Ventral abdominal wall defect correction in rats with contaminated meshes 1

PURPOSE: To investigate whether there is a difference between Marlex® and Dynamesh PP-light Marlex® meshes, in the abdominal wall defect correction, on rats in contaminated surgical site. METHODS: Twenty-eight Wistar rats were divided into two groups of 14, and four subgroups of seven animals. All subgroups underwent similar surgical procedure. One group received the mesh Marlex® and the other Dynamesh PP-light® for correction of the defect. Before implanting, the meshes went through a contamination process, on which was used standard solution containing 10 UFC of Escherichia coli. Fragments of the animal’s abdominal wall received macroscopic, microscopic and microbiological analysis. RESULTS: There was no statistical significance in the analysis of macroscopic variables. Accentuated inflammatory process was shown in all subgroups. The foreign body type reaction was mild in all subgroups, except Dynamesh®-14, which was moderate with no statistical significance. The microbiological analysis of the meshes was also similar between the subgroups. CONCLUSION: There was no difference between the meshes of Marlex® and Dynamesh PP-light® in the ventral abdominal wall defect correction on rats in contaminated surgical site.


Introduction
The incisional hernias or abdominal wall eventrations are frequent complications of abdominal surgery by laparotomy, estimating its incidence in 2-8%.Among the main risk factors for its appearance, are wound infection and obesity, among others [1][2][3] .
The surgical treatment is usually laborious and has high recurrence rates, even with experienced hands.There are several techniques for repair, from aponeurotic flaps to tension-free techniques, using synthetic prosthesis 4 .
The development of polypropylene prosthesis revolutionized surgery for repair of the abdominal wall hernia.
When compared to basic repairs, the tension-free techniques reduced recurrences and made possible the reconstruction of large ventral defects that were irreparable 5 .
Despite initial concerns about the possible rejection and infection, resulting from the use of prosthesis, there is evidence that tension-free hernioplasties using biomaterials, have significantly reduced recurrence and complication rates, making it accepted worldwide 6 .
The quality of synthetic meshes and surgical techniques has shown great developments in the past years.Materials such as polypropylene, polyglactin, polytetrafluoroethylene, woven polyester, polyvinylidene, among others, may be part, alone or in combination, of the composition of meshes currently used [7][8][9][10][11][12] .
Besides the material, which forms the mesh, its density also gives it particular characteristics.High density or microporous meshes (pores smaller than 10µm) may increase the chance of infection and fistula formation 13,14 .However, the low density or macroporous meshes (pores larger than 75 µm) prevents the development of infections 15,16 .
Often times the surgeon finds strangulated ventral hernias, urgently operated, with occlusion or with associated intestinal sub-occlusion, which favors the phenomenon of bacterial translocation.Considering several other clinical situations of possible peritoneal cavity contamination, such as those that occur directly through intestinal necrosis with perforations, fistulas and bacterial peritonitis, or in patients who underwent elective abdominal surgery with opening of the gastrointestinal tract and require associated herniorrhaphy, there is always the question as to whether or not use prosthesis, and what is the most appropriate prosthesis for this situation.
Thereby, this study aims to determine if there is a difference between the meshes of Marlex ® and Dynamesh PPlight ® in the ventral abdominal wall defect correction on rats in contaminated surgical site.

This study took place at the Experimental Surgery
Laboratory of the Faculdade Evangélica do Paraná, after approval by the Ethics Committee on Animal Research.
The sample consisted of 28 Wistar rats (Rattus novergicus albinus) males, adults, weighing between 280-358g, and the groups identified by picric acid staining at certain points, keeping the rats on day and night cycles of 12 h at room temperature of 24°C.Throughout the experiment, they received proper feed for the species and had free access to water.
For the experiment were used meshes of Marlex ® and DynaMesh -PP light ® (Figure 1).The mesh of Marlex ® consists of polypropylene monofilament classified as microporous (pores of 0.6 mm) and heavy grammage (95 g/m²).As Amid P 17 defined meshes with larger pores than 75 microns are considered macroporous.
These can be considered large pores when the pores are larger than 1.5 mm or 150 microns.Then the meshes used in the study are macroporous with large pores (DynaMesh PP light) and small pores (Marlex).The DynaMesh-PP light ® is composed of polypropylene monofilament, classified as macroporous (pores of 2.6 mm) and low grammage (36 g/m²).The meshes were cutted into pieces of 2.0x2.5cm, and contaminated with standard solution of Escherichia coli.The strains were cultivated in BHI broth; batch YF209, sterile, at 36° C. The solution was prepared 24 h before the experiment and reached a final concentration of 10ª UFC.
The 28 animals were split randomly into two groups of 14.The groups were given names related to the used meshes.The Marlex ® group was subdivided in two subgroups of seven animals.In the Marlex ® -7 subgroup, the euthanasia occurred on the 7 th day after surgery and in the Marlex ® -14 subgroup, the euthanasia on the 14 th day.The Dynamesh ® group was divided into two groups of seven.In the Dynamesh ® -7 subgroup, the euthanasia occurred on the 7 th postoperative day and in the Dynamesh ® -14 subgroup, on the 14 th day.
A digital scale weighed the animals and then forwarded them to the anesthesia.Each rat received a mixture of ketamine (90 mg/kg) and xylazine (10 mg/kg) intraperitoneally.The animals received anesthesia and considered ready for the surgical procedure as soon as they lost ocular and caudal reflexes.
The surgical procedures began with sub-xiphoid midline incision of approximately 3 cm length, followed by muscleaponeurotic resection of 1.0x1.5 cm on the upper left quadrant of the abdominal wall without opening of the peritoneum (Figure 2).Cutted meshes were submerged in the standard solution of Escherichia coli, stored in a Becker and applied in the muscular plane, with four separate points of polypropylene 4-0 (Figure 3).Then, the skin was closed with continuous suture, using polypropylene thread 4-0.To quantify the macroscopic analysis, was used the graduation 0=absent, 1=present, in the following items: skin necrosis; dehiscence in the surgical wound; seroma; hematoma at the surgical site; infection at the surgical site; and intra-abdominal adhesions.
Resection of a specimen, involving the complete operative area, was realized in all and with a cross section, they were divided into two equal parts.The cranial half was stored in sterile bottle with saline solution of 0.9% and sent for culture; the caudal, fixed in a cardboard template avoiding retraction into the bottle of formaldehyde (10% formalin) and sent for microscopic analysis.All vials received proper identification.
For microscopic analysis, the slides were processed in the usual way and stained with hematoxylin-eosin.In analysis, the slides were evaluated as following: acute inflammatory process or polymorphonuclear, and chronic or monomorfonuclear, and foreign body reaction or gigantocellular.
The cell count occurred by means of blind evaluation, where the pathologist had no knowledge of which subgroup he was evaluating.Based on the average of cells in the five largest power fields, ranked inflammation and gigantocellular in mild, moderate or marked on each animal (Figure 5).

Mild
Up to 5 neutrophils or lymphocytes Moderate 6 -20 neutrophils or lymphocytes Marked >20 neutrophils or lymphocytes

Mild
Up to 3 giant cells Moderate 4 -7 giant cells Marked >7 giant cells For microbiological analysis of the surgical specimen, was held culture in blood and MacConkey agar and considering positive if there was growth of Escherichia coli at any concentration after 48 h in culture medium.

Statistical analysis
The results of quantitative variables were described by

Results
The surgical procedures passed appropriately.The surgical average time was 10 min for each animal.There was an accidental opening of the parietal peritoneum in one animal from the Marlex ® -7 subgroup.In this animal, it was opted to perform the procedure in the upper right quadrant.
The animals had good postoperative evolution.One of the Marlex ® -7 subgroup died in the immediate postoperative period.

Macroscopic analysis
The Figure 6 illustrates the macroscopic appearance of the normal evolution of the healing process in the Marlex ® -14 and Dynamesh ® -14 subgroups.

Necrosis
One animal of the Marlex ® -7 subgroup evolved with necrosis of the entire surgical wound extension, beyond an underlying abscess.One animal of the Dynamesh ® -7 subgroup showed a necrotic spot in the pelvic region, associated with an intra-cavity abscess in this topography.There was no statistical significance in subsection necrosis in both groups (Figure 7).

Dehiscence of surgical wound
One animal from the Dynamesh ® -14 subgroup evolved with partial dehiscence with wound associated to the abscess at the surgical site.There was no statistical significance in the subsection wound dehiscence (Figure 8).

Seroma
Two animals from the Marlex ® -7 subgroup presented seroma at the implantation site of the surgical mesh.There was no statistical significance in the subsection seroma in both subgroups (Figure 9).

Hematoma at the surgical site
No animal showed hematoma.

Surgical site infection
Abscess in the abdominal wall was present in one animal from the Marlex ® 7-subgroup and one from the Marlex ® -14 subgroup (Figure 10).One animal from the Dynamesh ® -7 subgroup showed an intra-cavity abscess in the pelvic area and one from the Dynamesh-14 subgroup presented an abscess in the abdominal wall.There was no statistical significance in the subsection surgical site infection in both subgroups (Figure 11).

Intraperitoneal adhesions
No animal showed intraperitoneal adhesions.The average of foreign body giant cells was 1.54, with a median of 1.8.Characterized as severe inflammation and mild foreign body reaction (Figure 12B).The analysis of inflammatory cells showed no statistical difference between the subgroups of seven days.In the Marlex ® subgroup-14, the average polymorphonuclear cells was 24.48 and the median 24.6.The average mononuclear cells was 53.74 and the median 59.The average of foreign body giant cells was 3.17 with a median of 3.0.

Microscopic analysis
Characterized as severe inflammatory process with mild foreign body reaction (Figure 13A).
In the Dynamesh-14 subgroup, the average polymorphonuclear cells was 20.88 and the median 21.8 cells.
The average of mononuclear cells was 56.02 and the median was 53.4 cells.The average of foreign body giant cells was 4.37 with a median of 3.0 characterized as marked inflammation with moderate foreign body reaction (Figure 13B).The analysis of inflammatory cells showed no statistical difference between the subgroups.Comparing the Marlex ® -7 and Marlex ® -14 subgroups, it was noticed that the number of polymorphonuclear leukocytes and giant cells decreased while the mononuclear increased, featuring chronicity of the process (Figure 14A).

Microbiological analysis
Three animals from the Marlex ® 7-subgroup and three from the Dynamesh ® -7 subgroup presented positive cultures for Escherichia coli.There was no statistical difference between the subgroups.Three animals from the Marlex ® -14 subgroup and two from the Dynamesh ® -14 subgroup presented positive cultures for Escherichia coli.There was no statistical difference between the subgroups.

Discussion
The chosen animal for the research is the rat, due to its widely usage in studies involving meshes and repair of abdominal wall defects 8,17,18 .
There is no consensus in literature on the ideal prosthesis for use in contaminated environment.In most situations, one should choose low-density meshes with large pores and minimal surface area.Ideally, it should consist of a monofilament.If the mesh is placed in the peritoneal cavity, it needs a hybrid with a mesh of absorbable surfaces 6,19,20 .
The development of the polypropylene prosthetic revolutionized surgery for abdominal wall defects correction.The reduction in density of polypropylene, with the creation of lightweight meshes, theoretically reduced the foreign body reaction, causing less mesh contraction and providing better mesh incorporation in the abdominal wall, resulting in improved physiology of the abdominal wall [21][22][23] .Utiyama et al. 24 already showed no difference between polypropylene (high-density) and Ultrapro(r) (low-density) meshes at 21 days after surgery in extraperitoneal use in rats, comparing inflammatory response, mesh shortening, adhesions or complications.
This research focuses on experimental study of contaminated meshes in abdominal wall defect correction.This scenario simulates emergencies with incarcerated and strangulated hernias, in addition to elective situations where it is required to open the gastrointestinal tract, such as the paracolostomy hernia repair.The mesh used in the study was the polypropylene, chosen by being the most widespread and used both globally and in our field 25 .
The Escherichia coli is the most widely used bacteria in experimental studies that worked with contamination.Most studies use standard solutions containing this microorganism [25][26][27] .In 1989, Deitch et al. 28 in a clinical study of patients operated in emergency situations, with or without intestinal occlusion, noted that 59% had bacterial translocation and the more involved bacteria was Escherichia coli.
The option to use Escherichia coli for research, found that in cases of intestinal obstruction in strangulated hernias, the possibility of bacterial translocation can occur and this is the bacteria most often involved in these cases 28,29 and also is in the incidental openings of the human gastrointestinal tract.
The average time of the surgical procedure was approximately 10 min per animal.The most delicate moment of the surgery was the blunt dissection of the abdominal wall muscles, so that there was no violation of the parietal peritoneum, since the used meshes could not get in contact with the intra-abdominal viscera.Performing this blunt dissection with the aid of a swab was performed without difficulties.

Macroscopic evaluation
Barbuto et al. 30 studied the polypropylene meshes behavior in rats with and without induced peritonitis.It identified 50% of wound dehiscence in rats with peritonitis and 60% of those without peritonitis, with no statistical difference.This study observed incidence of 14.3% of wound dehiscence, which occurred in the Dynamesh ® -14 subgroup, a result similar to the study of found 12.5% of dehiscence of the surgical wound in Ultrapro-7 subgroup and 12.5% of wound dehiscence in the Proceed-28 subgroup.
The occurrence of seroma in the surgical site was early complication, as was present in 33.4% of the animals of the Marlex ® -7 subgroup, 28.6% of Dynamesh ® -7 and no animal of the 14 days subgroups.Our incidence of seroma was similar to that found by Klinge et al. 22 , 36% in the heavyweight meshes group and 20% in the lightweight.Greca et al. 21in study comparing meshes of heavyweight and lightweight, in dogs, observed a rate of 20% of seroma in both meshes.
Was not found the formation of hematomas in any animal, as well as Utrabo et al. 32 .Pundek et al. 31 had 12.5% incidence of hematoma in the Proceed-15 subgroup.Isa et al. 3 found 11% of hematoma in the Ultrapro-7 subgroup and 12.5% in the Proceed-7 subgroup.
Exactly one animal from each of the four subgroups of this study developed an abscess at the surgical site.Pundek et al. 31 and Utrabo et al. 32 did not show any cases of surgical site infection, as their paper did not involve any source of contamination of the surgical site.In disagreement with literature -since the rat is very resistant to infections -Isa et al. 3 showed high rates of early infection at the surgical site; the first author, with 55.6% of infection and the second, with 66.7% in the subgroups of seven days.
No animals in this study presented intra-peritoneal adhesions.We only compared this result with Utrabo et al. 32 because it was the only one to preserve the peritoneum, not allowing the prosthesis to come into direct contact with the intraabdominal viscera.The author found 18.75% of adhesions in the 30 days subgroup and 6.25% in 60 days, with no statistical significance.

Microscopic evaluation
In this study, was found no statistically significant difference in the evaluation of inflammatory response among Marlex ® -7 subgroups and Dynamesh ® -7, or between Marlex ® -14 and Dynamesh ® -14 21 .
Other studies with polypropylene meshes concluded that the higher the mesh grammage, greater the inflammatory response and surgical complications, determining a better biocompatibility of lightweight meshes 13,22,23 .
In disagreement, Weyhe et al. 19 in an experimental study in rats found worst biocompatibility of low-density meshes in comparison with high-density.

Microbiological evaluation
This study found positive cultures in 50% of the animals of the Marlex ® -7 subgroup and 42.9% of the Dynamesh ® -7 subgroup.In Marlex ® -14 subgroup, the positive cultures was 42.9%, and Dynamesh ® -14, 28.6%.However, at different times, Sebben et al. 25 obtained positive meshes cultures of 83% when the reoperation was in 24h, 33% when in 48h and 17% in 72h.As an overall result, his group showed 44% of positive meshes cultures.

FIGURE 1 -
FIGURE 1 -To the left, meshes of Marlex ® and to the right, Dynamesh PP-light ®

FIGURE 4 -
FIGURE 4 -Incision in "C" exposing the surgical site.

FIGURE 5 -
FIGURE 5 -Classification of the inflammatory and gigantocellular processes.
medians, minimum and maximum values, and the qualitative as frequencies and percentages.For comparing two groups in relation to qualitative variables considered the Fisher exact test and for the quantitative, non-parametric Mann-Whitney.Values of p<0.05 showed statistically significance.Analyzed the data with the computer program IBM SPSS Statistics v.20.
50.83 and the median 31.8.The average mononuclear cells was 40.09 and median of 39.1.The average foreign body giant cells was 3.23 with a median of 2.6.It characterized as inflammation of sharp intensity, with mild foreign body reaction (Figure 12A).In the Dynamesh ® -7 subgroup, the average of polymorphonuclear cells was 40.85 with a median of 32.8.The average of mononuclear cells was 53.82 with a median of 53.4.