Galea and periosteum flap filled with bone fragments in rabbits 1

PURPOSE: To study the bone viability of a vascularized galea and periosteum flap filled with bone fragments, as a substitute of the bone graft in facial reconstructive surgery. METHODS: Forty rabbits were studied, and divided in two groups. One had a simple galea and periosteum flap done and the other had the same flap done and filled with bone fragments of the calvaria. The bone formation was evaluated by radiographies, macroscopic and microscopic analysis. RESULTS: The bone neoformation in both groups with differences in bone morphology and structure especially at histological analysis. CONCLUSION: This study demonstrated osseous formation in both groups of galea and periosteum flaps, with and without bone fragments.


Introduction
Facial osseous defects caused by tumor resection, congenital malformation or trauma need reconstruction to recover the functional aspects of these lesions, like speech, breath, swallow, and also the aesthetic aspect of these patients [1][2][3][4][5] .
Facial tumors that demand extensive resection and even adjuvant treatments like radiotherapy, lead to function deformities and social problems because of the deformed appearance.The radiotherapy, necessary to complete the cancer treatment, interferes at the fracture healing and remodeling because it stimulates apoptosis and affects the blood supply 6,7 .
Congenital malformations like Franceschetti, Crouzon and Goldenhar Syndromes, facial clefts or mandible hypoplasia are even important as they need osseous tissue to reconstruct the defects.These malformations interfere at the social development of these children 8,9 .
Traumas are other important and usual reason of facial defects.Many times, only fracture correction is not enough to treat these patients.Some of them can lose an extensive part of soft and osseous tissue that demands more specific reconstructive techniques.
The bone graft is the most used procedure because it is autologous, has no rejection, but it has variable absorption rate at the post operative that may interferes at the final result.Besides, it leads to a donor site morbidity 2,5,11,[13][14][15][16][17][18] .
The temporoparietal fascia flap has demonstrated its adaptability to various reconstructive defects.It is a thin, pliable and resistant coverage to cartilage or osseous tissue and has an important blood supply by the superficial temporal vessels.It has a vascular pedicle and its length allows the flap to reach the ipsilateral defects of the orbitomaxillary region, oral cavity and mandible 19 .
Free flaps are used in difficult reconstructions.They provide reliable, well-vascularized soft tissue and less tissue absorption.However, its complications like arterial or venous thrombosis can lead to a total flap necrosis.Besides, it demands a proper medical team trained on microsurgery technique and special surgical tools 2,3,7,14,16,17,20 .
Alloplastic materials such as titanium, polymers, porous ceramics, hydroxyapatite, collagen sponges or hydrogels can also be used in reconstructive procedures, but there are risks like infection, they do not fully osseointegrate, can become unstable with cranial growth, leads to inflammatory response as a foreign body reaction and may occur extrusion 10,[21][22][23][24] .
Recent advances in tissue-engineering techniques have enabled new procedures to be developed for bone regeneration, and these procedures have the potential to improve on the present clinical strategies.Mesenchymal stem cells were differentiated into connective and mesenchymal tissue by Caplan 26 and Pittenger et al. 27 .These cells neoformation and regeneration capacity, special in craniofacial structures are natural.The adult mesenchymal cells derived from bone marrow, subcutaneous tissue, muscle, dental pulp, synovia and umbilical cord are potential donor for cellular therapy.The mesenchymal stem cell is multipotent and differentiates into osteoblasts, chondrocytes, adipocytes and myocytes 2,3,6,11,[25][26][27][28][29][30] .
However, the tissue-engineering techniques demand a specific laboratory, materials and prepared professional team.
It was imagined a galea and periosteum flap with a vascular pedicle and filled with bone fragments, which could be used at osseous facial deformities with a few donor area morbidity.

Methods
The Forty male, Norfolk rabbits of 70 days of life with 1730g to 3915g (mean 2655.75g± 179.29), were divided into two groups.
Control Group (CG), with ten animals, had a galea and periosteum cylinder flap done at the right side of calvaria and fixed at the frontal residual periosteum (Figure 1).The treated Group (TG), with 30 animals, had the same flap done and filled with fragments of bone harvested with an nº 23 surgical blade, from the other side of the cortical calvarial bone (Figure 2).The flap was fixed at the same place of the other group.
The flap was ressected and sent to histological analysis, stained with hematoxylin and eosin and Masson trichrome.The analysis was done with Leica DMLS microscope.A cellular qualitative analysis was done and the area and perimeter of the formed bone was measured.
The statistical analysis used chi-square tests and, when necessary, Fischer test.To compare the animal weight and bone formation, it was used logistic regression.To evaluate the examiners analysis was used the index of coincidence (Kappa) and ANOVA test was used to compare the area and perimeter values.
It was considered p<0.05.The analysis was done in S.A.S 9.3 software for Windows.

Results
All animals were submitted to surgical procedures and only one from TG, died at the anesthesia for the 30 th day radiography.
Thirty-nine rabbits had a radiographic exam at 30 days postoperative.The analysis had an important variation between the examiners.Control Group had from one animal (10%) to 90% of radiopacity at the lateral exam (Table 1).The examiner 1 and 2, and examiner 1 and 3 had low coincidence (kappa -0.02).
The coincidence of examiner 2 and 3 was high (kappa -0.86).At 60 th days radiography, 72.41%, 89.65% e 93.10% of the animals had radiopacity image, all at the same lateral incidence (Table 4).The coincidence of examiner 1 and 3, was low (kappa -0.10), 1 and 2 was considerable (kappa -0.25) and the coincidence of examiner 2 and 3 was substantial (kappa -0.78).No image was identified at dorsoventral view.There was no significant difference of bone formation between the groups at 30 days (p=0.2282,p=1.0000, p=1.0000).After 60 days, the lateral radiographic images demonstrated significant higher rates of radiopacity at TG (p=0.0266) for only one examiner.The other two examiners had no statistical difference between 30 and 60 days radiographies (p=0.5558,p=1.0000).
At the 60 days surgical procedure, all animals from both groups demonstrated good vascularized flaps and no tissue absorption, which suggests good vascular perfusion.
Eleven flaps, from the total thirty-nine animals, demonstrated a yellow tissue inside them; ten were from TG and one from CG.It was thin and spiky in eight and soft and irregular in three of them.The one from CG was soft and irregular (Figures 5 and 6).The histological analysis of CG demonstrated bone image with an eosinophilic material, osteblasts and fibroblasts in five samples (50%).The bone aspect was irregular and trabecular with osteoblast proliferation around it (Figure 7).At TG histological analysis, twenty samples (68.97%) had eosinophilic material with osteoblasts in more organized and regular form.Four images (20%) had vascular formation inside it (Figure 8).However, bone neoformation had no significant difference between the two groups (p=0.4459).
The osseus image area and perimeter were measured in all histological samples.At CG, the area was 2.447 to 321.751 mm 2 (mean = 90.512mm 2 ) and perimeter was 0.225 to 5.04 mm (mean = 1.82 mm).In TG, the area was 20.604 to 3041.709 mm 2 (mean = 450.419mm 2 ) and the perimeter was 0.79 to 18.448 mm (mean=3.59mm).There was no significant difference between the group's values (p=0.3979).

Discussion
The choice to use rabbits as experimental animals was made because it has an adequate calvaria size, larger than rats and it has a vascularized tissue by the temporal artery and vein.The rabbit has been used as experimental animal to bone defects and cranial reconstruction studies 24,31 .
The For this reason, it was used bone dust, fragments of the cortical calvarial bone, because it has osseous cells and essential proteins to bone formation induction, and minimum donor site morbidity.
The vascularized calvarial flaps are better than bone grafts because it has less absorption, better osseous integration and less infection rate 17,34 .
The periosteum is a vascularized membrane that covers the bone and is important at osseous healing.Studies demonstrated that the periosteum induces cartilage and bone formation, depending on the oxygen tension and protein levels.The fibroblasts from the outer layer of the periosteum and the osteoblasts from the inner layer provide a collagen matrix important at the healing process 9,32,33 .
The choice to use periosteum and calvaria flap was done because of the temporal vascular pedicle, which is flexible, can reach local areas and can be shaped in different forms, besides the regeneration capacity of the periosteum.
Macroscopic analysis demonstrated 100% vascularized tissue, no absorption of the flap and images inside the flap that suggested bone neoformation, one at CG (10%) and ten at TG (34.48%).
The radiographic results suggest osseous formation at the top of the calvaria where the flap was made, in lateral position images.At the dorsoventral view, there was no suggestive image of neoformed bone, may be because the osseous superposition of the rabbit calvaria that makes it difficult to differentiate osseous structures.
The histological analysis confirmed bone formation in both groups.However, the structure of this neoformed osseous tissue was different.In CG, 50% of the samples had irregular eosinofilic images with osteoblasts, which characterized a trabecular bone.And in TG, the eosinofilic image was organized, with regular borders and osteoblasts, characteristics of a more mature bone formation.
The area and perimeter had a large variety of values and did not demonstrate statistical difference between the groups.For this reason the bone was not considered by quantity but it was considered by quality and histological differences.
At this study, it was possible to demonstrate bone formation in vascularized flaps of galea and periosteum of rabbitt's calvaria and it was possible to differ tissue formation in two kinds of surgical techniques.However, the real viability analysis of this galea and periosteum flap filled with bone fragments could not be proved and demands more studies to reach a real conclusion.

Conclusion
This study demonstrated osseous formation in both groups, with and without bone fragments.

FIGURE 2 -
FIGURE 2 -A.Galea and periosteum cylinder confection and bone fragments harvested with an nº 23 surgical blade.B. Bone dust confection.

FIGURE 3 -
FIGURE 3 -A.Normal dorsoventral radiography.B. Lateral radiography of a normal animal.

FIGURE 5 -FIGURE 6 -
FIGURE 5 -Macroscopic image of a spiky tissue formation inside the flap.
morcelized bone graft used into vascularized flaps demonstrated that the osseous cells and proteins from growth factor family are necessary to progenitor cells induction of the periosteum.It is considered an ideal biomaterial to bone neoformation because it is a complement of growth factors stored in the noncalcified matrix of morcellized bone graft, such as bone morphogenetic proteins and proteins of the fibroblast growth factor family, induce progenitor cells within the periosteum to an osteogenic phenotype and result in new bone formation within the chambers 32,33 .
Radiography analysis at 30 days postoperative of Control Group.Yes-radiopaque image at right calvarial side.No-Any radiopaque image at right calvarial side.

TABLE 2 -
Radiopacity at lateral view at 60 th day x-ray of Control Group animals.Yes-radiopaque image at right calvarial side.No-Any radiopaque image at right calvarial side.

TABLE 3 -
Radiography analysis at 30 days postoperative of Treated Group.Yes-radiopaque image at right calvarial side.No-Any radiopaque image at right calvarial side.

TABLE 4 -
Radiopacity at lateral view in 60 th day x-ray of Treated Group animals.Yes-radiopaque image at right calvarial side.No-Any radiopaque image at right calvarial side.