Aim: To describe 1) The use of TNF blockers in early arthritis in daily clinical practice in France 2) To evaluate symptomatic, structural efficacy, and retention rate over 5 years 3) To evaluate predictive factors for TNF blocker response Type of study: Observational cohort study using cross-section and longitudinal data.

Description of the project methodology

• Patients: All patients in the ESPOIR cohort (multicentre French cohort study of early RA).A sub-analysis will be conducted among patients satisfying the ACREULAR 2010 criteria.

• Data collected: Patient characteristics, Clinical data regarding RA and related pathologies, Characteristics of treatments received The analysis will be conducted using data collected at baseline, 6, 12, 18, 24, 36, 48, 60 months.

• Analyses:

  1. Frequency of use of TNF blockers: the % of patients initiating TNF blockers will be calculated (Kaplan-Meier method), and the type of TNF blocker will be described, the route of administration, the dosage, the association with other DMARDs and the place of the TNF blockers in the treatment strategy during the first 5 years.
  2. Implementation of EULAR recommendations: the percentage of patients that initiate TNF blockers meeting the EULAR criteria for initiation will be estimated, and the concordance coefficient Kappa with regard to such fulfilment and the initiation of TNF blockers will be calculated, and disease severity outcome measures will be compared depending on the fulfilment or not.
  3. Identification of potential predictive factors for initiation of TNF blockers: a survival curve (Kaplan-Meier) will be performed. The baseline characteristics of the patients with regard to the initiation of TNF blocker during the first 5 years of the disease will be compared by univariate analysis and Log-rank test will be performed in all variables. A stepwise multivariate analysis (Cox analysis) will be performed.
  4. Therapeutical effect:the retention rate over time will be calculated, the changes in different variables will be compared in the group of patients who have received TNF blockers matched (using a propensity score) to 1,2 or 3 patients who have not. The DAS28 and HAQ will be assessed and compared at the short term (after at least 8 weeks of treatment) and long term (last available visit) in groups. The structural efficacy was evaluated by the radiographic progression at last available visit.The drug effect will be identically estimated depending on the TNF blocker used, by calculating the retention rate and comparing DAS28 at short term and long term.

  6) Identification of predictive factors for TNF blocker response: To evaluate the impact of baseline demographics and disease conditions on the DAS28 and HAQ response during the first 5 years will be compared by univariate and multivariate analysis.

Expected results:

Increase knowledge on the optimal use of TNF blocker and on predictive factors for TNF blocker response in early RA patients.

Inclusion Criteria:

• age 18 to 70 years
• more than 2 swollen joints for >6 weeks and <6 months
• suspected or confirmed diagnosis of RA
• no previous intake of DMARDs or steroids (except if <2 weeks).

Exclusion Criteria:

• Patients were excluded if the referring physician judged they had other clearly defined inflammatory rheumatic diseases.

- Patients: All patients in the ESPOIR cohort (multicentre French cohort study of early RA).A sub-analysis will be conducted among patients satisfying the ACREULAR 2010 criteria.