The clinical application of bone morphogenetic proteins (BMPs) offers solutions to many challenging problems in orthopaedics. However, a practical clinical problem is to obtain a controlled release of the BMPs. The attachment of heparin to biomaterials may result in an appropriate matrix for the binding, and sustained release of BMPs. Binding of growth factors to heparin stabilizes these growth factors, protects them from proteolytic degradation, and prolongs the half-life of BMPs in culture media 20-fold. We created a carrier based delivery system with a localized sustained release by loading a tricalciumphosphate/hydroxyapatite (TCP/HA) bone substitute coated with cross-linked collagen and heparin, with BMP-7.

TCP/HA granules (BoneSave^™, Stryker Orthopaedics) were coated with collagen, and subsequently the collagen was cross-linked in the presence (TCP/HA-Col-Hep) and absence (TCP/HA-Col) of heparin. BMP-7 was loaded onto the coated TCP/HA granules. Morphology of the coated collagen with and without heparin, and release kinetics of BMP-7 from the granules were analyzed. TCP/HA granules without coating were used as controls.

Analysis showed a highly porous collagen network on both TCP/HA-Col and TCP/HA-Col-Hep granules. Immersion of the granules in BMP-7 solution, resulted in the binding of 54±3% (62.9±5.4 ng BMP-7/mg granule) to the TCP/HA granules, 64±8% (69.0±9.6 ng BMP-7/mg granule) to the TCP/HA-Col granules, and 78±1% (92.9±4.8 ng BMP-7/mg granule) to the TCP/HA-Col-Hep granules. TCP/HA granules showed a burst release of BMP-7 within the first 4 h. TCP/HA-Col granules showed an initial burst release, followed by a more gradual release. In contrast, BMP-7 release from the TCP/HA-Col-Hep granules was sustained up to 21 days.

The sustained delivery system for BMP-7 developed in this study may provide a powerful tool for bone regeneration. This system could probably also be applied to deliver multiple growth factors that have affinities for heparin, which could for instance synergistically enhance osteogenesis by increasing vascularity.

Type I and II collagen-based scaffolds, with and without attached chondroitine sulphate (CS), were implanted without additional chondrocytes into full-thickness defects in the trochlea of young adult rabbits. We hypothesise that the chemical composition of the matrix will have a direct effect on the speed of repopulation and the phenotypic expression of the subchondral repair cells.

Evaluation of the repair process was performed with routine histology and with two quantitative histological grading systems, four and twelve weeks after implantation.

Four weeks after implantation, type I collagenous scaffolds were completely filled with a cartilage-like repair tissue. By contrast, type II collagenous scaffolds showed a superficial zone of cartilaginous tissue, and in many defects chondrocyte-like cells at the interface of the implant material with the subchondral bone. In collagen type II filled lesions larger areas of the scaffolds were completely devoid of repair tissue. Control defects showed a repair reaction that was very similar to that observed in defects filled with a type I scaffold.

After 12 weeks, the subchondral defect was largely replaced by bone and the differences between the scaffolds were less pronounced. The quantitative blind score of the sections confirmed that the scores of the control defect and of the collagen type I based scaffolds were slightly higher as compared to the type II based scaffolds. Irrespective of the type of scaffold, there was a trend that the scaffolds with CS scored slightly higher than those without CS.

We conclude that different types of scaffold induce different repair reactions. Collagen types I based scaffolds seem superior to guide progenitor cells from a subchondral origin into the defect. Repair cells in collagen type II based scaffolds seem to assume a chondrocyte-like phenotype, which could have a negative effect on the mobility of the repair cells.