Fabrication of polycaprolactone-silanated β-tricalcium phosphate-heparan sulfate scaffolds for spinal fusion applications.

Interbody spinal fusion relies on the use of external fixation and the placement of a fusion cage filled with graft materials (scaffolds) without regard for their mechanical performance. Stability at the fusion site is instead reliant on fixation hardware combined with a selected cage. Ideally, scaffolds placed into the cage should both support the formation of new bone and contribute to the mechanical stability at the fusion site. We recently developed a scaffold consisting of silane-modified PCL-TCP (PCL-siTCP) with mechanical properties that can withstand the higher loads generated in the spine. To ensure the scaffold more closely mimicked the bone matrix, we incorporated collagen (Col) and a heparan sulfate glycosaminoglycan sugar (HS3) with increased affinity for heparin-binding proteins such as bone morphogenetic protein-2 (BMP-2). The osteostimulatory characteristic of this novel device delivering exogenous BMP2 was assessed in vitro and in vivo as a prelude to future spinal fusion studies with this device. A combination of cell-free assays (BMP2 release), progenitor cell-based assays (BMP2 bioactivity, cell proliferation and differentiation), and rodent ectopic bone formation assays was used to assess the osteostimulatory characteristics of the PCL-siTCP-based scaffolds. Freshly prepared rat mesenchymal stem cells were used to determine reparative cell proliferation and differentiation on the PCL-siTCP-based scaffolds over a 28-day period in vitro. The bioactivity of BMP2 released from the scaffolds was assessed on progenitor cells over a 28-day period using ALP activity assays and release kinetics as determined by enzyme-linked immunosorbent assay. For ectopic bone formation, intramuscular placement of scaffolds into Sprague Dawley rats (female, 4 weeks old, 120-150 g) was achieved in five animals, each receiving four treatments randomized for location along the limb. The four groups tested were (1) PCL-siTCP/Col (5-mm diameter×1-mm thickness), PCL-siTCP/Col/BMP2 (5 µg), (3) PCL-siTCP/Col/HS3 (25 µg), and (4) PCL-siTCP/Col/HS3/BMP2 (25 and 5 µg, respectively). Bone formation was evaluated at 8 weeks post implantation by microcomputed tomography (µCT) and histology. Progenitor cell-based assays (proliferation, mRNA transcripts, and ALP activity) confirmed that BMP2 released from PCL-siTCP/Col/HS3 scaffolds increased ALP expression and mRNA levels of the osteogenic biomarkers Runx2, Col1a2, ALP, and bone gla protein-osteocalcin compared with devices without HS3. When the PCL-siTCP/Col/HS3/BMP2 scaffolds were implanted into rat hamstring muscle, increased bone formation (as determined by two-dimensional and three-dimensional µCTs and histologic analyses) was observed compared with scaffolds lacking BMP2. More consistent increases in the amount of ectopic bone were observed for the PCL-siTCP/Col/HS3/BMP2 implants compared with PCL-siTCP/Col/BMP2. Also, increased mineralizing tissue within the pores of the scaffold was seen with modified-tetrachrome histology, a result confirmed by µCT, and a modest but detectable increase in both the number and the thickness of ectopic bone structures were observed with the PCL-siTCP/Col/HS3/BMP2 implants. The combination of PCL-siTCP/Col/HS3/BMP2 thus represents a promising avenue for further development as a bone graft alternative for spinal fusion surgery.