Patients with rheumatoid arthritis are at risk of developing permanent joint damage and disability. This study hopes to identify the most effective way of using existing arthritis medication to minimise the chances of developing permanent disability. Patients will have their arthritis activity assessed using an ultrasound machine. If there is still evidence of active arthritis the participant's arthritis medication will be increased until the arthritis is in remission. The effectiveness of this approach will be compared to the traditional method of assessing arthritis using clinical examination.

Furthermore, it is extremely important to identify those patients most at risk of aggressive disease. The investigators hope to produce a more accurate measurement of disease prognosis by examining the relationship between a series of blood tests and how well controlled rheumatoid arthritis appears after 18 months of therapy. Some patients will also be asked to donate samples of joint fluid and joint lining for additional analysis.

PURPOSE 1 - to determine whether it is possible to achieve better control of inflammatory joint disease activity in early rheumatoid arthritis by using musculoskeletal ultrasound, instead of clinical examination, to identify the presence, or absence, of synovitis

NULL HYPOTHESIS 1 - using musculoskeletal ultrasound to confirm / refute the presence of ongoing synovitis will NOT allow better control of early rheumatoid arthritis nor prevent progression of destructive joint disease despite patients receiving more intensive disease modifying therapy regimens

PURPOSE 2 - to determine whether baseline measures of certain biochemical and pathological factors, associated with the development of inflammatory synovitis, are predictive of response to therapy in early rheumatoid arthritis and short term outcome measures of inflammatory joint disease activity, functional ability and quality of life

NULL HYPOTHESIS 2 - serial measures of biochemical and pathological factors, associated with the development of inflammatory synovitis, will NOT correlate with short term outcome measures of disease activity and therefore cannot be used to predict a patient's prognosis nor identify those at risk of progressive, destructive joint disease

TRIAL DESIGN - randomised, prospective single blinded trial of treatment strategy with a nested study correlating baseline measures and 18 month outcomes

Investigators will not be blinded to treatment group. Treatment decisions and escalation of therapy will be dictated by a standardised protocol. The sequence of therapy escalation will be identical for both groups. The groups will differ on the threshold needed to progress to the next treatment step

Assessors of disease activity, radiological and pathological outcomes will be blinded to treatment group and their findings will form the basis of each groups final outcome measures

TREATMENT PROTOCOL - the sequence of therapy escalation will be the same for each group. The groups differ by the 'trigger' required to progress to the next treatment step. Therapy will escalated in each group if the measured disease activity exceeds that groups threshold trigger. Changes in DMARD therapy doses and/or combinations take three months to reach maximum effect; therefore, at least a three month gap will be left between each treatment escalation

PRIMARY OUTCOME MEASURE

1. Magnetic Resonance Imaging of Dominant Wrist - baseline and 18 months. Images will be scored using the OMERACT RAMRIS(Rheumatoid Arthritis Magnetic Resonance Imaging Score) atlas. The change in each patient's synovitis and erosion scores will be pooled and compared for each intervention group
2. 44 joint Disease Activity Score - Mean change of DAS44 with time will represent the rate of response to treatment. Mean area under curve DAS44 will represent overall level of disease activity throughout the study period

SECONDARY OUTCOME MEASURES

1. European League Against Rheumatism Response Rates - A EULAR Good response is defined as a greater than 1.2 change in DAS44 and a final DAS44 less than 2.4. Disease remission is defined as DAS44 less than 1.6
2. Functional Measures - Health Assessment Questionnaire and EuroQoL-5D questionnaires at enrollment and then every 3 months.
3. Plain Xrays - plain xrays of hands, wrists and feet at baseline and 18 months. Change in Sharp score between baseline and 18 month films will be reported
4. Biomarker analysis - analyses will include specific genetic factors(genomic DNA), gene expression (RNA analysis), novel autoantibody assays, cytokine / emerging inflammatory protein profiling, lipid / lipoprotein based markers, metabolic assays and assessment of bone and cartilage turnover markers.

Samples will be collected at baseline, 3 months and 12 months (if not commenced on etanercept), immediately before commencing etanercept, 3 months and 6 months after commencing etanercept and 3 months after cessation. Final disease outcome measures for each patient will be correlated with baseline biomarker values to determine if any predictive relationships exist.

All values will be entered into a logistical regression analysis to try and create a statistical predictive model. Serial biomarker analyses will demonstrate how the different components of the pathogenetic process respond to the different stages of DMARD therapy.

Inclusion Criteria:

1. Patients attending Early RA clinics with newly diagnosed RA or anti-CCP +ve Undifferentiated Arthritis (UA)
2. Active disease (DAS28 > 3.2)
3. DMARD naïve or DMARD monotherapy for less than 6 weeks
4. Aged 18 or more

Exclusion Criteria:

1. Significant liver disease and/or abnormality of liver function tests
2. AST / ALT > x2 normal, Alkaline Phosphatase > x2.5 normal
3. Renal impairment - serum creatinine > 200 μmol/l, eGFR < 30
4. Cytopenias - white cell count < 4.0, haemoglobin < 10, platelet < 150
5. Pregnancy or planned pregnancy
6. Contraindication to MRI
7. Other co-morbid condition that in the opinion of the investigator would preclude the use of sequential or combination DMARD therapy