Bone formation process in porous calcium carbonate and hydroxyapatite Hajime Ohgushi,* Motoaki Okumura, Takafumi Yoshikawa, Keisuke Inoue, Norio Senpuku, and Susumu Tamai Department of Orthopedics, Nara Medical University, Nara, Japan Edwin C . Shors lnterpore International, lroine, California This study determined the bone formation i n porous calcium carbonate (CC) and porous hydroxyapatite (HA) in ectopic sites. The bone formation stimulus was derived from bone marrow cells. CC and HA in the shape of disks were implanted with or without rat marrow cells into subcutaneous sites of syngeneic rats. The CC and HA had identical microstructure: pore size was 190-230 pm, porosity was 50-60% and they were fully interconnected. Bone did not form in any implants without marrow cells (disks themselves), whereas bone consistently formed in the
pores of all implants with marrow cells after 4 weeks. The bone formation of both CC and HA occurred initially on surface of the pore regions and progressed toward the center of the pore. Scanning electron microscopy and electron-probe microanalysis revealed a continuum of calcium at the interfaces of both bone/ CC and bone/HA implants. These results indicate that the bone formation in calcium carbonate derived from marine corals is comparable to the bioactive hydroxyapatite.
IN T RODUCT ION
Hydroxyapatite ceramics (HA) are known to be biocompatible, osteoconductive and osteophilic with bone directly bonding to the surface of the HA.’-* Therefore, these ceramics are currently used as bone graft s ~ b s t i t u t e s .When ~-~ porous HA is placed next to viable bone, ingrowth of bone forming cells and bone formation are observed in the pores?-’ The clinical application of HA may be limited, however, because HA is considered to be essentially nonre~orbable.”’~ An ideal bone grafting material should be replaceable by the host bone. Therefore, the implant needs to be both biodegradable and osteoconductive. Among the biodegradable materials, the calcium carbonate (CC) skeleton of marine corals has been reported to be rapidly biodegradable and
*To whom correspondence should be addressed. Though the materials used in this study were supplied by Interpore International and Dr. Edwin C. Shors is an employee of the company, no benefit of any kind will be received either directly or indirectly by the authors.
Journal of Biomedical Materials Research, Vol. 26, 885-895 (1992) CCC 0021-9304/92/070885-11$4.00 0 1992 John Wiley & Sons, Inc.
OHGUSHI ET AL.
886
osteoconductive.'"-'' Recently, direct contact between the bone and calcite (Natural polycrystalin, metamorphic calcite CaC03) without interposition of soft tissue at the interface was r e p ~ r t e d .However, '~ the bone formation process and the nature of the interface between the bone and the calcium carbonate surface are ambiguous. Furthermore, the capacity of calcium carbonate to form bone, compared to the well known bioactive calcium phosphate ceramics such as HA, is still a major question. In a previous study, we reported that osteogenesis occurred in porous HA, replicated from the exoskeleton of coral by the replamineform technique, when combined with bone marrow cells. Bone was present in the pores of implants placed in subcutaneous site of This method is therefore ideal for determining the interaction between osteogenic cells derived from marrow cells and porous materials, without influences from preexisting host bone. In the current experiment, we used this method to analyze the effect of time and material on the degree of bone formation and on the interfacial chemistry between de nouo bone and disks of porous calcium carbonate and HA. The effects of implant geometry were controlled because the materials had identical microstructure, i.e., porosity, pore size, and interconnectivity. MET HODS
Implants of porous hydroxyapatite (HA) were made by converting the calcium carbonate skeleton of marine coral (genus Povites) to hydroxyapatite through a hydrothermal exchange reaction (Interpore 200 porous hydroxyapatite, Interpore International, Irvine, CA).I9Implants of calcium carbonate (CC) were manufactured using the calcium carbonate directly, as it is prepared prior to conversion. The implants chosen for this study have a pore volume of 50-60% and, fully interconnected pores, measuring 190-230 pm in diameter. The materials were cut into disks, 5 mm in diameter and 2 mm thick. The procedure of marrow cell preparation was detailed in previous report~.'~,*' Briefly, the diaphysis of the femora and tibiae from five Fisher rats (male, 8 weeks old) were cut using a surgairtome. The marrow plugs from the diaphysis were then hydrostatically forced into a 2-mL centrifuge tube containing phosphate buffered saline (PBS). The marrow was disaggregated by sequential passage through 18-G and 20-G needles. It was centrifuged (250g, 10 min) and 200 p L of the supernatant was disaggregated by vortex mixing. The CC and HA discs were soaked in the marrow cell suspension (6 to 8 discs/suspension). Syngeneic Fisher rats were anesthetized by intramuscular injection of pentobarbital (Nembutal, 3.5 mg/100 g BW.) following light ether inhalation. Four incisions (5 mm) were made on the back of the rats. Subcutaneous pouches were created by blunt dissection. Four ceramics were implanted: CC alone, CC with marrow, HA alone, and HA with marrow. Four to six rats were sacrificed each at 1, 2, 3, 4, and 8 weeks after surgery. Two rats were given f luorochrome labeling and harvested at 8 weeks specifically for undecalcified histology after plastic embedding. These rats were given one dose
BONE FORMATION IN POROUS CALCIUM CARBONATE
887
each of tetracycline (50 mg/kg subcutaneously) at 6 weeks and calcein (15 mg/kg, intravenously) at 7 weeks after implantation. For decalcified histological analysis, the harvested implants were fixed in 10% buffered formalin and decalcified (K-CX solution, Falma Co., Tokyo) about 12 h. Embedding was done in paraffin and the implants were cut parallel to the round face of the implant. Five-micrometer sections were then stained with hematoxylin and eosin. The quantity of bone and pore areas of the implants at 4 weeks after implantation were measured using a computerassisted system for histomorphometry (Olympus CIA-102, Olympus Optical Co., Tokyo).” The areas of 12 sections in each CC and HA implants with marrow cells were measured and analyzed statistically by Student’s t test. For undecalcified histological analysis, harvested implants were fixed in 70% ethanol, dehydrated in an alcohol series, defatted, embedded in methylmethacrylate, and cut into 7-pm sections using a microtome (Jung Model K). After staining with Villanueva bone stain, these specimens were observed under light microscopy or f luoromicroscopy. The implants embedded in methylmethacrylate were coated with carbon, and the implant-bone interface was analyzed using a scanning electron microscope connected to an x-ray wave dispersive spectrometer (EPMA8705, Shimadzu, Kyoto, Japan). The implant surface was first observed in the backscattered electron image mode and bone formation areas on the backscatter image were then observed under secondary electron imaging. Line scans for three elements (calcium, phosphorous, and magnesium) were obtained along a line on the bone-implant interface. Image formation and line scan were performed at an accelerating voltage of 10 kV. An electron beam, below 0.4 p m in diameter, was maintained at 2 X 10’ A. RESULTS
All implants without bone marrow cells showed fibrovascular invasion into pore regions, however they did not show any bone or cartilage formation (Fig. l, Table I). When the implants were combined with bone marrow cells, bone formation occurred (Fig. 2, Table I). Bone began to appear at about 3 weeks on the surface of the pore regions of the implants. It occurred through the process of membranous bone formation, thus cartilage did not appear. After 4 weeks, all of the implants with marrow cells showed bone formation and the amount of bone in the pores increased. These characteristic features of bone formation were identical in implants of either calcium carbonate (CC) or hydroxyaptite (HA). And the percentage of the pore space occupied by bone at 4 weeks after implantation for CC and HA were 22 2 1.9 and 19.1 2 2.9, respectively (values represent mean 2 SEM, p > 0.1). Thus, CC implants in the presence of bone marrow cells at these early intervals had comparable bone formation to that of HA. The tissue reactions against implants of CC and HA were different. As shown in Figures 1 and 2, many multinucleated giant cells were present in CC implants. These cells appeared in the presence and absence of the marrow cells, and seemed to phagocytize the pore surface of CC implants. In the
Figure 1. Four weeks after subcutaneous implantation of calcium carbonate (CC) and hydroxyapatite (HA) without marrow cells (Hematoxylin and eosin stain, original magnification X70). Left figure: CC implantation. Right figure: HA implantation. White area indicates the ghost of ceramic after decalcification process. Many multinucleated giant cells (arrows) are seen on the surface of CC.
BONE FORMATION IN POROUS CALCIUM CARBONATE
889
TABLE I Bone Formation Following Implantation ~
Weeks
CC Alone
1 2
014
3 4
014 016 012
8
CC with Marrow
HA Alone
HA with Marrow
014
Calcium carbonate (CC) and hydroxyapatite (HA) with or without marow cells were implanted into subcutaneous site of syngeneic rats. Ratio is number of implants with bone divided by number of implants. Only implants with marrow cells had bone formation, which was consistent in all implants after 4 weeks.
presence of marrow cells at 3 weeks after implantation, bone formation began to appear close to this phagocytic activity in some areas. In contrast, HA implants had few giant cells in the pore regions and did not show obvious phagocytes (Fig. 1).At 4 weeks, bone formation was clearly detected in many pore areas of both CC and HA implants with marrow cells (Fig. 2). There was no intervening fibrous tissue between the bone and the implant's pore surface. Although the bone1HA interface was smooth, bone1CC interface was irregular, probably due to the phagocytic activity by the giant cells (Fig. 2). The histological findings from both CC and HA implants indicated that bone formation occurred on the surface of pore regions and advanced toward the center of the pore. This was confirmed by both the decalcified and undecalcified sections. As shown in Figure 3, bone deposited directly onto the CC surface. Furthermore, f luorochrome labeling showed that tetracycline, administrated 6 weeks after implantation, was seen near the CC surface and calcein, administrated 7 weeks after implantation, was seen close to the center of the pore region (Fig. 3, right). These fluorescent lines were very sharp and not scattered. Thus, the bone formation begun on the surface of the coral by coordination of neighboring osteoblasts. The same pattern was seen in HA implants with marrow cell as reported previo~sly.'~ Thus, identical bone dynamics were seen in both CC and HA. As shown in scanning microscopy (Fig. 4) with x-ray dispersive analysis, the interface between the bone and CC was clearly identified by the transition of phosphorus levels. There was a sharp decrement of phosphorus in the CC. In contrast, a high level of calcium content was seen across the interface. In the soft tissue area, the level of all three elements (calcium, phosphorus, and magnesium) decreased to basal level. Thus, the direct contact of the de nouo bone to the surface of CC implants was shown. DISCUSSION
Our results clearly showed that porous calcium carbonate (CC) can be osteogenic in the presence of marrow cells. The bone began to appear at about 3 weeks and was consistently seen after 4 weeks (Table I, Fig. 2). The bone
Figure 2. Four weeks after subcutaneous implantation of calcium carbonate (CC) and hydroxyapatite (HA) with marrow cells (Hematoxylin and eosin stain). Left figures: CC implantation. Right figures: HA implantation. Upper figures: Lower magnification of the implants (original magnification X30). Lower figures: Higher magnification of rectangular area seen in corresponding upper figures. Dark gray area having osteocytic lacunae indicates bone appeared in porous area. White area indicates the ghost of ceramic after decalcification process. Multinucleated giant cell (arrow) is seen in close proximity to the bone formed in the pore area of CC.
Figure 3. Eight weeks after subcutaneous implantation of calcium carbonate (CC) with marrow cells. Left side is the undecalcified section under light microscopy (Villanueva bone stain, original magnification X180) and right side shows the same section under fluoromicroscopy. B indicates bone and C indicates CC. Arrow heads in left figure indicate the interface between bone and CC. Small arrows in right figure indicate the tetracycline labeling (administered 6 weeks after implantation) and large arrows in right figure indicate calcein labeling (administered 7 weeks after implantation).
rn
Figure 4. Backscattered electron image photograph of calcium carbonate (CC) with marrow cells 8 weeks after implantation. Bone (B; gray area), CC (C; white area) and soft connective tissue (S; black area) are seen. Arrows indicate the bone/CC interface. Right figure shows higher magnification of the rectangular area seen in corresponding left figure. Simultaneous line analysis of calcium (Ca), phosphorous (P), and magnesium (Mg) are shown as wave lines. Full scale for Ca and P is 1000 counts per second; and for Mg is 100 counts per second. Line analysis started from left (area of CC) to right (area of soft connective tissue seen as round shape). Bone area (B) exists between CC and soft tissue area (S). Note the continuous high level of calcium content at the interface between CC and bone. All of three elements (Ca, P and Mg) drop to basal level in the soft tissue area (S).
N
\D
m
BONE FORMATION IN POROUS CALCIUM CARBONATE
893
formation started on the pore surface by the coalescence of active osteoblasts (Fig. 3). Thus de novo bone formed directly in contact with the CC surface (Fig. 4).Our previ~us'"'~ and present findings showed that these characteristic features were also a property of porous hydroxyapatite. Thus, the bone forming response of CC is comparable to that of the well known bioactive hydroxyapatite (HA).The ideal bone substitute is a material which can be completely replaced by new bone. In contrast to HA, CC implants with or without marrow cells showed some degradation. Moreover, CC has the potential to completely degrade.'0z11r24 Thus, a biodegradable CC combined with bone marrow cells is one candidate for the ideal bone substitute. Our findings showed that its degradation is due to a phagocytic process mediated by multinucleated cells. Though we do not address whether or not the cells are osteoclast as proposed by Guillemin et al.," we have shown that the existence of multinucleated cells does not hamper the osteoblastic activity. This conclusion was based on the findings that bone formation could be detected in close proximity to multinucleated cells (Fig. 2). Concerning the nature of the interface between bone and calcium carbonate, Walker and Katz2' proposed two hypothesis about the adherence of bone to calcite (natural polycrystalin, metamorphic calcite CaC03, 100% dense). They postulated that carboxylate and sulfate groups in bone matrix can bond strongly to the exposed calcium sites of calcite to form stable calciumcarboxylate or calcium-sulfate compounds. Further, they postulated that bone adheres well to an overgrowth of hydroxyapatite on the surface of calcite. In this regard, others have reported the existence of a calcium/phosphate rich layer or crystal on the surface of bioactive substances21'22 and calcium carb ~ n a t e Thus, . ~ ~ we speculate that transformation of the calcium carbonate surface (degradation and new crystal formation) is a prerequisite for osteoblastic apposition and/or for differentiation of osteoprogenitor cells into osteogenic cells. These events ultimately lead to bone formation. Previous studies have shown that porous CC completely degrades after 3 months when placed in a bone forming defect.24If, however, bone formation does not occur in the CC before extensive degradation, the implant may not serve as a satisfactory bone graft substitute. To circumvent this limitation, an HA coating has been placed on the complex surfaces of porous CC. This recently developed technology of surface modification of porous CC to hydroxyapatite can control the rate of biodegradation by controlling the thickness of the HA coatingz4Thus, a biodegradable and bone forming cell loaded porous calcium carbonate/hydroxyapatite implant can be made. In summary, although the exact mechanism of bone formation is not clear, our data of bone dynamics (bonding ~ s t e o g e n e s i sFig. ; ~ ~ 3) in implants of calcium carbonate and the continuum of elevated calcium content at the interface of de novo bone and calcium carbonate (Fig. 4) suggest that calcium carbonate should be included in the category of bioactive materials. We thank Mr. H. Komi (Shimadzu Co., Kyoto, Japan) for SEM analysis.
OHGUSHI ET AL.
894
References
5. 6.
7. 8.
9.
10. 11.
12.
13.
14.
15.
16. 17.
18.
19.
K. deGroot, “Bioceramic consisting calcium phosphate salts,” Biomaterials, 1, 47-50 (1980). M. Jarcho, ”Calcium phosphate ceramics as hard tissue prosthetics,” Clin. Orthop., 157, 259-278 (1981). K. Kato, H. Aoki, T. Tabata, and M. Ogiso, ”Biocompatibility of apatite ceramics in mandibles,” Biomat. Med. Den Art. Org., 7, 291-297 (1979). C. A. van Blitterswijk, S.C. Hesseling, J. J. Grote, H. K. Koerten, and K. deGroot, ”The biocompatibility of hydroxyapatite ceramic: A study of retrieved human middle ear implants,” 1. Biomed. Muter. Res., 24, 433453 (1990). D. J. Sartoris, D.H. Gershuni, W.H. Akeson, R.E. Holmes, and D. Resnick, ”Coralline hydroxyapatite bone graft substitutes: Preliminary report of radiographic evaluation,” Radiology, 159,133-137 (1986). R.M. Meffert, J.R. Thomas, K.M. Hamilton, and C.N. Brownstein, ”Hydroxylapatite as an alloplastic graft in the treatment of human periodontal osseous defects,” J. Periodontol., 56, 63-73 (1985). R. Holmes, V. Mooney, R. Bucholz, and A. Tencer, “A coralline hydroxyapatite bone graft substitute,” Clin. Orthop., 188, 252-262 (1984). H. Ohgushi, V. M. Goldberg, and A. I. Caplan, ”Repair of bone defects with marrow and porous ceramic. Experiments in rats,” Acta Orthop. Scand., 60, 334-339 (1989). W. Renooij, A. Hoogenedoorn, W. J. Visser, R.F. Lentfereink, M.G. Schmitz, H. VanLeperen, S. J. Oldenburg, W. M. Janssen, L. M. Akkermans, and P. Wittebol, ”Bioresorption of ceramic strontium-85-labeled calcium phosphate implants in dog femora,” Clin. Orthop., 197,272-285 (1985). G. Guillemin, J. L. Patat, J. Fournie, and M. Chetail, ”The use of coral as a bone graft substitute,“ J. Biomed. h4ater. Res., 21, 557-567 (1987). G. Guillemin, A. Meunier, P. Dallant, I? Christel, J-C. Pouliquen, and L. Sedal, “Comparison of coral resorption and bone apposition with two natural corals of different porosities,” J. Biomed. Mater. Res., 23, 765-779 (1989). E. C. Shors, E. W. White, and G. Kopchok, “Biocompatibility, ostoconduction and biodegradation of porous hydroxyapatite, tricalcium phosphate, sintered hydroxyapatite, and calcium carbonate in rabbit bone defects,” Mater. Res. SOC. Symp. Proc., 110, 211-218 (1989). Y. Fujita, T. Yamamuro, T. Nakamura, S. Kotani, C. Ohtsuki, and T. Kokubo, “The bonding behavior of calcite to bone,” J Biomed. Muter. Res., 25, 991-1003 (1991). H. Ohgushi, V. M. Goldberg, and A. I. Caplan, ”Heterotopic osteogenesis in porous ceramics induced by marrow cells,” J. Orthop. Res., 7, 568578 (1989). H. Ohgushi, M. Okumura, S. Tamai, and E.C. Shors, “Marrow cell induced osteogenesis in porous hydroxyapatite and tricalcium phosphate: A comparative histomorphometric study of ectopic bone formation,” J. Biomed. Muter. Res., 24, 1563-1570 (1990). H. Ohgushi and M. Okumura, ”Osteogenic capacity of rat and human marrow cells in porous ceramics: Experiments in athymic nude mice,” Acta Orthop. Scand., 61, 431-434 (1990). M. Okumura, H. Ohgushi, and S. Tamai, ”Bonding osteogenesis in coralline hydroxyapatite combined with bone marrow cells,” Biomaterials, 12, 411-416 (1991). M. Okumura, H. Ohgushi, S. Tamai, and E.C. Shors, ”SEM-EPMA study of bone-hydroxyapatite interface,” Cells Mater., 1, 29-34 (1991). E. White and E.C. Shors, “Biomaterial aspects of Interpore-200 porous hydroxyapatite,“ Dent. Clin. North Am., 30, 49-67 (1986).
BONE FORMATION IN POROUS CALCIUM CARBONATE
895
20. M. M. Walker and J. L. Katz, ”Evaluation of bonding of bone to inorganic crystal surface,” Bull. Hosp. It. Dis., Orthop. Inst., XLIII, 103-108 (1983). 21. L. L. Hench, “Stability of ceramics in the physiological environment,” in Fundamental Aspects of Biocompufibility, Vol. 1, D. F. Williams (ed.), CRC Press, Boca Raton, 1981, pp. 67-85. 22. T. Kokubo, S. Ito, Z.T. Huang, T. Hayashi, S. Sakka, T. Kitugi, and T. Yamamuro, “Ca, P-rich layer formed on high-strength bioactive glassceramic A-W,” 1. Biomed. Muter. Xes., 24, 331-343 (1990). 23. J. de Kane and J. W. More, “The chemistry of orthophosphate uptake from seawater on to calcite and aragonite,” Geochim Cosmochim Acta, 42,1335-1340 (1978). 24. R. E. Holmes, E.C. Shors, G. Kopchok, L. R. Kitabayashi, and R.T. Laborde, ”Effect of implant biodegradation rate on bone repair,” Trans. 26th Ann. Meeting Sot. for Biornat., 13, 290 (1990). 25. J. F. Osborn and H. Newsely, ”Bonding osteogenesis induced by calcium phosphate ceramic implants,” in Biomateviuls 1980, G. D. Winter, D. F. Gibbons, and H. Plenk (eds.), John Wiley and Sons Ltd., New York, 1982, pp. 51-58. Received April 9, 1991 Accepted November 20,1991
pores of all implants with marrow cells after 4 weeks. The bone formation of both CC and HA occurred initially on surface of the pore regions and progressed toward the center of the pore. Scanning electron microscopy and electron-probe microanalysis revealed a continuum of calcium at the interfaces of both bone/ CC and bone/HA implants. These results indicate that the bone formation in calcium carbonate derived from marine corals is comparable to the bioactive hydroxyapatite.
IN T RODUCT ION
Hydroxyapatite ceramics (HA) are known to be biocompatible, osteoconductive and osteophilic with bone directly bonding to the surface of the HA.’-* Therefore, these ceramics are currently used as bone graft s ~ b s t i t u t e s .When ~-~ porous HA is placed next to viable bone, ingrowth of bone forming cells and bone formation are observed in the pores?-’ The clinical application of HA may be limited, however, because HA is considered to be essentially nonre~orbable.”’~ An ideal bone grafting material should be replaceable by the host bone. Therefore, the implant needs to be both biodegradable and osteoconductive. Among the biodegradable materials, the calcium carbonate (CC) skeleton of marine corals has been reported to be rapidly biodegradable and
*To whom correspondence should be addressed. Though the materials used in this study were supplied by Interpore International and Dr. Edwin C. Shors is an employee of the company, no benefit of any kind will be received either directly or indirectly by the authors.
Journal of Biomedical Materials Research, Vol. 26, 885-895 (1992) CCC 0021-9304/92/070885-11$4.00 0 1992 John Wiley & Sons, Inc.
OHGUSHI ET AL.
886
osteoconductive.'"-'' Recently, direct contact between the bone and calcite (Natural polycrystalin, metamorphic calcite CaC03) without interposition of soft tissue at the interface was r e p ~ r t e d .However, '~ the bone formation process and the nature of the interface between the bone and the calcium carbonate surface are ambiguous. Furthermore, the capacity of calcium carbonate to form bone, compared to the well known bioactive calcium phosphate ceramics such as HA, is still a major question. In a previous study, we reported that osteogenesis occurred in porous HA, replicated from the exoskeleton of coral by the replamineform technique, when combined with bone marrow cells. Bone was present in the pores of implants placed in subcutaneous site of This method is therefore ideal for determining the interaction between osteogenic cells derived from marrow cells and porous materials, without influences from preexisting host bone. In the current experiment, we used this method to analyze the effect of time and material on the degree of bone formation and on the interfacial chemistry between de nouo bone and disks of porous calcium carbonate and HA. The effects of implant geometry were controlled because the materials had identical microstructure, i.e., porosity, pore size, and interconnectivity. MET HODS
Implants of porous hydroxyapatite (HA) were made by converting the calcium carbonate skeleton of marine coral (genus Povites) to hydroxyapatite through a hydrothermal exchange reaction (Interpore 200 porous hydroxyapatite, Interpore International, Irvine, CA).I9Implants of calcium carbonate (CC) were manufactured using the calcium carbonate directly, as it is prepared prior to conversion. The implants chosen for this study have a pore volume of 50-60% and, fully interconnected pores, measuring 190-230 pm in diameter. The materials were cut into disks, 5 mm in diameter and 2 mm thick. The procedure of marrow cell preparation was detailed in previous report~.'~,*' Briefly, the diaphysis of the femora and tibiae from five Fisher rats (male, 8 weeks old) were cut using a surgairtome. The marrow plugs from the diaphysis were then hydrostatically forced into a 2-mL centrifuge tube containing phosphate buffered saline (PBS). The marrow was disaggregated by sequential passage through 18-G and 20-G needles. It was centrifuged (250g, 10 min) and 200 p L of the supernatant was disaggregated by vortex mixing. The CC and HA discs were soaked in the marrow cell suspension (6 to 8 discs/suspension). Syngeneic Fisher rats were anesthetized by intramuscular injection of pentobarbital (Nembutal, 3.5 mg/100 g BW.) following light ether inhalation. Four incisions (5 mm) were made on the back of the rats. Subcutaneous pouches were created by blunt dissection. Four ceramics were implanted: CC alone, CC with marrow, HA alone, and HA with marrow. Four to six rats were sacrificed each at 1, 2, 3, 4, and 8 weeks after surgery. Two rats were given f luorochrome labeling and harvested at 8 weeks specifically for undecalcified histology after plastic embedding. These rats were given one dose
BONE FORMATION IN POROUS CALCIUM CARBONATE
887
each of tetracycline (50 mg/kg subcutaneously) at 6 weeks and calcein (15 mg/kg, intravenously) at 7 weeks after implantation. For decalcified histological analysis, the harvested implants were fixed in 10% buffered formalin and decalcified (K-CX solution, Falma Co., Tokyo) about 12 h. Embedding was done in paraffin and the implants were cut parallel to the round face of the implant. Five-micrometer sections were then stained with hematoxylin and eosin. The quantity of bone and pore areas of the implants at 4 weeks after implantation were measured using a computerassisted system for histomorphometry (Olympus CIA-102, Olympus Optical Co., Tokyo).” The areas of 12 sections in each CC and HA implants with marrow cells were measured and analyzed statistically by Student’s t test. For undecalcified histological analysis, harvested implants were fixed in 70% ethanol, dehydrated in an alcohol series, defatted, embedded in methylmethacrylate, and cut into 7-pm sections using a microtome (Jung Model K). After staining with Villanueva bone stain, these specimens were observed under light microscopy or f luoromicroscopy. The implants embedded in methylmethacrylate were coated with carbon, and the implant-bone interface was analyzed using a scanning electron microscope connected to an x-ray wave dispersive spectrometer (EPMA8705, Shimadzu, Kyoto, Japan). The implant surface was first observed in the backscattered electron image mode and bone formation areas on the backscatter image were then observed under secondary electron imaging. Line scans for three elements (calcium, phosphorous, and magnesium) were obtained along a line on the bone-implant interface. Image formation and line scan were performed at an accelerating voltage of 10 kV. An electron beam, below 0.4 p m in diameter, was maintained at 2 X 10’ A. RESULTS
All implants without bone marrow cells showed fibrovascular invasion into pore regions, however they did not show any bone or cartilage formation (Fig. l, Table I). When the implants were combined with bone marrow cells, bone formation occurred (Fig. 2, Table I). Bone began to appear at about 3 weeks on the surface of the pore regions of the implants. It occurred through the process of membranous bone formation, thus cartilage did not appear. After 4 weeks, all of the implants with marrow cells showed bone formation and the amount of bone in the pores increased. These characteristic features of bone formation were identical in implants of either calcium carbonate (CC) or hydroxyaptite (HA). And the percentage of the pore space occupied by bone at 4 weeks after implantation for CC and HA were 22 2 1.9 and 19.1 2 2.9, respectively (values represent mean 2 SEM, p > 0.1). Thus, CC implants in the presence of bone marrow cells at these early intervals had comparable bone formation to that of HA. The tissue reactions against implants of CC and HA were different. As shown in Figures 1 and 2, many multinucleated giant cells were present in CC implants. These cells appeared in the presence and absence of the marrow cells, and seemed to phagocytize the pore surface of CC implants. In the
Figure 1. Four weeks after subcutaneous implantation of calcium carbonate (CC) and hydroxyapatite (HA) without marrow cells (Hematoxylin and eosin stain, original magnification X70). Left figure: CC implantation. Right figure: HA implantation. White area indicates the ghost of ceramic after decalcification process. Many multinucleated giant cells (arrows) are seen on the surface of CC.
BONE FORMATION IN POROUS CALCIUM CARBONATE
889
TABLE I Bone Formation Following Implantation ~
Weeks
CC Alone
1 2
014
3 4
014 016 012
8
CC with Marrow
HA Alone
HA with Marrow
014
Calcium carbonate (CC) and hydroxyapatite (HA) with or without marow cells were implanted into subcutaneous site of syngeneic rats. Ratio is number of implants with bone divided by number of implants. Only implants with marrow cells had bone formation, which was consistent in all implants after 4 weeks.
presence of marrow cells at 3 weeks after implantation, bone formation began to appear close to this phagocytic activity in some areas. In contrast, HA implants had few giant cells in the pore regions and did not show obvious phagocytes (Fig. 1).At 4 weeks, bone formation was clearly detected in many pore areas of both CC and HA implants with marrow cells (Fig. 2). There was no intervening fibrous tissue between the bone and the implant's pore surface. Although the bone1HA interface was smooth, bone1CC interface was irregular, probably due to the phagocytic activity by the giant cells (Fig. 2). The histological findings from both CC and HA implants indicated that bone formation occurred on the surface of pore regions and advanced toward the center of the pore. This was confirmed by both the decalcified and undecalcified sections. As shown in Figure 3, bone deposited directly onto the CC surface. Furthermore, f luorochrome labeling showed that tetracycline, administrated 6 weeks after implantation, was seen near the CC surface and calcein, administrated 7 weeks after implantation, was seen close to the center of the pore region (Fig. 3, right). These fluorescent lines were very sharp and not scattered. Thus, the bone formation begun on the surface of the coral by coordination of neighboring osteoblasts. The same pattern was seen in HA implants with marrow cell as reported previo~sly.'~ Thus, identical bone dynamics were seen in both CC and HA. As shown in scanning microscopy (Fig. 4) with x-ray dispersive analysis, the interface between the bone and CC was clearly identified by the transition of phosphorus levels. There was a sharp decrement of phosphorus in the CC. In contrast, a high level of calcium content was seen across the interface. In the soft tissue area, the level of all three elements (calcium, phosphorus, and magnesium) decreased to basal level. Thus, the direct contact of the de nouo bone to the surface of CC implants was shown. DISCUSSION
Our results clearly showed that porous calcium carbonate (CC) can be osteogenic in the presence of marrow cells. The bone began to appear at about 3 weeks and was consistently seen after 4 weeks (Table I, Fig. 2). The bone
Figure 2. Four weeks after subcutaneous implantation of calcium carbonate (CC) and hydroxyapatite (HA) with marrow cells (Hematoxylin and eosin stain). Left figures: CC implantation. Right figures: HA implantation. Upper figures: Lower magnification of the implants (original magnification X30). Lower figures: Higher magnification of rectangular area seen in corresponding upper figures. Dark gray area having osteocytic lacunae indicates bone appeared in porous area. White area indicates the ghost of ceramic after decalcification process. Multinucleated giant cell (arrow) is seen in close proximity to the bone formed in the pore area of CC.
Figure 3. Eight weeks after subcutaneous implantation of calcium carbonate (CC) with marrow cells. Left side is the undecalcified section under light microscopy (Villanueva bone stain, original magnification X180) and right side shows the same section under fluoromicroscopy. B indicates bone and C indicates CC. Arrow heads in left figure indicate the interface between bone and CC. Small arrows in right figure indicate the tetracycline labeling (administered 6 weeks after implantation) and large arrows in right figure indicate calcein labeling (administered 7 weeks after implantation).
rn
Figure 4. Backscattered electron image photograph of calcium carbonate (CC) with marrow cells 8 weeks after implantation. Bone (B; gray area), CC (C; white area) and soft connective tissue (S; black area) are seen. Arrows indicate the bone/CC interface. Right figure shows higher magnification of the rectangular area seen in corresponding left figure. Simultaneous line analysis of calcium (Ca), phosphorous (P), and magnesium (Mg) are shown as wave lines. Full scale for Ca and P is 1000 counts per second; and for Mg is 100 counts per second. Line analysis started from left (area of CC) to right (area of soft connective tissue seen as round shape). Bone area (B) exists between CC and soft tissue area (S). Note the continuous high level of calcium content at the interface between CC and bone. All of three elements (Ca, P and Mg) drop to basal level in the soft tissue area (S).
N
\D
m
BONE FORMATION IN POROUS CALCIUM CARBONATE
893
formation started on the pore surface by the coalescence of active osteoblasts (Fig. 3). Thus de novo bone formed directly in contact with the CC surface (Fig. 4).Our previ~us'"'~ and present findings showed that these characteristic features were also a property of porous hydroxyapatite. Thus, the bone forming response of CC is comparable to that of the well known bioactive hydroxyapatite (HA).The ideal bone substitute is a material which can be completely replaced by new bone. In contrast to HA, CC implants with or without marrow cells showed some degradation. Moreover, CC has the potential to completely degrade.'0z11r24 Thus, a biodegradable CC combined with bone marrow cells is one candidate for the ideal bone substitute. Our findings showed that its degradation is due to a phagocytic process mediated by multinucleated cells. Though we do not address whether or not the cells are osteoclast as proposed by Guillemin et al.," we have shown that the existence of multinucleated cells does not hamper the osteoblastic activity. This conclusion was based on the findings that bone formation could be detected in close proximity to multinucleated cells (Fig. 2). Concerning the nature of the interface between bone and calcium carbonate, Walker and Katz2' proposed two hypothesis about the adherence of bone to calcite (natural polycrystalin, metamorphic calcite CaC03, 100% dense). They postulated that carboxylate and sulfate groups in bone matrix can bond strongly to the exposed calcium sites of calcite to form stable calciumcarboxylate or calcium-sulfate compounds. Further, they postulated that bone adheres well to an overgrowth of hydroxyapatite on the surface of calcite. In this regard, others have reported the existence of a calcium/phosphate rich layer or crystal on the surface of bioactive substances21'22 and calcium carb ~ n a t e Thus, . ~ ~ we speculate that transformation of the calcium carbonate surface (degradation and new crystal formation) is a prerequisite for osteoblastic apposition and/or for differentiation of osteoprogenitor cells into osteogenic cells. These events ultimately lead to bone formation. Previous studies have shown that porous CC completely degrades after 3 months when placed in a bone forming defect.24If, however, bone formation does not occur in the CC before extensive degradation, the implant may not serve as a satisfactory bone graft substitute. To circumvent this limitation, an HA coating has been placed on the complex surfaces of porous CC. This recently developed technology of surface modification of porous CC to hydroxyapatite can control the rate of biodegradation by controlling the thickness of the HA coatingz4Thus, a biodegradable and bone forming cell loaded porous calcium carbonate/hydroxyapatite implant can be made. In summary, although the exact mechanism of bone formation is not clear, our data of bone dynamics (bonding ~ s t e o g e n e s i sFig. ; ~ ~ 3) in implants of calcium carbonate and the continuum of elevated calcium content at the interface of de novo bone and calcium carbonate (Fig. 4) suggest that calcium carbonate should be included in the category of bioactive materials. We thank Mr. H. Komi (Shimadzu Co., Kyoto, Japan) for SEM analysis.
OHGUSHI ET AL.
894
References
5. 6.
7. 8.
9.
10. 11.
12.
13.
14.
15.
16. 17.
18.
19.
K. deGroot, “Bioceramic consisting calcium phosphate salts,” Biomaterials, 1, 47-50 (1980). M. Jarcho, ”Calcium phosphate ceramics as hard tissue prosthetics,” Clin. Orthop., 157, 259-278 (1981). K. Kato, H. Aoki, T. Tabata, and M. Ogiso, ”Biocompatibility of apatite ceramics in mandibles,” Biomat. Med. Den Art. Org., 7, 291-297 (1979). C. A. van Blitterswijk, S.C. Hesseling, J. J. Grote, H. K. Koerten, and K. deGroot, ”The biocompatibility of hydroxyapatite ceramic: A study of retrieved human middle ear implants,” 1. Biomed. Muter. Res., 24, 433453 (1990). D. J. Sartoris, D.H. Gershuni, W.H. Akeson, R.E. Holmes, and D. Resnick, ”Coralline hydroxyapatite bone graft substitutes: Preliminary report of radiographic evaluation,” Radiology, 159,133-137 (1986). R.M. Meffert, J.R. Thomas, K.M. Hamilton, and C.N. Brownstein, ”Hydroxylapatite as an alloplastic graft in the treatment of human periodontal osseous defects,” J. Periodontol., 56, 63-73 (1985). R. Holmes, V. Mooney, R. Bucholz, and A. Tencer, “A coralline hydroxyapatite bone graft substitute,” Clin. Orthop., 188, 252-262 (1984). H. Ohgushi, V. M. Goldberg, and A. I. Caplan, ”Repair of bone defects with marrow and porous ceramic. Experiments in rats,” Acta Orthop. Scand., 60, 334-339 (1989). W. Renooij, A. Hoogenedoorn, W. J. Visser, R.F. Lentfereink, M.G. Schmitz, H. VanLeperen, S. J. Oldenburg, W. M. Janssen, L. M. Akkermans, and P. Wittebol, ”Bioresorption of ceramic strontium-85-labeled calcium phosphate implants in dog femora,” Clin. Orthop., 197,272-285 (1985). G. Guillemin, J. L. Patat, J. Fournie, and M. Chetail, ”The use of coral as a bone graft substitute,“ J. Biomed. h4ater. Res., 21, 557-567 (1987). G. Guillemin, A. Meunier, P. Dallant, I? Christel, J-C. Pouliquen, and L. Sedal, “Comparison of coral resorption and bone apposition with two natural corals of different porosities,” J. Biomed. Mater. Res., 23, 765-779 (1989). E. C. Shors, E. W. White, and G. Kopchok, “Biocompatibility, ostoconduction and biodegradation of porous hydroxyapatite, tricalcium phosphate, sintered hydroxyapatite, and calcium carbonate in rabbit bone defects,” Mater. Res. SOC. Symp. Proc., 110, 211-218 (1989). Y. Fujita, T. Yamamuro, T. Nakamura, S. Kotani, C. Ohtsuki, and T. Kokubo, “The bonding behavior of calcite to bone,” J Biomed. Muter. Res., 25, 991-1003 (1991). H. Ohgushi, V. M. Goldberg, and A. I. Caplan, ”Heterotopic osteogenesis in porous ceramics induced by marrow cells,” J. Orthop. Res., 7, 568578 (1989). H. Ohgushi, M. Okumura, S. Tamai, and E.C. Shors, “Marrow cell induced osteogenesis in porous hydroxyapatite and tricalcium phosphate: A comparative histomorphometric study of ectopic bone formation,” J. Biomed. Muter. Res., 24, 1563-1570 (1990). H. Ohgushi and M. Okumura, ”Osteogenic capacity of rat and human marrow cells in porous ceramics: Experiments in athymic nude mice,” Acta Orthop. Scand., 61, 431-434 (1990). M. Okumura, H. Ohgushi, and S. Tamai, ”Bonding osteogenesis in coralline hydroxyapatite combined with bone marrow cells,” Biomaterials, 12, 411-416 (1991). M. Okumura, H. Ohgushi, S. Tamai, and E.C. Shors, ”SEM-EPMA study of bone-hydroxyapatite interface,” Cells Mater., 1, 29-34 (1991). E. White and E.C. Shors, “Biomaterial aspects of Interpore-200 porous hydroxyapatite,“ Dent. Clin. North Am., 30, 49-67 (1986).
BONE FORMATION IN POROUS CALCIUM CARBONATE
895
20. M. M. Walker and J. L. Katz, ”Evaluation of bonding of bone to inorganic crystal surface,” Bull. Hosp. It. Dis., Orthop. Inst., XLIII, 103-108 (1983). 21. L. L. Hench, “Stability of ceramics in the physiological environment,” in Fundamental Aspects of Biocompufibility, Vol. 1, D. F. Williams (ed.), CRC Press, Boca Raton, 1981, pp. 67-85. 22. T. Kokubo, S. Ito, Z.T. Huang, T. Hayashi, S. Sakka, T. Kitugi, and T. Yamamuro, “Ca, P-rich layer formed on high-strength bioactive glassceramic A-W,” 1. Biomed. Muter. Xes., 24, 331-343 (1990). 23. J. de Kane and J. W. More, “The chemistry of orthophosphate uptake from seawater on to calcite and aragonite,” Geochim Cosmochim Acta, 42,1335-1340 (1978). 24. R. E. Holmes, E.C. Shors, G. Kopchok, L. R. Kitabayashi, and R.T. Laborde, ”Effect of implant biodegradation rate on bone repair,” Trans. 26th Ann. Meeting Sot. for Biornat., 13, 290 (1990). 25. J. F. Osborn and H. Newsely, ”Bonding osteogenesis induced by calcium phosphate ceramic implants,” in Biomateviuls 1980, G. D. Winter, D. F. Gibbons, and H. Plenk (eds.), John Wiley and Sons Ltd., New York, 1982, pp. 51-58. Received April 9, 1991 Accepted November 20,1991
