TISSUE ENGINEERED BONE REPAIR OF CALVARIAL DEFECTS USING-CULTURED PERIOSTEAL CELLS

Periosteum has been demonstrated to have cell populations, including c hondroprogenitor and osteoprogenitor cells, that can form both cartila ge and bone under appropriate conditions. In the present study, perios teum was harvested, expanded in cell culture, and used to repair criti cal size calvarial defects in a rabbit model. Periosteum was isolated from New Zealand White rabbits, grown in cell culture, labeled with th e thymidine analog bromodeoxyuridine for later localization, and seede d into resorbable polyglycolic acid scaffold matrices. Thirty adult Ne w Zealand White rabbits were divided into groups, and a single 15-mm d iameter full-thickness calvarial defect was made in each animal. In gr oup I, defects were repaired using resorbable polyglycolic acid implan ts seeded with periosteal cells. In group II, defects were repaired us ing untreated polyglycolic acid implants. In group III, the defects we re left unrepaired. Rabbits were killed at 4 and 12 weeks postoperativ ely. Defect sites were then studied histologically, biochemically, and radiographically. In vitro analysis of the cultured periosteal cells indicated an osteoblastic phenotype, with production of osteocalcin up on 1,25(OH)(2) vitamin D-3 induction. In vivo results at 4 weeks showe d islands of bone in the defects repaired with polyglycolic acid impla nts with periosteal cells (group I), whereas the defects repaired with untreated polyglycolic acid implants (group II) were filled with fibr ous tissue. Collagen content was significantly increased in group I co mpared with group II (2.90 +/- 0.80 rho g/mg dry weight versus 0.08 +/ - 0.11 mu g/mg dry weight, p < 0.006), as was the ash weight (0.58 +/- 0.11 mg/mg dry weight versus 0.35 +/- 0.06 mg/mg dry weight, p < 0.01 5). At 12 weeks there were large amounts of bone in group I, whereas t here were scattered islands of bone in groups II and III. Radiodensito metry demonstrated significantly increased radiodensity of the defect sites in group I, compared with groups II and III (0.740 +/- 0.250 OD/ mm(2) versus 0.404 +/- 0.100 OD/mm(2) and 0.266 +/- 0.150 OD/mm(2), re spectively, p < 0.05). Bromodeoxpuridine label, as detected by immunof luorescence, was identified in the newly formed bone in group I at bot h 4 and 12 weeks, confirming the contribution of the cultured perioste al cells to this bone formation. This study thus demonstrates a tissue -engineering approach to the repair of bone defects, which may have cl inical applications in craniofacial and orthopedic surgery.