Background Adipokines have recently emerged as pro-inflammatory mediators involved in the pathophysiology of rheumatoid arthritis (RA).
Objectives To determine whether serum adipokine levels independently predicted early rapid radiographic disease progression (RRP) in early RA.
Methods ELISA was used to assess baseline serum levels of total adiponectin, leptin and visfatin/NAMPT in the 632 patients from the French ESPOIR cohort who met the American College of Rheumatology-European League Against Rheumatism criteria for RA. We tested the association of serum adipokine levels and RRP defined as a change ≥ 5 in total Sharp-van der Heijde Score between inclusion and 1 year (ΔSHS≥5) corresponding to the destruction of one small joint (1) in univariate analysis and also after adjusting for all pertinent confounders (age, sex, body-mass-index, insulin resistance log(HOMA-IR), C-reactive protein level (logCRP), DAS28, log(HAQ score), rheumatoid factor and anti-CCP antibodies status, steroid use and radiographic evidence of RA damage at inclusion. Sensitivity versus the false positive frequency (1-specificity) for predicting radiographic disease progression with adiponectin levels was analysed by a receiver-operated characteristic (ROC) curve. The optimal cut-off point of adipokines level to identify patients with RRP was calculated using Youden index.
Results In univariate analyse, there was a trend for association between serum level of adiponectin and RRP (odds ratio OR [95%confidence interval]=1.47 [0.98-2.20]; p=0.06). In multivariate model, serum adiponectin level was independently associated with ΔSHS≥5 (adjusted OR=2.0 [1.14-3.52], p=0.0165) (Table). Conversely, leptin level was associated with RRP only in univariate analyse (OR=0.73 [0.54-0.97]; p=0.03) and visfatin/NAMPT level displayed no association. Considering the ROC curve, the best serum adiponectin cut-off value level to predict ΔSHS ≥ 5 was 1.60 μg/mL (sensitivity=0.74 and specificity=0.73 and area under the ROC curve=0.76).
Conclusions Serum adiponectin level is a simple and useful biomarker predicting early RRP in early RA independently of RA-confounding factors and metabolic status. Using the threshold of 1.6 μg/mL, patients with early RA at increased risk of early RRP can be identified.
Vastesaeger N et al. Rheumatology(Oxford) 2009,48:1114-21
Disclosure of Interest None Declared