Immunohistochemical comparison of lateral bone augmentation using a synthetic TiO2 block or a xenogeneic graft in chronic alveolar defects

Abstract Objectives To evaluate osteogenic markers and alveolar ridge profile changes in guided bone regeneration (GBR) of chronic noncontained bone defects using a nonresorbable TiO2 block. Materials and Methods Three buccal bone defects were created in each hemimandible of eight beagle dogs and allowed to heal for 8 weeks before GBR. Treatment was assigned by block randomization: TiO2 block: TiO2‐scaffold and a collagen membrane, DBBM particulates: Deproteinized bovine bone mineral (DBBM) and a collagen membrane, Empty control: Only collagen membrane. Bone regeneration was assessed on two different healing timepoints: early (4 weeks) and late healing (12 weeks) using several immunohistochemistry markers including alpha‐smooth muscle actin (α‐SMA), osteopontin, osteocalcin, tartrate‐resistant acid phosphatase, and collagen type I. Histomorphometry was performed on Movat Pentachrome‐stained and Von Kossa/Van Gieson‐stained sections. Stereolithographic (STL) models were used to compare alveolar profile changes. Results The percentage of α‐SMA and osteopontin increased in TiO2 group after 12 weeks of healing at the bone‐scaffold interface, while collagen type I increased in the empty control group. In the defect area, α‐SMA decreased in the empty control group, while collagen type I increased in the DBBM group. All groups maintained alveolar profile from 4 to 12 weeks, but TiO2 group demonstrated the widest soft tissue contour profile. Conclusions The present findings suggested contact osteogenesis when GBR is performed with a TiO2 block or DBBM particulates. The increase in osteopontin indicated a potential for bone formation beyond 12 weeks. The alveolar profile data indicated a sustained lateral increase in lateral bone augmentation using a TiO2 block and a collagen membrane, as compared with DBBM and a collagen membrane or a collagen membrane alone.


What is known
Lateral bone augmentation in chronic alveolar defects using a bone graft material usually leads to contact osteogenesis. Histological analysis may be used to describe the morphological situation, but gives limited information of the potential for bone formation.

What this study adds
Immunohistological data obtained by microtome sectioning MMA embedded samples indicates potential for further bone formation beyond 12 weeks healing.

| INTRODUCTION
Guided bone regeneration (GBR) employs a membrane as a mechanical barrier to avoid soft tissue involvement in the healing process. Thereby, the osteogenic potential that achieves bone augmentation arises from the bony defect's walls. The standard protocol commonly combines graft material with a membrane to create space for new bone formation and avoid soft tissue infiltration. Although some clinical studies have shown predictable bone gain, 1,2 others have reported less bone formation when using a graft material compared with when using the membrane alone. [3][4][5][6] In a previous in vivo experimental study, 7 GBR with a collagen membrane alone was compared with deproteinized bovine bone mineral (DBBM) and a ceramic TiO 2 scaffold. Less bone formation was observed in the TiO 2 and DBBM groups when compared with membrane alone group at the final follow-up time point after 12 weeks of healing. However, the groups using bone replacement grafts demonstrated increased volumetric lateral bone augmentation. In this study, however, these findings were assessed by microcomputed tomography and histomorphometry, where information could not be obtained on the osteogenic dynamics during this observation period.
Osteogenesis during GBR undergoes a complex process of finetuned coordinated phases. Initially, an inflammatory phase occurs where leukocytes, including macrophages, are recruited. Subsequently, new blood vessels form and osteoblasts deposit an extracellular matrix. A matrix maturation phase then follows before the final mineralization phase, where osteoblasts remodel woven bone into mature lamellar bone. 8,9 The different stages of osteoblast growth and differentiation can be identified either by specific gene expression or by quantifying protein secretion using histochemical methods. In the proliferation phase, there is a characteristic peak in collagen type 1 during the formation of bone extracellular matrix. Subsequently, during the matrix maturation phase, collagen type I decreases and osteopontin and osteocalcin increase, reaching their maximal expression during the mineralization phase. 10 Other histochemical markers like tartrate-resistant acid phosphatase (TRAP) and alpha-smooth muscle actin (α-SMA) represent biological cues for osteoclast activity and blood vessel formation, respectively. Hence, the quantification of these markers by histochemical analysis can study the stages of bone development, including the required neoangiogenesis and bone remodeling processes.
Therefore, the primary aim of this study was to assess the osteogenic potential by evaluation and quantification of osteogenic markers by immunohistochemistry (IHC), when using a bioabsorbable membrane alone compared with the use of either an additional TiO 2 block or DBBM particulates as bone replacement grafts. The secondary aim was to assess the changes in the alveolar ridge soft tissue profile from baseline to 4 weeks and 12 weeks by histomorphometry.

| Materials
A timeline of the study design is shown in Figure 1A. Porous ceramic TiO 2 scaffold blocks were produced by foam replication. Teeth extraction and defect creation were performed according to the protocol of Sanz and colleagues 11 Three standardized defects were created in each hemi mandible to a standardized box shape measuring 10 mm mesiodistally, 10 mm apicocoronally, and 5 mm buccolingually using bone burs under copious saline irrigation, and left to heal for 8 weeks prior to GBR procedures. After 8 weeks healing, GBR surgery was performed on the cortical bone of the one-wall defects. All recipient sites were perforated with round burs, treated with GBR materials ( Figure 1B)   At the allocated healing times the animals were euthanized with a lethal dose of sodium pentobarbital (40-60 mg/kg/i.v., Dolethal, Vetoquinol, France) and their mandibles were dissected and fixed in formalin.

| Histological preparation and histomorphometric analysis
Samples were dehydrated in an ascending series of alcohol and xylene baths before embedding in methyl methacrylate and polymerizing at F I G U R E 1 (A) Timeline of the study design. (B) Guided bone regeneration (GBR) procedure on noncontained defects. Showing the anterior defect with deproteinized bovine bone mineral particulates, middle defect with a TiO 2 block secured with a fixation screw and the empty posterior defect. Cortical perforations were performed at all recipient sites. All sites were covered with a collagen membrane stabilized by pin fixation. À20 C. The resulting embedded defect sites were divided into two halves. One was allocated for microcomputed tomography and undecalcified histomorphometry, and was utilized for the recently published results. 7  Movat Pentachrome stain was used to quantify collagen. 13 Von Kossa/Van Gieson staining was used to quantify the extracellular matrix mineralization. 14 Histomorphometry was performed on slides scanned using an AxioScan Z1 (Carl Zeiss, Germany) and analyzed using ImageJ (ImageJ 1.53f51, National Institutes of Health, USA). The Trainable Weka Segmentation plugin for ImageJ was used to quantify the stained areas, as described by Malhan and colleagues 15 In these slides, three regions of interest (ROI) were chosen. (1) The buccal half of the alveolar bone, including the grafted area, measured from the tip of the alveolar crest and extending 10 mm apically (ROItot).
(2) An area representing the interface between bone and the graft expanded 200 μm in both buccal and lingual directions (ROI400 μm).
In the empty control group, the area between bone and soft tissue was measured, and graft materials were excluded if present. (3) An area was determined from the same interface as for the ROI400 μm but only expanded 20 μm in both buccal and lingual directions (ROI 40 μm).

| Enzyme histochemical and immunohistochemical preparation and analysis
Sections were deplastified prior to staining. To show TRAP activity, sections were incubated in Sodium Acetate buffer, Naphthol-AS-TR phosphate (N6125-1G, Sigma, Germany) and Sodium Tartrate (Merck, Germany) at 37 C for 60 min.
To study the blood vessel formation, α-SMA, osteopontin, osteocalcin, and collagen type-I were diluted in DAKO-Diluent (S 0809), 1:400, 1:400, 1:1200, and 1:1200, respectively. Collagen type-I, α-SMA, osteocalcin, and osteopontin staining were quantified as described for histomorphometry, using the same ROIs. The number of osteoblasts was counted manually in TRAP-stained sections at the bone interface of ROItot. Blood vessels were classified and quantified as circular, intermediate, or irregular by α-SMA in ROI400 μm.

| Alveolar profile measurements
Individualized impression trays were fabricated for each animal. Silicon impressions of the mandible were taken using a light/heavy putty (Elite HD+, Zhermack spa, RO, Italy) prior to GBR procedure and at the end of study. Cast models were poured with stone (Fujirock type 4, GC. Corp, Tokyo, Japan), then optically scanned using a desktop 3D scanner (Zfx Evolution Scanner, Zimmer Dental, Bolzano, Italy) to obtain STL files.
MeshLab 2022.02 was used to align the images. 16 Buccolingual crosssections were made at the middle of the defect and exported to ImageJ for analysis. ROI was defined as the buccal half of the mandible, from the crest and 6 mm apically or until the mucogingival border. The impression taken prior to GBR procedure was set as a baseline and changes in area were measured in 2 mm increments in a coronoapical direction. 17

| Osteocalcin
Osteocalcin intensity was similar across all groups and time points.
For example, a homogenous staining of the prestine alveolar bone was observed, and no osteocalcin stain was seen in the grafted area.

| Tartrate-resistant acid phosphatase
Osteoclasts were identified in 35 out of 46 samples within the ROI.

| DISCUSSION
This study demonstrated that GBR in chronic defects using a TiO 2 block led to an increase of α-SMA and osteopontin from 4 to 12 weeks at the bone-scaffold interface. In the DBBM group, an increase in collagen type I was observed for ROItot, while the negative control group showed decreased α-SMA in ROItot and increased collagen type I at ROI40 μm.
The findings suggest contact osteogenesis when GBR is performed with a TiO 2 block or DBBM particulates. The increase in osteogenic markers indicates potential for bone formation beyond 12 weeks. In addition, the alveolar profile data showed a sustained lateral increase using a graft material against negative control.
These findings were in accordance with previous studies demonstrating osteoconductivity of TiO 2 blocks in preclinical models. 7,18 In addition, osteoblasts' enhanced osteopontin and vascular endothelial growth factor secretion by osteoblasts on TiO 2 have also been demonstrated in cell cultures compared with a SiO 2 surface. 19 The increased osteopontin at the bone border and around graft materials coincides with previous reports. [20][21][22] Osteopontin is required for bone formation and facilitates osteoblast and osteoclast adhesion. 23 area. However, the results could also indicate a lower osteogenic activity at 12 than 4 weeks at negative control sites. Bone graft materials have been shown to delay initial bone regeneration as compared with empty control sites. 6,26 Faster initial healing at empty control sites covered by a collagen membrane was expected compared with DBBM-grafted and TiO 2 -grafted sites.
One advantage of using graft materials in GBR is space maintenance under the cell occlusive membrane. 27,28 It has been shown by microCT measurements that both DBBM particulates and TiO 2 blocks preserved the space 12 weeks after GBR procedure. This study corroborated the results in alveolar contour change, which implied the soft tissue profile adapted to the graft materials. The STL images also made it possible to compare baseline alveolar shape prior to surgery F I G U R E 6 (A) Superimposed stereolithographic images of TiO 2 sample 4 weeks after guided bone regeneration. The baseline in purple and 4 weeks healing in green. Area difference in ROI is illustrated in yellow. Shown distance from the alveolar crest divides the top, mid, and low segments. (B) Area differences for alveolar contour. Statistical significance denoted by asterisks: *p < 0.05, **p < 0.01, ***p < 0.001. DBBM, deproteinized bovine bone mineral with the shape after 4 and 12 weeks. A minor resorption was found at the top increment in the negative control group, as expected when a flap was raised and the bone crest exposed to perform GBR. 29 A tissue loss at the coronal part was also found for the DBBM group but not for the TiO 2 group. This may be attributed to the different handling and properties of materials used in the chronic, noncontained defects. For example, in the middle and lower part of the ridge, the TiO 2 group was significantly wider than the empty control group. The DBBM group was not statistically significantly wider than the empty control group. This was expected, as a block graft material is more stable than a particulate. Graft dislocation following wound closure may also contribute to the reduced alveolar width at the coronal portion, especially for the particulates. As shown in an in vitro study by Mir-Mari and colleagues, 30 compressive forces on the augmented sites could not be totally avoided, even though a clinically tension-free flap closure was achieved. The authors reported enhanced stability of the particulates by application of fixation pins or by the use of a block graft as compared with particulated bone substitutes. The effect of graft dislocation was found to be substantiated in one-wall bone defects as compared with self-contained defects. 31 The authors found GBR with additional membrane fixation resulted in higher volume stability than without fixation and even better stability when a titaniumreinforced membrane or a bone block was used. In this study, the collagen membrane was secured with metal pins. A reinforced membrane may have been beneficial for the particulate group; however, that would add another variable in the healing process. Additionally, DBBM with a resorbable collagen membrane is a commonly used and well-documented procedure for augmentation and therefore chosen as a positive control. 32 The results from the alveolar ridge profile measurements in this study were partly in agreement with reported findings from Di Raimondo and colleagues, 17  Schwarz and colleagues reported osteocalcin antigen reactivity in the connective tissue adjacent to DBBM and beta-tricalcium phosphate granules. This was not seen in this study, where a stable intensity from 4 to 12 weeks was observed in mineralized bone only. The empty control group had the highest intensity in ROItot, but no differences were found otherwise. The lower intensity seen in the DBBM and TiO 2 groups compared with this study's empty control group may indicate early osteogenesis as osteocalcin is a marker of the later stages. 35 The comparable intensity of osteopontin was found across the groups at 8 and 16 weeks by Cha and colleagues. Osteopontin was situated around bone borders and graft particulates, according to this study. However, in this study, a higher intensity was found in the empty control group than the TiO 2 group at 4 weeks for ROItot, and the TiO 2 group demonstrated increased intensity from 4 to 12 weeks in both ROI40 and ROI400 μm.
Cha and colleagues reported no different TRAP counts for the DBBM group as compared with the empty control group, in agreement with this study. The biphasic calcium phosphate bone substitute demonstrated a significant increase in TRAP count from 8 to 16 weeks, which was hypothesized to be due to the resorption and following calcium and phosphate release from the biomaterial. As this study used a nonresorbable scaffold, no change in TRAP count was anticipated for the TiO 2 group. In addition to the different models used by Schwarz and colleagues and Cha and colleagues, these studies also applied different methods for immunohistochemical analysis. When compared with a previous study, 7 where methyl methacrylate sections were prepared by cutting and grinding, the use of microtome sectioned samples in this study presented several benefits.
Above all, the 5 μm thin sections could be deplastified and decalcified after microtome sectioning. Decalcifying is usually performed on the bulk sample prior to paraffin embedding and sectioning. However, decalcification would not affect the ceramic TiO 2 scaffold and would be impossible to cut when placed in soft decalcified tissue. The present method described by Malhan and colleagues 15 allowed for histochemical staining of TiO 2 containing samples without the need for specialized equipment like laser microtomes. 37 By decalcifying 5 μm sections, the process was also quicker than bulk decalcifying. In addition, this technique yielded a higher number of sections as no material was lost by cutting and grinding. Ultimately, this may reduce the required number of animals, according to the principles of humane experimental technique. 38 In this study, the thinner sections also resulted in better image quality. As previously reported, a dense structure was observed as a dark substance between the graft materials. 7 With the thinner sections, a collagen network in the extracellular matrix was clearly identified from Movat Pentachrom and Von Kossa/ Van Gieson stain. However, there were challenges with the methacrylate infiltration. Some samples were not adequately fixated. As a result, some sections had to be excluded in the analyses.
This study results should be interpreted with care due to the experimental nature of the study as well as the limited number of animals. In addition, the heterogeneity in study designs for GBR makes a comparison between studies challenging. The low number of treatment groups was also a limitation, and a DBBM material in block configuration could have served as a more relevant control in this study design.
Further studies should also evaluate if the regenerated tissue obtained with TiO 2 blocks will be stable over time and allow a reliable osseointegration of implants. Finally, despite the indications of osteogenic differentiation, a longer observation time is required to confirm future bone formation.

| CONCLUSION
In conclusion, within the study's limitations, the findings suggest contact osteogenesis when GBR is performed with a TiO 2 block or DBBM particulates. The increase in osteopontin markers indicates potential for osteogenesis beyond 12 weeks in this model. However, the alveolar profile data indicated a sustained lateral increase in lateral bone augmentation using a TiO 2 block and a collagen membrane, as compared with DBBM and a collagen membrane or a collagen membrane alone.