• Larsen score
• osteoprotegerin
• rheumatoid arthritis

Osteoprotegerin (OPG) is a soluble decoy receptor, produced by osteoblastic cells and in the inflamed synovium of RA by dendritic cells, B cells, and other immunocompetent cells.^1,2 It inhibits the differentiation of osteoclast precursor cells and the activation of mature osteoclasts by neutralising the receptor activator of NF-κB ligand (RANKL).³

RANKL, a member of the tumour necrosis factor family, is expressed on prae-osteoblasts and T lymphocytes. A soluble RANKL (sRANKL) can be produced by activated T lymphocytes or can be generated from the cell bound form by a protease. RANKL itself activates a receptor activator of NF-κB (RANK).⁴

RANKL, together with monocyte-colony stimulating factor, interleukin 1, and RANK is responsible for osteoclast formation and activation and inhibits osteoclast apoptosis. Thus OPG acts as antagonist to RANKL. An imbalance of this system may play a part in the skeletal complications of rheumatoid arthritis (RA).⁵

Our study aimed at comparing OPG and sRANKL in the serum of patients with RA.

We identified 44 patients with RA (24 female, 20 male, mean age at manifestation of RA 49 years) with 60 measurements. Sixteen patients received low dose steroids, five patients showed generalised osteoporosis (x ray and/or osteodensitometry).

The results were analysed by Spearman correlation statistics and Wilcoxon two sample test.

Serum OPG levels were measured in patients with RA using a sandwich-type enzyme linked immunosorbent assay (ELISA) based on two OPG-specific antibody preparations. The mean value of a healthy control group of 170 blood donors⁶ is 2.2 pmol/l (2.0 pmol/l for men, 2.4 pmol/l for women).

sRANKL was measured by an enzyme catalysed colour change detectable on a standard ELISA reader. To measure only the biologically active form(s) of sRANKL biosynthetic OPG/Fc was used as capture protein. The mean value of serum sRANKL levels in healthy subjects was calculated as 1.3 pmol/l (median 0.9).

We detected serum levels of OPG with a mean value of 4.2 pmol/l (SD 2.0) and serum levels of RANKL with a mean value of 0.9 pmol/l (SD 0.8). We found a significant correlation between OPG and erythrocyte sedimentation rate (ESR) and OPG and the Larsen score but no correlation between RANKL and OPG or between RANKL and the clinical and radiological measures (table 1).

No significantly different OPG levels were found either in patients receiving steroids or in patients with osteoporosis.

In RA RANKL leads to bone erosions by activation of osteoclasts, and this process is inhibited by OPG.⁷ Therefore OPG seems to play an important part in preventing erosions and osteoporosis in patients with RA. Kolarz et al suggested that patients with active inflammation may show higher OPG values owing to activation of several other cells.⁸ Haynes et al showed an increased expression of RANKL in tissues surrounding bone erosions and at the same time OPG was absent in tissues from patients with active RA.⁹

Despite the presence of raised serum OPG levels acting as protection mechanism, the local destructive effect of RANKL by activation of osteoclasts seems not to be fully balanced.

The up regulation of OPG might be a response to the inflammation; in contrast an up regulation of RANKL could not be found in the serum of patients with RA.

Haynes et al reported that OPG and RANKL behave differently depending on the cells which produce them.¹⁰ A further explanation may be the different strategies of both assays: the OPG assay measures free and bound OPG, the sRANKL assay only free sRANKL; complexes formed from OPG and sRANKL would therefore be detected only with the OPG, but not with the sRANKL assay.