From 1999–2003, all early rheumatoid arthritis patients (disease duration <1 year) who fulfilled the 1987 revised American College of Rheumatology (ACR) criteria for rheumatoid arthritis15
were asked to participate in this two‐year randomised, open‐label prospective multicentre strategy trial. Patients visited the outpatient clinic of one of the six rheumatology departments in the region of Utrecht, the Netherlands, collaborating in the Utrecht Rheumatoid Arthritis Cohort study group. Inclusion criteria were: symptoms <1 year and age >16 years. Exclusion criteria were: previous use of glucocorticoids or any DMARD, use of cytotoxic or immunosuppressive drugs within a period of three months before inclusion, alcohol abuse (>2 units per day), and psychological problems which would make adherence to the study protocol impossible. At baseline all patients were monitored for medical conditions that could interfere with MTX usage. This screening included a chest x
ray, liver enzymes (AST and ALT), albumin, hepatitis serology, serum creatinine and complete blood count. An independent person performed randomisation in blocks of nine per hospital. The medical ethics committees of all participating hospitals approved this study, and all patients gave written informed consent before entering the study.
A sample size of 100 patients per group was needed based on a difference in functional disability score of 0.25 (SD 0.7), with a power of 0.80 (β
0.2) and alpha of 0.05. It is estimated that in daily practice during the first two years of treatment approximately 30% of patients need to discontinue MTX because of inefficacy or adverse events.16
We therefore included 150 patients in each strategy group.
Patients were randomly allocated to one of two strategy groups: the conventional strategy group or the intensive strategy group. The starting dose of oral MTX was 7.5 mg/week. In both groups, the dosage of MTX was not changed if patients had responded compared with the previous visit (fig 1); otherwise the dosage was increased stepwise by 5 mg/week, to a maximum of 30 mg/week. If the maximum (tolerable) dose of MTX was reached and patients did not fulfil the criteria for sustained response (fig 1), MTX was administered subcutaneously (sc). For patients on MTX sc having an inadequate response, cyclosporine was added to the MTX, while the dosage of MTX was reduced to 15 mg/week (fig 1). The starting dose of cyclosporine was 2.5 mg/kg/day; this was increased stepwise by 0.5 mg/kg/day to a maximum of 4 mg/kg/day, if no response was reached. If patients fulfilled the criteria for sustained response, MTX was reduced stepwise by 2.5 mg/week as long as patients met these criteria; otherwise the dose of MTX was continued or increased again according to protocol.
Figure 1Protocol and response criteria for the intensive and conventional strategy group separately. The sustained response criteria had to be fulfilled for six months (three subsequent visits) in the conventional strategy group and for 12 weeks (more ...)
Folic acid was prescribed to every patient (0.5 mg/day). Use of non‐steroidal anti‐inflammatory drugs (NSAIDs) was allowed in both strategy groups. Intra‐articular injections were avoided in so far as possible because this might lead to bias with respect to treatment effect between the two treatment groups. Oral glucocorticoids were not allowed during the trial unless unavoidable, which then had to be approved by another rheumatologist participating in this study.
Conventional strategy group
Patients in this strategy group visited the outpatient clinic once every three months and were treated according to protocol. This strategy, which comprised dose adjustments based on the opinion of the individual rheumatologist at three‐monthly visits, was similar to common practice in the Netherlands in 1998 when this study was designed. The minimum time to reach the highest dose of 30 mg/week MTX was 52 weeks.
Intensive strategy group
The strategy approach in the intensive strategy group differed in three ways from that of the conventional strategy group, namely:
1. The use of a computer decision program (designed by JWG Jacobs). At each visit, data on swollen joint count, tender joint count, erythrocyte sedimentation rate (ESR), and visual analogue scale (VAS) for general well‐being were entered by the rheumatologist. The program then calculated whether or not predefined criteria of response to treatment were met. As ESR values were only known the next day, the participating rheumatologists informed their patient the following day by telephone whether a dose change was necessary or not.
2. The response criteria (fig 1).
3. The frequency of evaluations leading to therapeutic decisions; fast step‐up and fast step‐down of MTX dosage.
Patients in the intensive strategy group came to the outpatient clinic once every four weeks and the maximum dose of 30 mg/week MTX could be reached after 18 weeks.
At baseline and at the predefined assessment points, depending on the strategy group, the following clinical variables were assessed: ESR (mm/h1st
), number of swollen joints (0–38), number of tender joints (0–38), VAS for pain (mean score of VAS pain at night and VAS pain in the morning, 0–100 mm
most pain), VAS general well‐being (0–100 mm
worst score), and morning stiffness (0–180 min). The Dutch version of the Health Assessment Questionnaire (HAQ,17
) was filled out every three months (0–3
most functional disability). Rheumatoid factor status was defined to be positive or negative according to the latex fixation test or the Waaler‐Rose test (positive >20 U/l) at baseline. Radiographs of hands and feet were taken at inclusion, at one, and at two years. Radiographs were independently scored in chronological order according to the modified Sharp/van der Heijde method (range 0–44818
) by two investigators who were blinded for the treatment strategy.
Clinical efficacy and radiographic progression
The primary major outcome of this study was the number of patients in remission for at least three months at any time during the two year trial. Remission was defined as: no swollen joints, and at least two out of three of the following criteria: number of tender joints
and VAS general well‐being
20 mm; for both groups these criteria data were analysed at the same three month intervals.
Second major outcome was the area under the curve (AUC) standardised to time calculated for all clinical variables.19,20
The mean score of the previous and the next score imputed missing data between two visits. For 21% of the patients, functional disability scores were missing at baseline and these score were imputed by the mean baseline score of the strategy group.
Additionally, we calculated the mean change in disease activity at one and at two years and the individual patient improvement from baseline according to modified ACR50 criteria, because we did not measure VAS physician's global assessment. These latter criteria were met if patients had a 50% improvement in both number of swollen and tender joints and a 50% improvement in two out of four of the following variables: ESR, VAS pain, VAS general well‐being, and functional disability.
Adverse events and medication‐related toxicity
In both strategy groups, all adverse events were evaluated and recorded at every visit according to a predefined protocol. Liver toxicity was defined as an increase of transaminase enzymes (AST, ALT) above the upper limit of normal. Haematological abnormalities were defined as: anaemia (Hb<6.5 mmol/l), leucopenia (<3.5*109/l), thrombopenia (<150*109/l), pancytopenia (two out of three of these criteria), and other. Post‐dosing reactions were defined as reactions which occurred within 24 h after MTX intake (for example, arthralgia). Criteria for dose adjustments or discontinuation of MTX use because of adverse events were dictated by the study protocol.
Intention‐to‐treat analyses were performed, in which the last available data were carried forward. This was not done for radiographic damage because the scores at two years would then be underestimated. Median AUC for all clinical variables and median radiographic progression rate were compared between the two strategy groups using the Mann‐Whitney U test. Other continuous variables were tested using the independent t test in the case of normal distribution and Mann‐Whitney U test in the case of non‐normal distribution. For both remission rates and individual patient improvement rates, patients who had to withdraw in the first year of the study were designated as non‐responder in the first year. Patients who withdrew during the second year, but fulfilled the ACR50 criteria or the remission criteria during the first year, were considered as responders in the first year. For the two‐year analyses, all patients who withdrew from the study were considered as non‐responders. The χ2 test was used to test differences in remission and responder proportions between the two strategy groups.
For all analyses, the level of significance was set at p<0.05 and tested two‐sided using SPSS 12.0.