and stepmother have stood behind me and picked me up at the lowest points. I only wish my father coul d be here with me as I achieve this goal. My sisters and brother have shown great support and have been understanding. My partner, Mindy, has been t horoughly supportive and patient a nd has seen me through all of the hard times and made this possible.

PAGE 4

iv TABLE OF CONTENTS Page ACKNOWLEDGEMENTS...............................................................................................iii LIST OF TABLES............................................................................................................vii LIST OF FIGURES.........................................................................................................viii ABSTRACT....................................................................................................................... xi CHAPTER 1 INTRODUCTION TO SMALL-CA LIBER VASCULAR ALLOGRAFTS...............1 Purpose and Specific Aims of Study............................................................................5 Specific Aims........................................................................................................6 Hypothesis 1...................................................................................................6 Hypothesis 2...................................................................................................7 Hypothesis 3...................................................................................................7 Background and Significance................................................................................8 Indications......................................................................................................8 Anatomy and physiology.............................................................................11 Bypass graft choices...................................................................................................14 Autograft..............................................................................................................14 Synthetic..............................................................................................................16 Allograft..............................................................................................................17 Allograft Processing............................................................................................19 Sterilization and disinfection........................................................................19 Cryopreservation..........................................................................................21 Antigenicity.........................................................................................................25 Use of immunosuppressant and anticoagulant therapies..............................27 Denudation of cell layers..............................................................................29 Decellularization..........................................................................................29 Expected Outcomes....................................................................................................38 2 DECELLULARIZATION OF VEIN GRAFTS.........................................................39

PAGE 5

v Introduction.................................................................................................................39 Acellular Allografts....................................................................................................41 Decellularization Process............................................................................................44 Cell Removal Assessment...................................................................................44 Biological Assessment.........................................................................................50 Purpose of This Study.................................................................................................57 Materials and Methods...............................................................................................57 Cell removal........................................................................................................57 Histology.............................................................................................................58 Immunohistochemistry........................................................................................58 Optical evaluation................................................................................................59 Matrix integrity....................................................................................................59 Statistics...............................................................................................................59 Results........................................................................................................................ .60 3 STERILIZATION.......................................................................................................68 Introduction.................................................................................................................68 Disinfection and Sterilization Processes.....................................................................70 Processing and Determination of Survivors...............................................................75 Materials and Methods...............................................................................................77 Reduction Curves (Survivor Curves) for Spores.................................................77 Methods........................................................................................................77 Hydrogen peroxide.......................................................................................78 Peracetic acid................................................................................................78 Spike and recovery method..........................................................................79 Reduction curves..........................................................................................80 Data Presentation.................................................................................................80 Matrix Integrity...................................................................................................80 Results........................................................................................................................ .81 Suspension Sterilization......................................................................................81 Tissue Sterilization..............................................................................................84 4 PRESSURE-INDUCED INCREASE IN PORE VOLUME FOR INTRODUCTION OF DECELLULARIZATION AGENT INTO VEIN ALLOGRAFTS..........................................................................................................88 Derivation of Governing Equations............................................................................93 Materials and Methods.............................................................................................101 Histology...........................................................................................................102 Immunohistochemistry......................................................................................102 Optical Evaluation.............................................................................................103 Results.......................................................................................................................1 03 5 SUMMARY AND CONCLUSIONS.......................................................................105 Study Limitations......................................................................................................113

PAGE 6

vi Future Studies...........................................................................................................113 APPENDIX A DONOR INCLUSION LIST....................................................................................116 B MAJOR HISTOCOMPATIBILITY COMPLEX (HLA CLASS I AND CLASS II) IMMUNOHISTOCHEMISTRY PROTOCOL...................................................117 C ANTIGEN LEVELS FOR TRYPSIN......................................................................120 D ANTIGEN LEVELS FO R TRITON-X 100.............................................................122 E ANTIGEN LEVELS FOR SO DIUM DEOXYCHOLATE.....................................124 REFERENCE LIST.........................................................................................................126 BIOGRAPHICAL SKETCH...........................................................................................136

PAGE 7

vii LIST OF TABLES Table page 1-1. Features of typical arteries and veins........................................................................14 2-1. Processing parameters to evaluate decellularization of human saphenous vein.......58

PAGE 8

viii LIST OF FIGURES Figure page 1-1. The venous anastomosis in an end-to-side fashion for AV access..............................9 1-2. Types of composite sequential grafts........................................................................11 1-3. General structure of vessel walls...............................................................................12 1-4. Algorithm for determination of wh at graft to use in a distal bypass.........................16 2-1. Positive (left) and nega tive (right) control samples...................................................60 2-2. Result of the application of a threshold filter that lim its the pass of high intensity blues.........................................................................................................................6 1 2-3. The amount of class I MHC remaining in the tissue when compared to a negative control for trypsin and trypsin plus EDTA...............................................................62 2-4. The amount of class I MHC remaining in the tissue when compared to a negative control for Triton-X 100 and sodium deoxycholate.................................................62 2-5. The amount of class I MHC remaining in the tissue when compared to a negative control for Triton-X 100 and sodium deoxycholate combined................................63 2-6. Representative samples of tissue treated with Trypsin (left), Triton-X 100 (center) and sodium deoxycholate (right) and stained for class I MHC antigens....64 2-7. A comparison of the amount of cellularity (blue) visibl e in histology compared to the amount of antigens (red) present in the tissue. The samples were treated with 0.05% Trypsin at 37oC for 6 hours...................................................................65 2-8. A comparison of the amount of cellularity (blue) visibl e in histology compared to the amount of antigens (red) present in the tissue. The samples were treated with 0.05% Trypsin at 37oC for 24 hours.................................................................65 2-9. A comparison of the amount of cellularity (blue) visible in hi stology compared to the amount of antigens (red) present in the tissue. The samples were treated with 0.25% Cholate at 37oC for 6 hours...................................................................66

PAGE 9

ix 2-10. A comparison of the amount of cellularity (blue) visible in histology compared to the amount of antigens (red) present in the tissue. The samples were treated with 1.0% Cholate at 48oC for 6 hours.....................................................................66 2-11. Results of the collagen degrada tion assay after decellularization at 48oC for 24 hours.........................................................................................................................6 7 3-1. Bacillus stearothermophilus growth inhibition at diffe rent dilutions of hydrogen peroxide on TSA and blood agar plates...................................................................79 3-2. Bacillus stearothermophilus growth inhibition at differe nt dilutions of peracetic acid on blood agar plates..........................................................................................79 3-3. Survivor curve for 6% peroxide at 50oC resulting in a D-value of 9.87...................82 3-4. Reduction curve for 6% peroxide at 45oC resulting in a D-value of 23.0.................82 3-5. Reduction curve for 6% peroxide at 40oC resulting in a D-value of 47.39...............83 3-6. Linear regression of the log of the D-values at each temperature.............................83 3-7. Reduction curve for 0.05% PAA at 40oC resulting in a D-value of 1.90..................84 3-8. Survivor curve for 0.005% PAA at 40oC resulting in a D-value of 6.45..................84 3-9. Reduction curve for 6% hydrogen peroxide at 50oC resulting in a D-value of 6.63........................................................................................................................... 85 3-10. Reduction curve for 6% peroxide at 40oC resulting in a D-value of 15.17.............86 3-11. Collagen degradation of tissue trea ted with 6% hydrogen peroxide and 0.1% PAA at 50oC.............................................................................................................87 3-12. Collagen degradation of tissue trea ted with 6% hydrogen peroxide and 0.05% PAA at 40oC.............................................................................................................87 4-1. Analysis of the data to compare th e effects of pressure-induced stretch on the concentrations of macromolecules, LD L and albumin in the rabbit aorta...............91 4-2. Average profiles of rela tive concentrations of LDL a nd albumin as a function of distance from the lumen...........................................................................................91 4-3. Average profiles of re lative concentrations of LD L (a) and albumin (b) as a function of distance from the lumen........................................................................92 4-4. Transverse sectional vi ew of a blood vessel wall......................................................95 4-5. The approximation of the vascular wall structure with the fluid-wall model...........99

PAGE 10

x 4-6. Electrical analog of the filtration for the membrane (top-left) and for the wall (top-right). Electrical anal og of the mass transfer processes for the membrane (bottom-left) and for the wall (bottom-right).........................................................100 4-7. Histology of a vein pressurized at 100 mm Hg with 0.25% sodium deoxycholate.104 4-8. The percentage of tissue stained re d after treating at 0, 100 and 200 mm Hg for one to two hours at 37oC........................................................................................104 C-1. Vein segments treated with trypsin.........................................................................121 D-1. Vein segments treated with Triton-X 100..............................................................123 E-1. Vein segments treated with sodium deoxycholate..................................................125

PAGE 11

xi Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy OPTIMIZED STERILIZATION AND DECELLULARIZATION OF SMALLCALIBER VASCULAR ALLOGRAFTS WHILE PRESERVING MATRIX INTEGRITY By Donna M. K. Squillace December 2005 Chair: William L. Ditto Major Department: Biomedical Engineering Vascular disease is one of the most fre quent diseases worldwide and is a major public health priority causing coronary and pe ripheral occlusions leading to either heart attacks or limb amputations. Therapy involves bypass grafting with autogenous saphenous vein. More than 30% of patients have insufficient tissue typically requiring multiple revisions, exhausting usable autoge nous vessels. A viable alternative with similar benefits to the patient’s own tissue is a vascular allograft. The disadvantages to allografts are early stenosis and graft failu re related to a graft rejection response. Currently, endogenous cells are preserved to maintain functionality and prevent occlusions. These cells contai n antigens that elicit an immu ne response that may lead to thrombosis, intimal hyperplasia and eventual graft occlusion. Allograft tissue is processed aseptically, but is not considered sterile. Current methods of disinfection are in adequate resulting in a signi ficant discard rate due to positive cultures, hindering the supply which is already limited and not meeting the

PAGE 12

xii demand. Additionally, donor tissue carries a sign ificant risk of disease transmission due to false negative tests, human errors, and undetectable window periods that cannot be prevented through aseptic processing. This study addressed the issues of ster ilization and antigenicity through liquid chemical processing under controlled conditions Oxidants were investigated for their effect on reducing B. stearothermophilus a spore-forming organism, on spiked tissue while maintaining matrix integrity. Peracetic acid showed rapid inactivation rates at mild conditions and low concentrations that were favorable for tissue integrity. Hydrogen peroxide was less effective but has highly diffu sive properties, allowing deep penetration within the tissue. Combining the two oxi dants at mild conditions can achieve a significant reduction in bacteria while maintain ing the matrix integrity. The reduction in antigen laden material was investigated thr ough the use of detergents and enzymes under controlled conditions. Sodium deoxycholate achieved almost 100% reduction in antigen laden material but showed adverse effects on the matrix. Triton-X 100 was less efficient but had no effect on the matrix. Sodium deoxycholate at pressures above 0 mm Hg showed improved reductions. This study found the optimal treatment for a sterile, decellularized graft with the combination of chemicals at elevated temperatures and pressures.

PAGE 13

1 CHAPTER 1 INTRODUCTION TO SMALL-CALIBER ALLOGRAFTS Vascular disease is one of the most fr equent diseases worldwide and with the increasing age and longevity of the Ameri can population it will remain a major public health priority [1]. The mainstay of ther apy for patients with co ronary and peripheral occlusions is surgical bypa ss grafting with the patient’s own (autograft) saphenous vein or internal mammary artery. This procedure offers important benefits to the patient but incurs significant economic cost on the nationa l level. Commonly, insufficient autograft material is available and a patient may rece ive cadaver tissue (allograft). Unfortunately, allograft conduits have poor patency rate s resulting in numerous revision surgeries, amplifying the economic impact. A reduction in graft patency is a result of an immunological response to cellular material co ntained within and on the allograft leading to intimal hyperplasia and eventual graft occlus ion. The aim of this study is to develop a process to create an inert scaffold out of human saphenous vein in order to improve patency rates. By complete removal of all the endogenous material the tissue becomes nonantigenic and may allow recipient cell re population to occur immediately following implantation, bypassing the initial graft rejecti on response typically seen in allograft use. Allografts are plagued with complications le ading to early stenos is and graft failure thought to be related to an immune reaction si milar to graft rejection. Adaptation of vein

PAGE 14

2 allografts to the arterial environment has b een studied extensively in animal models and less in humans [1]. After implantation, the conduits undergo structural changes characterized by intimal hyperplasia and overa ll wall thickening. Ce llular events that occur after implantation can a ffect the patency by occluding the conduit. Occlusion or stenosis in coronary or peripheral circulati on is a common clinical occurrence present in small-caliber applications that necessitate s an additional intervention. Problems that cause occlusion of a graft are thrombosis a nd neointimal hyperplasia [2, 3]. Thrombosis occurs when platelets in circulating blood adhere to certain surfaces, then release chemicals to attract more platelets to form a large aggregate that generates thrombin [4]. Grafts most commonly fail due to the devel opment of fibrous intimal hyperplasia where there is an excess proliferati on of smooth muscle cells. This flow-restricting lesion may occur diffusely throughout the graft or, more commonly, at focal sites near anastomoses, particularly in compliance mismat ched synthetic grafts [1]. Current processing methods only treat th e tissue without rem oving a substantial amount of endogenous material that may harbor contaminants, such as blood and lipids, as well as antigen containing cells and cellular de bris that elicit an immune reaction. Processing methods and cryopreservation pres erve and retain endothelial cells and smooth muscle cells (SMC), both containi ng antigens that may require matching blood types for vascular reconstruc tions. Patients in need of an allograft must wait for a matching donor, and the waiting time for a compa tible (ABO relevant) graft depends on the rarity of the recipient’s blood type [5]. Rejection plays a significant role in failu re and leads to allosensitization [6]. Therefore, a great deal of effort is placed in determining an ABO-compatible donor as

PAGE 15

3 well as lymphocyte cross-matching [5]. The intimal layer of the conduit contains the antigen-present endothelial cells responsible for ABO relevancy. Stripping this layer of cells from the conduit removes the need fo r donor matching, but it is believed by some that a lack of surviving cells and endoth elial integrity may play a role in graft degeneration and early and late patency [5]. If injured and subconflu ent endothelial cells line the lumen, thrombosis and SMC growth is promoted leading to intimal hyperplasia [7]. Additionally, a process by which the grafts are decellularized will eliminate the need for blood-typing, reduce platelet activation, and prevent smoot h muscle cell proliferation that leads to neointimal hyperp lasia resulting in thrombosis. In the case of multiple revisions, insufficien t autograft material, or infection due to synthetic material, a vascular allograft is a medical necessity for life and/or limb saving operations. Currently, the demand for vascular allografts far exceeds the supply, forcing surgeons to pursue less favorable alternatives. Allograft tissue is obtained from cadavers and processed under aseptic conditions so it cann ot be considered a tr uly sterile conduit. Current processing at an AATB (American Association of Tissue Banks) accredited tissue bank uses an antimicrobial soak can resu lt in a substantial discard rate (27%) due to positive first and final cultu res from bacterial and fungal c ontamination resistant to the cocktail aimed at achieving sterility (persona l communication). Additionally, the risk of transmission of viruses and other pathogens can be reduced but not eliminated through donor screening. Therefore, vasc ular transplantation carries a significant risk of disease transmission to the recipient considering the occurrence of false negative test results, human errors in screening and processing, and virological testi ng within the window period of contraction and detec tion of diseases. More signif icantly, over the last several

PAGE 16

4 years multiple examples of donor to host disease transmission have been documented. These examples include transm ission of hepatitis C, b acteria, and fungi, and have resulted in serious morbidity and even mortal ity [8]. Unfortunatel y, traditional methods of sterilization, such as irra diation and ethylene oxide, have significant deleterious effects on tissue integrity that limit or pr eclude their use as allograft. The marketplace of surgeons acknowledges the inherent risk in dealing with human tissue and tends to exhaust autograft material instead [3]. Unfortunately, patients who do not have sufficient or acceptable autografts, or are unable to receive synthetic grafts are forced to resort to the use of allografts. To that end, the FDA has ta ken an interest in how human tissue might be held to a higher standard [9]. Therefore, a sterilization processes that is effective in inactivating endogenous ma terials as well as ach ieving a significant reduction in all possible orga nisms without adversely aff ecting the tissue matrix and functionality of the vessel would increase th e availability of allograft by reducing the discard rate and provide a safe alternative. The factors that determine a successful graft include the fo llowing; no aneurysms or dilations, no immunological reaction, no disease transmi ssion, no infection and longterm patency [10-14]. Once placed in arte rial circulation, it must be capable of withstanding long-term hemodynamic stress without mechanical failure. A failure of this type could be catastrophic and lead to morbidity, such as loss of limb, or even mortality. The availability, suturability, and simplicity of handling are desirable for minimizing operating time, risk and expense as well as long-term durability. Postimplantation, the graft should be fully biocompatible, resist ant to thrombosis and infection and be completely incorporated by the body to yield a neovessel resembling the native artery in

PAGE 17

5 structure and function [1, 6]. The graft should be porous enough to permit ingrowth of host’s cells, but still dura ble and suitable to maintain anastomosis integrity. Purpose and Specific Aims of Study The purpose of this research is to prov ide a process that can achieve complete removal of the cells containe d within the vascular tissue matrix and achieve a 6 log reduction in all possible organisms to reduce immune reaction and ensure the safety of the patient. Standard validation of sterilization processes introduce a 106 count of organisms to be inactivated to provide a f actor of safety above what is typically encountered in tissue contamin ation. Therefore, a 6 log reduction would result in complete inactivation of the inoculum. Hu man saphenous vein samples were introduced to a sequence of chemicals with mechanical stimulation in order to sterilize and decellularize the tissue. Since it is difficult to predict frequencies and concentrations of pathogens and infection windows may exceed the time between contraction and the time of death, the most difficult to inactivate spor e will be used to achieve the sterility assurance. The level of decellularization will be investigated with an immunohistochemistry stain for antigens locate d on the cell membranes within the tissue. The processing conditions will leave a neutra lized scaffold with minimal detrimental biochemical effects on the extracellular matr ix, no need for blood typing, and optimal biological functionality. Specimens used for this study will be donated human saphenous veins that are designated by informed consent to be used for research. The primary sources of the tissue will come from RTI Cardiovascular in Birmingham, AL and Regeneration Technologies, Inc. (RTI) owned Recovery Agencies. This study will use Greater Saphenous Veins not accepted for implantation, mostly because the inner diameter is out

PAGE 18

6 of specification (<3mm), but accepted for researc h. Donor tissue will be enrolled in this study only if it has consent for research, negati ve serological tests and medical director sign off at RTI. Only saphenous veins will be used since it is the most common allograft conduit currently used and relatively easy to recover since it is a superficial vein. Specific Aims Hypothesis 1 Through the use of controlled temperature in combination with measured enzyme and detergent concentrations, complete dece llularization will be achieved with minimal tissue damage. Additionally, the use of a chelating agent will enhance cell removal. Chapter 2 provides the results of the investigation of the use of detergents and enzymes at various conditions in order to ac hieve a completely decellularized scaffold while maintaining the matrix integrity. The parameters investigated will be chemical concentration, temperature and time. Sonication is also investigated as an alternative to shaking as a means of agitation. The rema ining cells and cellular debris will be investigated with histology and immunohist ochemistry. Using an immunohistochemical staining method for class I major histocompatible (MHC) antigens, the level of decellularization will be measur ed as the area fraction of red stained tissue. It will be shown that antigens still remain even though the tissue appears to be acellular. The effect on the extracellular matrix will be measured wi th an enzyme digestion assay that targets denatured collagen. If denatured collagen is present it may indicate later graft rupture or dilation or activation of enzy mes known to further degrade collagen postimplantation.

PAGE 19

7 Hypothesis 2 By increasing the temperature and increasi ng the concentration of oxidants used to inactivate bacteria, with the addition of s onication will decrease the time needed to achieve a 6 log reduction in a biological indicator. Since the frequencies and concentrations of various pathogens and polymicrobial infections on tissue remain unknown, it is difficu lt to predict what le vel of sterility is achievable. Using a biological indicator that is more difficult to inactivate than contaminants that may be detected on vasc ular tissue provides a process that can be applied across all possible microorga nisms. In Chapter 3, I chose Bacillus stearothermophilus which is a spore used for instrume nt sterilization si nce it is highly resistant to temperature and di fficult to kill. Suspensions of the spore are treated with hydrogen peroxide and peracetic acid at various concentra tions and temperatures to provide a baseline for tissue sterilization. Saphe nous vein segments were spiked with an inoculum of spores and treated in similar cond itions. The level of kill was measured with a plate count method where a Petri dish contai ning a growth media is inoculated and the surviving spores are counted as colonies. Hypothesis 3 Distension of venous tissue by pressurizing the lumen will expand the pore volume and allow more efficient penetration of deterg ents and reducing the time required to reach complete decellularization. Preliminary results from this study are cont ained in Chapter 4. The significance of a reduction in contact time is that the surfaces of the vein are in cons tant contact with the treatment agent, increasing the chance of dama ge to these areas while the agent is trying to diffuse into the matrix to reach the cells deep within. Additionally, a lower

PAGE 20

8 temperature was used to provide further stabil ity to the matrix. By applying pressure to the lumen, the extracellular matrix expands and allows for a volume flux. Saphenous vein segments are treated with a detergent at two pressures and for two time points. The analysis was conducted as was in Chapte r 2 with an immunohistochemical stain. Background and Significance Indications Small caliber grafts are used in bypass and revascularization procedures in the coronary and lower extremity circulation a nd for arteriovenous access for hemodialysis and cancer patients. Patients most affected w ith reconstructions are those with end stage renal disease (ESRD) requiring vascular access. The prevalence and numbers of patients that required hemodialysis have risen st eadily over the past two decades. The 1999 annual report of US Renal Data System (USRDS) reported more than 30,000 ESRD patients in the US undergo maintenance hemodi alysis annually [15]. It is the most common vascular operation in the country toda y with the leading cause of morbidity in ESRD patients related to vascular acce ss placement and resultant complications. The surgical procedure and graft prepar ation for AV access is shown in Figure 1. The femoral vein at a length less than 25 cm a nd diameter of 5 – 7 mm is usually used for this procedure. The multiple access operations required by each patient are a result of the poor overall patency rates of hemodialysis access. The increased number of AV graft operations versus primary AV fistulas is due to the increased age of the patients lacking superficial venous anatomy [15]. Multiple re visions cause a reduction in available access sites in the upper extremities, creating a n eed for alternatives for continuous access.

PAGE 21

9 Figure 1-1. The venous anastomosis in an end-to-side fashion for AV access [15]. A traditional coronary artery bypass graft (CABG) surgery is estimated to cost $20,000 and within ten years, it is estimated that nearly half of saphe nous vein grafts will require retreatment [16, 17]. In the U.S ., over 400,000 CABG procedures are performed annually with 20% being re-operations, often due to the use of unsuitable saphenous vein allografts [18]. In 2001, about 80% of bypass gr afts were left internal thoracic arteryLAD plus saphenous vein, 5 to 10% were bi lateral internal thor acic arteries plus saphenous vein, and 1% was allarterial conduits. Due to the invasiveness of this procedure, allograft is not popul ar with cardiac surgeons. In the lower extremities, critical limb ischemia is a result of femoropopliteal occlusive disease that obstru cts the blood flow into the lo wer limbs [4, 5, 11, 12, 19, 20]. It affects 10% of the population in the U.S. over 70 and 1 to 2% from 37 to 69 years of age [4]. About 82,000 people a year have di abetes-related leg and foot amputations which account for almost half of the non-trauma tic amputations in the U.S. [21, 22]. The average cost for primary amputation is approximately $40,000 with an additional $30,000 for rehabilitation and related medical surgic al care [23]. Femo ropopliteal disease is caused by atherosclerosis, which is a form of arteriosclerosis. Pla que from degenerated

PAGE 22

10 thickened arterial intima containing cholestero l lipoid material and lipophages is formed within the intima and inner media of arteries [19]. This may result in structures that impair flow and cause acute thrombosis. Pl aque that is removed from the wall during flow, called emboli, may lodge at sites of bifu rcation or trifurcation causing occlusion in the lower limbs. Occlusion occurs in the di stal common artery (34% ), distal popliteal artery (14%), or tibial arteries [4]. The vessels eventually become thrombosed due to turbulent flow [4, 24]. At early stages, occlusion forces blood to perfuse the limb via collateral pathways, thus compromising the circulation. Blood flow is sufficient at rest, but as activity increases, the supply of oxygen to leg mu scles is inadequate causing intermittent claudication [4]. As the disease progresses, outflow may be impaired even at rest, resulting in severe ischemia pain, tissue necr osis (gangrene) or cl audication resulting in the need for intervention [3, 5, 12-14, 20, 24-27] Lower extremity atherosclerosis is a marker for systemic atherosclerosis diseas e with significant incidence of systemic morbidity. About 30% of those affected will succumb to systemic atherosclerosis within five years [4]. Approximately 100,000 vascular reconstruc tive procedures for limb salvage are performed yearly in the United States [4]. Numerous studie s show the effectiveness of infrainguinal revasculariza tion in producing long-ter m patency and improved limb salvage rates [26]. The pro cedure involves increasing path ways and runoff for inflow from the common femoral artery to lower limb vessels (Figure 1-2) [12]. The different procedures involve above knee bypasses fr om the common femoral to the popliteal artery, or below knee bypasses from the common femoral artery to either the popliteal or

PAGE 23

11 tibial artery or from the superficial femoral artery to the popliteal artery [10, 12]. The choice of an above or below knee bypass proced ure depends on the extent of the popliteal artery disease [14]. Autologous saphenous vein is the graft of choice for critical limb ischemia, but up to 45% of the patients do not possess usable veins [5, 6, 26]. The need for alternative conduits has become increasi ngly evident consider ing the number of multiple procedures performed to salvage a single extremity [6, 12, 26]. Figure 1-2. Types of composite sequential graf ts: (A) bypass to the popliteal and one of the tibial arteries, (B1, B2) bypa sses to tibial ar teries [10]. Anatomy and physiology Blood vessel walls consists of three layers of which the proportions vary according to the size of the vessel (Figure 1-3) [28, 29] The inner most layer is known as the intima or tunica interna and is made up of an endothelial layer and basal lamina. The endothelial layer lines the lumen with the underlying connectiv e tissue containing variable amounts of elastic fibe rs [19, 28, 29]. This layer cons ists of flat cells that are metabolically active and produce a number of compounds that affect the vascular lumen diameter and control activati on of platelets [19, 28]. The cells act as a barrier that

PAGE 24

12 controls the entry of substances into the wall, but allows some to pass via transport system. The compounds secreted contro l blood coagulation and relaxation and contraction of the smooth muscle cells in the middle layer [19]. The basal lamina (basement membrane) is a thin layer of noncellular material with a filamentous texture underlying the endothelium. The principal component of this connective tissue is collagen, typically type IV [ 19, 28]. Separating the intima from the middle layer is a thick layer of elastic fibers called the internal elastic membrane [29]. Figure 1-3. General structure of vessel walls [29]. The intern al elastic lamina is present in veins but is less prominent as in arteries and sometimes not detectable. The middle layer is the media or tunica media. It is the thickest layer of the arteries and is smaller in their vein counterparts. It consists mostly of smooth muscle cells oriented in concentric sheets within connective tissue. Th is layer is involved in the

PAGE 25

13 contracting and relaxing of th e tissue and contains elastic sh eets and bundles of collagen fibrils, each of which contributes to the mech anical behavior of the vessels [28]. The amount of each substance affects the ultimat e mechanical properties and distinguishes arteries from veins. The collagen fibrils ar e made up mostly of type III with some type I and a trace of type V. The outer layer is divided from the media by the external elastic membrane which is a thin ba nd of elastic fibers [28, 29]. The outermost layer is the adventitia or tu nica externa. It forms a connective tissue sheath around the vessel connecting it to th e body forming the vascular network. It consists of mainly collagen fibers (simila r types and proportions as the media) and ground substance with scattered bands of elasti n fibers [30, 31]. Also contained within the tissue layer are fibroblasts, macropha ges, minute blood vessels (vasa vasorum), myelinated and nonmyelinated nerves. The similarities and differences in veins and arteries can be seen in Table 1-1. The relative wall thickness of veins is lower than in arteries. While the media is the largest layer in the arteries, the advent itia is the largest layer in the veins w ith a large amount of collagen, networks of elastic fibers and bundles of smooth muscle ce lls. The vein sees low pressures (5 to 75 mm Hg) compared to th e artery (80 to 120 mm Hg), therefore the media is thin with relatively few muscle cell s and very little elas tic tissue [28, 29]. The intimal layer of veins contains valves to prev ent backflow since the pressure in the veins is so low they cannot oppose gravity.

PAGE 26

14 Table 1-1. Features of typical ar teries and veins [29]. Feature Typical Artery Typical Vein General appearance in sectional view Usually round, relatively thick wall Usually flattened or collapsed with relatively thin wall Intima Endothelium Internal elastic membrane Usually rippled due to vessel constriction Present Often smooth Present may be poorly developed Media External elastic membrane Thick, dominated by smooth muscle and elastic fibers Present Thin, dominated by smooth muscle and collagen fibers Absent Adventitia Collagen and elastic fibers Collagen, elastic, smooth muscle fibers The muscular and elastic components permit controlled alterations in the diameter of the vessel as blood pressure or blood volum e changes. Arteries have thicker walls with the media containing more smooth muscle and elastic fibers. The collagen fibers are crimped while the elastic fibers recoil and restrict the lume n when not opposing pressure generated by the pumping of the heart to reta in its cylindrical sh ape. The endothelial lining shows pleats in a sectional view since it cannot contract along with the elastin. Veins have larger diameters and thinner walls than their corresponding arteries. Due to the lesser amount of elastin present the walls of the veins collapse when excised. Bypass graft choices Autograft The “gold” standard for vascul ar repair is the autograft. The saphenous vein is the most common small-caliber (<5 mm) graft for coronary bypass, arteriovenous access (AV), and lower-extremity reconstructions. Autograft has continuously demonstrated to

PAGE 27

15 perform better than all nonaut ogenous grafts. According to DeWeese [32], the advantage of autograft is its long-term patency even de spite poor outflow conditions. It is believed this may be attributed to the smooth inner lining or pseudointima, which appears almost immediately in vein autografts as opposed to allografts a nd, to a lesser degree, synthetic grafts [12, 24, 32]. Lower extremity, long-term results for infrapopliteal and even pedal arteries are excellent wi th patency at 5 years of 50 to 80% [1, 10, 20]. Unfortunately, up to 45% of patients seen with critical limb ischemia do not possess a usable greater saphenous vein and requ ire postoperative revision surgery in four to five years [1, 3, 10-12, 20, 24, 26]. Some pr oblems that preclude the use of an allautogenous graft are a possible si ze mismatch between the donor site vessel of the patient and the vessel to be repaire d, insufficient length or diameter prior use, previous graft problems such as infection in that site, fibrosis, or venous disease. As the number of procedures increase, as they do for patients with atherosclerosis and ESRD, the supply of autogr aft is greatly reduced. Ma ny patients lacking autograft with recurrent coronary dis ease are considered inoperable and those with distal lower extremity disease may suffer loss of limb [1 ]. Therefore, graft failure and conduit availability are important limitations for aut ograft uses and motivation for alternatives. Researchers and surgeons work to develop vi able alternatives such as prosthetics, composite prosthetic and aut ogenous grafts, allografts, and possibly xenografts [3, 10-12, 14, 24]. Figure 1-4 depicts an algorithm for dete rmining the type of graft to use in lowerextremity revascularization. The surgeon w ill perform the surgery with the ipsilateral greater saphenous vein or the contralatera l greater saphenous vein. Other autogenous options are the upper arm veins (cephalic and basilic), the lesser saphenous vein, or a

PAGE 28

16 composite of vein segments [14, 26]. Ectopi c (i.e., lesser saphenous and arm veins) or composite vein grafts are generally inferior to autogenous greater sa phenous vein grafts (40 – 60 % patency at 5 years) t hough still superior to synthetic grafts [1]. The point that should be taken from this is that surgeons will exhaust all possibl e autograft material before resorting to synthetics or allograft tissu e since the current state of these grafts has low rates of success. Figure 1-4. Algorithm for determination of wh at graft to use in a distal bypass [3]. Synthetic These patients without adequate autologous tissues may then be considered for an all-synthetic or a prosthetic -vein composite replacement, with materials such as polyurethanes, Dacron (polyethylene tetephtha te) or expanded polytetrafluoroethylene (ePTFE) [2, 33]. Synthetic material has the advantage of being an off-the-shelf product and is not a limited supply. For large-calib er arterial reconstr uctions, the long-term results for replacement of common femoral arte ries for either aneurismal or occlusive disease with available synthetics are generally excellent. However, pr osthetic grafts have proved to be unfavorable as small-caliber ar terial substitutes in demanding, low-flow environments resulting in occl usions [1, 2]. Synthetics have been used with moderate

PAGE 29

17 success in lower extremity bypasse s to the popliteal artery and some success to tibial vessels, but due to its inflexibility the eff ectiveness of below knee revascularization is inferior to all-autogenous graf ts used in a majority of st udies [26, 33]. The four year patency rate for ePTFE is only 12 to 14% in infrapopliteal and ao rta-coronary bypasses, and they are reconstructed twice as ofte n as autologous for arteriovenous (AV) access [15, 34]. Complications, such as infection and occlusion, lead to the need of graft removal, intravenous antibiotic treatment, debridement and drainage, placement of a catheter or an AV graft replacement. Due to multiple revisions in hemodialysis patients for AV access, available sites in the upper extremities become scarce. Some synthetics show good biocompatib ility, strength and deformation but eventually creep and lead to the developmen t of an aneurysm [33]. The most common failure modes for synthetic grafts are thrombosis due to the adhesive na ture of cells to the synthetic surface and development of focal anas tomotic strictures because of neointimal hyperplasia resulting from compliance mismat ch and resistance to incorporation, and poor outflow [2, 3, 15, 20]. In general, the gr afts never completely heal and have limited re-endothelialization mostly in the area of pa nnus ingrowth adjacent to anastomoses [1]. Additionally, infection is a co mmon problem with synthetics which causes sensitization and precludes the patient from revisions with synthetics. A major clinical problem with infection is it leads to high mortality rates, excessive bleeding, aort ointestinal fistulas, and major amputations of limbs [34]. Allograft Allograft is tissue taken from a donor of th e same species. Vascular allografts are often a medical necessity for patients who are either needing re vision surgery, do not have sufficient saphenous autografts, or are unable to receive synt hetic grafts due to

PAGE 30

18 sensitization or risk of renal fa ilure. In the early days of vasc ular surgery, allografts were used but were soon abandoned for poor long-term success due to the high rate of aneurysm formation and dissections of vessels with secondary dilatation and calcification due to biodegeneration [1, 5, 25]. Improvement s in recovery techniques, storage, and cryobiology revived their use. Allografts provide the advantage over autografts of reduced operating and anesthesia time, reduced operative trauma, limited incisions and reduced wound complications, all of which ar e important to consid er since the patients tend to be from the elder population [5, 6]. Hemodialysis patients frequently need a replacement graft as treatment for an infected arteriovenous (AV) acce ss graft [15]. Add itionally, SV allografts have been used to save limbs in below knee revascul arizations due to flexibility, compliance matching, size matching, and resistance to infecti on [25]. Vascular al lografts have been promoted by some studies as the best a lternative conduit for in frainguinal arterial reconstruction in infected fields when auto logous veins are unsatisfa ctory or unavailable [6]. For the coronary circulation, the inte rnal mammary artery (IMA) and the radial artery are used as bypass a llografts. The IMA has been shown to have long-term patency, but availability is se verely limited due to its short lengths and difficulty in recovering [1]. As mentioned above, two commonly addr essed disadvantages to the use of allograft tissue are the risk of contamination or disease tr ansmission and early to late occlusions. Good tissue banking practices ut ilized to prevent a tr ansmission occurrence include intensive screening of all potential donors, includi ng medical and social history and various required data to determine if th e donor is acceptable [35, 36]. Additionally,

PAGE 31

19 tissue and blood samples are screened with microbiological testing and serological testing, respectively. Th e blood samples are virologically sc reened for HIV, hepatitis B, hepatitis C, and CMV [5]. Tissue samples follow the grafts through all of the processing steps in order to detect potential tissu e infection following recovery [35]. The contamination problem of vascular allografts is centered on recovery, sterilization, and storage. Incidents of contamination can occur at recovery if the ba rrier to the digestive tract is compromised. Prepar ation and evaluation of the tissue for implantation requires extensive handling under aseptic conditions wh ich adds opportunities for contamination. Sterilization via antimicrobial solutions, ga mma-irradiation and ethylene oxide can prove to be inefficient or too destructive [37-41] The latter two have been abandoned due to tissue damage, leaving treatments by antimic robial solutions a nd cryopreservation. Allograft Processing Sterilization and disinfection The active surveillance activities and the outbreak investigation carried out by Centers for Disease Control (CDC) during 2002 highlight the fact that the spectrum of bacterial pathogens associated with allograft-associated in fections include Gram positive ( S. aureus and Enterococcus) and Gram-negative bacteria, especially those that ferment lactose [42]. Hence, any antimicrobial soluti on that is validated fo r processing allograft tissue should theoretically be active against any of the above pathogens. The importance of validating any sterilization technique agai nst a mixture of pat hogens is underscored by the fact that 19% of the all ograft-associated infections as certained during the active case finding period were polymicrobial. The fre quencies and concentrations of various pathogens in polymicrobial infections rema in unknown and will vary from recipient to recipient and, therefore, are not predictable. Based on data in the CDC investigation, the

PAGE 32

20 antimicrobial solution used should be validated against the following pathogens, at the very minimum: Staphylococcus aureus, Enterococcus faecalis, Pseudomonas aeruginosa, Escherichia coli, Clostridium species and Bacillus species [42]. There are no data to predict the relative proportions of the respective pathogens in the mixture being tested. The current common disinfection method is an antimicrobial soak in which the tissue is aseptically processed with antimicr obial and sometimes antif ungal cocktail [35]. A study to evaluate the effectiveness of the antibiotic cocktails used by a tissue bank for disinfection of human tissue was conducted by a GLP/GMP-compliant1 testing laboratory (AppTec, Marietta, GA). Saphenous vein samples were inoculated with 106 of aerobic, anaerobic and a mixture of test organisms suggested by the CDC for validation. The result was a one log reducti on or less with absolutely no effect detected in the Clostridium species [44]. Typically, the patient is safe since the donor will be discarded if representative samples tests positive for contaminants, but sometimes the method for testing sterility may be flawed [45]. Specif ically, residual antibiotics from the treatment process have been shown to inhibit growth in post-trea tment surveillance cultures producing falsenegative results leading to cases of sep tic arthritis, hepa titis C transmission, Streptococcus pyogenes (GAS), and Cl ostridium spp [ 8, 46]. Of note, cases of soft tissue allograft-associated disease transmission by s pore-forming organisms that are resistant to antimicrobial solutions have been repo rted within the last few years [8, 47] These tissues 1 GLP/GMP refers to Good Laboratory Practices and Good Manufacturing Practi ces, respectively. These are regulations set forth by the Food and Drug Administration (FDA) that requires manufacturers of human drugs and biological products, animal drugs, medical devices, and food additives to demonstrate safety and utility of their product 43. Administratin, F.a.D., Good Laboratory Practices (GLP) for non-clinical laboratory studies 21 CFR Part 58 p. Supporting Statement..

PAGE 33

21 have predominately been cartilage used in jo int repairs, but investigations prompted by these events have revealed transmissions and infections resulting from other soft tissue grafts such as tendons and saphenous veins Vascular tissue is less pr one to infection than cartilage, but it is still vulnerable to organi sms (including fungi and viruses) inherent in the donor or transmitted during recovery or processing that are resistant to the antimicrobials currently used for allograft di sinfection. Additionally, infection can occur and, in some cases, the rapid ons et of infection and the inf ecting organism itself suggest that graft to host disease tran smission was responsible rather than infection from some other source implementing the graft as the ca rrier [23, 48, 49]. More recently, analysis of the infecting organism's DNA has provided even more convincing evidence that donor to host transmission of inf ection has occurred [47] Cryopreservation Adverse results with the first cryopreserved vein caused enthusia sm to wane since techniques led to early thrombotic graft occlusion. Recent improvements in cryopreservation and controlled freezing have made it possible for arterial and venous allografts to be used as alternate vasc ular conduits [50]. Th e primary goal of a preservation technique is to maintain cell viability in order to increase vein graft availability by extending its shelf-life. In the process, the graft should have minimal antigenicity while viably functional and maintain structural integrity similar to native vein [6, 15]. The use of cryopreservation of vascular conduits for long-term storage has also been used as a means of viral inactiva tion in addition to reduc ing antigenicity, but the effectiveness is unclear w ith a small risk of viral dis ease transmission remaining. Cryopreserved veins are commercially availa ble from CryoLife, Inc. (Kennesaw, GA), Northwest Tissue Center (Seattle, WA ), and RTI Cardiovascular (Birmingham,

PAGE 34

22 AL). CryoLife has been processing cryopr eserved tissue for implantation since 1984 [35]. The common method is to place the vessels in a buffered solution containing cryoprotectants, such as DMSO, and then stor e in the vapor phase of liquid nitrogen (110 to –196oC) [6, 15]. An antimicrobial solution may be used initially to disinfect the tissue prior to storage. As the tissue is tr ansferred from the disi nfectant solution to the freezing solution, representative samples are re moved for sterility testing according to United States Pharmocoepia (USP) standards. The tissue is frozen in a controlled manner from +4oC to –80oC or below at an average rate between 0oC / minute and 5oC / minute. The cryopreserved vessels are then stored in freezers with the vapor phase of liquid nitrogen. Thawing and rins ing are complicated, non-ster ile procedures performed aseptically. The rinsing procedure must be performed immediatel y after the thawing procedure. The allograft pack age is first thawed at room temperature and then in warm saline (32oC to 42oC). The second event is supposed to rapidly thaw the tissue without adversely affecting cellular viability [6, 15]. The graft is then aseptically removed from packaging, flushed and soaked with a seri es of solutions to remove cryo-solution components before implanting. Freezing with cryoprotectants prevents the in tracellular vapor phase gradient that is created when the extracellular matrix freezes. The cytoplasm remains in liquid form and causes cellular membrane disruption from intr acellular swelling as a result of osmosis, producing a nonviable graft with structural damage. The cr yoprotectant protects the matrix and the cells by replacing the wate r [15]. Without cryoprotectants, saphenous vein allografts have been associated with high graft failure rates in infrainguinal bypass procedures [6]. At the same time, an ideal method of cryopreservation has not been

PAGE 35

23 determined. Experimental and clinical da ta suggests that current cryopreservation protocols may be responsible for making all ografts more brittle and could induce early graft rupture or dilatation [51-55]. Cryopr eserved veins are prone to fissuring during thawing which degrades the stre ngth and performance of the graf t. Therefore, aneurismal degeneration continues to be a matter of concer n. This may be due to preservation of the endothelial antigenic properties as wells as inadequacy of the structural and functional media characteristics (elasticity and compliance) [6]. In clinical surgery, there have been mediocre results leaving cryopreserved vein s as an alternative only when no other suitable conduit is available. A majority of the investigations on cryopreserved vessels show evidence of antigenicity and rejection. Th e development of intimal hyperp lasia that leads to graft failure is compounded and accelerated by im mune-mediated inflammatory response secondary to rejection, the eff ect of cryoprotectants, surgic al injuries and rheological mismatch [6]. Explanted tissue has shown ce llular infiltrates that correlates with wall hemorrhage and necrosis [6]. Venous allograf ts appeared to induce anti-HLA antibodies post implantation and an associated humoral immune response contributing to failure. Most inflammatory cells are activated by cytotoxic T cells, expressing CD3, CD8 and HLA-DR antigens. On the other hand, when compared to an acellular graft made of foreign material, PTFE, cryopreserved saphe nous vein allograft showed no significant difference in T-cell subpopulations or lymphocytotoxic antibodies [15]. Immunohistochemical examination showed mi nimal mononuclear cell infiltration in adventitial layer indicating no significant immunological activa tion or clinical rejection. Contradictory results on the pr eservation of the endothelial layer and smooth muscle

PAGE 36

24 layer show improvement in patency rates in animal studies while no clear benefit in humans has been discovered [6, 15]. This is most likely due to the freshness and young age of the tissue in an animal study compar ed to the older population of human donor tissue that may take as long as 24 hours to r ecover followed by additional time to process for implantation. This may be another indi cation why autograft pe rforms better than allograft. The ultimate immune response may also be directed against smooth muscle cells when the endothelial layer is absent or subconfluent. A llosensitization occurs due to expression of class I and II major histocom patibility complex (MHC) antigens present on preserved endothelial cells as well as non-MHC antigens that stimulate T-cell mediated rejection response [6, 15]. Immunologic mech anisms have been implicated in early thrombotic occlusion and poor patency rates in lower extremity re vascularization with cryopreserved allograft [1, 5, 25]. There has be en evidence of the development of intimal hyperplasia and late graft failure due to the replacement of the smooth muscle cell layer by fibrotic layers after implantation [6, 15]. Additionally, the structural integrity can be compromised with remodeling by structural reor ganization of the graft and thickening of the intima from smooth muscle cell proliferation. This is followed by a gradual decrease in compliance from fibrosis in the layers as well as immedi ate disintegration of immunocompetent cells and rel ease of proteases that probabl y cause elastin degradation [34]. Elmore et al found similar contractile characteristics and graft patency rates but some cellular differences when comparing noncryopreserved to cryopr eserved grafts in canine femoral arteries [15]. There was a decrease in smooth muscle cell relaxation

PAGE 37

25 response to nitric oxide (NO) in cryopreserved grafts as we ll as decreased relaxation to phenylephrine and endothelin [6, 15]. The disadvantage to cryopreservation is rela ted to the second issue raised with the occlusion of allograft veins that is most lik ely due to an immune reaction similar to a graft rejection response. Si nce endothelial cells, smooth muscle cells and fibroblasts contain antigens on their cell membranes, preservation of these components could theoretically have deleterious consequences by mediating im munological reactivity [25]. Additionally, preserving these cells creates th e need to consider ABO-compatibility and lymphocyte cross-matching [5]. Since there is a limited availability of allograft arteries, a long waiting list may be a problem limiting widespread use. Antigenicity Blood typing is required for vascular graf ts to prevent reje ction since simply inactivating contaminants is not sufficient protection. There is a limited availability of vein allografts due to ABO relevancy and a depletion in inventory because of positive cultures [5, 6]. At an AATB accredited tissue bank that utilizes antimicrobial soak and cryopreservation had a 2003 yield of saphenous vein grafts per donor of only 58% where about 236 saphenous veins were discarded due to positive first and final cultures. ABO matching contributes to this deficiency, but only ABO compatibility has been evaluated in most studies [6]. Simply matching blood type is not sufficient to prevent an immune response. Human leukocyte antigens (HLA), proteins found on most cells, are used by the immune system to distinguish between endogenous and foreign cells. Additionally, there are many HLA antigens, so HLA matching is not feasible for cl inical applications and the role of matching HLA and ABO for humans has not been determined [5, 6].

PAGE 38

26 Again, removing the cells from the graft woul d alleviate the problems with finding the appropriate donor. It is believed that the difference in all ograft healing compared to autograft healing is mainly due to this immunologic response. This contributes to the low patency rates seen in allograft use. Graf t rejection is a response to cl ass I and class II MHC antigens laden on the membranes of endothe lial cells [56]. This layer is in immediate contact with inflammatory cells in blood such as neut rophils, monocytes, and macrophages [29, 57]. Rejection plays a significant role in failure and leads to allosensitization [6]. Studies have found pathophysiological pr ocesses develop in the walls and gradually lead to stenosis, mainly at anastomosis, and segmenta l dilatation that may be a result of chronic transplant rejection. It appears from some studies that an intact endotheliu m is a prerequisite for improved long-term patency of autologous a nd alloplastic bypass grafts [5]. The endothelium is an active barri er that allows permeability to fluid and macromolecules while limiting entry of inflammatory cells (polymorphonuclear cells and monocytes) and immune effector cells (B and T lymphocytes). It plays a critical role in inflammation as well as maintaining a nonthrombogenic surface by maintaining normal platelet activation [58]. Cryopreserved veins boast at the most approximately 70% viability, leaving an insufficient endothelium that is not fully functional [51, 53]. Wh en dysfunction in the endothelium occurs, pathologica l processes result, leading to endothelial cell activation and thrombus formation. There is an in jury response with inflammation and wound healing where thrombin and fibrin are genera ted and platelets are activated. The graft begins to occlude with loss of pa tency and early graft failure.

PAGE 39

27 Chronic rejection also affects the vascular wall with medial destruction where the arterial wall dilates and possibl y ruptures [59]. Davies comp ared the healing response of allograft to autograft in a rabbit model with a common carotid artery bypass [60]. They found there were two phases in the response. The first phase consisted of significant thickening of the intima and media compared to autograft. The second phase showed a 51% decrease in the intimal thickening and a maintained medial thickening of 97% compared to autograft. The allograft ha d an exaggerated medial response while the autograft had an exaggerated intimal response. Overall, there was a net increase of 17% in allograft wall thickness compared to autogr aft. Even so, the lumen between the two sets of grafts was not signif icantly different after 28 days. A similar medial response was seen in an abdominal aortic graf t model in rats but the lumen appeared to occlude. There wa s medial cell loss and matrix degeneration, adventitial inflammation, and intimal hyperplasi a. The lumen narrowed by intimal cell proliferation that was found to express -actin, a marker for sm ooth muscle cells and fibroblasts. The smooth muscle cells were destroyed and the elastic lamina became dislocated and fibrosis was noted [59, 61, 62]. The inflammation in the adventitia disappeared along with the removal of the SM Cs [59, 61]. This gives credence to the antigen induced response by both the endothelial cells and SMCs. Use of immunosuppressant a nd anticoagulant therapies It is still questionable if cell viability is benefici al since it maintains its antigenicity which has led to the use of short term immunosuppressant and anticoagulation therapies [5, 6]. The use of immunosuppre ssion and anticoagulation therapies, along with cryopreservants with controlled freezing, has been shown to

PAGE 40

28 improve patency and graft acceptance to comb at cryopreserved conduits more thrombotic behavior early on [15]. Posner et al. showed improved one year patency with infrainguinal arterial reconstruction with low-dose immunosuppression treatment [63]. They also showed a substantial number of all ograft-related complications with early degeneration, pseudoaneurysm formation, and hemorrhaging. An improvement in graft patency in animal models and human arterial all ograft patency is se en with high-dose immunosuppressive treatment, whereas, the low-dose regimens were not enough to curtail the allogeneic response [6]. The more potent tr eatment may be effective, but the toxicity level may not be tolerated by older patients with multiple co-morbidities as well as increasing the risk of in fection, with reduced graft re jection, but disturbed wound healing, and possibly gangrene [5]. Therefore, patients with infected tissue necrosis, an indication for surgery, would be at greater ri sk if the immune system is compromised with treatment. One study where patients received high-dose immunosuppression for kidney transplants did not prevent rejection [6]. A histological and immunohistochemical analysis of the renal arterial allografts of the failed kidney transplants showed evidence of chronic rejection and intimal hype rplasia in the vessel walls. Most patients with critical limb ischemia will have reduced renal functi on. Exposing them to therapy may lead to further reduction of renal function [5]. In one study, treatment wa s discontinued in 12% of the patients due to gastro intestinal bleeding [6]. Th ese conditions preclude a large population of the patients needing revasculariz ation from receiving cryopreserved grafts if immunosuppression is required.

PAGE 41

29 Denudation of cell layers Experimental and clinical studies on cryopreserved veins show that many functional features of the endot helial cells and smooth muscle cells are preserved [6, 15]. These properties are prostacyclin and e ndothelin production, fibr inolytic activity, thrombotic properties, platelet deposition, a nd metabolic mechanisms. However, after implantation some studies show the presence of an intact endothelial layer while numerous animal studies and some clini cal studies shows moderate endothelial denudation to complete loss of endothelial cells after implantation [15]. It seems in some preservation protocols the endothelial laye r may no longer be bound to the graft and is removed in the circulatory flow negating the e fforts made to maintain and preserve that layer. The loss of the lining and smooth musc le cells are due to the inability of the preservation process to maintain complete ce ll viability resulting in a loss of adherence to the basement membrane. This is presumably due in part to a more rapid accumulation of low-density lipoprotein cholesterol compar ed with noncryopreserved human saphenous vein grafts [5, 6, 15]. Other possible factor s are toxicity of the preserving agents (particularly with antibiotics), cryoprese rvation events, spontaneous cell death, or destruction of the survivi ng cells by immunocompetent host immune cells [5]. Additionally, the ischemic time for allograft tissue varies up to a 24 hour period resulting in variable cell viability pr ior to cryopreservation. Decellularization The disadvantage to allograf t use is its biologically active components containing antigens that could lead to allosensitiza tion, resulting in an immune reaction and rejection. Cryopreserved grafts retain viable or nonviable cells wh ich still contain the antigenicity present in the protein structure of the cell surface. This has led to the

PAGE 42

30 investigation of decellularized conduits for vascular reconstruction. There is a general consensus among investigators that removing, in stead of merely inactivating, the cellular content may reduce or prevent an immunological response. Biologically active endothe lial cells express class I and class II major histocompatibility complex antigens that re sult in an immunological response [64]. Human studies using antibiot ics and cryopreservation to process allografts have demonstrated a panel of these reactive antigen s, some of which caused a broad recipient allosensitization [64, 65]. This is of particul ar interest in patients with end-stage renal disease awaiting a kidney transplant. The pr esence of alloantibody could preclude renal transplant due to cross-match since these pa tients are already immune compromised with decreased functioning of white blood cells [64, 66]. There is controversy in the literature as to the allosensitiz ation of cryopreserved grafts. The levels of panel reactive an tigens (PRA) in patients whom received cryopreserved vein grafts showed allosensitiz ation within approximately 3 months with an 84.1 % increase [64]. Although, when compared to dece llularized grafts (SynerGraft, CryoLife, Inc.) no significan t difference was found in primary, assisted primary or secondary patency. A study of cryopreserve d saphenous (SV) allograft in lower extremity revascularization yielded poor pa tency rates that were believed due to immunological reactions. These, and other recent studies demonstrated venous tissue expresses class I and class II major histocom patibility (MHC) antigens present on the preserved endothelial cells as well as non-MHC antigens that stimulate T-cell mediated rejection res ponse [67].

PAGE 43

31 If injured or subconfluent endothelial cel ls that line the lumen are present, thrombosis and smooth muscle cell (SMC) gr owth are promoted leading to intimal hyperplasia [7]. This is likely due to the ma jor antigenic determinants, class I and class II MHC, present on the SMC. Therefore a de nuded vessel leaves no protection from an immune reaction. It has been shown that the endothelial cells of cryopreserved vessels are not fully viable or confluent and slough o ff into the blood stream after implantation. Rejection plays a significan t role in failure and le ads to allosensitization. Allosensitization was demonstrat ed in patients who received cryopreserved allografts in a SynerGraft (CryoLife, Inc) study. Blood tests showed an increase in PRA levels partly due to exposure to the allogeneic antigens on the preserved endothelial cells. A study of cryopreserved saphenous vein allografts in lower extremity revascularization yielded poor patency rates that were likely due to immunological re actions. Another study using aortic valve allografts (AVA) showed ear ly failure caused by donor-specific immune response [68]. Preserved vessels used for hemodialysis access cau sed allosensitization with increased PRA levels. This can be a potentially serious problem for kidney transplant recipients since the presence of alloantibody c ould preclude a renal transplant due to cross-match incompatibility [64]. Chronic rejection can cause vascular wa ll destruction making the graft unsuitable for long-term applications. The medial laye r is destroyed with altered extracellular matrix leading to wall dilatation and rupture [ 59, 62]. This event could be catastrophic in a coronary site that may lead to morbidity or mortality. The cellular events that occur was modeled in abdominal aortic grafts in a rat model by seve ral researchers and summarized by Allaire [59]. Essentially, there is medial cell loss a nd matrix degradation,

PAGE 44

32 adventitial inflammation, and intimal hyperplasia. This is a typical response for immune injury in arterial allografts [59]. Thr ough the response, the media is thinned and the elastic lamina is dislocated [59, 62]. Smooth muscle cells are destroyed while inflammatory cells invade the adventitia, mo st likely due to the antigens present on the SMC since the inflammatory cells disappeared when the SMCs disappeared [59, 61, 62]. The response results in degradation of the matrix, fibrosis and functional failure. Additionally, the graft is o ccluded by narrowing of the lumen caused by the intimal cell proliferation thickening the vascular wall. Cu rrent strategies to reduce or prevent this cascade of events are immunosuppressives drugs or reducing the antigen icity of the graft with cross-linking agents, such as gluteral dehyde, or by sequential chemical treatments that promotes tissue decellularization. Decellularization involves the removal of major immunogenic components such as cells and their lipid membranes, membrane associated antigens and soluble proteins, and other lipids and more soluble glycosaminogl ycans [62, 69]. The resulting graft is an acellular scaffold with only insoluble structur al proteins, such as collagen for strength and elastin for distensibility [69, 70]. Primar ily present on the endothelial cells and also present on SMCs are class I and class II majo r histocompatibility complex antigens. A 10 and 20 week sheep study with an allograft patch model showed only 1 out of 8 sheep with a decellularized graft de monstrated a positive elevation in PRA levels to MHC I and none to MHC II. In the classically cryopr eserved group, 2/3 of the sheep showed an elevation [61]. Decellularized grafts have been evalua ted histologically, immunohistochemically, mechanically and with both animal studies a nd clinical studies. Hilbert evaluated the

PAGE 45

33 morphologic characteristics in a small-diamet er freeze-dried decellularized carotid artery of a goat decellularized with a nonionic detergent solution. Histological analysis showed complete endothelial cell removal while preserving the basement membrane. Additionally, no cells were seen in the intima, media or ad ventitia, although, there were remnants of smooth muscle cells in the medi a without nuclei. The extracellular matrix seemed well preserved, but oval to circular sp aces were occasionally noted in the media. The internal elastic lamina remained intact, as well as the layered elastic laminae in the media [70]. In an animal study conducted by Hilber t, all allografts were patent upon explantation with no significant dimensiona l changes or aneurism formation. The autografts showed a luminal surface lined wi th a layer of endothelial cells while the allograft showed a discontinuous layer. Even in these regions of incomplete endothelialization thrombi were not obser ved. Both types of grafts showed myofibroblasts, collagen and proteoglycans th at may be indicative of incorporation [70]. The allograft appeared relatively acellular in the media with focal cellular regions among the dense collagen and elastic lamellae co ntaining infiltrated myofibroblasts of the hosts. The authors believe that the number of host myofibroblasts in the media may have been significantly limited by the presence of the dense collagen bundles and smooth muscle cell remnants. Host cell migration wa s most apparent close to the anastomoses and was rich in proteoglycans. It appeared to occur in the adventitia along the length of the grafts. Histologically, there was no eviden ce of calcification and inflammatory cells were not present in the graft wall. Electr on microscopy did show calcification of minute

PAGE 46

34 remnants of cell membranes. Ingrowth of host blood vessels was not observed after 6 to 7 months [70]. Conklin investigated decellularization of a small-caliber xenograft with a porcine common carotid artery model [2]. This group followed cell lysis with multiple enzymatic digestions and detergent washes while ag itated. Histology and electron microscope showed complete removal of cellular compone nts while the extrace llular matrix appears to remain intact. Hematoxylin and eosin (H&E) stain showed no signs of remaining nuclear material in the walls of the vesse ls and TEM showed complete removal of cellular material from the media layers. Hi stology showed an intact internal elastic lamina along with elastin lamellae in the me dia. TEM showed the basic extracellular microstructure remained intact after processing. In addition to reducing immune reactions and decreasing the thrombogenicity of the luminal surface by decellularization, this gr oup wanted to ensure that the process would maintain a graft with similar mechan ical properties to th e native vessel, high strength and good handling characteristics. Pr ocessing with chemical s or treating with a cross-linker may affect the inte grity of the matrix in focal regions leading to aneurysm post implantation, weakening of the graft, or making it stiffer and more brittle. They investigated the compliance and burst strengt h of the vessels on a custom-built system that measures the diameter changes while increasing the intraluminal pressure. The compliance is expressed as the percentage di ameter change per pressure (mm Hg) change normalized to the compliance of fresh vessels. The average compliance was calculated as the slope of the linear re gression line in the physiologi cal pressure range (70-130 mm Hg).

PAGE 47

35 The mechanical analysis was performed on decellularized vessels, decellularized and heparin-treated vessels, fresh vessels alcohol preserved vessels, and ePTFE prosthetic grafts. The ePTFE grafts were the least compliant ( D/mm Hg 0.024%) and the decellularized grafts were slightly more compliant ( D/mm Hg 0.181%) than the fresh vessels ( D/mm Hg 0.172%). Alcohol caused th e tissue to stiffen some (0.160%) while Heparin treatment produced a much sti ffer graft (0.0975%). During burst testing, none of the fresh vessels burst within the limit of the pressu re transducer (2300 mm Hg). One out of four of the dece llularized vessels burst with in the limit at 1654 mm Hg. Although there may be some strength loss measured in that one vessel, there still is a high safety margin over ten times the physiologic pr essure. Unfortunately, only four samples were tested in this group. This group examined their decellularized process with heparin cross-linking on a carotid artery bypass in dogs. At 24 days, fi broblast-like cells a ppeared to densely populate the media and there were few endotheli al-like cells lining th e lumen. After two months, dense -actin staining suggested smooth mu scle cells densely populated the vessel wall and Factor VIII staining confirme d that endothelial cells lined the lumen. Teebken et al. developed an acellular va scular xenograft matrix from porcine thoracic aortas for the purpose of seed ing with human cells [71]. This group decellularized with enzymatic cell extrac tions using biological enzymes trypsin, ribonuclease (RNase) and deoxyrib onuclease (DNase). The idea is that these enzymes are capable of removing cell components as we ll as cellular antigens, lipids and to some extent glycosaminoglycans with limited toxi city. Additionally, th is group believes that extractions with detergents may disturb the endothelialization of the grafts both in vivo

PAGE 48

36 and in vitro Light microscopy showed no cell nucl ei or intracellu lar components in cross-sections of the aortic wall after processing. Immunohistochemical stains for fibroblasts and endothelial cells were all ne gative. SEM showed extracellular matrix fibers with openings of 1 to 10 m. Schaner studied the composition and stre ngth of decellularized human greater saphenous vein specimens to determine their potential as a vascul ar tissue-engineering scaffold [72]. This group used a detergent, sodium dodecyl sulfate (SDS), as a decellularizing agent. Transmural cell rem oval was found to be nearly complete (>94%) at a concentration of 0.075% SDS. The lu minal surface appeared completely devoid of endothelial cells at all concentrations tested. The collagen morphology appeared unchanged, the basement membrane remained in tact and the elastin staining decreased only slightly. This group also performed mechanical te sting to evaluate the effects of the process on the burst strength and suture retentio n strength. It was not ed that the vessels had normal consistency and good handling charac teristics. The burst strength of the decellularized graft was similar to the fresh vein (2480 460 mm Hg vs. 2380 620 mm Hg (p>0.05)) as well as the suture-holdi ng strength (185 30 gm vs. 178 66 gm (p>0.05)). The functionality was examined by placing decellulari zed canine jugular veins into a carotid interposition model. Each canine received an au tograft and either an allograft or a decellulari zed allograft on the contralateral si de. After two weeks, all grafts were assessed by Duplex imaging to be pate nt with no significant dilation, rupture or anastomotic false aneurysm.

PAGE 49

37 CryoLife, Inc. (Kennesaw, GA) is fili ng for FDA approval fo r a decellularization process on veins and heart valves called Syne rGraft [64, 67]. This process involves cell lysis in hypotonic sterile water, followed by e quilibration in a buffer. The tissue is then treated by an enzymatic digestion of the nucle ic acids with a combined solution of RNase and DNase. The grafts are then cryopreserve d for storage. Results on the CryoValve SG showed approximately 99% reduction in staining of the endothelial ce lls and interstitial cellular elements [67]. In a clinical study by Hawkins, 14 children (8.5 7.9 years) received decellularized, cryopreserved all ografts (CryoLife, Inc.), 6 were patch insertions and 8 with valved pulmonary allografts [67]. Th ese groups were compared to 20 historical control subjects (1.7 2.4 years) 8 with valves and 12 with allograft patch insertions. There was no attempt to match ABO blood t ypes in either group. The effect on the immunogenicity was measured at 1, 3 a nd 12 months by the frequency of HLA alloantibodies PRA: class I (HLA-A, HLAB and HLA-C) and class II (HLA-DR/DQ). Antibody levels were slightly higher from pre operative levels for both classes at all time points. The antibody levels were significantl y lower in the decellularized allograft group. There was a marked reduction in staining fo r class I and class II histocompatibility antigens. However, there was no work done in this study to determine whether reduced immunogenicity will truly allow tissue ingrow th and improved long-term durability in patients. Madden performed a clinical study of 20 patients with an upper extremity SynerGraft cadaver vein allograft for hem odialysis access [64]. The first 17 patients were matched (ABO) for blood types while th e 3 remaining patients were intentionally

PAGE 50

38 given ABO-incompatible SynerGraft allograf ts. Allosensitizati on was quantified using the PRA assay (American Society for Histoc ompatibility and Imm unogenetics laboratory protocol) on peripheral blood samples. None of the SynerGraft patients became allosensitized (PRA > 10%) at 10 months w ith a mean PRA of 3.2% (0 – 7%). All patients in the historic cryopr eserved group became allosensitized by 3.1 months with a mean PRA of 84.1%. All three patients with ABO-incompatible SynerGraft allografts showed no allosensitization or acute rejec tion-type reactions. Typically, removal of viable donor endothelial cells is believed to lead to increased thrombogenicity and decreased infection resistance. There was no significant difference between the groups in the primary, assisted primary or secondary patency rates and no gr afts were lost to infection. Expected Outcomes The deliverables from this study are cond itions and exposure times for a chemical sequence that will kill and remove infecti ous agents as well as endogenous cellular material without degradation of the matri x. Removal of the endothelium and other reaction causing cells wi ll eliminate the need for blood typing, reduce immune reaction, increase durability and longevity and redu ce the number of suitable donated grafts rejected due to positive culture.

PAGE 51

39 CHAPTER 2 DECELLULARIZATION OF VEIN GRAFTS Introduction Vascular disease is one of the most prevalent diseases worldwide with bypass or replacement as the mainstay of therapy for coronary and peripheral occlusions. Over 100,000 peripheral and over 400, 000 coronary bypasses are performed in the US annually [4, 27]. This results in a need for readily available small-diameter vascular grafts that are immunotoleran t and durable. For corona ry bypasses, the autograft saphenous vein or mammary artery are the standard and autologous saphenous vein remains the conduit of choice for infrapopliteal bypasses due to its flexibility [2, 70]. Using autologous tissue requires a second su rgery site and extended operating time and requires multiple revisions with a 4 year patency rate of only 40 to 70% [2]. Additionally, as many as 30% of patients will have unsuitable autogr aft material, forcing them to consider other options such as s ynthetics, allograft or xenograft [2, 73]. Allograft tissue has an advantage over synthe tics due to its ability to be innervated by host cells and resistance to in fection. The starting material is similar to the native tissue in its structure and f unction with the geometry and components necessary for cell differentiation. It contains the possibility of autologous cell infiltration creating a biologically active matrix with phenotypica lly appropriate cells. This allows for

PAGE 52

40 reparative and functional charact eristics not inherent in nonviable prosthetics. In the same manner, the disadvantage to allogr aft is its biologically active components containing antigens that could lead to allosensitization and chronic rejection. Allografts are plagued with complications leading to early stenosis and graft failure. Occlusion or stenosis in coronary or peripheral circulati on is a common clinical occurrence present in small-caliber applicatio ns that results in a dditional interventions [1]. Early thrombosis occurs when platel ets in circulating blo od adhere to certain surfaces, then release chemicals to attract more platelets to form a large aggregate that generates thrombin [67]. An animal study using allogenic va lve grafts in the descending aorta of rats showed thrombus formation after 21 days [68]. This is the most common failure mechanism for synthetics, but also occurs in allograft tissue. Cellular events that occur after implantation contribut e to this loss of patency. Adaptation of vein grafts to the arterial environment has been studied extensively in animal models but less in humans. Afte r implantation, the conduits undergo structural changes characterized by intimal hyperplasia an d overall wall thickening that is believed to be a result of the additional immunologi cal stimuli when compared to autograft [2, 7, 59, 60, 70]. Fibrous intimal hyperplasia develops from an excess of smooth muscle cell proliferation. This flow-re stricting lesion may occur diffu sely throughout the graft, or more commonly, at focal sites near anastomo ses. A study by Da vies et al. using a common carotid vein bypass graf t in rabbits was used to co mpare the differences between the healing response of autograft and allograft [60]. All grafts remained patent up to 28 days. Compared to the autogr afts, the allografts showed a 51% decrease in overall mean intimal thickness and a 97% increase in the ove rall mean medial thic kness, resulting in a

PAGE 53

41 net increase in the vascular wall thickness. Interestingly, the lumen of the autografts and allografts were not significantly different. It appeared that intim al hyperplasia occurred more in the autograft while the media of the a llograft had a greater response. Of note, the intimal hyperplasia occurred with int act endothelial cells on the surface. The traditional method of allograft storage utilizes cryopreservation for storage. The intent is to produce a viable construct with minimal antigenicity, optimal biologic functionality and structural integrity. Cryopreservation has been shown to yield improved patency rates over fresh and synthe tic grafts in hemodialysis access. The authors of this study believe th at it is infection resistant with a large portion of the endothelial cells viable at engraftment [64]. At the same time, clinical and experimental data suggest that cryopreservation may be re sponsible for rendering allografts more brittle and could induce early gr aft rupture or dilatation [5155]. Additionally, the viable cells remaining on the endothelium and within the vascular wall contain the antigens that in this instance may lead to the re jection response seen in allografts. Acellular Allografts The limitations with conventional cryopres erved small-diameter vascular grafts due to immunogenicity and rupture have led to the development of new decellularized vascular grafts. Fixing the tissue with cross-linkers is used to reduce the immunogenicity, but can be toxic or lead to aneurism formation. Decellularization involves the removal of major immunogenic co mponents leaving an acellular scaffold with only insoluble structur al proteins, such as colla gen and elasin [69, 70]. CryoLife, Inc. (Kennesaw, GA) has deve loped a decellularization process on veins and heart valves called SynerGraft [64, 67]. This process involves cell lysis in hypotonic sterile water, followed by equilibration in a buffer. The tissue is then treated by an

PAGE 54

42 enzymatic digestion of nucleic acids with a combined solution of ribonuclease (RNase) and deoxyribonuclease (DNase). The grafts are then cryopreserved for storage. Results on the CryoValve SG showed approximately 99% reduction in staining of the endothelial cells and interstitial cellular elements [67]. Maintaining the similarities to the native tissue of the biological composition and geometric design is essential to promote re -endothelialization a nd cell migration [61, 70, 71, 74]. The basement membrane on the lumi nal surface consists of type IV collagen with ligands for firm endotheli al cell and myofibroblast att achment and retention [62, 70, 72]. Preservation of this structure facilitates re-en dothelialization to reduce thrombogenicity and promotes migration of my ofibroblasts into th e vascular wall [59, 70, 75]. An insoluble extracellular matrix has b een shown to promote fi broblast proliferation and elastin synthesis and organization into fibers [59]. Another study showed the importance of complete cell removal where host myofibroblasts in the media appeared to have been significantly reduced by the pres ence of collagen bundles and smooth muscle cell remnants that restri cted migration [70]. There is concern that removal of the vi able endothelial cells could result in thrombogenicity and decreased infection resistan ce. It is believed by some investigators that a lack of surviving cells and endoth elial integrity may play a role in graft degeneration and early and late patency [5]. Madden showed no significant differences in patency or infection rates when comp aring CryoLife’s SynerGraft to their own cryopreserved grafts [64]. Virtually all cr yopreserved homografts ar e acellular within a year of implantation. It has been shown th at after implantation ce llular grafts become acellular over time while decellularized graf ts have a time-depende nt recellularization

PAGE 55

43 during the same period [61]. A cellular aortic allograft conduit tran splanted into a rat showed a complete loss of smooth muscle cells in the media after 21 days [68]. Inflammation along with the lack of cellular function limits tissue ingrowth, performance and reparation. The tissue typically scars a nd then mineralizes [ 61, 68]. A cryopreserved allograft pulmonary trunk implanted as a patch was found to become acellular over a 20 week period [61]. A fibrous sheath developed co nsisting of layers of fibroblasts that line the luminal side and encapsulate the advent itia. This event may represent the same foreign body response seen in surgical im plants. At the same time points, the decellularized tissue displayed time-dependent recellularization. Thus, the decellularized scaffold appeared to have an advantag e over the cellularized graft with faster repopulation and remodeling due to bypassing the in vivo decellularization step. With traditionally cryopreserved methods, the endothelial cells slough off and the interstitial cells are removed before the lumen is relined. These series of events could prevent or slow down the remodeling proce ss by exposing the cellularized venous wall, sparking an immune response that proceed s faster than the repopulation of the endothelium. At the same time, transanastom otic endothelialization has been limited in humans despite the success in animal m odels and is an ongoing challenge in the development of vascular grafts. A four year study in a canin e model with iliac and carotid arteries placed in the femoral and carotid positions showed complete endothelialization of the flow surface [62] The same group with the same model compared arteries to autograft saphe nous vein and showed no evidence of endothelialization in the allografts after 6 months while the autografts had substantial endothelial cell coverage. In another study by the same group all grafts remained patent

PAGE 56

44 at four weeks [76]. Therefore, results can be variable and unp redictable and appear to be time-dependent. Although, 4 of the 9 decellulari zed allografts were patent compared to 4 out of 7 cellular autografts. Cell repopulation in humans rarely occurs, but again may be attributed to the presence of necrotic cell debris or even a poptotic cells that block specific cell signals that cause autologous cell repopulati on [61]. Other studies have looked at the compressed nature of the media after decellu larization with dense collagen and elastin fiber networks reducing the porous network and preventing complete migration passed the intima [74]. Therefore, complete removal of antig ens is anticipated to decrease the immunological response, thus, reducing the ne ed for immunosuppressants and impacting the durability [7, 61, 62, 69, 75]. Remnants in the arteri al or venous wall may promote increased innate and cellular immune res ponses with inflammation and consequential scarring, contributing to failu re of organized migration of phenotypically appropriate cells. Additionally, cell remnants have been attributed to calcif ication of veins and valves [61, 70, 73, 77]. A sheep model showed calcification occurred in the classically cryopreserved tissue while decel lularized tissue stained nega tive except around the suture [61]. Decellularization Process Cell Removal Assessment The decellularization process in the liter ature follows three main steps, although the specifics of each step vary. The first step in the process is to osmotically lyse the cells contained within the tissue. This has been performed with sterile water, a hypotonic solution or a detergent solution [2, 64, 67, 72, 73, 78]. The cell inac tivation is followed by enzymatic digestion of nucleic acids. So me groups use detergents and some prefer

PAGE 57

45 biological enzymes with the inte ntion of protecting the matrix for re-endothelialization as well as limited toxicity. A dditionally, all methods use mechanical stripping such as shaking. The final cleansing step involves a wash out to remove any residual cellular elements and chemicals. All of the studies required a multiday process to achieve greater than 90% decellularization and result ed in variable matrix integrity. The most common detergents used we re sodium dodecyl sulfate (SDS), Octylphenol ethoxylate (Triton-X 100) and sodi um deoxycholate. SDS is a highly ionic detergent with an anionic hydrophobic ligand [79]. It lyses cell membranes and is believed to be uniform within all layers of vascul ar tissue [72]. It only requires a low concentration to remove endothelial cells, but removal of smooth muscle cells is dosedependent. It is an amphipatic molecule th at associates with pr oteins via its hydrophobic domain, leaving the hydrophilic region exposed. This possibly creates an altered internal charge state that leads to swelling of tissue by increased water binding. Collagen has more hydrophilic sites than elastin, thus has an affinity for water. Decreased thermal stability of collagen is due to th e disruption of hydrogen bonding [77, 79]. Krasovakaya showed the hydrolysis of el astin with pancreatic elastase was markedly accelerated if pretreat ed with SDS [77]. Due to the elastin being predominantly hydrophobic, SDS reduces its hydrophobicity, thus exposing it to an aqueous environment and making it more susceptible to elastases [77, 79]. Additionally, SDS is difficult to rinse from the fibe rs, so it could alter the mech anical properties by binding to the polypeptide chains within the fibers or bind ing into the interfiber spaces of the outer surfaces of the fibers [79].

PAGE 58

46 Residual amounts of SDS in the tissue can also be toxic to cell seeding in vivo and in vitro. Investigators have shown that porcine aortic roots treated with SDS were surrounded by nonviable endothelial cell fragment s that were seeded onto the matrix [80]. Within 24 hours of in cubation there was extensive cell lysis with patchy cell distribution. SDS has been shown to be the most effective in cell removal, but causes detrimental structural changes [69, 77]. C ourtman abandoned further use of SDS after witnessing a significant drop in the shrinkage temperature of bovine pericardium, as well as a three-fold increase in tissue thickness likely due to swelling [77]. A study with pulmonary porcine valve conduits decellu larized in 0.1% SDS showed complete decellularization but at the cost of a significantly disintegrate d matrix at as early as 24 hours incubation time [69]. Thermogravimetri c analysis (TGA) and differential scanning calorimetry (DSC) was used by Samouillan to analyze the effects of detergents on the extracellular matrix of pulmona ry aortic roots. They found that SDS had a destructive effect with destabilizing the coll agen triple helix and swelling of the elastin network [79]. Schaner, on the other hand, showed no significa nt alterations with their decellularization method [72]. These authors studied the composition and strength of a decellularized human greater saphenous vein specimen to de termine its potential as a vascular tissueengineering scaffold. Transmural cell rem oval was found to be nearly complete (>94%) at a concentration of 0.075% SDS. The luminal surface was completely devoid of endothelial cells at all concentrations test ed, but the cell removal in the vascular wall appeared to be concentration-dependent. Histology showed the collagen morphology was similar, the basement membrane was inta ct and the elastin staining only decreased slightly. To assess the effects on the st ructure of the vein, this group performed

PAGE 59

47 mechanical testing. The burst strength of th e decellularized graft wa s similar to the fresh vein (2480 460 mm Hg vs. 2380 620 mm Hg) as well as th e suture-holding strength (185 30 gm vs. 178 66 gm). The decellula rized vessels were placed as carotid interposition grafts and asse ssed by Duplex imaging after tw o weeks. All grafts were patent with no significant di latation or rupture. No wo rk was done to look at the inflammatory reaction or the re cellularization of the grafts. Nonionic detergents have been found, in most cases, to be effective in cell removal while maintaining the extrace llular matrix [69]. Triton-X 100 is an excellent detergent and emulsifying agent that has an effective performance across broad temperature ranges, is soluble in water and biodegr adable [81]. Additionally, it is compatible with anionic, cationic and other nonionic surf actants which makes it favorable to combine with other agents for a synergistic effect Courtman chose 1% TritonX 100 to decellularize canine iliac and carotid arties to be used in femora l interposition bypass graf ting [76]. After four weeks, angiogram showed that all grafts were patent. Further analysis showed no thrombus formation and no aneurism formation. There was no evidence of inflammatory cells except in small areas at the anastomoses. There were some cells on the lumen consistent with endothelial cells and some mesenchymal cells in the adventitia, but the media was completely acellular. There was a pannus ingrowth of smooth muscle cell at each anastomosis. Triton-X 100 showed no cha nge in cellularity in a rat aortic valve allografts (AVA) at 1% and little change at 5% for 24 hours [68]. Samouillan also looked at the effects of Triton on the extracel lular matrix with TGA and DSC and they found it had no effect on the structural integrit y and the collagen helix remained stable [79]. Cho achieved complete cell removal with 0.5% Triton and successful seeding of

PAGE 60

48 bone-marrow derived cells resulting in the typical cobblestone morphology visible in scanning electron microscopy [82]. This was confirmed with van Wildebrand factor and CD31 staining, which are phenotypic marker s for mature endothelial cells. Triton was combined in several studies with sodium deoxychol ate, resulting in complete decellularization with preservation of the matrix [69]. Kasimir found that pulmonary valve conduits could become comp letely acellular incubated for 24 hours at concentrations as low as 0.25%. Kasimir a dded EDTA to the solution, which may have improved cell removal. The matrix appeared well preserved with minimal structural alterations. Reider, a co-a uthor with Kasimir’s study, found incomplete removal of porcine aortic roots and pulm onary roots at the same cond itions but required intensive washing to remove the residual chemicals [80] When this group seeded the matrix, it found that the treatment enabled host rece llularization with a confluent layer of endothelial cells on the lumen surface. This was an improvement to their cell seeding on the grafts decellularized with SDS [80]. Detergent residuals can be cytotoxic a nd inhibit cell adhesion, migration and proliferation [61, 75]. SDS and Triton have been shown to be difficult to remove even after extensive washing. Regarding the toxic ity issue, biological enzymes have been investigated. Trypsin is a natural enzyme with limited toxicity, but has been found to only achieve partial decellulari zation and produce severe struct ural alterations [7, 69]. Although, other authors have shown preservati on of the matrix in human and porcine aortic tissue with Trypsin decel lularization [7, 69, 71]. It has typically been enhanced with EDTA to chelate Ca2+ thus increasing the efficiency of Trypsin. Teebken developed an acellular vascular xenograft matrix from porcine thoracic aortas for the purpose of

PAGE 61

49 seeding them with human endot helial cells [71]. Little evidence has been given to correlate results from the seeding in vitro and in vivo repopulation. This group used 0.1% Trypsin plus EDTA followed by additional biological enzymes, ribonuclease (RNase) and deoxyribonuclease (DNase). The idea is th at these enzymes are capable of removing cell components as well as cellular an tigens, lipids, and to some extent glycosaminoglycans with limited toxicit y. After two 24 hour extractions at 37oC in Trypsin and EDTA interrupted by incubation with RNase and DNase light microscopy showed no cell nuclei or intracel lular components in cross-sections of the aortic wall. Immunohistochemical stai ns for fibroblasts ( -actin) and endothelial cells (factor VIIIrelated antigen, CD31) were negative and the matrix appeared well preserved. Unfortunately, no mechanical testing or qua ntitative matrix analysis was performed. A study with allogeneic valve grafts d ecellularized with 0.05% Trypsin for 0.5 to 1.5 hrs showed complete loss of cells except in the media [68]. When followed by Triton for a 24 hour incubation period, the entire matr ix was acellular. In both groups, severe damage to the leaflets was noted with histol ogy. Another study usi ng porcine aortas with 0.1% Trypsin for 24, 48, 72 or 96 hours contrast s these findings [7]. This group found that cell removal was time-dependent with a majority of the cells removed at 48 hours with only minimal damage with an increas ing microporous struct ure, as assessed by SEM. Reider found incomplete cell remova l of porcine aortic roots at the same concentration of Trypsin comb ined with EDTA with a 48 hour incubatio n time [80]. In vitro cell seeding revealed cobblestone mor phology on the luminal surface, indicating a confluent endothelial cell layer while cells in the media attached in a patchy distribution and starting to synthesize collagen.

PAGE 62

50 Biological Assessment Rat aortas were grafted as autografts, decellularized autograf ts, allografts or decellularized allografts with an SDS protocol [59]. After two months, the grafts were explanted and examined for structural and ce llular changes. All autografts remained patent while 3 of the 13 allografts thrombos ed (78% patency). The untreated autograft showed an intact media with no intimal thickening but was surrounded by thick adventitial fibrosis. The d ecellularized autograft produced a compacted acellular media. The external elastic lamina was thinner and adventitial fibrosis wa s visible. Intimal thickening did occur but did not contain infl ammatory cells. The lumen was covered with endothelial cells. Unlike th e autograft, the intima contains cells that stained positive for -actin. Cellular and fibrillar elements were equally distributed over the length of the graft. The decellularization of the allograft mate rial did prevent dilation from occurring and the vessel appeared macroscopically si milar to the untreated autografts. The untreated allografts were slightly dila ted throughout the length compared to the autografts. There was reduced adventitial inflammatory in filtration and the elastin was preserved. Just as with the treated autogr aft, the treated allograft had a compacted acellular media. Although, it was noninflammat ory and significantly thinner and richer in elastin than the untreated allografts. Th e internal elastic lamina was approximately 90% preserved along the length but the external elastic lamina was thinner and embedded in adventitial fibrosis. The intima of th e treated allograft showed thickening with no inflammatory cells and a higher smooth muscle cell and collagen density than untreated

PAGE 63

51 allografts. The lumen was covered with e ndothelial cells with equal distribution of the extracellular matrix. The media of the untreated allografts was dislocated and significantly thinner with fewer smooth muscle cells and less elastin. It contained inflammatory infiltrates. The internal elastic lamina was fragmented ove r more than half of its length and the interlamellar fibers were disrupted. There was intimal thickening containing inflammatory cells and lymphocytes with leukoc ytes adhering to the endothelium. The adventitia contained large cellula r and fibrous inflammatory cells. Wilson used a multistep detergent-enzymatic extraction process with hypotonic and hypertonic solutions and combining SDS with Triton on canine iliac a nd carotid arteries [62]. These grafts were implanted into the carotid and femoral arteries of dogs, foregoing anticoagulant and antiplatelet therapy. After four years, they found no aneurism formation, no calcification, an intact elastin network and no degradation of the collagen in the wall. The luminal surface was comp letely endothelialized and there was no evidence of chronic rejection. The cumulativ e patency rate was 95% compared to 100% occlusion in three months with synthetic material. In a follow up study, this group compared autograft canine saphenous veins to decellularized allograft canine carotid arteries for a CABG model. After 6 mont hs, 4 out of 7 saphenous vein grafts were widely patent with the remaining three occl uded at the 2 week angiogram. The patent grafts showed substantial endot helial cell coverage on the lume n. For the allografts, 4 of 9 were widely patent with the remaining 5 thrombosed by the first 2 weeks. There was no inflammation, but minimal cell repopulati on and no evidence of endothelialization except near the anastomoses. The same group at the University of Toronto used canine

PAGE 64

52 iliac and carotid arteries for femoral interpos ition bypass grafts [76]. At 4 weeks, the grafts were widely patent with some cells on the lumen consistent with endothelial cells. The media was found to be acellular. A study with a proprietary process at Life Net (Virginia Beach, VA) using allograft pulmonary artery for a patch reconstruction to repair great vessels in sheep also compared cryopreserved allograft and cryopreserved decel lularized allografts as well as fresh decellularized allografts [61]. After 20 weeks, all patches incorporated into the great vessel repairs without aneuri sm formations or infection. There was no statistical difference in the explant and implant dimens ional ratios. The cryopreserved allografts became acellular over time and developed a fibrous sheath consisting of layers of fibroblast cells that line the luminal side and encapsulating the adventitia. This is similar to the foreign body response seen in surgical implants. The decellularized tissue showed time-dependent recellularization at 10 w eeks and 20 weeks. There was partial endothelialization along the lumen and cells infi ltrating the elastin bands within the wall. Staining for -actin indicated that most of the inf iltrating cells were biologically active myofibroblasts. The cryopreserved grafts were positive for calcification while none was evident in the decellu larized tissue. Grauss compared cellular and acellular valve grafts in the desce nding aorta of a rat for 21 days [68]. All cellularized explants showed leaflet deformation with a considerable decrease in colla gen and a slight decrease in elastin. There was a complete loss of smooth muscle cells in the media of the aortic root accompanied by multifocal disruption of elastin fibers. On the other ha nd, all of the decellulari zed explants showed leaflet preservation with elastic fibers in the media normally arranged and similar

PAGE 65

53 collagen content when compared to the nont ransplanted tissue. The media of the acellular allograft was also void of smooth muscle cells. Hilbert evaluated the morphologic characte ristics in a small-diameter freeze-dried decellularized carotid artery of a goat trea ted with a nonionic detergent solution [70]. This solution contained inhibitors, such as proteoase, reactive oxygen species and other free radicals. Histological analysis showed complete endothelial removal while preserving the basement membrane. Additi onally, no cells were seen in the intima, media or adventitia. Although, there were remn ants of smooth muscle cells in the media without nuclei. The extracellular matrix s eemed well preserved, but oval to circular spaces were occasionally noted in the media. The internal elastic lamina remained intact, as well as layered elastic laminae in the media. This group evaluated these decellularized freeze-dried graf ts as a carotid interposition graft in goats fo r 6 to 7 months and compared them to autologous cephalic vein grafts. All grafts we re patent upon explantation, a nd no significant dimensional changes or aneurism formations were noted in the allografts. There was marked variability in the thickness of the neointima within as well as between the allografts. There was also great variation along the length of the autografts. Although, the neointimal thickness of the allografts were comparatively thin at 133 234 m ( 1 to 640 m) compared to a thickness of 249 224 m (8 to 800 m) for the autografts. The autografts showed a luminal surface lined wi th a layer of endothelial cells while the allograft showed a discontinuous layer. Even in these regions of incomplete endothelialization, thrombi were not obser ved. Both types of grafts showed myofibroblasts, collagen and proteoglycans.

PAGE 66

54 Upon explantation, the allograft appeared relatively acellular in the media. Among the dense collagen and elasti c lamellae, there were focal cellular regions containing myofibroblasts of the host. The authors beli eve that the number of host myofibroblasts in the media may have been significantly limited by the presence of dense collagen bundles and smooth muscle cell remnants. Host cell migration was most apparent close to the anastomoses and was rich in proteoglycans. The variable regions of host cell migration from the neointima in to the media appeared to occur through fenestra tions in the internal elastic lamina. Most cell migr ation appeared to occur in the adventitia along the length of the grafts. The adventitia looked healed a nd repopulated by connec tive tissue cells. Histologically, there was no evidence of calci fication and inflammatory cells were not present in the graft wall. Electron microscopy showed calcification of minute remnants of cell membranes. Ingrowth of host blood vessels was not observed. Conklin investigated a small-caliber xenograft with a porcine common carotid artery model [2]. This group followed cell ly sis with multiple enzymatic digestions and detergent washes while agitating. Histology and electron microscopy showed complete removal of cellular components while the extrace llular matrix appears to remain intact. H&E stain showed no signs of remaining nuclear material in the walls of the vessels and TEM showed complete removal of cellular material from the medial layers. Histology showed an intact internal elastic lamina al ong with elastin lamellae in the media. TEM showed the basic extracellula r microstructure remained intact after processing. In addition to reducing immune reactions and decreasing thrombogenicity of the luminal surface by decellularization, this group wanted to ensure that the process would maintain a graft with similar mechanical proper ties to the native vessel, high strength and

PAGE 67

55 good handling characteristics. They investigat ed the compliance and burst strength of the vessels on a custom-built system that measur es the diameter changes while increasing the intraluminal pressure. The compliance is expressed as the percentage diameter change per mm Hg change in pressu re normalized to the complia nce of fresh vessels. The average compliance was calculated as the sl ope of the linear regression line in the physiological pressure ra nge (70 130 mm Hg). The mechanical analysis was performed on decellularized vessels, decellularized and heparin-treated vessels, fresh vessels alcohol preserved vessels, and ePTFE prosthetic grafts. The ePTFE grafts were the least compliant (0.025% and the decellularized grafts were slightly more compliant (0.182%) than the fresh vessels (0.172%). Alcohol caused the tissue to stiffen some (0.160%) while the heparin treatment produced a much stiffer graft (0.0975% ). During the burst testing, none of the fresh vessels burst within the limit of the pressure transduc er (2300 mm Hg). One out of four of the decellularized vessels burst w ithin the limit at 1654 mm Hg. Although there may be some strength loss measured in that one vessel, there still is a high safety of margin over ten times the physiological pressu re. Unfortunately, only four samples were tested in this group. Additionally, there wa s no difference found in the average suture retention strength between the fresh and the al cohol and heparin treated specimens. This only gives preimplantation mechanical resu lt with evidence of some effect. Once implanted and the cellular and remodeling respon se occurs, there could be an exaggerated effect on the samples that may cause minute damage in the vascular wall. This group examined their decellularized process with heparin cross-linking on a carotid artery bypass in dogs. Two animals eac h were euthanized at 24 and 67 days and

PAGE 68

56 the explants were prepared similarly for histology and stained with H&E or immunohistochemically for -actin or Factor VIII. At 24 days, fibroblast-like cells appeared to densely populate the media and th ere were few endothelial-like cells lining the lumen. After two months, the dense -actin staining suggested smooth muscle cells densely populated the vessel wa ll and Factor VIII staining confirmed that endothelial cells lined the lumen. In a clinical study conducted by Hawkins using the CryoLife process, 14 children (8.5 7.9 years) received d ecellularized, cryopreserved allografts, 6 were patch insertions and 8 with valved pulmonary allogr afts [67]. These groups were compared to 20 historical control subjects ( 1.7 2.4 years), 8 with valved and 12 with allograft patch insertions. There was no attempt to match ABO blood types in either group. The effect on the immunogenicity was measured at 1, 3 and 12 months by frequency of panelreactive HLA alloantibodies (PRA): class I (HLA-A, HLA-B, and HLA-C) and class II (HLA-DR/DQ). A flow cytome try technique was used to ca lculate the percentage of fluorescent positive beads, indicative of the pe rcentage of PRA. Antibody levels were significantly higher from the preimplantation le vels for both classes at all time points. The antibody levels were significantly lower in the decellularized allograft. There was a marked reduction in staining for class I a nd class II histocompa tibility antigens. However, there was no work done in th is study to determine whether reduced immunogenicity will truly allow tissue ingrow th and improved long-term durability in patients.

PAGE 69

57 Purpose of This Study This segment of the research was aimed at determining the appropriate processing conditions to balance decellularization with ma trix integrity. The tissue will be treated with different detergents and enzymes at va rious temperatures for specified times. The level of decellularization w ill be examined histologically and the amount of antigen containing material will be de termined immunohistochemica lly. The matrix integrity will be assessed with an enzyme digestion assay for collagen. The aim is to provide parameters that produce efficient cell and cellular debris remova l without damaging the insoluble scaffold. Materials and Methods Cell removal The effectiveness of cell and cellular de bris removal was evaluated using two detergents and two enzymes at various concentr ations (Table 2-1). The detergents used in this study were Triton-X 100 at 0.25% (v/v) and 1.0 % (v/v) and sodium deoxycholate (Cholate) at 0.25% (w/v) and 1.0% (w/v). Due to viscosity issues with sodium deoxycholate at low temperatures tests were repeated at 0.5 % (w/v). One treatment group combined Triton-X 100 with sodium deo xycholate. Trypsin alone at 0.05% (w/v) or 1.0% (w/v) and in combination with 0.02% (w/v) EDTA were the enzymes used in the decellularization process. ED TA was also combined with Triton for a comparison since the Trypsin samples appeared grossly degrad ed after treatment. A human saphenous vein sample 10 mm in length was placed in a solution at 37oC or 48oC for either 6 and 24 hours while continuously shaking. Samples were then rinsed in PBS at room temperature for 10 minutes while continuously shaking. Following the rinse step, the samples were cut in half and either prepared for histology or i mmunohistochemistry.

PAGE 70

58 Table 2-1. Processing parameters to evalua te decellularization of human saphenous vein. Chemical Concentrations (%) Trypsin 0.05 (w/v) 1.0 (w/v) Trypsin / EDTA 0.05 / 0.02 (w/v) 1.0 / 0.02 (w/v) Triton 0.25 (v/v) 1.0 (v/v) Cholate 0.25 (w/v) 1.0 (w/v) Triton / Cholate 0.25 / 0.25 1.0 / 0.5 Triton / EDTA 0.25 / 0.02 1.0 / 0.02 Histology Samples were placed in 10% buffered formalin immediately following the rinse step and remained there for at least 24 hours before further preparation. Then, the samples were embedded in paraffin and stai ned with hematoxylin and eosin (H&E) to determine cellularity. Immunohistochemistry Immunohistochemistry was used to determin e the level of antigen s remaining in the tissue after treatment. Only cl ass I MHC antibodies were used for evaluation since class I antigens are present in a much greater abundan ce than class II antigens. Samples were prepared in frozen blocks and sectioned with a cryostat. The sections were thawed and then rinsed for blocking and antibody staini ng. Sections were incubated for 30 minutes in primary antibody (mouse anti-HLA-ABC class I MHC) solution and then for anther 30 minutes in diluted biotinylated secondary an tibody (Vector ABC Elite kit) solution. An enzyme substrate, NovaRed (Vector), was used to stain the antibodies red. The counter stain was hematoxylin providing a blue cont rast. A full protocol can be found in Appendix B.

PAGE 71

59 Optical evaluation A Zeiss Axiophot 2 microscope with a mo torized Ludl scanning stage was used with tile field mapping at 5x to capture the entire image in one frame. Particle analysis was performed in Image J to determine the ra tio of antigen staining (red) to the total tissue sample. Matrix integrity One of the major structural components of the extracellular matr ix that gives veins their strength is collagen. Damage to collagen can result in mechanical failures such as increased compliance or aneurism formation. It can also result in a cellular response due the exposure of the triple helix when stabil ity is compromised. The amount of collagen denaturation was assessed to evaluate pro cessing conditions with a quantitative enzyme digestion assay. Saphenous vein samples ( 0.2 to 0.25 g) were treated with trypsin, a serine protease enzyme that is able to dige st only those collagen fibers that possess a break in the helix known as denaturation. Digested and undigested fractions are separated and hydrolyzed with concentrated hydrochloric acid (HCl) to release free amino acids from each fraction. Following neut ralization of the acid in 1 N NaOH, levels of hydroxyproline (an amino acid present in hi gh concentrations only in collagen) are assessed in each fraction by a colorime tric method at a wavelength of 550 nm (Chloramine T binding, and reduction of the substrate DAB to a colored end product). The level of denatured collagen in a given samp le is then expressed as a percent of the trypsin soluble fraction to the sum of both tr ypsin soluble and trypsin insoluble fractions. Statistics Data was analyzed with a two-sample student’s t-test with an value of 0.05 using Minitab Statistical softwa re. A p-value less than was considered a significant affect.

PAGE 72

60 Results The red staining as a result of the class I MHC antigens on the endothelial cells and the smooth muscle cells ranged from a bright red to a dark brown. A blue and green filter was applied to only allow red to pass. Obvi ous areas of antigen staining were removed due to this variation in re d hue. Therefore, only a blue filter was applied where a threshold value was determined when the blue -stained negative control disappeared while both green and red were allowed to fully pass to capture a greater portion of the antigen staining. All samples were compared to their negatives for a percentage of red staining. Figure 2-1 shows a tile mapped image of the a positive control and a negative control. Figure 2-2 shows the result afte r a threshold filter was appl ied. The negative control on the right is almost completely devoid of any particles, wherea s the positive control removes the background stained tissue and retains the stained antigens. Figure 2-1. Positive (left) and ne gative (right) control samples.

PAGE 73

61 Figure 2-2. Result of the application of a thre shold filter that limits the pass of high intensity blues. This procedure was applied to all images. There were three samples per treatment group and the result of each sample was averaged from three sections. The results for the Trypsin and Trypsin + EDTA treatment groups are shown in Figure 2-3 and the results for the Triton-X 100 and sodium deoxycholat e (cholate) treatment groups are shown in Figure 2-4 and Figure 2-5. The treatment groups are labeled according to the detergent or enzyme used followed by the processing conditions. For the Trypsin and Trypsin + EDTA groups, the concentrati ons ranged from 0.05% (L) to 1.0% (H), the processing temperatures were 37oC (L), 48oC (M) and 55oC (H), and the processing times were 6 hrs (L), 17 hrs (M) and 24 hrs (H). For the Tr iton and cholate groups, the concentrations ranged from 0.25% (L) to 1.0% (H), th e processing temperatures were 37oC (L) and 48oC (H), and the processing times were 6 hrs (L ) and 24 hrs (H). The reason the former groups had additional conditions was due to an error in the incubator temperature and an experimental error with the time. The sample s were run again at th e right conditions but the samples with the errors were retained for comparison.

PAGE 74

62 Trypsin and Trypsin/EDTA-5.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 45.00PC NC T (L L L) TE (L L L) T (L L M) TE (L L M) T (L L H) TE (L L H) T (L M L) T (L M H) T (L H L) T (L H H) T (H L L) T (H L M) T (H L H) T (H M L) TE (H M L) T (H M H) TE (H M H) T (H H L) TE (H H L) T (H H H) TE (H H H)Treatment% Red Figure 2-3. The amount of class I MHC rema ining in the tissue when compared to a negative control for trypsin and tryps in plus EDTA. The treatment group labels are T (Trypsin) and TE (Trypsin + EDTA) and the conditions are in order of concentration (L = 0.25%, H = 1.0%), temperature (L = 37oC, M = 48oC, H = 55oC) and time (L = 6 hrs, M = 17 hrs, H = 24 hrs). Triton and Cholate-5.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00PC NC TR (L L L) CH (L L L) TR (L L H) CH (L L H) TR (L H L) TR (L H H) TR (H L L) TR (H L H) TR (H H L) CH (H H L) TR (H H H) CH (H H H)Treatment% Red Figure 2-4. The amount of class I MHC rema ining in the tissue when compared to a negative control for Triton-X 100 and s odium deoxycholate. The treatment group labels are TR (Triton) and CH (c holate) and the conditions are in order of concentration (L = 0.25%, H = 1.0%), temperature (L = 37oC, H = 48oC) and time (L = 6 hrs, H = 24 hrs).

PAGE 75

63 Combination of Triton and Cholate-4.00 -2.00 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00PC NC TR LLL CH LLL TRC LLL TR LLH CH LLH TRC LLH TR HHH CH HHH TRC HHHTreatment% Red Figure 2-5. The amount of class I MHC rema ining in the tissue when compared to a negative control for Triton-X 100 and s odium deoxycholate combined. The treatment group labels are TR (Triton) CH (cholate), and TRC (Triton plus Cholate) and the conditions are in or der of concentration (L = 0.25%, H = 1.0%), temperature (L = 37oC, H = 48oC) and time (L = 6 hrs, H = 24 hrs). A significant reduction in red staining at 6 hours was only with an increase in temperature and concentration. Six gr oups in the Trypsin and Trypsin/EDTA experiments had a less then 5% red staining when compared to the negative controls. Treating the tissue with 1.0% Trypsin or 1.0%/0.02% Trypsin/EDTA at a minimum of 48oC and 17 hrs significantly reduced the level of antigens. Only at temperatures of 55oC was there a significant decrease in red staining within 6 hrs. The addition of EDTA has an inconsistent effect on the re moval of antigens from the tissue. For Triton and Cholate, all but three treatment groups we re below 5% and all of them were below 10% when compared to the controls. Cholate exhibited the least amount of staining with approximately 100% re moval at high concentrations and high temperatures. There was no statistical di fference in samples treated with Triton and Cholate combined compared to individually (Figure 2-5). The addition of EDTA to Triton did enhance the removal of cells at 0.25% for 6 hours at 37oC. Representative

PAGE 76

64 samples from each group are shown in Figure 2-6 for comparison. Additional representative samples for each chemical agen t at each of the processing conditions are illustrated in Appendix C-E. Figure 2-6. Representative samples of tissu e treated with Trypsin (left), Triton-X 100 (center) and sodium deoxycholate (r ight) and stained for class I MHC antigens. The results of the immunohist ochemical stain were compar ed to histology slides stained for cellularity (Figures 2-7 to 2-10). There is still ev idence of cells in the samples treated with Trypsin and that is correl ated with red stained tissue in the immunohistochemical slides. Figure 2-7 appe ars to have a reduction in the amount of cellular material present in the matrix, but the red stained segment shows a majority of the tissue still contains antigens. The extr acellular matrix is preserved in each segment except with 1.0% Cholate at 48oC. There are large holes within the venous wall and separation between the fibers.

PAGE 77

65 Figure 2-7. A comparison of the amount of cellularity (blue) visible in histology compared to the amount of antigens (red ) present in the ti ssue. The samples were treated with 0.05% Trypsin at 37oC for 6 hours. Figure 2-8. A comparison of the amount of cellularity (blue) visible in histology compared to the amount of antigens (red ) present in the ti ssue. The samples were treated with 0.05% Trypsin at 37oC for 24 hours.

PAGE 78

66 Figure 2-9. A comparison of the amount of cellularity (blue) visible in histology compared to the amount of antigens (red ) present in the ti ssue. The samples were treated with 0.25% Cholate at 37oC for 6 hours. Figure 2-10. A comparison of the amount of cellularity (blue) visible in histology compared to the amount of antigens (red ) present in the ti ssue. The samples were treated with 1.0% Cholate at 48oC for 6 hours. The conditions for each chemical that produced a reasonable reduction in red staining were analyzed quantitatively for their effect on collage n. Triton at 1.0% and Cholate at 0.25% were processed at 48oC for 24 hours under constant agitation. The samples were immediately placed in an ice bath after the 24 hours and then rinsed in cold phosphate buffered saline (PBS) to halt the de gradation process due to the temperature increase. The results were compared to the natural degradation of the untreated controls

PAGE 79

67 and are shown in Figure 2-11. A ratio of one indicates no difference from the untreated controls. A significant degrada tive effect was not seen with Triton, and cholate produced a more of an effect on the tissue matrix. The Trypsin showed le ss than half of the degradation of the untreated controls. Collagen Degradation Effects of Decellularization at 48oC 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20.25% Cholate1.0% Trypsin0.25% Triton% Sample Degradation / % Control Degradation Figure 2-11. Results of the collagen degr adation assay after decellularization at 48oC for 24 hours. The samples were donor matche d to reduce the inherent variability in tissue.

PAGE 80

68 CHAPTER 3 STERILIZATION Introduction Allograft tissue is obtained from cadaver s and processed under aseptic conditions so it cannot be considered a truly sterile conduit. Current processing methods at an AATB accredited tissue bank using an antimicrobi al soak usually discards approximately 27% of donor tissue due to pos itive first and final culture s from bacterial and fungal contamination. Additionally, the risk of tran smission of viruses and other pathogens can be reduced but not eliminated through donor screening. Limitations in current serological tests used to sc reen tissue donors result in highe r rates of bloodborne virus (BBV) among tissue donors than blood donors. Over the last several years multiple examples of donor to host disease transmissi on have been documented. These examples include transmission of hepati tis C (HCV), bacteria, and fungi, and have resulted in serious morbidity and even mortality [8 ]. A method for sterilization and viral inactivation that is effectiv e in removing and inactivating contaminants and endogenous materials without adversely affecting the tissu e matrix and functionality would increase availability and provide a sa fe alternative venous conduit. There would no longer be concern for the infection window period or erro rs in testing and fa lse negative results [83].

PAGE 81

69 Unfortunately, traditional methods of sterili zation, such as irradiation and ethylene oxide (ETO), have deleterious effects on vasc ular tissue that limit or preclude their use for allograft implantation. ETO use in hosp itals has decreased due to concern for the toxicity of residuals as well as its classifi cation as a carcinogenic and mutagenic agent [84, 85]. Ionizing radiation has been found to be effective ag ainst bacteria at 1.5 to 2.5 Mrad, but viral inactivation need s approximately 4 Mrad [86, 87] It has been shown that a dose greater than 3 Mrad has a deleterious effect on the biomechanical properties of human patellar tendon [88]. Additionally, soft tissue such as skin resulted in major structural changes after gamma irradiation [84]. The active surveillance activities and the outbreak investigation carried out by Centers for Disease Control (CDC) during 2002 highlights the fact that the spectrum of bacterial pathogens associated with allograft-associ ated infections include Gram-positive ( S. aureus and Enterococcus) and Gram-negative bacteria, especially those that ferment lactose [42]. The importance of validating a ny sterilization techni que against a mixture of pathogens is underscored by the fact that 19% of the allograft-associated infections ascertained during the active case finding peri od were polymicrobial. The frequencies and concentrations of various pathogens in polymicrobial infections remain unknown and will vary from recipient to recipient, therefor e, are not predictable. Based on data in the CDC investigation, the antimicrobial soluti on used should be validated against the following pathogens, at the very minimum: Staphylococcus aureus, Enterococcus faecalis, Pseudomonas aeruginosa, Escherichia coli, Clostridium species and Bacillus species [42].

PAGE 82

70 The current common disinfection method is an antimicrobial soak in which the tissue is aseptically processed with antimicr obial and sometimes antifungal cocktail and should not be considered sterile [35]. Additi onally, the method for testing sterility may be flawed [45]. Specifically, re sidual antibiotics from the treatment process have been shown to inhibit growth in post-treatment surveillance cu ltures producing false-negative results leading to cases of septic arthri tis, hepatitis C transmission, Streptococcus pyogenes (GAS), and Clostrid ium spp [8, 46]. Of note, cases of soft tissue allograftassociated disease transmission by spore-fo rming organisms that are resistant to antimicrobial solutions have been repo rted within the last few years [8, 47] Vascular tissue is vulnerable to organisms (including fungi and viruses) inherent in the donor or transmitted during procurement or processing that are resistant to the antimicrobials currently used for allograf t disinfection [23, 48, 49]. The fact that allograft-associated infecti ons result from an unpre dictable source of pathogens makes it difficult to validate a ster ilization process. One way to overcome a polymicrobial investigation is to use a representative bacter ium that is more difficult to kill than anything that will be detected on the tissue. The bacterium Bacillus stearothermophilus is a spore forming microbe that is resistant to a wide range of temperatures and processing conditions. A 6 log reduction in this spore would guarantee protection against the bacter ia typically found on contaminated vascular tissue. Disinfection and Sterilization Processes It is estimated that 60% of all human inf ections are caused by viruses [89]. Viruses are composed of DNA and RNA, lipids, and a protein membrane for protection. Their physical characteristics are cl assified as either lipophili c or hydrophilic. Lipophilic viruses have a lipid shell known as an envelope These viruses are ea sier to inactivate

PAGE 83

71 with removal of the envelope via a disinfectant. Such viruses include HIV, RSV and hepatitis B (HBV). Hydrophilic viruses are known as nonenveloped viruses since they do not have a shell. Instead, they have a very tough protein coat that is difficult for some disinfectants to enter. Such viruses include poliovirus, rh inovirus and hepatitis A (HAV). Transmission of viruses through blood and blood products has led to efforts in inactivation and removal of endogenous materi als and possible contaminants. The most widely used methods are pasteurization, chem ical inactivation, i rradiation and solvent extraction of essential lipids. Pasteurization involves destroying or retarding the growth of bacteria without destroying biological activity of the samp le by heating to a moderate degree (60 – 70oC) for a substantial period of time rath er than boiling [90]. This process is used to disinfect blood products such as Al pha-1 proteinase inhib itor concentrates from pooled human plasma [91]. Injected samples were maintained at 60oC for 10 hours and then tested for viral inactivation. The thermolabile viru ses, vesicular stomatitis virus, herpes simplex virus-1, visna, and HIV, were inactivated within 1 hour while poliovirus took almost 5 hours. Porcine parvovirus, one of the most thermo-stable viruses known, was reduced by 3 logs but there was still a 1.7 log infectivity rema ining after 10 hours. Clinical data 6 months post-infusion were all negative for hepatitis B (HBV), non-A nonB hepatitis, and HIV. Solvents and detergents alone or in combination are more effective than pasteurization, with success in disrupting enveloped viruse s but nonenveloped viruses are unaffected [83]. Triton-X 100 used with th e organic solvent tr i(n-butyl)phosphate (TNBP) was shown to inactivate very large quantities of HBV, BCV and HIV in pooled plasma [83, 92]. The samples we re incubated for 4 hours at 30oC in 1% TNBP and 1%

PAGE 84

72 Triton-X 100. This process achieved a greater than 6 log reduction in HBV and HIV and a greater than 5 log reduction in HCV. These results were tested in vitro by injecting chimpanzees with the plasma concentrates, and the animals were negative for HBV and HIV. Alcohols are known effective antiseptic agents used for hand sanitation and thermolabile instrument disinfection (70%) in hospitals worldwid e [93, 94]. The most widely used alcohol is isopr opyl alcohol. In dilutions of 60 to 95% it kills bacteria, mycobacteria, fungi and large or lipid-cont aining viruses [93]. However, it is not effective on hydrophilic viruses. Ethyl alcohol is used to inactivate hydrophilic viruses. A study compared the effectiveness of 70% alcohol in inactivation of HIV dried on a surface or in suspension [94]. The remaining 5.5 log titre after drying was completely inactivated within 1 min. This time was affect ed by an increase of 100% in protein load increasing the inactivation to 4 10 min, providi ng a significant barrier to dried viruses. Alternatively, the high titres of HIV in susp ension were rapidly inactivated independent of protein load. Of all microorganisms, spores are the most resistant to antimicrobial treatment. Spores are small, single-celled reproductive bod ies that are highly resistant to heat and are capable of growing into a new organism. Bacteria form spores and this dormant phase is a response to advers e environments as a means of protection. Recently, more attention has been placed on the need to inactivate spore forming bacteria since crossinfection by Clostridium spp. ( a species found in intestinal contents) has been reported on various soft tissue grafts that has resulted in graft removal or even death [42, 84].

PAGE 85

73 Oxidative processes involv e electrons from a reduc ing agent (Oxygen) being transferred to the oxidizing agent. This proc ess is used in sterilizing equipment with compounds such as hydrogen peroxide, peraceti c acid, peroxysulphates, chlorine dioxide and ozone. Hydrogen peroxide at a concentration of 3% is us ed as an antiseptic agent and combined with ultraviolet light to disinf ect cartons for food produc ts. Its sporicidal activity at higher concentrations has been investigated for me dical and dental instruments [95]. Suspensions of 106/ml Bacillus atrophaeus spores were maintained at room temperature in 7.5% hydrogen peroxide. No growth was detected after a 6 hour incubation period. Peracetic acid (PAA CH3C(O)OOH) is an organic oxidant that is highly effective against a wide variety of bacteria, fungi, viru ses and spores [84, 96]. It has been shown to be an effective antiviral agent for both enveloped and nonenveloped viruses at 0.2 to 0.35%. This chemical rapidly penetrates the microorganisms and interactions result in the release of oxygen free radical s causing the destruction of enzymes [96, 97]. It has been used for sterilization of bone, heart va lves and small intes tine without significant adverse effects on the morpohology and structur e [84, 96]. Another advantage is its low toxicity since its residues, oxygen, carbon dioxide and water, are natural and harmless. PAA combined with ethanol was examined for its effectiveness in inactivating a wide range of microorganisms in allogeneic bone tissue [85, 98]. In one study, nonviable organisms were detected after 2 and 4 hours in cubation periods. Ther e was a greater than 5 log reduction in Staphylococcus aureus, E. faecium, Pseudomonas aeruginosa, B. subitilis, Clostridium sporogenes, Mycobacterium terrae, and C. albicans [85]. The same group tested treatment of bone spongiosa cubes injected with three enveloped and three

PAGE 86

74 nonenveloped viruses and then treated with PAA with ethanol [98]. There was more than a 4 log reduction in all vi ruses except in one of the nonenveloped viruses, HAV. A greater than 7 log reduction was achieved in HAV after implementing a delipidating step. A porcine-derived scaffold made out of small intestine submucosa (SIS) was treated at room temperature for 5 min to 2 hours with a PAA/ et hanol (0.18% in 4.8% mixture) solution to inactivate viruses endoge nous to porcine [96]. Relevant enveloped, nonenveloped and model viruses represented diffe rent virus families. Enveloped viruses were inactivated more easily w ith all viruses inactivated with in 30 min. In a later study, this group examined the retention of endothe lial cells seeded onto the matrix after sterilizing with the PAA/ethanol (0.1%) soluti on [99]. The proteins present on the matrix that may contribute to site-specific remode ling are type I collagen, type IV collagen and fibronectin (FN). These proteins were well pr eserved and, therefore, retained their ability to bind cells. Peracetic acid was compared directly to hydrogen peroxide in a wastewater medium [97]. PAA has shown good disinfection against enteric bacteria in wastewaters, but viruses and bacterial spores are more resi stant. On the other hand, hydrogen peroxide is not typically used alone due to its sl ow disinfection action and low efficiency. Wastewater-like test medium containing E. coli Enterococcus faecalis or Salmonella enteritidis were treated for 10 min in PAA or hydrogen peroxide. PAA dose of 0.3% achieved a 2 to 3 log reduction in enteric b acteria while peroxide doses of 0.3 to 15% achieved below 0.2 log microbial reductions. Enterococcus faecalis is one of the organisms recommended for validation of a tissue disinfection process. The higher reactivity of PAA compared to hydrogen peroxi de may be due to its ability to better

PAGE 87

75 penetrate the cell membrane causi ng disruption as well as its ability to block enzymes and transport systems in microorganisms. Also, hydrogen peroxide is ineffective due to its purity. It is highly reactive due to hydroxyl radicals produced when ferrous iron (Fe2+) is added to the solution (Harbor-Weiss/Fent on reaction) [100]. Additionally, some organisms may be protected against hydrogen pe roxide due to catalase enzyme presence. Unfortunately, the effectiveness of these oxidant s in killing spores has to be balanced by their effect in damaging the tissue matrix. Collagen is the only pr otein susceptible to fragmentation by hydroxyl radicals. Huang et al. studied the sterilization of human donor skin with PAA [84]. They immersed skin in 0.1% PAA for 3 hours at room temperature under constant agitation. The samples retained their dermal structure and components of the basement membrane. The collagen fibers maintained their normal ar chitecture with fine and wavy elastin fibers located among them in a normal pattern pos t implantation. The extracellular components were only analyzed qualitatively using histol ogy and degradation was not assessed which can give an indication to the remodeling response. Human bone-patellar tendon-bone (BPTB) grafts were treated in a similar mi xture at room temperature under low pressure for 4 hours [101]. The tendons were mechanical ly tested with no significant difference found in the viscoelastic pr operties, stiffness or maximum loading properties when compared to untreated controls. Again, no conclusions can be dr awn about biological healing or remodeling. Processing and Determination of Survivors Sterilants can fully kill or remove bacteria and viruses to specified sterility levels under appropriate conditions. According to the FDA Guidance document for the use of liquid germicides, sterilization is associated w ith total absence of vi able organisms [102].

PAGE 88

76 The first step in evaluating a sterilization process is the selecti on of an appropriate microorganism(s) as a biological indicator. Bacillus stearothermophilus is a species of gram-positive bacteria found in soil, hot spri ngs and spoiled food products. It is not a common contaminate of tissue products, but it is used as an indicator in instrument sterilization equipment since it is the hardest to kill spore known as it is highly resistant to high temperatures. The effectiveness of a sterilant is defi ned in terms of decimal reduction time (Dvalue). This is the exposure time (t) requi red under certain conditions to cause one log reduction (n) (90%) of the in itial population. A suspensi on is expected to follow a predictable death rate regard ing a plot of the amount of survivors over time. The negative reciprocal of the slope of regression lines of survi vor curves gives the D-value where f oN N n n D t / log A study using B. stearothermophilus as one of its biological indicator for PAA and hydrogen peroxide calculated the D-values of vegetative and spore forming bacteria [103]. PAA was represented by a 1% soluti on of commercially available Minncare (0.45% PAA + 2.2% hydrogen peroxide, pH 2.3) An initial spore population of 104 to 105 CFU/ml (colony forming units) was placed in either Minncare or 1.5% to 26.5% hydrogen peroxide at 25oC and removed at regular interval s. For vegetative bacteria samples were removed every minute while samples were removed every 5 minutes for spore forming bacteria. The survivors were analyzed on tryptic soy agar (TSA) pour plates at various dilutions.

PAGE 89

77 The most resistant bacteria against Minncare were B. stearothermophilus, B. subtilis and E. coli with D-values 2 to 3 times higher than the more sensitive A. calcoaceticus, E. cloacae and S. aureus The presence of PAA in Minncare reduced the D-value of B. subtilis 10 times compared to hydrogen peroxide alone (5.9 min vs. 55.2 min). D-value for B. stearothermophilus was 4.7 min after being treated at a much higher concentration of hydrogen peroxide (26.5%). Materials and Methods Reduction Curves (Survivor Curves) for Spores Methods In order to determine the reduction in spor es after chemical processing, an effective and consistent method needed to be developed. Initially, a 107 spores/ml titre of B. stearothermophilus was used to enumerate the remaining population after treatment on both TSA media plates and sheep’s blood agar (BA) plates. It was determined that residual hydrogen peroxide inhi bited the growth at concen trations 0.6% or higher for TSA plates and 6% or higher for blood plates (Figure 3-1). It is believed that the catalase contained in the blood agar pl ates inactivated the hydrogen pe roxide, so TSA plates were abandoned for further studies. A similar inve stigation was performed on PAA, but no inhibition was observed within the processing concentrations (Figure 3-2). Even though the blood agar plates allowed for dilutions as low as 1:10, the titre used was too low to detect more than a 3 log reduction in the process limiting the data collection for the survivor curves. Therefore, the titre used in this study was a 109 CFU/ml solution of B. stearothermophilus to allow for a 4 log reduction detection limit. Consistency in replicates and countable plates without moisture damage were a problem. To ensure consistency, each vial wa s vortexed before each replicate and all of

PAGE 90

78 the material was removed from the pipette tip onto the plate. Dispos able plate spreaders were used to prevent cross contamination. Th e moisture needed to be maintained at a certain level because the pl ates would dry out at 60oC, but if there was too much moisture the bacteria would streak across the plate and prevent an accurate count. The first procedure implemented was propping the plat es open in a lamina r flow hood to allow excessive moisture to dry. The second me thod employed was placing the plates in a beaker with a porous covering to trap so me moisture while allowing the excess to evaporate. This resulted in consistent and evenly spread bacteria on the pour plates. Hydrogen peroxide Hydrogen peroxide at a concentr ation of 3% was used at 50oC and a concentration of 6% was used at temperatures of 40, 45 and 50oC to calculate a curve for modeling the temperature effects on spore kill. A vial wi th 0.9 ml of germicide was prewarmed in a water bath and then injected w ith 0.1 ml of spore solution (108 spores/ml). Samples were removed and immediately placed in an ice bath at 10, 15, 20 and 30 min. The vial was well-mixed on a vortex, and then diluted 1:10 in series then 0.1 ml were added to three media plates. The plates were pl aced in an incubator between 55oC and 60oC for 20 to 24 hours. Peracetic acid Peracetic acid at concentrations of 0.1%, 0.05% and 0.005% were used only at 40oC to calculate a curve for modeling the con centration effects on spore kill. Initial inactivation rates at 50oC were too rapid to detect. A vi al with 0.9 ml of germicide was pre-warmed in a water bath and then in jected with 0.1 ml of spore solution (108 spores/ml). Samples were removed and immedi ately placed in an ice bath at various time points depending on the concentration. The vial was well-mixed on a vortex, and then

PAGE 91

79 diluted 1:10 in series before plating. The plates were placed in an incubator between 55oC and 60oC for 20 to 24 hours. TSA Inhibition on Log plot1 10 100 1000 00.0060.060.66% Peroxide% Control BA Inhibition Logarithm 5/10/0510 100 1000 00.0060.060.66% Peroxide% Control Figure 3-1. Bacillus stearothermophilus growth inhibition at di fferent dilutions of hydrogen peroxide on TSA and blood agar plates. BA Inhibition on Log plot1 10 100 1000 00.0010.010.1% PAA% Control Figure 3-2. Bacillus stearothermophilus growth inhibition at di fferent dilutions of peracetic acid on blood agar plates. Spike and recovery method Before enumerating the population rema ining on tissue following chemical processing, the amount recoverable from the tissue must be determined. The recovery process developed involves the use of soni cation, mechanical shak ing and agitation on a vortex. Veins were cut into 30 mm segments and clamped at one end. The lumen of the vein was injected with 108 spores in a 0.1 ml inoculum. The other end was clamped and the veins were hung at 4oC for 15 minutes avoiding contact with any surfaces. The samples were then cut from the clamps and split open over a centrif uge tube containing 39.9 ml of Letheen broth containing Tween to ex pose the lumen to the recovery process.

PAGE 92

80 A new pair of scissors or a new scalpel blad e was used for each vein and the segments were cut over the broth to catch anything that may not have adhered to the surface of the tissue. The tubes were sonicated at room temperat ure for ten minutes to break the bacteria from the surfaces. The tubes were then placed on a shaker where they were vigorously agitated for another 10 min. The Letheen br oth containing the tissue sample was then vortexed for 30 s and diluted in a series of 1:10 dilutions with 0.1 ml of each dilution plated on blood agar plates in triplicate. The plates were incubated overnight at 60oC. Reduction curves The procedure for spike recovery was used to collect data for the reduction curves. Vein segments were inoculated and then treated in one of the following conditions. Hydrogen peroxide (6%) was analyzed at 40oC and 50oC while peracetic acid (0.1%) was only tested at 40oC. At each time point, three sample s were spiked and treated with one control for recovery efficiency. Data Presentation Typically, a survivor curve is constructed with the initial popul ation at time zero and the resultant populations at various time poi nts. Since not all data points could be completed in the same day, some reduction curves were constructed to compare log reduction in the controls for each experiment over time. This normalized the data to its own experimental episode. Matrix Integrity Achieving the appropriate log reduction mu st not come at the expense of the integrity of the extracellular matrix. The am ount of collagen denaturation as a result of chemical treatments was analyzed by a quantitative enzyme digestion assay. The effect

PAGE 93

81 of hydrogen peroxide at 6% at 40oC and 50oC on collagen was compared to PAA at 0.1% at the same specified temperatures. Saphenous vein samples (0.2 to 0.25 g) were treated with trypsin, a serine protease enzyme that is able to digest only those coll agen fibers that possess a break in the helix known as denaturation. Digested and undigest ed fractions are sepa rated and hydrolyzed with concentrated hydrochloric acid (HCl) to release free amino acids from each fraction. Following neutralization of the acid in 1 N NaOH, levels of hydroxyproline (an amino acid present in high concentrations only in collagen) are assessed in each fraction by a colorimetric method (Chloramine T binding, and reduction of the substrate DAB to a colored end product). The level of denature d collagen in a given sample is then expressed as a percent of the trypsin soluble fraction to th e sum of both trypsin soluble and trypsin insoluble fractions. Results Suspension Sterilization The D-value for 3% peroxide was too high at 50oC to be considered for further analysis. The D-value for 6% peroxide was reduced by approximately 80% by an increase in temperature from 40oC to 50oC. A 6 log reduction at 40oC would take almost 5 hours where it takes less than an hour at 50oC. Results for PAA we re only collected at 0.05% PAA or less and at 40oC since at higher concentratio ns and higher temperatures the spore inactivation was too rapid to de tect. A 6 log reduction at 0.05% PAA only takes 11 minutes and the Dvalue at 0.005% PAA at 40oC is still less than 6% peroxide at 50oC.

PAGE 94

82 6% Hydrogen Peroxide at 50oCy = -0.1013x + 8.4211 R2 = 0.97 4.00 4.50 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 05101520253035Time (min)Log (population) Figure 3-3. Survivor curv e for 6% peroxide at 50oC resulting in a D-value of 9.87. 6% Hydrogen Peroxide at 45oCy = 0.0435x 0.2305 R2 = 0.97 -0.5 0 0.5 1 1.5 2 2.5 3 010203040506070Time (min)Log Reduction Figure 3-4. Reduction curve for 6% peroxide at 45oC resulting in a D-value of 23.0.

PAGE 95

83 Log Reduction at 40oCy = 0.0211x 0.0553 R2 = 0.9371 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 05101520253035Time (min)Log Reduction Figure 3-5. Reduction curve for 6% peroxide at 40oC resulting in a D-value of 47.39. Predictive Modeling for 6% Peroxidey = -0.0756x + 4.7213 R2 = 0.9906 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 40424446485052Temp (C)Log (D) Figure 3-6. Linear regre ssion of the log of the D-values fo r each temperature. This gives an equation to interpolate and possibly extrapolate the D-values at various temperatures with 6% hydrogen peroxide.

PAGE 96

84 0.05% PAA at 40oCy = 0.5265x + 0.0492 R2 = 0.9395 0 0.5 1 1.5 2 2.5 3 3.5 4 01234567Time (min)Log reduction Figure 3-7. Reduction curve for 0.05% PAA at 40oC resulting in a D-value of 1.90. 0.005% PAA at 40oCy = -0.1551x + 8.0348 R2 = 0.999 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 024681012Time (min)Log (pop) Figure 3-8. Survivor curve for 0.005% PAA at 40oC resulting in a D-value of 6.45. Tissue Sterilization The recovery method developed in the st udy consistently achieved between 52 and 73% spike recovery. A control was used at each experimental episode to provide a correction factor for the treated samples. The recovered population divided by the correction factor provides th e surviving population after treat ment. The reduction curves for spiked tissue treated with 6% hydrogen peroxide at 50oC and 40oC are shown in

PAGE 97

85 Figure 3-9 and Figure 3-10, respectively. A ccording to the reduc tion curves, a 6 log reduction can be achieved in 1 hours at 40oC and in less than 40 minutes at 50oC. It was noted that the D-values of 6% peroxide with tissue were less than the D-values for spores treated in suspension. The D-value was decreased by approximately 4 minutes to 5.34. Human Tissue Treated with 6% Peroxide at 50 Cy = 0.1508x + 0.0885 R2 = 0.8099 0 1 2 3 4 5 6 05101520253035Time (min)Log Reduction Figure 3-9. Reduction curve fo r 6% hydrogen peroxide at 50oC resulting in a D-value of 6.63.

PAGE 98

86 Human Tissue Treated with 6% Peroxide at 40oCy = 0.0659x 7E-16 R2 = 0.9921 -0.5 0 0.5 1 1.5 2 2.5 05101520253035Time (min)Log Reduction Figure 3-10. Reduction curv e for 6% peroxide at 40oC resulting in a D-value of 15.17. Optimum spore kill is achieved at a high concentration of germicide and at a high temperature. The integrity of the tiss ue may be compromised at these extreme conditions. The ratio of degraded collagen after treatment in 6% hydrogen peroxide and 0.1% PAA at 50oC for one hour were compared to unt reated controls. The results are shown in Figure 3-11. There was not a statis tical difference between peroxide and PAA, but both showed significantly altered matrix properties. A reducti on in temperature to 40oC showed no significant difference in the coll agen degradation of samples treated with 6% peroxide or 0.05% PAA for one hour co mpared to the controls (Figure 3-12).

PAGE 99

87 Germicidal Effect on Collagen at 50 C0 0.5 1 1.5 2 2.5 3 3.5 4 4.56% Peroxide0.1% Peracetic% Sample Degradation / % Control Degradation Figure 3-11. Collagen degradation of ti ssue treated with 6% hydrogen peroxide and 0.1% PAA at 50oC. Both 6% hydrogen peroxide and 0.1% PAA had a detrimental effect on the collagen when compared to a donor matched control. Ratio 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Control6% Peroxide0.05% Peracetic TreatmentsRatio Figure 3-12. Collagen degradation of ti ssue treated with 6% hydrogen peroxide and 0.05% PAA at 40oC. Both 6% hydrogen peroxide and 0.05% PAA at 40oC were statistically similar to the cont rol of a donor matched control.

PAGE 100

88 CHAPTER 4 PRESSURE-INDUCED INCREASE IN PORE VOLUME FOR INTRODUCTION OF DECELLULARIZATION AGENT INTO VEIN ALLOGRAFTS A thorough review of decellu larization methods in the literature is provided in Chapter 2, detailing commons problems with th ese methods and its inability to achieve complete removal of cells and cellular debris as well as th e long processing times needed for successful decellularization. Without comple te removal of cells within the vein wall, an immune response is elicited and in some cas es can lead to evidence of calcification or inhibition of cell migration [1, 5, 25]. A dditionally, it is difficult to remove the decellularization chemicals that become embedde d within the matrix which has a residual toxic effect resulting in cell death and th e prevention of cell migration once implanted [70]. Processing times ranged from 24 hours to several days under physiological temperatures and continuous agitation. The extensive treatment times and higher temperatures may damage the vein surfaces si nce this is where maximum contact occurs. This may further lead to altered mechanical properties or a biol ogical response due to exposure of denatured collagen to blood flow. The collagen degradation assay used in this study measures the effective degradati on process providing an indication of the average damage across the vein wall. Intact collagen will therefore occur throughout the wall and may balance by denatured collagen on the surfaces, masking the possible

PAGE 101

89 degradative effects to the vein surfaces. I propose a more e fficient method to perfuse the venous wall that allows for shorter treatme nt times at physiological temperatures. Numerical models have been used to study the in vivo transport of water and macromolecules into and through arterial wa lls under physiological conditions. These methods are used to gain understa nding with respect to the migr ation of particles, such as lipoproteins, present in atherosclerosis, and in fluencing transmural pressure gradients. Macromolecule transport through healthy and lea ky clefts is believed to lead to intimal hyperplasia. Under normal conditions pre ssure is applied during blood flow, causing volume and particle flux and resulting in e xpansion or stretching of the extracellular matrix. By applying these concepts, the eff ect of applied pressure to the lumen on the transport of a surfactant thr ough the vein wall is investig ated to achieving a more efficient, less destructive process for rem oving cellular material. By pressure-induced stretching and pressure-induced convection, decellularization ag ents may more efficiently penetrate the venous wall to remove endogenous materials. Several factors contribute to water flow in living tissue, including smooth muscle cell metabolism and adsorption [104]. The ti ssue used in this study is nonviable, hence cells will not play a role in flow except as resistors or obstacles to flow as permeability will initially be low. The model used in th is dissertation, therefore, relies on pressure gradients and concentration gradients to define flow conditions. The effect of transmural pressure gradie nts on arterial transport of macromolecules and water has been studied to explain the occurrence of intimal hyperplasia and hypertension [84, 104-107]. Meyer et al. inve stigated the effects of pressure-induced stretch on low-density lipoprotein (LDL) and al bumin uptake in the rabbit aortic wall.

PAGE 102

90 First, they determined the external diameter as a function of applied pressure. They found that the diameter increased nonlinearly with pressure. Reconstruction of their data showed a perfect polynomial fit (R2 = 1) to a power of two (Figure 4-1). A correlation between pressure and macromolecule uptake was also observed. Albumin was shown to follow the same trend as the matrix diste nds with pressure (Figure 4-1) and when constrained to prevent distension the uptake was not influenced by pressure. Albumin showed a uniform increase across the aortic wa ll with an increase in pressure from 70 to 120 mm Hg, but little difference was seen with an increase fr om 120 mm Hg to 160 mm Hg. When a maximum distension is achieve d at which albumin uptake did not change despite a significant increas e in transmural pressure, indicating the importance of distension rather than transmural pressure. The concentration of LDL was much more pronounced in the intima compared to the medi a and levels were lower in comparison to albumin. This was most likely due the larger size of this molecule in comparison to albumin. The concentration of LDL was linea rly dependent on the pressure. Figure 4-2 from Meyer et al. shows an increase in concentration close to the lumen, but the concentration reaches a maximum in the media at 120 mm Hg and is no longer influenced by an increase in pressure [105]. This follows the trend of the distension of the aortic wall as the increase in diam eter above 120 mm Hg is small.

PAGE 103

91 R2 = 1 R2 = 0.9998 R2 = 10 10 20 30 40 50 60 70 050100150200Pressure (mm Hg)Concentration0 1 2 3 4 5 6 7Diameter (mm) LDL x10-3 Alb x10-3 Diameter Figure 4-1. Analysis of the da ta to compare the effects of pressure-induced stretch on the concentrations of macrom olecules, LDL and albumin in the rabbit aorta [105]. Figure 4-2. Average profiles of relative con centrations of LDL (a) and albumin (b) as a function of distance from the lume n, obtained from unwrapped aortas incubated for 30 minutes at 70, 120 and 160 mm Hg. The arrow indicates the medial-adventitial boundary [105].

PAGE 104

92 Further analysis of the e ffects of pressure-induced tr ansport performed done by Meyer et al. by wrapping the aortas with 4 and 5 mm sleeves. Figure 4-3 shows the effects on both LDL concentration and album in concentration across the wall. The unwrapped segments had higher concentrati ons near the lumen for LDL and higher concentrations almost uniformly across the wa ll for albumin. Their conclusion is that pressure-induced stretching of the wall is a ma jor determinant of arterial mass transport. Figure 4-3. Average profiles of relative concen trations of LDL (left) and albumin (right) as a function of distance from the lumen, obtained from unwrapped and wrapped aortas incubated for 30 minutes at 70 mm Hg (a), 120 mm Hg (b), or 160 mm Hg (c) [105]. Pressure-induced vector transport was examined in human saphenous veins in which biologically inert microspheres (100 nm ) were introduced to the vein wall at 100, 200 and 400 mm Hg and compared to 0 mm Hg [108]. More particles were found along

PAGE 105

93 the intima (thickness of 0.1 mm) of vessels perfused at pressu res above 0 mm Hg. Nearly twice the area of microspheres was found along the intima at 100 mm Hg and 400 mm Hg than at 0 mm Hg, although the percent ar eas at the intima were not significantly different between pressures of 100, 200 and 400 mm Hg. The microspheres in the media (thickness of 0.5 mm) were less than 0.1% on average. Ande r et al. hypothesize that at pressures greater than 100 mm Hg, medial tissue components become compacted that the internal elastic lamina (IEL) pores collapse [108]. The pores of the IEL for arteries are typically around 1 to 2 m which can be limiting to large particles [107]. There is evidence that an increase in tr ansmural pressure gradient that induces stretching of the matrix is indirectly related to mass transport. If it is assumed that interstitial fluid and extracellula r matrix is incompressible, this induced-stretch is needed to allow for volume change. There is also a limit to the influence of transport at maximum distension, whereby an increase in pr essure does not increase the fluid flux and may even inhibit flow due to compaction. A device was constructed th at allows constant flow of sodium deoxycholate at a constant pressure to increase the transport of the surfactant into the saphenous ve in wall in order to remove the cellular elements more rapidly than shaking. Preliminary results at pressures of 100 mm Hg and 200 mm Hg for one hour and two hours will be used to obtain th e physical parameters needed to calculate the process needed for complete decellularization. Derivation of Governing Equations The vein wall is divided into five layers: endothelial cell monolayer, intima, internal elastic lamina (IEL), media and adve ntitia (Figure 4-4). Since the endothelial layer acts as an active barrier to transport it is likely to offer more resistance to flow than

PAGE 106

94 bulk tissue [104]. The study mentioned above on rabbit aortic tissue found that removal of the endothelial cells increas ed the macromolecular upta ke at both 70 and 160 mm Hg of applied pressure [105]. The resistance due to this layer will not be a factor in this study since the grafts will be denuded or de-e ndothelialized and the la yer in contact with the luminal fluid will then be the intima. The intima itself provides very little resistance to flow and it has been shown that it provides no support to matrix stress while consolidation occurs at the interior wall at the border between the intima and rigid IEL [104]. When the intima (0.15 m thickness) is combined with the IEL (1.0 m thickness) there is a greater pressure drop ( P/L) than seen in the media. For 70 mm Hg applied pressure a pressure drop of 23 mm Hg wa s calculated over the intima for a 20 mm Hg/ m pressure gradient compared to a pressure gradie nt of 0.520 mm Hg/ m for the media. Therefore, the intimal layer will be treated as a soft, liquid-like layer and will be modeled as one layer with the IEL. The IEL provides a barrier or resistance to the flow between the intima and the media since it is basically an impermeable elastin barrier containing fenestral pores with diameters rangin g from 0.4 to 2.1 m and an area fraction of 0.002 to 0.2 (aorta of rat, sheep and dog) [ 107]. This layer is considered the primary resistance to flow as is evident in the literature above where accumulation of macromolecules was seen in the intima with a large drop in concentration in the media [105]. Although, the size of macromolecules are larg e compared to sodium deoxycholate, which may cause a greater resistance at the IEL pores. The media will be combined with the adventitia and mode led as a porous medium containing smooth muscle cells.

PAGE 107

95 Figure 4-4. Transverse sectional view of a blood vessel wall. The layer in contact with the lumen contains a single layer of endothelial cells. These cells will be removed so they will not contribute to the resistance to flow. Next to this layer is the intima containing microfibr ils of collagen and glycoproteins. The IEL is a layer of elastic fibers which se parates the intima from the media. The media is the next layer constituted principally of smooth muscle cells, collagen fibers, and glycoproteins. Th e adventitia, not shown, is the outer layer containing collagen, elas tin and muscle fibers [107]. There are three categories used to model ar terial transport [106]. The simplest is the wall-free model which treats the laye rs as membranes with suitable boundary conditions. The solution to this model re quires only a few parameters, such as, diffusivity, overall mass transfer coefficient and filtration velocity. This model cannot provide information on concentration profiles. The most complex model that accounts for each layer and their heterogeneity is th e multilayer. This model provides more realistic information on the transport dynamics but requires determination of a large number of parameters for each layer. A compromise is the fluid-wall model which accounts for the venous wall but assumes it is a homogeneous layer. This model will be explored further in this study.

PAGE 108

96 Flow through a porous media is mode led with Brinkman’s equation [107] f p fu K u P (1) 0 u (2) where the first term is the viscous term accounting for the no-slip boundary conditions over the IEL surfaces and the smooth muscle cells and the second term is Darcy’s law which characterizes flow in a porous medium away from the solid boundaries. Equation (2) is the equation of con tinuity (EOC) that expresse s the conservation of matter assuming an incompressible fluid (u ). I will assume the viscous flow is limited to the boundary layers on the smooth muscle cells and that these boundary layers are small and decrease over time as the matrix becomes decellularized. I will also assume that flow is one-dimensional across the layers in the wall and the interstitial fluid and matrix are incompressible. Therefore, volume change sa tisfies the diffusion and reduces the flow in the venous wall to Darcy’s law for filtration across the layers f Pu K x P (3) where Kp is the hydraulic resistance, Kp is Darcy’s permeability and uf is the filtration (interstitial) velocity. Darcy’s permeability is a parameter specifi c to a layer. It was observed with water flux in rabbit thorac ic aorta that the permeability of the intima (KPi) is 100-fold greater than the permeab ility of the media (KPm) [107]. The permeability for the media is actually redefined as an effective parameter (Kpeff) that accounts for the permeability of the matrix and the volume fraction of smooth muscle cells occupying the spaces between

PAGE 109

97 the matrix fibers. Therefore, equation (3) is rearranged to get the velocity profile of each layer as a function of the transmural pressure x P K upi fi (4) x P K upeff f (5) The interstitial velocity is typi cally very low on the order of 10-6 cm/s [107]. From Wang and Tarbell [109] the effectiv e permeability has the form ) ( 305828 0 1 305828 0 16 4 4F O F F F F K Kp peff (6) where F is the volume fraction of smooth muscle cell in the media. A typical value of F for aortic tissue is 0.4, reducing equation (6) to F F K Kp peff 1 1 (7) resulting in an effective permeability value th at is 42% of the permeability of the matrix. In models of arterial wall transport, the e ffective permeability is assumed to be constant. I will be removing the smooth muscle cells fr om the matrix, therefore, the effective permeability will change over time: ) ( 1 ) ( 1 t F t F K Kp peff (8) I have described the flow of fluid thro ugh the walls as a function of transmural pressure and porosity according to Darcy’s m odel. Now, we have to consider the chemical dynamics through the wall whic h are governed by convection-diffusion equations. Mass transport occurs by an external force, such as pressure, that causes fluid motion (convection) and/or movement of a flui d from an area of hi gher concentration to

PAGE 110

98 an area of lower concentration (diffusion). The volume flux and mass flux define the transport and are coupled by the Kedem-Katcha lsky equations. In the case of the fluidwall model (Figure 4-5), these equati ons describe the flux of fluid ( Jv) and the flux of chemicals ( Js) between two solutions with different concentrations and different pressures separated by a semi-permeable memb rane. The flux equations are then ) ( p L Jpeff v (9) c RT (10) v eff eff sJ c c f s c P J ) (2 1 (11) where Lpeff is the hydraulic conductivity (cm/s mm Hg), is the osmotic pressure difference, Peff (cm/s) is the permeability and seff is the sieving coefficient. Remember, the effective parameters depend on the porosit y of each medium. The sieving coefficient is an effect when the endothelial layer is present, so it will be neglected ( seff ~1). When the concentration of a single so lute is high it creates a large osmotic pressure difference and contributes to solven t flow. The concentrations used in this study will be assumed to be low enough to neglect the effects of osmotic pressure.

PAGE 111

99 Figure 4-5. The approximation of the vascular wall structur e with the fluid-wall model where the subscripts i, m and adv indicate the lumen, media and adventitia, respectively. The concentration is ci and the pressure is pi [106]. Equations (9) and (11) are redefined using a modified electric analogy derived by Prosi et al. [106]. Mass transf er is described as a potential flow due to pressure and concentration differences with parameters us ed to describe the resistances across the membrane (IEL) and across the porous layers in the wall (media). Figure 4-6 gives a graphical representation of the electric analogy for volume and mass flux. For the volume flux across a membrane with one solute we have, ) (2 1p p L Jpeff v (12) and for the filtration velocity through a porous layer we have, ) (2 1p p L K n upeff f (13) with L as the wall thickness. Thus, the resistance to flow through the membrane is R = 1/ Lpeff, while the resistance in the porous layer is R = L/Kpeff.

PAGE 112

100 Figure 4-6. Electrical analog of the filtration for the membrane (top-left) and for the wall (top-right). Electrical anal og of the mass transfer processes for the membrane (bottom-left) and for the wall (bottom-right) [106]. The solute dynamics are driven by the c oncentration gradient across a membrane coupled with equation (12) and described below as v m eff sJ c c f c c P J ) ( ) (2 1 2 1 (14) with the average concentration within the membrane, fm, defined by theoretical results by Kedem and Katchalsky as 2 1 2 1ln c c c c fm (15) The chemical filtrating in a porous layer in the direction normal to the layer surface is coupled with equation (13) a nd defined by the mass flux as ) ( ) (2 1 2 1c c nf u c c L D Jw f eff s (16) w f transpnf u c c R 2 1 v transpJ c c R ) / ln(2 1 peffL R 1 peffK L R effD R 1 effD L R

PAGE 113

101 where fw is the average concentration within the wall. The immunohistochemical stain used to determine class I MHC antigens in Ch apter 2 is used to calculate the average concentration in the wall. It is determined digitally by subtracting th e number of particles that stained red from the total number of pa rticles and then dividing the product of the total tissue area and a unit volume with the a ssumption that the con centration only varies in the direction normal to the wall surface. It is also assumed th at the concentration gradient is greatest within a small boundary la yer with the bulk of the unstained tissue region representing the bulk concentration ( c1) of sodium deoxycholate. The resistances to mass flux across a me mbrane and across the porous media are defined as an electrical circuit with two resi stances connected in pa rallel (Figure 4-6) and are defined as ) ( ) (2 1 2 1c c c c nf u L D Jw f eff s (17) ) ( 1 12 1c c R R Jtransp diff s (18) ) (transp diff transp diff totR R R R R (19) Rdiff is the resistance to diffusion and Rtransp is the resistance associated with the transport processes [106]. Materials and Methods Human saphenous vein from a cadaver was sectioned into 50 mm segments and placed on cannulas. The cannulas were attached to a frame that suspended the tissue in an ultrasonic bath containing 0.25% sodium deoxycholate set to 37oC. The inlet cannula, determined by the direction of the vein valv es, was attached to a tube inline with a

PAGE 114

102 pressure gage and a peristaltic pump. The outlet cannula was attached to tubing with an adjustable clamp that drained into a sink. The pump pulled sodium deoxycholate from a heated bell jar through the vein that was submerged in an ultrasonic bath. Pressure was monitored with the gauge and adjusted by eith er changing the flow ra te or adjusting the clamp at the end. The pressure was puls atile with a range of 10 mm Hg around the target applied pressure. There were three treatment groups with tw o veins per group. The first treatment group (T1) maintained an averag e pressure of 100 mm Hg at 37oC for one hour. The second treatment group (T2) was pressurized at 100 mm Hg for two hours. The pressure was increased to 200 mm Hg and maintained for two hours at 37oC for the third group (T3). A control group was trea ted at the same temperature for two hours with no pressure applied to the lumen (0 mm Hg). Following each treatment, the veins were flushed at the same pressure for 10 min. Each vein was s ectioned into three segments for histology and three segments for immunohistochemical staining for class I MHC antigens. Histology Samples were placed in 10% buffered fo rmalin immediately following the rinse step and remained there for at least 24 hour s before further preparation. Then, the samples were embedded in paraffin and stai ned with hematoxylin and eosin (H&E) to determine. Immunohistochemistry Immunohistochemistry was used to determin e the level of antigen s remaining in the tissue after treatment as a f unction of pressure and time. Only class I MHC antibodies were used for evaluation since class I antig ens are present in a much greater abundance than class II antigens. Samples were prepar ed in frozen blocks and sectioned with a

PAGE 115

103 cryostat. The sections were thawed and then rinsed for blocking and antibody staining. Sections were incubated for 30 minutes in primary antibody (mouse anti-HLA-ABC class I MHC) solution and then for another 30 mi nutes in diluted biotinylated secondary antibody (Vector ABC Elite kit) solution. An enzyme substrate, NovaRed (Vector), was used to stain the antibodies red. The count er stain was hematoxylin providing a blue contrast. A full protocol can be found in the appendix. Optical Evaluation A Zeiss Axiophot 2 microscope with a moto rized Ludl scanning stage was used for file field mapping at 5x to capture the entire image in one frame. Particle analysis was performed in Image J to determine the ratio of antigen staining (red) to the total tissue sample. Results Cells were still visible on the luminal surfaces as well as within the matrix for all groups. There was a reduction in the amount of red stained tissue in all treatment groups compared to the control group. An increas e in time from one hour to two hours at 100 mm Hg or an increase in pressure to 200 mm Hg did not significantly reduce the mean amount of antigens detected.

PAGE 116

104 Figure 4-7. Histology of a vein pressuri zed at 100 mm Hg with 0.25% sodium deoxycholate. Pressure-induced Decellularization0.00 5.00 10.00 15.00 20.00 ControlT1T2T3 Treatment Group% Red Figure 4-8. The percentage of tissue stained red after tr eating at 0, 100 and 200 mm Hg for one to two hours at 37oC.

PAGE 117

105 CHAPTER 5 SUMMARY AND CONCLUSIONS Efforts have been made to improve the pa tency rates of saphenous vein allografts for vascular reconstructions. Cryopreservation was developed to preserve the endothelial cells on the lumen to create a nonthromboge nic surface. Current protocols have not shown improved patency rates, but instead have exhibited an immune reaction similar to graft rejection that is believe d to cause intimal hyperplasia, occluding the graft as well as causing degradation of the matrix. Additiona lly, a sterilization me thod cannot be used since processing methods would c ounter the efforts made to pr eserve cells. Therefore, antimicrobial solutions are used to reduce th e possible contamination load on the tissue. These methods have been shown to be ineff ective in inactivating th e organisms suggested by the Centers for Disease Control (CDC) for a validated process [42, 44]. Investigators have addressed the patency issue by devel oping methods of decellu larization to remove the antigen laden endothelial cells and smooth muscle cells. The objective of my work was to determine the methods to create a biologically inert scaffold from human saphenous vein that has reduced antigen levels and is completely devoid of any possible contaminations. This was achieved by chemi cal decellularization a nd sterilization under controlled temperatur es and pressures.

PAGE 118

106 Donor screening and extensive serologi cal testing provides a high level of assurance to the recipi ent. Although, problems come with human errors, false-negative tests, and testing within the window of contraction. Furthe r complications come with the incompleteness of medical records and sec ond hand information on the donor’s social history from a relative or friend. Sterilizati on of vascular allografts add an additional level of protection to ensure that if there is an error the tissue will be inactivated of all possible contaminants. I used Bacillus stearothermophilus an organism that is thermostable and shown to be twice as resi stant to hydrogen peroxide and peracetic acid as organisms suggested by th e CDC [103]. A load of 106 was used to provide a factor of safety in the process since it is greater th an what would typically be seen in tissue contaminants. Sterilization of vascular allografts is not extensively covered in the literature, but other soft tissues, such as tendons, provide d some basis for choosing hydrogen peroxide and peracetic acid. To evaluate the effec tiveness of each oxidant at various conditions with survivor curves, a plate counting method had to be developed. Peroxide and PAA were plated at various dilutions with a dilution of organisms to produce about 100 CFU/ml on either tryptic soy agar (TSA) pour plates or sheep’s blood agar (BA) pour plates. PAA showed no inhibition in growth whereas peroxide showed inhibition at 6%. The peroxide was most likely affected by the catalase enzyme in the blood while PAA has been shown to inactivate most enzyme activ ities [97]. Since vascular tissue contains blood with catalase, I anticipated there may be a reduction in the inactivation rates of hydrogen peroxide. At the same time, blood co ntains iron that is known to significantly increase the effectiveness of hydrogen peroxide (Harbor-W eiss/Fenton) by the production

PAGE 119

107 of hydroxyl radicals. I discovere d that the inactivation rate significantly increased in the presence of tissue compared to in suspensi on with a significant reduction in the D-value at 40oC from 47 min/log to 15 min/log. Unfo rtunately, the amount of blood present in donor tissue may vary significantly giving rise to inconsistent results using hydrogen peroxide. If the tissue is fi rst treated to remove a signi ficant amount of blood and then iron is introduced in a controlled amount, this could increase the effectiveness in a consistent manner. Additionally, hydrogen peroxide has highly diffusive properties and permeates membranes at a rate comparable to that of diffusional wa ter flow [110]. The implication is that it could be more effective in reaching the deep re gions of the tissue if it is not inactivated along the way. The adventitia c ontains small blood vessels called vasa vasorum that may contain contaminan ts endogenous to the donor, highlighting the need for full penetration. An increase in temperature greatly in fluences the effectiveness of hydrogen peroxide to inactivate B. stearothermophilus with an 80% reduction in the D-value with a 10oC increase in temperature. Compared to the literature where Mazzola tested 26.5% peroxide on the same spore solution at room te mperature, the D-value calculated in this study with only 6% at an increased temperature of 40oC was similar. Data not presented in this study showed that pe roxide concentrations over 20% were damaging to tissue. Therefore, by increasing the temperature a lo wer concentration was effective and did not damage the matrix. Even with this increase in effectiveness, I found that peracetic acid was highly reactive with greater inactivati on rates at low temperatures and low concentrations while maintaining the integrity of the matrix. PAA was more effective at these conditions than

PAGE 120

108 hydrogen peroxide at all condi tions. The process for ster ilization of saphenous vein allografts was determined in this study to be treatment in 0.05% PAA at 40oC for 12 min. Treatment under these conditions for one hour were not damaging to the matrix so the functional integrity of the graft is maintained. Typically, contamination of vascular allografts occurs during graf t recovery and preparation and is located on the surface. Disease from a donor may be contained deep within the matrix in the vasa vasorum, therefore, it may require a longer period of time to achieve a 6 log reduction since the PAA has to traverse the matrix. In this st udy the bacteria was loaded into the lumen and dried to the surface for 15 minutes. Drying the spiked vein for longer periods of time or pressurizing the spiked lumen with air could cause the bacteria to penetrate further into the matrix. This would better represent a donor infection, thus a highe r level of sterility assurance. If this is the case, it could redefine how vascular tissue processes are validated. This highlights another need fo r pressurizing the lumen for more efficient penetration. Oxidants are more damaging to collagen at increased temperatures and increased times. Pressurizing the lumen would allow temperatures to be increased with shorter treatment times needed to achieve a 6 log reduction throughout the matrix. Additionally, hydrogen peroxide may be added to the process to reach deep within the matrix since it is much more diffusive than PAA. Cryopreserved allografts have been show n to only have 50 to 70% cell viability [51, 53, 54]. If the endothelia l lining is injured or subconf luent thrombosis and intimal hyperplasia occur. This also occurs as a re sult of graft rejection in response to the antigens on the endothelial cells and smooth muscle cells. Ther efore, an incomplete layer combined with antigen laden cells leads to early graft failure making cryopreservation

PAGE 121

109 unfavorable. Simply denuding the lumen is not sufficient since the smooth muscle cells deep within the tissue also contain antigens. This leads to time consuming treatment at high concentrations that damage the matrix. The decellularization methods used in the literature to remove antigens from vascular tissue require multiday processes and are only evaluated qualitatively. The chemicals used in the literature and used in this study for de cellularization included detergents and enzymes su ch as Trypsin, Triton-X 100 and sodium deoxycholate (cholate). All of the protocols reviewed required a minimum of 24 hours to achieve significant cell removal, with additional time for rinses and other related treatments. There is conflicting evidence of the effici ency of each agent to decellularize while maintaining the matrix integrity For instance, Teebken et al. found complete removal of the aortic wall of a pig with 0.1% Trypsin at 37oC for 24 hours while Grauss et al. found no change in cellularity in rat valve grafts with a ten-fold Trypsin concentration [2, 68]. It would seem that with a hi gher concentration and with the thinner matrix of the rat tissue the valve grafts would be acellular. The results in the literature are variable; somewhat resulting from the qualitative hi stological assessment that is a subjective analysis. As shown in Chapter 2 after a tr ypsin treatment, a graft that appears to be significantly reduced in cellular material acco rding to H&E may still contain a significant amount of antigens. Therefore, I used imm unohistochemistry to evaluate each treatment regime. My study only investigated up to 24 hours where I found that increasing the temperature to 48oC achieved almost complete removal for 1.0% Trypsin. Lower temperatures and lower concentrations show ed poor results in antigen removal. The

PAGE 122

110 effect of EDTA added to trypsin was inconsistent. The assessment of the matrix integrity showed less than half the amount of degrad ed collagen was present compared to the control sample, so it was inconclusive. The method used to determine matrix integrity is an enzyme digestion of denatured collagen strands by trypsin. Trypsin does not digest stable collagen, but the processing effects on collagen could not be assessed with this assay. The amount of denatured collagen co uld be determined by an analysis of the effluent following the decellularization treatme nt. Some researchers have determined qualitatively that the matrix was preserve d following treatment in Trypsin [7, 69, 71] while others have seen severe damage at physiological temperat ures for the extended periods of time needed for complete cell rem oval [71]. These analyses were performed qualitatively resulting in subjective assessments. Triton-X 100 has also shown variable result s in the efficiency of cell removal as well as extensive rinsing tim es required to remove residual chemicals, although all investigators agree that there is little damage to the extracellu lar matrix [68, 77, 80]. My data shows that complete cell removal is achieved with Triton at high temperatures within 24 hours. Additionally, the collagen de gradation assay showed that Triton did not affect the matrix integrity. On the other hand, cholate was the most efficient in cell removal with 100% removal at high temperat ures and long processing times and almost 100% removal at physiological conditions in only 6 hours. Higher temperatures can achieve complete cell removal, but there is a slight adverse affect on the collagen fibrils. Truly, the only was to assess the antigenicity of the graft and its pate ncy is to perform an animal study. The appropriate model is being investigated with sheep or dog as possibilities.

PAGE 123

111 In Chapter 4, I began to evaluate the c oncept of pressure-induced distension of the venous wall to allow for a volume exchange of fluid in the lumen and the interstitial fluid. Assuming the matrix is incompre ssible, the vein would expand and the pore volume would increase with increasing luminal pressure, getting the decellularizing agent deep into the tissue more efficiently. The goa l is to reduce the time needed for complete cell removal while protecting the vein surf ace and components close to the lumen from excessive contact times that may cause damage. The collagen degradation assay may only give an effective degradation result. The surfaces of the vein have the maximum contact time at the specified conditions while within the matrix there may be a reduced response due to the transport time. Therefore, the degradation on the surface may be masked by the better preserved integrity of the tissue within the wall. This is of importance because if there is recruitmen t of macrophages postimplantation, matrix metalloproteinases (MMPs) will be secreted and destroy the collagen and weaken the graft. A study using trypsin to remove any inherent or additional denatured collagen will be conducted to see if there is further re duction in the effects of an immunological reaction. Vein segments were treated in 0.25% chol ate under increased luminal pressure to determine the influence of convective transport from an increased volume flux. Cholate was chosen instead of Triton since it is more effective in shorter time periods, so it may be easier to detect a difference in only one and two hours. Additionally, since it causes more damage it could be determined if the e ffective degradation of the matrix could be reduced by this method. Due to a lack in tis sue availability the matrix degradation could not be assessed. Preliminary results for the decellularization effici ency showed almost

PAGE 124

112 50% less red stained tissue in samples trea ted at increased pressures. There was no significant difference in 100 mm Hg and 200 mm Hg, nor wa s there a difference when the treatment was increased by one hour. Hist ological analysis showed cells were still present on the endothelium. This could have hindered the results since this barrier produces a large resistance to flow. Further studies need to be done with an initial removal of the endothelial cells. Also, the sa mple size was small due to unavailability of tissue. Calculation of the physical paramete rs, such as permeability was unsuccessful. This was likely due to the fragmentation of the samples on the slides which sometimes made it difficult to measure regions. A study could be performed staining specifically for endothelial cells with Factor VIII antibodies and specifica lly for smooth muscle cells with -actin to determine the regions and the concentration gradient across each layer. It was shown that through increased temp eratures and concentrations that the killing potential of oxidants was improved. For decellularization, increasing the temperature had a greater affect on the effici ency of the detergents than increasing the concentration. This is likely due to the improved diffusive properties from the increased energy in the molecules. Increasing the lu minal pressure also improved the diffusive properties of cholate into the vein wall. A comparison of detergents and processing conditions for decellularization was improve d by the use of a quantitative method for measuring the concentration of red stained antigens in the vein wall. Additionally, a quantitative measure of the integrity of co llagen provided an object ive analysis of the preservation of the extracellu lar matrix. The results from this study provide a process for sterilization and decellulariza tion of saphenous vein allografts. Treatment in 0.05% peracetic acid for 12 min at 40oC will achieve a 6 log reduc tion in highly resistant

PAGE 125

113 Bacillus stearothermophilus Followed by treatment at 40oC for 6 hours in 0.25% sodium deoxycholate, the saphenous vein allogr aft will be completely decellularized with the matrix integrity well preserved. Study Limitations The main limitation to this study was tissue availability. As mentioned before, the supply of vascular allograft is limited a nd unable to meet the demands. The tissue I received was rejected tissue that had to have research consent from the donor’s next of kin. This further limited the availability of samples for my studies. There were several methods I used to compensate for this limita tion. I resorted to using cryopreserved saphenous vein allografts that were rejected due to a freezer failure. I only used this tissue for sterilization studies and reserved the fresh tissue for the decellularization studies. The importance of this assignment was that if any cell viability was lost during the graft preparation and cryopres ervation or as a result of th e sudden drop in temperature due to the freezer failure, it would result in a loss in cell attachment. The cells would be easier to remove and not repres ent the effects of treating fresh tissue. I received fresh tissue sporadically, so I had to prioritize my studies to produce results where I expected large detectable differences. Another limitation was small sample sizes that produced low power analyses. For example, in chapter 4 the effects of pressure-induced stretch was examined at 100 mm Hg for one hour and two hours while 200 mm Hg was only investigated for two hours. Future Studies Cholate was the most efficient in cell removal but Triton had no effect on the extracellular matrix. There was not a signifi cant difference in the concentration of red stained tissue between Triton al one and combined with cholate. At the same time, an

PAGE 126

114 analysis was not done on the collagen degradation when they were combined. A study could be conducted to examine the appropriate concentration ratio of Triton and cholate that would allow an increase in temperatur e and a reduction in processing time with 100% removal while preserving the matrix. Also alcohol has been shown to stiffen vein grafts [2]. Pretreatment with isopropyl al cohol may produce crosslinks that protect the tissue from degradation, a llowing increased processing temperatures and reduced processing times. A study looking at this effect would also need to determine if these crosslinks would prevent the agents from penetrating the matrix. An important study is to combine the decellu larization process with the sterilization process to investigate a synergistic effect that produces both an acellular and sterile graft. The first step would be a delipidating and de -endothelialization st ep using Triton with EDTA. Triton would preserve the matrix to allow further pro cessing and EDTA was shown to improve cell removal when added to Triton. Removing the blood and lipids initially will improve the efficiency of the oxi dants. The next step would be treatment with peracetic acid to inactivate the majority any possi ble contaminants. Since hydrogen peroxide is more diffusive, it w ould be used next to get to th e deep regions of the matrix. Since PAA inactivates a majority of the bi ological material, pe roxide would not be inactivated while traversing the matrix. Th e decellularization step would follow with cholate or Triton and cholate. The inac tivation by the oxidant s would improve the efficiency of the decellularization process by reducing or eliminating cell viability. Successive rinses in phosphate buffered saline (PBS) would be the last steps to remove any residual chemicals. All steps would be performed under at least 100 mm Hg while sonicated.

PAGE 127

115 Collagen degradation gives some indication of matrix integrity but further studies are needed to compare the gross mechanical properties following treatment. The compliance of the graft is measured as the slope of the linear regi on of a plot of the change in diameter over the change in applied pressure. An increase in compliance would indicate a degradation of the matrix and altered flow dynamics once implanted. The burst test gives the burst strength which is a measure of the maximum pressure the vein can withstand before an aneurysm forms and bursts and the vessel leaks. A reduction in this value also indicates damage within the matrix, but more importantly determines if the vein can withstand physiological pressures once implanted. The vein undergoes an arterialization pr ocess once implanted. In order to bypass this process and have faster incorporation, the sterilization and d ecellularization would be applied to arterial vascular allografts. Additionally, th e process can be applied to xenograft conduits and then analyzed for its a ffect on alpha (1,3-di)-galactosyl present in animal tissue but not human tissue. Fu rther improvement in incorporation and remodeling could result from cell seeding or perfusing the tissue with the same pressureinduced stretch method with growth factors or pharmacological agents. This could be followed up with a mild crosslinking treatment to simulate a controlled release and exposure of these agents to the cells of the recipient.

PAGE 128

116 APPENDIX A DONOR INCLUSION LIST Greater Saphenous Vein: Age criteria: MALE: 15 y/o 65 y/o FEMALE: 15 y/o 39 y/o Time constraints: Flush time shall be time when saphenous vein is initially perfused Papaverine is mixed 1 ml (30 mg/ml) for every 250 ml of sterile isotonic solution [0.12 mg/cc] Saphenous vein recovery times Must be flushed prior to 12 hours post asystole or last seen/known alive time Must wait minimum 15 minutes post infusion before commencement of dissection Rule-out saphenous veins if infusion can not meet 12 hour flushing time limit Recovery times same as for heart (refer to above) Perfuse using a 60 cc syringe only

PAGE 129

117 APPENDIX B MAJOR HISTOCOMPATIBILITY COMP LEX (HLA CLASS I AND CLASS II) IMMUNOHISTOCHEMISTRY PROTOCOL 1. Remove slides from freezer. Allow sections to come to room temperature. 2. Rinse slides for 5 minutes in dH2O. 3. Blocking. Submerse slides in 3% H2O2 for 5 minutes. 4. Primary antibody prep. Prepare a 1:100 dilution of primary antibodies in antibody diluent (Note: primary antibody solutions need to be prepared 30 minutes before use) Sigma mouse anti-HLA-ABC (Class I) Dako mouse anti-human HLA DP,DQ,DR (Class II) 5. Wash slides for 5 minutes in PBS. (Note: must use fresh PBS for each wash.) 6. Blocking serum prep. Prepare diluted normal blocking serum (Vector ABC Elite kit): add 3 drops of stock normal serum to 10 ml of PBS in mixing bottle 7. Blotting. Carefully blot slides dry without touching the tissue. 8. Pap pen. Quickly (so tissue doesn’t dry out ) use a pap pen to draw a circle around the tissue, and place slides in an incubation chamber. 9. Blocking. Incubate sections for 20 minutes with diluted normal blocking serum. (Note: keep incubation chamber closed during all in cubation periods to keep tissue from drying out.) 10. Blotting. Carefully blot excess serum from slides without touching tissue. 11. Primary antibody. Incubate sections for 30 minutes in primary antibody solutions (use separate slides for Class I and Class II antibodies).

PAGE 130

118 12. Wash slides for 5 minutes in PBS. (Note: designate Coplin jars for each primary antibody used.) 13. Secondary antibody prep. Prepare diluted bioti nylated secondary antibody solution (Vector ABC Elite kit): add 3 drops of normal blocking serum stock to 10 ml of PBS in mixing bottle add 1 drop of biotinylated antibody stock 14. Secondary antibody. Incubate sections for 30 minutes in diluted biotinylated secondary antibody solution. 15. ABC Reagent prep. Prepare VECTASTAIN Elite ABC Reagent: add exactly 2 drops of Reagent A to 5 ml of PBS in the mixing bottle add exactly 2 drops of Reagent B to the same mixing bottle, mix immediately, and allow to stand for 30 minutes before use. 16. Wash slides for 5 minutes in PBS. 17. ABC Reagent. Incubate sections for 30 mi nutes in VECTASTAIN Elite ABC Reagent solution. 18. Wash slides for 5 minutes in PBS. 19. Enzyme Substrate prep. Prepare NovaRed (Vector) substrate: add 3 drops of Reagent 1 to 5 ml of distilled water, and mix well add 2 drops of Reagent 2 and mix well add 2 drops of Reagent 3 and mix well add 2 drops of Hydrogen Peroxi de solution and mix well. (Note: must use this solution immediatel y. This substrate could be carcinogenic so caution should be taken.) 20. Enzyme Substrate. Incubate sections in N ovaRed for 40 seconds. Dump excess stain onto a towel and place slide in tap water. Stain slides a few at a time. (Note: bleach items that come in contact with substrate, including microscope.) 21. Counterstain sections in hematoxylin for 25-30 seconds (for Hematoxylin 1) or for 15 seconds (in Hematoxylin 2) (dip a few times). 22. Rinse slides for 1 minute in running tap water. 23. Clarifier 2. Dip slides in Clarifier 2 solution for 5 dips. 24. Rinse slides for 1 minute in running tap water. 25. Bluing. Dip slides in Bluing solution for 5 dips.

PAGE 131

119 26. Rinse slides for 1 minute in running tap water. 27. Dehydrate with series of alcohols. 95% alcohol for 10 dips 100% alcohol for 10 to 15 dips x 3 alcohols 28. Clear with series of xylene. Dip slides in xylene for 1 minute x 3 xylenes 29. Coverslip. Use xylene mounting agent to coverslip.

PAGE 132

120 APPENDIX C ANTIGEN LEVELS FOR TRYPSIN (A) (B) (C) (D)

PAGE 133

121 (E) (F) (G) (H) Figure C-1. Vein segments treated w ith trypsin. (A) 0.05% trypsin at 37oC for 6 hours, (B) 0.05% trypsin at 37oC for 24 hours, (C) 0.05% trypsin at 48oC for 6 hours, (D) 0.05% trypsin at 48oC for 24 hours, (E) 1.0% trypsin at 37oC for 6 hours, (F) 1.0% trypsin at 37oC for 24 hours, (G) 1.0% trypsin at 48oC for 6 hours, and (H) 1.0% trypsin at 48oC for 24 hours.

PAGE 134

122 APPENDIX D ANTIGEN LEVELS FOR TRITON-X 100 (A) (B) (B) (D)

PAGE 135

123 (E) (F) (G) (H) Figure D-1. Vein segments treated w ith Triton-X 100. (A) 0.25% Triton at 37oC for 6 hours, (B) 0.25% Triton at 37oC for 24 hours, (C) 0.25% Triton at 48oC for 6 hours, (D) 0.25% Triton at 48oC for 24 hours, (E) 1.0% Triton at 37oC for 6 hours, (F) 1.0% Triton at 37oC for 24 hours, (G) 1.0% Triton at 48oC for 6 hours, and (H) 1.0% Triton at 48oC for 24 hours.

PAGE 136

124 APPENDIX E ANTIGEN LEVELS OF SODIUM DEOXYCHOLATE (A) (B) (B) (D)

PAGE 137

125 (E) (F) (G) (H) Figure E-1. Vein segments treated with sodium deoxycholate. (A) 0.25% Cholate at 37oC for 6 hours, (B) 0.25% Cholate at 37oC for 24 hours, (C) 0.25% Cholate at 48oC for 6 hours, (D) 0.25% Cholate at 48oC for 24 hours, (E) 1.0% Cholate at 37oC for 6 hours, (F) 1.0% Cholate at 37oC for 24 hours, (G) 1.0% Cholate at 48oC for 6 hours, and (H) 1.0% Cholate at 48oC for 24 hours.

PAGE 138

126 REFERENCE LIST 1. Conte, M S, The ideal small arterial substitute: a search for the Holy Grail? Faseb J, 1998. 12(1): p. 43-5. 2. Conklin, B S, Richter, E R, Kreutziger, K L, Zhong, D S, Chen, C, Development and evaluation of a novel dece llularized vascular xenograft. Med Eng Phys, 2002. 24(3): p. 173-83. 3. DeMasi, R J, Snyder, S O, The current status of prosth etic-vein composite grafts for lower extremity revascularization. Surg Clin North Am, 1995. 75(4): p. 74152. 4. Stillman, R M, I nfrainguinal occlusive disease eMedicine, May, 1995. http://www.emedicine.com/med/topic2719.htm 5. Prager, M, Holzenbein, T, Aslim, E, Do menig, C, Muhlbacher, F, Kretschmer, G, Fresh arterial homograft transplantation: a novel concept for critical limb ischemia. Eur J Vasc Endovasc Surg, 2002. 24: p. 314-321. 6. Timaran, C, Goldman, M H, Saphenous vein allografts fo r infrainguinal arterial reconstruction: current ro le as vascular conduits. Adv Vasc Surg, 2002. 10: p. 183 197. 7. Bader, A, Steinhoff, G, Strobl, K, Sc hilling, T, Brandes, G, Mertsching, H, Tsikas, D, Froelich, J, Haverich, A, Engineering of human vascular aortic tissue based on a xenogeneic starter matrix. Transplantation, 2000. 70(1): p. 7-14. 8. Centers for Disease Control (CDC), Hepatitis C virus transmission from an antibody_negative organ and tissue d onor United States, 2000 2002. MMWR Weekly, 2003. 52(13): p. 273-276. 9. Food and Drug Administration (FDA), Human Cells, Tissues, and Cellular and Tissue-Based Products 2004. 10. Bastounis, E, Georgopoulos, S, Maltez os, C, Alexiou, D, Chiotopoulos, D, Bramis, J, PTFE-vein composite grafts for critical limb ischaemia: a valuable alternative to all-autogenous in frageniculate reconstructions. Eur J Vasc Endovasc Surg, 1999. 18(2): p. 127-32.

PAGE 139

127 11. Benedetti-Valentini, F, Gossetti, B, Irace, I, Martinelli, O, Gattuso, R, Composite grafts for critical limb ischemia. Cardio Vasc Surg, 1996. 4(3): p. 372-376. 12. DeLaurentis, D, Freidmann, P, Arterial reconstruction about and below the knee: another look. Am J Surg, 1971. 121(April): p. 392 397. 13. Lord, J W, Jr, Sadranagani, B, Bajwa, G, Rossi, G, New technique for construction of coposite Dacron vein grafts for femoro-distal popliteal bypass in the severely ischemic leg. Ann Surg, 1975. 181(5): p. 670-5. 14. Oppat, W F, Pearce, W H, McMill an, J S, McCarthy, W J, Yao, J S, Natural history of composite sequential bypass: ten years' experience. Arch Surg, 1999. 134(7): p. 754-7; discussion 757-8. 15. Lin, P, Bush, R L, Weiss, V J, Lumsden, A B, Vascular access: the utility of cryopreserved vein allograft. Adv Vasc Surg, 2001. 9: p. 109 119. 16. Safer SVG Interventions Medtech Insight. 2001. p. 256-7. 17. Mizell, J, Maglish, B, Matheny, R, Minimally Invasive Direct Coronary Artery Bypass Graft Surgery: An Introduc tion for Critical Care Nurses. Critical Care Nurse, 1997. 17(3): p. 46-55. 18. CryoLife, Vascular Tissue CryoVein 2002. http://www.cryolife.com/products.htm 19. Dorland's medical dictionary 25th ed. 1995, Philadelphia: W B Saunders Company. 20. Linton, R R, Wirthlin, L S, Femoropopliteal composite Dacron and autogenous vein bypass grafts. Arch Surg, 1973. 107: p. 748-753. 21. Centers for Disease Control and Prevention, 1999 Diabetes Surveillance Report 1999, Centers for Disease Control and Prevention, Atlanta, GA. 22. Centers for Disease Control and Prevention, Diabetes: Disabling, Deadly, and on the Rise 2002, Centers for Disease Contro l and Prevention, Atlanta, GA. 23. Clark, D, Enduring and Cost Effective Limb Salvage Hyperbaric Medicine Update, 1998. p. 61-66. 24. Alexander, J J, Wells, K E, Yuhas, J P, Piotrowski, J J, The role of composite sequential bypass in the treatment of mu ltilevel infrainquinal arterial occlusive disease. Am J Surg, 1996. 172: p. 118-122.

PAGE 140

128 25. Castier, Y, Leseche, G, Palombi, T, Petit, M D, Cerceau, O, Early experience with cryopreserved arterial allografts in below-knee revascularization for limb salvage. Am J Surg, 1999. 177(3): p. 197-202. 26. Faries, P L, LoGerfo, F W, Arora, S, Pulling, M C, Rohan, D I, Akbari, C M, Campbell, D R, Gibbons, G W, Pomposelli, F B, Arm vein conduit is superior to composite prosthetic-autogenous grafts in lower extremity revasculariztion. J Vasc Surg, 2000. 31(6): p. 1119-1126. 27. Londrey, G L, Ramsey, D E, Hodgson, K J, Barkmeier, L D, Sumner, D S, Infrapopliteal bypass for severe ischem ia: comparison of autogenous vein, composite, and prosthetic grafts. J Vasc Surg, May 1991. 13(5): p. 631-636. 28. Fung, Y C, Biomechanics: Mechanical pr operties of living tissues Second ed. 1993, New York: Springer-Verlag. 29. Martini, F H, Fundamentals of anatomy and physiology Fifth ed. 2001, Upper Saddle River, New Jersey: Prentice-Hall, Inc. 30. Fung, Y C, Liu S Q, Determination of the mechanical properties of the different layers of blood vessels in vivo. Proc Natl Acad Sci U S A, 1995. 92(6): p. 216973. 31. McClurken, M, Lipson, D, Collagen shrinkage and vessel sealing 2001, Dover: TissueLink Medical, Inc. http://www.tissuelink.com/pdf/Collage n_Shrinkage_Vessel_Sealing_WP_A.pdf 32. DeWeese, J A, Autogenous venous bypass gr afts--still the standard. Cardiovasc Surg, 1994. 2(2): p. 180-6. 33. Gupta, B S, Kasyanov, V A, Biomechanics of human common carotid artery and design of novel hybrid textil e compliant vascular grafts. J Biomed Mater Res, 1997. 34(3): p. 341-349. 34. Gabriel, M, Kostrzewa, A, Sobieska, M, Immune response after cryopreserved aortic allograft replacement for major vascular infection. Transplant Proceedings, 2002. 34: p. 713-714. 35. Patient Information Intro CryoLife. http://www.cryolife.com. 36. Woll, J E, Kasprisin, D, Standards for Tissue Banking American Association of Tissue Banks (AATB), 2001, McLean, VA. 37. Almond, C H, Anido, H, Seaber, A V, Johnson, R E, Mackenzie, J W, Aortic homograft preservation with dimethyl sulfoxide. Arch Surg, 1966. 93(4): p. 656658.

PAGE 141

129 38. Beach, P M, Bowman, F O, Kaiser, G A, Parodi, E, Malm, J R, Aortic valve replacement with frozen irradiated homografts. Circulation, 1972. 45(supp I): p. 29-35. 39. Bolooki, H, Rubinson, R M, Prochazka, J, Jude, J R, A simple method of aortic homograft valve sterilization and preservation. J Thorac Cardiovasc Surg, 1972. 63(2): p. 249-257. 40. Bowman, F O, Malm, J R, Harris, P D, Kaiser, G A, Kovalik, A T, Further evaluation of aortic valve homografts sterilized by electon-beam energy. Circulation, 1969. 39(supp I): p. 57-60. 41. Innes, B J, Gustafson, D P, Morris, E V, Thomson, N B, Jr, Effect of sterilization methods on endothelium of hom ograft and heterograft valves. J Surg Res, 1971. 11(3): p. 111-118. 42. Centers for Disease Control (CDC), Update: allograft-a ssociated bacterial infections United States. MMWR Weekly Report, 2002. 51: p. 207-210. 43. Food and Drug Administratin (FDA), Good Laboratory Practices (GLP) for nonclinical laboratory studies 21 CFR Part 58 p. Supporting Statement. 44. Alexander, K, Cadaveric Cardiovascular Tissue Disinfection Performance Qualification 2004, AppTec: Maretta, GA. 45. Waterworth, P M, Lockey, E, Berry, E M, Pearce, H M, A critical investigation into the antibiotic sterilizati on of heart valve homografts. Thorax, 1974. 29: p. 432-436. 46. Centers for Disease Control (CDC), Septic arthritis following anterior cruciate ligament reconstruction using tendon a llograft Florida and Louisiana, 2000. MMWR Weekly, 2001. 50(48): p. 1081-3. 47. Centers for Disease Control (CDC), Invasive Streptococcus pyogenes after allograft implantation Colorado, 2003. MMWR Weekly, 2003. 52(48): p. 11731176. 49. Reller, L B, Maddoux, G L, Eckman, M R, Pappas, G, Bacterial endocarditis caused by Oerskovia turbata. Ann Intern med, 1975. 83(5): p. 664-666. 50. Silver, M D, Tuffnell, P G, Bigelow, W G, Endocarditis caused by Paecilomyces varioti affecting an aortic valve allograft. J Thorac Cardiovasc Surg, 1971. 61(2): p. 278-281.

PAGE 142

130 51. Pukacki, F, Jankowski, T, Gabriel, M, Oszkinis, G, Krasinski, Z, Zapalski, S, The mechanical properties of fresh and cr yopreserved arterial homografts. Eur J Vasc Endovasc Surg, 2000. 20(1): p. 21-4. 52. Bellon, J M, Gimeno, M J, Pascual, G, Garcia-Honduvilla, N, Dominguez, B, Bujan, J, Arterial damage induced by cryopres ervation is irreversible following organ culture. Eur J Vasc Endovasc Surg, 1999. 17(2): p. 136-43. 53. Faggioli, G L, Gargiulo, M, Giardino, R, Pasquinelli, G, Preda, P, Fini, M, Corbascio, M, Stella, A, D'Addato, M, Ricotta, J J, Long-term cryopreservation of autologous veins in rabbits Cardiovasc Surg, 1994. 2(2): p. 259-65. 54. Galambos, B, Csonge, L, von Versen, R, Olah, A, Tamas, L, Zsoldos, P, Preservation of vein allograft viability during long-term storage Eur Surg Res, 2005. 37(1): p. 60-7. 55. Pascual, G, Garcia-Honduvilla, N, Rodrigu ez, M, Turegano, F, Bujan, J, Bellon, J M, Effect of the thawing proce ss on cryopreserved arteries. Ann Vasc Surg, 2001. 15(6): p. 619-27. 56. Ruddle, A C, George, S, Armitage, W J, MacGowan, A, McCulloch, S, Brookes, S T, Mitchell, D C, Venous allografts prepared from stripped long saphenous vein. Is there a need fo r antibiotic sterilisation? Eur J Vasc Endovasc Surg, 1998. 15(5): p. 444-8. 57. Lin, P H, Brinkman, W T, Terramani, T T, Lumsden, A B, Management of infected hemodialysis access grafts us ing cryopreserved human vein allografts. Am J Surg, 2002. 184: p. 31-36. 58. Mirelli, M, Stella, A, Faggioli, G L, Scolari, M P, Iannelli, S, Freyrie, A, Buscaroli, A, De Santis, L, Rest a, F, Bonomini, V, D'Addato, M, Immune response following fresh arterial hom ograft replacement fo r aortoiliac graft infection. Eur J Vasc Endovasc Surg, 1999. 18: p. 424-429. 59. Silverstein, R L, The Vascular Endothelium in Inflammation: Basic Principles and Clinical Correlates 1999, Philadelphia: Lippincott Williams & Wilkins: p. 207-225. 60. Allaire, E, Guettier, C, Bruneva l, P, Plissonnier, D, Michel, J B, Cell-free arterial grafts: morphologic characteristics of aor tic isografts, allogr afts, and xenografts in rats. J Vasc Surg, 1994. 19(3): p. 446-56. 61. Davies, M G, Dalen, H, Svendsen, E, Hagen, P O, Functional and histological differences in autogenous and a llogenic vein grafts: two different vasculopathies? J Surg Res, 1997. 69(1): p. 14-22.

PAGE 143

131 62. Ketchedjian, A, Jones, A L, Krueger, P, Robinson, E, Crouch, K, Wolfinbarger, L, Jr, Hopkins, R, Recellularization of decellularize d allograft scaffolds in ovine great vessel reconstructions. Ann Thorac Surg, 2005. 79(3): p. 888-96; discussion 896. 63. Wilson, G J, Courtman, D W, Klement, P, Lee, J M, Yeger, H, Acellular matrix: a biomaterials approach for coronary ar tery bypass and heart valve replacement. Ann Thorac Surg, 1995. 60(2 Suppl): p. S353-8. 64. Posner, M P, Makhoul, R G, Altman, M, Kimball, P, Cohen, N, Sobel, M, Dattilo, J, Lee, H M, Early results of infrageniculate arterial reconstruction using cryopreserved homograft saphenous condu it (CADVEIN) and combination lowdose systemic immunosuppression. J Am Coll Surg, 1996. 183(3): p. 208-216. 65. Madden, R, Lipkowitz, G, Benedetto, B, Kurbanov, A, Miller, M, Bow, L, Decellularized cadaver vein allografts us ed for hemodialysis access do not cause allosensitization or prec lude kidney transplantation. Am J Kidney Dis, 2002. 40(6): p. 1240-3. 66. Hawkins, J A, Breinholt, J P, Lambert, L M, Fuller, T C, Profaizer, T, McGough, E C, Shaddy, R E, Class I and class II anti-HLA an tibodies after implantation of cryopreserved allograft mate rial in pediatric patients. J Thorac Cardiovasc Surg, 2000. 119(2): p. 324-330. 67. End-stage kidney disease MedlinePlus. p. Medical dictionary. http://www.nlm.nih.gov/medlineplus/ency/article/000500.htm 68. Hawkins, J A, Hillman, N D, Lambert, L M, Jones, J, Di Russo, G B, Profaizer, T, Fuller, T C, Minich, L L, Williams, R V, Shaddy, R E, Immunogenicity of decellularized cryopreserved allografts in pediatric cardiac surgery: comparison with standard cryopreserved allografts. J Thorac Cardiovasc Surg, 2003. 126(1): p. 247-52; discussion 252-3. 69. Grauss, R W, Hazekamp, M G, van Vliet, S, Gittenberger-de Groot, A C, DeRuiter, M C, Decellularization of rat aortic va lve allografts reduces leaflet destruction and extracellul ar matrix remodeling. J Thorac Cardiovasc Surg, 2003. 126(6): p. 2003-10. 70. Kasimir, M T, Rieder, E, Seebacher, G, Silberhumer, G, Wolner, E, Weigel, G, Simon, P, Comparison of different decellularization procedures of porcine heart valves. Int J Artif Organs, 2003. 26(5): p. 421-7. 71. Hilbert, S L, Boerboom, L E, Livesey, S A, Ferrans, V J, Explant pathology study of decellularized carotid artery vascular grafts. J Biomed Mater Res A, 2004. 69(2): p. 197-204.

PAGE 144

132 72. Teebken, O E, Bader, A, Steinhoff, G, Haverich, A, Tissue engineering of vascular grafts: human cell seeding of decellularised porcine matrix. Eur J Vasc Endovasc Surg, 2000. 19(4): p. 381-6. 73. Schaner, P J, Martin, N D, Tulenko, T N, Shapiro, I M, Tarola, N A, Leichter, R F, Carabasi, R A, Dimuzio, P J, Decellularized vein as a potential scaffold for vascular tissue engineering. J Vasc Surg, 2004. 40(1): p. 146-53. 74. Huynh, T, Abraham, G, Murray, J, Brockbank, K, Hagen, P O, Sullivan, S, Remodeling of an acellular collagen gra ft into a physiolog ically responsive neovessel. Nat Biotechnol, 1999. 17(11): p. 1083-6. 75. Lu, Q, Ganesan, K, Simionescu, D T, Vyavahare, N R, Novel porous aortic elastin and collagen scaffolds for tissue engineering. Biomaterials, 2004. 25(22): p. 5227-37. 76. Uchimura, E, Sawa, Y, Taketani, S, Yamanaka, Y, Hara, M, Matsuda, H, Miyake, J, Novel method of preparing acellular card iovascular grafts by decellularization with poly(ethylene glycol). J Biomed Mater Res A, 2003. 67(3): p. 834-7. 77. Courtman, D W, Errett, B F,Wilson, G J, The role of crosslin king in modification of the immune response elicited against xenogenic vascular acellular matrices. J Biomed Mater Res, 2001. 55(4): p. 576-86. 78. Courtman, D W, Pereira, C A, Kashef V, McComb, D, Lee, J M, Wilson, G, Development of a pericardial acellular matrix biomaterial: biochemical and mechanical effects of cell extraction. J Biomed Mater Res, 1994. 28(6): p. 655-66. 79. Sievers, H H, Stierle, U, Schmidtke, C, Bechtel, M, Decellularized pulmonary homograft (SynerGraft) for reconstruction of the right ventricular outflow tract: first clinical experience. Z Kardiol, 2003. 92(1): p. 53-9. 80. Samouillan, V, Dandurand-Lods, J, Lamu re, A, Maurel, E, Lacabanne, C, Gerosa, G, Venturini, A, Casarotto, D, Gherardini, L, Spina, M, Thermal analysis characterization of aortic tissues for cardiac valve bioprostheses. J Biomed Mater Res, 1999. 46(4): p. 531-8. 81. Rieder, E, Kasimir, M T, Silberhumer, G, Seebacher, G, Wolner, E, Simon, P, Weigel, G, Decellularization protocols of porcine heart valves differ importantly in efficiency of cell remova l and susceptibility of the ma trix to recellularization with human vascular cells. J Thorac Cardiovasc Surg, 2004. 127(2): p. 399-405. 82. DOW Surfactants, Triton X-100 Surfactant: Product Information p. 2. http://www.dow.com

PAGE 145

133 83. Cho, S W, Park, H J, Ryu, J H, Kim, S H, Kim, Y H, Choi, C Y, Lee, M J, Kim, J S, Jang, I S, Kim, D I, Kim, B S, Vascular patches tissue-engineered with autologous bone marrow-derived cells and decellularized tissue matrices. Biomaterials, 2005. 26(14): p. 1915-24. 84. Horowitz, B, Ben-Hur, E., Strategies for viral inactivation. Curr Opin Hematol, 1995. 2(6): p. 484-92. 85. Huang, Q, Dawson, R A, Pegg, D E, Kearney, J N, Macneil, S, Use of peracetic acid to sterilize human donor skin for produc tion of acellular dermal matrices for clinical use. Wound Repair Regen, 2004. 12(3): p. 276-87. 86. Pruss, A, Baumann, B, Seibold, M, Ka o, M, Tintelnot, K, von Versen, R, Radtke, H, Dorner, T, Pauli, G, Gobel, U B, Validation of the sterilization procedure of allogeneic avital bone transplants using peracetic acid-ethanol. Biologicals, 2001. 29(2): p. 59-66. 87. Fideler, B M, Vangsness, C T, Jr, Moore, T, Li, Z, Rasheed, S, Effects of gamma irradiation on the human immunodeficiency virus. A study in frozen human bonepatellar ligament-bone grafts ob tained from infected cadavera. J Bone Joint Surg Am, 1994. 76(7): p. 1032-5. 88. Pruss, A, Kao, M, Gohs, U, Koscielny, J, von Versen, R, Pauli, G, Effect of gamma irradiation on human cortical bone transplants contaminated with enveloped and non-enveloped viruses. Biologicals, 2002. 30(2): p. 125-33. 89. Fideler, B M, Vangsness, C T, Jr, Lu, B, Orlando, C, Moore, T, Gamma irradiation: effects on biomechanical properties of human bone-patellar tendonbone allografts. Am J Sports Med, 1995. 23(5): p. 643-6. 90. Germ Protection Center, Introduction to viruses VNAA (Visiting Nurse Associations of America), 2005. http://www.vnaa.org/vnaa/gen/Germ_Pr otection_Center___INTRODUCTION_T O_VIRUSES,html~germ_protectio n_center_viruses.html.aspx 91. Dox, I G, Melloni, B J, Eisner, G M, The HarperCollins Illustrated Medical Dictionary 1993, New York: HarperCollins Publisher, Inc. 92. Mitra, G, Dobkin, M B, Louie, R E, Mozen, M M, Inactivation of viruses in therapeutic products derived from human plasma. Am J Med, 1988. 84(6A): p. 87-90. 93. Horowitz, B, Bonomo, R, Prince, A M, Chin, S N, Brotman, B, Shulman, R W, Solvent/detergent-treated plas ma: a virus-inactivated subs titute for fresh frozen plasma. Blood, 1992. 79(3): p. 826-31.

PAGE 146

134 94. Block, S S, Disenfection, Sterilization, and Preservation Fifth ed. 2001, Philadelphia: Lippincott Williams & Wilkins. 95. van Bueren, J, Larkin, D P, Simpson, R A, Inactivation of human immunodeficiency virus type 1 by alcohols. J Hosp Infect, 1994. 28(2): p. 137-48. 96. Acosta-Gio, A E, Rueda-Pati no, J L, Sanchez-Perez, L, Sporicidal activity in liquid chemical products to sterilize or high-level di sinfect medical and dental instruments. Am J Infect Control, 2005. 33(5): p. 307-9. 97. Hodde, J, Hiles, M, Virus safety of a porcine-deri ved medical device: evaluation of a viral inactivation method. Biotechnol Bioeng, 2002. 79(2): p. 211-6. 98. Koivunen, J, Heinonen-Tanski, H, Inactivation of enteri c microorganisms with chemical disinfectants, UV irradiatio n and combined chemical/UV treatments. Water Res, 2005. 39(8): p. 1519-26. 99. Pruss, A, Gobel, U B, Pauli, G, Kao, M, Seibold, M, Monig, H J, Hansen, A, von Versen, R., Peracetic acid-ethanol treatment of allogeneic av ital bone tissue transplants--a reliable sterilization method. Ann Transplant, 2003. 8(2): p. 34-42. 100. Hodde, J P, Record, R D, Tullius, R S, Badylak, S F, Retention of endothelial cell adherence to porcine-derived extrace llular matrix after disinfection and sterilization. Tissue Eng, 2002. 8(2): p. 225-34. 101. Reinhart, D, Control of odors from construc tion and demolition (C&D) debris landfills 2004, University of Florida: Gainesville. http://www.floridacenter.org/publ ications/Reinhart%200232004-04.pdf 102. Scheffler, S U, Scherler, J, Pr uss, A, von Versen, R, Weiler, A, Biomechanical comparison of human bone-patellar tendon-bone grafts after st erilization with peracetic acid ethanol. Cell Tissue Bank, 2005. 6(2): p. 109-15. 103. United States Department of Health and Human Services, Guidance on the Content and Format of Premarket Notifi cation [510(k)] Submissions for Liquid Chemical Germicides 1996. 104. Mazzola, P G, Penna, T C, Martins, A M, Determination of decimal reduction time (D value) of chemical agents used in hospitals for disinfection purposes. BMC Infect Dis, 2003. 3(1): p. 24. 105. Kenyon, D E, A mathematical model of wate r flux through aortic tissue. Bull Math Biol, 1979. 41(1): p. 79-90.

PAGE 147

135 106. Meyer, G, Merval, R, Tedgui, A, Effects of pressure-induced stretch and convection on low-density lipoprotein and albumin uptake in the rabbit aortic wall. Circ Res, 1996. 79(3): p. 532-40. 107. Prosi, M, Zunino, P, Perktold, K, Quarteroni, A, Mathematical and numerical models for transfer of low-density lipoprot eins through the arterial walls: a new methodology for the model se t up with applications to the study of disturbed lumenal flow. J Biomech, 2005. 38(4): p. 903-17. 108. Tada, S, Tarbell, J M, Fenestral pore size in the in ternal elastic lamina affects transmural flow distribution in the artery wall. Ann Biomed Eng, 2001. 29(6): p. 456-66. 109. Ander, S, MacLennan, M, Bentil, S, Leavitt, B, Chesler, N, Pressure-induced vector transport in human saphenous vein. Ann Biomed Eng, 2005. 33(2): p. 2028. 110. Wang, D M, Tarbell, J M, Modeling interstitial flow in an artery wall allows estimation of wall shear st ress on smooth muscle cells. J Biomech Eng, 1995. 117(3): p. 358-63. 111. Henzler, T, Steudle, E, Transport and metabolic degradation of hydrogen peroxide in Chara corallina: model ca lculations and measurements with the pressure probe suggest transport of H2O2 across water channels. J Exp Bot, 2000. 51(353): p. 2053-66.

PAGE 148

136 BIOGRAPHICAL SKETCH Donna M. K. Squillace graduated from the University of South Carolina in Columbia, SC, with a Bachelor of Science in chemical engineering in May of 1997. She later continued her studies in chemical e ngineering at the University of Florida completing a Master of Science in December of 1999. Concurrently, she worked as a Research Scientist at Regeneration Technologi es, Inc., in Alachua, FL, while completing her doctoral degree at the University of Florida.


Permanent Link: http://ufdc.ufl.edu/UFE0012420/00001

Material Information

Title: Optimized Sterilization and Decellularization of Small-Caliber Vascular Allografts while Preserving Matrix Integrity
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0012420:00001

Permanent Link: http://ufdc.ufl.edu/UFE0012420/00001

Material Information

Title: Optimized Sterilization and Decellularization of Small-Caliber Vascular Allografts while Preserving Matrix Integrity
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0012420:00001


This item has the following downloads:


Full Text











OPTIMIZED STERILIZATION AND DECELLULARIZATION OF SMALL-
CALIBER VASCULAR ALLOGRAFTS WHILE PRESERVING MATRIX
INTEGRITY















By

DONNA M. K. SQUILLACE


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

DONNA M. K. SQUILLACE















ACKNOWLEDGEMENTS

I would like to thank my family for all of their support and patience. My mother

and stepfather, and my father and stepmother have stood behind me and picked me up at

the lowest points. I only wish my father could be here with me as I achieve this goal.

My sisters and brother have shown great support and have been understanding. My

partner, Mindy, has been thoroughly supportive and patient and has seen me through all

of the hard times and made this possible.





















TABLE OF CONTENTS

Page

A C K N O W L E D G E M E N T S ............................................................................................... iii

LIST OF TABLES ................ .................. ................ .... ....... .............. .. vii

L IST O F FIG U R E S .............. .............................. ............. ........... ... ....... viii

ABSTRACT ........ .............. ............. ...... ...................... xi

CHAPTER

1 INTRODUCTION TO SMALL-CALIBER VASCULAR ALLOGRAFTS ...............1

Purpose and Specific A im s of Study ................................................. .....................5
S p ecific A im s ..................................................... ................ .. 6
H y p oth esis 1 ............................................... ....................... 6
H ypothesis 2 ..................................................... ......... .. ......... 7
H y p oth esis 3 .............................................. ....................... 7
Background and Significance.................................. ........... ............ 8
In d ic atio n s .................................................... ................ 8
A natom y and physiology .................................................. .................. 11
B pass graft choices ........................................................ .. ............ 14
A utograft ......................................................................................14
Synthetic .................................................... ........ .. ................... 16
A llograft ........................................... ............................ 17
A llograft Processing .................................... ..... ... .. .. .. .......... ....19
Sterilization and disinfection.................................... ........................ 19
C ryopreservation ........................ .... .............. ...................... .....2 1
A ntig en city ......................................................................... 2 5
Use of immunosuppressant and anticoagulant therapies............................27
D enudation of cell layers......................................... .......... ............... 29
D ecellularization .................................. ................. ..... ....... 29
Expected O utcom es .......................................... .. .. ......... ......... 38

2 DECELLULARIZATION OF VEIN GRAFTS ........................................................ 39









Introduction ........................................................................................................ 39
A cellular A llografts ............................................... .. .... ......... ........ 41
D ecellu larization P rocess......................................... .............................................44
Cell Removal Assessment ........... .............. ................... 44
Biological Assessment ........... ..................... ............... 50
Purpose of This Study .......... ........................................ ...............57
M materials and M methods ....................................................................... ..................57
C ell rem o v al ................................................................5 7
H isto lo g y ....................................................... 5 8
Immunohistochemistry .............. ........................ .. ..... ............... 58
O ptical ev alu ation .......... .................................................. ........ ............ 59
M atrix integrity ............. .......... .... ...... ........ .. ...... .............. 59
S statistics ...................... .. ............. ..................................................... 5 9
R e su lts ............. ......... .. ............. .. .........................................................6 0

3 STERILIZATION ............................................................................ .. ......... ................68

Introduction ............... ..... .. .. ..................................................................68
Disinfection and Sterilization Processes ..................................70
Processing and D eterm nation of Survivors ........................................ .............. 75
M materials and M methods ..............................................77
Reduction Curves (Survivor Curves) for Spores .........................................77
Methods ............... ......... .................. 77
Hydrogen peroxide ..................... ..............................78
Peracetic acid................................... ............... 78
Spike and recovery method ................ .............. ............ 79
Reduction curves .............. ..... ......... ................80
D ata P re sen tatio n ........................................................................................... 8 0
M atrix In teg rity ............................................................................................. 8 0
Results ............. .. .. .... ............ ............. ...............81
Suspension Sterilization ........................................................ 81
T issu e Sterilization ................................84.............................

4 PRESSURE-INDUCED INCREASE IN PORE VOLUME FOR
INTRODUCTION OF DECELLULARIZATION AGENT INTO VEIN
A L L O G R A F T S .................................................................................................... 8 8

Derivation of Governing Equations ............. ....................... .................93
M materials and M methods ...........................................................101
H istology ................................... .. ... .... ...... .. ............102
Im m unohistochem istry .......... ......................... .................. ................. .102
O optical E valuation .................... ........................ ............... ....... ........ 103
Results .......................... ............. ..................... 103

5 SUMMARY AND CONCLUSIONS ............................................................ 105

Study Limitations ............................. ........................................113



v









Future Studies ................................................................. ...... ........ 113

APPENDIX

A D O N O R IN CLU SIO N L IST ......................................................... .....................116

B MAJOR HISTOCOMPATIBILITY COMPLEX (HLA CLASS I AND CLASS
II) IMMUNOHISTOCHEMISTRY PROTOCOL .............................117

C ANTIGEN LEVELS FOR TRYPSIN ............................................ ...............120

D ANTIGEN LEVELS FOR TRITON-X 100 ..................................................122

E ANTIGEN LEVELS FOR SODIUM DEOXYCHOLATE ...............................124

R E F E R E N C E L IS T .............................................................................. .................... 12 6

BIOGRAPHICAL SKETCH .............. ........... ................ 136















LIST OF TABLES

Table p

1-1. Features of typical arteries and veins. ............................................ ............... 14

2-1. Processing parameters to evaluate decellularization of human saphenous vein. ......58















LIST OF FIGURES


Figure p

1-1. The venous anastomosis in an end-to-side fashion for AV access ...........................

1-2. Types of composite sequential grafts. .............................................. ............... 11

1-3. General structure of vessel w alls. ........................................................................12

1-4. Algorithm for determination of what graft to use in a distal bypass .......................16

2-1. Positive (left) and negative (right) control samples........................................60

2-2. Result of the application of a threshold filter that limits the pass of high intensity
b lu es. ............................................................................... 6 1

2-3. The amount of class I MHC remaining in the tissue when compared to a negative
control for trypsin and trypsin plus EDTA...................................................... 62

2-4. The amount of class I MHC remaining in the tissue when compared to a negative
control for Triton-X 100 and sodium deoxycholate.................... ... .............62

2-5. The amount of class I MHC remaining in the tissue when compared to a negative
control for Triton-X 100 and sodium deoxycholate combined..............................63

2-6. Representative samples of tissue treated with Trypsin (left), Triton-X 100
(center) and sodium deoxycholate (right) and stained for class I MHC antigens. ...64

2-7. A comparison of the amount of cellularity (blue) visible in histology compared to
the amount of antigens (red) present in the tissue. The samples were treated
with 0.05% Trypsin at 370C for 6 hours........................................ ............... 65

2-8. A comparison of the amount of cellularity (blue) visible in histology compared to
the amount of antigens (red) present in the tissue. The samples were treated
with 0.05% Trypsin at 370C for 24 hours...................................... ............... 65

2-9. A comparison of the amount of cellularity (blue) visible in histology compared to
the amount of antigens (red) present in the tissue. The samples were treated
with 0.25% Cholate at 370C for 6 hours......................................... .............. 66









2-10. A comparison of the amount of cellularity (blue) visible in histology compared
to the amount of antigens (red) present in the tissue. The samples were treated
w ith 1.0% Cholate at 48 C for 6 hours................................... ....... ............... 66

2-11. Results of the collagen degradation assay after decellularization at 480C for 24
h ou rs. ................................................................................ 6 7

3-1. Bacillus stearothermophilus growth inhibition at different dilutions of hydrogen
peroxide on TSA and blood agar plates. ...................................... ............... 79

3-2. Bacillus stearothermophilus growth inhibition at different dilutions of peracetic
acid on blood agar plates ................................................ .............................. 79

3-3. Survivor curve for 6% peroxide at 500C resulting in a D-value of 9.87. ................82

3-4. Reduction curve for 6% peroxide at 450C resulting in a D-value of 23.0 ...............82

3-5. Reduction curve for 6% peroxide at 400C resulting in a D-value of 47.39...............83

3-6. Linear regression of the log of the D-values at each temperature.............................83

3-7. Reduction curve for 0.05% PAA at 400C resulting in a D-value of 1.90 ................84

3-8. Survivor curve for 0.005% PAA at 400C resulting in a D-value of 6.45 ..................84

3-9. Reduction curve for 6% hydrogen peroxide at 500C resulting in a D-value of
6 .6 3 .......... .............................. ................................................ 8 5

3-10. Reduction curve for 6% peroxide at 400C resulting in a D-value of 15.17............86

3-11. Collagen degradation of tissue treated with 6% hydrogen peroxide and 0.1%
P A A at 5 0 C ....................................................... ................ 8 7

3-12. Collagen degradation of tissue treated with 6% hydrogen peroxide and 0.05%
P A A at 4 0 C ....................................................... ................ 8 7

4-1. Analysis of the data to compare the effects of pressure-induced stretch on the
concentrations of macromolecules, LDL and albumin in the rabbit aorta ..............91

4-2. Average profiles of relative concentrations of LDL and albumin as a function of
distance from the lum en. ............................................... ............................... 91

4-3. Average profiles of relative concentrations of LDL (a) and albumin (b) as a
function of distance from the lumen. ............................................ ............... 92

4-4. Transverse sectional view of a blood vessel wall...................................................95

4-5. The approximation of the vascular wall structure with the fluid-wall model. ..........99









4-6. Electrical analog of the filtration for the membrane (top-left) and for the wall
(top-right). Electrical analog of the mass transfer processes for the membrane
(bottom-left) and for the wall (bottom-right). .................................... ......... ...... 100

4-7. Histology of a vein pressurized at 100 mm Hg with 0.25% sodium deoxycholate. 104

4-8. The percentage of tissue stained red after treating at 0, 100 and 200 mm Hg for
one to tw o hours at 37 C .............................................. ............................. 104

C-1. Vein segments treated with trypsin ........................... ..................................... 121

D-1. Vein segments treated with Triton-X 100. .................................. ............... 123

E-1. Vein segments treated with sodium deoxycholate ............................................125















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

OPTIMIZED STERILIZATION AND DECELLULARIZATION OF SMALL-
CALIBER VASCULAR ALLOGRAFTS WHILE PRESERVING MATRIX
INTEGRITY

By

Donna M. K. Squillace

December 2005

Chair: William L. Ditto
Major Department: Biomedical Engineering

Vascular disease is one of the most frequent diseases worldwide and is a major

public health priority causing coronary and peripheral occlusions leading to either heart

attacks or limb amputations. Therapy involves bypass grafting with autogenous

saphenous vein. More than 30% of patients have insufficient tissue typically requiring

multiple revisions, exhausting usable autogenous vessels. A viable alternative with

similar benefits to the patient's own tissue is a vascular allograft. The disadvantages to

allografts are early stenosis and graft failure related to a graft rejection response.

Currently, endogenous cells are preserved to maintain functionality and prevent

occlusions. These cells contain antigens that elicit an immune response that may lead to

thrombosis, intimal hyperplasia and eventual graft occlusion.

Allograft tissue is processed aseptically, but is not considered sterile. Current

methods of disinfection are inadequate resulting in a significant discard rate due to

positive cultures, hindering the supply which is already limited and not meeting the









demand. Additionally, donor tissue carries a significant risk of disease transmission due

to false negative tests, human errors, and undetectable window periods that cannot be

prevented through aseptic processing.

This study addressed the issues of sterilization and antigenicity through liquid

chemical processing under controlled conditions. Oxidants were investigated for their

effect on reducing B. stearothermophilus, a spore-forming organism, on spiked tissue

while maintaining matrix integrity. Peracetic acid showed rapid inactivation rates at mild

conditions and low concentrations that were favorable for tissue integrity. Hydrogen

peroxide was less effective but has highly diffusive properties, allowing deep penetration

within the tissue. Combining the two oxidants at mild conditions can achieve a

significant reduction in bacteria while maintaining the matrix integrity. The reduction in

antigen laden material was investigated through the use of detergents and enzymes under

controlled conditions. Sodium deoxycholate achieved almost 100% reduction in antigen

laden material but showed adverse effects on the matrix. Triton-X 100 was less efficient

but had no effect on the matrix. Sodium deoxycholate at pressures above 0 mm Hg

showed improved reductions. This study found the optimal treatment for a sterile,

decellularized graft with the combination of chemicals at elevated temperatures and

pressures.




















CHAPTER 1
INTRODUCTION TO SMALL-CALIBER ALLOGRAFTS

Vascular disease is one of the most frequent diseases worldwide and with the

increasing age and longevity of the American population it will remain a major public

health priority [1]. The mainstay of therapy for patients with coronary and peripheral

occlusions is surgical bypass grafting with the patient's own (autograft) saphenous vein

or internal mammary artery. This procedure offers important benefits to the patient but

incurs significant economic cost on the national level. Commonly, insufficient autograft

material is available and a patient may receive cadaver tissue (allograft). Unfortunately,

allograft conduits have poor patency rates resulting in numerous revision surgeries,

amplifying the economic impact. A reduction in graft patency is a result of an

immunological response to cellular material contained within and on the allograft leading

to intimal hyperplasia and eventual graft occlusion. The aim of this study is to develop a

process to create an inert scaffold out of human saphenous vein in order to improve

patency rates. By complete removal of all the endogenous material the tissue becomes

nonantigenic and may allow recipient cell repopulation to occur immediately following

implantation, bypassing the initial graft rejection response typically seen in allograft use.

Allografts are plagued with complications leading to early stenosis and graft failure

thought to be related to an immune reaction similar to graft rejection. Adaptation of vein









allografts to the arterial environment has been studied extensively in animal models and

less in humans [1]. After implantation, the conduits undergo structural changes

characterized by intimal hyperplasia and overall wall thickening. Cellular events that

occur after implantation can affect the patency by occluding the conduit. Occlusion or

stenosis in coronary or peripheral circulation is a common clinical occurrence present in

small-caliber applications that necessitates an additional intervention. Problems that

cause occlusion of a graft are thrombosis and neointimal hyperplasia [2, 3]. Thrombosis

occurs when platelets in circulating blood adhere to certain surfaces, then release

chemicals to attract more platelets to form a large aggregate that generates thrombin [4].

Grafts most commonly fail due to the development of fibrous intimal hyperplasia where

there is an excess proliferation of smooth muscle cells. This flow-restricting lesion may

occur diffusely throughout the graft or, more commonly, at focal sites near anastomoses,

particularly in compliance mismatched synthetic grafts [1].

Current processing methods only treat the tissue without removing a substantial

amount of endogenous material that may harbor contaminants, such as blood and lipids,

as well as antigen containing cells and cellular debris that elicit an immune reaction.

Processing methods and cryopreservation preserve and retain endothelial cells and

smooth muscle cells (SMC), both containing antigens that may require matching blood

types for vascular reconstructions. Patients in need of an allograft must wait for a

matching donor, and the waiting time for a compatible (ABO relevant) graft depends on

the rarity of the recipient's blood type [5].

Rejection plays a significant role in failure and leads to allosensitization [6].

Therefore, a great deal of effort is placed in determining an ABO-compatible donor as









well as lymphocyte cross-matching [5]. The intimal layer of the conduit contains the

antigen-present endothelial cells responsible for ABO relevancy. Stripping this layer of

cells from the conduit removes the need for donor matching, but it is believed by some

that a lack of surviving cells and endothelial integrity may play a role in graft

degeneration and early and late patency [5]. If injured and subconfluent endothelial cells

line the lumen, thrombosis and SMC growth is promoted leading to intimal hyperplasia

[7]. Additionally, a process by which the grafts are decellularized will eliminate the need

for blood-typing, reduce platelet activation, and prevent smooth muscle cell proliferation

that leads to neointimal hyperplasia resulting in thrombosis.

In the case of multiple revisions, insufficient autograft material, or infection due to

synthetic material, a vascular allograft is a medical necessity for life and/or limb saving

operations. Currently, the demand for vascular allografts far exceeds the supply, forcing

surgeons to pursue less favorable alternatives. Allograft tissue is obtained from cadavers

and processed under aseptic conditions so it cannot be considered a truly sterile conduit.

Current processing at an AATB (American Association of Tissue Banks) accredited

tissue bank uses an antimicrobial soak can result in a substantial discard rate (27%) due

to positive first and final cultures from bacterial and fungal contamination resistant to the

cocktail aimed at achieving sterility (personal communication). Additionally, the risk of

transmission of viruses and other pathogens can be reduced but not eliminated through

donor screening. Therefore, vascular transplantation carries a significant risk of disease

transmission to the recipient considering the occurrence of false negative test results,

human errors in screening and processing, and virological testing within the window

period of contraction and detection of diseases. More significantly, over the last several









years multiple examples of donor to host disease transmission have been documented.

These examples include transmission of hepatitis C, bacteria, and fungi, and have

resulted in serious morbidity and even mortality [8]. Unfortunately, traditional methods

of sterilization, such as irradiation and ethylene oxide, have significant deleterious effects

on tissue integrity that limit or preclude their use as allograft.

The marketplace of surgeons acknowledges the inherent risk in dealing with human

tissue and tends to exhaust autograft material instead [3]. Unfortunately, patients who do

not have sufficient or acceptable autografts, or are unable to receive synthetic grafts are

forced to resort to the use of allografts. To that end, the FDA has taken an interest in how

human tissue might be held to a higher standard [9]. Therefore, a sterilization processes

that is effective in inactivating endogenous materials as well as achieving a significant

reduction in all possible organisms without adversely affecting the tissue matrix and

functionality of the vessel would increase the availability of allograft by reducing the

discard rate and provide a safe alternative.

The factors that determine a successful graft include the following; no aneurysms

or dilations, no immunological reaction, no disease transmission, no infection and long-

term patency [10-14]. Once placed in arterial circulation, it must be capable of

withstanding long-term hemodynamic stress without mechanical failure. A failure of this

type could be catastrophic and lead to morbidity, such as loss of limb, or even mortality.

The availability, suturability, and simplicity of handling are desirable for minimizing

operating time, risk and expense as well as long-term durability. Postimplantation, the

graft should be fully biocompatible, resistant to thrombosis and infection and be

completely incorporated by the body to yield a neovessel resembling the native artery in









structure and function [1, 6]. The graft should be porous enough to permit ingrowth of

host's cells, but still durable and suitable to maintain anastomosis integrity.

Purpose and Specific Aims of Study

The purpose of this research is to provide a process that can achieve complete

removal of the cells contained within the vascular tissue matrix and achieve a 6 log

reduction in all possible organisms to reduce immune reaction and ensure the safety of

the patient. Standard validation of sterilization processes introduce a 106 count of

organisms to be inactivated to provide a factor of safety above what is typically

encountered in tissue contamination. Therefore, a 6 log reduction would result in

complete inactivation of the inoculum. Human saphenous vein samples were introduced

to a sequence of chemicals with mechanical stimulation in order to sterilize and

decellularize the tissue. Since it is difficult to predict frequencies and concentrations of

pathogens and infection windows may exceed the time between contraction and the time

of death, the most difficult to inactivate spore will be used to achieve the sterility

assurance. The level of decellularization will be investigated with an

immunohistochemistry stain for antigens located on the cell membranes within the tissue.

The processing conditions will leave a neutralized scaffold with minimal detrimental

biochemical effects on the extracellular matrix, no need for blood typing, and optimal

biological functionality.

Specimens used for this study will be donated human saphenous veins that are

designated by informed consent to be used for research. The primary sources of the

tissue will come from RTI Cardiovascular in Birmingham, AL and Regeneration

Technologies, Inc. (RTI) owned Recovery Agencies. This study will use Greater

Saphenous Veins not accepted for implantation, mostly because the inner diameter is out









of specification (<3mm), but accepted for research. Donor tissue will be enrolled in this

study only if it has consent for research, negative serological tests and medical director

sign off at RTI. Only saphenous veins will be used since it is the most common allograft

conduit currently used and relatively easy to recover since it is a superficial vein.

Specific Aims

Hypothesis 1

Through the use of controlled temperature in combination with measured enzyme

and detergent concentrations, complete decellularization will be achieved with minimal

tissue damage. Additionally, the use of a chelating agent will enhance cell removal.

Chapter 2 provides the results of the investigation of the use of detergents and

enzymes at various conditions in order to achieve a completely decellularized scaffold

while maintaining the matrix integrity. The parameters investigated will be chemical

concentration, temperature and time. Sonication is also investigated as an alternative to

shaking as a means of agitation. The remaining cells and cellular debris will be

investigated with histology and immunohistochemistry. Using an immunohistochemical

staining method for class I major histocompatible (MHC) antigens, the level of

decellularization will be measured as the area fraction of red stained tissue. It will be

shown that antigens still remain even though the tissue appears to be acellular. The effect

on the extracellular matrix will be measured with an enzyme digestion assay that targets

denatured collagen. If denatured collagen is present it may indicate later graft rupture or

dilation or activation of enzymes known to further degrade collagen postimplantation.









Hypothesis 2

By increasing the temperature and increasing the concentration of oxidants used to

inactivate bacteria, with the addition of sonication will decrease the time needed to

achieve a 6 log reduction in a biological indicator.

Since the frequencies and concentrations of various pathogens and polymicrobial

infections on tissue remain unknown, it is difficult to predict what level of sterility is

achievable. Using a biological indicator that is more difficult to inactivate than

contaminants that may be detected on vascular tissue provides a process that can be

applied across all possible microorganisms. In Chapter 3, I chose Bacillus

stearothermophilus which is a spore used for instrument sterilization since it is highly

resistant to temperature and difficult to kill. Suspensions of the spore are treated with

hydrogen peroxide and peracetic acid at various concentrations and temperatures to

provide a baseline for tissue sterilization. Saphenous vein segments were spiked with an

inoculum of spores and treated in similar conditions. The level of kill was measured with

a plate count method where a Petri dish containing a growth media is inoculated and the

surviving spores are counted as colonies.

Hypothesis 3

Distension of venous tissue by pressurizing the lumen will expand the pore volume

and allow more efficient penetration of detergents and reducing the time required to reach

complete decellularization.

Preliminary results from this study are contained in Chapter 4. The significance of

a reduction in contact time is that the surfaces of the vein are in constant contact with the

treatment agent, increasing the chance of damage to these areas while the agent is trying

to diffuse into the matrix to reach the cells deep within. Additionally, a lower









temperature was used to provide further stability to the matrix. By applying pressure to

the lumen, the extracellular matrix expands and allows for a volume flux. Saphenous

vein segments are treated with a detergent at two pressures and for two time points. The

analysis was conducted as was in Chapter 2 with an immunohistochemical stain.

Background and Significance

Indications

Small caliber grafts are used in bypass and revascularization procedures in the

coronary and lower extremity circulation and for arteriovenous access for hemodialysis

and cancer patients. Patients most affected with reconstructions are those with end stage

renal disease (ESRD) requiring vascular access. The prevalence and numbers of patients

that required hemodialysis have risen steadily over the past two decades. The 1999

annual report of US Renal Data System (USRDS) reported more than 30,000 ESRD

patients in the US undergo maintenance hemodialysis annually [15]. It is the most

common vascular operation in the country today with the leading cause of morbidity in

ESRD patients related to vascular access placement and resultant complications.

The surgical procedure and graft preparation for AV access is shown in Figure 1.

The femoral vein at a length less than 25 cm and diameter of 5 7 mm is usually used for

this procedure. The multiple access operations required by each patient are a result of the

poor overall patency rates of hemodialysis access. The increased number of AV graft

operations versus primary AV fistulas is due to the increased age of the patients lacking

superficial venous anatomy [15]. Multiple revisions cause a reduction in available access

sites in the upper extremities, creating a need for alternatives for continuous access.














Grand








Vain '

Figure 1-1. The venous anastomosis in an end-to-side fashion for AV access [15].

A traditional coronary artery bypass graft (CABG) surgery is estimated to cost

$20,000 and within ten years, it is estimated that nearly half of saphenous vein grafts will

require retreatment [16, 17]. In the U.S., over 400,000 CABG procedures are performed

annually with 20% being re-operations, often due to the use of unsuitable saphenous vein

allografts [18]. In 2001, about 80% of bypass grafts were left internal thoracic artery-

LAD plus saphenous vein, 5 to 10% were bilateral internal thoracic arteries plus

saphenous vein, and 1% was all-arterial conduits. Due to the invasiveness of this

procedure, allograft is not popular with cardiac surgeons.

In the lower extremities, critical limb ischemia is a result of femoropopliteal

occlusive disease that obstructs the blood flow into the lower limbs [4, 5, 11, 12, 19, 20].

It affects 10% of the population in the U.S. over 70 and 1 to 2% from 37 to 69 years of

age [4]. About 82,000 people a year have diabetes-related leg and foot amputations

which account for almost half of the non-traumatic amputations in the U.S. [21, 22]. The

average cost for primary amputation is approximately $40,000 with an additional $30,000

for rehabilitation and related medical surgical care [23]. Femoropopliteal disease is

caused by atherosclerosis, which is a form of arteriosclerosis. Plaque from degenerated









thickened arterial intima containing cholesterol lipoid material and lipophages is formed

within the intima and inner media of arteries [19]. This may result in structures that

impair flow and cause acute thrombosis. Plaque that is removed from the wall during

flow, called emboli, may lodge at sites of bifurcation or trifurcation causing occlusion in

the lower limbs. Occlusion occurs in the distal common artery (34%), distal popliteal

artery (14%), or tibial arteries [4]. The vessels eventually become thrombosed due to

turbulent flow [4, 24].

At early stages, occlusion forces blood to perfuse the limb via collateral pathways,

thus compromising the circulation. Blood flow is sufficient at rest, but as activity

increases, the supply of oxygen to leg muscles is inadequate causing intermittent

claudication [4]. As the disease progresses, outflow may be impaired even at rest,

resulting in severe ischemia pain, tissue necrosis (gangrene) or claudication resulting in

the need for intervention [3, 5, 12-14, 20, 24-27]. Lower extremity atherosclerosis is a

marker for systemic atherosclerosis disease with significant incidence of systemic

morbidity. About 30% of those affected will succumb to systemic atherosclerosis within

five years [4].

Approximately 100,000 vascular reconstructive procedures for limb salvage are

performed yearly in the United States [4]. Numerous studies show the effectiveness of

infrainguinal revascularization in producing long-term patency and improved limb

salvage rates [26]. The procedure involves increasing pathways and runoff for inflow

from the common femoral artery to lower limb vessels (Figure 1-2) [12]. The different

procedures involve above knee bypasses from the common femoral to the popliteal

artery, or below knee bypasses from the common femoral artery to either the popliteal or









tibial artery or from the superficial femoral artery to the popliteal artery [10, 12]. The

choice of an above or below knee bypass procedure depends on the extent of the popliteal

artery disease [14]. Autologous saphenous vein is the graft of choice for critical limb

ischemia, but up to 45% of the patients do not possess usable veins [5, 6, 26]. The need

for alternative conduits has become increasingly evident considering the number of

multiple procedures performed to salvage a single extremity [6, 12, 26].

t- II


4i '1 j











Figure 1-2. Types of composite sequential grafts: (A) bypass to the popliteal and one of
the tibial arteries, (B B2) bypasses to tibial arteries [10].

Anatomy and physiology

Blood vessel walls consists of three layers of which the proportions vary according

to the size of the vessel (Figure 1-3) [28, 29]. The inner most layer is known as the

intima or tunica internal and is made up of an endothelial layer and basal lamina. The

endothelial layer lines the lumen with the underlying connective tissue containing

variable amounts of elastic fibers [19, 28, 29]. This layer consists of flat cells that are

metabolically active and produce a number of compounds that affect the vascular lumen

diameter and control activation of platelets [19, 28]. The cells act as a barrier that











controls the entry of substances into the wall, but allows some to pass via transport


system. The compounds secreted control blood coagulation and relaxation and


contraction of the smooth muscle cells in the middle layer [19]. The basal lamina


(basement membrane) is a thin layer of noncellular material with a filamentous texture


underlying the endothelium. The principal component of this connective tissue is


collagen, typically type IV [19, 28]. Separating the intima from the middle layer is a


thick layer of elastic fibers called the internal elastic membrane [29].
















(b,) C.pl .,y
cocnfl o2001 BenBjln CJmnrngs. an Impdnt of AdonWe~ sley LogWn, ian
Valve

Endothellum Endothelium
Tunica
Basement F inlima Basement
membranes a membrane
Inlemal
lamina
Smooth
muscle
Tunica
media Extemal
elastic
lamina
Tunica advenlitia
(a) Vein (b) Muscular artery


Figure 1-3. General structure of vessel walls [29]. The internal elastic lamina is present
in veins but is less prominent as in arteries and sometimes not detectable.

The middle layer is the media or tunica media. It is the thickest layer of the arteries


and is smaller in their vein counterparts. It consists mostly of smooth muscle cells


oriented in concentric sheets within connective tissue. This layer is involved in the









contracting and relaxing of the tissue and contains elastic sheets and bundles of collagen

fibrils, each of which contributes to the mechanical behavior of the vessels [28]. The

amount of each substance affects the ultimate mechanical properties and distinguishes

arteries from veins. The collagen fibrils are made up mostly of type III with some type I

and a trace of type V. The outer layer is divided from the media by the external elastic

membrane which is a thin band of elastic fibers [28, 29].

The outermost layer is the adventitia or tunica externa. It forms a connective tissue

sheath around the vessel connecting it to the body forming the vascular network. It

consists of mainly collagen fibers (similar types and proportions as the media) and

ground substance with scattered bands of elastin fibers [30, 31]. Also contained within

the tissue layer are fibroblasts, macrophages, minute blood vessels vasaa vasorum),

myelinated and nonmyelinated nerves.

The similarities and differences in veins and arteries can be seen in Table 1-1. The

relative wall thickness of veins is lower than in arteries. While the media is the largest

layer in the arteries, the adventitia is the largest layer in the veins with a large amount of

collagen, networks of elastic fibers and bundles of smooth muscle cells. The vein sees

low pressures (5 to 75 mm Hg) compared to the artery (80 to 120 mm Hg), therefore the

media is thin with relatively few muscle cells and very little elastic tissue [28, 29]. The

intimal layer of veins contains valves to prevent backflow since the pressure in the veins

is so low they cannot oppose gravity.









Table 1-1. Features of typical arteries and veins [29].

Feature Typical Artery Typical Vein
General appearance in Usually round, relatively Usually flattened or
sectional view thick wall collapsed with relatively
thin wall
Intima
Endothelium Usually rippled due to Often smooth
vessel constriction

Internal elastic membrane Present Present may be poorly
developed
Media Thick, dominated by Thin, dominated by
smooth muscle and elastic smooth muscle and
fibers collagen fibers

External elastic membrane Present Absent
Adventitia Collagen and elastic fibers Collagen, elastic, smooth
muscle fibers


The muscular and elastic components permit controlled alterations in the diameter

of the vessel as blood pressure or blood volume changes. Arteries have thicker walls

with the media containing more smooth muscle and elastic fibers. The collagen fibers are

crimped while the elastic fibers recoil and restrict the lumen when not opposing pressure

generated by the pumping of the heart to retain its cylindrical shape. The endothelial

lining shows pleats in a sectional view since it cannot contract along with the elastin.

Veins have larger diameters and thinner walls than their corresponding arteries. Due to

the lesser amount of elastin present the walls of the veins collapse when excised.

Bypass graft choices

Autograft

The "gold" standard for vascular repair is the autograft. The saphenous vein is the

most common small-caliber (<5 mm) graft for coronary bypass, arteriovenous access

(AV), and lower-extremity reconstructions. Autograft has continuously demonstrated to









perform better than all nonautogenous grafts. According to DeWeese [32], the advantage

of autograft is its long-term patency even despite poor outflow conditions. It is believed

this may be attributed to the smooth inner lining or pseudointima, which appears almost

immediately in vein autografts as opposed to allografts and, to a lesser degree, synthetic

grafts [12, 24, 32]. Lower extremity, long-term results for infrapopliteal and even pedal

arteries are excellent with patency at 5 years of 50 to 80% [1, 10, 20].

Unfortunately, up to 45% of patients seen with critical limb ischemia do not

possess a usable greater saphenous vein and require postoperative revision surgery in four

to five years [1, 3, 10-12, 20, 24, 26]. Some problems that preclude the use of an all-

autogenous graft are a possible size mismatch between the donor site vessel of the patient

and the vessel to be repaired, insufficient length or diameter, prior use, previous graft

problems such as infection in that site, fibrosis, or venous disease.

As the number of procedures increase, as they do for patients with atherosclerosis

and ESRD, the supply of autograft is greatly reduced. Many patients lacking autograft

with recurrent coronary disease are considered inoperable and those with distal lower

extremity disease may suffer loss of limb [1]. Therefore, graft failure and conduit

availability are important limitations for autograft uses and motivation for alternatives.

Researchers and surgeons work to develop viable alternatives such as prosthetics,

composite prosthetic and autogenous grafts, allografts, and possibly xenografts [3, 10-12,

14, 24]. Figure 1-4 depicts an algorithm for determining the type of graft to use in lower-

extremity revascularization. The surgeon will perform the surgery with the ipsilateral

greater saphenous vein or the contralateral greater saphenous vein. Other autogenous

options are the upper arm veins cephalicc and basilic), the lesser saphenous vein, or a










composite of vein segments [14, 26]. Ectopic (i.e., lesser saphenous and arm veins) or

composite vein grafts are generally inferior to autogenous greater saphenous vein grafts

(40 60 % patency at 5 years) though still superior to synthetic grafts [1]. The point that

should be taken from this is that surgeons will exhaust all possible autograft material

before resorting to synthetics or allograft tissue since the current state of these grafts has

low rates of success.

Distal Bypass Required

Autogcnws Vi in Available?

I, IUVlS OMea SV lfcd Fidd7
21, Cftseolw a 3SV s
3. Af Mlha i
A. cephalJc SVA AMMKVi AiMMI*W

4. LawWSV X

a. KqUtir +t-AVF
b cmigr +-R


Figure 1-4. Algorithm for determination of what graft to use in a distal bypass [3].

Synthetic

These patients without adequate autologous tissues may then be considered for an

all-synthetic or a prosthetic-vein composite replacement, with materials such as

polyurethanes, Dacron (polyethylene tetephthate) or expanded polytetrafluoroethylene

(ePTFE) [2, 33]. Synthetic material has the advantage of being an off-the-shelf product

and is not a limited supply. For large-caliber arterial reconstructions, the long-term

results for replacement of common femoral arteries for either aneurismal or occlusive

disease with available synthetics are generally excellent. However, prosthetic grafts have

proved to be unfavorable as small-caliber arterial substitutes in demanding, low-flow

environments resulting in occlusions [1, 2]. Synthetics have been used with moderate









success in lower extremity bypasses to the popliteal artery and some success to tibial

vessels, but due to its inflexibility the effectiveness of below knee revascularization is

inferior to all-autogenous grafts used in a majority of studies [26, 33]. The four year

patency rate for ePTFE is only 12 to 14% in infrapopliteal and aorta-coronary bypasses,

and they are reconstructed twice as often as autologous for arteriovenous (AV) access

[15, 34]. Complications, such as infection and occlusion, lead to the need of graft

removal, intravenous antibiotic treatment, debridement and drainage, placement of a

catheter or an AV graft replacement. Due to multiple revisions in hemodialysis patients

for AV access, available sites in the upper extremities become scarce.

Some synthetics show good biocompatibility, strength and deformation but

eventually creep and lead to the development of an aneurysm [33]. The most common

failure modes for synthetic grafts are thrombosis due to the adhesive nature of cells to the

synthetic surface and development of focal anastomotic strictures because of neointimal

hyperplasia resulting from compliance mismatch and resistance to incorporation, and

poor outflow [2, 3, 15, 20]. In general, the grafts never completely heal and have limited

re-endothelialization mostly in the area of pannus ingrowth adjacent to anastomoses [1].

Additionally, infection is a common problem with synthetics which causes sensitization

and precludes the patient from revisions with synthetics. A major clinical problem with

infection is it leads to high mortality rates, excessive bleeding, aortointestinal fistulas,

and major amputations of limbs [34].

Allograft

Allograft is tissue taken from a donor of the same species. Vascular allografts are

often a medical necessity for patients who are either needing revision surgery, do not

have sufficient saphenous autografts, or are unable to receive synthetic grafts due to









sensitization or risk of renal failure. In the early days of vascular surgery, allografts were

used but were soon abandoned for poor long-term success due to the high rate of

aneurysm formation and dissections of vessels with secondary dilatation and calcification

due to biodegeneration [1, 5, 25]. Improvements in recovery techniques, storage, and

cryobiology revived their use. Allografts provide the advantage over autografts of

reduced operating and anesthesia time, reduced operative trauma, limited incisions and

reduced wound complications, all of which are important to consider since the patients

tend to be from the elder population [5, 6].

Hemodialysis patients frequently need a replacement graft as treatment for an

infected arteriovenous (AV) access graft [15]. Additionally, SV allografts have been

used to save limbs in below knee revascularizations due to flexibility, compliance

matching, size matching, and resistance to infection [25]. Vascular allografts have been

promoted by some studies as the best alternative conduit for infrainguinal arterial

reconstruction in infected fields when autologous veins are unsatisfactory or unavailable

[6]. For the coronary circulation, the internal mammary artery (IMA) and the radial

artery are used as bypass allografts. The IMA has been shown to have long-term

patency, but availability is severely limited due to its short lengths and difficulty in

recovering [1].

As mentioned above, two commonly addressed disadvantages to the use of

allograft tissue are the risk of contamination or disease transmission and early to late

occlusions. Good tissue banking practices utilized to prevent a transmission occurrence

include intensive screening of all potential donors, including medical and social history

and various required data to determine if the donor is acceptable [35, 36]. Additionally,









tissue and blood samples are screened with microbiological testing and serological

testing, respectively. The blood samples are virologically screened for HIV, hepatitis B,

hepatitis C, and CMV [5]. Tissue samples follow the grafts through all of the processing

steps in order to detect potential tissue infection following recovery [35]. The

contamination problem of vascular allografts is centered on recovery, sterilization, and

storage. Incidents of contamination can occur at recovery if the barrier to the digestive

tract is compromised. Preparation and evaluation of the tissue for implantation requires

extensive handling under aseptic conditions which adds opportunities for contamination.

Sterilization via antimicrobial solutions, gamma-irradiation and ethylene oxide can prove

to be inefficient or too destructive [37-41]. The latter two have been abandoned due to

tissue damage, leaving treatments by antimicrobial solutions and cryopreservation.

Allograft Processing

Sterilization and disinfection

The active surveillance activities and the outbreak investigation carried out by

Centers for Disease Control (CDC) during 2002 highlight the fact that the spectrum of

bacterial pathogens associated with allograft-associated infections include Gram positive

(S. aureus and Enterococcus) and Gram-negative bacteria, especially those that ferment

lactose [42]. Hence, any antimicrobial solution that is validated for processing allograft

tissue should theoretically be active against any of the above pathogens. The importance

of validating any sterilization technique against a mixture of pathogens is underscored by

the fact that 19% of the allograft-associated infections ascertained during the active case

finding period were polymicrobial. The frequencies and concentrations of various

pathogens in polymicrobial infections remain unknown and will vary from recipient to

recipient and, therefore, are not predictable. Based on data in the CDC investigation, the









antimicrobial solution used should be validated against the following pathogens, at the

very minimum: Staphylococcus aureus, Enterococcus faecalis, Pseudomonas aeruginosa,

Escherichia coli, Clostridium species and Bacillus species [42]. There are no data to

predict the relative proportions of the respective pathogens in the mixture being tested.

The current common disinfection method is an antimicrobial soak in which the

tissue is aseptically processed with antimicrobial and sometimes antifungal cocktail [35].

A study to evaluate the effectiveness of the antibiotic cocktails used by a tissue bank for

disinfection of human tissue was conducted by a GLP/GMP-compliant1 testing laboratory

(AppTec, Marietta, GA). Saphenous vein samples were inoculated with 106 of aerobic,

anaerobic and a mixture of test organisms suggested by the CDC for validation. The

result was a one log reduction or less with absolutely no effect detected in the

Clostridium species [44].

Typically, the patient is safe since the donor will be discarded if representative

samples tests positive for contaminants, but sometimes the method for testing sterility

may be flawed [45]. Specifically, residual antibiotics from the treatment process have

been shown to inhibit growth in post-treatment surveillance cultures producing false-

negative results leading to cases of septic arthritis, hepatitis C transmission,

Streptococcus pyogenes (GAS), and Clostridium spp [8, 46]. Of note, cases of soft tissue

allograft-associated disease transmission by spore-forming organisms that are resistant to

antimicrobial solutions have been reported within the last few years [8, 47]. These tissues


1 GLP/GMP refers to Good Laboratory Practices and Good Manufacturing Practices, respectively. These
are regulations set forth by the Food and Drug Administration (FDA) that requires manufacturers of human
drugs and biological products, animal drugs, medical devices, and food additives to demonstrate safety and
utility of their product
43. Administration, F.a.D., Good Laboratory Practices (GLP) for non-clinical laboratory studies 21
CFR Part 58. p. Supporting Statement..









have predominately been cartilage used in joint repairs, but investigations prompted by

these events have revealed transmissions and infections resulting from other soft tissue

grafts such as tendons and saphenous veins. Vascular tissue is less prone to infection than

cartilage, but it is still vulnerable to organisms (including fungi and viruses) inherent in

the donor or transmitted during recovery or processing that are resistant to the

antimicrobials currently used for allograft disinfection. Additionally, infection can occur

and, in some cases, the rapid onset of infection and the infecting organism itself suggest

that graft to host disease transmission was responsible rather than infection from some

other source implementing the graft as the carrier [23, 48, 49]. More recently, analysis of

the infecting organism's DNA has provided even more convincing evidence that donor to

host transmission of infection has occurred [47].

Cryopreservation

Adverse results with the first cryopreserved vein caused enthusiasm to wane since

techniques led to early thrombotic graft occlusion. Recent improvements in

cryopreservation and controlled freezing have made it possible for arterial and venous

allografts to be used as alternate vascular conduits [50]. The primary goal of a

preservation technique is to maintain cell viability in order to increase vein graft

availability by extending its shelf-life. In the process, the graft should have minimal

antigenicity while viably functional and maintain structural integrity similar to native

vein [6, 15]. The use of cryopreservation of vascular conduits for long-term storage has

also been used as a means of viral inactivation in addition to reducing antigenicity, but

the effectiveness is unclear with a small risk of viral disease transmission remaining.

Cryopreserved veins are commercially available from CryoLife, Inc. (Kennesaw,

GA), Northwest Tissue Center (Seattle, WA), and RTI Cardiovascular (Birmingham,









AL). CryoLife has been processing cryopreserved tissue for implantation since 1984

[35]. The common method is to place the vessels in a buffered solution containing

cryoprotectants, such as DMSO, and then store in the vapor phase of liquid nitrogen (-

110 to -1960C) [6, 15]. An antimicrobial solution may be used initially to disinfect the

tissue prior to storage. As the tissue is transferred from the disinfectant solution to the

freezing solution, representative samples are removed for sterility testing according to

United States Pharmocoepia (USP) standards. The tissue is frozen in a controlled manner

from +4C to -800C or below at an average rate between 0C / minute and 5C / minute.

The cryopreserved vessels are then stored in freezers with the vapor phase of liquid

nitrogen. Thawing and rinsing are complicated, non-sterile procedures performed

aseptically. The rinsing procedure must be performed immediately after the thawing

procedure. The allograft package is first thawed at room temperature and then in warm

saline (32C to 420C). The second event is supposed to rapidly thaw the tissue without

adversely affecting cellular viability [6, 15]. The graft is then aseptically removed from

packaging, flushed and soaked with a series of solutions to remove cryo-solution

components before implanting.

Freezing with cryoprotectants prevents the intracellular vapor phase gradient that is

created when the extracellular matrix freezes. The cytoplasm remains in liquid form and

causes cellular membrane disruption from intracellular swelling as a result of osmosis,

producing a nonviable graft with structural damage. The cryoprotectant protects the

matrix and the cells by replacing the water [15]. Without cryoprotectants, saphenous

vein allografts have been associated with high graft failure rates in infrainguinal bypass

procedures [6]. At the same time, an ideal method of cryopreservation has not been









determined. Experimental and clinical data suggests that current cryopreservation

protocols may be responsible for making allografts more brittle and could induce early

graft rupture or dilatation [51-55]. Cryopreserved veins are prone to fissuring during

thawing which degrades the strength and performance of the graft. Therefore, aneurismal

degeneration continues to be a matter of concern. This may be due to preservation of the

endothelial antigenic properties as wells as inadequacy of the structural and functional

media characteristics (elasticity and compliance) [6]. In clinical surgery, there have been

mediocre results leaving cryopreserved veins as an alternative only when no other

suitable conduit is available.

A majority of the investigations on cryopreserved vessels show evidence of

antigenicity and rejection. The development of intimal hyperplasia that leads to graft

failure is compounded and accelerated by immune-mediated inflammatory response

secondary to rejection, the effect of cryoprotectants, surgical injuries and theological

mismatch [6]. Explanted tissue has shown cellular infiltrates that correlates with wall

hemorrhage and necrosis [6]. Venous allografts appeared to induce anti-HLA antibodies

post implantation and an associated humoral immune response contributing to failure.

Most inflammatory cells are activated by cytotoxic T cells, expressing CD3, CD8 and

HLA-DR antigens. On the other hand, when compared to an acellular graft made of

foreign material, PTFE, cryopreserved saphenous vein allograft showed no significant

difference in T-cell subpopulations or lymphocytotoxic antibodies [15].

Immunohistochemical examination showed minimal mononuclear cell infiltration in

adventitial layer indicating no significant immunological activation or clinical rejection.

Contradictory results on the preservation of the endothelial layer and smooth muscle









layer show improvement in patency rates in animal studies while no clear benefit in

humans has been discovered [6, 15]. This is most likely due to the freshness and young

age of the tissue in an animal study compared to the older population of human donor

tissue that may take as long as 24 hours to recover followed by additional time to process

for implantation. This may be another indication why autograft performs better than

allograft.

The ultimate immune response may also be directed against smooth muscle cells

when the endothelial layer is absent or subconfluent. Allosensitization occurs due to

expression of class I and II major histocompatibility complex (MHC) antigens present on

preserved endothelial cells as well as non-MHC antigens that stimulate T-cell mediated

rejection response [6, 15]. Immunologic mechanisms have been implicated in early

thrombotic occlusion and poor patency rates in lower extremity revascularization with

cryopreserved allograft [1, 5, 25]. There has been evidence of the development of intimal

hyperplasia and late graft failure due to the replacement of the smooth muscle cell layer

by fibrotic layers after implantation [6, 15]. Additionally, the structural integrity can be

compromised with remodeling by structural reorganization of the graft and thickening of

the intima from smooth muscle cell proliferation. This is followed by a gradual decrease

in compliance from fibrosis in the layers as well as immediate disintegration of

immunocompetent cells and release of proteases that probably cause elastin degradation

[34]. Elmore et al found similar contractile characteristics and graft patency rates but

some cellular differences when comparing noncryopreserved to cryopreserved grafts in

canine femoral arteries [15]. There was a decrease in smooth muscle cell relaxation









response to nitric oxide (NO) in cryopreserved grafts as well as decreased relaxation to

phenylephrine and endothelin [6, 15].

The disadvantage to cryopreservation is related to the second issue raised with the

occlusion of allograft veins that is most likely due to an immune reaction similar to a

graft rejection response. Since endothelial cells, smooth muscle cells and fibroblasts

contain antigens on their cell membranes, preservation of these components could

theoretically have deleterious consequences by mediating immunological reactivity [25].

Additionally, preserving these cells creates the need to consider ABO-compatibility and

lymphocyte cross-matching [5]. Since there is a limited availability of allograft arteries,

a long waiting list may be a problem limiting widespread use.

Antigenicity

Blood typing is required for vascular grafts to prevent rejection since simply

inactivating contaminants is not sufficient protection. There is a limited availability of

vein allografts due to ABO relevancy and a depletion in inventory because of positive

cultures [5, 6]. At an AATB accredited tissue bank that utilizes antimicrobial soak and

cryopreservation had a 2003 yield of saphenous vein grafts per donor of only 58% where

about 236 saphenous veins were discarded due to positive first and final cultures. ABO

matching contributes to this deficiency, but only ABO compatibility has been evaluated

in most studies [6]. Simply matching blood type is not sufficient to prevent an immune

response. Human leukocyte antigens (HLA), proteins found on most cells, are used by

the immune system to distinguish between endogenous and foreign cells. Additionally,

there are many HLA antigens, so HLA matching is not feasible for clinical applications

and the role of matching HLA and ABO for humans has not been determined [5, 6].









Again, removing the cells from the graft would alleviate the problems with finding the

appropriate donor.

It is believed that the difference in allograft healing compared to autograft healing

is mainly due to this immunologic response. This contributes to the low patency rates

seen in allograft use. Graft rejection is a response to class I and class II MHC antigens

laden on the membranes of endothelial cells [56]. This layer is in immediate contact with

inflammatory cells in blood such as neutrophils, monocytes, and macrophages [29, 57].

Rejection plays a significant role in failure and leads to allosensitization [6]. Studies

have found pathophysiological processes develop in the walls and gradually lead to

stenosis, mainly at anastomosis, and segmental dilatation that may be a result of chronic

transplant rejection.

It appears from some studies that an intact endothelium is a prerequisite for

improved long-term patency of autologous and alloplastic bypass grafts [5]. The

endothelium is an active barrier that allows permeability to fluid and macromolecules

while limiting entry of inflammatory cells (polymorphonuclear cells and monocytes) and

immune effector cells (B and T lymphocytes). It plays a critical role in inflammation as

well as maintaining a nonthrombogenic surface by maintaining normal platelet activation

[58]. Cryopreserved veins boast at the most approximately 70% viability, leaving an

insufficient endothelium that is not fully functional [51, 53]. When dysfunction in the

endothelium occurs, pathological processes result, leading to endothelial cell activation

and thrombus formation. There is an injury response with inflammation and wound

healing where thrombin and fibrin are generated and platelets are activated. The graft

begins to occlude with loss of patency and early graft failure.









Chronic rejection also affects the vascular wall with medial destruction where the

arterial wall dilates and possibly ruptures [59]. Davies compared the healing response of

allograft to autograft in a rabbit model with a common carotid artery bypass [60]. They

found there were two phases in the response. The first phase consisted of significant

thickening of the intima and media compared to autograft. The second phase showed a

51% decrease in the intimal thickening and a maintained medial thickening of 97%

compared to autograft. The allograft had an exaggerated medial response while the

autograft had an exaggerated intimal response. Overall, there was a net increase of 17%

in allograft wall thickness compared to autograft. Even so, the lumen between the two

sets of grafts was not significantly different after 28 days.

A similar medial response was seen in an abdominal aortic graft model in rats but

the lumen appeared to occlude. There was medial cell loss and matrix degeneration,

adventitial inflammation, and intimal hyperplasia. The lumen narrowed by intimal cell

proliferation that was found to express a-actin, a marker for smooth muscle cells and

fibroblasts. The smooth muscle cells were destroyed and the elastic lamina became

dislocated and fibrosis was noted [59, 61, 62]. The inflammation in the adventitia

disappeared along with the removal of the SMCs [59, 61]. This gives credence to the

antigen induced response by both the endothelial cells and SMCs.

Use of immunosuppressant and anticoagulant therapies

It is still questionable if cell viability is beneficial since it maintains its

antigenicity which has led to the use of short term immunosuppressant and

anticoagulation therapies [5, 6]. The use of immunosuppression and anticoagulation

therapies, along with cryopreservants with controlled freezing, has been shown to









improve patency and graft acceptance to combat cryopreserved conduits more thrombotic

behavior early on [15].

Posner et al. showed improved one year patency with infrainguinal arterial

reconstruction with low-dose immunosuppression treatment [63]. They also showed a

substantial number of allograft-related complications with early degeneration,

pseudoaneurysm formation, and hemorrhaging. An improvement in graft patency in

animal models and human arterial allograft patency is seen with high-dose

immunosuppressive treatment, whereas, the low-dose regimens were not enough to

curtail the allogeneic response [6]. The more potent treatment may be effective, but the

toxicity level may not be tolerated by older patients with multiple co-morbidities as well

as increasing the risk of infection, with reduced graft rejection, but disturbed wound

healing, and possibly gangrene [5]. Therefore, patients with infected tissue necrosis, an

indication for surgery, would be at greater risk if the immune system is compromised

with treatment. One study where patients received high-dose immunosuppression for

kidney transplants did not prevent rejection [6]. A histological and immunohistochemical

analysis of the renal arterial allografts of the failed kidney transplants showed evidence of

chronic rejection and intimal hyperplasia in the vessel walls. Most patients with critical

limb ischemia will have reduced renal function. Exposing them to therapy may lead to

further reduction of renal function [5]. In one study, treatment was discontinued in 12%

of the patients due to gastrointestinal bleeding [6]. These conditions preclude a large

population of the patients needing revascularization from receiving cryopreserved grafts

if immunosuppression is required.









Denudation of cell layers

Experimental and clinical studies on cryopreserved veins show that many

functional features of the endothelial cells and smooth muscle cells are preserved [6, 15].

These properties are prostacyclin and endothelin production, fibrinolytic activity,

thrombotic properties, platelet deposition, and metabolic mechanisms. However, after

implantation some studies show the presence of an intact endothelial layer while

numerous animal studies and some clinical studies shows moderate endothelial

denudation to complete loss of endothelial cells after implantation [15]. It seems in some

preservation protocols the endothelial layer may no longer be bound to the graft and is

removed in the circulatory flow negating the efforts made to maintain and preserve that

layer. The loss of the lining and smooth muscle cells are due to the inability of the

preservation process to maintain complete cell viability resulting in a loss of adherence to

the basement membrane. This is presumably due in part to a more rapid accumulation of

low-density lipoprotein cholesterol compared with noncryopreserved human saphenous

vein grafts [5, 6, 15]. Other possible factors are toxicity of the preserving agents

(particularly with antibiotics), cryopreservation events, spontaneous cell death, or

destruction of the surviving cells by immunocompetent host immune cells [5].

Additionally, the ischemic time for allograft tissue varies up to a 24 hour period resulting

in variable cell viability prior to cryopreservation.

Decellularization

The disadvantage to allograft use is its biologically active components containing

antigens that could lead to allosensitization, resulting in an immune reaction and

rejection. Cryopreserved grafts retain viable or nonviable cells which still contain the

antigenicity present in the protein structure of the cell surface. This has led to the









investigation of decellularized conduits for vascular reconstruction. There is a general

consensus among investigators that removing, instead of merely inactivating, the cellular

content may reduce or prevent an immunological response.

Biologically active endothelial cells express class I and class II major

histocompatibility complex antigens that result in an immunological response [64].

Human studies using antibiotics and cryopreservation to process allografts have

demonstrated a panel of these reactive antigens, some of which caused a broad recipient

allosensitization [64, 65]. This is of particular interest in patients with end-stage renal

disease awaiting a kidney transplant. The presence of alloantibody could preclude renal

transplant due to cross-match since these patients are already immune compromised with

decreased functioning of white blood cells [64, 66].

There is controversy in the literature as to the allosensitization of cryopreserved

grafts. The levels of panel reactive antigens (PRA) in patients whom received

cryopreserved vein grafts showed allosensitization within approximately 3 months with

an 84.1 % increase [64]. Although, when compared to decellularized grafts (SynerGraft,

CryoLife, Inc.) no significant difference was found in primary, assisted primary or

secondary patency. A study of cryopreserved saphenous (SV) allograft in lower

extremity revascularization yielded poor patency rates that were believed due to

immunological reactions. These, and other recent studies demonstrated venous tissue

expresses class I and class II major histocompatibility (MHC) antigens present on the

preserved endothelial cells as well as non-MHC antigens that stimulate T-cell mediated

rejection response [67].









If injured or subconfluent endothelial cells that line the lumen are present,

thrombosis and smooth muscle cell (SMC) growth are promoted leading to intimal

hyperplasia [7]. This is likely due to the major antigenic determinants, class I and class II

MHC, present on the SMC. Therefore a denuded vessel leaves no protection from an

immune reaction. It has been shown that the endothelial cells of cryopreserved vessels

are not fully viable or confluent and slough off into the blood stream after implantation.

Rejection plays a significant role in failure and leads to allosensitization.

Allosensitization was demonstrated in patients who received cryopreserved allografts in a

SynerGraft (CryoLife, Inc) study. Blood tests showed an increase in PRA levels partly

due to exposure to the allogeneic antigens on the preserved endothelial cells. A study of

cryopreserved saphenous vein allografts in lower extremity revascularization yielded

poor patency rates that were likely due to immunological reactions. Another study using

aortic valve allografts (AVA) showed early failure caused by donor-specific immune

response [68]. Preserved vessels used for hemodialysis access caused allosensitization

with increased PRA levels. This can be a potentially serious problem for kidney

transplant recipients since the presence of alloantibody could preclude a renal transplant

due to cross-match incompatibility [64].

Chronic rejection can cause vascular wall destruction making the graft unsuitable

for long-term applications. The medial layer is destroyed with altered extracellular

matrix leading to wall dilatation and rupture [59, 62]. This event could be catastrophic in

a coronary site that may lead to morbidity or mortality. The cellular events that occur

was modeled in abdominal aortic grafts in a rat model by several researchers and

summarized by Allaire [59]. Essentially, there is medial cell loss and matrix degradation,









adventitial inflammation, and intimal hyperplasia. This is a typical response for immune

injury in arterial allografts [59]. Through the response, the media is thinned and the

elastic lamina is dislocated [59, 62]. Smooth muscle cells are destroyed while

inflammatory cells invade the adventitia, most likely due to the antigens present on the

SMC since the inflammatory cells disappeared when the SMCs disappeared [59, 61, 62].

The response results in degradation of the matrix, fibrosis and functional failure.

Additionally, the graft is occluded by narrowing of the lumen caused by the intimal cell

proliferation thickening the vascular wall. Current strategies to reduce or prevent this

cascade of events are immunosuppressives drugs or reducing the antigenicity of the graft

with cross-linking agents, such as gluteraldehyde, or by sequential chemical treatments

that promotes tissue decellularization.

Decellularization involves the removal of major immunogenic components such

as cells and their lipid membranes, membrane associated antigens and soluble proteins,

and other lipids and more soluble glycosaminoglycans [62, 69]. The resulting graft is an

acellular scaffold with only insoluble structural proteins, such as collagen for strength

and elastin for distensibility [69, 70]. Primarily present on the endothelial cells and also

present on SMCs are class I and class II major histocompatibility complex antigens. A

10 and 20 week sheep study with an allograft patch model showed only 1 out of 8 sheep

with a decellularized graft demonstrated a positive elevation in PRA levels to MHC I and

none to MHC II. In the classically cryopreserved group, 2/3 of the sheep showed an

elevation [61].

Decellularized grafts have been evaluated histologically, immunohistochemically,

mechanically and with both animal studies and clinical studies. Hilbert evaluated the









morphologic characteristics in a small-diameter freeze-dried decellularized carotid artery

of a goat decellularized with a nonionic detergent solution. Histological analysis showed

complete endothelial cell removal while preserving the basement membrane.

Additionally, no cells were seen in the intima, media or adventitia, although, there were

remnants of smooth muscle cells in the media without nuclei. The extracellular matrix

seemed well preserved, but oval to circular spaces were occasionally noted in the media.

The internal elastic lamina remained intact, as well as the layered elastic laminae in the

media [70].

In an animal study conducted by Hilbert, all allografts were patent upon

explanation with no significant dimensional changes or aneurism formation. The

autografts showed a luminal surface lined with a layer of endothelial cells while the

allograft showed a discontinuous layer. Even in these regions of incomplete

endothelialization thrombi were not observed. Both types of grafts showed

myofibroblasts, collagen and proteoglycans that may be indicative of incorporation [70].

The allograft appeared relatively acellular in the media with focal cellular regions

among the dense collagen and elastic lamellae containing infiltrated myofibroblasts of the

hosts. The authors believe that the number of host myofibroblasts in the media may have

been significantly limited by the presence of the dense collagen bundles and smooth

muscle cell remnants. Host cell migration was most apparent close to the anastomoses

and was rich in proteoglycans. It appeared to occur in the adventitia along the length of

the grafts. Histologically, there was no evidence of calcification and inflammatory cells

were not present in the graft wall. Electron microscopy did show calcification of minute









remnants of cell membranes. Ingrowth of host blood vessels was not observed after 6 to

7 months [70].

Conklin investigated decellularization of a small-caliber xenograft with a porcine

common carotid artery model [2]. This group followed cell lysis with multiple enzymatic

digestions and detergent washes while agitated. Histology and electron microscope

showed complete removal of cellular components while the extracellular matrix appears

to remain intact. Hematoxylin and eosin (H&E) stain showed no signs of remaining

nuclear material in the walls of the vessels and TEM showed complete removal of

cellular material from the media layers. Histology showed an intact internal elastic

lamina along with elastin lamellae in the media. TEM showed the basic extracellular

microstructure remained intact after processing.

In addition to reducing immune reactions and decreasing the thrombogenicity of

the luminal surface by decellularization, this group wanted to ensure that the process

would maintain a graft with similar mechanical properties to the native vessel, high

strength and good handling characteristics. Processing with chemicals or treating with a

cross-linker may affect the integrity of the matrix in focal regions leading to aneurysm

post implantation, weakening of the graft, or making it stiffer and more brittle. They

investigated the compliance and burst strength of the vessels on a custom-built system

that measures the diameter changes while increasing the intraluminal pressure. The

compliance is expressed as the percentage diameter change per pressure (mm Hg) change

normalized to the compliance of fresh vessels. The average compliance was calculated as

the slope of the linear regression line in the physiological pressure range (70-130 mm

Hg).









The mechanical analysis was performed on decellularized vessels, decellularized

and heparin-treated vessels, fresh vessels, alcohol preserved vessels, and ePTFE

prosthetic grafts. The ePTFE grafts were the least compliant (AD/mm Hg 0.024%) and

the decellularized grafts were slightly more compliant (AD/mm Hg 0.181%) than the

fresh vessels (AD/mm Hg 0.172%). Alcohol caused the tissue to stiffen some (0.160%)

while Heparin treatment produced a much stiffer graft (0.0975%). During burst testing,

none of the fresh vessels burst within the limit of the pressure transducer (2300 mm Hg).

One out of four of the decellularized vessels burst within the limit at 1654 mm Hg.

Although there may be some strength loss measured in that one vessel, there still is a high

safety margin over ten times the physiologic pressure. Unfortunately, only four samples

were tested in this group.

This group examined their decellularized process with heparin cross-linking on a

carotid artery bypass in dogs. At 24 days, fibroblast-like cells appeared to densely

populate the media and there were few endothelial-like cells lining the lumen. After two

months, dense a-actin staining suggested smooth muscle cells densely populated the

vessel wall and Factor VIII staining confirmed that endothelial cells lined the lumen.

Teebken et al. developed an acellular vascular xenograft matrix from porcine

thoracic aortas for the purpose of seeding with human cells [71]. This group

decellularized with enzymatic cell extractions using biological enzymes trypsin,

ribonuclease (RNase) and deoxyribonuclease (DNase). The idea is that these enzymes

are capable of removing cell components as well as cellular antigens, lipids and to some

extent glycosaminoglycans with limited toxicity. Additionally, this group believes that

extractions with detergents may disturb the endothelialization of the grafts both in vivo









and in vitro. Light microscopy showed no cell nuclei or intracellular components in

cross-sections of the aortic wall after processing. Immunohistochemical stains for

fibroblasts and endothelial cells were all negative. SEM showed extracellular matrix

fibers with openings of 1 to 10 |tm.

Schaner studied the composition and strength of decellularized human greater

saphenous vein specimens to determine their potential as a vascular tissue-engineering

scaffold [72]. This group used a detergent, sodium dodecyl sulfate (SDS), as a

decellularizing agent. Transmural cell removal was found to be nearly complete (>94%)

at a concentration of 0.075% SDS. The luminal surface appeared completely devoid of

endothelial cells at all concentrations tested. The collagen morphology appeared

unchanged, the basement membrane remained intact and the elastin staining decreased

only slightly.

This group also performed mechanical testing to evaluate the effects of the

process on the burst strength and suture retention strength. It was noted that the vessels

had normal consistency and good handling characteristics. The burst strength of the

decellularized graft was similar to the fresh vein (2480 460 mm Hg vs. 2380 620 mm

Hg (p>0.05)) as well as the suture-holding strength (185 30 gm vs. 178 66 gm

(p>0.05)). The functionality was examined by placing decellularized canine jugular

veins into a carotid interposition model. Each canine received an autograft and either an

allograft or a decellularized allograft on the contralateral side. After two weeks, all grafts

were assessed by Duplex imaging to be patent with no significant dilation, rupture or

anastomotic false aneurysm.









CryoLife, Inc. (Kennesaw, GA) is filing for FDA approval for a decellularization

process on veins and heart valves called SynerGraft [64, 67]. This process involves cell

lysis in hypotonic sterile water, followed by equilibration in a buffer. The tissue is then

treated by an enzymatic digestion of the nucleic acids with a combined solution of RNase

and DNase. The grafts are then cryopreserved for storage. Results on the CryoValve SG

showed approximately 99% reduction in staining of the endothelial cells and interstitial

cellular elements [67].

In a clinical study by Hawkins, 14 children (8.5 7.9 years) received

decellularized, cryopreserved allografts (CryoLife, Inc.), 6 were patch insertions and 8

with valved pulmonary allografts [67]. These groups were compared to 20 historical

control subjects (1.7 2.4 years), 8 with valves and 12 with allograft patch insertions.

There was no attempt to match ABO blood types in either group. The effect on the

immunogenicity was measured at 1, 3 and 12 months by the frequency of HLA

alloantibodies PRA: class I (HLA-A, HLA-B and HLA-C) and class II (HLA-DR/DQ).

Antibody levels were slightly higher from preoperative levels for both classes at all time

points. The antibody levels were significantly lower in the decellularized allograft group.

There was a marked reduction in staining for class I and class II histocompatibility

antigens. However, there was no work done in this study to determine whether reduced

immunogenicity will truly allow tissue ingrowth and improved long-term durability in

patients.

Madden performed a clinical study of 20 patients with an upper extremity

SynerGraft cadaver vein allograft for hemodialysis access [64]. The first 17 patients

were matched (ABO) for blood types while the 3 remaining patients were intentionally









given ABO-incompatible SynerGraft allografts. Allosensitization was quantified using

the PRA assay (American Society for Histocompatibility and Immunogenetics laboratory

protocol) on peripheral blood samples. None of the SynerGraft patients became

allosensitized (PRA > 10%) at 10 months with a mean PRA of 3.2% (0 7%). All

patients in the historic cryopreserved group became allosensitized by 3.1 months with a

mean PRA of 84.1%. All three patients with ABO-incompatible SynerGraft allografts

showed no allosensitization or acute rejection-type reactions. Typically, removal of

viable donor endothelial cells is believed to lead to increased thrombogenicity and

decreased infection resistance. There was no significant difference between the groups in

the primary, assisted primary or secondary patency rates and no grafts were lost to

infection.

Expected Outcomes

The deliverables from this study are conditions and exposure times for a chemical

sequence that will kill and remove infectious agents as well as endogenous cellular

material without degradation of the matrix. Removal of the endothelium and other

reaction causing cells will eliminate the need for blood typing, reduce immune reaction,

increase durability and longevity and reduce the number of suitable donated grafts

rejected due to positive culture.




















CHAPTER 2
DECELLULARIZATION OF VEIN GRAFTS

Introduction

Vascular disease is one of the most prevalent diseases worldwide with bypass or

replacement as the mainstay of therapy for coronary and peripheral occlusions. Over

100,000 peripheral and over 400, 000 coronary bypasses are performed in the US

annually [4, 27]. This results in a need for readily available small-diameter vascular

grafts that are immunotolerant and durable. For coronary bypasses, the autograft

saphenous vein or mammary artery are the standard and autologous saphenous vein

remains the conduit of choice for infrapopliteal bypasses due to its flexibility [2, 70].

Using autologous tissue requires a second surgery site and extended operating time and

requires multiple revisions with a 4 year patency rate of only 40 to 70% [2].

Additionally, as many as 30% of patients will have unsuitable autograft material, forcing

them to consider other options such as synthetics, allograft or xenograft [2, 73].

Allograft tissue has an advantage over synthetics due to its ability to be innervated

by host cells and resistance to infection. The starting material is similar to the native

tissue in its structure and function with the geometry and components necessary for cell

differentiation. It contains the possibility of autologous cell infiltration creating a

biologically active matrix with phenotypically appropriate cells. This allows for









reparative and functional characteristics not inherent in nonviable prosthetics. In the

same manner, the disadvantage to allograft is its biologically active components

containing antigens that could lead to allosensitization and chronic rejection.

Allografts are plagued with complications leading to early stenosis and graft

failure. Occlusion or stenosis in coronary or peripheral circulation is a common clinical

occurrence present in small-caliber applications that results in additional interventions

[1]. Early thrombosis occurs when platelets in circulating blood adhere to certain

surfaces, then release chemicals to attract more platelets to form a large aggregate that

generates thrombin [67]. An animal study using allogenic valve grafts in the descending

aorta of rats showed thrombus formation after 21 days [68]. This is the most common

failure mechanism for synthetics, but also occurs in allograft tissue. Cellular events that

occur after implantation contribute to this loss of patency.

Adaptation of vein grafts to the arterial environment has been studied extensively

in animal models but less in humans. After implantation, the conduits undergo structural

changes characterized by intimal hyperplasia and overall wall thickening that is believed

to be a result of the additional immunological stimuli when compared to autograft [2, 7,

59, 60, 70]. Fibrous intimal hyperplasia develops from an excess of smooth muscle cell

proliferation. This flow-restricting lesion may occur diffusely throughout the graft, or

more commonly, at focal sites near anastomoses. A study by Davies et al. using a

common carotid vein bypass graft in rabbits was used to compare the differences between

the healing response of autograft and allograft [60]. All grafts remained patent up to 28

days. Compared to the autografts, the allografts showed a 51% decrease in overall mean

intimal thickness and a 97% increase in the overall mean medial thickness, resulting in a









net increase in the vascular wall thickness. Interestingly, the lumen of the autografts and

allografts were not significantly different. It appeared that intimal hyperplasia occurred

more in the autograft while the media of the allograft had a greater response. Of note, the

intimal hyperplasia occurred with intact endothelial cells on the surface.

The traditional method of allograft storage utilizes cryopreservation for storage.

The intent is to produce a viable construct with minimal antigenicity, optimal biologic

functionality and structural integrity. Cryopreservation has been shown to yield

improved patency rates over fresh and synthetic grafts in hemodialysis access. The

authors of this study believe that it is infection resistant with a large portion of the

endothelial cells viable at engraftment [64]. At the same time, clinical and experimental

data suggest that cryopreservation may be responsible for rendering allografts more

brittle and could induce early graft rupture or dilatation [51-55]. Additionally, the viable

cells remaining on the endothelium and within the vascular wall contain the antigens that

in this instance may lead to the rejection response seen in allografts.

Acellular Allografts

The limitations with conventional cryopreserved small-diameter vascular grafts

due to immunogenicity and rupture have led to the development of new decellularized

vascular grafts. Fixing the tissue with cross-linkers is used to reduce the

immunogenicity, but can be toxic or lead to aneurism formation. Decellularization

involves the removal of major immunogenic components leaving an acellular scaffold

with only insoluble structural proteins, such as collagen and elasin [69, 70].

CryoLife, Inc. (Kennesaw, GA) has developed a decellularization process on veins

and heart valves called SynerGraft [64, 67]. This process involves cell lysis in hypotonic

sterile water, followed by equilibration in a buffer. The tissue is then treated by an









enzymatic digestion of nucleic acids with a combined solution of ribonuclease (RNase)

and deoxyribonuclease (DNase). The grafts are then cryopreserved for storage. Results

on the CryoValve SG showed approximately 99% reduction in staining of the endothelial

cells and interstitial cellular elements [67].

Maintaining the similarities to the native tissue of the biological composition and

geometric design is essential to promote re-endothelialization and cell migration [61, 70,

71, 74]. The basement membrane on the luminal surface consists of type IV collagen

with ligands for firm endothelial cell and myofibroblast attachment and retention [62, 70,

72]. Preservation of this structure facilitates re-endothelialization to reduce

thrombogenicity and promotes migration of myofibroblasts into the vascular wall [59, 70,

75]. An insoluble extracellular matrix has been shown to promote fibroblast proliferation

and elastin synthesis and organization into fibers [59]. Another study showed the

importance of complete cell removal where host myofibroblasts in the media appeared to

have been significantly reduced by the presence of collagen bundles and smooth muscle

cell remnants that restricted migration [70].

There is concern that removal of the viable endothelial cells could result in

thrombogenicity and decreased infection resistance. It is believed by some investigators

that a lack of surviving cells and endothelial integrity may play a role in graft

degeneration and early and late patency [5]. Madden showed no significant differences

in patency or infection rates when comparing CryoLife's SynerGraft to their own

cryopreserved grafts [64]. Virtually all cryopreserved homografts are acellular within a

year of implantation. It has been shown that after implantation cellular grafts become

acellular over time while decellularized grafts have a time-dependent recellularization









during the same period [61]. A cellular aortic allograft conduit transplanted into a rat

showed a complete loss of smooth muscle cells in the media after 21 days [68].

Inflammation along with the lack of cellular function limits tissue ingrowth, performance

and reparation. The tissue typically scars and then mineralizes [61, 68]. A cryopreserved

allograft pulmonary trunk implanted as a patch was found to become acellular over a 20

week period [61]. A fibrous sheath developed consisting of layers of fibroblasts that line

the luminal side and encapsulate the adventitia. This event may represent the same

foreign body response seen in surgical implants. At the same time points, the

decellularized tissue displayed time-dependent recellularization. Thus, the decellularized

scaffold appeared to have an advantage over the cellularized graft with faster

repopulation and remodeling due to bypassing the in vivo decellularization step.

With traditionally cryopreserved methods, the endothelial cells slough off and the

interstitial cells are removed before the lumen is relined. These series of events could

prevent or slow down the remodeling process by exposing the cellularized venous wall,

sparking an immune response that proceeds faster than the repopulation of the

endothelium. At the same time, transanastomotic endothelialization has been limited in

humans despite the success in animal models and is an ongoing challenge in the

development of vascular grafts. A four year study in a canine model with iliac and

carotid arteries placed in the femoral and carotid positions showed complete

endothelialization of the flow surface [62]. The same group with the same model

compared arteries to autograft saphenous vein and showed no evidence of

endothelialization in the allografts after 6 months while the autografts had substantial

endothelial cell coverage. In another study by the same group all grafts remained patent









at four weeks [76]. Therefore, results can be variable and unpredictable and appear to be

time-dependent. Although, 4 of the 9 decellularized allografts were patent compared to 4

out of 7 cellular autografts. Cell repopulation in humans rarely occurs, but again may be

attributed to the presence of necrotic cell debris or even apoptotic cells that block specific

cell signals that cause autologous cell repopulation [61]. Other studies have looked at the

compressed nature of the media after decellularization with dense collagen and elastin

fiber networks reducing the porous network and preventing complete migration passed

the intima [74].

Therefore, complete removal of antigens is anticipated to decrease the

immunological response, thus, reducing the need for immunosuppressants and impacting

the durability [7, 61, 62, 69, 75]. Remnants in the arterial or venous wall may promote

increased innate and cellular immune responses with inflammation and consequential

scarring, contributing to failure of organized migration of phenotypically appropriate

cells. Additionally, cell remnants have been attributed to calcification of veins and

valves [61, 70, 73, 77]. A sheep model showed calcification occurred in the classically

cryopreserved tissue while decellularized tissue stained negative except around the suture

[61].

Decellularization Process

Cell Removal Assessment

The decellularization process in the literature follows three main steps, although

the specifics of each step vary. The first step in the process is to osmotically lyse the

cells contained within the tissue. This has been performed with sterile water, a hypotonic

solution or a detergent solution [2, 64, 67, 72, 73, 78]. The cell inactivation is followed

by enzymatic digestion of nucleic acids. Some groups use detergents and some prefer









biological enzymes with the intention of protecting the matrix for re-endothelialization as

well as limited toxicity. Additionally, all methods use mechanical stripping such as

shaking. The final cleansing step involves a wash out to remove any residual cellular

elements and chemicals. All of the studies required a multiday process to achieve greater

than 90% decellularization and resulted in variable matrix integrity.

The most common detergents used were sodium dodecyl sulfate (SDS),

Octylphenol ethoxylate (Triton-X 100) and sodium deoxycholate. SDS is a highly ionic

detergent with an anionic hydrophobic ligand [79]. It lyses cell membranes and is

believed to be uniform within all layers of vascular tissue [72]. It only requires a low

concentration to remove endothelial cells, but removal of smooth muscle cells is dose-

dependent. It is an amphipatic molecule that associates with proteins via its hydrophobic

domain, leaving the hydrophilic region exposed. This possibly creates an altered internal

charge state that leads to swelling of tissue by increased water binding. Collagen has

more hydrophilic sites than elastin, thus has an affinity for water. Decreased thermal

stability of collagen is due to the disruption of hydrogen bonding [77, 79].

Krasovakaya showed the hydrolysis of elastin with pancreatic elastase was

markedly accelerated if pretreated with SDS [77]. Due to the elastin being predominantly

hydrophobic, SDS reduces its hydrophobicity, thus exposing it to an aqueous

environment and making it more susceptible to elastases [77, 79]. Additionally, SDS is

difficult to rinse from the fibers, so it could alter the mechanical properties by binding to

the polypeptide chains within the fibers or binding into the interfiber spaces of the outer

surfaces of the fibers [79].









Residual amounts of SDS in the tissue can also be toxic to cell seeding in vivo and

in vitro. Investigators have shown that porcine aortic roots treated with SDS were

surrounded by nonviable endothelial cell fragments that were seeded onto the matrix

[80]. Within 24 hours of incubation there was extensive cell lysis with patchy cell

distribution. SDS has been shown to be the most effective in cell removal, but causes

detrimental structural changes [69, 77]. Courtman abandoned further use of SDS after

witnessing a significant drop in the shrinkage temperature of bovine pericardium, as well

as a three-fold increase in tissue thickness likely due to swelling [77]. A study with

pulmonary porcine valve conduits decellularized in 0.1% SDS showed complete

decellularization but at the cost of a significantly disintegrated matrix at as early as 24

hours incubation time [69]. Thermogravimetric analysis (TGA) and differential scanning

calorimetry (DSC) was used by Samouillan to analyze the effects of detergents on the

extracellular matrix of pulmonary aortic roots. They found that SDS had a destructive

effect with destabilizing the collagen triple helix and swelling of the elastin network [79].

Schaner, on the other hand, showed no significant alterations with their decellularization

method [72]. These authors studied the composition and strength of a decellularized

human greater saphenous vein specimen to determine its potential as a vascular tissue-

engineering scaffold. Transmural cell removal was found to be nearly complete (>94%)

at a concentration of 0.075% SDS. The luminal surface was completely devoid of

endothelial cells at all concentrations tested, but the cell removal in the vascular wall

appeared to be concentration-dependent. Histology showed the collagen morphology

was similar, the basement membrane was intact and the elastin staining only decreased

slightly. To assess the effects on the structure of the vein, this group performed









mechanical testing. The burst strength of the decellularized graft was similar to the fresh

vein (2480 460 mm Hg vs. 2380 620 mm Hg) as well as the suture-holding strength

(185 30 gm vs. 178 66 gm). The decellularized vessels were placed as carotid

interposition grafts and assessed by Duplex imaging after two weeks. All grafts were

patent with no significant dilatation or rupture. No work was done to look at the

inflammatory reaction or the recellularization of the grafts.

Nonionic detergents have been found, in most cases, to be effective in cell removal

while maintaining the extracellular matrix [69]. Triton-X 100 is an excellent detergent

and emulsifying agent that has an effective performance across broad temperature ranges,

is soluble in water and biodegradable [81]. Additionally, it is compatible with anionic,

cationic and other nonionic surfactants which makes it favorable to combine with other

agents for a synergistic effect. Courtman chose 1% Triton-X 100 to decellularize canine

iliac and carotid parties to be used in femoral interposition bypass grafting [76]. After four

weeks, angiogram showed that all grafts were patent. Further analysis showed no

thrombus formation and no aneurism formation. There was no evidence of inflammatory

cells except in small areas at the anastomoses. There were some cells on the lumen

consistent with endothelial cells and some mesenchymal cells in the adventitia, but the

media was completely acellular. There was a pannus ingrowth of smooth muscle cell at

each anastomosis. Triton-X 100 showed no change in cellularity in a rat aortic valve

allografts (AVA) at 1% and little change at 5% for 24 hours [68]. Samouillan also

looked at the effects of Triton on the extracellular matrix with TGA and DSC and they

found it had no effect on the structural integrity and the collagen helix remained stable

[79]. Cho achieved complete cell removal with 0.5% Triton and successful seeding of









bone-marrow derived cells resulting in the typical cobblestone morphology visible in

scanning electron microscopy [82]. This was confirmed with van Wildebrand factor and

CD31 staining, which are phenotypic markers for mature endothelial cells.

Triton was combined in several studies with sodium deoxycholate, resulting in

complete decellularization with preservation of the matrix [69]. Kasimir found that

pulmonary valve conduits could become completely acellular incubated for 24 hours at

concentrations as low as 0.25%. Kasimir added EDTA to the solution, which may have

improved cell removal. The matrix appeared well preserved with minimal structural

alterations. Reider, a co-author with Kasimir's study, found incomplete removal of

porcine aortic roots and pulmonary roots at the same conditions but required intensive

washing to remove the residual chemicals [80]. When this group seeded the matrix, it

found that the treatment enabled host recellularization with a confluent layer of

endothelial cells on the lumen surface. This was an improvement to their cell seeding on

the grafts decellularized with SDS [80].

Detergent residuals can be cytotoxic and inhibit cell adhesion, migration and

proliferation [61, 75]. SDS and Triton have been shown to be difficult to remove even

after extensive washing. Regarding the toxicity issue, biological enzymes have been

investigated. Trypsin is a natural enzyme with limited toxicity, but has been found to

only achieve partial decellularization and produce severe structural alterations [7, 69].

Although, other authors have shown preservation of the matrix in human and porcine

aortic tissue with Trypsin decellularization [7, 69, 71]. It has typically been enhanced

with EDTA to chelate Ca2+ thus increasing the efficiency of Trypsin. Teebken developed

an acellular vascular xenograft matrix from porcine thoracic aortas for the purpose of









seeding them with human endothelial cells [71]. Little evidence has been given to

correlate results from the seeding in vitro and in vivo repopulation. This group used 0.1%

Trypsin plus EDTA followed by additional biological enzymes, ribonuclease (RNase)

and deoxyribonuclease (DNase). The idea is that these enzymes are capable of removing

cell components as well as cellular antigens, lipids, and to some extent

glycosaminoglycans with limited toxicity. After two 24 hour extractions at 370C in

Trypsin and EDTA interrupted by incubation with RNase and DNase, light microscopy

showed no cell nuclei or intracellular components in cross-sections of the aortic wall.

Immunohistochemical stains for fibroblasts (ca-actin) and endothelial cells (factor VIII-

related antigen, CD31) were negative and the matrix appeared well preserved.

Unfortunately, no mechanical testing or quantitative matrix analysis was performed.

A study with allogeneic valve grafts decellularized with 0.05% Trypsin for 0.5 to

1.5 hrs showed complete loss of cells except in the media [68]. When followed by Triton

for a 24 hour incubation period, the entire matrix was acellular. In both groups, severe

damage to the leaflets was noted with histology. Another study using porcine aortas with

0.1% Trypsin for 24, 48, 72 or 96 hours contrasts these findings [7]. This group found

that cell removal was time-dependent with a majority of the cells removed at 48 hours

with only minimal damage with an increasing microporous structure, as assessed by

SEM. Reider found incomplete cell removal of porcine aortic roots at the same

concentration of Trypsin combined with EDTA with a 48 hour incubation time [80]. In

vitro cell seeding revealed cobblestone morphology on the luminal surface, indicating a

confluent endothelial cell layer while cells in the media attached in a patchy distribution

and starting to synthesize collagen.









Biological Assessment

Rat aortas were grafted as autografts, decellularized autografts, allografts or

decellularized allografts with an SDS protocol [59]. After two months, the grafts were

explanted and examined for structural and cellular changes. All autografts remained

patent while 3 of the 13 allografts thrombosed (78% patency). The untreated autograft

showed an intact media with no intimal thickening but was surrounded by thick

adventitial fibrosis. The decellularized autograft produced a compacted acellular media.

The external elastic lamina was thinner and adventitial fibrosis was visible. Intimal

thickening did occur but did not contain inflammatory cells. The lumen was covered

with endothelial cells. Unlike the autograft, the intima contains cells that stained positive

for ca-actin. Cellular and fibrillar elements were equally distributed over the length of the

graft.

The decellularization of the allograft material did prevent dilation from occurring

and the vessel appeared macroscopically similar to the untreated autografts. The

untreated allografts were slightly dilated throughout the length compared to the

autografts. There was reduced adventitial inflammatory infiltration and the elastin was

preserved. Just as with the treated autograft, the treated allograft had a compacted

acellular media. Although, it was noninflammatory and significantly thinner and richer

in elastin than the untreated allografts. The internal elastic lamina was approximately

90% preserved along the length but the external elastic lamina was thinner and embedded

in adventitial fibrosis. The intima of the treated allograft showed thickening with no

inflammatory cells and a higher smooth muscle cell and collagen density than untreated









allografts. The lumen was covered with endothelial cells with equal distribution of the

extracellular matrix.

The media of the untreated allografts was dislocated and significantly thinner with

fewer smooth muscle cells and less elastin. It contained inflammatory infiltrates. The

internal elastic lamina was fragmented over more than half of its length and the

interlamellar fibers were disrupted. There was intimal thickening containing

inflammatory cells and lymphocytes with leukocytes adhering to the endothelium. The

adventitia contained large cellular and fibrous inflammatory cells.

Wilson used a multistep detergent-enzymatic extraction process with hypotonic and

hypertonic solutions and combining SDS with Triton on canine iliac and carotid arteries

[62]. These grafts were implanted into the carotid and femoral arteries of dogs, foregoing

anticoagulant and antiplatelet therapy. After four years, they found no aneurism

formation, no calcification, an intact elastin network and no degradation of the collagen

in the wall. The luminal surface was completely endothelialized and there was no

evidence of chronic rejection. The cumulative patency rate was 95% compared to 100%

occlusion in three months with synthetic material. In a follow up study, this group

compared autograft canine saphenous veins to decellularized allograft canine carotid

arteries for a CABG model. After 6 months, 4 out of 7 saphenous vein grafts were

widely patent with the remaining three occluded at the 2 week angiogram. The patent

grafts showed substantial endothelial cell coverage on the lumen. For the allografts, 4 of

9 were widely patent with the remaining 5 thrombosed by the first 2 weeks. There was

no inflammation, but minimal cell repopulation and no evidence of endothelialization

except near the anastomoses. The same group at the University of Toronto used canine









iliac and carotid arteries for femoral interposition bypass grafts [76]. At 4 weeks, the

grafts were widely patent with some cells on the lumen consistent with endothelial cells.

The media was found to be acellular.

A study with a proprietary process at LifeNet (Virginia Beach, VA) using allograft

pulmonary artery for a patch reconstruction to repair great vessels in sheep also compared

cryopreserved allograft and cryopreserved decellularized allografts as well as fresh

decellularized allografts [61]. After 20 weeks, all patches incorporated into the great

vessel repairs without aneurism formations or infection. There was no statistical

difference in the explant and implant dimensional ratios. The cryopreserved allografts

became acellular over time and developed a fibrous sheath consisting of layers of

fibroblast cells that line the luminal side and encapsulating the adventitia. This is similar

to the foreign body response seen in surgical implants. The decellularized tissue showed

time-dependent recellularization at 10 weeks and 20 weeks. There was partial

endothelialization along the lumen and cells infiltrating the elastin bands within the wall.

Staining for c-actin indicated that most of the infiltrating cells were biologically active

myofibroblasts. The cryopreserved grafts were positive for calcification while none was

evident in the decellularized tissue.

Grauss compared cellular and acellular valve grafts in the descending aorta of a rat

for 21 days [68]. All cellularized explants showed leaflet deformation with a

considerable decrease in collagen and a slight decrease in elastin. There was a complete

loss of smooth muscle cells in the media of the aortic root accompanied by multifocal

disruption of elastin fibers. On the other hand, all of the decellularized explants showed

leaflet preservation with elastic fibers in the media normally arranged and similar









collagen content when compared to the nontransplanted tissue. The media of the

acellular allograft was also void of smooth muscle cells.

Hilbert evaluated the morphologic characteristics in a small-diameter freeze-dried

decellularized carotid artery of a goat treated with a nonionic detergent solution [70].

This solution contained inhibitors, such as proteoase, reactive oxygen species and other

free radicals. Histological analysis showed complete endothelial removal while

preserving the basement membrane. Additionally, no cells were seen in the intima,

media or adventitia. Although, there were remnants of smooth muscle cells in the media

without nuclei. The extracellular matrix seemed well preserved, but oval to circular

spaces were occasionally noted in the media. The internal elastic lamina remained intact,

as well as layered elastic laminae in the media.

This group evaluated these decellularized freeze-dried grafts as a carotid

interposition graft in goats for 6 to 7 months and compared them to autologous cephalic

vein grafts. All grafts were patent upon explanation, and no significant dimensional

changes or aneurism formations were noted in the allografts. There was marked

variability in the thickness of the neointima within as well as between the allografts.

There was also great variation along the length of the autografts. Although, the

neointimal thickness of the allografts were comparatively thin at 133 234 jtm ( 1 to 640

[tm) compared to a thickness of 249 224 |tm (8 to 800 [tm) for the autografts. The

autografts showed a luminal surface lined with a layer of endothelial cells while the

allograft showed a discontinuous layer. Even in these regions of incomplete

endothelialization, thrombi were not observed. Both types of grafts showed

myofibroblasts, collagen and proteoglycans.









Upon explanation, the allograft appeared relatively acellular in the media. Among

the dense collagen and elastic lamellae, there were focal cellular regions containing

myofibroblasts of the host. The authors believe that the number of host myofibroblasts in

the media may have been significantly limited by the presence of dense collagen bundles

and smooth muscle cell remnants. Host cell migration was most apparent close to the

anastomoses and was rich in proteoglycans. The variable regions of host cell migration

from the neointima in to the media appeared to occur through fenestrations in the internal

elastic lamina. Most cell migration appeared to occur in the adventitia along the length of

the grafts. The adventitia looked healed and repopulated by connective tissue cells.

Histologically, there was no evidence of calcification and inflammatory cells were not

present in the graft wall. Electron microscopy showed calcification of minute remnants

of cell membranes. Ingrowth of host blood vessels was not observed.

Conklin investigated a small-caliber xenograft with a porcine common carotid

artery model [2]. This group followed cell lysis with multiple enzymatic digestions and

detergent washes while agitating. Histology and electron microscopy showed complete

removal of cellular components while the extracellular matrix appears to remain intact.

H&E stain showed no signs of remaining nuclear material in the walls of the vessels and

TEM showed complete removal of cellular material from the medial layers. Histology

showed an intact internal elastic lamina along with elastin lamellae in the media. TEM

showed the basic extracellular microstructure remained intact after processing.

In addition to reducing immune reactions and decreasing thrombogenicity of the

luminal surface by decellularization, this group wanted to ensure that the process would

maintain a graft with similar mechanical properties to the native vessel, high strength and









good handling characteristics. They investigated the compliance and burst strength of the

vessels on a custom-built system that measures the diameter changes while increasing the

intraluminal pressure. The compliance is expressed as the percentage diameter change

per mm Hg change in pressure normalized to the compliance of fresh vessels. The

average compliance was calculated as the slope of the linear regression line in the

physiological pressure range (70 130 mm Hg).

The mechanical analysis was performed on decellularized vessels, decellularized

and heparin-treated vessels, fresh vessels, alcohol preserved vessels, and ePTFE

prosthetic grafts. The ePTFE grafts were the least compliant (0.025% and the

decellularized grafts were slightly more compliant (0.182%) than the fresh vessels

(0.172%). Alcohol caused the tissue to stiffen some (0.160%) while the heparin

treatment produced a much stiffer graft (0.0975%). During the burst testing, none of the

fresh vessels burst within the limit of the pressure transducer (2300 mm Hg). One out of

four of the decellularized vessels burst within the limit at 1654 mm Hg. Although there

may be some strength loss measured in that one vessel, there still is a high safety of

margin over ten times the physiological pressure. Unfortunately, only four samples were

tested in this group. Additionally, there was no difference found in the average suture

retention strength between the fresh and the alcohol and heparin treated specimens. This

only gives preimplantation mechanical result with evidence of some effect. Once

implanted and the cellular and remodeling response occurs, there could be an exaggerated

effect on the samples that may cause minute damage in the vascular wall.

This group examined their decellularized process with heparin cross-linking on a

carotid artery bypass in dogs. Two animals each were euthanized at 24 and 67 days and









the explants were prepared similarly for histology and stained with H&E or

immunohistochemically for a-actin or Factor VIII. At 24 days, fibroblast-like cells

appeared to densely populate the media and there were few endothelial-like cells lining

the lumen. After two months, the dense a-actin staining suggested smooth muscle cells

densely populated the vessel wall and Factor VIII staining confirmed that endothelial

cells lined the lumen.

In a clinical study conducted by Hawkins using the CryoLife process, 14 children

(8.5 7.9 years) received decellularized, cryopreserved allografts, 6 were patch

insertions and 8 with valved pulmonary allografts [67]. These groups were compared to

20 historical control subjects (1.7 2.4 years), 8 with valved and 12 with allograft patch

insertions. There was no attempt to match ABO blood types in either group. The effect

on the immunogenicity was measured at 1, 3 and 12 months by frequency of panel-

reactive HLA alloantibodies (PRA): class I (HLA-A, HLA-B, and HLA-C) and class II

(HLA-DR/DQ). A flow cytometry technique was used to calculate the percentage of

fluorescent positive beads, indicative of the percentage of PRA. Antibody levels were

significantly higher from the preimplantation levels for both classes at all time points.

The antibody levels were significantly lower in the decellularized allograft. There was a

marked reduction in staining for class I and class II histocompatibility antigens.

However, there was no work done in this study to determine whether reduced

immunogenicity will truly allow tissue ingrowth and improved long-term durability in

patients.









Purpose of This Study

This segment of the research was aimed at determining the appropriate processing

conditions to balance decellularization with matrix integrity. The tissue will be treated

with different detergents and enzymes at various temperatures for specified times. The

level of decellularization will be examined histologically and the amount of antigen

containing material will be determined immunohistochemically. The matrix integrity

will be assessed with an enzyme digestion assay for collagen. The aim is to provide

parameters that produce efficient cell and cellular debris removal without damaging the

insoluble scaffold.

Materials and Methods

Cell removal

The effectiveness of cell and cellular debris removal was evaluated using two

detergents and two enzymes at various concentrations (Table 2-1). The detergents used

in this study were Triton-X 100 at 0.25% (v/v) and 1.0 % (v/v) and sodium deoxycholate

(Cholate) at 0.25% (w/v) and 1.0% (w/v). Due to viscosity issues with sodium

deoxycholate at low temperatures, tests were repeated at 0.5% (w/v). One treatment

group combined Triton-X 100 with sodium deoxycholate. Trypsin alone at 0.05% (w/v)

or 1.0% (w/v) and in combination with 0.02% (w/v) EDTA were the enzymes used in the

decellularization process. EDTA was also combined with Triton for a comparison since

the Trypsin samples appeared grossly degraded after treatment. A human saphenous vein

sample 10 mm in length was placed in a solution at 37C or 48C for either 6 and 24

hours while continuously shaking. Samples were then rinsed in PBS at room temperature

for 10 minutes while continuously shaking. Following the rinse step, the samples were

cut in half and either prepared for histology or immunohistochemistry.









Table 2-1. Processing parameters to evaluate decellularization of human saphenous vein.
Chemical Concentrations (%)
Trypsin 0.05 (w/v) 1.0 (w/v)
Trypsin / EDTA 0.05/0.02 1.0/0.02
(w/v) (w/v)
Triton 0.25 (v/v) 1.0 (v/v)
Cholate 0.25 (w/v) 1.0 (w/v)
Triton / Cholate 0.25/0.25 1.0/0.5
Triton / EDTA 0.25 / 0.02 1.0/0.02
Histology

Samples were placed in 10% buffered formalin immediately following the rinse

step and remained there for at least 24 hours before further preparation. Then, the

samples were embedded in paraffin and stained with hematoxylin and eosin (H&E) to

determine cellularity.

Immunohistochemistry

Immunohistochemistry was used to determine the level of antigens remaining in the

tissue after treatment. Only class I MHC antibodies were used for evaluation since class I

antigens are present in a much greater abundance than class II antigens. Samples were

prepared in frozen blocks and sectioned with a cryostat. The sections were thawed and

then rinsed for blocking and antibody staining. Sections were incubated for 30 minutes

in primary antibody (mouse anti-HLA-ABC class I MHC) solution and then for anther 30

minutes in diluted biotinylated secondary antibody (Vector ABC Elite kit) solution. An

enzyme substrate, NovaRed (Vector), was used to stain the antibodies red. The counter

stain was hematoxylin providing a blue contrast. A full protocol can be found in

Appendix B.









Optical evaluation

A Zeiss Axiophot 2 microscope with a motorized Ludl scanning stage was used

with tile field mapping at 5x to capture the entire image in one frame. Particle analysis

was performed in Image J to determine the ratio of antigen staining (red) to the total

tissue sample.

Matrix integrity

One of the major structural components of the extracellular matrix that gives veins

their strength is collagen. Damage to collagen can result in mechanical failures such as

increased compliance or aneurism formation. It can also result in a cellular response due

the exposure of the triple helix when stability is compromised. The amount of collagen

denaturation was assessed to evaluate processing conditions with a quantitative enzyme

digestion assay. Saphenous vein samples (0.2 to 0.25 g) were treated with trypsin, a

serine protease enzyme that is able to digest only those collagen fibers that possess a

break in the helix known as denaturation. Digested and undigested fractions are

separated and hydrolyzed with concentrated hydrochloric acid (HC1) to release free

amino acids from each fraction. Following neutralization of the acid in 1 N NaOH, levels

of hydroxyproline (an amino acid present in high concentrations only in collagen) are

assessed in each fraction by a colorimetric method at a wavelength of 550 nm

(Chloramine T binding, and reduction of the substrate DAB to a colored end product).

The level of denatured collagen in a given sample is then expressed as a percent of the

trypsin soluble fraction to the sum of both trypsin soluble and trypsin insoluble fractions.

Statistics

Data was analyzed with a two-sample student's t-test with an a value of 0.05 using

Minitab Statistical software. A p-value less than a was considered a significant affect.









Results

The red staining as a result of the class I MHC antigens on the endothelial cells and

the smooth muscle cells ranged from a bright red to a dark brown. A blue and green filter

was applied to only allow red to pass. Obvious areas of antigen staining were removed

due to this variation in red hue. Therefore, only a blue filter was applied where a

threshold value was determined when the blue-stained negative control disappeared while

both green and red were allowed to fully pass to capture a greater portion of the antigen

staining. All samples were compared to their negatives for a percentage of red staining.

Figure 2-1 shows a tile mapped image of the a positive control and a negative control.

Figure 2-2 shows the result after a threshold filter was applied. The negative control on

the right is almost completely devoid of any particles, whereas the positive control

removes the background stained tissue and retains the stained antigens.


Figure 2-1. Positive (left) and negative (right) control samples.





















Figure 2-2. Result of the application of a threshold filter that limits the pass of high
intensity blues.

This procedure was applied to all images. There were three samples per treatment

group and the result of each sample was averaged from three sections. The results for the

Trypsin and Trypsin + EDTA treatment groups are shown in Figure 2-3 and the results

for the Triton-X 100 and sodium deoxycholate (cholate) treatment groups are shown in

Figure 2-4 and Figure 2-5. The treatment groups are labeled according to the detergent or

enzyme used followed by the processing conditions. For the Trypsin and Trypsin +

EDTA groups, the concentrations ranged from 0.05% (L) to 1.0% (H), the processing

temperatures were 37C (L), 48C (M) and 55C (H), and the processing times were 6 hrs

(L), 17 hrs (M) and 24 hrs (H). For the Triton and cholate groups, the concentrations

ranged from 0.25% (L) to 1.0% (H), the processing temperatures were 37C (L) and 48C

(H), and the processing times were 6 hrs (L) and 24 hrs (H). The reason the former

groups had additional conditions was due to an error in the incubator temperature and an

experimental error with the time. The samples were run again at the right conditions but

the samples with the errors were retained for comparison.








62



Trypsin and Trypsin/EDTA


4500

4000

3500

3000

2500

2000

1500

1000

500-

000

5 00


Treatment


Figure 2-3. The amount of class I MHC remaining in the tissue when compared to a
negative control for trypsin and trypsin plus EDTA. The treatment group
labels are T (Trypsin) and TE (Trypsin + EDTA) and the conditions are in
order of concentration (L = 0.25%, H = 1.0%), temperature (L = 370C, M
480C, H = 55C) and time (L = 6 hrs, M = 17 hrs, H = 24 hrs).


Triton and Cholate


2000


1000



5 0 0
oo--.-J----_Jr--I--= = =--

Snn = t = U -


Treatment


Figure 2-4. The amount of class I MHC remaining in the tissue when compared to a
negative control for Triton-X 100 and sodium deoxycholate. The treatment
group labels are TR (Triton) and CH (cholate) and the conditions are in order
of concentration (L = 0.25%, H = 1.0%), temperature (L = 370C, H = 480C)
and time (L = 6 hrs, H = 24 hrs).


T I I = = -J -J TT T


. .U . . ..U. . .


- j Lu i- -


1I- u.










Combination of Triton and Cholate


1600 .


2600



400
-2 00 2 --- ---

Treatment
Figure 2-5. The amount of class I MHC remaining in the tissue when compared to a
negative control for Triton-X 100 and sodium deoxycholate combined. The
treatment group labels are TR (Triton), CH (cholate), and TRC (Triton plus
Cholate) and the conditions are in order of concentration (L = 0.25%, H =
1.0%), temperature (L = 37C, H = 48C) and time (L = 6 hrs, H = 24 hrs).

A significant reduction in red staining at 6 hours was only with an increase in

temperature and concentration. Six groups in the Trypsin and Trypsin/EDTA

experiments had a less then 5% red staining when compared to the negative controls.

Treating the tissue with 1.0% Trypsin or 1.0%/0.02% Trypsin/EDTA at a minimum of

48C and 17 hrs significantly reduced the level of antigens. Only at temperatures of 55C

was there a significant decrease in red staining within 6 hrs. The addition ofEDTA has

an inconsistent effect on the removal of antigens from the tissue.

For Triton and Cholate, all but three treatment groups were below 5% and all of

them were below 10% when compared to the controls. Cholate exhibited the least

amount of staining with approximately 100% removal at high concentrations and high

temperatures. There was no statistical difference in samples treated with Triton and

Cholate combined compared to individually (Figure 2-5). The addition of EDTA to

Triton did enhance the removal of cells at 0.25% for 6 hours at 37C. Representative









samples from each group are shown in Figure 2-6 for comparison. Additional

representative samples for each chemical agent at each of the processing conditions are

illustrated in Appendix C-E.








.* -. '.


Figure 2-6. Representative samples of tissue treated with Trypsin (left), Triton-X 100
(center) and sodium deoxycholate (right) and stained for class I MHC
antigens.

The results of the immunohistochemical stain were compared to histology slides

stained for cellularity (Figures 2-7 to 2-10). There is still evidence of cells in the samples

treated with Trypsin and that is correlated with red stained tissue in the

immunohistochemical slides. Figure 2-7 appears to have a reduction in the amount of

cellular material present in the matrix, but the red stained segment shows a majority of

the tissue still contains antigens. The extracellular matrix is preserved in each segment

except with 1.0% Cholate at 480C. There are large holes within the venous wall and

separation between the fibers.












aT W',',t
-4 ts


:-*r,
S ~.'
s'L


* y.9


Figure 2-7. A comparison of the amount of cellularity (blue) visible in histology
compared to the amount of antigens (red) present in the tissue. The samples
were treated with 0.05% Trypsin at 370C for 6 hours.


Figure 2-8. A comparison of the amount of cellularity (blue) visible in histology
compared to the amount of antigens (red) present in the tissue. The samples
were treated with 0.05% Trypsin at 370C for 24 hours.


_


I :













Uh



i :


Figure 2-9. A comparison of the amount of cellularity (blue) visible in histology
compared to the amount of antigens (red) present in the tissue. The samples
were treated with 0.25% Cholate at 370C for 6 hours.





'+ ,'=,


Figure 2-10. A comparison of the amount of cellularity (blue) visible in histology
compared to the amount of antigens (red) present in the tissue. The samples
were treated with 1.0% Cholate at 480C for 6 hours.

The conditions for each chemical that produced a reasonable reduction in red

staining were analyzed quantitatively for their effect on collagen. Triton at 1.0% and

Cholate at 0.25% were processed at 480C for 24 hours under constant agitation. The

samples were immediately placed in an ice bath after the 24 hours and then rinsed in cold

phosphate buffered saline (PBS) to halt the degradation process due to the temperature

increase. The results were compared to the natural degradation of the untreated controls







67



and are shown in Figure 2-11. A ratio of one indicates no difference from the untreated


controls. A significant degradative effect was not seen with Triton, and cholate produced


a more of an effect on the tissue matrix. The Trypsin showed less than half of the


degradation of the untreated controls.


Collagen Degradation Effects of Decellularization at 480C


.0 1 8


14
o 12


08
06
.04
02

0 25% Cholate 1 0% Trypsin 0 25% Triton


Figure 2-11. Results of the collagen degradation assay after decellularization at 480C for
24 hours. The samples were donor matched to reduce the inherent variability
in tissue.




















CHAPTER 3
STERILIZATION

Introduction

Allograft tissue is obtained from cadavers and processed under aseptic conditions

so it cannot be considered a truly sterile conduit. Current processing methods at an

AATB accredited tissue bank using an antimicrobial soak usually discards approximately

27% of donor tissue due to positive first and final cultures from bacterial and fungal

contamination. Additionally, the risk of transmission of viruses and other pathogens can

be reduced but not eliminated through donor screening. Limitations in current

serological tests used to screen tissue donors result in higher rates of bloodborne virus

(BBV) among tissue donors than blood donors. Over the last several years multiple

examples of donor to host disease transmission have been documented. These examples

include transmission of hepatitis C (HCV), bacteria, and fungi, and have resulted in

serious morbidity and even mortality [8]. A method for sterilization and viral

inactivation that is effective in removing and inactivating contaminants and endogenous

materials without adversely affecting the tissue matrix and functionality would increase

availability and provide a safe alternative venous conduit. There would no longer be

concern for the infection window period or errors in testing and false negative results

[83].









Unfortunately, traditional methods of sterilization, such as irradiation and ethylene

oxide (ETO), have deleterious effects on vascular tissue that limit or preclude their use

for allograft implantation. ETO use in hospitals has decreased due to concern for the

toxicity of residuals as well as its classification as a carcinogenic and mutagenic agent

[84, 85]. Ionizing radiation has been found to be effective against bacteria at 1.5 to 2.5

Mrad, but viral inactivation needs approximately 4 Mrad [86, 87]. It has been shown that

a dose greater than 3 Mrad has a deleterious effect on the biomechanical properties of

human patellar tendon [88]. Additionally, soft tissue such as skin resulted in major

structural changes after gamma irradiation [84].

The active surveillance activities and the outbreak investigation carried out by

Centers for Disease Control (CDC) during 2002 highlights the fact that the spectrum of

bacterial pathogens associated with allograft-associated infections include Gram-positive

(S. aureus and Enterococcus) and Gram-negative bacteria, especially those that ferment

lactose [42]. The importance of validating any sterilization technique against a mixture

of pathogens is underscored by the fact that 19% of the allograft-associated infections

ascertained during the active case finding period were polymicrobial. The frequencies

and concentrations of various pathogens in polymicrobial infections remain unknown and

will vary from recipient to recipient, therefore, are not predictable. Based on data in the

CDC investigation, the antimicrobial solution used should be validated against the

following pathogens, at the very minimum: Staphylococcus aureus, Enterococcus

faecalis, Pseudomonas aeruginosa, Escherichia coli, Clostridium species and Bacillus

species [42].









The current common disinfection method is an antimicrobial soak in which the

tissue is aseptically processed with antimicrobial and sometimes antifungal cocktail and

should not be considered sterile [35]. Additionally, the method for testing sterility may be

flawed [45]. Specifically, residual antibiotics from the treatment process have been

shown to inhibit growth in post-treatment surveillance cultures producing false-negative

results leading to cases of septic arthritis, hepatitis C transmission, Streptococcus

pyogenes (GAS), and Clostridium spp [8, 46]. Of note, cases of soft tissue allograft-

associated disease transmission by spore-forming organisms that are resistant to

antimicrobial solutions have been reported within the last few years [8, 47]. Vascular

tissue is vulnerable to organisms (including fungi and viruses) inherent in the donor or

transmitted during procurement or processing that are resistant to the antimicrobials

currently used for allograft disinfection [23, 48, 49].

The fact that allograft-associated infections result from an unpredictable source of

pathogens makes it difficult to validate a sterilization process. One way to overcome a

polymicrobial investigation is to use a representative bacterium that is more difficult to

kill than anything that will be detected on the tissue. The bacterium Bacillus

stearothermophilus is a spore forming microbe that is resistant to a wide range of

temperatures and processing conditions. A 6 log reduction in this spore would guarantee

protection against the bacteria typically found on contaminated vascular tissue.

Disinfection and Sterilization Processes

It is estimated that 60% of all human infections are caused by viruses [89]. Viruses

are composed of DNA and RNA, lipids, and a protein membrane for protection. Their

physical characteristics are classified as either lipophilic or hydrophilic. Lipophilic

viruses have a lipid shell known as an envelope. These viruses are easier to inactivate









with removal of the envelope via a disinfectant. Such viruses include HIV, RSV and

hepatitis B (HBV). Hydrophilic viruses are known as nonenveloped viruses since they do

not have a shell. Instead, they have a very tough protein coat that is difficult for some

disinfectants to enter. Such viruses include poliovirus, rhinovirus and hepatitis A (HAV).

Transmission of viruses through blood and blood products has led to efforts in

inactivation and removal of endogenous materials and possible contaminants. The most

widely used methods are pasteurization, chemical inactivation, irradiation and solvent

extraction of essential lipids. Pasteurization involves destroying or retarding the growth

of bacteria without destroying biological activity of the sample by heating to a moderate

degree (60 70C) for a substantial period of time rather than boiling [90]. This process

is used to disinfect blood products such as Alpha-1 proteinase inhibitor concentrates from

pooled human plasma [91]. Injected samples were maintained at 600C for 10 hours and

then tested for viral inactivation. The thermolabile viruses, vesicular stomatitis virus,

herpes simplex virus-1, visna, and HIV, were inactivated within 1 hour while poliovirus

took almost 5 hours. Porcine parvovirus, one of the most thermo-stable viruses known,

was reduced by 3 logs but there was still a 1.7 log infectivity remaining after 10 hours.

Clinical data 6 months post-infusion were all negative for hepatitis B (HBV), non-A non-

B hepatitis, and HIV.

Solvents and detergents alone or in combination are more effective than

pasteurization, with success in disrupting enveloped viruses but nonenveloped viruses are

unaffected [83]. Triton-X 100 used with the organic solvent tri(n-butyl)phosphate

(TNBP) was shown to inactivate very large quantities of HBV, BCV and HIV in pooled

plasma [83, 92]. The samples were incubated for 4 hours at 300C in 1% TNBP and 1%









Triton-X 100. This process achieved a greater than 6 log reduction in HBV and HIV and

a greater than 5 log reduction in HCV. These results were tested in vitro by injecting

chimpanzees with the plasma concentrates, and the animals were negative for HBV and

HIV.

Alcohols are known effective antiseptic agents used for hand sanitation and

thermolabile instrument disinfection (70%) in hospitals worldwide [93, 94]. The most

widely used alcohol is isopropyl alcohol. In dilutions of 60 to 95% it kills bacteria,

mycobacteria, fungi and large or lipid-containing viruses [93]. However, it is not

effective on hydrophilic viruses. Ethyl alcohol is used to inactivate hydrophilic viruses.

A study compared the effectiveness of 70% alcohol in inactivation of HIV dried on a

surface or in suspension [94]. The remaining 5.5 log titre after drying was completely

inactivated within 1 min. This time was affected by an increase of 100% in protein load

increasing the inactivation to 4 10 min, providing a significant barrier to dried viruses.

Alternatively, the high titres of HIV in suspension were rapidly inactivated independent

of protein load.

Of all microorganisms, spores are the most resistant to antimicrobial treatment.

Spores are small, single-celled reproductive bodies that are highly resistant to heat and

are capable of growing into a new organism. Bacteria form spores and this dormant

phase is a response to adverse environments as a means of protection. Recently, more

attention has been placed on the need to inactivate spore forming bacteria since cross-

infection by Clostridium spp. ( a species found in intestinal contents) has been reported

on various soft tissue grafts that has resulted in graft removal or even death [42, 84].









Oxidative processes involve electrons from a reducing agent (Oxygen) being

transferred to the oxidizing agent. This process is used in sterilizing equipment with

compounds such as hydrogen peroxide, peracetic acid, peroxysulphates, chlorine dioxide

and ozone. Hydrogen peroxide at a concentration of 3% is used as an antiseptic agent

and combined with ultraviolet light to disinfect cartons for food products. Its sporicidal

activity at higher concentrations has been investigated for medical and dental instruments

[95]. Suspensions of 106/ml Bacillus atrophaeus spores were maintained at room

temperature in 7.5% hydrogen peroxide. No growth was detected after a 6 hour

incubation period.

Peracetic acid (PAA CH3C(O)OOH) is an organic oxidant that is highly effective

against a wide variety of bacteria, fungi, viruses and spores [84, 96]. It has been shown

to be an effective antiviral agent for both enveloped and nonenveloped viruses at 0.2 to

0.35%. This chemical rapidly penetrates the microorganisms and interactions result in

the release of oxygen free radicals causing the destruction of enzymes [96, 97]. It has

been used for sterilization of bone, heart valves and small intestine without significant

adverse effects on the morpohology and structure [84, 96]. Another advantage is its low

toxicity since its residues, oxygen, carbon dioxide and water, are natural and harmless.

PAA combined with ethanol was examined for its effectiveness in inactivating a

wide range of microorganisms in allogeneic bone tissue [85, 98]. In one study, nonviable

organisms were detected after 2 and 4 hours incubation periods. There was a greater than

5 log reduction in Staphylococcus aureus, E. faecium, Pseudomonas aeruginosa, B.

subitilis, Clostridium sporogenes, Mycobacterium terrae, and C. albicans [85]. The same

group tested treatment of bone spongiosa cubes injected with three enveloped and three









nonenveloped viruses and then treated with PAA with ethanol [98]. There was more than

a 4 log reduction in all viruses except in one of the nonenveloped viruses, HAV. A

greater than 7 log reduction was achieved in HAV after implementing a delipidating step.

A porcine-derived scaffold made out of small intestine submucosa (SIS) was

treated at room temperature for 5 min to 2 hours with a PAA/ ethanol (0.18% in 4.8%

mixture) solution to inactivate viruses endogenous to porcine [96]. Relevant enveloped,

nonenveloped and model viruses represented different virus families. Enveloped viruses

were inactivated more easily with all viruses inactivated within 30 min. In a later study,

this group examined the retention of endothelial cells seeded onto the matrix after

sterilizing with the PAA/ethanol (0.1%) solution [99]. The proteins present on the matrix

that may contribute to site-specific remodeling are type I collagen, type IV collagen and

fibronectin (FN). These proteins were well preserved and, therefore, retained their ability

to bind cells.

Peracetic acid was compared directly to hydrogen peroxide in a wastewater

medium [97]. PAA has shown good disinfection against enteric bacteria in wastewaters,

but viruses and bacterial spores are more resistant. On the other hand, hydrogen peroxide

is not typically used alone due to its slow disinfection action and low efficiency.

Wastewater-like test medium containing E. coli, Enterococcusfaecalis or Salmonella

enteritidis were treated for 10 min in PAA or hydrogen peroxide. PAA dose of 0.3%

achieved a 2 to 3 log reduction in enteric bacteria while peroxide doses of 0.3 to 15%

achieved below 0.2 log microbial reductions. Enterococcusfaecalis is one of the

organisms recommended for validation of a tissue disinfection process. The higher

reactivity of PAA compared to hydrogen peroxide may be due to its ability to better









penetrate the cell membrane causing disruption as well as its ability to block enzymes and

transport systems in microorganisms. Also, hydrogen peroxide is ineffective due to its

purity. It is highly reactive due to hydroxyl radicals produced when ferrous iron (Fe2+) is

added to the solution (Harbor-Weiss/Fenton reaction) [100]. Additionally, some

organisms may be protected against hydrogen peroxide due to catalase enzyme presence.

Unfortunately, the effectiveness of these oxidants in killing spores has to be balanced by

their effect in damaging the tissue matrix. Collagen is the only protein susceptible to

fragmentation by hydroxyl radicals.

Huang et al. studied the sterilization of human donor skin with PAA [84]. They

immersed skin in 0.1% PAA for 3 hours at room temperature under constant agitation.

The samples retained their dermal structure and components of the basement membrane.

The collagen fibers maintained their normal architecture with fine and wavy elastin fibers

located among them in a normal pattern post implantation. The extracellular components

were only analyzed qualitatively using histology and degradation was not assessed which

can give an indication to the remodeling response. Human bone-patellar tendon-bone

(BPTB) grafts were treated in a similar mixture at room temperature under low pressure

for 4 hours [101]. The tendons were mechanically tested with no significant difference

found in the viscoelastic properties, stiffness or maximum loading properties when

compared to untreated controls. Again, no conclusions can be drawn about biological

healing or remodeling.

Processing and Determination of Survivors

Sterilants can fully kill or remove bacteria and viruses to specified sterility levels

under appropriate conditions. According to the FDA Guidance document for the use of

liquid germicides, sterilization is associated with total absence of viable organisms [102].









The first step in evaluating a sterilization process is the selection of an appropriate

microorganisms) as a biological indicator. Bacillus stearothermophilus is a species of

gram-positive bacteria found in soil, hot springs and spoiled food products. It is not a

common contaminate of tissue products, but it is used as an indicator in instrument

sterilization equipment since it is the hardest to kill spore known as it is highly resistant

to high temperatures.

The effectiveness of a sterilant is defined in terms of decimal reduction time (D-

value). This is the exposure time (t) required under certain conditions to cause one log

reduction (n) (90%) of the initial population. A suspension is expected to follow a

predictable death rate regarding a plot of the amount of survivors over time. The

negative reciprocal of the slope of regression lines of survivor curves gives the D-value

where

t=Dxn
n= log(No /f)

A study using B. stearothermophilus as one of its biological indicator for PAA and

hydrogen peroxide calculated the D-values of vegetative and spore forming bacteria

[103]. PAA was represented by a 1% solution of commercially available Minncare

(0.45% PAA + 2.2% hydrogen peroxide, pH 2.3). An initial spore population of 104 to

105 CFU/ml (colony forming units) was placed in either Minncare or 1.5% to 26.5%

hydrogen peroxide at 250C and removed at regular intervals. For vegetative bacteria

samples were removed every minute while samples were removed every 5 minutes for

spore forming bacteria. The survivors were analyzed on tryptic soy agar (TSA) pour

plates at various dilutions.









The most resistant bacteria against Minncare were B. \iW i/r ithetim,,phihi1, B.

subtilis and E. coli with D-values 2 to 3 times higher than the more sensitive A.

calcoaceticus, E. cloacae and S. aureus. The presence of PAA in Minncare reduced the

D-value of B. subtilis 10 times compared to hydrogen peroxide alone (5.9 min vs. 55.2

min). D-value for B. stearothermophilus was 4.7 min after being treated at a much higher

concentration of hydrogen peroxide (26.5%).

Materials and Methods

Reduction Curves (Survivor Curves) for Spores

Methods

In order to determine the reduction in spores after chemical processing, an effective

and consistent method needed to be developed. Initially, a 107 spores/ml titre of B.

stearothermophilus was used to enumerate the remaining population after treatment on

both TSA media plates and sheep's blood agar (BA) plates. It was determined that

residual hydrogen peroxide inhibited the growth at concentrations 0.6% or higher for

TSA plates and 6% or higher for blood plates (Figure 3-1). It is believed that the catalase

contained in the blood agar plates inactivated the hydrogen peroxide, so TSA plates were

abandoned for further studies. A similar investigation was performed on PAA, but no

inhibition was observed within the processing concentrations (Figure 3-2). Even though

the blood agar plates allowed for dilutions as low as 1:10, the titre used was too low to

detect more than a 3 log reduction in the process limiting the data collection for the

survivor curves. Therefore, the titre used in this study was a 109 CFU/ml solution of B.

stearothermophilus to allow for a 4 log reduction detection limit.

Consistency in replicates and countable plates without moisture damage were a

problem. To ensure consistency, each vial was vortexed before each replicate and all of









the material was removed from the pipette tip onto the plate. Disposable plate spreaders

were used to prevent cross contamination. The moisture needed to be maintained at a

certain level because the plates would dry out at 600C, but if there was too much moisture

the bacteria would streak across the plate and prevent an accurate count. The first

procedure implemented was propping the plates open in a laminar flow hood to allow

excessive moisture to dry. The second method employed was placing the plates in a

beaker with a porous covering to trap some moisture while allowing the excess to

evaporate. This resulted in consistent and evenly spread bacteria on the pour plates.

Hydrogen peroxide

Hydrogen peroxide at a concentration of 3% was used at 500C and a concentration

of 6% was used at temperatures of 40, 45 and 50C to calculate a curve for modeling the

temperature effects on spore kill. A vial with 0.9 ml of germicide was prewarmed in a

water bath and then injected with 0.1 ml of spore solution (108 spores/ml). Samples were

removed and immediately placed in an ice bath at 10, 15, 20 and 30 min. The vial was

well-mixed on a vortex, and then diluted 1:10 in series then 0.1 ml were added to three

media plates. The plates were placed in an incubator between 55C and 60C for 20 to 24

hours.

Peracetic acid

Peracetic acid at concentrations of 0.1%, 0.05% and 0.005% were used only at

40C to calculate a curve for modeling the concentration effects on spore kill. Initial

inactivation rates at 500C were too rapid to detect. A vial with 0.9 ml of germicide was

pre-warmed in a water bath and then injected with 0.1 ml of spore solution (108

spores/ml). Samples were removed and immediately placed in an ice bath at various time

points depending on the concentration. The vial was well-mixed on a vortex, and then










diluted 1:10 in series before plating. The plates were placed in an incubator between

55C and 60C for 20 to 24 hours.

TSA Inhibition on Log plot BA Inhibition Logarithm 5/1005











Figure 3-1. Bacillus stearothermophilus growth inhibition at different dilutions of
hydrogen peroxide on TSA and blood agar plates.

BA Inhibition on Log plot








% PAA
Figure 3-2. Bacillus stearothermophilus growth inhibition at different dilutions of
peracetic acid on blood agar plates.

Spike and recovery method

Before enumerating the population remaining on tissue following chemical

processing, the amount recoverable from the tissue must be determined. The recovery

process developed involves the use of sonication, mechanical shaking and agitation on a

vortex. Veins were cut into 30 mm segments and clamped at one end. The lumen of the

vein was injected with 108 spores in a 0.1 ml inoculum. The other end was clamped and

the veins were hung at 40C for 15 minutes avoiding contact with any surfaces. The

samples were then cut from the clamps and split open over a centrifuge tube containing

39.9 ml of Letheen broth containing Tween to expose the lumen to the recovery process.









A new pair of scissors or a new scalpel blade was used for each vein and the segments

were cut over the broth to catch anything that may not have adhered to the surface of the

tissue.

The tubes were sonicated at room temperature for ten minutes to break the bacteria

from the surfaces. The tubes were then placed on a shaker where they were vigorously

agitated for another 10 min. The Letheen broth containing the tissue sample was then

vortexed for 30 s and diluted in a series of 1:10 dilutions with 0.1 ml of each dilution

plated on blood agar plates in triplicate. The plates were incubated overnight at 600C.

Reduction curves

The procedure for spike recovery was used to collect data for the reduction curves.

Vein segments were inoculated and then treated in one of the following conditions.

Hydrogen peroxide (6%) was analyzed at 400C and 500C while peracetic acid (0.1%) was

only tested at 400C. At each time point, three samples were spiked and treated with one

control for recovery efficiency.

Data Presentation

Typically, a survivor curve is constructed with the initial population at time zero

and the resultant populations at various time points. Since not all data points could be

completed in the same day, some reduction curves were constructed to compare log

reduction in the controls for each experiment over time. This normalized the data to its

own experimental episode.

Matrix Integrity

Achieving the appropriate log reduction must not come at the expense of the

integrity of the extracellular matrix. The amount of collagen denaturation as a result of

chemical treatments was analyzed by a quantitative enzyme digestion assay. The effect









of hydrogen peroxide at 6% at 400C and 50C on collagen was compared to PAA at

0.1% at the same specified temperatures.

Saphenous vein samples (0.2 to 0.25 g) were treated with trypsin, a serine protease

enzyme that is able to digest only those collagen fibers that possess a break in the helix

known as denaturation. Digested and undigested fractions are separated and hydrolyzed

with concentrated hydrochloric acid (HC1) to release free amino acids from each fraction.

Following neutralization of the acid in 1 N NaOH, levels of hydroxyproline (an amino

acid present in high concentrations only in collagen) are assessed in each fraction by a

colorimetric method (Chloramine T binding, and reduction of the substrate DAB to a

colored end product). The level of denatured collagen in a given sample is then

expressed as a percent of the trypsin soluble fraction to the sum of both trypsin soluble

and trypsin insoluble fractions.

Results

Suspension Sterilization

The D-value for 3% peroxide was too high at 500C to be considered for further

analysis. The D-value for 6% peroxide was reduced by approximately 80% by an

increase in temperature from 40C to 50C. A 6 log reduction at 400C would take almost

5 hours where it takes less than an hour at 50C. Results for PAA were only collected at

0.05% PAA or less and at 400C since at higher concentrations and higher temperatures

the spore inactivation was too rapid to detect. A 6 log reduction at 0.05% PAA only

takes 11 minutes and the D-value at 0.005% PAA at 400C is still less than 6% peroxide at

50C.









82





6% Hydrogen Peroxide at 500C

900
y=-0 1013x +8 4211
8 50 R2 = 0 97
S550

800

7 50















750~~~ ~~~~~~~~ = 0----0435x------------------2305-----
| 700



cm 600
-,

5 50

500

450

400
0 5 10 15 20 25 30 35
Time (min)




Figure 3-3. Survivor curve for 6% peroxide at 500C resulting in a D-value of 9.87.



6% Hydrogen Peroxide at 45C


y =0 0435x- 0 2305
R2= 097
25






S15-
"o

S1
0
05



10 20 30 40 50 60

-05
Time (min)




Figure 3-4. Reduction curve for 6% peroxide at 450C resulting in a D-value of 23.0.









83





Log Reduction at 400C


07


06


05


C 04

0
- 03

0
a 02


01


0


-01


Time (min)




Figure 3-5. Reduction curve for 6% peroxide at 400C resulting in a D-value of 47.39.


Predictive Modeling for 6% Peroxide


42 44 46
Temp (C)


48 50 52


Figure 3-6. Linear regression of the log of the D-values for each temperature. This gives

an equation to interpolate and possibly extrapolate the D-values at various

temperatures with 6% hydrogen peroxide.


y = 0 0211x 00553
R2 = 0 9371


















5 10 15 20 25 30


y = -0 0756x + 4 7213
1 60 R2 = 0 9906
1 60

1 40

1 20

1 00

080

n n


0 uu00








84



0.05% PAA at 400C


Time (min)

Figure 3-7. Reduction curve for 0.05% PAA at 400C resulting in a D-value of 1.90.


0.005% PAA at 400C


y= -01551x+80348
R2 = 0 999
800

7 00

6 00

0500

o 400
-j


6
Time (min)


Figure 3-8. Survivor curve for 0.005% PAA at 400C resulting in a D-value of 6.45.


Tissue Sterilization


The recovery method developed in the study consistently achieved between 52 and


73% spike recovery. A control was used at each experimental episode to provide a


correction factor for the treated samples. The recovered population divided by the


correction factor provides the surviving population after treatment. The reduction curves


for spiked tissue treated with 6% hydrogen peroxide at 500C and 400C are shown in











Figure 3-9 and Figure 3-10, respectively. According to the reduction curves, a 6 log


reduction can be achieved in 1 12 hours at 400C and in less than 40 minutes at 50C. It


was noted that the D-values of 6% peroxide with tissue were less than the D-values for


spores treated in suspension. The D-value was decreased by approximately 4 minutes to


5.34.


Human Tissue Treated with 6% Peroxide at 50 C

y= 01508x+ 00885
R2= 0 8099
5


4
.3



2-J





0 5 10 15 20 25 30 35
Time (min)


Figure 3-9. Reduction curve for 6% hydrogen peroxide at 500C resulting in a D-value of
6.63.












Human Tissue Treated with 6% Peroxide at 400C

25
y =0 0659x- 7E-16
R2 = 0 9921

1 5 -------------------^^---------




S1
0



05-



5 10 15 20 25 30

-05
Time (min)


Figure 3-10. Reduction curve for 6% peroxide at 400C resulting in a D-value of 15.17.


Optimum spore kill is achieved at a high concentration of germicide and at a high


temperature. The integrity of the tissue may be compromised at these extreme


conditions. The ratio of degraded collagen after treatment in 6% hydrogen peroxide and


0.1% PAA at 50C for one hour were compared to untreated controls. The results are


shown in Figure 3-11. There was not a statistical difference between peroxide and PAA,


but both showed significantly altered matrix properties. A reduction in temperature to


40C showed no significant difference in the collagen degradation of samples treated with


6% peroxide or 0.05% PAA for one hour compared to the controls (Figure 3-12).








87



Germicidal Effect on Collagen at 50 C


45
0
4
-
o
S35

g 3
S-


3-
o 25

o 2

15


E
S05

0-


6% Peroxide


0 1% Peracetic


Figure 3-11. Collagen degradation of tissue treated with 6% hydrogen peroxide
and 0.1% PAA at 500C. Both 6% hydrogen peroxide and 0.1% PAA had a detrimental
effect on the collagen when compared to a donor matched control.

Ratio


1.4
1.2
1
0.8
0.6
0.4
0.2
0


Control


6% Peroxide 0.05% Peracetic

Treatments


Figure 3-12. Collagen degradation of tissue treated with 6% hydrogen peroxide
and 0.05% PAA at 400C. Both 6% hydrogen peroxide and 0.05% PAA at 400C were
statistically similar to the control of a donor matched control.




















CHAPTER 4
PRESSURE-INDUCED INCREASE IN PORE VOLUME FOR INTRODUCTION OF
DECELLULARIZATION AGENT INTO VEIN ALLOGRAFTS

A thorough review of decellularization methods in the literature is provided in

Chapter 2, detailing commons problems with these methods and its inability to achieve

complete removal of cells and cellular debris as well as the long processing times needed

for successful decellularization. Without complete removal of cells within the vein wall,

an immune response is elicited and in some cases can lead to evidence of calcification or

inhibition of cell migration [1, 5, 25]. Additionally, it is difficult to remove the

decellularization chemicals that become embedded within the matrix which has a residual

toxic effect resulting in cell death and the prevention of cell migration once implanted

[70]. Processing times ranged from 24 hours to several days under physiological

temperatures and continuous agitation. The extensive treatment times and higher

temperatures may damage the vein surfaces since this is where maximum contact occurs.

This may further lead to altered mechanical properties or a biological response due to

exposure of denatured collagen to blood flow. The collagen degradation assay used in

this study measures the effective degradation process providing an indication of the

average damage across the vein wall. Intact collagen will therefore occur throughout the

wall and may balance by denatured collagen on the surfaces, masking the possible