As biologic treatments are expensive and associated with some concerns regarding long-term safety, investigator hypothesize that early tapering and then withdrawal of biological agent, in an homogenous group of children with juvenile idiopathic arthritis achieving inactive disease, is safe and not inferior to the maintenance of stable treatment intensity over 24 weeks. In addition, investigator also hypothesize that an earlier tapering of treatment is associated with a better quality-of-life and a general cost saving effect. MRP8/14 will be studied as a potential biomarker for the risk of relapse. A study for biologic agent, anti-biologic agent antibodies and a pharmacogenomic approach will complete the research, as pharmacokinetic study during withdrawal of biologic treatment are rare in children.

Juvenile idiopathic arthritis (JIA) is characterized by chronic arthritis of unknown etiology starting before the age of 16. There are four to five thousand paediatric patients with JIA in France. Most of these patients are diagnosed with oligoarticular or rheumatoid factor negative polyarticular JIA. The prognosis of the disease has dramatically improved thanks to the introduction of biologic agents in patients with an extended oligoarticular or rheumatoid factor negative polyarticular JIA and inadequate response to methotrexate. Inactive disease and long-lasting clinical remission are achieved in most cases. "Treat to target" approaches are increasingly recommended, with earlier introduction of biologics, however the way to taper or withdraw treatment in patients achieving inactive disease is not codified. As biologic treatments are expensive and associated with some concerns regarding long-term safety, this study aim to test, in a randomized fashion, the hypothesis that early tapering of biologic agents (i.e. increasing the intervals between injections as soon as inactive disease is documented) is safe and non-inferior to the maintenance of stable treatment intensity over 24 weeks, and therefore test the possibility of early biologic agent withdrawal. It will also study concentrations of different biological agent, the occurrence of anti-drugs antibodies while tapering and then withdrawing biologics, and their possible association with a higher risk of relapse. In addition, investigators will test if the serum level of proteins 100 (MRP8/14) could be predictive of flares. Finally, pharmaco-economic analyses and quality of life studies will be conducted.

Inclusion Criteria:

• Patient aged 2 to 17 years and treated with etanercept or tocilizumab or adalimumab, or patient aged 6 to 17 years and treated with abatacept.
• Patient with an oligoarticular or polyarticular rheumatoid factor negative JIA
• Patient treated with biologic treatment for persistent arthritis according to the marketing authorization.
• Patient who achieved inactive disease within two years of treatment with the last biologic agent administered, according to Wallace criteria : no joints with active arthritis, no active uveitis (as defined by the SUN Working Group), ESR or CRP level within normal limits in the laboratory where tested (or, if elevated, not attributable to JIA), physician's global assessment of disease activity score (< 10/100 visual analogue scale), and duration of morning stiffness < ou = 15 minutes (within 7 days before the visit).
• Patient with inactive disease achieved for less than 12 months.
• Patient with stable doses of non-steroidal anti-inflammatory drugs, Methotrexate (maximum 20 mg/m2/week), and other non biologic DMARD for at least one month before inclusion
• Patient without steroids or joint injection or live vaccines injection for at least one month.
• Signed informed consent by both parents (or legal guardian) and patient's agreement.
• Patient affiliated to the National Health Assurance system.

Exclusion Criteria:

• Patient with systemic form, rheumatoid factor positive, psoriatic or associated with enthesitis related JIA.
• Patient undergoing biologic therapy due to JIA-associated uveitis or with active uveitis at time of randomization.
• Patient with any contraindication to continue ongoing biologic treatment, notably ongoing uncontrolled infection, suspicion or evidence of demyelinating disease of the central nervous system.
• Patient previously treated with the same biotherapy for which dose decreasing or biotherapy withdrawal was already tested in the past for inactive disease and then reintroduced.
• Pregnancy or absence of effective contraception (including abstinence) in a pubertal patient.
• Patient suffering from tuberculosis.
• Patient with moderate to severe cardiac failure (NYHA class III / IV).