Comparative study of intraperitoneal adhesions related to light-weight polypropylene mesh and type I polymerized and purified bovine collagen coated light-weight polypropylene mesh in rabbits.

PURPOSE
To compare the effectiveness of light-weight polypropylene mesh coated with polymerized and purified bovine type I collagen (Surgidry HNB) in the treatment of abdominal wall defect and the degree of adhesion formation.


METHODS
Two types of polypropylene mesh were implanted after creation of defect measuring 6.0cm X 5.5cm in the anterior abdominal wall of 32 male New Zealand breed rabbits, divided in two groups (n = 32): (1) light-weigh macroporous polypropylene, (2) type I polymerized and purified bovine collagen coated light-weigh macroporous polypropylene. These animals were further accessed for adhesions, histological evaluation of inflammation and wall's thickness.


RESULTS
The percentage of the area adhered in group 1 (62.31 ± 16.6) was higher compared to group 2 (22.19 ± 14.57) (p <0.05). There was an association between the percentage of the covered area by adhesions and the type of adhesion, toughness and the scores obtained by the adhesion score by correlation analysis (p <0.05). There was no difference between the groups in any variables in relation to the degree of inflammation.


CONCLUSION
The purified type I bovine collagen coated light-weigh polypropylene mesh showed to be effective in the repair of abdominal wall defects and reducing adhesion formation.

peritoneum mesothelial cells over the mesh surface. The peritoneum regenerates itself over the mesh in 7 days, avoiding the formation of adhesions in the covered area 1,3,5,6. Studies that relate the treatment of abdominal wall defects to the formation of adhesions are mostly performed in animals and widely observed in medical literature 5,6,8 . The macroscopic quantification, due to the direct observation of the adhesions, makes studies in human beings ethically prohibitive, once a new surgical procedure would be necessary, allowing the formation of new adhesions. Therefore, studies performed in humans have the bias of being performed under the treatment of complications caused by initial repairs, such as hernia recurrence or intestinal obstruction caused by adhesions 3,4,6,7 .
The use of polymerized and purified type I bovine pericardium organic matrix for the correction of abdominal wall defects in rabbits was studied, and it was come to the conclusion that this material is not effective due to the high incidence of hernia formation. However, this material showed to be biocompatible, causing minimal inflammatory and foreign body reaction, having a better architectural arrangement of the collagen fibers, allowing its adequate use as a barrier against the formation of adhesions [4][5][6]8 .
In virtue of the absence of an "ideal" prosthesis, the material used must be individualized, taken each case into consideration. This study aims to bring a new association of materials to be used for the adequate treatment of the abdominal wall defects and reduction of formation of intrabdominal adhesions. The composed mesh produced for this study is the result of the combination of a polypropylene macroporous light-weight mesh -Repol Mesh -and a polymerized and purified type I bovine pericardium organic matrix sterilized with gama radiation, combined together with a Vycril ® 5-0

■ Introduction
The popularization of laparascopic correction of incisonal hernias with intraperitonial fixation of mesh raised the concerns towards the formation of postsurgical adhesions. The lack of and "ideal" mesh lead to the development of a series of composed meshes, mixing different types of materials 1-4 . The main advantage of the composed meshes it is the possibility of intraperitoneal placement, with minimal adhesion formation. Regardless of the large spectrum of brand options available, almost all manufacturers still use the same three basic materials -polypropylene, polyester or polytetrafluoroethylene (PTFE). These materials are used in combination one with another or with a variety of different materials, such as titanium, omega 3, monocryl, polyvinylidene-fluoride (PVDF) and hyaluronate [2][3][4] .
Every mesh may produce adhesions once put in contact with the intestine, but the extension of the adherence is determined by the width of its pores, filament structure and the area of the surface in contact with the bowels 1,2,4 . Standart meshes or heavy-weight tend to produce a rather intense fibrotic tissue response, granting its adhesion to the abdominal wall, but, also, to the intrabdominal structures. Nevertheless, microporous meshes and ePTFE based meshes present with lower fibrotic tissue growth, decreasing the formation of adhesions, but at the same time reducing its ability to adhere to the abdominal wall 2,3,6 .
Those opposed findings illustrate the difficulty within the manufacturing of an ideal prosthesis that presents adequate adherence to the abdominal without compromising intraabdominal structures. Composed meshes intent to provide an additional surface, that allows safe placement when in contact with intraabdominal structures due to the inflammatory reaction that induces the proliferation of suture. This composed mesh, under study, has never been described in medical literature so far.

■ Methods
This work was carried out in accordance with that recommended by the International Standards for the Protection of Animals and the Brazilian Animal Experimentation Code (1988), and was approved by the Committee of Ethics in Animal Experimentation, Universidade Federal de Minas Gerais of the protocol number 099/2011.
Thirty-two New Zeland male rabbits, three months of age and with weights above two kilograms, acquired from the Experimental Veterinary Farm were studied. All of the rabbits were identified and placed in the Biothery of the School of Medicine, one animal per cage. They received daily rations for rabbits and filtered water ad libitum.
The rabbits were anesthetized with an intramuscular injection in the gluteal region with 5% ketamine hydrochloride (Ketamin-S ® (+), Cristália, Itapira-SP) at a dose of 35 mg/kg (0.7 ml/kg), coupled with 2% xylazine hydrochloride (Rompun ® , Bayer, Sao Paulo-SP) at a dose of 6 mg/kg (0.3 ml/kg). When necessary, half of the initial dose of the anesthesia was applied. During the entire period of anesthesia, the heart and respiration rate were observed, as were the rabbit's voluntary movements, in an attempt to detect complications.
After the trichotomy of the abdomen, antisepsis was carried out, using a 2% degerming chlorhexidine solution followed by a 70% alcohol solution and the setting of surgical fields. A defect measuring 6cm x 5.5cm was created in the ventral abdominal wall with removal of a muscle-aponeurotic sheaf, using single card template for all animals. The rabbits were then divided into two groups of 16 animals by a random drawing (n=32) for placement of the mesh in the intraperitoneal space: -Group 1: placement of light-weigh macroporous polypropylene mesh, 8.0cm x 7.5cm on the borders of the abdominal wall within 1cm of the edge of the mesh, with a 3-0 monofilament polypropylene suture.
-Group 2: placement of type I polymerized and purified bovine collagen coated light-weigh macroporous polypropylene, 8.0cm x 7.5cm on the borders of the abdominal wall within 1cm of the edge of the mesh using only the polypropylene layer as anchorage, with a 3-0 monofilament polypropylene suture.
After the surgery and during the entire follow-up period, the rabbits received rations and filtered water ad libitum, and were kept in individual cages, under appropriate conditions of hygiene, ventilation and natural illumination.
At the end of the follow-up period, on the 90th postoperative day, the animals were killed with inhalation of carbon dioxide in a closed chamber following an intramuscular injection of 2 ml of xylazine (10mg/kg).
A laparotomy in U, released a quadrangular portion of the abdominal wall, to study the occurrence of intrabdominal adhesions within evaluation of the following aspects: intrabdominal organs with adhesions, presence of vascularization, level of resistance, percentage of the mesh's surface covered by adhesions and incorporation of the mesh in the edges of wall defect.
The analyses of percentage of coverage by adhesions were calculated using the IMAGEJ software 1.47V (Wayne Rasband, National Institutes of Health, USA), and two images, taken with a Canon EOS 1100D camera with EFD 18-55mm lenses, of each animal. Mascroscopic evaluation of the adhesions were made using the adhesion score described by the Surgical Membrane Study Group 9 (Table 1). Opaque, large vessels present 4 Tenacity None 0 Adhesions falls apart 1

Adhesions lysed with traction 2
Adhesions requiring sharp dissection 3

Possible total 11
The removed abdominal wall was then prepared for histological study and stained with hematoxylin and eosin (HE) and Masson trichrome staining. The microscopic evaluation with HE and Masson's staining was performed to quantify the foreign body gigantic cells, granulomas, chronic inflammation, acute inflammation and neovascularization. A measurement of the wall thickness was also performed.
The analyses were performed using SPSS (Statistical Package for Social Science) version 20.0, 2012 of Microsoft Excel database tables. The variables were analyzed regarding their distribution and supposition of normality was verified by the Shapiro-Wild test. For the analysis of the qualitative variables Pearson's Chi-square distribution and Fisher's exact test were used, and the quantitative analyses were performed using Student's t test for the independent samples.
For the correlation analysis Pearson's coefficients were adopted in the cases in which the variables were not categorized. For the total punctuation score, categorized in low and high risk, the Biserial Correlation Coefficient, which is an estimate of Person's Linear Correlation Coefficient, was used.

■ Results
All of the animals recovered spontaneously from the surgeries and survived the three-month experiment and presented no late complications, such as hematomas, surgical wound infection, fistulas, incisional hernia relapse or mesh extrusion. Two animals, one from each group, presented surgical wound seroma, with spontaneous regression and no need of surgical approach.

Area covered by adhesions
The percentage of the mesh's surface covered by adhesions was smaller in the composite mesh group when compared to the polypropylene group (63.31% vs. 22.91%; p<0.05) ( Table 3). Moreover, when analyzing both groups total scores according to the Adhesion Score of the Surgical Membrane study group, it was possible to observe that 93.75% of the animals belonging to polypropylene group presented with a total score above 5, however only 6.25% of the animals from the group that received the composite mesh presented a comparable score (Tables 4 and 5).

Level of adherences -Tenacity
According to the classification of adhesions, it was possible to observe that the groups presented different tenacity levels. From the polypropylene mesh group, 2 animals presented each levels 1 and 2 of adherences, giving a 6.3% for the respective categories. Meanwhile 14 animals presented level 3 adherences (87.5%). Amongst the composite mesh group it was seen a larger occurrence of adhesions levels 1 and 2 (10 and 5 animals each, comprehending, statistically, 62.5% and 31,3% of the total) and a reduced number of level 3 adhesions, with only one animal affected (6,3%) ( Figure 2).

Correlation analysis
The correlation analysis was made between the average area of mesh covered by adhesions and the following variants: type of adhesions, tenacity and adherence score. The average of the percentage of the surface covered by adhesions (41.9%), was used to distribute all animals (n=32) within 2 groups: below and above the average.
The first correlation, between the tenacity and the percentage of mesh covered by adhesions, showed that the animals with adhesions of tenacity ranked 1 were the ones that had adhesions below the average. None of the animals presented a score 0, confirming the correlation between the increase of adhesion surface and the tenacity score (Table 6). The second correlation used the Chisquare test to associate the type of adhesion and with percentage of the mesh surface covered by the adhesions. In the same way, that the animals that presented a smaller coverage area by the adhesions were ranked score 1, showing a direct relation between the increase of adhesion surface and the type of adhesion score (Table 7). The third, and last correlation, was made between the total result obtained according to the score of adhesions and the percentage of the mesh surface covered by the adhesions. This last correlation proved that the increase in the surface covered by adhesions is directly related to a higher adhesion score (Table 8).

Microscopic analysis of inflammation degree
There was no statistical difference between none of the evaluated variables, according to modified Hooker's classification: giant foreign body cells, granulomas, chronic infection, acute inflammation and neovascularization.

■ Discussion
Large sized animals are the most adequate for the development of abdominal wall defect models. The advantaged seen in utilizing small animals, such as rats and rabbits, lies in the advantage of lower cost, small sized accommodations and more accessible materials 10 .
In this study this fact was taken into consideration, and this acknowledgment made the choice of rabbits suitable for the development of the study. The choice weighted the benefits of the viability of the surgical procedure, without recurring to special surgical equipment, and the smaller level of biological complexity regarding the surgical procedure itself and the post-operative care necessary for the amount of animals required 10 .
In the present study both meshes were placed in the abdominal cavity, according to the laparoscopic technique, having it fixated to the edges of the abdominal defects and with a 1 cm overlap, minimum (respecting Pascal's principle). The lightweight polypropylene meshes were effective for the repair of the abdominal wall due to the outcome of no incisional hernia being developed by any of the animals submitted to the application of this prosthesis.
The composite mesh produced for this study was effective for the reduction of the area adhered do the surface of the mesh. This result was impaired only by the exposition of a certain amount of polypropylene into de cavity causing omental, intestinal and liver adherences. Ideally, the mesh would be completely covered by bovine collagen, however the high costs and technological complexity of the production did not justify the cost-benefit of the attempt, once no studies proved its full effectiveness until the present moment.
Type 1 bovine collagen was effective, even with a long observational period, in the reduction of extension and degree of adherences according to Jenkins classification. For the correlation analysis, the animals were sorted in a manner that the results showing no adhesions or adhesions with small clinical significance (level I) were in the same group, and in those animals the composite mesh showed a positive correlation in all variables evaluated. Therefore, the composite mesh, besides the reduction of adherence area, also presented looser adhesions.
As predicted, there was no statistical difference in any of the analyzed variables, according to Hooker's modified classification. The inflammatory process induced by the presence of the lightweight polypropylene mesh was the same in both groups, and the organic collagen layer acted as a barrier preventing the formation of intrabdominal adherences between the intrabdominal organs and the polypropylene mesh, modulating the biological response to the presence of the mesh, but without interfering on its efficiency on the repair of the abdominal defect and its resistance to the mechanic tensions of the abdominal wall.

■ Conclusions
The type I bovine collagen and lightweight polypropylene composite mesh is an effective prosthetic for the repair of abdominal wall defects associated with lesser adhesions, considering that there is no perfect model or manufactured mesh that suits the prerogative of not producing any foreign body reaction, according to modern literature. Although the animal model utilized does not suits the requirements of extrapolating this study results for other biological models, such as humans, this experimental study is the first that evaluates and analyses both materials assembled and may influence the investment in advanced technology for the production of a polypropylene prosthetic mesh fully coated by bovine collagen, so that further studies might be produced.