Hydroxyapatite (HAp; Ca₁₀[PO₄]₆[OH]₂) has long been used for the reconstruction of bony defects in orthopedic surgery and in the field of dentistry [12]. HAp shows excellent biocompatibility and osteoconductivity, but a limitation of its use in bony defect reconstruction is that the material remains stable in the bone as a foreign body. HAp is poorly resorbed in vivo, and it can become a source of infection [345]. Although bone apatite is generally thought to be composed of HAp, in fact, human bone is not stoichiometric HAp; instead, it is carbonate apatite (CO₃Ap; Ca_10−a[PO₄]_6−b[CO₃]_c[OH]_2−d), which contains 6%–9% by weight of carbonate in its apatite crystal structure [6]. Since CO₃Ap is more soluble in acidic solution than HAp, CO₃Ap is easily resorbed by osteoclasts [78]. Doi et al. [9] used a sintering method to fabricate a CO₃Ap block. This sintered CO₃Ap block contained up to 6% by weight of carbonate in its apatitic structure. Ishikawa et al. succeeded in fabricating low-crystalline CO₃Ap without a sintering step by utilizing a dissolution-precipitation reaction involving precursors such as low-crystalline calcite [10]. In their study, since the CO₃Ap block was fabricated in the aqueous solution, it contained larger amounts of carbonate in its apatitic structure than the block that was fabricated using a sintering method, and its crystallinity was also similar to that of bone. It was also shown that CO₃Ap upregulated the osteoblastic differentiation of human bone marrow cells earlier than sintered HAp [11]. CO₃Ap showed faster bone formation than Bio-Oss^® (Geistlich‐Pharma, Wolhusen, Switzerland) in a rabbit femur [12]. Taking into account these results, it is evident that CO₃Ap possesses higher osteoinductivity than sintered HAp or BioOss^®, and CO₃Ap may be a superior bone substitute material and scaffold for bone regeneration than those alternatives. This evidence encouraged us to use CO₃Ap granules in the treatment of human bone defects. Kudoh et al. previously reported the safety and efficacy of CO₃Ap granules in single-stage sinus floor augmentation for cases in which the residual alveolar bone height was between 3.7 and 6.0 mm [13]. The limitation of their study was that only radiographic images were evaluated, and no histomorphometric evaluations were performed. Therefore, little was determined about the resorption behavior and tissue response of CO₃Ap in the human body.

The purpose of this study is to clarify the safety and efficacy of CO₃Ap granules in sinus floor augmentation for a severely atrophic maxilla with a residual alveolar bone height in the range of 1.0–5.0 mm, with delayed implant placement. This study is the first histological evaluation of low-crystalline CO₃Ap granules using bone biopsy specimens in humans.

Table 2 summarizes the patients, implants (length, diameter, and installation torque), and amount of CO₃Ap granules used. The amount of CO₃Ap used in each case was between 0.4–2.5 cm³ (1.3±0.6 cm³). No perforation of the sinus membranes occurred during the augmentation. In all cases, the postoperative healing of the augmentation was uneventful, with no abnormal bleeding, pain, or swelling due to the CO₃Ap granules. At 8±2 months after augmentation, it was possible to place 17 implants with lengths of 6.5–11.5 mm. The implant installation torque ranged from 12.0 to 50.0 Ncm, and the mean value was 25.1±13.2 Ncm. All implants achieved osseointegration.

Table 3 and Figure 3 show the changes in the mean EBH. Preoperative residual bone height in the molar region of the maxilla ranged from 1.0 to 5.0 mm, and the mean height was 3.5±1.3 mm. The mean EBH increased to 13.3±1.7 mm immediately after augmentation with the CO₃Ap granules, then decreased to 10.7±1.9 mm and 9.6±1.4 mm at 7±2 months and 18±2 months after augmentation, respectively. The mean EBH at 7±2 months after augmentation was 19.5% lower than that immediately after augmentation, and the difference was significant (P<0.01). The EBH at 18±2 months after sinus floor augmentation had decreased from that at 7±2 months after augmentation, and this decrease was statistically significant (P<0.01).

Clinically, no abnormal bone resorption was observed for the sites augmented with CO₃Ap granules, and a bone height that could support implants was maintained. No implant failures or complications were observed up to 18±2 months after augmentation. Furthermore, no postoperative infection, redness, or persistent pain, such as pain due to an allergic reaction, was observed at the augmented sites under consideration. Therefore, the overall survival rate of the implants was 100% at approximately 31 months after implant placement.

Figure 4 shows surgical views of the maxillary sinus floor augmentation and implant placement in case 10. In this case, the preoperative residual bone heights of the planned implant placement sites (#15, 16, and 17) were 3.0, 1.4, and 1.0 mm, respectively (Figure 5A and E). CO₃Ap granules (Figure 4A) were filled into the elevated space immediately after augmentation (Figure 4B and C). The EBH increased to 14.0, 17.0, and 15.0 mm at sites #15, 16, and 17, respectively. At implant placement, 8 months after augmentation, some CO₃Ap granules were observed on the surface of the augmented site (Figure 4D). New bone had formed around the CO₃Ap granules and was combined with them. Although the EBH had decreased from 14.0 to 10.9 mm at site #15, from 17.0 to 11.5 mm at site #16, and from 15.0 to 10.3 mm at site #17 by 7 months after augmentation, an implant with a length of 10.0 mm could be placed into sites #15 and #16, and an implant with a length of 8.0 mm could be placed into site #17 (Figure 4D). By 18 months after augmentation, the EBH had decreased to 10.0 mm at sites #15 and 16 and to 8.3 mm at site #17. From the panoramic radiographs (Figure 5B, C, and D), the boundary line between the residual bone and the CO₃Ap granules could be detected immediately after augmentation (see arrowheads in Figure 5B); however, it gradually became obscure and could no longer be detected at 18 months post-augmentation (Figure 5D). Figure 5E and F show CT images of the CO₃Ap-grafted site before and 7 months after augmentation, respectively. Residual CO₃Ap granules were observed as high-density areas in contact with the new bone. Figure 5g shows a micro-CT image of a bone biopsy specimen at 7 months after augmentation. The residual CO₃Ap granules were in direct contact with the new bone without the presence of intermediate soft tissue.

The images in Figure 6 show bone biopsy sections from case 8 at 7 months after augmentation. The amounts of new bone and residual CO₃Ap were 43.8% and 13.2%, respectively (Figure 6A). Few inflammatory cells or foreign body giant cells were observed around the CO₃Ap granules (Figure 6B). New bone and CO₃Ap granules were in direct contact with each other without any intermediate fibrous tissue (Figure 6C), and the CO₃Ap granules were in the process of being replaced by new bone (Figure 6D). In the indicated area of Figure 6d, osteoblasts were observed on the surface of the CO₃Ap granules (Figure 6E). In the serial section, using Villanueva-Goldner stain, it was observed that new bone (red) was gradually being replaced with mature bone (green) (Figure 6F).

Table 4 shows the amounts of new bone and residual CO₃Ap in biopsy specimens subjected to histological evaluation. Bone formation was observed in all cases; the amount of new bone ranged from 11.7% to 63.7%, and the amount of residual CO₃Ap ranged from 0.0% to 35.3%, with individual variations among the cases. The mean amounts of new bone and CO₃Ap were 33.8%±15.1% and 15.3%±11.9%, respectively.

Bone biopsy examination revealed few inflammatory cells or foreign body giant cells, and new bone had formed around the residual CO₃Ap granules in all cases. The CO₃Ap granules were in direct contact with the new bone, with no intermediate fibrous tissue. The results obtained in this study clearly demonstrate the excellent biocompatibility and osteoconductivity of low-crystalline CO₃Ap granules in humans, indicating that CO₃Ap granules are a safe and useful bone substitute in two-stage sinus floor augmentation.

Implant insertion torque value, or installation torque value, has been a significant clinical indicator of primary implant stability and superior long-term implant survival rates [1516]. Johansson et al. [17] reported that implant insertion torque values for failed implants were lower than the corresponding values for surviving stable implants. In our study, the installation torque values ranged from 12.0 to 50.0 Ncm (25.1±13.2 Ncm), and the values varied widely. No relationship was observed between the installation torque value and either the ratio of newly formed bone or the height of residual bone (data not shown). Dos Anjos et al. [18] performed implant placement 8 months after sinus floor augmentation with Bio-Oss^® and measured the installation torque. Their data indicated that all implants placed in the maxillary sinus presented higher installation torque values (35.0±9.8 Ncm) than those present in our results. However, installation torque value has been found to be affected by implant design and surgical technique. O'Sullivan et al. [19] measured the installation torque values of different implant designs in the maxillary bone of human cadavers, and the Brånemark Mark IV tapered implant showed a significantly higher installation torque than did the straight implants under study, including the Brånemark Standard implant, the Mark II implant, and the Osseotite implant. Tabassum et al. [20] demonstrated that implants placed with the undersized surgical technique showed higher installation torque values than those placed with the press-fit technique. Although the installation torque values from Dos Anjos's report were higher than those in our study, it is worth noting that they used tapered implants in all cases. In contrast, we used several kinds of implants, because multiple institutions were involved in the trial. Therefore, the variation in the installation torque values that we observed could very well be due to the variety of implant designs and surgical techniques used. In our study, all implants achieved osseointegration, and the overall survival rate at about 31 months after implant placement was 100%.

Figure 7 summarizes the reports of EBH reduction after 2-stage sinus floor augmentation using various bone substitutes. Deppe et al. [21] performed sinus floor augmentation solely with autogenous cortical bone obtained from the iliac crest or the mandible. The EBH reduction rates for the autogenous cortical bone graft at 11 months after augmentation were 21.5% in the iliac crest group and 15.9% in the mandibular bone group. Kim et al. [22] performed sinus floor augmentation with a mixture of Bio-Oss^® and a small amount of autogenous cortical bone. EBH reduction from immediately after augmentation to 1 year after augmentation was 7.8%. In our study, EBH reduction at 18±2 months after augmentation was 27.8%. Taking into account these results, we conclude that the order of resorption speed after augmentation for the graft materials was, from fastest to slowest, autogenous cortical bone, then CO₃Ap granules, and then Bio-Oss^®.

Regarding the changes in EBH after implant placement, Deppe et al. [21] showed that EBH reduction for autogenous cortical bone grafts was 14.4% in the iliac crest group and 8.4% in the mandible group at 5.5 months after implant placement, where the EBH at the time of implant placement was assumed to be 100%. Hieu et al. [23] performed sinus floor augmentation with Bio-Oss^® alone. EBH reduction at 10 months after implant placement was 6.5% compared to EBH reduction of 10.3% at 10±2 months after implant placement in our study. Therefore, the pace of EBH reduction after implant placement of CO₃Ap granules was almost equal to that of autogenous cortical bone and faster than that of Bio-Oss^®.

Figure 8 summarizes the reports on the rate of new bone formation in biopsy specimens from histological evaluation after sinus floor augmentation with autogenous bone and Bio-Oss^®. Zerbo et al. [24] performed histomorphometric analysis and evaluated new bone formation after augmentation with autogenous cortical bone alone. The graft bone was harvested from the mandible, bone biopsy was performed at 6 months after augmentation, and the rate of new bone formation was 41.0%±10.0%. John and Wenz [25] performed histomorphometric analyses on bone biopsies harvested at 3 to 8 months after augmentation using mandibular bone alone. They reported that the bone formation rate for the mandibular bone was 53.5%±2.5%. It has been proposed that the rate of new bone formation from autogenous cortical bone alone is approximately 40%–50%. New bone formation rates for Bio-Oss^® were reported by John and Wenz [25], Sartori et al. [26], and Tadjoedin et al. [27] to be 29.5%±7.4%, 29.8%±2.6%, and 22.9%±2.5%, respectively. Based on these reports, the rate of new bone formation with Bio-Oss^® is approximately 20%–30%. In our study, the amount of new bone at 7±2 months after augmentation was 33.8%±15.1%. A comparison with other graft materials, such as autogenous cortical bone and Bio-Oss^®, was not performed in this study. However, Fujisawa et al. [12] demonstrated that a significantly larger amount of new bone formation was elicited in the cortical portion of a bone defect in a rabbit femur when reconstructed with CO₃Ap granules than with Bio-Oss^® at 8 weeks after implantation. Therefore, we surmise that the rate of new bone formation for CO₃Ap granules is higher than that for Bio-Oss^®.

John and Wenz [25] reported that the bone formation rate of a 2:1 mixture of Bio-Oss^® and mandibular bone was 32.2%±6.9%. Tadjoedin et al. [27] demonstrated a relationship between the rate of new bone formation and the ratio of Bio-Oss^® to autogenous cortical bone. The bone formation rates for 20%, 50%, 80%, and 100% Bio-Oss^® were 37.3%±4.4%, 33.6%±2.7%, 24.7%±2.4%, and 22.9%±2.5%, respectively. This revealed that the bone formation rate was inversely correlated with the amount of Bio-Oss^®. These studies demonstrated that the rate of new bone formation for a combination of Bio-Oss^® and autogenous cortical bone was approximately 24%–37%. In the same manner, adding autogenous cortical bone to CO₃Ap granules would be expected to increase the rate of new bone formation compared to using CO₃Ap granules alone.

Hallman et al. [28] and John and Wenz [25] performed sinus floor augmentation with Bio-Oss^® and reported the amounts of residual Bio-Oss^® at implant placement to be 11.8%±3.6% and 14.9%±6.5%, respectively. In our study, the amount of residual CO₃Ap granules at 7±2 months after augmentation was 15.3%±11.9%, and the values varied widely. Fujisawa et al. [12] showed that the CO₃Ap granules became smaller in size over time, and they also showed that the resorption speed for CO₃Ap granules was faster than that for Bio-Oss^® at 8 weeks after implantation, where implantation was made to a bony defect of a rabbit femur. Therefore, it is reasonable to expect that resorption of CO₃Ap granules would also be faster than that of Bio-Oss^® even in humans.

Galindo-Moreno et al. [29] and Moy et al. [30] reported the results of a histomorphometric study of bone biopsy specimens augmented with Bio-Oss^® and HAp, respectively. Inflammatory cells and foreign body giant cells were rarely observed in the bone biopsy specimens of these materials. The new bone was in direct contact with the augmentation materials, without an intermediate layer of fibrous tissue. Bone biopsy of the CO₃Ap granules showed similar findings (Figure 5), and these results indicate that CO₃Ap granules possess excellent biocompatibility, like Bio-Oss^® and HAp.

Although Bio-Oss^® is widely used as a viable bone substitute, it is derived from a bovine source. Therefore, Bio-Oss^® carries some risk of transmission of infectious diseases, such as bovine spongiform encephalopathy, via prions from cows to humans [31]. Moreover, there remains a possibility of the transmission of new infectious diseases stemming from unknown pathogens. Since Bio-Oss^® is a natural material, there are also slight differences in stability and composition across product lots. In contrast, CO₃Ap granules are a completely artificial synthetic bone substitute; therefore, they have never contained any pathogens, including prions, and they offer stable properties with no variation across product lots.

In conclusion, this study is the first in-human demonstration of the behavior of low-crystalline CO₃Ap granules from bone biopsy specimens. The results of this clinical study suggest that low-crystalline CO₃Ap is a safe and beneficial bone substitute in two-stage sinus floor augmentation. Radiographic and histological examination of the bone biopsy specimens suggest that CO₃Ap granules possess excellent biocompatibility without any risk of allergic reaction or immunological rejection, and they achieve adequate EBH for implant placement. In all cases, the postoperative recovery from augmentation with CO₃Ap granules was uneventful. All implants achieved osseointegration and clinically showed no mobility at 31 months after implant placement. However, further study with a larger sample size and long-term follow-up with patients is needed for a more thorough evaluation.