Introduction

The present invention relates to a tissue sealant for use in the formation of an anastomosis in the gastrointestinal tract, a method for treating cancer in the gastrointestinal tract, and the use of fibrin and one or more human or recombinant growth factor (s) for the manufacture of a tissue sealant.

Background of the invention

Gastrointestinal resections, specifically colon resections, are frequently performed due to a diseased segment, either benign or malignant. The remaining bowel ends need to be reconnected by an anastomosis and heal to restore bowel function. Proper wound healing is an integral part of the patients' recovery and is especially important for those critically ill. Disruption of its course may result in a delayed healing process, dehiscence, infection and death. Generally, the healing process of acute wounds involves a complex, dynamic, well-orchestrated series of events ¹ . Colonic anastomoses tend to heal slower and are usually accompanied by a higher rate of complications in comparison to other segments of the gastrointestinal tract ^2-3 . Colonic anastomoses are usually created either with staplers or are hand-sewn ^4-5 , both techniques were shown to evoke an inflammatory response due to a foreign body reaction at the site of the anastomosis ² _' ⁶ _' ⁷ that further compromises healing and may contribute to post-operative complications ^8-9 . The most feared complication after colonic resection remains anastomotic leakage with an incidence of up to 25% ^10~12 . Leakage is a significant contributor to the overall morbidity and mortality from colon resection. Additional complications following bowel resections include ileus, bleeding and strictures. While strictures are usually a late complication, ileus and bleeding often occur within the first week following the operation ^13-16 . The colon, like most of the gastrointestinal tract consists of four layers: mucosa, submucosa, muscularis propria, and serosa. The healing process following a surgical resection is initiated by the full-thickness injury to the intestine and involves a series of events including three well-defined phases: hemostasis & inflammation (0-3 days), proliferation (3-14 days), and remodeling & scar maturation (day 15 and onwards) ¹ _' ³ . The submucosa provides the intestine with sufficient structural integrity and strength to protect against intraluminal pressures ¹⁷ _' ¹⁸ . The mucosa is repaired by migration and hyperplasia of epithelial cells, which cover the granulation tissue of the wound and thus seal the defect and create a barrier to the luminal contents ¹⁹ . This sealing can be complete in as little as three days if the layers of the bowel wall are directly apposed.

In the prior art, several suggestions exist for reconnecting the bowel ends after resection of a part of the intestine. US2015351730A discloses devices for sealing two bowels stumps together on the external circumference using a sealing cuff to ensure that the bowel stumps remain in position. A sealant can be applied on the external part of the intestine between the intestine and the sealing cuff.

Also, devices that use a medicament in conjunction with staples are known, see e.g. US2017056566A, which discloses an implantable adjunct that can be configured to be applied to tissue by a surgical stapler in conjunction with staples. The adjunct can have one or more medicaments releasably retained therein. Examples of medicaments include an antimicrobial agent, an antifungal agent, an antiviral agent, an antiinflammatory agent, a growth factor, an analgesic, an anticancer agent, an anti-adhesion agent, a tissue matrix degradation inhibitor, a nutrient, an oxygen expelling agent, an amino acid, glutamine, insulin, butyrate, and dextran.

US2015115014A discloses circular anastomosis stapler kits that deliver a therapeutic agent followed by stapling of tissue. The therapeutic agent can be a drug, an enzyme, a growth factor, a peptide, a protein, a nutrient, an excipient, a cell, or combinations thereof. US2003065345A shows a medical device which may be coated to minimize or substantially eliminate a biological organism's reaction to the introduction of the medical device to the organism. The coating may be a variety of therapeutic drugs, agents or compounds .

Fresno L et al discloses in The Veterinary Journal 185 (2010) 32-327 intestinal anastomosis procedures that uses platelet- rich plasma (PRP) . The platelets were harvested by centrifugation using a two-step centrifugation procedure. After the final centrifugation, the platelet pellet was resuspended in plasma. Thus, the content of fibrin was on the same level as in normal plasma, i.e. 3-4 mg/ml. After activation with calcium chloride, the intestinal edges were immerged in the activated PRP for 10 min until the PRP had completely gelled. An end-to-end anastomosis of the intestinal edges was then performed using sutures. The study reaches the conclusion that PRP appeared to increase granulation tissue and fibrosis, but did not influence anastomotic breaking strength. Yol, S. et al in Journal of Surgical Research 146, 190-194 (2008) also investigate the effect of PRP on tissue maturation and wound healing in colonic anastomosis. The bursting pressure was measured day 7 after operation and compared to a control group. The results showed a statistically significant difference between the control group and the PRP group.

While methods exist in the prior art for performing intestine resection there is still room for improvement, notably in the resection of colon in the treatment of colonic cancer, also known as colorectal cancer or bowel cancer. The present invention is suggesting a new strategy, which has the potential to obtain rapid anastomotic healing and lower amount and severity of complications after resection. A main complication to be addressed in the present invention is the anastomotic leakage. While leakage is undesired for all parts of the intestine it is particularly risky for the patient when the anastomotic leakage results from colon resection due to the high prevalence of infectious microorganisms. Summary of the invention

The present invention relates to a tissue sealant for use in the formation of an anastomosis in the gastrointestinal tract, wherein the tissue sealant comprises 8 mg/ml or more fibrin and one or more human or recombinant growth factor (s) .

In a certain embodiment, the invention relates to that gastrointestinal tract stump ends are formed by resecting a segment of the gastrointestinal tract, the tissue sealant is applied to one or both gastrointestinal stump end(s), and the gastrointestinal stump ends are connected to form the anastomosis in the gastrointestinal tract.

Contrary to the majority of procedures of the prior art, that applies the tissue sealant to the exterior parts of the gastrointestinal tract, the present invention suggests applying the tissue sealant to one or both gastrointestinal stump end(s) to assist in the connection of the gastrointestinal stump ends when the anastomosis is formed. In addition, the present invention applies a concentration of fibrin in the tissue sealant, which is significantly higher than the concentration in plasma. It is currently believed that the higher concentration of fibrin provides a sustained release matrix for accommodating growth factors and/or platelets. As the healing progresses, the fibrin matrix will be degraded by the digestion of naturally occurring enzymes like plasmin, thereby liberating platelets and activating the growth factors to the anastomosis area. The sustained release ensures that growth factors are provided to the anastomosis area during an extended time period of the healing process. The platelet-rich plasma (PRP) used in the prior art as the sealant releases the platelets and the derived growth factors relatively fast. It has been estimated that the platelets deliver about 90% of their pay load of platelet derived growth factors within 60 min. Swift M J, et al shows that the 80% release threshold of the platelet derived growth factors bFGF and TGF-βΙ was passed after 2 hours.

The presence of a higher fibrin concentration in the activated product also results in a highly viscous and sticky sealant, that remains on the stump ends after application without draining off. This effect is notably pronounced when the polymerisation of the sealant occurs simultaneous with the application to the stump end(s) . The viscous and sticky sealant provides for an effective initial attachment of the stump ends.

Test results reported herein surprisingly shows that application of the fibrin sealant of the invention resulted in a higher bursting pressure. Furthermore, sealant-treated animals showed a greater degree of mucin production in the early phases. This was accompanied by a more mature collagen phenotype in sealant-treated animals with a greater degree of early angiogenic response along with infiltration of a higher density of immunomodulatory M2 macrophages. These findings suggest a more rapid anastomotic healing.

In addition, the application of a tissue sealant to the stump ends results in fewer and/or less severe complications. Thus, at the end of the period of the experiments reported in the examples, no adhesion to the abdominal wall was observed. To obtain a reinforced sealing, the tissue sealant may further be applied around the anastomosis in a distance above and below the anastomotic line, such as 0.5 cm to 5 cm, suitably 1 to 3 cm above and/or below the anastomotic line.

Various reasons for surgical removal of a segment of the gastrointestinal tract exist. In one embodiment of the invention the resected segment of the gastrointestinal tract is diseased or expected to be diseased. Examples of diseases include the presence of benign or malignant cancer cells. In a certain embodiment of the invention gastrointestinal resection is performed to treat colorectal cancer.

Other diseases which may be treated by resecting a segment of the gastrointestinal tract, include inflammatory conditions of the gastrointestinal tract, such as inflammatory bowel disease, colorectal polyps, colonic diverticular disease, Crohn's disease and ulcerative colitis. Crohn's disease may affect the small intestine and large intestine, the mouth, esophagus, stomach and the anus, whereas ulcerative colitis primarily affects the colon and the rectum.

In another embodiment of the invention the resected segment of the gastrointestinal tract includes traumatized tissue, e.g. caused by blunt or penetrating trauma. Specific examples include accidents, combat injuries, or criminal acts. In a further embodiment of the invention, the resected segment is healthy, and the segment is removed due to an obesity surgery, such as bariatric surgery. Bariatric surgery may be performed by resecting and rerouting the small intestine to the small stomach pouch, i.e. gastric bypass surgery. Other types of bariatric surgery include biliopancreatic diversion, such as a duodenal switch in which the part of the stomach along the greater curve is resected.

In a further embodiment, the present invention is applied in the treatment of intussusception. Intussusception is a medical condition in which a part of the intestine folds into the section next to it. It typically involves the small intestine and less commonly the large intestine. Specific examples of intussusception include entering of the ileum into the cecum, or the prolapsing of ileum or jejunum into itself. Thus, according to the present invention the resected segment includes the intussusception.

Various types of tissue sealants, also called tissue glues or surgical tissue adhesives, useful according to the present invention are commercially available and include fibrin sealants, cyanoacrylate sealants, gelatin-based sealants, polyurethan-based sealants, PEG-based sealants, polyester- baser based sealants, polyvinylalcohol-based sealants, oxidized regenerated methylcellulose sealants, glutaraldehyde-based adhesives, and human fibrinogen and thrombin fleece. In a preferred embodiment of the present invention, the tissue sealant is a fibrin sealant, i.e. a tissue sealant comprising fibrin. A Fibrin sealant has the advantage of participating in the hemostasis. The fibrin sealant is generally derived from mammal blood to form bovine fibrin, human fibrin, equine fibrin, porcine fibrin, etc. Human fibrin is preferably used to increase the biocompatibility . In a preferred aspect, the fibrin sealant is derived from the patient's own blood, i.e. the fibrin is autologous. The risk of mammal-borne contaminants is eliminated, which will protect the patient and the surgeon against e.g. viral diseases not yet identified. Furthermore, the use of autologous fibrin further increases the biocompatibility and reduces the risk of viral infections.

In a preferred embodiment, the tissue sealant contains one or more human or recombinant growth factor (s) . The growth factors are capable of stimulating cell growth, proliferation, healing and cellular differentiation of the gastrointestinal stump ends, thereby promoting the healing process. The growth factors may emanate from a biological or synthetic origin. In a preferred embodiment, the human growth factors are derived from platelets.

In a preferred embodiment, the tissue sealant is enriched with platelets, also called thrombocytes. The function of platelets is to help stop the bleeding by clumping and clotting blood vessel injuries. When forming an anastomosis, the platelets present in the tissue sealant will increase the healing activity at the intestine stump ends. The patient's own circulating platelets may also be attracted to the anastomosis site and assist the platelets of the tissue sealant in the hemostasis of the anastomosis. In a preferred aspect of the invention the human growth factors at least partly emanate from the platelets of the tissue sealant.

The platelets may be derived from a human blood source. In a preferred embodiment, the platelets of the tissue sealant are derived from the patient's own blood. It is generally believed that fewer complications are experienced with autologous platelets due to the absent or reduced risk of immunological reactions. In a certain embodiment the concentration of platelets in the tissue sealant is at least twice the concentration of platelets in the patient's own blood. The increased amount of platelets in the fibrin matrix provides sufficient growth factors during the sustained release period. In an embodiment of the invention the concentration of platelets in the tissue sealant is at least three times the concentration of platelets in the patient's own blood. Usually, the concentration of the platelets in the tissue sealant is no more than 10 times the concentration of platelets in the patient's own blood, such as no more than 7 times. In a certain aspect of the invention, the concentration of platelets in the tissue sealant is about 3 to about 9 times, such as about 5 to 7 times, the concentration of platelets in the patient's own blood.

In a certain aspect of the invention at least a portion of the fibrin is fibrin I monomer maintained at a pH lower than pH 5. The advantage is that the fibrin I polymer is soluble under mild acid conditions and thus easily can be applied to the stump ends. In a preferred embodiment, at least 90% of the fibrin in the tissue sealant is fibrin I.

According to the invention the amount of fibrin is at least 8 mg/mL. To ensure more suitable sustained-release matrix a higher concentration of fibrin I may be selected, such as a concentration above 10 mg/mL, such as above 12 mg/mL, such as above 14 mg/mL such as above 16 mg/mL. Suitably, the concentration of fibrin I does not exceed 50 mg/mL to reduce the risk of the sealant from becoming too viscous. Suitably, the concentration of the fibrin I does not exceed 30 mg/mL. In a preferred embodiment of the invention the concentration of fibrin I in the tissue sealant is 15 to 25 mg/mL.

In certain embodiments of the present invention the tissue sealant may comprise an anti-fibrinolytic agent, such as tranexamic acid to reduce the risk of blood loss during surgery and to reduce the decomposition of fibrin. Generally, the amount of tranexamic acid, when present is at least 0.05 mg/mL, such as at least a 0.1 mg/mL. The optional addition of anti-fibrinolytic agent such as tranexamic acid may be used to adjust the delivery time for the tissue factors. The time period of highest importance for obtaining a successful healing process is from day 0 to day 7. Thus, the amount of tranexamic acid may be adjusted so that a steady release of growth factors optionally derived from platelets is obtained during the first 7 days after surgery. In a suitable embodiment of the invention, the amount of tranexamic acid in the tissue sealant of the invention is adjusted so that the recommended i.v. dose of lOmg/kg 3-4 daily is not exceeded. Preferably, the concentration is not above 500 mg/ml in the sealant of the invention to prevent too slow degradation of the fibrin. At present, it is believed that a suitable concentration of anti-fibrinolytic agent is 1.5 mg/mL to 100 mg/mL .

In a preferred embodiment, part or all said fibrin I monomer after or simultaneously with application to the stump end(s) is polymerized to fibrin II by activation with an alkaline agent. The advantage is that the fibrin II aggregates and forms a physical connection between the intestine stump end ( s ) .

The high concentration of fibrin in the tissue sealant according to the invention result in effective connection between the 2 stump ends as illustrated by the increased burst pressure obtained in the experiments reported in the appending examples. Furthermore, a high elasticity of the anastomosis in the early stages is obtained for the tissue sealant according to the invention, which will reduce the risk of leakage when partly degraded food objects are passing the gastrointestinal tract.

In a preferred aspect, the tissue sealant is Vivostat® PRF® (platelet rich fibrin) sealant. During the preparation of Vivostat® PRF®, citrate is transferred to the Vivostat® PRF®- Preparation Unit, and 120 mL of blood is collected from the patient into the Preparation Unit. The pH 4 syringe is placed in the Preparation Unit, which is then loaded into the Vivostat® Processor Unit. Upon activation of the automated process, the blood is warmed to 36°C and it is separated in the upper reservoir chamber of the Preparation Unit by centrifugation . Plasma (containing the platelet fraction) is then passed through three channels to the reaction (mixing) chamber of the Preparation Unit, and batroxobin that has been released from the cartridge contained in the Preparation Unit, is mixed with the separated platelet rich plasma. In the plasma mixture that contains acid-soluble fibrin I, the fibrin I polymerizes and is isolated on the walls of the chamber by repeated centrifugation . Excess fibrinogen and platelet depleted serum are then withdrawn back into the upper reservoir (waste) chamber of the Preparation Unit, leaving concentrated fibrin I and platelets in the reaction chamber. This platelet-rich fibrin concentrate is then mixed with 3.5 mL of pH 4 sodium acetate buffer to dissolve the fibrin I molecules. The resulting acidified platelet-rich fibrin I solution is transferred into the bottom of the Preparation Unit. Any residual pH 4 buffer is dumped into a separate bottom in the Preparation Unit and the acidified platelet- rich fibrin I solution is then withdrawn into the empty pH 4 buffer syringe.

The concentrated fibrin I monomer solution is stable for up to 8 hours at room temperature or for several days at -20°C. The fibrin I monomer can be transferred to the desired applicator device for use in the present invention.

The concentrated fibrin I monomer solution is co-applied to the stump end(s) with an alkaline buffer (carbonate/- bicarbonate buffer, pH 10) to activate the fibrin I. Thus, described herein is also a method for applying fibrin I to one or both stump ends formed by resecting a segment of the gastrointestinal tract, comprising the step of spraying the tissue sealant of the invention comprising 8 mg/ml or more fibrin I to the stump end(s) and simultaneously spraying an alkaline buffer to activate the fibrin I. The ratio between the concentrated fibrin I monomer solution and the alkaline buffer is usually in the range of 4-10:1. The co-spraying of the fibrin I solution and the alkaline buffer result in an almost instant polymerization of the fibrin I monomer to the polymer. The almost instant polymerization provides the possibility of fast surgical intervention, which will reduce the time the patient is fully anesthetized.

In a specific embodiment, to initiate the application process, both the carefully mixed fibrin syringe and the syringe containing pH 10 carbonate/bicarbonate buffer are loaded into the Applicator Unit. Upon activation of the device's priming sequence, the plungers of the syringes move upwards at a set rate to deliver the two solutions through a multi-lumen tube until the system is primed (i.e., both lumens are completely full from syringe to tip of the Spraypen Applicator) . Upon further activation of the Spraypen Applicator, air, pressurized by the pump in the Application Unit, passes through a third lumen and exits together with the two solutions at the application tip where a fine spray is formed. Fibrin I in the now alkalized Vivostat® PRF® polymerizes, and after Vivostat® PRF® is applied, fibrinopeptide B (FPB) is released and react with endogenous thrombin to form fibrin II, which is crosslinked and polymerized by the activity of endogenous FXIIIa.

When the fibrin I is converted to the fibrin II the stump end(s) will become attached to each other. In a suitable embodiment, a connecting device is used in addition to the tissue sealant to secure the physical integrity of the anastomosis. The connecting device may be a suture hand-sewed by the surgeon to connect the stump end(s) . However, it is generally preferred to use one or more staple (s) to connect the stump end(s) in the anastomosis. A number of specialized staplers commercially available may be used to obtain the reconnection . Examples of useful stapling devices include Chex™ LC 60 linear stapler, Chex™ circular stapler obtainable from Frankenman Ltd; Echelon Flex™ GST System; DST Series™ EEA™ staplers available from Medtronic; and Circular stapler Circulo 3R and Pro - H View available from Sferamed.

In case a connecting device is not used, fasting may be continued to reduce the risk of leakage.

In a preferred aspect, the connecting device is wholly or partly covered with fibrin to increase the biocompatibility and reduce the risk of infection. Thus, the present invention also relates to a connecting device, such as a staple, covered by the tissue sealant according to the present invention. As an example, the staples may be covered by the same fibrin sealant as used in the formation of the anastomosis. Preferably, the fibrin is autologous, which will reduce the risk of rejection.

The tissue sealant may be co-administered with a desired substance selected among drugs, structural reinforcing substances, cells, or a mixture thereof.

In an aspect of the invention, the desired substance for coadministration is an antibiotic agent, an antifibrinolytic agent, a chemotherapeutic agent, or a pain relief medicament. An antibiotic agent is generally used as a preventive treatment in case of a leak of content from the anastomosis. An anti- fibrinolytic agent may be used to inhibit the fibrinolysis. A chemotherapeutic agent may be used in the treatment of cancer, such as colorectal cancer as a supplement to the removal of a segment of the gastrointestinal tract. A pain relief medicament may be used to increase the patient well-being or recovery.

In an embodiment of the invention the co-administered cells are stem cells, such as bone marrow or embryonic stem cells, or cells derived from adipose tissue. The co-administered cells may be useful in the formation of tissue at or around the anastomosis. In a certain embodiment, the stem cells are selected as allogenic or autologous stem cells for delivery to a patient also treated with chemotherapy or radiation.

As the healing process takes several weeks it may be advantageous to have a retarded or sustained release of the co-administrated desired substance. In an embodiment of the invention, the co-administrated desired substance is embedded in the fibrin II structure for sustained release. To further sustain the release of the co-administered desired substance, Anti-fibrinolytic agents may be added to the tissue sealant to prevent degradation of fibrin.

In other embodiments of the invention it is desired to have a fast initial burst release of a desired substance. In this case, a proteolytic enzyme, such as plasmin or trypsin, may be added to the tissue sealant to speed up the degradation process .

The present invention also relates to a method for treating cancer in the gastrointestinal tract, comprising the steps of: a) resecting a segment of the gastrointestinal tract, thereby producing a proximal gastrointestinal stump end and a distal gastrointestinal stump end,

b) Applying the tissue sealant as disclosed above to the proximal gastrointestinal stump end and/or the distal gastrointestinal stump end, and

c) Connecting the proximal gastrointestinal stump end and the distal gastrointestinal stump end to form an anastomosis .

The present invention also relates to the use of fibrin for the manufacture of a tissue sealant as disclosed above, for treatment or prevention of cancer in the gastro-intestinal tract .

Example

The study was designed to test the ability of the tissue sealant Vivostat® PRF® to reinforce colonic anastomosis in a pig colon, the effect on the healing process, and the complications, including leakage and bleeding post-surgery.

The objective was to evaluate the impact of the Vivostat® PRF® sealant on left side colonic anastomotic healing by postoperative complication rate and wound healing course.

Materials and methods

Study design and animal groups:

At sacrifice:

Bacteriology from pelvis

pressure: anastomosis with sealant #1,5,9,13

Study flow for each animal

Resu lts

1. Animal well-being

All animals underwent a health test prior to the study surgery and were acclimated for 7 days. The average baseline bodyweight and rectal temperature levels of the animals at the 1 ^st surgery were 47.3 kg (range, 44.5-49.5 kg) and 38.1°C (range, 36.2-39.0 °C) , respectively. The mean duration under anesthesia was 108 minutes (range, 66-158 minutes) . Other well-being signs monitored for the study animals included food consumption, alertness and bowel habits which were normal for all animals throughout the housing period. The animal's weight was reduced for the 4-day group - an average loss of 2 kg (range, 1.5-2.5 kg) . The 10-day and 30-day group exhibited a weight gain of 5 kg (range, 4-5.5 kg) and 14 kg (range, 12.5-16kg), respectively. These changes in weight are typical for pigs following abdominal surgery at these time points . The average bodyweight of the animals at sacrifice for the 4- day group was 43.3 kg (range, 43.0-44.0 kg), for the 10-day group 52.8 kg (range, 51.5-54.0 kg) and for the 30-day group 62.3 kg (range, 62.0-62.5 kg) . The mean duration of the sacrifice was 33.5 minutes (range, 10-57 minutes) .

None of the animals had an elevated fever, reduced appetite, changes in bowel habits, unexpected changes in weight or other clinical signs of anastomotic leakage during the housing period .

2. Intra and post-operative

2.1 Surgical technique

All anastomoses were created at 20 cm from the anal verge. The proximal stump closure was created using a manual purse- string device and the distal stump closure was created using the Chex™ LC 60 linear stapler (Frankenman Ltd.) . The circular anastomosis was created using the Chex™ CS 28 mm circular stapler (Frankenman Ltd.) .

There was no tension noted in any of the anastomoses created during the study, both the proximal and the distal donuts were complete for all anastomoses. Sutures were not added over any of the anastomoses.

A leak test using air insufflation in the colon and water in the abdomen was negative for leakage in all anastomoses following closure of the circular stapler. Blood loss was monitored during the surgery and there was no excessive blood loss (defined as blood loss over 500 ml) in any of the animals. 2.2 Sea lant appl ication

Blood was drawn from all animals for sealant creation prior to any antibiotics given, to prevent interference with sealant creation. The sealant was placed between the two bowel stumps; 1.5ml on each stump in all animals. The remaining amount of sealant (average, 2.5 ml; range, 2.1-3.5 ml) was placed as reinforcement over the anastomosis so that it covered 2 cm above and 2 cm below the anastomotic line. This was achieved in all animals except one - in animal number 8, 4-day group, approximately 80% of the external circumference of the anastomosis was covered due to a technical malfunction, see adverse event description below. Overall, the sealant was found easy to use by the veterinary doctor applying the material and the sealant was visible on the anastomosis during application which assisted in proper placement of the sealant. The time to complete the anastomosis was 8.6 minutes (range, 4.5-13.6 minutes) . This was defined as the time from creation of two bowel stumps to the time of retraction of the circular stapler. It took an average of 2.2 minutes (range, 1.4-2.9 minutes) to complete the application of sealant between the two bowel stumps, and an average of 1.8 minutes (range, 1.3-

3.3 minutes) to complete the application of sealant over the anastomosis. Of note is that the 3.3 minutes instance to complete the application was in a case of technical issue (see adverse events below for animal number 8, 4-day group) .

2.3 Bacteriology

A swab test for bacteriology testing was taken from all animals upon opening of the abdominal cavity and was negative in all cases. A second swab test for bacteriology testing was taken following the creation of the anastomosis and this was positive for E. coli growth in 3 out of the 16 animals. In one animal, there was sparse population (animal number 2, no sealant in the acute group) in the other two animals there was substantial growth (animals 7 and 8 in the 4-day group, both with sealant) . The final swab test for bacteriology testing was done upon entering the abdomen at sacrifice for all animals except the acute group and was negative in all cases. The swabs from the 4-day group at sacrifice were delayed in refrigeration and were tested for bacteriology after a few days.

2.4 Adhesions at sacrifice

Upon opening of the abdomen at sacrifice an evaluation of adhesions was done, both at the abdominal wall as well as intra-abdominal adhesions in the anastomotic area.

Abdominal wall adhesions: the 30-day group had no adhesions on the abdominal wall. The 10-day group and 4-day group had filmy/ mild adhesions on the abdominal wall most often of small bowel on the suture line area in all specimens. This is a typical finding of abdominal surgery in pigs.

Intra-abdominal adhesions in the anastomotic area were evaluated using a published score [Zuhlke et al . (2005) Ann Surg.; 241: 534-40]. Briefly, the abdominal adhesions were qualitatively scored using a scale of 1-4 (none = 0, filmy = 1, mild = 2, moderate = 3, and severe = 4) . Overall, there were very few abdominal adhesions following the surgery. 63% of the animals had adhesions, 75% of the non-sealant group and 58% of the sealant group. The overall average score for abdominal adhesions was 0.81 filmy adhesions.

In the 30-day group and 10-day group each, 3 animals had mild intra-abdominal adhesions and one animal in each group (sealant animal in both groups) had no intra-abdominal adhesions. In the 4-day group one animal had colon, small bowel and bladder adhesions to anastomotic area. It seemed like there may have been a micro-leak on the anastomosis site causing these adhesions (animal number 7) .

There were no intra-abdominal abscesses or peri-colonic abscesses in any of the study animals at sacrifice.

2.5 Adverse events and Complications

Adverse events and complications are described according to time of event.

• In one animal (animal number 13, 30-day group) the colon was relatively narrow. It was difficult to insert the circular stapler anvil. The stapler was "stuck" - could not retract as usual and required some force to retract.

• Surgical sterility was compromised in one animal (animal number 3, acute group) . Of note, that none of the bacteriology tests in this animal were positive for bacteria .

• The leak test was done in all animals prior to placement of the sealant apart for one animal (animal number 9, 10-day group) . • A slight burn on the bladder serosa was inadvertently done in one animal at the time of surgery (animal number 6, 4-day group) .

• In one animal (animal number 8, 4-day group) there was a technical malfunction. During application, there was a block and sealant did not flow. Initially, only 60% coverage over the anastomosis was achieved. After noticing that there was sealant left in the unit, the application kit was changed and the application continued. After using this additional sealant (0.5 ml) an approximate total coverage of 80% was reached around the anastomosis.

• Upon sacrifice animal number 10 (10-day group) had signs of scarring on the spleen. This was a non-sealant animal.

• One animal (animal 7, 4-day group) showed leakage signs in gross pathology. This animal and one more (animals 5, 4-day group) showed signs of micro-leak on histology. Both animals were sealant animals.

• One delivery of swab tests for bacteriology testing following sacrifice of the 4-day group was delayed in reaching the bacteriology lab and although the swabs were kept refrigerated for 7 days, the results may be unreliable (all 4 swabs were negative for growth) .

3. Anastomosis mecha n ical characteristics

The acute group underwent bowel preparation before surgery and sacrifice. The survival group's animals underwent bowel preparation before surgery but not before sacrifice. The anastomosis was harvested 5 cm above and below the anastomotic line and the proximal side was marked with a suture. The anastomotic burst pressure was measured ex-vivo on the day of sacrifice in one sealant animal in each group. For the non- sealant animals, following harvesting of the anastomosis, an additional anastomosis was created and the burst pressure was measured ex-vivo for this additional anastomosis.

Burst pressure measurement method: a tube connected to an inflation device was inserted through the distal end of the anastomosis and a zip tie was placed over it. An additional tube connected to a hand-operated mercury sphygmomanometer was inserted through the proximal end and a zip tie was placed over it. The anastomosis with the connected tubing was placed in a temperature controlled bath of 37°C. The actual temperature at the time of the burst testing was measured and was indeed found to be 37°C on average (range 36.8-37.3°C) . The pressure required to burst the anastomoses with sealant of the acute, 4 day, 10 day and 30 day anastomoses was 100 mmHg, 100 mmHg, 210 mmHg and 160 mmHg respectively. The average burst pressure for the additional acute anastomosis of the non-sealant animals was 37.5 mmHg (range, 20-60 mmHg) . All anastomoses burst as bubbles from the stapler line apart for one animal (animal number 13, 30-day group), which burst from the native tissue. The burst pressures for sealant and non-sealant animals are shown below in table 1.

Table 1:

Day 0 Day 4 Day 10 Day 30 Average

Sealant 100 100 210 160 141 mmHg

Non- 40 30 60 20 37.5 sealant

mmHg The data for the burst pressure indicates that the pigs treated with the tissue sealant according to the invention obtains a superior adhesion and sealing compared to the non- sealant group throughout the entire period.

Chart of acute burst pressure results for sealant and non sealant animals compared to normal intraluminal bowe pressure is shown in Fig. 1.

4. Gross pathology

As noted above, overall, the animals exhibited very little abdominal adhesions at sacrifice. This was noted by the study veterinary doctor to be an unusual result. Only 38% of sealant animals exhibited adhesions. In addition, the anastomotic line was barely visible on several of the anastomoses. In gross pathology, sealant was visible on the acute and 4-day group at sacrifice, however, it was not visible on the 10-day group and 30-day group. There was no evidence of hemorrhage in any of the animals.

5. H istological findings

General impression: The histological findings in the study were typical and as expected for colorectal anastomosis in a pig model. There was no significant difference in the histological parameters between sealant and non-sealant animals. The animals in which sealant was used showed no tissue damage or abnormal reaction as a result of the sealant. The acute group samples presented with acute hemorrhage with no other findings. There were remnants of sealant material on the serosal surface of the samples, which fell off during trimming and histological processing. The 4-day group samples presented with necrosis of the area of anastomosis and marked granulocyte infiltration (acute inflammation) with early macrophage infiltration and early fibroblast proliferation. In two samples (animals 5 and 7), necrosis was extensive in the area of anastomosis, full thickness in some of the cuts, suggestive of a micro-leak of intestinal fluid. In these samples, bacteria were observed in the sealant and there were aggregations of neutrophils infiltrating it. In some 4-day samples the sealant was still present, and there was evidence of minimal focal proliferation of fibroblasts in the area with focal neutrophilic infiltration, typical findings of anastomosis at this time point. In the 10-day group samples, the findings in all samples were very similar with evidence of early organization of the anastomotic area by fibrous tissue attempts of regeneration of the epithelium and decreased inflammation. Serosal fibrosis and adhesions were observed in some of the samples. Sealant was not observed in the 10-day group samples. In the 30-day group samples, regeneration was almost complete; the findings were typical of this time point.

5.1 Thickness of anastomotic line

The thickness of the anastomotic line was measured in millimeters. It is an approximate measurement taken at the level of the submucosa. A thinner anastomotic line indicates maturation of the scar. The anastomotic line was not measured for the anastomoses that underwent burst pressure, as this mechanical stretching does not allow for an accurate measurement of the natural anastomotic line. There was no fibrosis in the acute and the 4-day samples, the apposed edges of intestine were necrotic. In these cases, the area of necrosis, if present, or the distance between the two edges was given for these time points. The average anastomotic line width was 2.2 mm, 2.8 mm and 1.1 mm for all animals, sealant animals and non-sealant animals, respectively.

Table 2:

5.2 Maturation of fibrous tissue

The maturation of the fibrous tissue indicates organization of the granulation tissue (4: no maturation of fibrous tissue; 3: minimal maturation 2: some maturation; 1: distinct maturation 0: mature connective tissue almost normal) . Fibrosis was not present in the acute samples and a few reactive fibroblasts were present in the 4-day samples.

The average maturation fibrosis was scored 2.9, 2.9, 2.7 for all animals, sealant animals and non-sealant animals, respectively . Table 3:

5.3 Foreign body reaction

Foreign body reaction is the end-stage response of the inflammatory and wound healing responses. It is characterized by the presence of micro-granulomas , macrophages or giant cells surrounding food material of debris. Granulomas may form with increased/persistent inflammation or in cases of leakage or perforation, surrounding food material in the wall of the intestine or on the serosa and peripheral fat. Foreign body reaction around the staples is not scored as such. After complete healing of the anastomotic line the staples were mostly encapsulated or extruded in the lumen of the intestine and were no longer observed on histological sections.

The average foreign body reaction was scored 0.44, 0.39, 0.58 for all animals, sealant animals and non-sealant animals, respectively .

5.4 Inflammation

Inflammation is the infiltration by granulocytes, (neutrophils and eosinophils) and/or by mononuclear cells (lymphocytes, plasma cells, histiocytes) to the anastomotic area. Granulocytes increase in the acute phase as part of the healing process and are expected to decrease in the subacute and chronic samples. Mononuclear cells increase in the subacute samples, but should be in very small numbers in the chronic samples. The average inflammation score based on both mononuclear cells and granulocytes was scored 1.2, 1.3, 1.0 for all animals, sealant animals and non-sealant animals, respectively.

Table 4 :

5.5 Epithelization

Epithelization is attempts of mucosa adjacent to the anastomotic line to cover the defect on the mucosal surface due to the creation of the anastomosis (the scale for epithelization was 0 to 4 from 0: Continuous epithelium with proprial regeneration to 4: ulcerated surface, no epithelialization) . The average epithelization was scored 2.7, 2.9, 2.3 for all animals, sealant animals and non-sealant animals respectively .

Table 5 :

* 2 micro-leaks in this group

Discussion and conclusion

In the current study, we have investigated the wound healing processes of anastomoses which have been reinforced with the Vivostat® PRF sealant using a porcine model of left-sided colon anastomosis.

It is well-established that dehiscence of the large bowel anastomosis increases operative mortality and for survivors it significantly enhances morbidity and prolongs hospital stay. For this reason, we have focused in this investigation on the post-operative assessment of the colonic anastomotic healing and set our objective on evaluating the impact of the Vivostat® PRF sealant on anastomotic outcomes namely postoperative complication rate and wound healing course. To a limited extent, due to the small sample size, we were able to compare this impact to anastomoses created as per the standard of care. We primarily monitored for animal wellbeing, mechanical integrity of the anastomosis and histopathological parameters. In this study, we demonstrated that the anastomosis reinforced with the Vivostat® PRF sealant have a similar wound healing course to anastomosis without reinforcement and that the sealant was associated with limited intra- or post-operative complications.

Clinically, in every gastrointestinal anastomosis sutures or staples cause preliminary compression of the intestinal ends until wound healing finally links the adapted bowel ends. Our current histological findings are consistent with this healing course and show that the Vivostat® PRF sealant did not change or delay this process. In fact, healing of the tested samples showed minimal inflammation, early maturation of the fibrous tissue, and fast mucosal regeneration with evident re-epithelialization . Our observations demonstrated that the presence of sealant on the anastomosis resulted with a substantial increase in the bursting pressure strength of the left-colon anastomoses in the acute group - as compared with the non-sealant samples and the bursting pressure strength in the chronic groups was well above the intraluminal bowel pressure as reported in the literature ²⁰ .

The histological samples of two animals exhibited signs of micro-leak (animal number 5 and 7 from the 4-day group) . Both animals did not show any clinical signs of leakage (elevated temperature, reduced appetite, reduction in weight) and for both animals no intraabdominal and pericolonic abscess was found nor did they show any signs of sepsis. For animal number 5 there was no indication other than the histology for a leakage event. The animal wellbeing as well as gross pathology did not result in any suspicion for a leakage event. This anastomosis (animal number 5) underwent burst testing and the burst occurred on the anastomotic line which did not seem to be different from the other samples. For animal number 7, there were more adhesions upon sacrifice than the other animals at the same time point (see section 2.4 above) and the anastomotic line seemed wider. Of note, that for this animal the second swab test, following creation of the anastomosis was positive for e. coli. The histological evaluation shows that the bacterial colonies due to the micro- leak were restricted to the sealant area. Numerous neutrophils were seen infiltrating the necrotic area of anastomosis and were present in small numbers in the sealant. It is possible that this micro-leak was due to intra-operative contamination and that the sealant had a containing affect which avoided the presentation of an intra-abdominal abscess or full blown sepsis as the presenting signs of the leakage.

We conclude that anastomosis reinforced with the Vivostat® PRF sealant show similar bowel recovery after left-sided colon anastomosis in a porcine model as anastomosis created with the standard of care and no reinforcement with a substantially higher burst pressure. Our findings suggest that the sealant is able to contain contamination or leakage from the bowel, thus mitigating the clinical presentation of leakage.

Anastomosis histology

Mucin

The mucin content levels are shown in table 6, where a higher mucin percentage is considered a positive parameter for healing. The mucin content of the colonic crypt base was calculated using digital images of the PAS/Alcein blue- stained slides, where the bottom half of crypts within 2 mm of the anastomotic line with a clear long axis were chosen as the processing region of interest (ROI) . The Image-Pro Plus was filtered with median of 3 x 3 two passes and segmented for mucin. The mucin content was semi-quantitatively assessed within 2 mm of the anastomosis in measurable crypts so as to calculate a percentage by dividing the mucin stain by the total tissue area.

Table 6. Mucin content: Differences between sealant-treated and non-treated anastomoses.

The mean mucin percentage for matrix-treated cases was greater than the control anastomoses (54.5% vs. 42.3%) with higher mucin content in the earlier phase of the anastomosis for the non-sealant treated animals and a more constant mucin content for the sealant-treated group. In correlating these two parameters, on day 0 there was a consistent starting epithelial line thickness with 53% mucin content in both groups. By day 4 the epithelial thickness had increased slightly in the sealant-treated cases with mucin maintenance (excluding one animal #11 for possible contamination as suggested by the presence of increased epithelization and variable amounts of Collagen III) . By day 10 no differences were noted in the epithelial thickness but mucin production remained high in the sealant-treated cases with by 30 days an overall thicker epithelial layer in matrix-treated cases and similar mucin reduction in both groups.

Collagen deposition

A qualitative collagen maturity analysis was made with polarized light using the Picrosirius red (PSR) stain, detecting the relative amount of collagen III (immature collagen) as thin fibres with green birefringence. Re- epithelialization of the mucosa adjacent to the anastomotic line was semi-quantitatively determined with an epithelialisation score (0-4) where 0 = a continuous epithelial barrier with regeneration in the deeper layers and 4 = an ulcerated non-epithelialized surface.

Picrosirius Red staining at day 0 showed no immature collagen III (thin fibers with green birefringence) in any of the sites examined. By day 4 there was only a minimal amount of collagen III in all the sites examined with no difference noted between sealant-treated and control groups. By day 10, sealant- treated sites showed variable amounts of collagen III with less immature collagen deposition in those anastomoses with sealant. At day 30, there was a predominance of collagen fibers evident as orange birefringence with only rare small bundles of immature collagen evident in both matrix-treated and control anastomosis. In this qualitative assessment one animal was excluded because possible intra-operative contamination in the Day 10 group (#7) but the data would suggest that overall collagen III is reduced more in the sealant-treated sites, with a similarity between the groups by 30 days. This would imply an earlier conversion in the presence of the sealant from an immature to a more mature (Collagen 1) phenotype .

Inflammatory Infiltrates

Measurement of M2 macrophages (CD163/CD68+ and cMAF+ ) and non-M2 macrophages (CD163/CD68+, cMAF-) was performed in the mucosal, mural and serosal regions on double-stained images with Photoshop cell tagging and red-green fluorochrome separation of the M2 and non-M2 cell densities. A morphometric approach was used to semi-quantitatively assess differences in the degree of inflammatory cell infiltrate, where the pathologist was blinded to the treatment and times of tissue harvest. The type and number of inflammatory cells (granulocytes, mononuclear cells) were assessed on 5-7 H&E- stained, randomly selected high power fields (x 4000) and captured with a Q imaging camera (Olympus BH-2 microscope) . Images were assessed with an image analyzer (MetaMorph imaging system, Universal Imaging Corporation, West Chester PA) .

Infiltrates of M2 macrophages (which are involved in tissue repair and which are predominantly associated with an immunomodulatory Th2-type cytokine production profile) are inversely correlated with the expression of pro-inflammatory mediators generated from non-M2 macrophages. Table 8 shows the M2 and non-M2 densities in the matrix-treated and control groups at days 4, 10 and 30 along with the total macrophage density .

Table 8: M2 Macrophage Density (per mm ² ) : Differences between sealant-treated and non-treated anastomoses at mucosal, mural and serosal levels.

Day 30

Mean

Sealant 49 420 78 134 74 163 Mean No

Sealant 88 41 85 40

*Numbers represent the %age of M2 Macrophages measured as a density (cells/mm ² ) / % Total macrophages

Non-M2 macrophage density was lower than the M2 density when the sealant was used although there was no appreciable difference between the two groups by 30 days. Early on, total macrophage density was highest in the mucosa and muscularis in the control animals with no appreciable differences in M2 macrophage infiltration. By Day 10, the mucosa and the muscularis of control animals showed a marked reduction in the total macrophage density with higher levels of M2 macrophages in the mucosa and the muscularis of matrix-treated compared with control cases. This would suggest a different dynamic of pro-inflammatory macrophage infiltration with Sealant treatment. Conclusion

There was a slightly higher early mucin production combined with mature collagen deposition, a more extensive angiogenic response and a greater percentage infiltration into the mucosa and the muscularis of immunomodulatory macrophages. These findings suggest a shift in the inflammatory responses with application of this novel sealant towards a more rapid anastomotic healing.

References

1. McKinley AJ, Krukowski ZH. Intestinal anastomoses Surgery

2006; 24(7): 224-228.

2. Thompson SK, Chang EY, Jobe BA. Clinical review: Healing in gastrointestinal anastomoses, part I. Microsurgery

2006; 26(3) : 131-6.

3. Enoch S, Leaper DJ. Basic science of wound healing

Surgery 2005; 23(2):37-42.

4. Mortensen NJ, Ashraf S. Intestinal anastomosis. In Ashley

SW, ed. ACS Surgery: Principles and Practice, Vol. 5.

Philadelphia, PA: BC Decker Inc., 2008

5. Patel KA, Thomas WEG. Sutures, ligatures and staples.

Surgery 2008; 26(2):48-53.

6. Koruda MJ, Rolandelli RH . Experimental studies on the healing of colonic anastomoses. J Surg Res 1990;

48 (5) : 504-15.

7. Senagore A, Milsom JW, Walshaw RK, et al . Direct comparison between Czerny-Lembert and circular-stapled anastomotic techniques in colorectal anastomosis: a similar pattern of healing for both. Dis Colon Rectum 1992; 35 (9) : 862-9.

8. Witte MB, Barbul A. Repair of full-thickness bowel injury. Crit Care Med 2003; 31(8 Suppl ) : S538-46. van der Bilt JD, Borel Rinkes IH. Surgery and angiogenesis . Biochim Biophys Acta 2004; 1654 ( 1 ): 95-104. . Matthieesn P, et al . Risk Factors for Anastomotic Leakage After Anterior Resection of the Rectum. Colorectal disease 2004 6 .-462-469.

. Chittmittrapap S, et al . Anastomotic Leakage Following Surgery for Esophageal Atresia. Journal of Pediatric Surgery, 1992; 27(1) : 29-32

. Chambers W.M, Mortensen N . J . McC . Postoperative Leakage and Abscess Formation after Colorectal Surgery. Best Practice & Research Clinical Gastroenterology 2004; 18 (5): 865-880.

. Balemba OB, Grondahl ML, Mbassa GK, et al . The organisation of the enteric nervous system in the submucous and mucous layers of the small intestine of the pig studied by VIP and neurofilament protein immunohistochemistry. J Anat 1998; 192 ( Pt 2):257-67. . Karaosmanoglu T, Muftuoglu S, Dagdeviren A, et al . Morphological changes in the myenteric plexus of rat ileum after transection and end-to-end anastomosis. J Anat 1996; 188 ( Pt 2):323-31.

. Katsui R, Kojima Y, Kuniyasu H, et al . A new possibility for repairing the anal dysfunction by promoting regeneration of the reflex pathways in the enteric nervous system. Am J Physiol Gastrointest Liver Physiol 2008; 294 (4) :G1084-93.

. Hendriks T, Mastboom WJ. Healing of experimental intestinal anastomoses. Parameters for repair. Dis Colon Rectum 1990; 33 ( 10 ) : 891- 901.

. Neutzling CB, Lustosa SA, Proenca IM, da Silva EM, Matos D. Stapled versus handsewn methods for colorectal anastomosis surgery. Cochrane Database Syst.

Rev 2012; 2: CD003144.

18. Hesp FL, Hendriks T, Lubbers EJ, deBoer HH. Wound healing in the intestinal wall. A comparison between experimental ileal and colonic anastomoses. Dis Colon Rectum 1984;

27 (2) : 99-104.

19. Munireddy S, Kavalukas SL, Barbul A. Intra-abdominal Healing: Gastrointestinal Tract and Adhesions. Surg Clin North Am 2010.

20.Kozarek RA, Earnest DL, Silverstein ME, et al. Air pressure-induced colon injury during diagnostic colonoscopy. Gastroenterology 1980;78:7-14