In Japan, the system for bone banking is inadequate; thus, the frequency of use of artificial bone is high in comparison with that in North America and Europe. Accordingly, research on new materials for artificial bone is actively pursued. There are two types of bone formation: osteoconduction and osteoinduction. With osteoconduction, bone tissue grows along the material, whereas with osteoinduction, tissue containing no bone cells differentiates to form new bone tissue in muscle or under the skin. Artificial bone such as hydroxyapatite is superior in affinity to bone; it has good osteoconductivity but it lacks bone-inductive ability. The bone induction research of M.R. Urist1 in the 1960s showed that osteoinduction was dependent on a protein derived from the bone matrix, which he designated as bone morphogenetic protein (BMP). In 1988, J.M. Wozney and colleagues2 succeeded in cloning BMP, and recombinant BMP has been applied clinically in vertebral body fusion and the treatment of pseudoarthrosis, among other conditions. However, in Japan, the clinical use of BMPs is not permitted. The issue of why we cannot use BMPs in Japan is a very important one; but that is not the purpose here, which is instead to discuss the osteoinduction of biomaterials.

Generally, it is thought that osteoinduction is the fundamental difference between autograft bone and artificial bone. However, there have been a number of reports that bone is induced inside artificial bone with a special porous structure implanted into muscle. Ripamonti3 from South Africa was the first to report osteoinduction in the pores of hydroxyapatite made of coral. In his experiments, the hydroxyapatite was implanted in the muscles of baboons. Collaborative research between a Dutch group and a group in Chengdu, China,4,5 revealed that similar bone formation was seen with porous ceramics of various types of hydroxyapatite, tricalcium phosphate, and Bioglass® (NovaBone Products, Alachua, FL, USA). Fujibayashi and others in our department6 also reported that porous titanium processed with alkali-heat treatment has osteoinduction. This material contains neither calcium nor phosphate, suggesting that these elements may not be essential for inducing bone formation. The bone induction ability of a material may thus be dependent on both the macroporous structure and nanolevel microstructure of the surface. Furthermore, there are several additional features to consider. This phenomenon does not occur in rodents, in contrast with BMP, where small amounts of BMP are sufficient for osteoinduction in rodents but a large amount is necessary in dogs, sheep, and primates. In addition, bone tissue remains inside the pores for more than 1 year.6,7 For bone that does not receive loading, it is common sense to think that it should be resorbed. However, with porous materials implanted inside muscle, no loading is received; nevertheless, bone tissue persists for 1 year or more. The current explanation for this phenomenon is that BMP adsorption occurs as part of the mechanism of bone induction, but evidence that clearly indicates that this mechanism does not exist, and actual elucidation of the mechanism is still awaited.

From the point of view of the study of biomaterials, it is interesting that the porous structure plays an important role in the materials’ function. Recently, understanding has been gained into how blood stem cells from the marrow choose the place where they exist, their so-called niche. The responses to the porous material of artificial bone suggest the possibility that the well-controlled structure of biomaterials induces cell and tissue differentiation. The understanding of this phenomenon will open new fields of research into interactions between materials and living tissue.