Biomechanical analysis of fracture risk associated with tibia deformity in children with osteogenesis imperfecta: a finite element analysis.

OBJECTIVES: Osteogenesis imperfecta (OI) frequently leads to long-bone bowing requiring a surgical intervention in severe cases to avoid subsequent fractures. However, there are no objective criteria to decide when to perform such intervention. The objective is to develop a finite element model to predict the risk of tibial fracture associated with tibia deformity in patients with OI. METHODS: A comprehensive FE model of the tibia was adapted to match bi-planar radiographs of a 7 year-old girl with OI. Ten additional models with different deformed geometries (from 2° to 24°) were created and the elasto-plastic mechanical properties were adapted to reflect OI conditions. Loads were obtained from mechanography of two-legged hopping. Two additional impact cases (lateral and torsion) were also simulated. Principal strain levels were used to define a risk criterion. RESULTS: Fracture risks for the two-legged hopping load case remained low and constant until tibia bowing reached 15° and 16° in sagittal and coronal planes respectively. Fracture risks for lateral and torsion impact were equivalent whatever the level of tibial bowing. CONCLUSIONS: The finite element model of OI tibia provides an objective means of assessing the necessity of surgical intervention for a given level of tibia bowing in OI-affected children.

Spatial control of bone formation using a porous polymer scaffold co-delivering anabolic rhBMP-2 and anti-resorptive agents.

Current clinical delivery of recombinant human bone morphogenetic proteins (rhBMPs) utilises freeze-dried collagen. Despite effective new bone generation, rhBMP via collagen can be limited by significant complications due to inflammation and uncontrolled bone formation. This study aimed to produce an alternative rhBMP local delivery system to permit more controllable and superior rhBMP-induced bone formation. Cylindrical porous poly(lactic-co-glycolic acid) (PLGA) scaffolds were manufactured by thermally-induced phase separation. Scaffolds were encapsulated with anabolic rhBMP-2 (20 µg) ± anti-resorptive agents: zoledronic acid (5 µg ZA), ZA pre-adsorbed onto hydroxyapatite microparticles, (5 µg ZA/2% HA) or IkappaB kinase (IKK) inhibitor (10 µg PS-1145). Scaffolds were inserted in a 6-mm critical-sized femoral defect in Wistar rats, and compared against rhBMP-2 via collagen. The regenerate region was examined at 6 weeks by 3D microCT and descriptive histology. MicroCT and histology revealed rhBMP-induced bone was more restricted in the PLGA scaffolds than collagen scaffolds (-92.3% TV, p < 0.01). The regenerate formed by PLGA + rhBMP-2/ZA/HA showed comparable bone volume to rhBMP-2 via collagen, and bone mineral density was +9.1% higher (p < 0.01). Local adjunct ZA/HA or PS-1145 significantly enhanced PLGA + rhBMP-induced bone formation by +78.2% and +52.0%, respectively (p ≤ 0.01). Mechanistically, MG-63 human osteoblast-like cells showed cellular invasion and proliferation within PLGA scaffolds. In conclusion, PLGA scaffolds enabled superior spatial control of rhBMP-induced bone formation over clinically-used collagen. The PLGA scaffold has the potential to avoid uncontrollable bone formation-related safety issues and to customise bone shape by scaffold design. Moreover, local treatment with anti-resorptive agents incorporated within the scaffold further augmented rhBMP-induced bone formation.