The morselized and impacted bone graft. Animal experiments on proteins, impaction and load.

The results of primary hip replacements are good. However, dealing with a loose prosthesis has been problematic, especially when major bone deficiencies are encountered. These problems appear to have been solved by the introduction of the Slooff-Ling method of using morselized and impacted allograft chips. The clinical results are excellent in the hands of the innovators. However, it remains confusing that a thick layer of dead, broken, immunogenic tissue taken from another individual does not resorb and collapse during remodeling. Still harder to understand is the impression, as judged by radiography, that this thick layer seems to incorporate and remodel up to a distance of perhaps 10 mm or more from the host bone, whereas the ingrowth distance into a non-morselized graft is limited to a few mms. To clarify the biological basis of the morselized and impacted grafts better, the present study was stated. Three hypotheses were initially proposed to explain the good clinical results: 1. Morselization releases growth factors present in the graft (osteoinduction). 2. Impaction makes it easier for the ingrowing bone to climb up into the graft (osteoconduction). 3. The compliance or elasticity of the graft allows the load to produce deformations that stimulate bone formation (mechanical load). In the first studies, bone chambers were implanted in rats and the distance of new bone ingrowth into a graft in the chamber was measured. In Paper I, a morselized graft was deproteinized by slow heating under high pressure. Ingrowing bone did not reach so far into the deproteinized graft as into a non-treated one. We concluded that the proteins present in the graft partly determine how far ingrowing new bone will extend into a graft. In Paper II, a cancellous graft was impacted so that the bone volume fraction of the graft rose from 35% in the unimpacted to 65%. The impacted grafts were compared to unimpacted ones and it was shown that impaction reduced the ingrowth of new bone into a graft in the chamber at six weeks. In Paper III, this somewhat unexpected finding was further studied. Syngeneic and allogeneic grafts showed a reduced ingrowth distance at six weeks when impacted, compared to unipacted controls. However, the reduction was not found when the time for ingrowth was extended to 12 weeks, indicating a possible catch-up phenomenon. Moreover, an exogenously applied growth factor, osteogenic protein-1, was found to have increased the ingrowth distance of new bone into impacted grafts at six weeks. In Paper IV, a rabbit knee prosthesis was developed to study the effect of a mechanical load on the remodeling of a morselized and impacted graft. All rabbits had their tibial marrow cavity cleansed of cancellous bone, which was replaced by a morselized and impacted bone graft. Six rabbits received a complete tibial prosthesis with a tibial load-bearing tray and a stem transferring the load to the impacted graft with each step made by the rabbit. Another six rabbits had only the stem, without the tibial tray, inserted into the impacted graft. With this design, the load from walking was not transferred to the graft, since there was no joint surface replacement to transfer the load to the stem and the graft. Thus, the graft was loaded in rabbits receiving a full prosthesis, whereas it was unloaded in the animals receiving only the stem. New bone formation and resorption of the graft were increased in the loaded grafts, and we concluded that a load increases the rate or speed of remodeling. In Paper V, four patients were operated on for vertebral fractures. The fractures were stabilized by plates and the vertebral bodies packed with autogenous morselized graft. After 1.5 years, when the fractures were clinically and radiographically healed, a biopsy was taken. It was found, that even after such a long time, large areas remained unremodeled and sometimes even unrevascularized. In some parts, necrotic graft trabeculae were embedded in fibrous vasc