Global ETD Search

Return to search

Investigation of activated remodelling in the healing of experimental stress fractures and the influence of anti-inflammatory treatments

Investigation of activated remodelling in the healing of experimental stress fractures and the influence of anti-inflammatory treatments Lisa Jane Kidd Abstract Targeted and focal remodelling are important processes in bone homeostasis and pathology. However, the factors that initiate and direct remodelling to repair microcracks, or respond to excess loading are still poorly understood. The rat ulna-loading (RUL) model has been widely used to examine modelling and remodelling responses to axial cyclic loading. However the model has not yet been fully characterised. Stress fractures are common amongst athletes, dancers and military recruits, but there is almost no information available on the mechanism of healing of these fractures. Although cyclooxygenase-2 (COX-2) is a key mediator of bone resorption and bone formation, very little information is available on the effect of non-steroidal antiinflammatories (NSAIDs) on stress fracture healing. Remodelling may play a role in the pathogenesis of stress fractures, and there is growing interest in the potential use of bisphosphonates to prevent them. Nonetheless, the effect of bisphosphonates on stress fracture healing is not known. PMX53 is a C5a receptor antagonist developed as a novel anti-inflammatory agent. It is effective against inflammatory arthritis, but has not been tested in any fracture models. The aims of this study were to undertake a detailed examination of the histology, histomorphometry and gene expression of the healing and remodelling process initiated by RUL, and to use this model to determine the effects of selective and non-selective NSAIDs, a bisphosphonate and PMX53 on stress fracture healing. To characterise the RUL model, fatigue fractures were created by loading ulnae until displacement was observed to increase by between 4% and 50%. Ulnae were bulk-stained in basic fuchsin and processed for undecalcified histology. For all remaining experiments, loading was stopped when the displacement had increased by 10%. For detailed histology and histomorphometry, ulnae were decalcified, paraffin embedded and stained with toluidine blue, saffranin-O or for tartrate resistant acid phosphatase (TRAP). Ulnae were examined at 1, 2, 4, 6, 8, and 10 weeks after loading. The effects of DFU (a selective COX-2 inhibitor, 2 mg/kg po), ibuprofen (a non-selective NSAID, 30 mg/kg po) and PMX53 (10 mg/kg po) were examined at 2, 4 and 6 weeks after loading. Effects of risedronate (a bisphosphonate) were examined at a high (1.0 mg/kg po) and low dose (0.1 mg/kg po) at 2, 6 and 10 weeks after loading. RUL did not create isolated intracortical microcracks, but curvilinear fatigue fractures that occurred at a standard position in the medial cortex of the distal ulna diaphysis. These stress fractures induced rapid periosteal woven bone formation and direct intracortical remodelling along the fracture line that originated at the periosteum and progressed towards the medullary cavity. Basic multicellular units (BMUs) could be followed through serial sections extending along the fracture line towards the centre of the bone. Quantitative, real-time PCR was performed at 4 hours, 24 hours, 4 days, 7 days and 14 days after fatigue fracture. Following each period, bones were dissected and mRNA was extracted using standard protocols. Gene expression was compared between loaded and unloaded ulnae and to an unloaded control group. Four hours after loading, there was a marked, 220-fold increase (P<0.0001) in expression of Interleukin-6 (IL-6). There were also prominent peak increases in mRNA expression for Osteoprotegerin (OPG), cyclooxygenase-2 (COX-2), and vascular endothelial growth factor (VEGF) (all P<0.0001). At 24 hours there was a peak increase in mRNA expression for IL-11 (73-fold increase, P<0.0001). At 4 days there was a significant increase in mRNA expression for Bcl-2, COX-1, bone morphogenic protein (BMP)-2, insulin-like growth factor (IGF)-1, osteopontin (OPN), and stromal cell derived factor SDF-1. At 7 days there was a significant increase in mRNA expression of Receptor activator of nuclear factor kappa β ligand (RANKL) and OPN. The dramatic, early up-regulation of IL-6 and IL-11 suggests they play a central role in initiating signalling events for stress fracture healing. Treatment with PMX53 did not affect any measures of woven bone formation or stress fracture remodelling. There were no treatment effects of Ibuprofen or DFU on the area of woven bone. DFU treatment resulted in a significant reduction in the area of porosity (resorption) and BMU area along the fracture line at 2 weeks after fracture. Ibuprofen treatment resulted in a significant reduction in length and area of BMUs and new bone formation along the fracture line at 6 weeks (p < 0.05). This is the first report to demonstrate a negative effect on stress fracture healing of both a selective COX-2 inhibitor and a non-selective NSAID. These data confirm the importance of cyclooxygenase in bone resorption and formation during remodelling. Bisphosphonates are potent inhibitors of osteoclastic bone resorption Two, 6 and 10 weeks after loading, measures of resorption and new bone formation were significantly reduced along the fracture line by high dose risedronate treatment, but not by the low dose. Only the porosity along the fracture line 2 weeks after loading was significantly reduced by the low dose risedronate. The low dose more closely resembles the clinical dose used to treat patients. Woven bone formation and consolidation were not affected by the low or high doses of risedronate. In conclusion, fatigue fractures in the rat ulna are highly reproducible, induce exuberant periosteal woven bone formation, and heal by direct remodelling along the fracture line. Remodelling is associated with gene expression for molecules typically associated with bone resorption and formation, angiogenesis and cell signalling. Remodelling of the stress fracture line was adversely affected by treatment with selective and non-selective COX inhibitors, by high dose treatment with risedronate, but not by PMX53, a C5a antagonist.