Machined Implants Research Articles

Histologic Evaluation of Early Human Bone Response to Different Implant Surfaces

Studies have demonstrated that roughened dental implant surfaces show firmer bone fixation and an increased percentage of bone-to-implant contact (BIC%) compared to commercially pure titanium-surface (machined) implants. Therefore, the purpose of this study was to evaluate the influence of implant-surface topography on human bone tissue after 2 months of unloaded healing. Fourteen subjects with a mean age of 46.87 +/- 9.45 years received two microimplants each (2.5 mm in diameter and 6 mm in length), one test (sandblasted acid-etched surface) and one control (machined surface), either in the mandible or in the maxilla. After a healing period of 2 months, the microimplants and surrounding tissues were removed with a trephine bur and prepared for histologic analysis. All microimplants, except for one of the controls, were clinically stable after the healing period. Histometric evaluation indicated that the mean BIC% was 23.08% +/- 11.95% and 42.83% +/- 9.80% for machined and rough microimplant surfaces, respectively (P = 0.0005). The bone area within the threads was also higher for sandblasted-surface implants (P = 0.0005). The mean percentage of bone density did not differ between the two groups (P = 0.578). Data from the present histological study suggest that the sandblasted acid-etched implant provides a better human bone tissue response than machined implants under unloaded conditions after a healing period of 2 months.

Bone Contact Around Osseointegrated Implants: A Histologic Study of Acid-Etched and Machined Surfaces

Current literature reports that surface acid etching can improve bone--implant contact (BIC). The aim of this study was to evaluate the differences of BIC between acid-etched (Osseotite) and machined surface implants. Six white New Zealand mature rabbits were used in the present investigation. Each rabbit received two implants, specially made with two surfaces on it (one acid-etched and one machined) into each tibia. A total of 24 implants were inserted. Two animals each were killed at 15, 30, and 60 days after implant placement. Histomorphometry of BIC was statistically evaluated. The acid-etched surfaces showed a greater bone contact percentage than the machined ones during the early phase of healing (15 days), which was not statistically significant. On the other hand, after 30 and 60 days, the differences of BIC between acid-etched and machined surfaces were statistically significant. The acid-etched surface appears to improve BIC at a later stage of healing.

Biological Factors Involved in the Osseointegration of Oral Titanium Implants With Different Surfaces: A Pilot Study in Minipigs

The stability of titanium implants is determined by the rigid load-bearing connections that are formed by the bone, a process that involves a complex network of cells, pro- and anti-inflammatory mediators, and growth factors. The osseointegration processes at the interfaces of machined and porous implants were studied using molecular and histological techniques. Two machined and two porous titanium implants were inserted into the tibiae of four minipigs. The animals were sacrificed at 15, 30, 60, and 90 days post-implantation. The levels of bone morphogenetic protein (BMP)-4, transforming growth factor (TGF)-beta1, and tumor necrosis factor (TNF)-alpha were quantified in the peri-implant osseous samples. The levels of interleukin (IL)-1beta, IL-6, IL-10, and TNF-alpha in the serum were also assessed. Histomorphological analysis showed evidence of bone ossification around the porous implant at 60 days. Surrounding the machined implants, highly sclerotic fibrous pads started the healing response at 90 days, and the levels of TGF-beta1 and BMP-4 began to increase at 60 days, at which time bone ossification around the porous implants was already evident. TNF-alpha was not present in the bone next to the implants. The serum levels of cytokines IL-1beta, IL-6, and IL-10 were not increased. The serum level of TNF-alpha increased during the healing process. We observed that the levels of BMP-4 and TGF-beta1, which play essential roles in the osteogenesis process, increased earlier around the porous implants than around the machined implants. Similarly, the ossification process was initiated earlier at the surfaces of the porous implants than at the surfaces of the machined implants.

Predisposing conditions for retrograde peri‐implantitis, and treatment suggestions

Recent case reports introduced the term retrograde peri-implantitis as a lesion (radiolucency) around the most apical part of an osseointegrated implant. It develops within the first months after insertion. This retrospective study aimed to find predisposing conditions for such peri-apical lesions and to evaluate treatment strategies. All single implants (426 in the upper, 113 in the lower jaw, all Brånemark system type) placed at the department of Periodontology of the University Hospital (Catholic University Leuven) were included in this retrospective evaluation to check the incidence of retrograde peri-implantitis. Eventual predisposing factors such as patient characteristics (age, medical history), recipient site (local bone quality and quantity, cause of tooth loss), periodontal and endodontic conditions of neighboring teeth, implant characteristics (length, surface characteristics), and surgical aspects (guided bone regeneration, osseous fenestration, or dehiscency) were considered. Moreover, implants with retrograde peri-implantitis were followed longitudinally to verify their treatment outcome by means of different parameters (Periotest values (PTV), marginal bone level, radiological size of peri-apical defect). Seven implants in the upper (1.6%) and 3 in the lower jaw (2.7%) showed retrograde peri-implantitis, before or at abutment connection. In comparison with successful implants, such peri-apical lesions occurred preferably at sites with a history of an obvious endodontic pathology of the extracted tooth to be replaced. The incidence of retrograde peri-implantitis was significantly higher (P<0.0001) for TiUnite implants when compared with the machined implants (8/80 vs. 2/459). The machined implant surface, however, showed a higher failure rate (6.8%) than the TiUnite implants (2.5%). Failures with machined surfaces preferably occurred at extraction sites of teeth with a history of endodontic pathology or sites adjacent to teeth with an obvious endodontic pathology. No other predisposing factors could be identified. A curettage of the peri-apical lesions and the use of a bone substitute material prevented further progression of such lesions in the upper jaw (implants maintained their marginal bone and low PTV scores). A treatment in the lower jaw was less successful. Within the limitations of a retrospective study, these results seem to indicate that retrograde peri-implantitis is provoked by remaining scar or granulomatous tissue at the recipient site: endodontic pathology of extracted tooth (scar tissue-impacted tooth) or possible endodontic pathology from a neighboring tooth.

Periimplant Soft Tissue Barrier at Experimental One‐Piece Mini‐implants with Different Surface Topography in Humans: A Light‐Microscopic Overview and Histometric Analysis

Following connection to the oral cavity, osseointegrated dental implants and surrounding tissues are exposed to microbiologic and biomechanical challenges. The establishment of a firm functional periimplant soft tissue barrier (PSTB) is considered to be important to protect the implant's interface from invasion of bacteria. The current knowledge on the histologic architecture of the PSTB is mainly based on animal experiments. The aim of this study was to histologically characterize the PSTB formed in humans around experimental one-piece mini-implants with different surface topography. Five patients received a total of 12 experimental titanium, one-piece mini-implants with an oxidized (n = 4), an acid-etched (n = 4), or a machined (n = 4) surface distal to therapeutic implants. Following transmucosal healing of 8 weeks and at abutment connection of the regular implants, the mini-implants were harvested with a layer of surrounding hard and soft tissue. The specimens were fixed and processed for histologic sectioning according to standard procedures. The most central bucco-oral section cut in the long axis was used for morphologic analyses of the PSTB. The vertical soft tissue morphology was quantified using histometric measurements. The overall height of the soft tissue, that is, the biologic width, was around 4 to 4.5 mm and consisted of an epithelial and a supracrestal connective tissue barrier. The junctional epithelium established the attachment to the implant surface, whereas the collagen fibers and fibroblasts of the connective tissue seal were oriented parallel to the implant. The epithelial attachment was shorter at the oxidized and acid-etched surfaces compared with the machined surfaces. Accordingly, the oxidized and acid-etched mini-implants exhibited a longer zone of connective tissue seal. The periimplant soft tissue formed at the experimental one-piece mini-implants in humans was of a character similar to that described in animal studies. The oxidized and acid-etched implants revealed less epithelial downgrowth and longer connective tissue seal than machined implants.

Early Function of Splinted Implants in Maxillas and Posterior Mandibles, Using Brånemark System® Tiunite™ Implants: An 18‐Month Prospective Clinical Multicenter Study

Immediate or early loading of implants placed in maxillas and posterior mandibles has been a concern as bone density is often low in these areas, making it difficult to establish good initial implant stability. By adapting implant design and insertion protocols, however, high initial implant stability may be achieved in these regions. Further, a modified implant surface texture has been proved to help in maintaining stability during the initial healing period. The aim of the present study was to investigate the clinical performance of oxidized titanium implants (TiUnite, Nobel Biocare AB, Gothenburg, Sweden) when used for early function in the maxilla and in the posterior mandible, locations where the bone density often is low. A further aim was to evaluate the marginal bone level at oxidized implants and compare it with that of machined-surface implants used in a previous study. Thirty-one patients were consecutively included in the study, and 37 edentulous areas in maxillas and posterior mandibles were treated. Bruxism and uncontrolled periodontal disease were exclusion criteria. Temporary prostheses were generally placed within 9 days but not after 16 days from implant placement. A previous study applying the same study design and clinical protocol but using machined-surface implants was used for comparisons. Of the 111 implants installed, 1 failed, giving an overall survival rate of 99.1% after 18 months. The prosthesis survival was 100%. The marginal bone resorption was 0.8 mm (standard deviation [SD], 1.0), as opposed to 1.6 mm (SD, 1.3) in the previous study with machined-surface implants, but was not statistically significantly different (p = .10). The present clinical protocol (aiming at high primary stability) and the use of oxidized titanium implants for early functional loading in the maxilla and the posterior mandible resulted in a high implant survival rate and a favorable marginal bone level during a follow-up of 18 months. The difference in marginal bone resorption between the oxidized implants in the present study and the machined implants from a previous investigation with the same study design was not statistically significant.

A removal torque of the laser-treated titanium implants in rabbit tibia

The purpose of the present study is to evaluate the significance of different surface textures by comparison of the removal forces for laser-treated and machined titanium screw 8 weeks after the installation in rabbit tibia. A total of 14 screw shaped, commercially pure titanium implants with a length of 5 mm, a diameter of 3.75 mm were grouped as follows: Group A: seven implants left as-machined; Group B: seven implants treated with laser method (CSM implant, CSM company, Daegu, Korea) Topographic evaluation was performed with scanning electron microscope (Hitachi S-4200, Japan) to compare the surface structure of laser-treated and machined ones. Installation procedures were done according to Branemark protocol after pre-threading, machined implants were inserted in right tibia metaphysics and laser-treated surface implants were inserted in left ones. Eight weeks post surgically seven rabbits were sacrificed. The implant sites were exposed, and the bone and soft tissues that had formed on top of the implants were carefully removed. Subsequently, the force needed to unscrew the implants ( n=14) was measured using a digital torque gauge (Mark-10 corporation, USA). Scanning electron micrographs of the laser-treated and machined control groups demonstrated created a deep and regular honey-comb pattern with small pore, while machined treatment created the typical microscopically grooved and relatively smooth surface characteristic. Eight weeks after implant placement, the average removal torque was 23.58±3.71 N cm for the machined implants, 62.57±10.44 N cm for the laser-treated implants. The torque measurements yielded statistically significant differences between the machined group and the laser-etched group ( p=0.00055) (Wilcoxon's signed-rank test). The laser-treated group achieved higher removal torque values compared to the machined control group.

Mechanical and histomorphometric evaluations of titanium implants with different surface treatments inserted in sheep cortical bone

Improvement of the implant–bone interface is still an open problem and the interest in chemical modification of implant surfaces for cementless fixation has grown steadily over the past decade. Mechanical and histomorphometric investigations were performed at different times on implants inserted into sheep femoral cortical bone to compare the in vivo osseointegration of titanium screws (∅ 3.5×7 mm length) with different surface treatments. After 8 weeks of implantation, the push-out force of anodized and hydrothermally treated implants (ANODIC) was significantly higher than that of machined implants (MACH) (36%, p<0.0005), whereas a decrease of 39% was observed for acid-etched implants (HF) when compared to other surface treatments. After 12 weeks of implantation, the push-out force values of HF implants were still significantly lower than those observed for MACH (−19%, p<0.01) and hydroxyapatite vacuum plasma-sprayed implants (HAVPS, −25%, p<0.0005), and the highest push-out force was found in HAVPS ( p<0.001) implants. After 8 and 12 weeks of implantation, the AI of HF implants was significantly ( p<0.05) lower (∼−25%) than that of MACH, HAVPS and ANODIC implants. In conclusion, results appear to confirm that there are no specific differences between ANODIC and HAVPS implants in terms of behavior. Moreover, although MACH implants show some surface contaminating agents, they appear to ensure good osseointegration within 12 weeks both mechanically and histomorphometrically, as do ANODIC and HAVPS implants. However, further studies are required to investigate bone hardness and mineralization around implants.

Bone response to machined and resorbable blast material titanium implants: an experimental study in rabbits.

The aim of the present study was a comparison of implants' responses to a machined surface and to a surface sandblasted with hydroxyapatite (HA) particles (resorbable blast material [RBM]). Threaded machined and RBM, grade 3, commercially pure, titanium, screw-shaped inplants were used in this study. Twenty-four New Zealand white mature male rabbits were used. The inplants were inserted into the articular femoral knee joint according to a previously described technique. Each rabbit received 2 inplants, 1 test (RBM) and 1 control (machined). A total of 48 implants (24 control and 24 test) were inserted. The rabbits were anesthetized with intramuscular injections of fluanisone (0.7 mg/ kg body weight) and diazepam (1.5 mg/kg b.wt.), and local anesthesia was given using 1 mL of 2% lidocaine/adrenalin solution. Two rabbits died in the postoperative course. Four animals were euthanatized with an overdose of intravenous pentobarbital after 1, 2, 3, and 4 weeks; 6 rabbits were euthanatized after 8 weeks. A total of 44 implants were retrieved. The specimens were processed with the Precise 1 Automated System to obtain thin ground sections. A total of 3 slides were obtained for each implant. The slides were stained with acid and basic fuchsin and toluidine blue. The slides were observed in normal transmitted light under a Leitz Laborlux microscope, and histomorphometric analysis was performed. With the machined implants, it was possible to observe the presence of bone trabeculae near the implant surface at low magnification. At higher magnification many actively secreting alkaline phosphatasepositive (ALP+) osteoblasts were observed. In many areas, a not yet mineralized matrix was present. After 4 to 8 weeks, mature bone appeared in direct contact with the implant surface, but in many areas a not yet mineralized osteoid matrix was interposed between the mineralized bone and implant surface. In the RBM implants, many ALP+ osteoblasts were present and in direct contact with the implant surface. In other areas of the implant perimeter it was possible to observe the formation of an osteoid matrix directly on the implant surface. Mature bone with few marrow spaces was present after 4 to 8 weeks. Beginning in the third week, a statistically significant difference (P < .001) was found in the bone-implant contact percentages in machined and RBM implants. It must be stressed that these results have been obtained in a passive, nonloaded situation.

Peri-implant health around screw-shaped c.p. titanium machined implants in partially edentulous patients with or without ongoing periodontitis.

The relationship between periodontitis and peri-implantitis remains a matter of debate. The present study compared, "within" randomly chosen partially edentulous patients (n=84 subjects, 97 jaws), the marginal bone loss around teeth and implants during 5 years (range 3 to 11 years) following the first year of bone remodelling. The patients had all been rehabilitated by means of screw-shape c.p. titanium implants with a machined surface (Brånemark system). During the 5 years observation interval, periodontal parameters (marginal bone and attachment loss, the latter for teeth only) were collected together with data on confounding factors (smoking, oral hygiene, tooth loss). Marginal bone loss was measured through long-cone intra-oral radiographs. The mean "interval" bone loss was significantly (P=0.0001) higher around teeth (0.48+/-0.95 mm) than around implants (0.09+/-0.28 mm). The corresponding data for the "worst" performing tooth (0.99+/-1.25 mm) and implant (0.19+/-0.32 mm) per subject showed the same tendency. Neither attachment nor bone loss around teeth correlated with marginal bone loss around implants. This study indicated that the rate of bone loss around screw-shape c.p. titanium implants with a machined surface (Brånemark system implants) was not influenced by the progression rate of periodontal destruction around the remaining teeth within the same jaw.

Dedicated to the Memory of Marshall R. Urist

BACKGROUND:Previous experimental studies have shown a higher degree of bone-implant contact for surface-enlarged implants compared with machined implants. Yet, there is insufficient evidence that such implants show higher stability and an increased survival rate. PURPOSE:The purpose of this investigation was to study the integration and stability of grit-blasted implants with retention elements on the implant neck, with and without marginal bone defects, compared with machined implants without retention elements. MATERIALS AND METHODS:After tooth extraction of the mandibular premolars in six dogs, two grit-blasted, partly microthreaded Astra Tech implants and one standard Brånemark implant were bilaterally placed in each dog. On one side, 3 yen 3 mm large buccal defects were created, to expose three to four implant threads. The contralateral side served as control, and no defects were made. The animals were sacrificed after 4 months of healing. Implant stability was measured using resonance frequency analysis at implant installation and after 4 months of healing. Histologic and histomorphometric evaluation was made after 4 months of healing. RESULTS:Resonance frequency analysis indicated that all implants in the test and control groups were osseointegrated after 4 months, with a tendency toward higher implant stability for the Astra Tech implants. There was a statistically significant higher increase in resonance frequency for the Astra test implants compared with their corresponding controls. Histology and histomorphometry showed well-integrated implants with varying degrees of bone repair at the defect sites. The greater bone-implant contact for the Astra implants was statistically significant. No significant difference between the implants in amount of bone filling the threads was recorded. CONCLUSIONS:The Astra Tech implants tested showed a higher degree of bone-implant contact and higher level of bone regenerated at defect sites compared with the Brånemark implants. Resonance frequency analysis demonstrated a significantly higher increase in the Astra test implants compared with their control groups than did the Brånemark test implants versus their controls.

Bone response to machined cast titanium implants

The aim was to evaluate the bone response to machined cast titanium (Ti) implants. Commercially pure (c.p.) machined Ti implants served as controls. Analyses of the surface composition and topography by Auger electron spectroscopy (AES) and scanning electron microscopy (SEM) revealed no differences comparing the two materials. Cast screw-shaped and identical machined Ti implants were inserted in the tibial metaphysis of 6 rabbits. After 3 and 6 months, the amount of bone within threads and the degree of bone-implant contact were histomorphometrically evaluated. The bone area of cast Ti implants was 45% after 3 months and 62% after 6 months. The corresponding values for machined Ti implants were 51% and 58%, respectively. The total bone-implant contact for cast Ti implants was 19% (25% control implants) after 3 months and 45% (37% for control implants) 6 months after implantation. No statistically significant differences were observed between the two materials at any time interval. The present experimental results indicate that machined cast Ti implants integrate equally well in bone as machined c.p. Ti implants do.

Bone response to surface-modified titanium implants: studies on the early tissue response to machined and electropolished implants with different oxide thicknesses

The bone formation around titanium implants with varied surface properties is investigated. Machined and electropolished samples with and without thick, anodically formed surface oxides were prepared, surface characterized and inserted in the cortical bone of rabbits (1, 3 and 6 weeks). Scanning electron microscopy, scanning Auger electron spectroscopy and atomic force microscopy revealed marked differences in oxide thickness, surface topography and roughness, but no significant differences in surface chemical composition, between the different groups of implants. Light microscopic morphology and morphometry showed that all implants were in contact with bone and had a large proportion of bone within the threads at 6 weeks. The smooth, electropolished implants, irrespective of anodic oxidation, were surrounded by less bone than the machined implants after 1 week. After 6 weeks the bone volume as well as the bone-implant contact were lower for the merely electropolished implants than for the other three groups. Our study shows that a high degree of bone contact and bone formation are achieved with titanium implants which are modified with respect to oxide thickness and surface topography. However, the result with the smooth (electropolished) implants indicates that a reduction of surface roughness, in the initial phase, decreases the rate of bone formation in rabbit cortical bone.

Anchorage of TiO2-blasted, HA-coated, and machined implants: an experimental study with rabbits.

The purpose of this study was to evaluate the histometrical and biomechanical anchorage of TiO2-blasted implants and TiO2-blasted implants coated with hydroxyapatite. The control implants were machined. Twenty-six rabbits had a total of 156 implants placed in the proximal part of the tibia. Each rabbit had a machined, a TiO2-blasted, and a TiO2-blasted, HA-coated implant placed in each tibia. After a healing period of 3 and 12 weeks, respectively, the implants placed in the right tibia were used for removal torque test, and the implants placed in the left tibia were used for histomorphometrical measurements. Preoperatively, implants from the same batches were examined topographically with a TopScan 3D system. The TiO2-blasted implants demonstrated significantly higher removal torque values than the machined implants, and they also had a significantly more irregular surface. Furthermore, significantly higher bone-to-implant contact length fractions were measured adjacent to the TiO2-blasted implants in contrast to the machined implants. The advantages of a TiO2-blasted surface were more pronounced after 3 weeks than after 12 weeks. The results demonstrated that it was possible to influence the anchorage of implants by altering the surface structure morphology. The new method with TiO2 blasting on the titanium surface improves the anchorage of implants but is not yet practicable for HA coating.

Cast titanium as implant material

The tissue response in rats to implants made of machined and cast titanium was evaluated after 1 and 12 weeks. The implants consisted of a circular plate portion, located in the abdominal wall, and a cylindrical rod portion protruding into the peritoneal cavity. The chemical and topographical surface properties of the two types of implants differed considerably. The implants with surrounding tissue were processed en bloc for light and electron microscopy. The bulk metal was removed by an electrochemical procedure which permitted the sectioning and evaluation of the intact implant-tissue interface. The general distribution of macrophages and fibroblasts was the same around the plate portion of both types of implants. Macrophages constituted the predominating cell type with the highest concentration in the innermost cell zone closest to the implant. The number of macrophages per section area was significantly higher around machined implants. Multinuclear giant cells, always located at the implant surface, were more frequent around cast implants. The majority of the intraperitoneal rod portions were partially (1 week) or completely (12 weeks) covered by tissue; partial or complete overgrowth of tissue was rare for machined rod portions. Imaging electron energy loss spectroscopy demonstrated the presence of titanium in macrophages in the peripheral part of the tissue capsule around cast, but not machined implants. We conclude that the tissue responses to the two types of titanium implants differed considerably in the two biological environments (soft tissue in abdominal wall; peritoneal cavity) examined and that the response in one environment does not predict the response in the other. We also believe that improvements have to be made in the casting procedure in order to reduce surface roughness and contamination before cast implants can be used in clinical applications.

Bone response to surface modified titanium implants: studies on electropolished implants with different oxide thicknesses and morphology

In a series of experimental studies, bone formation was analysed around systematically modified titanium implants. In the present study, machined, electropolished and anodically oxidized implants were prepared, surface characterized and inserted in the cortical bone of rabbits (7wks and 12wks). SEM, scanning Auger electron spectroscopy and atomic force microscopy revealed no differences in surface composition but marked differences in oxide thickness, surface topography and roughness. Light microscopic morphology and morphometry showed that all implants were in contact with bone, and had a large proportion of bone within the threads. The smooth, electropolished implants were surrounded by less bone than the machined implants with similar oxide thickness, (4–5 nm) and the anodically oxidized implants with thicker oxides (21 nm and 180 nm, respectively) after 7wks. These studies show that a high degree of bone contact and bone formation can be achieved with titanium implants which are modified with respect to oxide thickness and surface topography. However, it appears that a reduction of surface roughness may influence the rate of bone formation in rabbit cortical bone. Biomaterials (1994) 15, (13) 1062–1074

Quantitative comparison of screw-shaped commercially pure titanium and zirconium implants in rabbit tibia

Screw-shaped implants with an outer diameter of 3.7 mm and a total length of 8 mm were made from rods of commercially pure titanium (c.p. Ti) and commercially pure zirconium (c.p. Zr). Prior to insertion in rabbit tibia for 12 weeks, the surface roughness of two randomly chosen c.p. Ti and two c.p. Zr machined implants of the batch were numerically and visually described with a TopScan 3-D system. Irrespective of implant material, they demonstrated rather similar surface roughnesses, for example, the R a and R q values (that describe the average deviation from the mean line) were 0.79 and 1.18 for c.p. Ti versus 0.56 and 0.86 for c.p. Zr. Each rabbit had four implants inserted; two made of c.p. Ti in one leg and two made of c.p. Zr in the other. The bone tissue reactions to the materials were qualitatively and quantitatively examined. Removal torque tests were performed on the distal implants and histomorphometrical evaluations of the proximal ones. No qualitative or quantitative bone differences were observed. Removal torque measurements demonstrated a mean of 26.2 N cm±9.6 for c.p. Ti versus a mean of 25.9 N cm±7.1 for the c.p.Zr implants. Histomorphometrical comparisons of the bone-to-metal contact revealed a mean of 29 ±10.2% for c.p. Ti and a mean of 19±5.5% for the c.p. Zr samples. The mean bone area in the threads around the c.p. Ti samples was 49±16.6% compared to 43±7.9% for the c.p. Zr group. Commercially pure titanium and zirconium implants seemed to be well accepted by the bone bed after 12 weeks of insertion in rabbit bone.