Article Text

Download PDFPDF

Optimal use of methotrexate: the advantages of tight control
  1. J W J Bijlsma1,
  2. M E Weinblatt2
  1. 1
    Department of Rheumatology & Clinical Immunology, UMC Utrecht, Utrecht, the Netherlands
  2. 2
    Department of Rheumatology and Immunology, Brigham & Women’s Hospital, Boston, MA, USA
  1. Professor J W J Bijlsma, UMC Utrecht, PO Box 85500, 3508 GA Utrecht, the Netherlands; j.w.j.bijlsma{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

The CAMERA study shows that using methotrexate in a tight control setting might lead to considerable improvement in disease activity in early rheumatoid arthritisSee linked article p 1443

The outlook for patients who are in 2007 newly diagnosed with rheumatoid arthritis has improved dramatically.1 Important elements in this improvement are: (1) the concept of a “window of opportunity”: treat early and treat effectively;2 (2) early use of disease modifying anti-rheumatic drugs (DMARDs);3 (3) combination treatment, including glucocorticoids;4 5 (4) the use of biologicals;6 and (5) optimising therapy to the individual patient.

The CAMERA study reported in this issue by the Utrecht group See pages 1443 deals with this last item: optimising therapy to the individual patient.7 In this randomised controlled trial the drug treatment given to patients with early rheumatoid arthritis was the same in both groups, but the frequency and tightness of control were different. This report has some important messages that merit further comment:

1. The goal of tight control is feasible with the anchor drug methotrexate in a much larger proportion of patients than previously thought.

2. Using the same drug, but seeing the patient more often, and thus being able to adjust that specific drug more often, is much more efficacious then seeing the patient less frequently.

3. In a busy practice, tight control, aided by a computer-assisted protocol, is feasible.


There is, as yet, no definition of “tight control” in rheumatoid arthritis. Elements of tight control are:

  • seeing the patient frequently, to enable frequent adjustments to the therapy;

  • improvement is not enough: the aim is low disease activity, or preferably even remission;

  • being as objective as possible in the judgment of disease activity; not the doctor’s or patient’s impression, but an objective measurement of different aspects of disease activity, such as joint scores, visual analogue scale for well being, acute phase reactant in the blood;

  • being rigorous in adjusting therapy—for example, a DAS of 1.7 is not 1.6; do not wait, but increase treatment according to protocol (a computer program may help in this respect);

  • seeing the patient frequently until disease activity is under control.


As a definition of tight control we might therefore suggest that it is a treatment strategy for the individual rheumatoid arthritis patient, aiming for a predefined level of low disease activity or remission, within a limited period of time (as soon as possible).8

Six years ago a remarkable observation was made in the Netherlands when the first TNF-α blocker became available. There were serious concerns at government level regarding the costs of this new treatment. Therefore the Dutch Society for Rheumatology developed guidelines for when to use TNF-α blockers: patients needed first to be treated with at least two DMARDs in adequate dosages, including methotrexate, in a dosage of 25 mg/week for at least three months before they could qualify for a TNF-α blocker. When estimates were made, it was thought that about 10% of the rheumatoid arthritis patients would need to be treated with TNF-α blockers; however, when methotrexate dosage was increased in order to qualify for this new treatment, many more patients responded to the higher methotrexate dosage. In addition, it also proved possible to increase the dosage of methotrexate quite rapidly when patients and doctors were anxious to reach that suggested higher dosage. In the first years only half of the expected number of rheumatoid arthritis patients were treated with TNF-α blockers, indicating that the potency of methotrexate was underused.

As reported in this issue, the CAMERA study7 is a good example of optimising methotrexate treatment in early rheumatoid arthritis. Patients with early (disease duration <1 year) rheumatoid arthritis were randomly assigned to the intensive strategy or to the conventional strategy group. Both groups received methotrexate, the aim of treatment being remission. The intensive strategy group was seen every month and treatment was tailored to the individual patient on the basis of predefined response criteria, using a computerised decision program. Patients in the conventional group came to the outpatient clinic once every three months and were treated according to common practice. By increasing the dosage of methotrexate in the intensive strategy group monthly, a dosage of 30 mg/week was reached within four months, leading to a much higher percentage of patients with low disease activity than previously reported. Remarkably, the tolerance of methotrexate at these higher dosages was comparable with that of lower dosages, apart from liver toxicity (23.2% vs 18.6%). In the study period of two years 50% of patients in the intensive strategy versus 37% of patients in the conventional strategy group achieved at least one period of remission, lasting at least three months. As expected, the difference was most pronounced in the first year (35% vs 14%).

The correct dose of methotrexate is a critical issue to maximise control of disease activity in rheumatoid arthritis. Even though dose ranging studies with methotrexate in rheumatoid arthritis are limited, there is sufficient evidence that the effective dose for most patients is around 10 mg/m2.

In randomised controlled studies in which the methotrexate dose was escalated to 20 mg/week, the effects on signs and symptoms of rheumatoid arthritis (as measured by the ACR 20 response criteria) were not statistically different between anti-TNF therapy as monotherapy (etanercept or adalimumab) and methotrexate 20 mg/week at 6–12 months.911


On both sides of the Atlantic discussions are ongoing about how to define remission of rheumatoid arthritis in the 21st century.12 13 Components of this definition are low disease activity, or even absence of disease activity, and no radiographic progression. There are of course good arguments to include no radiographic progression in an operational definition of remission. However it is far from practical to use a definition including radiographs for the day-to-day management of patients with early rheumatoid arthritis. But what about the use of disease activity scores in individual patients? For instance, the widely-used DAS is validated for groups but not for individual patients. When defining a patient in remission, the DAS has more face validity than the DAS28, just because the feet are not scored in the latter. Different components of the DAS have different influence on the total score, and in the lower range the influence of erythrocyte sedimentation rate is much greater than in the higher range. However, when evaluating a patient to decide whether treatment should be adapted, the patient’s own perception of pain can be a major issue, especially when a pain syndrome has developed. Patients can score a DAS >2, just based on their pain perception, even though the doctor treating them sees no disease activity. When this leads to an increase in treatment—for example, by starting a TNF-α blocker—over-treatment might happen. Other aspects, such as regression to the mean of the DAS, might also be relevant in interpreting changes in the DAS.14 Clearly, more work needs to be done to define the best way to evaluate individual patients’ remission or low disease state in such a way that individual treatment adaptations can be advised.

In the CAMERA study the reported cumulative dosage of methotrexate was comparable in both the conservative and intensive strategy group, which is remarkable. One explanation seems to be that because remission was quite often reached in the intensive strategy group, the dosage of methotrexate could be reduced in a substantial number of patients. It is also relevant to note that the cumulative use of NSAIDs was lower in the intensive strategy group. With regard to radiological progressions, the cumulative probability plot showed that in both strategy groups approximately 50% of patients did not progress over two years, but for the other 50% the progression rate was higher in the conventional strategy group than in the intensive strategy group. If the study protocol prescribed that the methotrexate dosage should not be decreased when remission was reached, perhaps this difference in radiographic progression would have been even more pronounced. This finding suggests that tight control is a good concept in bringing the disease under control, but that to maintain this control other aspects also need to be taken into account.

The CAMERA study clearly shows that using methotrexate in a tight control setting might lead to considerable improvement in disease activity in early rheumatoid arthritis. A programme to increase the dose of methotrexate to 20–25 mg/week in a more rapid manner will also lead to earlier identification of patients who are partial responders on methotrexate and might benefit from combination therapy. Tight control seems to be a promising new paradigm for reaching the aim of low disease activity—or, even better, remission in the first two years of treatment.


View Abstract


  • Competing interests: None declared.

Linked Articles