S. Agarwal1, J. Peterson1, S. Loder1, O. Eboda1, C. Brownley1, K. Ranganathan1, D. Fine1, K. Stettnichs1, A. Mohedas2, P. Yu2, S. Wang1, S. Buchman1, P. Cederna1, B. Levi1 1University Of Michigan,Surgery,Ann Arbor, MI, USA 2Brigham And Women’s Hospital,Boston, MA, USA
Introduction: Treatment options for heterotopic ossification (HO) including surgical excision and radiation cause tissue damage and result in recurrence. Prophylactic treatment with non-steroidal anti-inflammatory drugs (NSAIDs) including aspirin and celecoxib has been reported to decrease ectopic bone formation in patients after orthopedic procedures although the pathway remains unexplored. Here we demonstrate that administration of celecoxib, a Cox-2 specific inhibitor, decreases HO formation following trauma independent of bone morphogenetic protein receptor (BMPR) function.
Methods: For our HO model, male C57BL/6 mice underwent Achilles tenotomy of the left hindlimb with 30% total body surface area partial-thickness burn over the dorsum. Mice were administered intraperitoneal celecoxib or carrier daily. HO was quantified by micro-CT imaging at 2 week intervals up to 9 weeks after trauma (threshold Hounsfield units 1250). Wound-healing was assessed by daily imaging. To analyze BMP signaling, we used the BRE-luc reporter in C2C12 cells in vitro. C2C12 cells were administered indomethacin (cox-1/cox-2 inhibitor) in the presence of BMP2, BMP4, BMP6, or BMP9 followed by quantification of luciferase activity.
Results: Administration of celecoxib resulted in an 80 percent decrease in HO formation at 7 weeks (5.06 mm3 v. 1.30 mm3, p<0.05) and 77 percent decrease at 9 weeks (5.61 mm3 v. 1.55 mm3) after trauma. In contrast to the carrier-treated group, HO formation was undetectable for the first 3 weeks in the celecoxib-treated group. All Achilles tenotomy incision and dorsal burn sites showed grossly normal healing. Mesenchymal stem cells from burned mice treated with 1 uM celecoxib in vitro demonstrated 60% less alkaline phosphatase staining and 75% less alizarin red staining than untreated cells. Finally, in vitro administration of indomethacin to C2C12 cells with the BRE-luc reporter resulted in no significant decrease in luciferase activity, suggesting that Cox-2 inhibition does not inhibit BMP receptor function.
Conclusion: We demonstrate that Cox-2 inhibition decreases HO volume in a burn/trauma model. Decreased mineral deposition also occurs in the in vitro setting with mesenchymal stem cells, suggesting a direct effect on the cells responsible for bone formation. Furthermore, the BRE-luc reporter assay demonstrates that Cox-2 inhibition likely does not impart its effect through the BMP pathway. Our findings suggest that therapeutic targets for HO need not be limited to the BMP pathway, and that the Cox-2 enzyme deserves further attention. Patients at risk for HO following trauma may benefit from early celecoxib treatment with minimal impact on wound healing.
