Evaluation of 2 novel approaches for assessing the ability of demineralized freeze-dried bone allograft to induce new bone formation

Background: Because of the wide variation in the ability of human demineralized freeze-dried bone allograft (DFDBA) to reproducibly induce new bone formation, there is a need for a reliable measure of bone induction activity. In this study we examined an immature osteoprogenitor cell line for its potential utility in measuring the activity of DFDBA in vitro. Methods: We characterized the response of 2T9 cells, an immature osteoprogenitor cell line derived from the calvariae of transgenic mice containing the SV40 T-antigen driven by the mouse bone morphogenetic protein (BMP)-2 promoter, to recombinant human BMP-2 by measuring alkaline phosphatase specific activity, osteocalcin production, and matrix mineralization, Responses were compared to those obtained with 1,25-(OH)(2)D-3. In addition, 2T9 cells were cultured with active or inactive human DFDBA in the presence or absence of BMP-2. We also tested the hypothesis that radio-opacity of tissue following implantation of DFDBA in vivo correlates with the ability of human DFDBA to induce new bone. DFDBA from 9 different donors, stratified by age, were implanted subcutaneously in the thorax of 18 nude (nu/nu) mice. Tissue was harvested at 36 days postoperatively and examined histologically and biochemically for calcium and phosphorus uptake. Results: 2T9 cells exhibited a dose- and time-dependent response to soluble BMP-2. Proliferation was decreased and alkaline phosphatase activity, osteocalcin production, and mineralized nodule formation were increased. The effects were dose- and time-dependent. Peak effects on alkaline phosphatase and osteocalcin were noted on day 8, whereas mineral deposition did not begin to occur until day 12, 1,25-(OH)2D3 did not regulate these effects unless used with BMP-2, When the cells were exposed to active or inactive DFDBA in the presence or absence of BMP-2, no effect on 2T9 cell differentiation was observed. This indicated that DFDBA released no soluble factors with bone inductive ability and that if any active factors were adsorbed to the DFDBA, they were inactivated. When DFDBA was implanted subcutaneously in the thorax of nude mice, there was no histologic evidence of new bone formation. However, there was a donor age-dependent decrease in Ca and P uptake of the implanted tissue, reflecting a donor age-dependent decrease in remineralization of DFDBA. Conclusions: These data indicate that cell culture assays like the one used in this study may not be appropriate indicators of bone induction ability by DFDBA since soluble factors may not be responsible for bone induction in vivo. Nonetheless, in vitro assays are still needed. While Ca and P uptake by DFDBA-implanted tissue in the present study correlated with the age-dependent decrease in bone induction at intramuscular sites in a previously reported study, these data show that early x-rays may actually detect remineralization and not new bone formation. Thus, assessment of bone induction ability may still depend on histologic analysis of animal models.