Osteotropic biomolecules must be introduced into tissues via a “carrier”, which functions to retain the biomolecules at the desired site. This is important because local cell populations must be exposed to bioactive molecules at suprathreshold concentrations and for a period of time sufficient to initiate cellular responses.
Three approaches to delivering biomolecules to the tissue-implant interface have been described in the literature. The simplest is to dip an implant into a solution containing the biomolecule of interest just prior to surgical insertion of the device. The primary problem with this method is that it does not provide control over the release of the biomolecule; within minutes there is in an initial burst release, followed by no further release or a decreasing release rate. Although some degree of response is observed, because of rapid desorption and clearance of biomolecules the responses are not optimal. Another commonly used approach involves the use of biodegradable polymers to release proteins at the tissue-implant interface. Such devices can be formulated to reduce or even eliminate the initial burst release and to exhibit relatively linear release through approximately 11 wk. The third method, described here, involves direct immobilization of biologically active molecules on the surface of bulk biomaterial samples. Thus, extremes in biomolecule delivery have been described: quick burst release, slow extended release, and no release. None of these methods, however, really mimic the way biomolecules are synthesized during embryogenesis and wound healing. A temporal sequence of multiple growth factors is observed during fracture healing. Analysis of callus formation following fracture has shown that platelet-derived growth factor predominates in the first three days, fibroblast growth factor from 3 to 6 days, insulin-like growth factor from 6 to 9 days, and transforming growth factor from 9 to 21 days in the wound healing cycle [Int. J. Dev. Biol. 37, 573, 1993]. Other osteogenic [J. Bone Miner. Metab. 18, 63, 2000], chondrogenic [J. Bone Miner. Res. 15, 1014, 2000], angiogenic [Clin. Orthop., 224, 2000], and neurotrophic [Bone 26, 625, 2000] molecules are also expressed in fracture callus. Controlled release techniques have the potential to recreate naturally occurring concentration profiles of cytokines and growth factors.
We developed three systems for creating temporally varying profiles of one or more biologically active proteins. Depending on the polymer used, time-dependent concentration profiles in three times scales, ranging from three or four days to three or four weeks, were achieved by multilayering and heterogeneously loading the layer. Representative results for a three-layered devices are shown in the figure to the right. Distinct release peaks are evident. One or more biomolecules can be released with each peak, and the concentration can be controlled by changing the loading of each layer.