Influence of collagen membrane on bone quality in titanium meshes reconstructions-study in rats Running title-Influence of collagen membrane on bone quality Summary sentence-No influence of collagen membrane on guided bone regeneration with titanium meshes

BACKGROUND
New bone formation and tissue remodeling are the major challenges in today's implantology. Titanium meshes have demonstrated reconstructive potential for vertical bone gain. However, the soft tissue healing is technically sensitive to the surgical procedure. The combined usage of collagen membrane and specification of the meshes may ensure greater predictability. Therefore, this study aims to evaluate the influence of collagen membrane on the quality of the new bone formation in guided bone regeneration procedures with different titanium meshes.


METHODS
Twenty-eight Wistar rats were randomly allocated into four main experimental groups, according to mesh pore size in μm: Group P300 (titanium meshes, with 0.3 mm thickness and 3 mm pore size; n = 7); Group P175 (titanium meshes, with 0.3 mm thickness and 1.75 mm pore size; n = 7); Group P85: (titanium meshes, with 0.04 mm thickness and 0.85 mm pore size; n = 7); Group P15: (titanium meshes. with 0.04 mm thickness and 0.15 pore size; n = 7). The femurs of each animal were subdivided into test and control groups: Test (T): bovine bone graft associated with porcine collagen and collagen membrane was used; Control (C): bovine bone graft associated with porcine collagen was used without association with collagen membrane. Bone quality evaluation by in vivo microtomography and histologic analysis were performed.


RESULTS
Bone volume formation was similar between groups (p > 0.05). However, the titanium meshes with pore size >1mm demonstrated higher mineral bone density in comparison to meshes with pore size <1mm (p < 0.05), regardless the combined usage of collagen membrane. All groups showed a spongy bone formation after 30 days.


CONCLUSION
Combined usage of collagen membrane in guided bone regeneration procedures with titanium mesh did not show improvements in new bone quality in rat femur model. However, titanium mesh pore size specifications may have influence in bone quality. This article is protected by copyright. All rights reserved.


Introduction
Dental implants for tooth replacement have been supported by studies [1][2][3] , but in some cases, residual bone volume is often insufficient for optimal rehabilitation. Thus, restoration of bone volume has become necessary to achieve success in rehabilitation with implants 4 .
One of the techniques that allows gain and maintenance of bone tissue is guided bone regeneration (GBR), in which a mechanical barrier is positioned to prevent rapid fibroblasts proliferation, allowing defect osteoprogenitor cells to repopulate the area, initiating bone formation process [5][6][7] . GBR is recognized as an effective and predictable method to ensure bone formation and, in many cases, is associated with bone grafts or substitutes, which act as osteoconductors 8 .
Studies comparing the best type of barrier to be used have been published in literature since the last 60 years 4,6,[9][10][11][12] . Occlusive membranes in association with bone graft material demonstrated a gain of bone tissue in several studies [11][12][13][14] . Absorbable membranes, for example, maintain a temporary barrier between 6 to 8 weeks 15 , and eliminate need for further surgical procedure for removal. However, it has been shown that occlusive membranes without titanium reinforcement, made from soft materials such as collagen membranes, tend to collapse in large reconstructions because they do not have adequate resistance to space maintenance 9 , which is essential in vertical bone reconstructions 10 . Titanium mesh barriers have been shown that bone volume to be created can be planned prior to the surgical procedure and mesh can be molded to maintain volume during healing period without graft compression by the flap 16 . The presence of pores prevents soft tissue growth internally and allows interstitial fluid diffusion 14 . Studies shows satisfactory results for bone reconstructions with titanium meshes 9, 17-19 . This article is protected by copyright. All rights reserved.
However, morphology factors such as pore size and space maintenance, are discussed to ensure greater predictability. Studies in literature have been demonstrated space maintenance importance 9,[20][21][22][23] and also that barriers needs to be malleable enough to promote required geometry to ensure bone gain in height and thickness 17 . Furthermore, it has been suggested that larger diameter pore size allows new bone angiogenesis and better nutrients diffusion. On the other hand, smaller diameter pore size completely occlusive meshes may limit neovascularization process, but also restrict fibrous connective tissue invasion 9,24 .
Authors have suggested that bone growth occurs in 50 µm diameter pore size meshes 25 .
Up until present time, the ideal mechanical barrier for GBR remains in studies, aiming analyze factors such as, occlusivity, stability, ideal pore size, peripheral sealing between barrier and bone tissue, blood supply required and providing proliferation of osteoprogenitor cells 26 . However, some clinical and preclinical studies have demonstrated limited bone regeneration and soft tissue infiltration when occlusive membranes were not used in association to titanium mesh 24, 27, 28 .
Lim et al. used titanium mesh after implant installation and, despite high exposure rate, authors did not observe bone regeneration in many samples, suggesting the additional use of occlusive membrane 28 . The presence of soft tissue layer below titanium mesh and lack of mineralization soft tissue evidence, demonstrates possibility on using occlusive membrane 24,29,30 . Despite mechanical barrier need to obtain GBR criteria, good results were demonstrated through after use of different pore size titanium meshes 8,9,16,[30][31][32][33] , which shows that results are inconclusive.
Thus, the present study aims to evaluate collagen membrane influence in GBR when associated with titanium meshes. Moreover, evaluate differences in pore size and thickness of titanium meshes. This article is protected by copyright. All rights reserved.

Ethical aspects and financial support
Initially, this research project was sent to University of Ribeirão Preto ethics committee (CEP / UNAERP), which was duly approved and registered through code ComÉt: 15/2015. ARRIVE guidelines were consulted in reporting this study 34 .

Samples characterization
Twenty-eight male Wistar rats (Rattus norvegicus albinus) with an average weight of 410.8 grams were used in this study, all originated from UNAERP central laboratory. The animals were kept in appropriate plastic boxes with food and water ad libitum before and during experimental period, and remained in UNAERP`s laboratory in a 12 hours cycle environment of light and temperature between 22 and 24ºC.

Experimental groups
Animals were randomly allocated into four main experimental groups, according to titanium meshes use:  Test (T): femur in which bovine bone graft associated with porcine collagen ‡ ‡ (BC) and collagen membrane § § were used.
 Control (C): femur in which only BC was used.
To determine collagen membrane use, right and left femurs of each animal were randomized using the Microsoft Excel 14.0.7 program *** . Randomization was only revealed after opening surgical wound and complete visualization of femur.

Surgical Procedure
After weighing, animals received general anesthesia, obtained by the association of 0.08 ml / 100g of Ketamine Hydrochloride and 0.04 ml / 100g of Xylazine Hydrochloride, via intramuscular injection. Subsequently, tricotomy was performed with electric cutter and then local asepsis using a 2% chlorhexidine solution.
Initial incision was made parallel to long axis of the femur, in thigh outer portion, using # 3 scalpel handle, mounted with # 15 scalpel blade † † † ( Figure 1A between each other, which was measured with a periodontal probe ( Figure 1B and C). Then each block of bone graft BC ( Figure 1D) was equally divided into four samples, so that one BC was enough for four femurs. BC sample was then positioned over perforations and titanium mesh was positioned and fixed with two gingival screws of 5 mm high and 1.6 mm in diameter **** . In groups P300 and P175, screws were positioned on femur opposite sides ( Figure 1E and G). In groups P85 and P15, screws were positioned only at the top of femur ( Figure 1F and H). Following randomization, 10 mm x 15 mm collagen membrane was positioned above titanium mesh, only in femur test ( Figure 1I). Surgical area was closed using absorbable sutures † † † † .
After surgery, animals received a single intramuscular antibiotic dose of 24,000 IU / kg Penicillin G-benzathine at a dose of 0.01 ml per 100 g of the body weight and dipyrone 500 mg in water. Twenty-four hours after surgical procedure, animals were anesthetized using the same anesthesia technique previously reported. Then, each animal was positioned for in vivo computerized microtomograph analysis, thus determining baseline analysis.  Image J program § § § § § was used.

Analysis of results
Statistical analysis was performed with GraphPad Prism version 7 statistical program ****** . Microtomographic data and histological measurements of bone type and BC/femur interface were compared between groups through ANOVA two criteria test, with Tukey post test. A significance level of 5% was used for all statistical analyzes.

Results
Surgical procedure occurred with some intercurrences. Group P175 and P85 lost two animals in each group. P15 group lost one animal, totaling 23 animals in the experiment.

Volumetric measurements
Numerical values were distributed in Table 1 for each three-dimensional parameter evaluated. Initially, a lower tissue graft volume was observed in the P300 and P175 groups.
After the 30-day period, tissue volume was similar in all groups. In all volumetric parameters after 30 days, no statistically significant differences were observed.
In relation to BMD after 30 days, denser tissue was observed in the titanium meshes with pore size > 1 mm (p <0.05), with mean and standard deviation of 0.1 ± 0.03 and 0.11 ± 0.02 in control femur, and 0.01± 0.03 and 0.10 ± 0.02 in test femur, for groups P300 and P175 respectively (Table 1).
Tridimensional reconstructions demonstrated titanium meshes, screws and bone graft set at baseline and 30 days. This article is protected by copyright. All rights reserved.

Bone type analysis
Percentage of non-mineralized areas were 37±18. 8

Discussion
In this present study, four types of titanium mesh were associated with bone graft and collagen membrane in order to evaluate additional use of occlusive membrane influence in titanium mesh's reconstructions and evaluate the influence of mesh thickness and pore size in new bone formation. Pore size ranged from 15 to 300 μm, and thickness from 4 and 30 μm.
The model used in the study was a rat femur, due to greater disposition in adapt titanium meshes. In addition, exophytic bone formation model, to gain bone tissue out of skeletal framework, was used as opposed to bone formation surrounded by defect walls or alveoli.
Thus, new bone tissue was formed over the outer surface of femur, without creating a defect.
The collagen membrane used in the present study is absorbable, composed of type I and III porcine collagen, in a double layer. The inner surface, facing bone tissue, consists of This article is protected by copyright. All rights reserved.
disorganized collagen fibers, which allow osteoblasts proliferation. The outer surface facing soft tissue is dense and prevents fibroblasts proliferation 35 . In vivo study in rat femur demonstrated that membrane absorption process is initiated after 4 weeks and completed after 6 weeks 15 .
It was suggested that occlusive barrier uses in association with titanium meshes could minimize soft tissue formation in bone defect, aiming acquire a biomaterial or combination of biomaterials that would allow bone tissue formation of better quality 8,9,16,30,33  This article is protected by copyright. All rights reserved.
Bone graft used in this study is composed by a combination of inorganic bovine bone particles and a purified collagen matrix at 9: 1 (BC). When compared to β-tricalcium phosphate for GBR with collagen membrane in rat calvaria, BC demonstrated small amount of new bone formed besides presence of structures similar to bone tissue, but without osteocytes, suggesting to be grafts remnants. They also observed that, after 10 weeks, BC did not develop complete closure of defect 37  After histological analysis, our results showed that percentage of non-mineralized areas in ranged from 33.3% to 45.1%. According to Bonucci 39 , the main difference between compact and spongy bone depends on its porosity. The percentage of voids related to osteon channels, osteocyte canaliculi and Volkmann channels will determine this porosity. In compact bones, voids range from 5 to 30%. If presence of voids is greater than 30%, bone may be characterized as spongy, with density ranging from 0.1 to 0.9 g / cm 3 . Results of BMD and percentage of non-mineralized areas suggest that new bone type was spongy for all groups, regardless collagen membrane and pore size. In addition, gaps of empty osteocytes were also observed, suggesting remaining graft after 30 days, according to Kato et al. 37 .
When evaluating BC/femur interface, it was observed that 73.5% to 88.3% of this interface was complete, and there was no statistically significant difference between groups.
This suggests that, even in exophytic areas, BC may be integrated into femoral bone tissue.
Also, 33% of samples with pore size > 1 mm and 20% of samples with por size < 1 mm presented isolated formation of bone tissue in concentric lamellae (osteon). It was suggested that, in order to obtain osteon formation, it would be ideal to use meshes with pore size of at least 150 µm 25 . In our results, even in low proportion, it was possible to observe osteons formation in intermediate stages, in all pore size, with or without collagen membrane.
This article is protected by copyright. All rights reserved.
Through in vivo µ-CT results, it was observed that there was no volumetric difference in new bone formed using different pore size. Numerically, P15 mesh showed a higher percentage of bone volume (30.9 ± 12.33), despite it's thickness, when compared to other groups (P300 = 26.8 ± 4.6, P175 = 27.5 ± 11.3, P85 = 22.3 ± 9.6). Rakhmatia et al. 9 also did not observe significant difference in volumetric parameters due to pore size after 8 weeks in rats. In this case, titanium mesh was used without bone graft in a 7 mm defect in rat calvaria.
Also, they observed that mesh thickness between 100 and 200 μm would be ideal for larger bone reconstructions, avoiding collapse within the defect.
In relation to BMD, it was observed that meshes with larger pore size presented higher density, and difference was statistically significant. Studies suggested that titanium meshes with larger pore size would have a lower mineral density, different from demonstrated by our results 9 . Numerically, Rakhmatia et al 9 observed density between 8 and 10 g/cm 3 in meshes with pore size between 20 and 100 μm in rats. Our values ranged from 0.04 to 0.11 g / cm 3 , and in P300 and P175 meshes density was 0.10 and 0.11 g / cm 3 (control), respectively, and 0.11 g / cm 3  interference of 30 μm thickness. We suggest that, when using thinner meshes for large vertical bone reconstructions, screw tents should be used in order to avoid collapse and reduction of bone gain 40 . Different rates of bone formation and BMD observed in the literature studies and our results suggest that mesh pore size may not be the only factor interfering new bone formation. Factors such as physical characteristics of material, thickness, chemical composition, biocompatibility 41 , besides pore shape, manufacturer, vascularization and use of bone graft to serve as a framework.

Conclusions
Despite the experimental study limitations and supported by the findings of this investigation, data demonstrated that combined usage of collagen membrane in guided bone regeneration procedures with titanium mesh did not show improvements in new bone quality in rat femur model. However, titanium mesh pore size specifications have influence in bone quality.  This article is protected by copyright. All rights reserved. Tables   GROUPS   TISSUE  VOLUME  (TV, mm  Bone mineral density (g/cm 3 ). Significant differences between meshes pore size > 1mm and meshes pore size < 1 mm (Anova, Tukey, p <0.05). Level of significance was considered 95% (Anova, Tukey).