CA2373706A1 - Fibrin sealants providing less inflammatory response and methods using same - Google Patents

Abstract

In accordance with the present invention novel fibrin sealants which provoke less of any inflammatory response than prior art sealants, and methods of using such sealants, are disclosed. It has now been found that fibrin sealants can be prepared in a number of ways to selectively include or exclude components which are presently understood to upregulate or downregulate the body's natural inflammatory response. The novel sealants of this invention are useful in any instance where fibrin sealants have known utility. Nonlimiting examples include use in hemostasis, sealing of fluid or air leaks, delivery of cells, tissue repair and wound healing, drug delivery and particularly useful in the prevention of post surgical adhesion formation.

Description

FIBRIN SEALANTS PROVIDING LESS INFLAMMATORY
RESPONSE AND METHODS USING SAME
Field of the Invention This invention relates to the prevention of post surgical adhesions and, more particularly, concerns enhanced methods and compositions using a fibrin sealant to prevent such adhesions.
Background of the Invention Formation of tissue adhesions between adjacent tissues is an adverse side effect of many surgical procedures. It is believed that following abdominal surgery the incidence of peritoneal adhesion formation may be as high as 900. These post surgical adhesions often lead to pain, discomfort, immobility and, in the case of gynecological surgery, female infertility. Also, post surgical adhesions may even result in life-threatening bowel strangulation is some instances. Thus, there exists a large unmet need for a way to prevent post surgical adhesions, given the potential discomfort and medical risks they pose following routine surgery and the accompanying financial burden of remedying this adverse effect.
A number of pathways have been investigated to reduce or prevent post surgical adhesions. One proposed mechanism involves the reduction of the inflammatory reaction at the wounded site following surgery. In this regard, the use of corticosteroids, NSAIDS, histamine antagonists and calcium channel blockers has bee suggested. Another prevention method recommends the inhibition of coagulation at the wounded site using, e.g., heparin or oral anticoagulants. Still further methods entail the promotion of fibrinolysis through appropriate administration of, for example, fibrinolysin, Streptokinase, Urokinase and t-PA. More recently, barrier materials have been suggested for the prevention of post surgical adhesions. These include, but are not limited to, amniotic membrane, rubber, silver foil, Teflon, dextran, hyaluronic acid, Surgigel~ (regenerated cellulose), Interceed~
(TC7 oxidized, regenerated cellulose), Polaxamer 907 (temperature dependent polymer), Gore-Texo (expanded polytetrafluorethylene) and SepraFilm~ (hyaluronic acid derivative film).
The above materials and methods fall into two basic categories; those which do not work effectively, and those which may provide some efficacy but are limited in either their use or applicability. Accordingly, new efforts have continued in this area.
The role of fibrin sealants in surgical procedures generally has been reviewed for many years. The role of fibrin sealants preventing or reducing of post surgical adhesions has been investigated over the past few years and is not totally understood at this time. Indeed, the literature in this area seems somewhat divided. Several researchers have reported an improvement in post surgical adhesion prevention using fibrin sealants. For example, the incidence of parovarian adhesions was significantly reduced in women subjected to laser vaporization of endometriomas and fibrin sealant (Dormer, J. et al.; J Gynecol Surg 7:163, 1991), incisioned and colonic adhesions formation was reduced in abraded rabbit uterine horns (Chmielewski, G. et al.; The American Surgeon Vol 58 No. 9; 590, 1992), intrabdominal adhesion formation was reduced in rats (deVirgilio, et al.;
Aarch Surg Vol 125; 1378, 1990), epidural scar formation in rats was reduced after laminectomy (Vaquero, J. et al.; Acta Nevrochir (alien) 120; 159-163, 1993) and flexor tendon adhesion formation in rabbits was reduced following a partial laceration of the flexor tendon (Fryknra, E. et al.; Journal of Hand Surgery Vol. 18A, No. 1; 68, 1993).

Others reporting on the use of fibrin sealants to prevent post surgical adhesions have not been so positive. For example, fibrin sealant was found to have no statistically significant effect in preventing perivascular adhesions following arteriotemics of the femoral and carotid arteries of dogs (Dickinson, C. et al., Vascular Surgery; 15, 1993);
fibrin sealant did not prevent adhesion formation in colonic anastomoses in the rat (van der Ham, A., et al.; J Surgical Research, 55; 256-260, 1993) and fibrin sealant did not reduce post surgical adhesions following ovarian reconstruction in the rabbit (Bilgin, T., et al.; Gynecol Obstet Invest 39; 186-187, 1995).
WO 92/22312 to Wadstrom discloses combinations of fibrin sealants and biocompatible polymers reportedly useful for the prevention of post surgical adhesions. WO 92/22312 reports that fibrin sealants alone provide a wound healing effect which results in strong scar formation and does not prevent adhesions. The viscosity enhancing polymers are useful to allow application of the otherwise watery sealant components to vertical surfaces and also prevents adhesions. These polymers are high molecular polyglycors or polysaccharides.
WO 96/22115 discloses a self supporting sheet material of cross-linked fibrin having a particular range of pore sizes to prevent or reduce post surgical adhesions. The use of spray application to prepare these and similar sheets is reported in WO 98/02098. These sheets are described as generally non-adherent and non-hemostatic and are preferably used in conjunction with a known fibrin sealant which does possess adherence and hemostatic capabilities. These sheets are prepared using high concentrations of fibrinogen and high dose thrombin to obtain the desired structure regarding pore size and distribution (preferably <20 Vim, more preferably <5 Vim, most preferably <1 Vim). Also, the sheet is pre-formed outside of the body and needs to be applied as a solid material, i.e., WO 00/72856 CA 02373706 2001-11-13 pCT~S00/15139 not applied as a spray or liquid.
Since fibrin sealants are used increasingly for hemostasis and fluid and air leakage in surgical procedures, it is important that a sealant is used which does not promote adhesion formation. Also, since fibrin is a part of the natural healing process it seems desirable to attempt to use fibrin in resolving the post surgical adhesions problem, as well. It would be preferable to be able to accomplish this without polymer additives, without high concentrations of fibrinogen, without subjecting the patient to high dose thrombin and, preferably, without the need for preformed sheets which must, in turn, be used with standard sealants.
A co-pending U.S. application, 60/136,901, filed concurrently herewith and entitled "Prevention of Post Surgical Adhesions Using a Fibrin Monomer Sealant" discloses that fibrin polymers, which may also be useful as fibrin sealants, are extremely useful in the prevention of post surgical adhesions when the polymer is formed of more chemically-natural materials. This is understood to refer to materials which most closely resemble the body's own clotting materials preferably patient-derived, i.e., autologous. That co-pending application describes such materials as involving one or more of the following:
1) the application a fibrin monomer to form a fibrin polymer;
2) the use of autologous fibrin-polymer-forming materials;
3) the inclusion of autologous plasma proteins;
9) the substantial avoidance of exogenous or added thrombin or other fibrinopeptide A and/or B-cleaving enzyme; and/or 5) the avoidance of solvent/detergent or other potentially denaturing process steps.
It has been found that one, some or all of the above _ q _ WO 00/72856 CA 02373706 2001-11-13 pCT~S00/15139 parameters can reduce the incidence of post surgical adhesions. Although the exact mechanism is not known, it is believed that several, if not each, of these parameters may reduce the inflammatory response which, if overly accelerated, may result in post surgical adhesion formation.
As mentioned above, the use of antiinflammatory agents has been tried to reduce the incidence of post surgical adhesion formation, although the problem still persists.
Fibrin sealants are believed to play a positive role in wound healing but prior art sealants include several factors which may overly accelerate the inflammatory response and cause post surgical adhesion formation.
Brief Description of the Figures Fig. 1 Illustrates a rabbit paw flexor tendon per the experiments of Example 2.
Fig. 2 Is a graph showing variation of dG'/dt with time per Example 2.
Fig. 3 Is a graph illustrating the adhesion experiment of Example 3.
Fig. 4 Is a graph comparison of VivostatT"', Tissucol~
and Beriplast~ kinetic rate constants based on a second order reaction over the initial 200 seconds of polymerisation.
Fig. 5 Is a graph illustrating adhesion strength at first breaking point at various polymerisation times for of VivostatT"', Tissucol~ and Beriplast~.
Fig. 6 Is a graph illustrating elongation at first breaking point at various polymerisation times for VivostatT"~, Tissucol~ and Beriplast~.
Summary of the Invention In accordance with the present invention novel fibrin sealants which provoke less of any inflammatory response than prior art sealants, and methods of using such sealants, are disclosed. It has now been found that fibrin sealants can be prepared in a number of ways to selectively include or exclude components which are presently understood to upregulate or downregulate the body's natural inflammatory response. The novel sealants of this invention are useful in any instance where fibrin sealants have known utility. Nonlimiting examples include use in hemostasis, sealing of fluid or air leaks, delivery of cells, tissue repair and would healing, drug delivery and particularly useful in the prevention of post surgical adhesion formation.
Detailed Description of the Invention The present invention discloses fibrin sealants designed to produce less of an inflammatory response than their prior art counterparts. Accordingly, they reduce the incidence of post surgical adhesion formation while providing excellent hemostasis, adherence and fluid sealing benefits. One key to the inflammatory response is understood to be the production of nitric oxide. The fibrin sealants of the present invention downregulate nitric oxide formation and the inflammatory response, generally, in several ways. Each of these factors are believed to provide reduction in post surgical adhesion formation and the present invention involves methods using one, some or all of these factors.
Accordingly, methods employing one, some or all of these aspects are considered to be a part of the present invention.
As mentioned above, it is reasonable that the fibrin polymer resulting from the present methods is able to reduce or prevent post surgical adhesions over a variety of fibrin concentrations and need not be a high concentration material when compared to the prior art. Further, there is no need to pre-form sheet-like materials ex vivo for application to a surgical wound site and no need for additional sealant layers to be used. The present fibrin sealant, when applied using the methods described herein, possesses not only good barrier qualities to resist post surgical adhesion formation but also does less to induce their formation. At this same time it provides hemostasis, fluid sealing, adherence to the tissue and enhanced cell migration at the wound site but not into the sealant. This enhanced cell migration is believed to provide improved angiogenesis and tissue repair.
Most prior art sealants are fibrin II polymers, formed by the cleavage of fibrinopeptides A and B from fibrinogen. U.S.
5,750,657 describes novel fibrin monomer-based sealants which can be either fibrin I or fibrin II. It has now been found that fibrin I polymer sealants provide unexpectedly improved reduction of post surgical adhesion formation when compared to prior art sealants. Fibrin I, which is the result of the cleavage of fibrinopeptide A from fibrinogen, provides less of an inflammatory response than does fibrin II. It is believed that the fibrin I downregulates nitric oxide formation when compared fibrin II polymers. Nitric oxide induces, and at high levels can overly accelerate, the natural inflammatory and cell migratory responses, as well as the accretion of collagen. These are essential steps to formation of post surgical adhesions. It may explain why the preferred embodiments in WO 98/02098 which utilizes a fibrin II polymer sheet (as evidenced from the combination of fibrinogen and thrombin) must be preformed and cured ex vivo to avoid formation of post surgical adhesions. With the present sealants and methods, the fibrin is applied or sprayed onto the desired site and the resulting fibrin I polymer provides hemostasis and prevents post surgical adhesions. In accordance with the present methods the fibrin I polymer can also slowly convert to fibrin II over time, e.g., over a period of a few hours. Animal studies have now found that post surgical adhesions are present in at least a weak form within 10 minutes following injury and are relatively _ 7 _ WO 00/72856 CA 02373706 2001-11-13 pCT/[JS00/15139 tenacious at 1 hour post-injury. It is also believed that primary fibrinous adhesions form during the first 16 hours post-injury in what can be called a post surgical adhesion modeling stage. Further, a post surgical adhesion remodeling stage occurs from 16 hours onward in which fibrinous adhesions mature into fibrous adhesions. The presence of a fibrin I (or substantially fibrin I) polymer during the first 10 minutes to 2 or 3 hours post-injury provides notable improvement in post surgical adhesion prevention compared to other prevention methods. The preferred fibrin sealant used in the present methods is initially formed as a fibrin I polymer which gradually changes to a fibrin II polymer or a fibrin I/fibrin II mix over the 2-3 hour period post-application. The important feature is that during the initial period when adhesion formation and potential ingrowth into the sealant are understood to occur, the polymer is preferably a fibrin I.
The above-mentioned U.S. 5,750,657 describes methods of making and using fibrin monomer to make fibrin sealant.
Fibrin monomer-based sealants for improved methods of preventing post surgical adhesions are claimed in the co-pending CV0279 referred to earlier. One method of preparing suitable fibrin monomer solutions involves forming a non-crosslinked fibrin I polymer and thereafter solubilizing the polymer. This can be done so as to reduce the level of fibrinopeptide A which would otherwise be expected in a fibrin polymer-forming process, i.e., by decanting the remaining plasma after enzymatic catalysis of FPA cleavage but before solubilization. Fibrinopeptide A can help trigger the natural inflammatory response. While the extent to which fibrinopeptide A induces the inflammatory response may or may not be singularly sufficient to promote the formation of post surgical adhesions, its cumulative effect with other inflammatory inducers in prior art sealants must be considered. The converse of this is that fibrin sealants with _ g _ WO 00/72856 CA 02373706 2001-11-13 pCT/US00/15139 reduced amounts of fibrinopeptide A, i.e., reduced from that which would be expected in the theoretical polymerization of fibrin polymer, are now believed to reduce the inflammatory response and therefore lower the incidence of post surgical adhesion formation. Accordingly, part of the present invention involves fibrin sealants with reduced amounts of fibrinopeptide A and, in particular, methods of using such sealants preferably in a method of reducing the incidence of post surgical adhesion formation. Preferably, the sealant formulation applied to the patient should contain less than 500 of the theoretical FPA expected.
Other downregulators of nitric oxide formation can also be employed in the fibrin sealants of the present invention.
The present inventors now have a process for preparing fibrin monomer solutions wherein ascorbate (an ascorbic acid) is used in the process and at least a portion of the ascorbate ends up in the fibrin polymer sealant. Ascorbate is also a downregulator of nitric oxide, and sealant preparations including ascorbate are part of the present invention.
In the formation of a fibrin polymer to set as a fibrin sealant and/or as a barrier to post surgical adhesion formation, it is now understood that other plasma proteins and factors, e.g., factor Xa, tPa, when bound onto the fibrin polymer can also increase the inflammatory response. However, their mere presence is not believed to induce the inflammatory response if they are not bound into the fibrin polymer. A
major player in binding these and other factors onto the fibrin polymer is fibronectin. Some sealants in the prior art add fibronectin to increase these levels. The natural fibronectin levels in human plasma are between 300 and 400 /.cg/ml. It has now been found that fibrin sealants with reduced fibronectin levels can provide a reduction in the formation of post surgical adhesions. Further, lower levels of fibronectin appear to result in a fibrin polymer with greater resistance to cell penetration. Thus, fibrin sealants with reduced levels of fibronectin are especially useful in methods where the prevention of post surgical adhesions is desired and beneficial.
Further in accordance with the present invention sealants with tPa, preferably not bound into the fibrin polymer, are believed to enhance a more controlled inflammatory response resulting in wound healing/tissue repair, but not in post surgical adhesion formation. Preferably, these levels of tPa are combined with the lower levels of fibronectin, as discussed above, to avoid binding the tPa onto the polymer.
Also related to this is PAI-1 levels which, of course, affect the tPa levels. Preferably, sealants in accordance with the present invention are substantially free of PAI-1.
The present invention provides fibrin sealants which are found to promote wound healing while preventing post surgical adhesions. Given that some upregulation of the inflammatory response is believed necessary for wound healing, but too much upregulation may result in adhesion formation, it is remarkable that a single fibrin sealant product can provide both of these levels.
The sealants of the present invention can also be applied with additives, e.g., hydrocolloids, polymers, drugs and the like.
The fibrin sealants of the present invention may also include one or more additional autologous plasma proteins.
Autologous plasma proteins, preferably selected from those which would otherwise be present in a natural clotting process can be added to the sealant components before or during application to the surgical wound site in accordance with the present methods. Alternatively, they can be coharvested in the process of producing the one or more autologous blood components which will form the fibrin sealant. For example, a process to prepare autologous fibrinogen may result in a natural cocktail including the desired fibrinogen and several other proteins and/or growth factors. Rather than purify the fibrinogen so as to remove the additional autologous materials, they can simply be co-applied as a whole with the fibrinogen. Similarly, autologous materials coharvested in autologous thrombin processes can be employed. Alternatively, and preferably, in a fibrin monomer sealant method many autologous blood proteins are coharvested and may be co-applied with the monomer. GB 97/11927.5 describes such compositions useful as fibrin sealants for hemostasis and sealing. Thus, blood components (preferably autologous), such as fibrin monomer, may include one or more proteins which are added or co-harvested and are selected from the group consisting of prothrombin, factor XIII (activatable), plasminogen, fibronectin, antithrombin III and factor X and which may be autologous. Preferably, compositions useful herein include about 10-30 mg/ml of autologous fibrin monomer, about 10-40 ~g/ml of autologous prothrombin and about 100-200 E,cg/ml of plasminogen. They may further include 5-100 /.cg/ml of activable factor XIII and/or 45-150 /.cg/ml of fibronection and/or 2.0-7.O~cg/ml of factor X and/or 50-200 ~g/ml of antithrombin III, wherein all such components may be autologous to the patient receiving the anti-adhesion treatment.
The present sealants are extremely useful for hemostasis, sealing of other fluid or air leaks, wound healing and prevention of post surgical adhesions.
In preferred embodiments the materials and methods of the present invention are used in confirmation with the teachings of co-pending U.S. Application 60/136,901 and/or 60/136,902.
The present invention will now be described with reference to the following examples.

A fibrin sealant in accordance with the present methods was prepared as described by Edwardson et al. in U.S. 5,750,657 and using a process and apparatus as disclosed by Holm, inter alia, in U.S. 5,741,428, U.S. 5,603,845, U.S. 5,824,230 and U.S. 5,958,253.
Freshly drawn anti coagulated whole blood (120m1 plus 17m1 40 trisodium citrate USP) was centrifugally separated and the resulting plasma (60m1) reacted with biotin-batroxobin for 10 minute at 37°C. The acid soluble fibrin I polymer produced was isolated by centrifugation and dissolved in 3.5 - 5.2m1 0.2M sodium acetate buffer (pH4) containing calcium ions.
Approximately 6m1 of concentrated fibrin I (20 + 2mg/ml) which is stable for several days at -20°C resulted. Trace amounts of biotin-batroxobin were removed by addition of freeze-dried avidin covalently coupled to agarose, which hydrates the fibrin I. Within 5 minutes the biotin-batroxobin:avidin-agarose was removed by filtration and the concentrated fibrin I monomer was transferred to the desired applicator devices for use in the following experiments.
The resulting F1 monomer solution was co-applied with a carbonate/bicarbonate buffer (pHlO) in a ratio of 7:1 (Fl:pHlO).

WO 00/72856 CA 02373706 2001-11-13 pCT/[JS00/15139 This study was constructed to assess the effect of the fibrin monomer sealant of Example 1 (hereinafter "F1 monomer sealant") on post-surgical flexor tendon adhesion formation in a rabbit experimental model. Further subdivision into mobilized and immobilized postoperative groups allowed assessment of any synergy between the F1 monomer sealant and a method well documented to reduce adhesion formation, namely early active mobilization.
Materials and Methods Animal Model and Surgical Procedure 20 Murex Lop rabbits were used for this study. They were of equal sex distribution and the body weight ranged from 2500 to 4500 grams. There was one death prior to end point assessment making the total number of rabbits 19 (9 male, 10 female).
The animals were obtained at least seven days prior to surgery from Murex BioTech Ltd (Dartfort, Kent) to allow for acclimatization. Throughout the study period they were housed in single cages and fed and watered ad libitum. Regular assessment of the animals' general condition and surgical wound were carried out in accordance with the U.K. Home Office "Guide for the Care and the Use of Laboratory Animals" 1996.
The surgical procedure was carried out in a fully equipped operating theatre, which was "Good Laboratory Practice"
compliant. Induction of anaesthesia was by Hypnorm~ (Janseen Copenhagen, Denmark, 0.2m1/kg, im) followed by Diazeparm (Phoenix, 0.5m1/kg, iv). Maintenance was via an anaesthetic mask, which delivered 2o halothane (Zeneca) and oxygen flowing at 21/min. The condition of the anaesthetised animals was continuously monitored using a pulse and oxygen saturation probe. Recovery from anaesthesia was encouraged with pure oxygen delivery via the facemask.
Prior to the start of the procedure the left front paw's flexor aspect was shaved with hair clippers. After anaesthetic induction the operative site was prepared with chlorhexidine in alcohol and iodine in alcohol scrubs. The field was isolated with sterile drapes. With the aid of an operating microscope, the second and fourth digits of the left front paw were longitudinally incised over the base of the proximal phalanx. Blunt dissection in the midline revealed the digital sheath and its tendinous contents. The sheath was then opened between pulleys A2 and A3 (a point corresponding to the middle of the proximal phalanx). Flexor digitorum profundus was exposed (Figure 1). This long flexor tendon of WO 00/72856 CA 02373706 2001-11-13 pCT~S00/15139 digits two and four then received a standa°rd s~urg~cal sii~3ur_y~~.
on its volar aspect with a 15 blade. The injury measured 5 mm in length and exposed the core substance of the tendon. Prior to wound closure the flexor digitorum profundus wounds of digit two and four were either treated with F1 monomer sealant (Figure 3) or received no treatment. Application of F1 monomer sealant (between 0.1 - 0.4 mls) was through a variable fine jet applicator as disclosed in WO 97/20585 and WO
98/20931. The resultant coating was allowed to polymerise in air for 3 minutes and the injured tendon was then returned to the base of the wound. The operated untreated digits also received 3 minutes of air exposure. All second digits were immobilised with the additional surgical procedure of proximal tendon transaction. This was performed through a transverse skin incision just distal to the carpal tunnel. Both flexor digitorum profundus and flexor digitorurn superficialis to digit two were sharply transacted in the palm thus immobilising the digit. All skin incisions were closed with subcuticular interrupted horizontal mattress sutures (4/0 Vicryl (Ethicon)). The wound was then dressed with Cicatrin~
(Wellcome) antimicrobial powder and sprayed with Opsite~
(Smith and nephew). No external dressings were applied. All animals received buprenorphine (0,01-0,05 mg/kg) for postoperative analgesia. After recovery animals were allowed to move about as normal in their cages. At 14 days post surgery, the animals were euthanased using a lethal barbiturate intravascular injection.
Biomechanical Assessment of Adhesion Development Adhesion development was assessed by the use of a tensiometer in all operated groups. In addition the animals unoperated right front paw (digits two and four) were assessed in the same way so as to provide an unoperated control group for comparative analysis. Double blind biomechanical assessment was therefore conducted on 5 groups.
Group 1 Unoperated controls.
Group 2 Immobilised (digit 2) operated and Fl monomer sealant tested.
Group 3 Mobilised (digit 4) operated and F1 monomer sealant tested.
Group 4 Immobilised (digit 2) operated and no further treatment.
Group 5 Mobilised (digit 4) operated and no further treatment.

The tensiometer (NE Holm A/S, Denmark) measured the force in grams required to pull the flexor digitorurn profundus tendon from its sheath. The freshly culled animals' front left and right second and fourth digit were each dissected and the flexor digitorum superficialis and flexor digitorum profundus were transacted proximal and distal to the operative injury site. The proximal dissection culminated in transacting the two tendons approximately 15 mm proximal to the mouth of the digital sheath. The distal dissection culminated in the flexor digitorum profundus tendon being transacted between the A3 and A4 pulley making sure that this was proximal to the insertion of the vincular vessels. The proximal stump of the flexor digitorurn profundus tendon was then transfixed with a silk 2/0-stay suture. With the nail of the relevant digit held rigid in a clamp, the silk tie was then connected to the tensiometer. The force required to pull the tendon free from the sheath was recorded in grams and was indicative of adhesion build-up.
Statistical assessment of the observed data was carried out using a robust regression technique in a Stata Release 6 statistical software package. This analysis accounted for the structure of the data, which consisted of several measurements per animal. This technique specifies that there is inter but not necessarily intra animal observation independence. Robust estimates of the standard errors of regression coefficients were calculated using the Huber/White/sandwich estimator.
This takes into account the potential lack of independence from the same sample. The residual variances were not constant between groups on the original scale of measurement.
They were therefore not normally distributed. The statistical analysis was therefore performed after applying a loglo transformation. Regression analysis obtained estimates of geometric mean tension and 95% confidence intervals per group.
To obtain these values the coefficients and confidence intervals on the logarithmic scale were transformed back into the original scale of measurement (Table 1 and 2). The raw data has been graphed on a box and whisker plot (Sigma Plot version 4.0). The boxes correspond to the interquartile range (the central 500 of the data) with an internal line to mark the median. The mean is represented with a dotted line. The length of the whiskers are 1.5 times the interquartile range.
Values outside the whiskers have been plotted individually.
Results This study used 20 rabbits. One died prior to biomechanical assessment. They were randomly assigned to one of four treatment groups or an unoperated control group. The potential maximum number of observations of the 19 rabbits was 76, four per rabbit. However out of these only 67 were WO 00/72856 CA 02373706 2001-11-13 pCT/US00/15139 suitable for statistical analysis (880) due for example to the tendon snapping in the mobile group, or incomplete distal transaction prior to tensiometer pull. Statistical assessment was therefore performed on the following number of tendons per group. Group 1 (n = 36) Group 2 (n = 7) Group 3 (n = 6) Group 4 (n = 10) and finally Group S (n = 8).
The raw tensiometer pulls mean value for each group is graphically represented in graph 1. Interpretation of the raw data revealed an overall reduction in pull required to remove the tendon from its sheath by 75.60 when comparing the total treated F1 monomer sealant groups. When separated into mobile and immobile the reduction from untreated to treated groups was 79.7% and 76.80 respectively.
As previously stated the group data was not normally distributed. Meaningful interpretation of the raw data required application of a logarithmic scale. Table 1 shows the geometric mean tension values and 95% confidence intervals for the mean per group. Comparison was made between operated groups and unoperated controls using robust regression analysis. This analysis demonstrated no statistical difference when comparing the F1 monomer sealant treated groups (immobilised p=0.42, and mobilised p=0.47) with the normal unoperated control pulls. This implies there was no significant difference in adhesion formation between F1 monomer sealant treated injuries and unoperated controls.
There was however a highly significant difference with comparisons made between the operated untreated groups and the unoperated controls (immobilised p<0.001, and mobilised p<0.001). This indicated that without treatment with F1 monomer sealant, surgical injury produced significant increases in adhesion formation, as evaluated by tensiometer pull in comparison to the unoperated controls. Further analysis of the data shows that between the operated groups with the same type of post operative mobilisation, the addition of F1 monomer sealant makes a significant difference (difference between F1 monomer sealant immobilised and untreated immobilised p=0.03, difference between F1 monomer sealant mobilised and untreated immobilised p=0.03). After performing regression analysis on the loglo of the raw data, percentage change in mean tension from unoperated control side was as follows:
Group 2 Immobilised operated and F1 monomer sealant treated (increased by 72%) Group 3 Mobilised operated F1 monomer sealant treated (increased by 390) WO 00/72856 CA 02373706 2001-11-13 pCT/US00/15139 Group 4 Immobilised operated and no further treatment (increased by 9800) Group 5 Mobilised operated and no further treatment (increased by 9900) From these data it can be concluded that both the mobile and immobile Fl monomer sealant treated groups were not significantly different to unoperated control group (P=0.47 and P=0.42 respectively Table 2). The raw data graph (Graph 1) shows that F1 monomer sealant works in combination with active mobilisation to reduce the force required to pull the tendon from the sheath, indicating a considerable reduction in adhesion generation in treated cases. This is contrasted with the comparison made between the unoperated control group and the operated untreated group. Both the mobile and immobile untreated groups form significantly more adhesion than the unoperated control (P<0.001 respectively (Table 2)). Again the raw data graph does show the benefits of mobilisation in the reduction of adhesion formation (Graph 1).

Q' 200 a~
._ c _o_ N
n~ 100 Group Graph I
Box and Whisker plot of raw data. The boxes correspond to the interquartile range (the central SO% of the data). The solid internal line represents the median value. The dotted horizontal line is the mean value. The length of the whiskers are plotted individually. Points outside these are individually graphed.

Control Treat Imm Treat Mod un~rea~ ~rrun vmucw m~~..

Group Geometric Mean 95~ confidence interval Tension for the mean Unoperated control 1.5 0.9 to 2.9 Vivostat Immobilised 4.7 -1. to 26.3 Vivostat Mobilised 3.2 -0.5 to 12.6 No Treatment Immobilised 45.1 13.0 to 142.0 No Treatment Mobilised 23.4 7.1 to 66.3 Table 1 The geometric mean tension and 95o confidence intervals for the mean per group Unoperated Vivostat Vivostat No Treatment control Immobilised Mobilised Immobilised Vivostat 0.42 Immobilised Vivostat 0.47 0-77 Mobilised No Treatment <0.001 0.03 0.004 Immobilised No Treatment <0.001 0.13 . .

Mobilised Table 2 P Values from the robust regression analysis comparing every pair of group means.

FXAMPT,F 3 The clinical performance of fibrin sealants is influenced by physical properties such as elasticity, tensile strength, and ability to adhere to human tissue. These properties are related to the internal structure of the fibrin sealant that builds as it polymerises. Analysis of the minimum polymerisation time to achieve a functional fibrin clot is clinically important. Instant tissue-fibrin sealant adhesion is desirable to ensure that the fibrin sealant functions on contact and remains at the site of application without being washed away by blood or displaced by movement of the target tissue (e. g., the heart or lungs). The physical characteristics of fibrin sealants are related to the extent of fibrin crosslinking. Determination of the polymerisation rate allows calculation of the minimum time required to produce a functional clot. The adhesion characteristics to vital human tissue and kinetics of polymerisation between 20 and 300 seconds post-application of VivostatT"' Fibrin 1 monomer based sealant have been analysed and compared to those obtained for two conventional fibrin sealants, Tissucol~ and Beriplast~. Mathematical analysis of the experimental data revealed that polymerisation of VivostatT"' sealant followed first order kinetics whereas that of Beriplast~ and Tissucol~
followed second order kinetics. This study demonstrates that VivostatT"' sealant polymerises faster than conventional fibrin sealants.
1. Materials and Methods Solutions were prepared from fresh blood donations as described in Example 1 and used within I hour. Sealant was delivered using the SpraypenT"' applicator in combination with the automated VivostatT"' application unit as disclosed by Holm et al. in WO 97/20585 and WO 98/20931. Conventional fibrin sealants Tissucol~ (Baxter) (1 ml and 2 ml kits) and Beriplast~ (Aventis) (1 ml and 3 ml kits) were prepared and applied according to the manufacturers' instructions.
Tissucol~ and Beriplast~ were applied using the Duploject~ and Pantaject~ applicators, respectively, fitted with a needle or a spray head.
Torsion rheometry experiments were performed using a controlled stress rheometer, Carri-Med CSL 100 auto gap.
Approximately 0.5 ml of fibrin sealant was applied on to the bottom of the rheometer. The top was a 2 cm diameter fine-hatched plate and the sample platform consisted of a 2 cm diameter fine-hatched plate over a Peltier baseplate. The temperature of the rheometer was set at 37°C. Oscillation experiments were performed using a constant'osci'lTatio~ ~tc~rqu-e-of 15 uNm at a frequency of 0.1 Hz for 11 minutes.
Adhesion experiments were performed using a recently described model involving use of vital human tissue (Kjaergaardetal, Eur. Surg. Res. 1999). The human tissue samples were greater saphenous vein grafts left over from coronary artery bypass grafting. To ensure tissue vitality, all samples were kept in physiological saline solution and used within 24 hours of harvesting. The vein graft was split longitudinally and the split graft was cut into 1 cm2 samples that were fixed to the sample holder using Gore-Tex V5 retaining sutures. The two samples were brought into close proximity at an angle of 45°, and 0.1 ml of fibrin sealant was sprayed on the tissue surfaces, where the adventitia was exposed. The two tissue samples were brought into contact without external pressure and were left polymerised until the adhesion experiment began.
Tests were performed using a Nene universal testing machine model M5 at a speed of 10 mm per minute. Adhesion strength defined as force divided by the cross-sectional area of the sample, adhesion energy (i.e., area under the experimental curve as seen in Figure 3), and elongation (i.e., extension reached by the specimen) were calculated for each experiment.
Mean values were calculated at each polymerisation time.
Confidence intervals (i.e., 95o CI) were also calculated for each parameter as 95% CI=1.95~CV/fn, where n is the number of samples analysed and CV=6.60, 8.35, and 9.81% for the adhesion strength, extension and adhesion energy, respectively.

2. Results Sixteen solutions, each prepared from fresh blood donations from different donors, were sprayed on to the rheometer stage, and the rheology of the fibrin clot formed was studied over a period of 11 minutes. The average fibrin I concentration was 22.20 mg/ml + 12.7o(CV). Four samples of Tissucol~ and Beriplast~ were also analysed the rheometer.
Curve fitting analysis of dG'/dt against time revealed that the polymerisation of sealant followed first order kinetics as shown in equation (1) where t is the time, (dG'ldt)o is the value of the derivative at t=0 and k is the kinetic rate constant:
dG' (dG') _ e_kt ( 1 ) dt (dt)o A single curve showing dG'/dt versus time was prepared averaging the values for the sixteen samples at each time point. The values obtained for k and (dG'/dt) o were 1.17-10-2+0. 05 ~ 10-2 s-1 and 3. 17+0. 15 pa s-1, respectively (r=0. 9926;
X~=0.0683).
Analogous mathematical analysis was applied to samples of Tissucol~ and Beriplast~ fibrin sealants. An attempt to fit the experimental data to equation (1) gave a very poor correlation (r=0.8300-0.9400). In this case, dG'/dt followed second order kinetics as shown in equation (2) where t is the time, (dG'ldt)o is the value of the derivative at t=0 and k is the kinetic rate constant in pascals~l.
(dG. ) - 1 (2) (dt ) kt 1 ' (dG') (dt )o Kinetic parameters for Tissucol~ and Beriplast~ were calculated by fitting the average dG'/dt curve to equation (2). Table 3 shows the results obtained following this procedure. For conventional fibrin sealants, the mixing efficiency of the two components of the fibrin sealant had a great influence on the speed of formation of the clot. Spray delivery systems increased the kinetic rate constant by almost 43o for Tissucol~ and by 18% for Tissucol~.
The kinetics of polymerisation of Tissucol~ and BeriplastO
were compared over the first 200 seconds of polymerisation.
Over this period of time and for the sake of comparison, the experimental curve can be fitted using the second order kinetics model represented by equation (2) with a good correlation (r=0.9974). In this case, the kinetic rate constant exhibited a similar absolute value (i.e., k=1.06.10-2_+0.0610-Zpa-1~ to that obtained using equation (1) . Results are summarised in Figure 4. As shown, the second order kinetic rate constant was between 6 and 20 times greater than the other fibrin sealants.
Adhesion experiments were performed on samples of sprayed fibrin sealant in order to compare similar application systems. Most of the samples showed a first breaking point (i.e., intersection of points A and B, see Figure 3) before the maximum force was reached. This point was detected by a sudden reduction in adhesion force (Figure 3) and was visually observed as a partial breaking of the sample. This represents the limit of the elastic character of the sealant and the first failure of the system, which has clear implications from a clinical point of view. Elongation at first breaking point (i.e., elongation at B) and adhesion strength at first breaking point (i.e., force at point A divided by the cross-sectional area) were measured for each sample. Maximum adhesion strength and elongation as well as adhesion energy (i.e., area under the experimental curve) were also calculated. Two identical adhesion experiments were prepared at each time point and the mean value recorded for each sealant sample. Results are summarised in Tables 4-6.
Reached a high degree of elasticity much faster than either Tissucol~ or Beriplast~. This is consistent with the kinetic analysis showing that the polymerisation rate was much faster than that of competitor products. As a result, superior instant adhesion to tissue (Figure 5) as well as better elastic characteristics (Figure 6).

Table 3 Kinetic parameters and regression analysis correlation for Tissucol~ or Beriplast~ fibrin sealants Applicator (dG' /dt)o/pa s-1 k 103/pa-' r XZ
Tissucol~
Needle 14.15+0.43 0.98+0.02 0.9893 3.20 Spray 54.19+4.71 0.84+0.04 0.9934 5.53 Beriplast~
Needle 19.75+1.65 1.11+0.07 0.9739 8.80 Spray 29.22+4.35 0.84+0.04. 0.9615 6.56 Table 4 Adhesion properties + 95o CI for VivostatTM fibrin sealant Time (seconds) Adhesion strength 36.35+3.31 34.83+3.17 37.80+3.44 53.13+4.84 (gcm z) Elongation (mm) 18.36+2.11 15.73+1.81 19.20+1.63 11.56+1.33 Elongation at 1.54+0.18 1.62+0.19 1.90+0.22 1.88+0.22 first break (mm) Adhesion strength 33.65+3.06 39.83+3.17 34.67+3.16 34.67+3.16 at first break (gcm-z) Adhesion energy 3.12+0.42 2.81+0.38 2.97+0.33 2.99+0.40 (mJ) WO 00/72856 CA 02373706 2001-11-13 pCT~S00/15139 Table 5 Adhesion properties + 95% CI for 'Tissucol~ fibrin sealant Time (seconds) 20 40* 120 300 Adhesion strength 24.99+2.68 90.92+5.26 65.39+5.95 81.11+7.38 (gcm 2) Elongation (mm) 9.45+0.51 6.38+1.04 8.19+0.94 11.65+1.34 Elongation at 0.80+0.09 0.29+0.05 0.21+0.02 0.39+0.09 first break (mm) Adhesion strength 21.91+1.95 17.39+3.32 19.37+1.76 27.53+2.51 at first break (gcm-z) Adhesion energy 0.68+0.09 1.65+0.32 2.91+0.33 6.29+0.85 (mJ) * Single experiment WO 00/72856 CA 02373706 2001-11-13 pCT~S00/15139 Table 6 Adhesion properties + 95o CI for Beriplast~ fibr~~in sealant Time (seconds) AdheSlOn Strength 49.95+4.55 93.61+8.52 131.69+11.98 168.35+15.32 ( g cm-' ) Elongation (mm) 6.60-20.76 9.23+0.49 11.89+1.36 18.48+2.13 Elongation at first 0.96+0.05 0.19+0.02 1.17+0.13 1.25+0.19 break (mm) Adhesion strength 22.43+2.09 39.67+3.16 30.59+2.78 75.96+6.87 at first break (cm-2) Adhesion energy (mJ) 1.41+0.19 1.58+0.21 7.53+1.02 7.59+1.03 WO 00/72856 CA 02373706 2001-11-13 pCT~S00/15139 FXAMPT,F 4 The purpose of this Example 4 is to evaluate the ability of three fibrin sealants in the prevention of post surgical adhesion (PSA) in the peritoneal cavity of rabbits following surgical injury to the uterine horn and the opposing ipsilateral peritoneal wall. This example evaluate syringe-applied sealants as follows:
a) the VivostatT"' fibrin I monomer-based fibrin sealant as used in the earlier examples;
b) Tissucol~, a fibrin sealant which was commercially available in Europe through Immuno AG in Austria, and which is a two component sealant system involving the coapplication of pooled human fibrinogen (in an aprotinim-containing solution) and a bovine thrombin component in a calcium chloride solution; and c) Cyroprecipitate (CYRO), a cryoprecipitate concentrated blood component containing fibrinogen from a single donor and co-applied with bovine thrombin.
The Tissucol~ and CRYO were all evaluated with 4 units and 500 units of bovine thrombin.
Model In summary a standard abrasion injury was inflicted on each uterine horn and ipsilateral peritoneal] wall, which would naturally lie in juxtaposition.
The injured areas were then held apposed using positional sutures, placed outside the experimental site, with treated sites separated by a layer of fibrin sealant. Experimental sites were then left in vivo for the length of recovery time dictated by the study protocol. Due to the level of injury induced and injured surfaces being apposed this rabbit uterine horn abrasion model is a severe experimental PSA model, with a "worse case" scenario created.
To ensure that injuries of consistent standard area could be induced, templates, 25 mm x 3 mm, designed to fit the uterine horn or peritoneal wall, were specially constructed.
Abrasions were induced using a scraper with a standard depth (1.5mm) which fitted the exact area of the template. With the additional parameter of using a set number of scrapes for each injury site, injuries were maximally standardised within the limits of biological variation.

Uterine horns were chosen as the experimental site, together with the peritoneal wall, due to their size, structure and location within the peritoneal cavity. Heavy handling of tissues has been shown to induce PSA formation (Boys, 1942;
Connolly & Smith, 1960) and hence excessive tissue handling needed to be avoided. In addition, ex-breeding rabbits were also chosen for these investigations due to their large size facilitating appropriate procedures.
Procedure Premedication was administrated with hypnorm (Fentanyl citrate 0.315mg/ml and fluanisone lOmg/ml. Supplied by Janssen Saunderton, High Wycombe, Buckinghamshire.) (0.2m1/kg body weight intramuscular) to the right gluteous maximus muscle.
Surgical anaesthesia was induced with hypnorm (0.3m1/kg body weight intramuscular) and diazepam (5mg/ml diazepam. Supplied by Phoenix Pharmaceuticals Limited, Gloucester.) (2.5mg/kg body weight intravenous). Full sterile operating procedures were observed. Particular attention was paid to the washing of surgical gloves in sterile water to remove all traces of particulate matter (starch, powder) which might, if transferred to the operative site, induce granuloma or adhesion formation.
Fur was shaved from the surgical area and the area scrubbed with alcoholic chlorhexidine followed by iodine. Laparotomy was performed using cutting diathermy, by a single incision in the midline, from lower liver margin to the level of the iliac fossa, through skin and muscle to peritoneum, which was divided with scissors. Retraction was applied to laparotomy wound edges to allow access to uterine horns.
Experimental areas were selected on the serosal surface of each uterine horn and corresponding internal ipsilateral surfaces of the peritoneum, which would naturally lie in juxtaposition with each other. A standard template, constructed specifically either for the uterine horn or peritoneal wall, was placed on the selected areas and each area abraded with the sharp edge of a scraping tool. A
standard number of 10 scrapes was used to cause homogeneous punctate bleeding but not so far as to cause frank contiguous hemorrhage. Hence a standard, and reproducible injury was produced in each case.
Single throws of Stannius positional suture (Ethibond 6/0 or similar) were placed 5mm outside each end of abraded areas (peritoneal wall and uterine horn). These sutures passed only through the serosa of the peritoneum and corresponding uterine horn. For control groups, sutures were tightened to bring the two juxtapositional abraded areas together and maintain contact between the two areas. Timing was commenced from this point. For treated groups positional sutures were tightened to bring the two abraded areas close together but not in contact and fibrin sealant was then applied to abraded areas.
Sutures were then drawn together to bring the two areas into contact, using as much tension as was needed to create reasonable contact but avoiding tight sutures. Timing commenced from this point.
In groups which sampled up to 30 minutes post injury, rabbits were kept anaesthetised, in the supine position with the laparotomy incision held closed with tissue clamps. In groups which sampled from 1 hour post injury onwards, the laparotomy was closed and reopened after the relevant time period.
Laparotomy wound closure was in two layers: first with 2/0 plain catgut on atraumatic half round needle for the peritoneum and muscle layers, using transplant longflow, overunder, non-interrupted crossed sutures; and secondly with 2/0 Prolene or similar on half round cutting needle using interrupted mattress sutures for the skin. Immediately after closure of the wound, wound dressing was applied on and around the wound.
At 19 days post surgery animals were anesthetized (as for the surgical procedure). Laparotomy was reopened along the original incision with experimental areas identified and macroscopic observations recorded and photographed. The abraded area of the peritonem together with a border of unabraded tissue in excess of 5 mm and the associated uterine horn were resected.
A summary of the experimental groups is in Table 7 below.
Treatment Group Num ber of Mean Volume Mean Fibrin n AnimalsExperimentalOf Sealant I COTIC

sites Applied (ml) (mg/ml) ~ 13 N/A N/A

Control 6 12 0.930.06 14.902.41 Novel Fibrin Sealant (Needle Application) 11 50-28.75 l 17 TissuCOl~ 6 : .
Tota 4 Units Bovine Thrombin 5 0.500.03 - 500 Units Bovine Thrombin 6 1.120.11 6 Total: 10.64 CRYO 6 0.470.02 0.84 - 4 Units Bovine Thrombin - 500 Units Bovine Thrombin 6 0.330.03 Results Macroscopic Results All of the control groups (13/13) showed PSAs joining the abraded areas of the peritoneal wall and uterine horn. The macroscopic incidence of experimentally induced PSAs for treatment groups was 8.3o for VivostatT"' sealant, 16.70 CRYO 9 units thrombin, 66.70 CRYO 500 units thrombin, 1000 Tissucol~
9 units thrombin and 50o Tissucol~ 500 units thrombin as shown in Graph 2, below.
Graph 2. Macroscopic PSA incidence.

Microscopic Quantitative Results Mean volumes ef PSAs were 73.93mm3 for the control group, 1130mm3 for the VivostatT"' sealant treated group, 21.06 and 179.98mm3 for the CRYO treated groups with 9 and 500 units thrombin respectively and 106.12 and 69.92mm3 for the Tissucol~
treated groups with 9 and 500 units thrombin respectively (Graph 3). Significant reduction in mean PSA volume compared to controls was seen with VivostatT"' sealant (P<0.001) and CRYO
(P=0.0152) 9 units thrombin treated groups only.
Assuming that the control group demonstrated 1000 PSA
formation, the percentage volume of PSAs for each treatment was 15.280 VivostatT"' sealant, 28.99° CRYO (9 units thrombin), 293.960 CRYO (500 units thrombin), 193.550 Tissucol~ (9 units thrombin) and 93.900 Tissucol~ (500 units thrombin) (Graph 9).
Graph 3. t9ean PSA volumes together with SEMs and statistical differences.

Graph 4. Mean PSA percentage volume.
Thus, the Vivostat fibrin I monomer based sealant demonstrated a superior prevention of PSAs in this model by providing a nearly 85o reduction in PSAs over the control and a significant improvement over the other sealants tested.
Table 8 Post Surgical Incidence/Reduction atment % PSA
T

re Incidence Reduction VIVOSTAT 15.28 84.72 CRYO + 9 units 28.99 71.51 CRYO + 500 units 243.46 -193.96 Tissucol + 9 units 143.55 -43.55 Tissucol + 500 units 93.90 6.10 _ 13 _ FY11MPT.F 5 The aim of this example was to assess the ability of VivostatTM
Fibrin Monomer-Based Fibrin Sealant made from human blood (prepared as in Example 1) to reduce or prevent post surgical adhesions in a rat caecal abrasion model.
Experimental Procedure Sixteen, female, Sprague Dawley rats were randomised into two groups. Each group received a standardized caecal and opposing peritoneal abrasion wounding which was either left untreated or sprayed with VivostatT"~ Fibrin Sealant. The abrasions were then apposed with sutures. The animals were then allowed to recover and maintained in the animal unit for 14 days.
On day 19, the control and VivostatT"' Fibrin Sealant treated animals were euthanased and the experimental sites removed.
The wounds were examined grossly, histopathologically and stereologically.
Results The abrasion procedures resulted in an inadvertent perforation of the caecum in two animals (which was repaired using a purse string suture of 4/0 vicryl (Ethicon UK). The sites were still used for evaluation. All other surgical procedures were completed uneventfully.
There were no abnormal clinical signs observed and no mortalities.
The fibrin I solution concentrations used on the treated wounds ranged from 13.46 to 16.02 mg/ml. A mean volume of 0.79m1 VivostatT"' Fibrin Sealant was applied via the spray application to experimental sites.
Macroscopic examination of the wounds on day 19 showed adhesions in 8/8 control wounds and 0/8 VivostatT"" Fibrin Sealant treated wounds. All adhesions were dense, tenacious and fibrous.
Mean volume of post surgical adhesions measured stereologically was 89.91mm3 for control cases and 17.82mm3 for VivostatT"' Fibrin Sealant treated cases. There were fibrinous like connections, which were not considered to be adhesions, in the VivostatT"' Fibrin Sealant treated cases, but they were recorded as adhesion volumes for the purpose of analysis.

Conclusion The mean volume of adhesions was significantly less (P< 0.1) in the VivostatT"' Fibrin Sealant treated group compared to the control group. VivostatT"' Fibrin Sealant is an effective agent for the reduction of post-surgical adhesions when compared to controls in this rat caecal abrasion model.

FXAMPT.F F, This example is designed to evaluate the formation/prevention of PSAs in the stomach, colon and caecum of the pig by introducing a surgical injury to those sites similar to the rabbit uterine from model above. Pig models previously used to investigate PSAs demonstrate that pigs from PSAs in response to injury or trauma by identical pathogenesis to humans. In this example control (untreated/injured) animals were compared to those treated with of VivostatT"' Fibrin Monomer-Based Fibrin Sealant as prepared in Example 1, but wherein the sources were (a) human and (b) from the pig being treated, i.e., autologous sealant.
Pre-medication was administered using Ketamine (Ketalar-50mg/ml ketamine hydrochloride. Supplied by Parke-Davis, Pontypool, Gwent.)(5mgs/Kg) plus Xylazine (Rompun 2%-Xylazine hydrochloride 23.32mg/ml (equivalent to 20mg/ml xylazine) and 1Mg/ml methyl 9-hydroxy-benzoate (preservative). Supplied by Bayer Plc., Animal Health Business Group, Bury St. Edmunds, Suffolk.)(lmg/kg)) intramuscularly in the gluteous maximus muscle. At the operating suite, pigs were induced to and maintained at full anaesthesia with Halothane at 4o in oxygen and nitrous oxide delivered at 1.5 litres per minute and 0.5 litres respectively per minute via mask to the snout.
Once anaesthetised pigs were transferred to lay in the supine position on the operating table, where they were secured by soft tapes to each limb. A plastic ear tag bearing the unique pig identification number for the study was secured to one of the animals' ears. For each pig, the surgical site was shaved and scrubbed with chlorhexidine in alcohol followed by swabbing twice with iodine in alcohol. Sterile operative procedures were adhered to from this point on.
The animals were draped. Cutting diathermy and scissors were used to expose the spiral colon and the stomach through a midline laparotomy starting at the level of the distal xyphoid process of the sternum and extending distally 10 - l2cms.
One area was selected on the lateral surface of the spiral colon or on the blunt end of the caecum and the medio-lateral serosal aspect of the stomach such that these areas each laid naturally against the ipsilateral peritoneal wall. Areas were then selected on the ipsilateral peritoneal wall. As with the rabbit uterine horn abrasion model, each selected area was abraded, using a template and scraper, a standardised number of 12 scrapes which caused homogeneous punctate bleeding but not so far as to cause frank contiguous haemorrhage was performed.

A single throe: suture (Ethibond 0.6 or similar) was placed at each end of t~:G experimental site but outside of the abraded areas. These sutures passed only through the internal serosa of the peritoneum and then passed through the serosa at the respective ends of the apposed abraded areas on the stomach or colon. Sutur a mere tightened to bring the two apposed abraded areas -lose together but not in contact. For treated cases, human er porcine fibrin sealant was applied to the abraded areas (mean volume of 1.3m1 per experimental site), whereas no treatment was applied in control cases. Positional sutures were then tightened to bring the two abraded areas into close contact.
The laparotomy was closed in two layers, first with 2/0 Dexon Plus on an atraumatic half round needle for peritoneum and muscle layers, second with 2/0 Prolene or similar on half round cutting needle for skin. The first suture layer was of "transplant" iongflow, over-under, non-interrupted crossed sutures, the second layer was of interrupted cruciate mattress sutures.
At 7 days post-surgery the animals were scarified and evaluated as follows. Animals were anaesthetised as for the surgical procedure. Laparotomy was reopened along the original incision, with experimental areas identified and macroscopic observations recorded and photographed.
The colon, caecum and stomach experimental sites, joined to the peritoneum by positional sutures at each end of the site, were resected. Euthanasia was achieved by high dose intravenous pentabarbitone (Expiral-Pentobarbitone sodium BP
200mg/ml. Supplied by Sanofi Animal Health Ltd., Watford, Hertfordshire.) (150mg/kg body weight)).
Resected tissues were trimmed of excess adipose tissue, pinned flat on stiff card, to retain a standard, lifelike tissue positional relationship, and immersion fixed in loo neutral buffered formal saline for at least 24 hours at room temperature.
The table below summarizes the experiment.

Treatment Number Injury Number of Volume of Fibrin I
Group of Site Experimental Fibrin Concn Animals Sites sealant (mg/ml) Applied (ml ) Control 3 Stomach 3 N/A N/A

Colon 4 Caecum 2 Total: 9 Human 3 Stomach 2 1.560.06 18.012.29 fibrin Colon 4 sealant Caecum 2 Total: 8 Porcine 2 Stomach 1 1.250.05 19.952.93 fibrin Colon 2 sealant Caecum 2 Total: 5 Results All animals demonstrated general PSAs between the peritoneal suture line and underlying tissues: most frequently the omentum, ileum, caecum, spleen and liver respectively. All these PSAs were separated by blunt dissection and did not interfere with experimental sites.
All control experimental sites were adhered, many being quite severe. 5 out of the 8 sites treated with human fibrin sealant were clear of adhesions, with the remaining sites joined by either PSAs or fibrin sealant. 2 out of the 4 porcine fibrin sealant treated sites were identified as non-adhered with the other sites joined by either PSAs or fibrin sealant.
Mean volume of PSAs for the control group was 207.25mm3 for the colon, 31.33mm3 for the caecurn and 298.11 mm3 for the stomach with an overall mean of 181.78mm3. Human fibrin sealant treated group demonstrated mean volumes of 42.83mm3, 8.96mm3 and 209.47mm3 for the colon, caecum and stomach, respectively, with an overall volume of 74.77mm3. Whereas no PSAs were present in colon and caecum (Figure 8.5) experimental sites treated with porcine fibrin sealant, with 6.25mm3 for the stomach and a group mean of 1.27mm3 (Graph 5).
Assuming that the control group demonstrated 1000 volume of PSA, reduction with human fibrin sealant was 79, 71 and 18%
for the colon, caecum and stomach, respectively, with an overall reduction of 590. Porcine fibrin sealant demonstrated 1000 reduction for both colon and caecum, 97o for the stomach and overall 99% reduction of PSAs compared to controls (Graph 6) .
- 3a -___.- --.---, ._-- _ __-___-.__ , r--I
I
300 i E
E 2~ i v E

I
I
i Cobra Caecum Stanaoh Overall Mean Experimental Site Graph 5 .Mean PSA volumes together with SEMs.

Ho so v~ ~o eo >' ~' S0 0 Gaecvm aiomacn vvera~ mean cot«, Experimental Slte Graph 6 Mean PSA percentages.

C . ~~
f1 wrr~n FbmS..lant p Pasin. Fibrin Swlant

Claims (13)

Claims

1. A method of preventing post surgical adhesions at a surgical site using fibrin sealant wherein said fibrin sealant is prepared so as to induce little or no inflammatory response at said surgical site.

2. The method of claim 1 wherein said fibrin sealant downregulates nitric oxide formation.

3. The method of claim 1 wherein at least a portion of said sealant is a fibrin I polymer.

4. The method of claim 3 wherein said sealant forms a fibrin I polymer upon application to said site and wherein part or all of said fibrin I is thereafter slowly converted fribrin II over time.

5. The method of claim 1 wherein said sealant contains ascorbate sufficient to downregulate nitric oxide formation.

6. The method of claim 1 wherein said fibrin sealant has reduced levels of fibronectin.

7. A fibrin sealant which downregulates nitric oxide formation.

8. A fibrin sealant which contains agents added to downregulate nitric oxide.

9. A fibrin sealant which reduces the inflammatory response.

10. A fibrin sealant being a fibrin polymer, at lest a portion of which is fibrin I polymer.

11. A fibrin sealant which includes less than a theoretical amount of fibrinopeptide A.

12. A fibrin sealant which contains ascorbate in an amount sufficient to downregulate nitric oxide.

13. A fibrin sealant which includes reduced levels of fibronectin.