Fabrication and Histological Evaluation of Porous Carbonate Apatite Block from Gypsum Block Containing Spherical Phenol Resin as a Porogen

The utility of carbonate apatite (CO3Ap) as a bone substitute has been demonstrated. The feasibility of fabricating macroporous CO3Ap was evaluated through a two-step dissolution–precipitation reaction using gypsum as the precursor and spherical phenol resin as the porogen. Porogen-containing gypsum was heated to burn out the porogen and to fabricate macroporous structures. Gypsum transformed into CaCO3 upon immersion in a sodium carbonate solution, while maintaining its macroporous structure. Next, CaCO3 transformed into CO3Ap upon immersion in a Na2HPO4 solution while maintaining its macroporous structure. The utility of the macroporous CO3Ap for histologically reconstructing bone defects was evaluated in rabbit femurs. After 4 weeks, a much larger bone was formed inside the macroporous CO3Ap than that inside non-macroporous CO3Ap and macroporous hydroxyapatite (HAp). A larger amount of bone was observed inside non-macroporous CO3Ap than inside macroporous HAp. The bone defects were completely reconstructed within 12 weeks using macroporous CO3Ap. In conclusion, macroporous CO3Ap has good potential as an ideal bone substitute.


Introduction
In orthopedic surgery, bone grafts have been used for the treatment of pseudarthrosis or bone defects due to bone fracture or tumor resection. Autograft is still the gold standard. However, although autografts possess good osteoconduction and osteoinduction and induce osteogenesis without causing an immunological response, more skin incision is required and bone harvest is limited [1][2][3][4]. Therefore, artificial bone substitutes are often used to fill bone defects in combination with or without autografts.
Hydroxyapatite (Hap; Ca 10 (PO 4 ) 6 (OH) 2 ) has been widely used as an artificial bone substitute because it exhibits high tissue compatibility and good osteoconduction. Contrary to autografts, HAp is hardly absorbed at the bone defect and remains there for a long time [5]. It should be noted that the inorganic component of the bone is not pure HAp but AB-type carbonate apatite (CO 3 Ap; Ca 10-a (PO 4 ) 6-b (CO) c ), which contains 4-9% carbonate ions (CO 3 2− ) in an apatitic structure [6].
In addition, resorption is known to be quicker with an increase in substitution of CO 3 2− at the A-site or OH site and/or at the B-site or PO 4 site [7]. Furthermore, compared to the crystal structure of HAp, the structure of CO 3 Ap more closely resembles bone apatite [8]. Thus, CO 3 Ap is expected to be an ideal bone substitute.

Fabrication of Cylindrical CO 3 Ap and HAp Blocks
Cylindrical porous CO 3 Ap blocks (6 mm in diameter and 3 mm in thickness) were fabricated using calcium sulfate hemihydrate (CaSO 4 ·1/2H 2 O, Wako Pure Chemical Industries Ltd., Osaka, Japan) and spherical phenol resin (LPS-C100; Lygnyte Inc., Osaka, Japan), with an average diameter of 100 µm by compositional transformation through dissolution-precipitation reactions, which converted CaSO 4 to CaCO 3 and CaCO 3 to CO 3 Ap. CaSO 4 ·1/2H 2 O powder and spherical phenol resin were mixed so that the proportion of spherical phenol resin was 30% and 40% by mass. The mixture or CaSO 4 ·1/2H 2 O powder alone was further mixed with distilled water at a water-to-powder ratio of 0.23. The paste thus prepared was packed into a split plastic mold (6 mm in diameter and 3 mm in height), and both sides were covered with glass plates and kept at room temperature for 3 h for hardening. The hardened sample was heated to 700 • C at a rate of 0.13 • C/min and maintained at 700 • C for 3 h to burn out the spherical phenol resin.
The CaSO 4 blocks were then immersed in a solution of 2 mol/L sodium carbonate (Na 2 CO 3 ) and 2 mol/L sodium hydrogen carbonate (NaHCO 3 ) adjusted to pH 9 at 90 • C for 24 h to fabricate CaCO 3 blocks through the dissolution-precipitation reaction using CaSO 4 blocks as the precursor. The solution is denoted as Na-H-CO 3 in the remainder of the text.
Next, CaCO 3 blocks were immersed in 1 mol/L disodium hydrogen phosphate (Na 2 HPO 4 ) at 80 • C for three weeks to fabricate CO 3 Ap blocks through the dissolution-precipitation reaction using CaCO 3 blocks as a precursor.
Cylindrical porous HAp blocks (6 mm in diameter and 3 mm in height) were fabricated as a control. HAp powder (HAP-200; Taihei Chemicals, Saitama, Japan) was mixed with spherical phenol resin porogens so that the mass proportion of the porogens was 40%. The mixture was pressed at 20 MPa with a stainless-steel mold using an oil pressure press machine (MT-50HD; NPa System, 3 of 16 Saitama, Japan). The compacts were heated at 0.13 • C/min to 1000 • C and maintained at 1000 • C for 3 h in an electronic furnace (SBV-1515D; Motoyama, Osaka, Japan) to burn out the porogens and to sinter the HAp powder. The mass content of spherical phenol resin in the original precursor is stated in parentheses. For example, CO 3 Ap (30%) indicates that the CO 3 Ap block was made using 30% spherical phenol resin by mass in the raw material, i.e., CaSO 4 ·1/2H 2 O.

Characterization of Samples
The composition of the samples was analyzed by X-ray diffraction (XRD), Fourier transform infrared (FT-IR) spectroscopy, and elemental analysis. XRD patterns were recorded using a powder X-ray diffractometer (D8 Advance A25; Bruker AXS GmbH, Karlsruhe, Germany) with CuKα radiation operated at a tube voltage of 40 kV and a tube current of 40 mA during continuous scanning at 2θ, ranging from 10 • to 40 • at a scanning rate of 2 • /min. FT-IR spectra were measured by a KBr disc method using a spectrometer (SPECTRUM 2000LX; Perkin Elmer Co. Ltd., Kanagawa, Japan).
The surface morphologies of the CO 3 Ap and HAp blocks were observed using a scanning electron microscope (SEM) (S-3400N; Hitachi High-Technologies, Tokyo, Japan) at an accelerating voltage of 15 kV after applying a gold-palladium coating with a magnetron sputtering machine (MSP-1S; Vacuum Device Co., Ibaraki, Japan).
The distribution of the interconnected pore size was evaluated through the penetration of Hg vapor using porosimetry.
The overall porosity was calculated by dividing the apparent density of the sample (d samp ) by the theoretical density of HAp (d HAp ), as shown below in Equation (1).
The mechanical strength of the samples was evaluated in terms of diametral tensile strength (DTS). The diameter and height of each sample were measured using a micrometer (MDC-25M; Mitutoyo Co. Ltd., Kanagawa, Japan). A load was applied to crush each sample in a universal testing machine (AGS-J; Shimadzu Corp., Kyoto, Japan) at a crosshead speed of 1 mm/min. Each DTS value represents the mean of at least eight samples.

Surgical Procedure
Sixteen 18-week-old white, male Japanese rabbits with an average weight of 3.1 kg (Japan SLC Inc., Hamamatsu, Japan) were subjected to animal studies. The experiments were conducted according to the Guide for the Care and Use of Laboratory Animals, National Research Council, USA; the protocols were approved by the Animal Care and Use Committee of Kyushu University (approval number A-28-144-1). The CO 3 Ap and HAp blocks were sterilized by heating at 170 • C for 3 h.
The rabbits were anesthetized by intramuscular injection of ketamine-xylazine (35-10 mg/kg). A medial longitudinal skin incision was made at the distal femur of both legs, and the medial epicondyle of the femur was exposed. A 5.8 mm hole was drilled into the epiphysis of the distal femur and extended to 6 mm. After the CO 3 Ap and HAp blocks were inserted, the periosteum, fascia, and skin were closed by suturing. All animals were allowed unrestrained movement after recovery from anesthesia.

Image Analysis
Four and twelve weeks after implantation, the rabbits were euthanized and the distal femurs were harvested (n = 4 in each group) while removing the surrounding soft tissue. The distal femur including the CO 3 Ap and HAp blocks were scanned using micro-computed tomography (µ-CT) (Skyscan 1075 KHS; Skyscan, Kontich, Belgium) at a source voltage of 60 kV and a source current of 170 µA with a 0.5 mm aluminum filter. Slices obtained from µ-CT measurements were used to reconstruct 2D and 3D images using analysis software. The volume of the remaining blocks was calculated by quantitative 3D evaluation. Before implantation, the volume of each block was calculated using its diameter and thickness. The residual rate (%) was calculated using Equation (2).
Residual rate (%) = (volume of remaining block/volume of block before implantation) × 100 (2) The trabecular thickness (Tb, Th) and trabecular number (Tb, N) were analyzed using software provided with the µ-CT scanner. The trabecular thickness and trabecular number were calculated using Equations (3) and (4), respectively.

Histological Procedures
The obtained samples were fixed with a solution containing 4% paraformaldehyde and 5% glutaraldehyde in 0.1 mol/L phosphate-buffered saline at pH 7.4 for one week. After decalcification with Plank-Rychlo (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) solution for 24 h, the samples were dehydrated by a graded series of ethanol and N-butyl glycidyl ether followed by embedding in an epoxy resin (Quetol 651; Nisshin EM, Tokyo, Japan). The samples were sectioned at a thickness of 1 µm using an ultramicrotome (ULTRACUT S; Reichert-Nissei, Tokyo Japan), followed by staining with hematoxylin and eosin (H&E). Next, the specimens were examined under a microscope (BZ-X710; KEYENCE, Osaka, Japan) to detect grafted material and new bone formation. The amount of new bone was estimated as a percentage of the sample insertion area using a BZX analyzer attached to the microscope.

Statistical Analysis
The Mann-Whitney U test was performed to assess the difference in the residual rate, new bone formation, trabecular thickness, and trabecular number. Significant difference was defined as p < 0.05. Statistical analyses were performed using the JMP software (Version 12.0; SAS Institute, Cary, NC, USA).  (Figure 1g-i). The macrostructure of the samples was the same before and after immersion in Na-H-CO 3 and Na 2 HPO 4 solutions, whereas the microstructure was different. Needle-like crystals typical of CaSO 4 were observed before immersion. When the CaSO 4 blocks were immersed in Na-H-CO 3 solution, rhombohedral and hexagonal crystals were observed. When the samples were further immersed in Na 2 HPO 4 solution, fine aggregated crystals were observed. Figure 2 summarizes the pore size distribution measured by mercury intrusion porosimetry. In the case of CO 3 Ap (30%) and CO 3 Ap (40%), interconnected pores with diameters of 5-30 µm were fabricated in addition to pores that were 0.2-2 µm in diameter, which were observed also in CO 3 Ap (0%). Figure 2 summarizes the pore size distribution measured by mercury intrusion porosimetry. In the case of CO3Ap (30%) and CO3Ap (40%), interconnected pores with diameters of 5-30 μm were fabricated in addition to pores that were 0.2-2 μm in diameter, which were observed also in CO3Ap (0%).  Figure 2 summarizes the pore size distribution measured by mercury intrusion porosimetry. In the case of CO3Ap (30%) and CO3Ap (40%), interconnected pores with diameters of 5-30 μm were fabricated in addition to pores that were 0.2-2 μm in diameter, which were observed also in CO3Ap (0%). in Na-H-CO3 ( Figure 3b) and Na2HPO4 solution ( Figure 3c) and those of sintered HAp blocks ( Figure  3d). The XRD patterns of standard CaSO4 ( Figure 3e) and CaCO3 ( Figure 3f) are also presented for comparison. As shown in this Figure, CaCO3 blocks were fabricated through the dissolutionprecipitation reaction when CaSO4 blocks were immersed in Na-H-CO3 solution. Next, apatite blocks were fabricated through the dissolution-precipitation reaction when CaCO3 blocks were immersed in Na2HPO4 solution (data not shown). Apatite blocks fabricated in Na2HPO4 solution using CaCO3 blocks as a precursor exhibited broader peaks compared with the peaks for sintered HAp, indicating that the crystallinity of apatite formed in the Na2HPO4 solution was lower or closer to that of bone apatite. The XRD patterns were the same regardless of the amount of added spherical phenol resin.  Apatite blocks fabricated in Na 2 HPO 4 solution using CaCO 3 blocks as a precursor exhibited broader peaks compared with the peaks for sintered HAp, indicating that the crystallinity of apatite formed in the Na 2 HPO 4 solution was lower or closer to that of bone apatite. The XRD patterns were the same regardless of the amount of added spherical phenol resin.  (Figure 4f) are also presented for comparison. When CaCO 3 blocks were immersed in Na 2 HPO 4 solution, the appearance of CO 3 peaks that were attributed to AB-type CO 3 Ap ( ) [36,37] indicated that AB-type CO 3 Ap was formed. No absorption peak due to apatitic OH was found. In contrast, sintered HAp exhibited absorption peaks attributed to OH ( ), but no peaks attributed to apatitic CO 3 were observed.

Results
The porosity and DTS values of CaSO 4 , CaCO 3 , CO 3 Ap, and HAp blocks are summarized in Table 1. The porosity of these blocks decreased with increase in the amount of spherical phenol resin added. The porosity of CaCO 3 blocks was higher compared to those of CaSO 4 and CO 3 Ap blocks regardless of the amount of spherical phenol resin added because of the presence of micropores in addition to macropores. The DTS value decreased with increasing porosity. The DTS value of the CaCO 3 blocks was lower than that of the CaSO 4 blocks, while that of the CO 3 Ap blocks was higher than that of the CaCO 3 blocks.  (Figure 4f) are also presented for comparison. When CaCO3 blocks were immersed in Na2HPO4 solution, the appearance of CO3 peaks that were attributed to AB-type CO3Ap (▲) [36,37] indicated that AB-type CO3Ap was formed. No absorption peak due to apatitic OH was found. In contrast, sintered HAp exhibited absorption peaks attributed to OH (△), but no peaks attributed to apatitic CO3 were observed. The porosity and DTS values of CaSO4, CaCO3, CO3Ap, and HAp blocks are summarized in Table 1. The porosity of these blocks decreased with increase in the amount of spherical phenol resin added. The porosity of CaCO3 blocks was higher compared to those of CaSO4 and CO3Ap blocks regardless of the amount of spherical phenol resin added because of the presence of micropores in addition to macropores. The DTS value decreased with increasing porosity. The DTS value of the CaCO3 blocks was lower than that of the CaSO4 blocks, while that of the CO3Ap blocks was higher than that of the CaCO3 blocks.    Figures 5 and 6 summarize the µ-CT observations for CO 3 Ap (5a-c, 5e-g, 6a-c, 6e-g) and HAp blocks (5d, 5h, 6d, 6h) in the sagittal (5a-d, 6a-d) and coronal planes (5e-h, 6e-h) at 4 and 12 weeks, respectively. The edges of CO 3 Ap (30%) and CO 3 Ap (40%) were rounded at 4 weeks after implantation, whereas the CO 3 Ap (0%) and HAp (40%) blocks had a rectangular shape even at 12 weeks in the coronal plane. At 12 weeks, both CO 3 Ap (30%) and CO 3 Ap (40%) were resorbed almost completely and replaced with new bone. Spotty high-density regions were observed in CO 3 Ap (30%) and CO 3 Ap (40%) at 4 weeks and in CO 3 Ap (0%) at 12 weeks, particularly on the cancellous bone side. This suggested that osteointegration occurred earlier in CO 3 Ap (30%) and CO 3 Ap (40%) compared to that in the CO 3 Ap (0%) and HAp blocks (40%). Additionally, a greater amount of spotty high-density regions was found on the cancellous bone side than at the cortex. In contrast, spotty low-density regions were formed in the HAp blocks at 12 weeks after implantation. weeks in the coronal plane. At 12 weeks, both CO3Ap (30%) and CO3Ap (40%) were resorbed almost completely and replaced with new bone. Spotty high-density regions were observed in CO3Ap (30%) and CO3Ap (40%) at 4 weeks and in CO3Ap (0%) at 12 weeks, particularly on the cancellous bone side. This suggested that osteointegration occurred earlier in CO3Ap (30%) and CO3Ap (40%) compared to that in the CO3Ap (0%) and HAp blocks (40%). Additionally, a greater amount of spotty high-density regions was found on the cancellous bone side than at the cortex. In contrast, spotty low-density regions were formed in the HAp blocks at 12 weeks after implantation.    Figure 7 shows the amount of residual materials in the CO3Ap and HAp blocks. The amounts of residual CO3Ap (30%) and CO3Ap (40%) were statistically lower (p < 0.05) than those of CO3Ap (0%) and HAp (40%). In addition, the amount of remaining CO3Ap (0%) was statistically lower (p < 0.05) than that of HAp (40%). CO3Ap (30%) and CO3Ap (40%) blocks were almost completely resorbed at 12 weeks after implantation.   Figure 7 shows the amount of residual materials in the CO 3 Ap and HAp blocks. The amounts of residual CO 3 Ap (30%) and CO 3 Ap (40%) were statistically lower (p < 0.05) than those of CO 3 Ap (0%) and HAp (40%). In addition, the amount of remaining CO 3 Ap (0%) was statistically lower (p < 0.05) than that of HAp (40%). CO 3 Ap (30%) and CO 3 Ap (40%) blocks were almost completely resorbed at 12 weeks after implantation.
indicate the implant insertion portions. Figure 7 shows the amount of residual materials in the CO3Ap and HAp blocks. The amounts of residual CO3Ap (30%) and CO3Ap (40%) were statistically lower (p < 0.05) than those of CO3Ap (0%) and HAp (40%). In addition, the amount of remaining CO3Ap (0%) was statistically lower (p < 0.05) than that of HAp (40%). CO3Ap (30%) and CO3Ap (40%) blocks were almost completely resorbed at 12 weeks after implantation.   For macroporous CO3Ap or CO3Ap (30%) and CO3Ap (40%), tissues including new bone were formed inside the CO3Ap block. A much greater amount of tissues and cells including new bone, vessels, and red blood cells were found inside the CO3Ap (40%) block compared to those found in the CO3Ap (30%) block. In the case of CO3Ap (0%), no tissue penetration was observed at this stage. In the case of HAp (40%), fibrous tissue penetration was observed, whereas no bone tissue was seen inside the block.
Osteoclasts were also found inside the CO3Ap (30%) and CO3Ap (40%) blocks, and resorption occurred well from the trabecular bone side.
At 12 weeks, almost all the CO3Ap (30%) and CO3Ap (40%) blocks were resorbed. Moreover, the original trabecular structure was reconstructed, indicating that remodeling was completed within 12 weeks. New bone was also found inside the CO3Ap (0%) and HAp (40%) blocks. As can be seen in Figure 9h, new bone was spherical and stained more strongly by hematoxylin in the case of HAp (40%). Figure 10 illustrates the amount of new bone as a function of implantation period. The largest bone was formed in the case of CO3Ap (40%), followed by CO3Ap (30%) at 4 weeks. At 12 weeks after implantation, the amount of new bone inside the sample was increased compared with that at 4 weeks, except for that in CO3Ap (40%). For macroporous CO 3 Ap or CO 3 Ap (30%) and CO 3 Ap (40%), tissues including new bone were formed inside the CO 3 Ap block. A much greater amount of tissues and cells including new bone, vessels, and red blood cells were found inside the CO 3 Ap (40%) block compared to those found in the CO 3 Ap (30%) block. In the case of CO 3 Ap (0%), no tissue penetration was observed at this stage. In the case of HAp (40%), fibrous tissue penetration was observed, whereas no bone tissue was seen inside the block.
Osteoclasts were also found inside the CO 3 Ap (30%) and CO 3 Ap (40%) blocks, and resorption occurred well from the trabecular bone side.
At 12 weeks, almost all the CO 3 Ap (30%) and CO 3 Ap (40%) blocks were resorbed. Moreover, the original trabecular structure was reconstructed, indicating that remodeling was completed within 12 weeks. New bone was also found inside the CO 3 Ap (0%) and HAp (40%) blocks. As can be seen in Figure 9h, new bone was spherical and stained more strongly by hematoxylin in the case of HAp (40%). Figure 10 illustrates the amount of new bone as a function of implantation period. The largest bone was formed in the case of CO 3 Ap (40%), followed by CO 3 Ap (30%) at 4 weeks. At 12 weeks after implantation, the amount of new bone inside the sample was increased compared with that at 4 weeks, except for that in CO 3 Ap (40%). Figure 11 shows the trabecular thickness (Tb, Th) and trabecular number (Tb, N) calculated in the implant insertion portions at 12 weeks to estimate the remodeling of new bone. At the trabecular bone of the femur epicondyle, femur Tb, Th and Tb, N were calculated as a control (Figure 8). There was no significant difference between Tb, Th and Tb, N in the implant insertion portion and that in the lateral side. This indicated that remodeling was almost complete within 12 weeks in the cases of CO 3 Ap (30%) and CO 3 Ap (40%) blocks.  Figure 11 shows the trabecular thickness (Tb, Th) and trabecular number (Tb, N) calculated in the implant insertion portions at 12 weeks to estimate the remodeling of new bone. At the trabecular bone of the femur epicondyle, femur Tb, Th and Tb, N were calculated as a control (Figure 8). There was no significant difference between Tb, Th and Tb, N in the implant insertion portion and that in the lateral side. This indicated that remodeling was almost complete within 12 weeks in the cases of CO3Ap (30%) and CO3Ap (40%) blocks.

Discussion
The results obtained in this study clearly demonstrated that porous CaSO 4 blocks can be fabricated by heating the set CaSO 4 ·2H 2 O containing spherical porogens since CaSO 4 blocks are stable even at 700 • C. Porous CO 3 Ap blocks were fabricated through dissolution-precipitation reactions using CaSO 4 blocks while maintaining the latter's macroscopic structure. First, the CaSO 4 block transformed into CaCO 3 while maintaining its macroscopic structure when immersed in Na-H-CO 3 solution (Figures 1, 3 and 4). Next, the CaCO 3 block transformed into CO 3 Ap while maintaining its macroscopic structure when immersed in Na 2 HPO 4 solution. The CO 3 Ap fabricated in this study was AB-type CO 3 Ap or CO 3 Ap found in the human bone. Notably, compared to HAp, this CO 3 Ap is more similar to bone apatite not only with respect to composition but also with respect to crystallinity. Both types of CO 3 Ap, i.e., bone apatite and CO 3 Ap, fabricated in this study are formed in an aqueous solution, whereas HAp is fabricated in an electronic furnace by sintering at high temperature. Therefore, the crystallinity of CO 3 Ap is close to that of bone apatite whereas that of HAp is extremely high, as shown in Figure 3. The dissolution-precipitation procedure is simple and enables the fabrication of artificial bone substitutes that are similar to bone in terms of composition and crystallinity.
The porosity and DTS were controllable by adjusting the amount of spherical phenol resin added. The DTS values of CO 3 Ap (0%), CO 3 Ap (30%), and the HAp block were 3.5 ± 0.3 MPa, 1.4 ± 0.2 MPa, and 1.8 ± 0.3 MPa, respectively. Kopperdahl and Keaveny reported that the DTS of the human trabecular bone lies between 1.3-3.5 MPa [38]. Thus, the DTS values of CO 3 Ap (0%), CO 3 Ap (30%), and the HAp block are nearly equal to that of human trabecular bone.
In addition, the histological results obtained in this study demonstrated the utility of microporous structures. Although CO 3 Ap (0%) was gradually replaced by new bone (Figure 10), the process took time. As shown in Figure 2, only micropores smaller than 2 µm were observed in the case of CO 3 Ap (0%). These micropores may be useful in facilitating osteoclastic resorption by decreasing the density of CO 3 Ap. However, they are too small for cells to penetrate the interior of the sample. Thus, osteoclasts need to resorb CO 3 Ap from the surface of the block, followed by new bone formation by osteoblasts. In contrast, macropores that are 5-100 µm in diameter are present in addition to micropores (smaller than 2 µm) in CO 3 Ap (30%) and CO 3 Ap (40%) blocks. The macropores allow cells and tissues to penetrate the interior of the sample. Although both CO 3 Ap (30%) and CO 3 Ap (40%) have macropores, more tissues penetrated CO 3 Ap (40%). As shown in Figure 8, tissues penetrated from the trabecular bone but not from the periosteum side. Due to cell and tissue penetration, replacement of CO 3 Ap by new bone, similar to remodeling, was remarkable for CO 3 Ap (30%) and CO 3 Ap (40%). At 4 weeks, the remaining CO 3 Ap (30%) and CO 3 Ap (40%) was approximately one fifth of that of CO 3 Ap (0%). The amount of new bone was much greater in CO 3 Ap (40%) compared to that in CO 3 Ap (30%), as shown in Figure 10. Additionally, the amount of new bone was much greater in CO 3 Ap (30%) compared to that in CO 3 Ap (0%). It should be noted that the amount of new bone formed in CO 3 Ap (0%) was greater compared to that in HAp (40%). This difference confirmed that the osteoconductivity of CO 3 Ap was much higher than that of HAp. Although the detailed mechanism of the higher osteoconductivity of CO 3 Ap has not yet been elucidated, one of the reasons behind this phenomenon could be the upregulation of differentiation from bone marrow to osteoblasts in the case of CO 3 Ap [32].
At 12 weeks after implantation, CO 3 Ap (30%) and CO 3 Ap (40%) were completely resorbed, as shown in Figures 7 and 9. If CO 3 Ap (30%) and CO 3 Ap (40%) were resorbed linearly with time, similar to the resorption of CO 3 Ap (0%) and HAp (40%), complete resorption may have been achieved at approximately 8 weeks. This resorption rate seems sufficient for clinical purposes. Approximately 30% of CO 3 Ap (0%) was resorbed at 12 weeks. Although this value is much smaller than that for CO 3 Ap (30%) and CO 3 Ap (40%), the rate is twice that for HAp (40%) and acceptable for clinical applications. As a result of complete resorption and new bone formation, the bone defect area was reconstructed completely. Figure 11 summarizes the comparison of trabecular thickness and trabecular number between the reconstructed side and the lateral side used as a control. There was no statistically significant difference between them, indicating the complete reconstruction of the bone defect using CO 3 Ap (30%) and CO 3 Ap (40%).
In this study, spherical phenol resin with a diameter of 100 µm was used as a porogen. No optimization of the macropore size was carried out even though osteoconduction is known to be governed by pore size. Several researchers suggested that 0.3-0.5 mm is the optimal size range for osteoconduction [39,40]. At the same time, some studies proposed that pore sizes greater than 0.4 mm are less conducive to new bone formation, as evidenced by the accumulation of adipocytes and bone marrow [41] and reduced mechanical properties [39]. In general, 80-100 µm is the minimal pore size required for osteoconduction [42][43][44]. Further studies are needed to optimize the 3D interconnected porous CO 3 Ap based on the results of this investigation.
The total porosity of the HAp (40%) blocks fabricated was 64.2 ± 1.0%, which was close to that of the CO 3 Ap block (30%) at 67.5 ± 0.7%. The DTSs of HAp (40%) and CO 3 Ap (30%) were 1.8 ± 0.3 MPa and 1.4 ± 0.2 MPa, respectively, with no statistically significant difference. On the other hand, the resorption rate of CO 3 Ap (30%) was approximately eight times greater than that of HAp (40%) at 4 weeks and approximately six times greater at 12 weeks. This difference in resorption rate was probably due to the osteoclasts and not to chemical dissolution because, among calcium phosphates, apatite (including CO 3 Ap) is the most thermodynamically stable phase [9].
From the histological images, implant resorption, vascularization, and new bone formation were observed inside the CO 3 Ap (30%) blocks at 4 weeks and normal trabecular bone structure was seen at 12 weeks. Implant resorption occurred well from the trabecular side. This indicated that the resorption of the CO 3 Ap block was due to osteoclasts because, in the natural bone, remodeling occurs more actively in the trabecular bone than in the cortical bone. In the HAp (40%) block, cell infiltration occurred at 4 weeks and even bone formation was observed in the macropores at the center of the samples at 12 weeks; however, new bone appeared to be immature. However, when the CO 3 Ap (0%) block was inserted, new bone was formed only around the samples at 4 weeks. In general, 3D macroporous scaffolds provide greater advantages in the repair of bone defects because the interconnected macroporous structures facilitate nutrient delivery, cell migration, and eventually vascularization and bone ingrowth [45][46][47].
Porous blocks with organized macropores have been fabricated by 3D printing [48][49][50][51]. Currently, this method cannot produce macropores smaller than~350 µm [48][49][50][51]. In contrast, the method proposed in this study can control the macropore size of porous blocks, within the range of a few micrometers to several hundred micrometers, by using spherical phenol resin with the desired diameter. Wang et al. reported that the trabecular thickness was 0.083 mm when the porous block fabricated by 3D printing was implanted [51]. In this study, when CO 3 Ap (30%) and CO 3 Ap (40%) blocks were implanted, the trabecular thicknesses were 0.15 and 0.17 mm, respectively. These findings demonstrate that CO 3 Ap blocks have a higher osteogenic ability than porous blocks fabricated by 3D printing.

Conclusions
Macroporous CO 3 Ap blocks were fabricated through a two-step dissolution-precipitation reaction using gypsum as a precursor and spherical phenol resin as a porogen. Material resorption and new bone formation were quicker in the case of CO 3 Ap (30%) and CO 3 Ap (40%) compared to those in CO 3 Ap (0%) and HAp (40%). Treatment of the bone defect was completed within 12 weeks through reconstruction using CO 3 Ap (30%) and CO 3 Ap (40%). Macropores are useful for the acceleration of this process, which is similar to bone remodeling.