Intended for healthcare professionals


Treating rheumatoid arthritis

BMJ 2007; 335 doi: (Published 12 July 2007) Cite this as: BMJ 2007;335:56
  1. Paul Emery, professor and head of academic unit of musculoskeletal disease1,
  2. Tore K Kvien, professor and head of rheumatology department2
  1. 1Leeds University, Leeds LS7 4S
  2. 2Department of Rheumatology, Diakonhjemmet Hospital and Faculty of Medicine, University of Oslo, Norway
  1. P.Emery{at}

    Antitumour necrosis factor can produce remission if started early

    Rheumatoid arthritis is a chronic inflammatory disease affecting the synovial joints. It is characterised by persistent inflammation and destruction of bone and joints. It affects physical functioning and other dimensions of quality of life. Symptoms can be treated with analgesics and non-steroidal anti-inflammatory drugs, which block cyclo-oxygenase. The underlying disease process can be affected by drugs that block or reduce the concentration of cytokines, which are known as disease modifying antirheumatic drugs. These drugs are either conventional small molecules or biological molecules.

    The outcome of treatment for patients with rheumatoid arthritis has improved considerably during the past 20 years.1 People have speculated about the potential to reverse the disease for some time. Treatments that aim to induce remission have been called for since the start of the 1990s.2 Several approaches have been investigated, including step-up (adding one drug to another) and step-down (starting with a combination of drugs and withdrawing individual ones) regimens3 and giving three disease modifying antirheumatic drugs at the same time.4 All approaches have reduced inflammatory activity, delayed radiographic progression, and improved function and quality of life. However, only a few patients have achieved remission.5

    The benefits of cytokine blockers—for example, drugs that antagonise tumour necrosis factor (TNF)—have exceeded initial predictions. Blocking TNF was predicted to reduce symptoms, but to have little structural effect on bone and cartilage, as evidence from animal models indicated that interleukin-1 was a more important bone cytokine than TNF. Clinical studies have shown, however, that TNF blocking agents have long lasting benefit, particularly when combined with methotrexate, and have a major effect on bone damage.6 There is thus little doubt that the current most effective treatment for patients with rheumatoid arthritis is a combination of methotrexate and anti-TNF. But questions remain, such as whether all patients need this treatment to achieve an adequate response and whether it can be cost effective.

    The effects of TNF on remission have been assessed in two studies.7 8 The first found that 12 months of anti-TNF given early on in the disease process could produce remission, which was sustained one year after ceasing treatment.7 The second found that six months of TNF blockade showed only short term benefit.8

    Against this background, a single blinded pragmatic study with four treatment arms was started in the Netherlands, and the longer term data from this study have now been published.9 The first two arms tested the efficacy of conventional disease modifying antirheumatic drugs, with either a switching or step-up protocol, while the third and fourth arms compared anti-TNF plus methotrexate with combination treatment plus corticosteroids.

    A crucial feature of the study was that participants were all treated according to prespecified targets (treatment-to-target strategy)—initial drug regimens were adjusted if participants failed to reach low disease activity (disease activity score ≤2.4). By definition, all patients should have achieved the same end point, as they would have been switched to a biological treatment if they did not respond. The design allowed the investigators to analyse how frequently the strategy could succeed and to examine subsets of patients for the benefits of strategies such as switching or stepping down drugs.

    What were the main findings of this study? Firstly, by using an aggressive treatment-to-target strategy, remission can be achieved in about 40% of patients. Secondly, the trial arms that included anti-TNF reduced inflammation more rapidly, reduced long term disease activity more effectively, and had bigger structural benefits than those that did not. Thirdly, induction with anti-TNF allowed more patients to come off the biological treatment and remain in remission. Fourthly, the escalating monotherapy arms (which measured the likelihood of a good response to individual agents) showed that while around 30% of patients started on methotrexate responded well, subsequent responses to further conventional disease modifying antirheumatic drugs were rare. Thus, most patients who fail methotrexate monotherapy will benefit from a direct switch to anti-TNF plus methotrexate, rather than trying other conventional antirheumatic drug regimens.

    These long term data confirm the success of the treatment-to-target strategy, which has improved outcome in other pragmatic trials with conventional disease modifying antirheumatic drugs.10 Furthermore, anti-TNF plus methotrexate was the most effective and least toxic treatment.

    Another interesting message to come out of the study relates to prognosis. Patients entered into this study had established early rheumatoid arthritis and a high proportion had structural damage, indicating disease with a relatively poor prognosis. It is interesting that patients who declined entry into the study, and mostly had milder disease at baseline (and should therefore have had a better outcome), ended up doing much worse than the included patients (F Breedveld, personal communication, 2007). One lesson is that if patients with a poor prognosis could be identified early, then they would be the ones most likely to benefit from more aggressive approaches.

    Data suggest that patients with a poor prognosis can be identified by a combination of features including a high acute phase response, autoantibodies to rheumatoid factor or cyclic citrullinated peptides (or both), and evidence of radiological erosions at baseline. By using more sensitive imaging techniques, it might be possible to identify more people at high risk of a severe disease course.11

    Are we now in a position to recommend that patients with a poor prognosis should receive anti-TNF as first line treatment? If the cost of the drugs was minimal, the only relevant comparator would be combination regimens with corticosteroids and methotrexate as the anchoring drugs. Cost effectiveness analyses are under way and will influence how widespread such an approach is likely to be.

    The success of the treatment-to-target strategy supports its use in clinical practice whenever possible. The data also suggest that patients with rheumatoid arthritis can achieve remission, but for them to cease treatment and remain in remission they must be treated early on in the disease process.7 12 However, sufficient uncertainty exists to warrant further double blinded trials and analyses of their costs.


    • Competing interests: PE and TKK have provided expert advice and undertaken clinical trials for producers of biological agents.

    • Provenance and peer review: Commissioned; not externally peer reviewed.


    View Abstract