ARTHRITIS & RHEUMATISM Vol. 54, No. 1, January 2006, pp 76–81 DOI 10.1002/art.21528 © 2006, American College of Rheumatology Anti–Tumor Necrosis Factor Therapy Increases Synovial Osteoprotegerin Expression in Rheumatoid Arthritis Anca Irinel Catrina, 1 Erik af Klint, 2 Sofia Ernestam, 3 Sergiu-Bogdan Catrina, 1 Dimitrios Makrygiannakis, 2 Ileana Ruxandra Botusan, 2 Lars Klareskog, 2 and Ann-Kristin Ulfgren 2 Objective. Treatment of rheumatoid arthritis (RA) with tumor necrosis factor (TNF)–blocking agents, including etanercept and infliximab, has resulted in reductions in the radiographic progression of RA. How- ever, the exact mechanism by which this protection occurs has not been determined. In order to add to such knowledge, we investigated the effect of anti-TNF ther- apy on the expression of osteoprotegerin (OPG) and receptor activator of NF-B ligand (RANKL) in syno- vial tissue. Methods. The expression of OPG and RANKL in synovial biopsy specimens was evaluated by immunohis- tochemistry. Serial synovial biopsy specimens were ob- tained from 18 patients with RA, before and after treatment with etanercept (9 patients) or infliximab (9 patients). Biopsy specimens were evaluated by double- blind semiquantitative analysis and image analysis. The in vitro effect of TNF antagonists on the RANKL/OPG expression in osteoblasts and endothelial cells was evaluated by Western blotting. Statistical analysis was performed using Wilcoxon’s signed rank test, followed by the Bonferroni correction for multiple comparisons of paired samples. The results of in vitro experiments were evaluated by one-way analysis of variance, with Tukey’s post hoc test. Results. Treatment with both infliximab and et- anercept increased the expression of OPG in synovial tissue. After 8 weeks of treatment, neither infliximab nor etanercept influenced RANKL expression. In both groups of patients, the RANKL:OPG ratio decreased following therapy. In vitro, both of the TNF antagonists mimicked the in vivo effect, inducing a decrease in the RANKL:OPG ratio in TNF-primed osteoblasts and en- dothelial cells. Conclusion. Therapy with TNF antagonists in RA modulates the OPG/RANKL system, a potential mech- anism that could explain the retardation of radio- graphic damage observed following anti-TNF therapy. Rheumatoid arthritis (RA) is a chronic inflamma- tory disease associated with skeletal complications, such as focal bone erosions at the site of inflammation and sys- temic osteopenia, that lead to joint-related disability. The formation and activation of osteoclasts at the cartilage– pannus junction appear to be essential for RA-associated bone loss. These processes are driven by the interaction between receptor activator of NF-B ligand (RANKL) and its decoy receptor, osteoprotegerin (OPG) (1). A link between bone biology and the immune system is consistent with the finding that several cyto- kines known to mediate immune functions are also implicated in bone resorption. For instance, tumor ne- crosis factor (TNF) is one of the more potent osteoclas- togenic cytokines produced in the setting of inflamma- tion and is able to induce RANKL expression on the surface of stromal/osteoblast cells (2). The importance Supported by grants from the Swedish Medical Research Council, the Swedish Rheumatism Association, AFA Insurance, the Freemason Lodge Barnhuset, Stockholm, the King Gustav V 80-Year Foundation, and the So ¨derberg Foundation. Dr. Botusan’s work was supported by a Marie Curie fellowship within the Eurogendis program. 1 Anca Irinel Catrina, MD, PhD, Sergiu-Bogdan Catrina, MD, PhD: Karolinska University Hospital, Karolinska Institutet, Solna, Swe- den, and Carol Davila University of Medicine, Bucharest, Romania; 2 Erik af Klint, MD, Dimitrios Makrygiannakis, MD, Ileana Ruxandra Botusan, MD, Lars Klareskog, MD, PhD, Ann-Kristin Ulfgren, PhD: Karolinska University Hospital, Karolinska Institutet, Solna, Sweden; 3 Sofia Ernes- tam, MD: Karolinska University Hospital, Karolinska Institutet, Hud- dinge, Sweden. Dr. A. I. Catrina has received consulting fees (less than $10,000 per year) from Centocor. Dr. Klareskog has been a consultant for but did not receive any fee from Centocor and Wyeth. Address correspondence and reprint requests to Anca Irinel Catrina, MD, PhD, Rheumatology Research Laboratory, CMM L8:04, Karolinska University Hospital, S-17176 Stockholm, Sweden. E-mail: Anca.Catrina@ki.se. Submitted for publication February 8, 2005; accepted in revised form September 23, 2005. 76