The objective of the study is to identify and validate predictive markers of anakinra responsiveness in RA patients by a transcriptomic approach.

Patients with active RA (ACR criteria) were given subcutaneous anakinra (100mg/d) 3 associated with metotrexate. Anakinra efficacy was evaluated at week 12, using the EULAR response criteria.

A blood sample was collected just before the onset of anakinra treatment and total RNAs were extracted from the peripheral blood mononuclear cells. The [33P] radiolabeled mRNAs were hybridized (duplicate or triplicate) over a set of 10.000 human cDNA probes spotted at a high density on nylon membranes. Data were normalized and filtered to allow the comparison between RNA samples. Statistical analyses were performed with the R software and hierarchical clustering was performed with the Cluster and Tree View softwares.

Inclusion Criteria:

• Is age 18 years old or older Satisfies the 1987 American College of Rheumatology revised criteria for Rheumatoid arthritis Has active disease at the time of randomization as indicated by a DAS28> 5.1 Has a disease at least refractory to DMARDs whose methotrexate or leflunomide Is capable of understanding and signing an informed consent form Agrees to use a medically accepted form of contraception during the study

Exclusion Criteria:

• Is pregnant or breast-feeding or without contraception Has significant concurrent medical diseases including cancer or a history of cancer within 5 years of entering the study, uncompensated congestive heart failure, significant active infection or any underlying diseases that could predispose subjects to infections (whose tuberculosis) Has renal disease (creatinin clearance level < 30 ml/min) Has allergy to anakinra Has leukopenia (white blood cells < 1.5 x 109 /l

Patients who Satisfied the 1987 American College of Rheumatology revised criteria for Rheumatoid arthritis