Rheumatoid arthritis (RA) is a chronic, inflammatory disease of unknown aetiology. Medical treatment aims at controlling synovitis to prevent joint destruction and subsequent loss of function.1

The optimum treatment strategy for recent onset rheumatoid arthritis has not yet been established. Early treatment with disease-modifying antirheumatic drugs (DMARDs) aiming at tight disease control improves the long-term outcome.2^–5 Lately, studies of tumour necrosis factor (TNF)α inhibitors have shown convincing clinical and radiographic disease control in early RA.6^–8 One question that arises, therefore, is whether similar treatment goals can be reached with conventional DMARD treatment.

The CIMESTRA (Ciclosporine, Methotrexate, Steroid in RA) study is a randomised double-blind study in which two treatment regimens aiming at prompt suppression of joint inflammation are compared in patients with untreated rheumatoid arthritis of duration less than 6 months.9 Early, intensive step-up treatment with methotrexate and intra-articular betamethasone produced rapid and sustained disease control at 12 months follow-up. Addition of ciclosporine improved the clinical response further and reduced the need for intra-articular glucocorticosteroids. About 40% of the patients were in Disease Activity Score (DAS) remission after 1 year, and erosive progression was halted in both treatment arms.

In this paper we report the results of the second year double-blind extension of the study. During the second year, methotrexate and intra-articular glucocorticoid injections were continued, while hydroxychlorochine was added in all patients and ciclosporine was tapered to zero. The aim was to investigate whether the clinical and radiographic results after 1 year could be maintained at 2 years.

PATIENTS AND METHODS

Study design

The study design and results of the first year of this investigator-initiated, multi-centre, randomised, double-blind, parallel-group, placebo-controlled trial that included 160 consecutive patients with early active RA have been published previously and are only briefly outlined below.9 During the second year of the study, double-blinding was maintained.

Inclusion and exclusion criteria

At study entry, all participants fulfilled the American College of Rheumatology (ACR) 1987 revised criteria for RA,10 had active disease of duration <6 months with at least two swollen joints, and were aged 18–75 years. Baseline characteristics of the study population are presented in table 1.

Treatment strategy

The CIMESTRA trial consisted of two treatment arms: (1) methotrexate plus ciclosporine (combination therapy group), and (2) methotrexate plus placebo-ciclosporine (monotherapy group). If patients presented with swollen joints during the first year, they were treated with intra-articular betamethasone injections (7 mg/litre, maximum four joints or 4 ml per visit), and the methotrexate dose was escalated by 2.5 mg/week every 4 weeks from 7.5 mg/week up to maximum 20 mg/week, followed by stepwise increase (0.5 mg/kg) in ciclosporine/placebo-ciclosporine every 4 weeks from 2.5 mg/kg up to maximum 4.0 mg/kg of body weight.9

During the second year, patients were seen in the outpatient clinic every 4 weeks. Ciclosporine/placebo-ciclosporine was tapered to zero by 0.5 mg/kg every 4 weeks from week 76, while methotrexate was continued. The long-term strategy was to withdraw methotrexate in patients who were in remission from year 3 and onwards. Hydroxychlorochine 200 mg/day was added in all patients at week 68, irrespective of disease activity, because a previous study had shown that after withdrawal of methotrexate, hydroxychlorochine extended the response seen with combination therapy.11 If swollen joints were present at any visit, intra-articular injections of betamethasone were given and methotrexate dose was escalated. Patients who did not achieve an ACR20 treatment response (20% improvement in symptoms) despite 20 mg methotrexate per week were switched to parenteral methotrexate for 3 months, followed by triple therapy (methotrexate, sulphasalazine and hydroxychlorochine) for 3 months and after this switched to TNF inhibitor treatment if disease activity persisted.

Joints were evaluated and injections given by independent, blinded, and trained assessors. All patients received folic acid (1.2 mg daily, 6 days per week) as well as calcium and vitamin D supplementation. Patients with a Z score of <0 in the femoral neck or lumbar spine at study entry received alendronate 10 mg/day. Mild analgesics were given on demand. Oral glucocorticoids were not allowed at any time.

Outcome measures

The primary efficacy end point was the ACR20 response at 2 years. Secondary end points included remission, cumulative dose of betamethasone, and radiographic progression. Remission was defined according to the ACR criteria for remission in RA12 and needed to be present at both week 100 and week 104 to be acknowledged. Remission was also defined according to the DAS in 28 joints (DAS28 <2.6).13 Disability was assessed with the Health Assessment Questionnaire (HAQ) at each visit.14

Conventional radiographs of the hands (posteroanterior and Nørgaard15 projections), wrists (posteroanterior and lateral projections), and forefeet (anteroposterior view) were obtained at baseline and at 1 and 2 years. The radiographs were scored by an independent senior musculoskeletal radiologist (AaV), blinded to treatment group assignment but not to the chronologic order of the images, using the Sharp–van der Heijde scoring method.16 The primary radiographic end point was change in total Sharp–van der Heijde score (TSS) from baseline.17 The estimated yearly rate of progression in the TSS was calculated according to the duration of disease and the baseline TSS for each patient.18

Treatment adjustments for adverse events

The dosage of ciclosporine/placebo-ciclosporine was reduced if the serum creatinine increased by >30% compared with baseline and was withdrawn in the case of a persistent increase. Hypertension (blood pressure >140/90 mmHg) was treated with amlodipine 5–10 mg/day and if necessary with metoprolol 50–200 mg/day, and the ciclosporine/placebo-ciclosporine dosage was reduced until the blood pressure was normalised. Side effects of methotrexate were handled in accordance with local guidelines. In the case of severe toxicity, the patient was excluded from the study and the study medication withdrawn.

Ethical considerations

All patients gave a renewed written informed consent to participate in the second year of the trial. The year 2 protocol was approved by the national health authorities and ethics committees in all five participating counties (ethics committee reference no. M-1959-98). The trial was performed in accordance with the Declaration of Helsinki and the International Conference on Harmonisation 1996 revised Guidelines for Good Clinical Practice in the European Community. The trial was registered at http://www.clinicaltrials.gov (reference no. NCT00209859).

Statistical analysis

A total of 57 patients in each arm completing year 2 gave 69% power to detect a 20% difference in the ACR20 response rate between the two arms at α = 0.05 (two-sided test). This was calculated on an expected response rate of 70% in the monotherapy group.19

Comparisons between groups were made with the Fisher exact test for dichotomous responses and the Mann–Whitney U test for non-dichotomous responses. Changes over time were analysed with the McNemar test for dichotomous responses and the Wilcoxon signed rank sum test for non-dichotomous responses. Data are reported as the mean (SD) for variables in which normal distribution was found; otherwise the data are reported as the median (interquartile range (IQR)).

Analysis was by intent-to-treat. We used the last observation carried forward approach for missing data. Intent-to-treat analysis without last observation carried forward and a completers’ analysis were also performed and gave similar results (data not shown). The R software package20 was used for the statistical analysis, which was performed by an independent statistician.

RESULTS

Patient characteristics

A total of 80 patients were included in each of the two therapy groups at baseline (fig 1). A total of 7 patients in the monotherapy group did not sign the informed consent (hypertension: n = 1; lack of efficacy: n = 3; patient request: n = 3); in the combination therapy group it was 13 (hypertension: n = 4; increased serum creatinine: n = 1; concomitant medical diseases: n = 3; moved: n = 1, patient request: n = 4). The repeated intra-articular injections were in no case the reason for withdrawal from the study. A total of 61 patients in the combination group and 56 patients in the monotherapy group entered the second year of the study, and a total of 104 completed year 2 (fig 1). At baseline, a high proportion had risk factors for erosive disease: IgM-rheumatoid factor (RF) positive: 65%; anti-cyclic citrullinated peptide (CCP) positive: 59%.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Flow chart showing the distribution of the study patients from initial contact to completion of the study. HT, hypertension; i.a., intra-articular; ITT, intent to treat; LOE, lack of efficacy; RA, rheumatoid arthritis; SAE, serious adverse event.

The median dosage of methotrexate at 2 years was 17.5 mg/week (IQR 12.5–20) in both treatment groups (table 1). In accordance with the protocol, ciclosporine/placebo-ciclosporine had been withdrawn at week 104 in all patients. The cumulated dose of betamethasone during year 2 was 1.5 ml (IQR 0–5.125) in the combination therapy group compared to 2 ml (0–7.25) in the monotherapy group (p = 0.18), which was significantly less than during year 1, where it was 9 ml (6.5–16.5) and 12 ml (7.5–18), respectively; p<0.001 for difference between years in the two arms.

Radiographs were obtained at baseline in 155 patients (77 receiving combination therapy and 78 receiving monotherapy), and in 160 patients at 1 and 2 years. Some of the completers were on escape medication of parenteral methotrexate (four in the combination group, five in the monotherapy group), triple therapy (two in the combination and two in the monotherapy group), and none received TNF inhibitor treatment. At 2 years, a total of 20 non-completers (n = 11 withdrawn during year 1 (n = 5 in the combination and n = 6 in the monotherapy group) and n = 9 during year 2 (n = 6 in the combination and n = 3 in the monotherapy group)) had initiated biological treatment after exclusion from the study.

Clinical efficacy

The proportion of patients achieving an ACR20 response at 2 years was higher in the combination group (88% of patients) than in the monotherapy group (73%); p = 0.04 (fig 2). Similarly, the proportion of patients achieving ACR50 response was higher for the combination therapy group (79%) than for the monotherapy group (62%) p = 0.03, whereas there was no significant difference for ACR70 response (fig 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Outcome measures. American College of Rheumatology (ACR)20, ACR50, and ACR70 response rates during the 2-year trial. Dark columns: At 1 year (52 weeks), Light columns: At 2 years (104 weeks). Left columns: combination therapy group (methotrexate and ciclosporine); right columns: monotherapy group (methotrexate and placebo-ciclosporine). *p<0.05, between combination and monotherapy groups (Mann–Whitney U test) or within groups between time points (Wilcoxon signed rank sum test).

Significantly more patients in the combination therapy group achieved ACR50 after 2 years than after 1 year (p = 0.04, McNemar), and in the monotherapy group significantly more patients achieved ACR50 (p<0.05), ACR70 (p = 0.03) and DAS remission (p = 0.02).

The proportion of patients who were in remission according to the ACR criteria at both week 100 and week 104 (including the absence of glucocorticoid injections after week 96) in the combination therapy and monotherapy groups was 41% and 35%, respectively (not significant between groups). The achievement of DAS remission at 2 years was 51% and 50% in the two groups, respectively. A total of 16% in each treatment group were in DAS remission at all visits during year 2. A total of 15% in the combination therapy group and 10% in the monotherapy group were in ACR remission throughout year 2 (not significant, Fisher exact test).

The reductions in numbers of swollen and tender joints, visual analogue scale (VAS) scores, HAQ score and serum C-reactive protein (CRP), which were observed during year 1, were sustained during year 2 (table 1). At 2 years, 60 patients in the combination group and 58 patients in the monotherapy group had no swollen joints (40 joints count) (p = 0.86 between groups), 48 and 42 had HAQ scores of ⩽0.25 (p = 0.43), and 44 and 38 had pain scores of ⩽10 mm on a 100 mm VAS (p = 0.43). In all, 34 patients in the combination group and 27 in the monotherapy group had no swollen joints at any visit during year 2 (not significant).

Radiographic evaluation of joint damage

Sharp–van der Heijde scores at baseline, 1 year and 2 years together with estimated and observed progression rates are shown in table 2. The increase in TSS from baseline to year 2 was mean 1.42 (median: 0; range 0–20) in the combination group vs 2.03 (0; −1 to 43) in the monotherapy group (p = 0.49, Mann–Whitney U test). There were no differences in radiographic scores between the treatment groups either at baseline or after 2 years. Figure 3 shows the changes in radiographical scores during the 2 years of the trial. At baseline, 61% of the patients had erosive disease (TSS>0), while corresponding values after 1 and 2 years were 64% and 66%.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3 Estimated and observed radiographic changes, scored according to the Sharp–van der Heijde method, in the two treatment groups at 24, 52 and 104 weeks. A. ΔTotal Sharp score. B. ΔErosion score. C. ΔJoint space narrowing score. D. Probability plot showing the change in total Sharp–van der Heijde score from baseline to week 104.

The estimated yearly rates of progression in TSS were mean 22 (median 5; IQR 0–32) for the combination therapy group; and mean 16 (median 11; IQR 0–21) for the monotherapy group. The estimated yearly progression in erosion scores (ES) was 12 (7; 0–20) for the combination group and 10 (0; 0–15) for the monotherapy group. For the joint space narrowing scores (JSN) it was 10 (0; 0–9) for the combination group and 6 (0; 0–4) for the monotherapy group.

Adverse events

Serious adverse events leading to study withdrawal during the second year of the study were seen in three patients in the combination therapy group (malignancy: two (B cell lymphoma and bladder cancer); infections: one (suspected, but not confirmed infectious arthritis); other: zero). There were no serious adverse events in the monotherapy group during year 2. No deaths occurred. Adverse events during the second year are shown in table 3.

The median increase in serum creatinine level at 2 years was 7% (IQR 0 to 16%) in the combination therapy group (this was unchanged from year 1) and 4% (−3 to 8%) in the monotherapy group (2% increase per year); p = 0.006 between groups. Serum creatinine levels increased more than 30% over baseline levels in 11 patients in the combination therapy group and 3 patients in the monotherapy group (range 33–63% and 39–46%, respectively). In four of these patients (all receiving combination therapy), ciclosporine was withdrawn because of persistently high serum creatinine levels (33–46% increase), after which the serum creatinine levels decreased to normal.

Table 4 shows the use of antihypertensive treatment in the two treatment groups at time of inclusion, at 1 year and at 2 years.

DISCUSSION

The target of the CIMESTRA study, to suppress synovitis immediately after diagnosis, was established in patients with active rheumatoid arthritis by an aggressive DMARD step-up treatment combined with intra-articular betamethasone. This treatment strategy resulted in rapid and sustained disease control over 2 years without significant drug toxicity. The ACR20 and ACR50 responses were persistently 15% higher in the methotrexate–ciclosporine group than in the methotrexate only group and increased during the second year of the study. Thus, by the end of the study, close to 90% in the combination and 75% in the monotherapy groups had achieved an ACR20 response. In both treatment groups, 50% were in DAS remission, and the rate of radiographic progression was suppressed to a minimum.

Several factors may have contributed to the favourable outcomes of the CIMESTRA study. The aggressive intervention was initiated at a very early stage of the disease with a median disease duration of 3 months. T cells are thought to play a prominent role at the onset of RA.21 This concept is supported by the superior clinical response in the ciclosporin-methotrexate group. However, although the addition of ciclosporine reduced signs and symptoms of inflammation, the remission rate and erosive progression did not differ between treatment groups.

Intra-articular injections of betamethasone yielded prompt and long-lasting clinical response even at low doses and were well tolerated by the patients. In the present study, the cumulated dose of betamethasone was low during the first year, and even lower during the second year. A low dose of oral prednisolone has been demonstrated to retard radiographic progression in some,22 23 but not all24 studies of early rheumatoid arthritis. These studies support the suggestion that injections with intra-articular betametasone early in the disease not only give symptomatic relief, but in addition may also contribute to the retardation of bone erosions observed in the present study. Furthermore, the study set-up aiming at tight disease control with frequent outpatient visits may in itself have favoured a good outcome, as demonstrated in the TICORA study.2 Thus, targeted intra-articular steroid with frequent follow-up seems to be the most effective strategy along with disease-modifying therapy. Many of the side effects of high dose systemic corticosteroids are avoided and patients derive benefit in their most affected joints.

In the FIN-RACo study, triple therapy (methotrexate, sulphasalazine and hydroxychlorochine) was compared to monotherapy.19 The ACR50 response rates, and the proportion of patients who achieved ACR remission in the CIMESTRA combination treatment group were equal to those reported with triple therapy in the FIN-RACo study. By contrast, remission was more frequent in the CIMESTRA monotherapy group (35% vs 18%),19 which may be explained by the preferential use of sulphasalazine as monotherapy in the FIN-RACo study. In addition, the consistent use of intra-articular injections of betamethasone may have contributed to the high response rate in the CIMESTRA monotherapy group. In a 2-year study of triple therapy,25 the ACR20, ACR50 and ACR70 response rates in patients with advanced rheumatoid arthritis, who had not previously received methotrexate, were 83%, 67% and 33%. This was somewhat lower than in the present study and stresses the importance of early intervention. In the BeSt study, different treatment regimens in early rheumatoid arthritis were compared.4 With intensive monitoring and treatment adjustments, approximately 80% of the patients achieved an ACR20 response after 2 years in the groups treated with conventional DMARD therapy and in the group treated with the TNF inhibitor infliximab. This supports the hypothesis that excellent clinical disease control may be achieved in early rheumatoid arthritis with intensive use of conventional DMARDs.

The mean increase in total Sharp–van der Heijde score in the CIMESTRA study was 1.6 points after 2 years in both treatment groups. Based on recommendations in the literature,26 the x rays in the present study were read in sequence, which increases sensitivity to change. The radiographic progression rate in the CIMESTRA study was similar to what has been reported in studies of TNF inhibitors in early rheumatoid arthritis.6 7 27 The patients in the latter had more severe disease than in the present study, but by contrast, the x ray readers were blinded with regard to the time point at which the images had been obtained, which may bias the results downward.26 The progression of radiographic erosions in the CIMESTRA study were significantly less than reported in the FinRACo and COBRA studies.19 28 The COBRA study compared an intensive, short-term combination therapy with sulphasalazine monotherapy, and data at 5 years showed a persistent reduction in the rate of radiographic progression in the combination therapy group.28 In all, 12% of the COBRA patients had no radiographic progression during the 4–5 year follow-up period. In the CIMESTRA study about 70% of the patients had an unchanged total Sharp score after 2 years. In the BeSt study, the radiographic progression rate was similar to what was found in the CIMESTRA study, and radiographic progression was suppressed equally well with conventional DMARD therapy as with TNF inhibition.4 The results lend support to the view that pharmacologic resetting of the radiographic progression may be accomplished with intensive DMARD and corticosteroid therapy at an early stage of the disease.28

Hydroxychlorochine at 200 mg daily was added for all patients at week 68. This may have contributed to the improved clinical response between year 1 and year 2. More patients in the methotrexate–ciclosporine group either declined to enter year 2 or dropped out, which may reflect lower patient compliance. The percentage of patients who completed the 2-year trial was 70% and 60% in the monotherapy and combination therapy groups, respectively, which is equivalent to what was found in a 2-year study of TNF inhibitor therapy in early rheumatoid arthritis.7

Some aspects of the design of the present study should be noted. Because of the double-blind design, both treatment arms received intra-articular betamethasone, which makes us unable to assess its relative significance in the CIMESTRA regimen. Ciclosporine has been abandoned by some rheumatologists with reference to its toxicity. The concern of clinicians is often that longer-term therapy might lead to renal deterioration. For this reason the undoubted beneficial effects of a combination of methotrexate and ciclosporin have not been widely embraced by the rheumatological community. The present study demonstrates that if patients are closely monitored with regard to serum creatinine and blood pressure, there was no evidence of sustained side effects from ciclosporine. The disease duration in the present study was short, and in clinical practice patients often have longer duration than 3 months before they are referred to a rheumatologist. Randomisation was successful, except that significantly more patients in the combination group were anti-CCP positive, which is associated with a more aggressive disease course. This may have contributed to the lack of radiographic benefit from additional ciclosporine.

In conclusion, aggressive conventional therapy with intra-articular betamethasone and methotrexate effectively and safely maintained disease control and minimised radiographic progression for 2 years in patients with early, aggressive rheumatoid arthritis. Addition of ciclosporine during the first year improved ACR20 and ACR50 further after 2 years as compared with 1 year, but did not have any additional effect on the remission rate and radiographic outcome.