Cartilage and Bone Biomarkers in Rheumatoid Arthritis: Prediction of 10-Year Radiographic Progression - OP 0128
Authors: Syversen, Goll, Odegard, Gaarder, Haavardsholm, van der Heijde, Landewe, Kvein
Background:
Bone and cartilage destruction, in the form of marginal erosions and cartilage loss, are features of rheumatoid arthritis. In theory, elevated markers of bone and cartilage breakdown may be present in synovial fluid or in circulation at an early stage, permitting early detection of damage. Here, Syversen et al explore the ability of various bone and cartilage biomarkers to predict 10-year radiographic progression in an inception cohort of RA patients.
Methods:
Consecutive RA patients with RA disease duration < 4 years were recruited from a community-based outpatient clinic population in the Netherlands, Norway, and France and enrolled in the European Research on Incapacitating Disease and Social Support (EURIDISS). Hand and feet radiographs were obtained at baseline and after 5 and 10 years of follow-up. Stored baseline serum sample were assayed for bone and cartilage biomarkers: Receptor Activator for Nuclear Factor κ B Ligand (RANKL), osteoprotegerin (OPG), Collagen Type 1 C-telopeptide (CTX-1), chitinase 3-like protein 1 (YKL40), cartilage oligomeric matrix protein (COMP), and C2C (a breakdown product of Type II cartilage) and assessed for their ability to predict 10-year radiographic progression.
Results:
238 patients with RA less than 4 years were enrolled. Only 125 had complete follow-up radiographs to 10 years. While only 50% of patients were treated with DMARDs, 59% developed erosive changes over 10 years. Several biomarkers (ESR, CRP, YKL40, and CTX1) were significantly higher at baseline in future progressors compared to non-progressors. However, only CTX1 was a significant univariate predictor of 10-year radiographic progression using simple linear regression. Using multivariable linear regression, CTX1 remained significantly associated with 10-year radiographic progression in the final model including RF, anti-CCP antibody status, baseline erosions, and baseline CRP. However, the predictive ability of CTX1 was weak compared to commonly used predictors (RF, CCP, baseline erosions, and baseline CRP).
Conclusions:
Among select biomarkers of bone and cartilage breakdown and repair, only baseline CTX1 (and only weakly) was significantly associated with 10-year radiographic progression in early RA.
Editorial Comment:
Although firmly based in theory, the use of serum biomarkers to predict events at the level of the synovium is problematic and has not proven successful in a variety of settings. It is possible that assaying the same biomarkers in synovial fluid or synovial biopsy specimens would yield more robust results. However, these samples are more difficult to obtain, more expensive to process, and are problematic to obtain longitudinally. Linear regression was used to determine significance in this study. It is possible that prediction may not be uniform over the concentration of the biomarker, but may be significant over a threshold value. However, no data to this effect were presented. In addition, a previous study identified the predictive value of the ratio of RANKL to OPG for erosive disease in RA, raising the possibility that combinations of biomarkers, if based in biologic plausibility, could be more predictive than single biomarkers.