Ex vivo gene therapy in autologous critical-size craniofacial bone regeneration.

In therapeutic bone repairs, autologous bone grafts, conventional or vascularized allografts, and biocompatible artificial bone substitutes all have their shortcomings. The bone formed from peptides [recombinant human bone morphogenetic proteins (BMPs)], demineralized bone powder, or a combination of both is small in size. Tissue engineering may be an alternative for cranial bone repair. In this study, the authors developed an animal model to test the hypothesis that replication-defective, adenovirus-mediated human BMP-2 gene transfer to bone marrow stromal cells enhances the autologous bone formation for repairing a critical-size craniofacial defect. The mesenchymal stromal cells of miniature swine were separated from the iliac crest aspirate and expanded in monolayer culture 1 month before implantation. The cultured mesenchymal stromal cells were infected with recombinant, replication-defective human adenovirus BMP-2, 7 days before implantation. Bilateral 2 x 5-cm2 cranial defects were created, leaving no osteogenic periosteum and dura behind. Mesenchymal stromal cells at 5 x 10(7)/ml were mixed with collagen type I to form mesenchymal stromal cell/polymer constructs. Mesenchymal stromal cells used for the control site were infected with adenovirus beta-Gal under the same conditions. After 6 weeks and 3 months, 10 miniature swine were euthanized and the cranium repair was examined. Near-complete repair of the critical-size cranial defect by tissue-engineered mesenchymal stromal cell/collagen type I construct was observed. The new bone formation area (in square centimeters) measured by three-dimensional computed tomography demonstrated that the improvement from 6 weeks to 3 months was significantly greater on the experimental side than on the control side (2.15 cm2 versus 0.54 cm2, p < 0.001) and significantly greater at 3 months than at 6 weeks (2.13 cm2 versus 0.52 cm2, p < 0.001). The difference between the experimental and control groups was significant at 3 months (mean difference, 2.13 cm2; p < 0.001). The maximal compressive strength of the new bone was similar to that of the normal cranial bone when evaluated by biomechanical testing (cranium bone versus tissue-engineered bone, 88.646 +/- 5.121 MPa versus 80.536 +/- 19.302 MPa; p = 0.227). Adenovirus was absent from all constructs by immunochemical staining at 6 weeks and 3 months after implantation. The successful repair of cranial defects in this experiment demonstrates the efficacy of the integration of the autologous stem cell concept, gene medicine, and polymers in producing tissue-engineered bone.