Effect of hydrogen peroxide on osteoinduction by demineralized bone

The osteoinductive capacity of demineralized bone matrix (DBM) has led to wide use of this material for surgical reconstruction. Preparation of DBM often includes sterilization with ethylene oxide, disinfection with various chemical agents, or irradiation. Exposure to hydrogen peroxide (H2O2) is used for both sterilization and bleaching of bone, the latter primarily for cosmetic reasons. We investigated the effect of H2O2, on the osteoinductive capacity of DBM. Cortical bone implants prepared from rat femurs were placed into 3% H2O2 solution. Control specimens were not exposed to H2O2. Bones were then lipid-extracted, demineralized, sterilized with ethylene oxide, and freeze-dried in an identical manner. Allografts were implanted into rat hosts for 1 to 3 weeks. Osteoinduction proceeded rapidly in implants not exposed to H2O2, with chondrocytes and new bone appearing in the implant. After 3 weeks, perforations in the implant were largely replaced with new bone. In contrast, osteoinduction did not occur in implants treated with H2O2. Perforations in H2O2-treated implants were filled with vascularized fibrous tissue, but no cartilage or bone. These findings reveal that H2O2 used for disinfection or bleaching of DBM can abolish its osteoinductive capacity in rats.     

Effects of lathyritic drugs and lathyritic demineralized bone matrix on induced and sustained osteogenesis

Demineralized bone matrix was implanted in normal and lathyritic rats. At 2 weeks, the bone that formed in the lathyritic animals had an elevated alkaline phosphatase activity and a reduced calcium content compared with the controls. Four weeks after implantation, these biochemical parameters were reversed, with a decrease in alkaline phosphatase activity and an increase in calcium content to control levels. the histology of the recovered implants revealed new bone formation. Lathyritic demineralized bone matrix was prepared from bones of rats fed β-aminopropionitrile for 2 weeks (2-week BAPN-DBM) or 4 weeks (4-week BAPN-DBN), and was implanted in normal rats. Two weeks after implantation, both preparations of lathyritic demineralized bone matrix demonstrated early bone formation, although alkaline phosphatase activity and calcium content were reduced. By 4 weeks after implantation, no biochemical or histological evidence of bone formation remained at the site of the 4-week BAPN-DBM implants; continued but reduced bone formation was observed at the site of the 2-week BAPN-DBM implants. Reconstitution of inactivated normal demineralized bone matrix with the guanidine-soluble extracts restored the osteoinductive capacity. However, reconsistution of inactivated lathyritic demineralized bone matrix (4-week BAPN-DBM) failed to restore the osteoinductive capacity. These results indicate that the degree of crosslinking of the collagen matrix that acts as a carrier for osteoinductive proteins plays a key role in inducing and sustaining osteogenesis.     

Platelet-derived growth factor inhibits demineralized bone matrix-induced intramuscular cartilage and bone formation. A study of immunocompromised mice

BACKGROUND: Platelet-derived growth factor (PDGF) has been proposed as a therapeutic agent to promote bone-healing. The purpose of this study was to examine the effect of PDGF on the ability of human demineralized bone matrix to induce bone formation in a nude-mouse muscle-implantation model. We also examined whether platelet-rich plasma, which contains PDGF, also modulates osteoinduction in this model. METHODS: Human demineralized bone matrix, previously shown to be osteoinductive in the calf muscles of nude mice, was mixed with PDGF-BB (0, 0.1, 1, and 10 microg/10 mg of demineralized bone matrix) and was implanted bilaterally in the calf muscles of immunocompromised (nu/nu) mice (six mice in each group). Heat-inactivated demineralized bone matrix was used as a control. Tissue was harvested at fourteen, twenty-eight, and fifty-six days after implantation. Platelet-rich plasma was prepared from the blood of a healthy donor with use of the Harvest PRP preparation device, activated with thrombin, and mixed with active and inactive demineralized bone matrix. Fifty-six days post-implantation, tissues were harvested. Osteoinduction was assessed with use of a qualitative scoring system and with quantitative histomorphometry. RESULTS: Cartilage was present at fourteen days in all tissues that had received an implant, but the amount decreased as the PDGF concentration increased. PDGF reduced bone formation at twenty-eight days in a dose-dependent manner. This inhibitory effect was resolved by fifty-six days, except in tissues in which demineralized bone matrix and 10 microg of PDGF had been implanted. In sites treated with 10 microg of PDGF, the area of new bone was decreased and the area of bone marrow was reduced at twenty-eight and fifty-six days. PDGF also appeared to retard resorption of demineralized bone matrix in a dose-dependent manner. Platelet-rich plasma reduced osteoinduction by human demineralized bone matrix that had high osteoinductive activity and had no effect on osteoinduction by demineralized bone matrix with low activity. CONCLUSIONS: PDGF inhibits, in a dose-dependent manner, intramuscular osteoinduction and chondrogenesis by demineralized bone matrix in immunocompromised mice. Platelet-rich plasma also reduces the osteoinductivity of active demineralized bone matrix.     

The effect of various sterilization procedures on the osteoinductive properties of demineralized bone matrix]

To minimize potential infection following the transplantation of allogeneic bone, extremely rigorous selection of donors and careful processing and storage of samples are required. Other major problems related to allogeneic transplants, such as reduced osteogenic properties and immunological reactions, led to the development of demineralized bone matrix (DBM). This osteoinductive bone extract is largely free of antigens and is easy to produce. However, to eliminate the potential risk of infection, DBM should be sterilized prior to implantation. The purpose of this study was to investigate the influence of different sterilization techniques on the osteoinductive properties of DBM. A series of 76 cortical defects (drill holes) 0.6 cm in diameter in the tibiae of 11 Merino sheep were filled with DBM in addition to autogeneic and allogeneic cancellous bone. Prior to implantation DBM was sterilized by autoclaving, gamma irradiation, or application of ethylene oxide or ethyl alcohol. A further 12 drill holes were left empty as controls. The formation of new bone was examined 3 and 6 weeks postoperatively, using histological, fluorescent-optical and microradiographical techniques. The amount of newly formed bone was also quantified. Apart from autoclaved DBM all matrix grafts showed excellent new bone formation following sterilization, by far exceeding the formation with allogeneic cancellous bone.     

Implants of heterologous demineralized bone matrix for induction of posterior spinal fusion in rats.

The authors tested the osteoinductive capacity of powdered heterologous (bovine) demineralized bone matrix in rats. The first part of the study concerned a monolateral posterior spinal implant after decortication of three vertebrae, using as a control area the animal's contralateral side, in which neither bone graft nor any other material were placed. In another group of rats, a comparative evaluation was made of powdered heterologous demineralized bone matrix and fresh autologous bone. In the same animal, autologous bone was implanted to realize a thoracic posterior fusion and demineralized bone matrix, to induce a posterior fusion in the lumbar area. All data obtained suggested a good osteoinductive activity of heterologous powdered demineralized bone matrix. The two posterior spinal fusions done in the same animal with heterologous demineralized bone matrix or autologous bone, respectively, had similar callus development and required the same time for formation.     

An evaluation of human demineralized bone matrices in a rat femoral defect model

The osteoconductive and osteoinductive potential of two human allogeneic demineralized bone matrix putties were compared in a critical-sized athymic rat femoral defect model. Defects were treated with (1) a demineralized bone matrix in a hyaluronic acid carrier, (2) a demineralized bone matrix in a glycerol carrier, (3) a hyaluronic acid carrier alone, or (4) with no implant. Radiographic examinations and histologic analyses were done at 4, 8, and 16 weeks postoperatively. Eight of the 48 defects treated with a demineralized bone matrix and none of the 36 surgical controls showed complete radiographic healing by 16 weeks and no statistically significant difference between the radiographic scores for the two demineralized bone matrix preparations was found. On histologic review, both preparations of demineralized bone matrix had passive remineralization. The largest foci of endochondral ossification were seen in limbs treated with a demineralized bone matrix in a hyaluronic acid carrier. The 8-mm rat femoral defect allows for stringent assessment of the osteoinductive potential of bone graft substitutes. Hyaluronic acid and glycerol are viable carriers for demineralized bone matrices. As both de-mineralized bone matrices tested provided an adequate osteoconductive matrix and showed some, although limited, osteoinductive capacity, these materials should be used in clinical practice only as bone graft extenders or enhancers.     

Ethene oxide and bone induction: Controversy remains

There is controversy as to whether ethene oxide (“ethylene oxide”, EO) sterilization destroys the bone-inducing capacity of demineralized bone matrix (DBM) or not. Correctly performed studies seem to support both opinions. Bone conductive properties of fresh frozen, defatted bone grafts are greatly impaired by EO sterilization, whereas purified inductive proteins resist EO. Studies showing destruction of osteoinductive capacity used nonpulverized DBM, whereas the others used powder. This could be the key to resolving the controversy, because if EO treatment reduces the cells' ability to penetrate a cortical graft and to reach inductive proteins inside it, it may appear noninductive after EO sterilization, even though BMP molecules may be intact. On the other hand, cells could easily penetrate the powder implants.

We compared the effect of EO sterilization on the inductive capacity of demineralized cortical bone with that of DBM powder, using allogeneic material in rats. Cortical pieces lost all inductive capacity by EO sterilization, whereas the powder yielded a calcium content which was at best one fourth of the un-sterilized. The concentrations of residual EO, ethene chlorohydrin and ethene glucol at implantation were far below approved levels. Another difference between studies is the humidity during EO treatment. In our hands, humidification reduced bone yield by half.

In conclusion, EO sterilization may impair the biological performance of bone inductive implants by reducing cell penetration into bulk material. However, DBM powder, when correctly sterilized, also yielded scanty amounts of bone.     

Rapid quantitative bioassay of osteoinduction.

We developed a reproducible, relatively rapid bioassay that quantitatively correlates with the osteoinductive capacity of demineralized bone matrix obtained from human long bones. We have found that Saos human osteosarcoma cells proliferate in response to incubation with demineralized bone matrix and that an index of this proliferative activity correlates with demineralized bone matrix-induced osteogenesis in vivo. The bioassay (Saos cell proliferation) had an interassay coefficient of variation of 23 +/- 2% and an intra-assay coefficient of 11 +/- 1%. Cell proliferation was normalized to a standard sample of demineralized bone matrix with a clinically high osteoinductive capacity, which was assigned a value of one. The Saos cell proliferation for each sample was related to the standard and assigned a value placing it into the low (0.00-0.39), intermediate (0.40-0.69), or high (0.70-1.49) osteoinductive index group. Osteoinduction of human demineralized bone matrix was quantitated by expressing new bone formation as a function of the total bone volume (new bone plus the demineralized bone powder). The demineralized bone matrix was placed in pouches formed in the rectus abdominis muscles of athymic rats, and endochondral bone formation was assessed at 35 days following implantation, when marrow spaces in the ossicles were formed by new bone bridging the spaces between demineralized bone matrix particles. The proliferative index correlated with the area of new bone formation in histological sections of the newly formed ossicles. When the proliferative index (the osteoinductive index) was divided into low, intermediate, and high groups, the correlation between it and new bone formation (osteoinduction) was 0.850 (p < 0.0005) in 25 samples of demineralized bone matrix. There was no overlap in the osteoinduction stimulated between the samples with low and high osteoinductive indices. We conclude that the proliferation assay is useful for the routine screening of bone allograft donors for osteoinductive potential. Furthermore, the two-dimensional area of new bone formation, as it relates to total new bone area, is a quantitative measure of osteoinduction.     

Osteogenesis in rats with an inductive bovine composite

Subcutaneous (S.C.) implantation of allogeneic demineralized bone matrix in rats results in endochondral bone formation. In contrast, implants of bovine demineralized bone matrix in rat S.C. tissue show inconsistent cartilage and bone formation, presumably due to an intense inflammatory reaction at the implant site. To overcome this response, a partially purified bone inducing extract was prepared from bovine bone by a series of steps that included demineralization, guanidine/HCl extraction, gel filtration, and cation exchange chromatography. To develop a carrier, the inactive guanidine/HCl-extracted matrix was then trypsinized to remove the inflammatory and immunogenic components, thus yielding a predominantly collagenous matrix. Bovine composites were prepared by combining different amounts of the bone inducing extract with a carrier that consisted of the trypsinized bone matrix and purified soluble bovine dermal collagen. Subcutaneous implantation of the composite preparation resulted in dose-dependent endochondral bone formation in rats. The inductive activity and the low-level inflammatory response were comparable to allogeneic implants.     

Chemotaxis of muscle-derived mesenchymal cells to bone-inductive proteins of rat

Summary In this investigation we examined the chemotaxis of muscle-derived mesenchymal cells from neonatal rats to partially purified extracts of demineralized bone matrix with osteoinductive properties. Using the modified Boyden chamber assay and muscle-derived mesenchymal cells obtained from neonatal Long-Evans rats, we tested the chemotactic properties of the 4 M guanidine-HCl extract from demineralized rat bone matrix and fractions thereof that were derived from sequential Sepharose CL-6B, TSK-3000 and HPLC-C₁₈ chromatography. We have identified that those fractions exhibiting chemotactic properties were also osteoinductive. Therefore, demineralized bone matrix serves as its own soluble signal and insoluble substratum in the inductive process leading to endochondral bone formationin vivo.     

Antibacterial and osteoinductive properties of demineralized bone matrix treated with silver

The problems incurred by storage of demineralized bone allograft material and its potential use in contaminated operative sites make an antibacterial property desirable. Silver was considered for this role because of its wide spectrum of antibacterial susceptibility, low incidence of resistance, and its ability to persistently inhibit bacteria after binding to collagen matrices. Demineralized bone matrix prepared from rat diaphyseal bone segments was treated by exposure fo AgCl, AgNO3 and NaNO3 solutions prior to lyophilization. The resulting material was tested for bacterial inhibition after incubation in saline solutions for various times and showed inhibition persisting for at least four weeks (Ag-treated material only). Silver treating the matrix was found to partially inhibit the osteoinductive capacity at 10(-3) and 10(-2) M but not at 10(-5) M as measured by intramuscular implantation in the rat for six weeks. Control and NaNo3-treated specimens showed normal bone growth as measured by ashing and by 99mTc binding, and confirmed by radiologic densities. Histologic sections showed dense microdeposits on dense material predominately near the decalcified bone surfaces, but also within the matrix. The results suggest that pretreatment with silver at concentrations in the 10(-4) range would render the implant material antibacterial, protect its sterility, and leave the osteoinductive capacity intact.     

Platelet-rich plasma inhibits demineralized bone matrix-induced bone formation in nude mice

BACKGROUND: It is unclear whether platelet-rich plasma is a clinically effective adjunct to osteoinductive agents such as demineralized bone matrix. It contains platelet-derived growth factor (PDGF), which decreases osteoinduction by human demineralized bone matrix in nude-mouse muscle, suggesting that platelet-rich plasma may also have a negative impact. This study tested the hypothesis that platelet-rich plasma reduces demineralized bone matrix-induced bone formation and that this effect varies with donor-dependent differences in platelet-rich plasma and demineralized bone matrix. METHODS: Human platelet-rich plasma was prepared from blood from six men (average age [and standard error of the mean], 29.2 +/- 2.4 years). Platelet numbers were determined, and growth factors were quantified before and after platelet activation. Human demineralized bone matrix from two donors (demineralized bone matrix-1 and demineralized bone matrix-2) was mixed with activated platelet-rich plasma and was implanted bilaterally in the gastrocnemius muscle in eighty male nude mice (eight implants per variable). Fifty-six days after implantation, the hindlimb calf muscles were harvested for histological analysis. Osteoinduction was evaluated with use of a qualitative score and morphometric measurements of ossicle size, new bone formation, and residual demineralized bone matrix. RESULTS: Compared with platelet-poor plasma, platelet-rich plasma preparations exhibited a fourfold increase in the platelet count, a fifteenfold increase in the amount of transforming growth factor-beta, a sixfold increase in the amount of PDGF-BB, a fivefold increase in the amount of PDGF-AA, and a twofold increase in the amount of PDGF-AB. Demineralized bone matrix-1 was more osteoinductive than demineralized bone matrix-2, as determined on the basis of a greater ossicle area. The effect of platelet-rich plasma was either neutral or inhibitory depending on the demineralized bone matrix batch. When used with demineralized bone matrix-1, platelet-rich plasma did not alter the qualitative score or overall ossicle size, but it decreased the new bone area. When used with demineralized bone matrix-2, platelet-rich plasma reduced the qualitative score, ossicle area, and new bone area and increased the amount of residual demineralized bone matrix. The effects on osteoinduction also varied with the donor of the platelet-rich plasma. CONCLUSIONS: Platelet-rich plasma decreased the osteoinductivity of demineralized bone matrix implanted in immunocom-promised mice, and the activities of both demineralized bone matrix and platelet-rich plasma were donor-dependent.     

In vivo analysis of the half-life of the osteoinductive potential of demineralized bone matrix using diffusion chambers

Summary Subcutaneous implantation of demineralized bone matrix (DBM) from rat initiates a sequence of developmental events that results in endochondral bone formation. This investigation examined the modification of the osteoinductive potential of DBM during the intial stages of this developmental cascade. Diffusion chambers (DC), constructed with filters of known pore size, permitting or excluding cells from entering the chambers, and containing DBM were subcutaneously implanted into Long-Evans male rats for specific time periods (1–7 days). DC were recovered and the osteoinductive potential of the matrix from these chambers was then tested by subcutaneous implantation and assaying the resulting day 11 plaque tissue enzymatically for alkaline phosphatase activity, and histologically for evidence of chondrogenesis and osteogenesis. The possible modification of DBM by local systemic factors (enzymatic degra-dation) or contact by polymorphonuclear leukocytes (PMNs) was also investigated. We have concluded from this study that the osteoinductive potential of DBM has a half-life of 5–7 days following implantation and although the enzymes collagenase, elastase, and trypsin abolished this activity, pepsin significantly enhanced it. Culture of PMNs with matrix prior to its implantation appeared to have little effect. Furthermore, during the initial stages of matrix-induced endochondral bone formation, DBM serves as both the instructive inducer and permissive substratum required in this process.     

Bone induction by fetal and adult human bone matrix in athymic rats

In the rat the intramuscular implantation of demineralized rat bone matrix induces local bone formation. In adult primates, however, allogenous bone matrix induces little or no bone formation in extraskeletal sites. To assay inductive properties, human demineralized bone matrix from 6 adult donors and 4 fetuses was implanted intramuscularly in athymic rats for 6 weeks. Fetal and adult matrix implants yielded about the same amount of bone: about half of the bone yield from rat or rabbit matrix in the same model. We conclude that human bone matrix has inductive properties and that failures to induce bone formation in adult primates may be due to an inability by the recipients to respond to inductive stimuli of adult bone matrix.     

Bone induction by fetal and adult human bone matrix in athymic rats

In the rat the intramuscular implantation of demineralized rat bone matrix induces local bone formation. In adult primates, however, allogenous bone matrix induces little or no bone formation in extraskeletal sites. To assay inductive properties, human demineralized bone matrix from 6 adult donors and 4 fetuses was implanted intramuscularly in athymic rats for 6 weeks. Fetal and adult matrix implants yielded about the same amount of bone: about half of the bone yield from rat or rabbit matrix in the same model. We conclude that human bone matrix has inductive properties and that failures to induce bone formation in adult primates may be due to an inability by the recipients to respond to inductive stimuli of adult bone matrix.     

Induction of endochondral bone by demineralized bone matrix from diabetic rats

Summary In this investigation we examined the osteoinductive potential of demineralized bone matrix derived from chronically diabetic (streptozotocin-induced) rats. Long-Evans rats (28–31 days) were made diabetic with a single injection of streptozotocin (65 mg/kg) and provided food and waterad lib for 2 months. Diaphyseal shafts of femurs and tibias removed from the diabetic rats and their sibling controls were dehydrated, pulverized, sieved to 74–420 μm particles, and demineralized Matrix was then bioassayed for its ability to induce endochondral bone on day 11 following subcutaneous implantation over the thorax of Long-Evans rats. The resulting plaques of tissue were subjected to histological analysis, determination of alkaline phosphatase activity, and calcium content. Bone matrix derived from diabetic animals proved to be a significantly better inducer of endochondral bone than did control matrix.     

Ethylene oxide does not extinguish the osteoinductive capacity of demineralized bone: A reappraisal in rats

We examined the influence of ethylene oxide (EO) and gamma irradiation on the osteoinductive capacity of demineralized bone. Demineralized bone powder prepared from Wistar rats was exposed to EO (55 °C or 40 °C) or gamma irradiation (25 KGy) or was preserved in ethanol. Sterilely-prepared bones served as controls. The powder was packed in a gelatin capsule and implanted for 6 weeks in muscles of 6-week-old female rats. Exposure of demineralized bone particles to EO 55 °C resulted in an almost complete loss of osteoinductivity. Irradiated bones lost about 40% of their osteoinductive capacity, while sterilization with EO at 40 °C resulted in only a slight alteration of the osteoinductivity, as assessed by the recovered weight ratio, calcium content, alkaline phosphatase activity measurements and histo-morphometry. Ethanol treatment had no influence on the new bone yield when compared to controls.

As EO exposure at 40 °C is a true sterilization procedure, it can be recommended in a clinical setting for its small effect on osteoinductive capacity as assessed experimentally in rats.     

An unexpected outcome during testing of commercially available demineralized bone graft materials: how safe are the nonallograft components?

STUDY DESIGN: Radiographic and histologic analyses of commercially available bone graft materials were performed. OBJECTIVE: To compare the osteoinductive efficacy of commercially available demineralized bone matrix material. SUMMARY OF BACKGROUND DATA: The relative in vivo bone formation and toxicology of the nonallograft components the make up various commercially available demineralized bone matrix products are not known. METHODS: An in vivo bone formation model was used in 30 athymic rats. Six different bone grafting materials were tested in subcutaneous and intermuscular locations. After 4 weeks, radiographic and histologic testing of bone formation was performed. RESULTS: Eight of nine rats implanted with Grafton demineralized bone matrix products died 1 to 4 days after implantation of the bone graft material. None of the remaining 10 animals implanted with the four other grafting materials died. The experiment was modified and completed with a lower dose of bone graft material. Pathologic analysis indicated that the cause of death was hemorrhagic necrosis of the kidneys, most likely caused by a toxic effect on the glomeruli and tubules. A possible causative factor may have been the glycerol in the graft material. CONCLUSIONS: Although the volume of Grafton product per kilogram of body weight used in this study was approximately eight times the maximum volume used in humans, the authors believe that this data must be reported because this product is used substantially in clinical settings. In addition, the osteoinductive performance and relative safety of the nonallograft components in all commercially available demineralized bone grafts are not known.     

Influence of indomethacin on induced heterotopic bone formation in rats. Importance of length of treatment and of age

The effect of indomethacin on heterotopic and orthotopic bone formation in rats was analyzed with respect to (1) length of treatment after implantation, (2) duration of the indomethacin induced inhibition of heterotopic bone formation, and (3) influence of age of the implant recipient. Three weeks after implantation of demineralized bone matrix, the ash weight of implants from rats receiving indomethacin 2 mg/kg body weight during the entire experiment was 31% lower than that of controls. Animals treated for only six days after implantation exhibited an almost equally pronounced inhibition. However, six weeks after implantation, the inhibition caused by six days of indomethacin treatment had almost dissipated. In older rats the implants of demineralized bone matrix induces smaller volumes of new bone than in younger rats, but indomethacin causes approximately the same degree of inhibition of osteoinduction. Orthotopic bone is not affected by indomethacin treatment. This study shows that a short period of indomethacin treatment at the time of implantation of demineralized bone matrix is sufficient to reduce experimental bone formation, but the inhibitory effect slowly diminishes if the inductive process is continuous. The results indicate that the inhibition of heterotopic new bone formation by indomethacin may be mediated through reduction of the initial inflammatory response or by reduced mesenchymal cell proliferation.     

Osteoinductivity of commercially available demineralized bone matrix. Preparations in a spine fusion model

BACKGROUND: Although autogenous bone is the most widely used graft material for spinal fusion, demineralized bone matrix preparations are available as alternatives or supplements to autograft. They are prepared by acid extraction of most of the mineralized component, with retention of the collagen and noncollagenous proteins, including growth factors. Differences in allograft processing methods among suppliers might yield products with different osteoinductive activities. The purpose of this study was to compare the efficacy of three different commercially available demineralized bone matrix products for inducing spinal fusion in an athymic rat model. METHODS: Sixty male athymic rats underwent spinal fusion and were divided into three groups of eighteen animals each. Group I received Grafton Putty; Group II, DBX Putty; and Group III, AlloMatrix Injectable Putty. A control group of six animals (Group IV) underwent decortication alone. Six animals from each of the three experimental groups were killed at each of three intervals (two, four, and eight weeks), and the six animals from the control group were killed at eight weeks. At each of the time-points, radiographic and histologic analysis and manual testing of the explanted spines were performed. RESULTS: The spines in Group I demonstrated higher rates of radiographically evident fusion at eight weeks than did the spines in Group III or Group IV (p < 0.05). Manual testing of the spines at four weeks revealed variable fusion rates (five of six in Group I, two of six in Group II, and none of six in Group III). At eight weeks, all six spines in Group I, three of the six in Group II, and no spine in Group III or IV had fused. Histologic analysis of the spines in Groups I, II, and III demonstrated varying amounts of residual demineralized bone matrix and new bone formation. Group-I spines demonstrated the most new bone formation. CONCLUSIONS: This study demonstrated differences in the osteoinductive potentials of commercially available demineralized bone matrices in this animal model.