gms | German Medical Science

Objectives: Recent studies failed to demonstrate a clear clinical advantage of bone grafts (BGs) combined with BMP over BGs alone for the treatment of osseous defects. Within this context, it has been suggested that BMPs might lead to accelerated graft resorption. Whilst the effects of BMPs on mesenchymal cells have been extensively studied in various investigative approaches, their influence on the hematopoietic cell system including osteoclasts remains surprisingly elusive.

Methods: In the present study, we transplanted human cancellous bone with (BMP+) or without (BMP-) rhBMP-7 subcutaneously into immuno-incompetent murine hosts (NOD/Scid, n= 2x6). This ectopic model was chosen to exclude any effects of loading on the osteogenic capacity of the BGs and to study the biological effects of rhBMP-7 on the grafts only. Fourteen weeks post transplantation, specimens were harvested and further analysed by micro-CT, histology, immunohistochemistry, histomorphometry and flow cytometry. This transplantation model enabled us to clearly differentiate between newly formed bone (murine) and BGs (human) by immunohistochemical staining for human-specific extracellular matrix proteins such as collagen I and osteocalcin. Staining for TRAP, CD45, CD3, CD19/20 and CD68 was performed to identify cells of the hematopoietic lineages. Flow cytometry analyses were performed on the retrieved specimens and the corresponding mouse femora to analyse their cellular composition and content of hematopoietic cells.

Results and conclusion: BGs implanted along with rhBMP-7 recapitulated the morphology of a physiological organ bone with a cortex-like outer structure, an inner trabecular network and a physiological bone marrow compartment. In contrast, BGs implanted without rhBMP-7 were not remodelled and were loosely connected with fibrous tissue. Though overall bone volume was significantly higher in the BMP+ group than in the BMP- group (835 vs 365 mm3, p<0.001), bone volume to total volume was significantly lower (31 vs 47 %, p=0.002). Bone marrow spaces were larger in the BMP+ group (0.47 vs 0.28 mm, p=0.002). Histomorphometry revealed that the ratio of BGs to total area was significantly smaller in the BMP+ group (0.14 vs 0.34, p=0.004) whilst the ratio of newly formed bone to total area was higher (0.12 vs 0.001, p<0.001). Numbers of osteoclasts per bone perimeter were significantly higher in the BMP+ group (p=0.002). Immunohistochemical staining and flow cytometry demonstrated that the bone marrow spaces in the BMP+ group but not in the BMP- group were filled with a functional hematopoietic cell system comparable to the one found in the mouse femur.

To our knowledge, we are the first to show that rhBMP-7 supports the development of a functional bone marrow compartment comparable to the one found in native bone. Furthermore, our results clearly demonstrate that though rhBMP-7 induces new bone formation, the resorption rate of BGs is higher than the rate of new bone formation.