Bone augmentation by octacalcium phosphate and collagen composite coated with poly‐lactic acid cage

Abstract Objective Although octacalcium phosphate and collagen composite (OCP/Col) has demonstrated excellent bone regeneration, it has never achieved bone augmentation. The present study investigated whether it could be enabled by OCP/Col disks treated with parathyroid hormone (PTH) and covered with a poly‐lactic acid (PLA) cage. Materials and methods The prepared OCP/Col disks with three different types of PLA cages (no hole, one large hole, several small holes) were implanted into subperiosteal pockets in rodent calvaria. Histological, and histomorphometric analyses were conducted at 12 weeks after implantation. Results Implants with all PLA cage variants achieved sufficient bone augmentation, and analyses showed that new bone was formed from the original bone and along the PLA cage. While the PLA cage variant with no holes sporadically evoked new bone formation even at the central area of the roof of the PLA cage, the PLA cage variants with holes had no new bone in the area of the hole or beneath the periosteum. Conclusions These results suggest that sufficient bone augmentation could be achieved by treating the OCP/Col disks with PTH and covering them with a PLA cage, and periosteum might not have been involved in the bone formation in this experiment.


| INTRODUCTION
After tooth extraction, the residual alveolar ridge generally has limited bone volume because of bone resorption (Tallgren, 2003). Atrophic alveolar ridges might impede proper placement of dental implants in their longterm functional position, with an acceptable aesthetic profile of the final prosthesis (Buser, Martin, & Belser, 2004). To overcome these challenges, autologous bone grafting has been used for the reconstruction and augmentation of such alveolar ridges to facilitate implantation (Clavero & Lundgren, 2003). However, collecting autologous bones damages the healthy body since there are problems associated with the use of autologous bone in terms of donor site pain, morbidity, and infection (Jensen & Terheyden, 2009). In contrast, hydroxyapatite (HA) and β-tricalcium phosphate (β-TCP) have been examined experimentally as alternatives for autologous bone augmentation (Pinholt, Ruyter, Haanaes, & Bang, 1992;Zerbo et al., 2004). Yet, autologous bone grafting is still considered the clinical "gold standard" and the most effective method for bone regeneration (Hjorting-Hansen, 2002).
However, OCP/Col discs alone did not enhance appositional bone formation because of compression stress under the periosteum (Suzuki et al., 2009). Another study has shown strong indications that supporting the OCP/Col discs with a polytetrafluoroethylene (PTFE) ring with higher modulus than OCP/Col alleviates mechanical stress and enables bone formation (A. . It was also suggested that OCP/Col with the single local administration of teriparatide (TPTD), a crucial regulator of calcium and phosphate metabolism and functions (Sibai, Morgan, & Einhorn, 2011), enhanced bone regeneration in a rodent calvarial critical-sized bone defect more than OCP/Col per se (Kajii et al., 2018). Therefore, it was hypothesized that bone augmentation could be achieved by covering the parathyroid hormone (PTH) treated OCP/Col discs with poly-lactic acid (PLA) cages. To study this hypothesis, these samples were implanted into subperiosteal pockets of rodent calvarium. Additionally, it was also investigated if bone regeneration was influenced by the condition of contact between the periosteum and OCP/Col. Previous studies have indicated that bone augmentation instigated by OCP/Col is predominantly initiated from the surface of original bone side and no bone formation has been observed from the periosteal side (A. Matsui, Anada, et al., 2010;Suzuki et al., 2009), even though it has been suggested that periosteal cells formed bone (Nakahara et al., 1990).

| Preparation of OCP/Col
The preparation of octacalcium phosphate and collagen composite (OCP/Col) was described previously (Kajii et al., 2018). Briefly, OCP was prepared by direct precipitation, and sieved granules (particle sizes 300-500 μm) were produced. Collagen was prepared from NMP collagen PS (Nippon Meat Packers, Tsukuba, Ibaraki, Japan), which is a lyophilized powder of pepsin-digested atelocollagen isolated from porcine dermis. Sieved granules of OCP were added to concentrated collagen and mixed, and the weight percentage of OCP in OCP/Col was 77%. The OCP/Col mixture was then lyophilized, and a disk was molded (diameter 9 mm, thickness 1.5 mm). OCP/Col disks were prepared by dehydrothermal treatment (150 C, 24 hr) in a vacuum drying oven. OCP/Col disks were sterilized using electron beam irradiation.

| Implantation procedure
Twelve-week-old male Wistar rats (SLC Corp., Hamamatsu, Shizuoka, Japan) were used. The principles of laboratory animal care as well as national laws were followed. All procedures were approved by the Animal Research Committee of Tohoku University (2016BeA-001). Experimental rats were anesthetized with intraperitoneal dexmedetomidine hydrochloride (0.05 mg/kg), midazolam (0.12 mg/kg), and butorphanol tartrate (0.15 mg/kg). First, an accurate skin incision was made on the forehead region and the dissected skin was ablated. Then, periosteum of the calvarium was ablated after the incision reached the skull, and a subperiosteal pocket was prepared on the calvarium (Figure 2a). An OCP/Col disk treated with PTH (1.0 μg/0.1 ml) was covered with a PLA cage (sample), and it was placed on the calvarium (Figure 2b).
The sample was inserted into the created subperiosteal pocket by moving it backward on the calvarium and placed under the periosteum so that it was in contact with the bone surface of the calvarium ( Figure 2c). After that the sample was covered with the ablated periosteum and the incision was sutured with absorbable thread, to prevent the sample from escaping the subperiosteal pocket ( Figure 2d).
Finally, the dissected skin was repositioned and sutured with silk thread. Each five experimental rats per group were randomly divided into three groups (N, B, and S groups), and in rats in each group were fixed 12 weeks after implantation (n = 5 × 3).

| Micro-CT examination
4 and 12 weeks after implantation, in vivo microcomputed tomography (CT) analysis of the rat calvarium was performed using an X-ray CT system (Latheta LCT-200; Hitachi Aloka Medical, Tokyo, Japan) after intraperitoneal injection of sodium pentobarbital (50 mg/kg), as described previously (Kajii et al., 2018). The calvariums were scanned continuously in increments of 120 μm, with pixel size 60 μm. CT images were acquired using the following parameters: 50 kVp tube voltage, 500 μA tube current. After completion of the CT analysis at 12 weeks, rats were euthanized by an intraperitoneal injection of an overdose of sodium pentobarbital. After sacrifice, the implants were resected together with the surrounding bones and tissues and fixed in 4% paraformaldehyde in 0.1 M phosphate-buffered saline, pH 7.4.

| Tissue preparations and quantitative micrograph analysis
The samples were decalcified in 10% ethylenediaminetetraacetic acid in 0.01 M phosphate buffer, pH 7.4 at 4 C for 4-6 weeks after radiographs had been taken. Specimens were embedded in paraffin, and F I G U R E 1 Poly-lactic acid (PLA) cages of the N, B, and S groups. The PLA cage was 10 mm outside diameter, 8.5 mm inside diameter, 2.5 mm outside height, and 1.5 mm inside height. The cage covered the upper surface and side of the octacalcium phosphate and collagen composite (OCP/Col) disc. The PLA cages had three variants (no hole (N), one large hole (B): φ6 mm × 1, and several small holes (S): φ1 mm × 7), Bars: 3 mm F I G U R E 2 Implantation procedure of samples. The periosteum of the calvarium was ablated after the incision reached the skull, and a subperiosteal pocket was prepared on the calvarium (a). An octacalcium phosphate and collagen composite (OCP/Col) disk treated with parathyroid hormone (PTH) (1.0 μg/0.1 ml) was covered with a poly-lactic acid (PLA) cage (sample), and it was placed on the calvarium (b). The sample was inserted into the created subperiosteal pocket by moving it backward on the bone surface and placed under the periosteum so that it was in contact with the bone surface of the calvarium (c). After that the sample was covered with the ablated periosteum and the incision was sutured with absorbable thread to prevent the sample from escaping the subperiosteal pocket (d). Bars: 3 mm the center was extracted, and sectioned coronally. Sections were stained with hematoxylin and eosin, and photographs were taken using a photomicroscope (Leica DM2500, Leica Microsystems Japan, Tokyo, Japan). Histomorphometrical analysis of the histological sections of

| Statistical analysis
At 12 weeks after implantation, n-Bone% was analyzed statistically using Excel v. X. (Microsoft Co., Redmond, WA). All values are reported as mean ± SD. The χ 2 test was used to investigate whether each group had a normal distribution, and Bartlett's test was employed to examine the homogeneity of variance across samples.
One-way analysis of variance or the Kruskal-Wallis test was used to compare means among groups. Significance was accepted at p < .05. If significant differences were detected in the mean values, Tukey-Kramer or Scheffe's multiple comparison analysis was used as a post hoc test.

| Micro-CT analysis
The top side of the images indicate the skin side, whereas the bottom side shows the original bone side. In the N and S groups, radiopacity in the OCP/Col implanted area was more abundant than radiolucency in the area at 4 weeks after implantation. The radiopacity was dominant in marginal areas and near the original bone and increased with time. After 12 weeks, the increased radiopacity was greater in the marginal area than in the central area. In the B group, radiolucency in the OCP/Col implanted area was more abundant than radiopacity in the area at 4 weeks after implantation. Although the radiopacity was dominant in marginal areas and increased with time, it was lower than the radiopacity in other groups. After 12 weeks, the radiopacity had increased in the OCP/Col implanted area, but was still lower than in the other groups (Figure 4). ). The percentage of newly formed bone (n-Bone%) was calculated as the area of newly formed bone/the area surrounded by the cage (c, d) × 100 for all areas and the area in each region F I G U R E 4 Micro-computed tomography (CT) at 4 and 12 weeks after implantation. The top of the images indicate the skin side, and the bottom is the original bone side. In the N and S groups, radiopacity in the octacalcium phosphate and collagen composite (OCP/Col) implanted area was more abundant than radiolucency in the area at 4 weeks after implantation. After 12 weeks, the increased radiopacity in marginal area was greater than that in central area. In the B group, radiolucency in the OCP/Col implanted area was more abundant than radiopacity in the area at 4 weeks after implantation. After 12 weeks, the radiopacity had increased in the OCP/Col implanted area, but it was still lower than other groups. Bars: 3 mm F I G U R E 5 Histological results at 12 weeks after implantation for the total area of experimental group. The top of the figure indicates the skin side, and the bottom is the original bone side. It was observed bone augmentation in all groups (N, B, and S), and newly formed bone was extended from the original bone toward the skin side and developed along the poly-lactic acid (PLA) cage. Bars: 3 mm; B, newly formed bone; P, PLA cage; F, fibrous tissue  (Figures 5 and 6).

| Histomorphometrical examination
The n-Bone% of the total area in each section (LM, LC, UM, and UC) of the experimental groups is shown in Figures 7 and 8 and Table 1.
Although the n-Bone% in total area of the S group (52.0 ± 7.4%) was higher than those of the N group (50.3 ± 13.4%) and B group (41.0 ± 7.3%), there was no significant difference among these groups ( Figure 7). In every group, the n-Bone% of LM was highest, followed in order by LC, UM, and UC. In the N group, the n-Bone% of UC (23.0 ± 15.2%) was significantly lower than those of LM (72.5 ± 14.2%) and LC (60.5 ± 13.2%). In the B group, the n-Bone% of UC (13.4 ± 7.5%) was significantly lower than those of LM (64.6 ± 13.4%), LC (55.4 ± 8.5%), and UM (34.0 ± 10.7%). In addition, the n-Bone% of UM (34.0 ± 10.7%) was significantly lower than those of LM (64.6 ± 13.4%) and LC (55.4 ± 8.5%). In the S group, the n-Bone% of UC (27.6 ± 11.6%) was significantly lower than those of LM F I G U R E 7 Quantitative analysis of newly formed bone of total area of experimental group. Although n-Bone% of S group in total area was higher than those of N group and B group, there was no significant difference among these groups F I G U R E 8 Quantitative analysis of newly formed bone in each section by experimental group. In every group, the percentage of newly formed bone (n-Bone%) in the lower marginal (LM) area was highest, followed in order by lower central (LC), upper marginal (UM), and upper central (UC). In the N group, the n-Bone% of UC was significantly lower than those of LM and LC. In the B group, the n-Bone% of UC was significantly lower than those of LM, LC, and UM. Additionally, the n-Bone% of UM was significantly lower than those of LM and LC. In the S group, the n-Bone% of UC was significantly lower than those of LM and LC, and the n-Bone% of UM was significantly lower than that of LM T A B L E 1 Quantitative analysis of the percentage of newly formed bone (n-Bone%) for the total area and in each section by experimental group (75.3 ± 10.4%) and LC (61.2 ± 9.1%), and the n-Bone% in the S group of UM (42.7 ± 13.0%) was significantly lower than that of LM (75.3 ± 10.4%) (Figure 8).

| DISCUSSION
This study confirmed bone augmentation of nearly double the thickness of original bone, when PTH treated and PLA covered OCP/Col discs were implanted into the subperiosteal pocket on a rodent calvarium. Although this study provided different bone augmentation models, it surpassed the results of previous studies of bone augmentation using OCP/Col (A. Matsui, Anada, et al., 2010;Suzuki et al., 2009). It was reported that single local administration of TPTD with OCP/Col promoted bone regeneration in rodent calvarial defects more than OCP/Col without TPTD (Kajii et al., 2018). Administration of TPTD is expected to accelerate remodeling of newly formed bone by enhancing osteogenic and osteoclastic activities (Morimoto et al., 2014), and thus adding TPTD to OCP/Col is hypothesized to promote In all experimental groups, the radiopaque area increased with time on the original bone side and in the marginal area from 4 to 12 weeks after implantation. In the central area of the skin side, radiolucency was dominant in B group, whereas radiopaque and radiolucent areas were mixed in the N and S group at 12 weeks after implantation. However, it was reported that it would be more suitable to apply histomorphometric analysis of newly formed bone than radiomorphometric analysis by micro-CT images (Kajii et al., 2018) since it is difficult to distinguish newly formed bone and converted apatite by radiographic examination. Therefore, histomorphometric analysis was prioritized in this study. In all the experimental groups, the newly formed bone extended from the original bone toward the skin side, and was abundant on the original bone side and in the marginal area. It has previously been shown that OCP acts together with osteoblasts, bone lining cells and their closely committed progenitors on the original bone to enhance bone formation from the surface of the original bone . Thus, it can be assumed that in this study osteoprogenitors might have interacted with OCP/Col and consequently enhanced bone augmentation. Additionally, using a PTFE ring to cover the OCP/Col, has been shown to enhance appositional bone formation by alleviating mechanical stress on OCP/Col (A. Matsui, Anada, et al., 2010). Since some of the new bone formation inside the PLA cage occurred in contact with the biocompatible PLA, it is likely that bone formation was initiated by osteoprogenitors that had invaded into the OCP/Col from the blood stream (Miettinen, Makela, Vainio, Rokkanen, & Tormala, 1992 Matsui, Anada, et al., 2010). In contrast, fibrous tissue invasion from the small holes in the S group may have had partial influence on delaying new bone formation. Although it was previously reported that the periosteum was involved in new bone formation (Nakahara et al., 1990), no bone had formed near the hole area of the PLA cage. This suggests that, in case of this experiment, periosteum might not have been involved in bone formation. Although this study demonstrated sufficient bone augmentation initiated by OCP/Col treated with PTH and covered with a PLA cage, PLA remains in the body for a long time. Additionally, the bone augmentation using this conventional OCP/Col (OCP/Col in this study) was insufficient because of the poor mechanical properties of this material (A. Matsui, Anada, et al., 2010). To achieve bone augmentation using OCP/Col itself, a new OCP/Col should be developed to maintain the shape and improve the mechanical properties after implantation.

ACKNOWLEDGMENTS
This study was supported in part by JSPS KAKENHI Grant Numbers 16H03159, 16K11741, and 18K19891.

CONFLICT OF INTEREST
One of the authors (S.K.) has obtained patents on OCP/Col in Japan (#5046511) and combination of calcium phosphate containing porous composite and PTH in Japan (#6094716).