Previous studies on surface structural changes in vitro as well as in vivo of bioactive A-W-type glass-ceramics and Bio-glass-type glasses showed that the essential condition for glasses and glass-ceramics to bond to living bone is formation of a bonelike apatite layer on their surfaces in the body. Gross et al., however, had explained the bone-bonding mechanism of Ceravital-type apatite-containing glass-ceramic without mentioning formation of the surface apatite layer. In the present study, apatite formation on the surface of one of Ceravital-type glass-ceramics was investigated in vitro as well as in vivo. An apatite-containing glass-ceramic of the composition Na2O 5, CaO 33, SiO2 46, Ca(PO3)2 16 wt%, which was named KGS by Gross et al., was soaked in an acellular simulated body fluid which had ion concentrations almost equal to those of the human blood plasma. The same kind of glass-ceramic was implanted into a rabbit tibia. Thin-film x-ray diffraction, Fourier transform infrared reflection spectroscopy, and scanning electron microscopic observation of the surfaces of the specimens soaked in the simulated body fluid showed that Ceravital-type glass-ceramic also forms a layer of carbonate-containing hydroxyapatite of small crystallites and/or a defective structure on its surface in the fluid. Electron probe x-ray microanalysis of the interface between the glass-ceramic and the surrounding bone showed that a thin layer rich in Ca and P is present at the interface. These findings indicated that Ceravital-type glass-ceramics also form the bonelike apatite layer on its surface in the body and bond to living bone through the apatite layer.

It has been proposed that the formation of a surface apatite layer in vivo on surface active ceramics is an essential condition for chemical bonding between ceramics and bone tissue. To clarify the difference in bone-bonding mechanisms between surface active ceramics and bioresorbable ceramics, two experiments were performed using plates of dense beta-tricalcium phosphate (beta-TCP). First, plates of beta-TCP were implanted subcutaneously in rats for 8 weeks. Surface change due to bioresorption was observed with scanning electron microscopy. Formation of the apatite layer on the surface was investigated using thin-film x-ray diffraction and Fourier transform infrared reflection spectroscopy. Second, plates of beta-TCP were implanted in tibiae of rabbits for 8 and 25 weeks and subjected to the detaching test to measure bone-bonding strength. beta-TCP bonded strongly to bone. Undecalcified sections of the interface of bone and beta-TCP were examined with SEM-EPMA. However, by physicochemical methods, no formation of surface apatite layer was observed. These results suggest that beta-TCP bonds to bone through microanchoring between bone and rough surface of resorbed beta-TCP.

The interface between bioactive ceramics and bone was studied by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). The materials were apatite-wollastonite-containing glass ceramic (A-W.GC) as a representative surface-active ceramic, and calcite and beta-tricalcium phosphate (beta-TCP) as resorbable ceramics. Particles of these materials, ranging between about 100 microns and 300 microns in diameter, were implanted into rat tibiae, and specimens were prepared for observation at 8 weeks after implantation. Both SEM and TEM demonstrated that A-W.GC was bonded to bone through a thin Ca-P-rich layer consisting of fine apatite crystals apparently different from those of bone in shape, size, and orientation. Collagen fibers of the bone reached the surface of this layer, and chemical bonding between A-W.GC and the bone was speculated. Calcite and beta-TCP, on the other hand, made direct contact with the bone, and no apatite layer was present at the interface. The surfaces of the implants became rough due to degradation, and bone grew into the finest surface irregularities. However, we were unable to demonstrate any continuity of crystals between the resorbable implants and bone by high-resolution TEM. Accordingly, the bonding strength was considered to be mainly attributable to mechanical interlocking.

The interfaces between four kinds of surface-active ceramic and bone were studied by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) using undecalcified specimens. The materials were Bioglass-type glass (Bioglass), Ceravital-type glass-ceramic (KGS), apatite- and wollastonite-containing glass-ceramic (A-W.GC) and hydroxyapatite (HA). Particles of these materials, ranging between about 100 and 300 microns in diameter, were implanted into rat tibiae, and specimens were prepared for observation at 8 weeks after implantation. All materials were observed to bond to bone through a collagen-free layer consisting of fine apatite crystals distinct from those in bone. The crystals of this apatite layer and those of bone were intermingled at their interface, suggesting chemical bonding. In Bioglass, which had only a glassy phase, several tens of microns of the material surface had changed to such an apatite layer. In KGS and A-W.GC, which had macrocrystals in the glassy phase, an intervening apatite layer about 0.5 micron thick was observed between the materials and bone. Furthermore, fine apatite crystals were also observed among the macrocrystals near the surface of the materials. In HA, which had no glassy phase, an intervening apatite layer was much less distinct and sometimes absent. These differences were considered to be attributable to the differences in chemical composition, crystallization, and solubility of the materials.

Plates of calcite (CaCO3) were implanted in rabbit tibiae, and their biocompatibility and bonding ability to bone were studied. The plates were also implanted subfascially in rabbit muscle for 8 weeks, and changes on their surfaces in the body were examined. Contact microradiography and Giemsa surface stain demonstrated direct bonding between calcite and bone without interpositions. The average failure load of the interface between calcite and bone was 4.11 kg, indicating an adequate strength of bonding. However, a Ca-P-rich layer, which formed on the surfaces of other bioactive ceramics in vivo, was not detected by a scanning electron microscope-electron probe x-ray microanalyzer. Scanning electron micrographs of the surface of calcite implanted subfascially for 8 weeks showed marked degradation and a rough surface. However, the surface apatite layer was not detected by thin-film x-ray diffraction analysis and Fourier transform infrared reflection spectroscopy. Calcite is a biodegradable material that bonds to bone without a surface apatite layer. The mechanical bonding provided by the anchoring effect of the newly formed bone into the surface roughness of calcite is considered to be a major factor in calcite-bone bonding.