Effect of COX-2 inhibition on tendon-to-bone healing and PGE2 concentration after anterior cruciate ligament reconstruction

Martin Sauerschnig, Josef Stolberg-Stolberg, Carmen Schmidt, Valerie Wienerroither, Michael Plecko, Karin Schlichting, Carsten Perka, Christian Dynybil

Non-steroidal anti-inflammatory drugs are commonly used to reduce pain and inflammation in orthopaedic patients. Selective cyclooxygenase-2 (COX-2) inhibitors have been developed to minimize drug-specific side effects. However, they are suspected to impair both bone and tendon healing. The objective of this study is to evaluate the effect of COX-2 inhibitor administration on tendon-to-bone healing and prostaglandin E (PGE2) concentration. Thirty-two New Zealand white rabbits underwent reconstructions of the anterior cruciate ligaments and were randomized into four groups: Two groups postoperatively received a selective COX-2 inhibitor (Celecoxib) on a daily basis for 3 weeks, the two other groups received no postoperative COX-2 inhibitors at all and were examined after three or 6 weeks. The PGE2 concentration of the synovial fluid, the osseous integration of the tendon graft at tunnel aperture and midtunnel section, as well as the stability of the tendon graft were examined via biomechanic testing. After 3 weeks, the PGE2 content of the synovial fluid in the COX-2 inhibitor recipients was significantly lower than that of the control group (p = 0.018). At the same time, the COX-2 inhibitor recipients had a significantly lower bone density and lower amount of new bone formation than the control group (p = 0.020; p = 0.028) in the tunnel aperture. At the 6-week examination, there was a significant increase in the PGE2 content within synovial fluid of the COX-2 inhibitor recipients (p = 0.022), whose treatment with COX-2 inhibitors had ended 3 weeks earlier; in contrast, the transplant stability decreased and was reduced by 37% compared to the controls. Selective COX-2 inhibitors cause impaired tendon-to-bone healing, weaken mechanical stability and decrease PGE2 content of the synovial fluid. The present study suggests a reluctant use of COX-2 inhibitors when tendon-to-bone healing is intended.