Article Text

Download PDFPDF

Opposite relationships between circulating Dkk-1 and cartilage breakdown in patients with rheumatoid arthritis and knee osteoarthritis
  1. N Voorzanger-Rousselot1,
  2. N Charni Ben-Tabassi1,
  3. P Garnero1,2
  1. 1
    Molecular Markers, CCBR-SYNARC, Lyon, France
  2. 2
    INSERM Research Unit 664, Lyon, France
  1. Correspondence to Dr P Garnero, CCBR-SYNARC, 16 rue Montbrillant, 69003 Lyon, France; patrick.garnero{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Recent animal and clinical studies have suggested that Dkk-1, a secreted inhibitor of the canonical Wnt signalling pathway, could play an important role in mediating the alterations of joint tissue turnover associated with rheumatoid arthritis (RA) and osteoarthritis. Diarra et al1 showed that blockage of Dkk-1 abolished bone erosion in an inflammatory mouse model and that circulating Dkk-1 was increased in patients with active RA. More recently, we reported that increased Dkk-1 was associated with a higher risk of radiological bone erosion in patients with early RA receiving anti-tumour necrosis factor therapy.2 On the other hand, elevated circulating levels of Dkk-1 were reported to be moderately associated with reduced radiological progression of hip osteoarthritis in one study of elderly women.3 The relationship between Dkk-1 and cartilage turnover in patients with RA and osteoarthritis is, however, unknown.

In this study we compared circulating Dkk-1 measured by ELISA (Biomedica, Vienna, Austria)2 in 55 patients with active RA (65% women; mean age 55.4 years (SD 14.1), median disease duration 11 years), 85 subjects with knee osteoarthritis (74% women; mean age 62.8 years (SD 7.7), radiological Kellgren–Lawrence score II–III, median disease duration 3.45 years) and 93 healthy sex and age-matched controls and investigated its association with biochemical markers of cartilage degradation. All patients met the American College of Rheumatology criteria.

Circulating Dkk-1 levels were on average 479% higher (p<0.001) in patients with RA, but 37% (p<0.0001) lower in subjects with knee osteoarthritis (table 1) compared with healthy controls (3724 pg/ml; SD 1179). In patients with RA, Dkk-1 levels were associated with pain (r  =  0.40, p<0.001) and disease activity scores in 28 joints (r  =  0.25, p = 0.06), but not with C-reactive protein levels (p = 0.58) as previously reported.2 Dkk-1 were also positively associated with serum levels of the type II collagen degradation markers Helix-II (r  =  0.49, p<0.001) and C2C (r  =  0.60, p<0.001).4 The 33 (60%) RA patients with Dkk-1 levels above the 95th percentile of healthy controls had 38% (p<0.05) higher Helix-II levels than the 21 patients with low Dkk-1 values. Conversely, in patients with knee osteoarthritis, Dkk-1 levels correlated negatively with C2C (r  =  −0.24, p = 0.031) but not with Helix II (p = 0.93). No correlation was found between Dkk-1 and C-reactive protein in osteoarthritis subjects (p = 0.35).

Table 1

Serum concentrations of Dkk-1 and markers of cartilage degradation in patients with knee osteoarthritis and RA

This study showed that the relationship between Dkk-1 and cartilage degradation varies between RA and osteoarthritis patients, probably reflecting difference in the underlying mechanisms of joint destruction. In RA, animal and human studies have shown that tumour necrosis factor induces the overexpression of Dkk-1 by synovial fibroblasts,1 leading to increased osteoclastic activity and resorption of calcified cartilage. This process involves cysteine proteases, including the osteoclastic cathepsin K and collagenases. This could explain the positive relationship between Dkk-1 and both Helix-II and C2C, which are markers of type II collagen degradation mediated by cathepsins and collagenases, respectively.5 Conversely, in osteoarthritis, superficial cartilage is mostly affected, a process believed to involve collagenases secreted by chondrocytes.6,7 In this respect it is interesting to note that in a mouse model of osteoarthritis8 and in human osteoarthritis patients9,10 the loss of sFRP-3, another inhibitor of the Wnt signalling pathway, was associated with increased cartilage damage, possibly by modulating the activity of collagenases. This may explain the negative association between Dkk-1 and C2C in patients with knee osteoarthritis.


The authors would like to thank Dr Gerhard Hawa from Biomedica for supplying Dkk-1 assays.


View Abstract


  • Competing interests None.