Step-down approach using either cyclosporin A or methotrexate as maintenance therapy in early rheumatoid arthritis

Authors


Abstract

Objective

To evaluate the feasibility and outcome of the step-down approach using either cyclosporin A (CSA) or methotrexate (MTX) as maintenance therapy following 6 months treatment with these 2 agents in combination in early, nonerosive rheumatoid arthritis (RA).

Methods

Fifty-seven patients younger than 65 years with early, nonerosive RA were first treated with CSA and MTX in combination for 6 months. They were then randomly stepped down to single-agent maintenance treatment for another 18 months. Safety, clinical efficacy, survival on treatment, and radiographic progression were evaluated.

Results

When being treated with combination therapy, 7 of the 57 patients (12.3%) withdrew because of adverse events. Of the remaining 50 patients, 42 (84.0%) were American College of Rheumatology (ACR) 20% responders, 30 (60.0%) were ACR 50% responders, and 23 (46.0%) were ACR 70% responders. At month 6, 22 patients were randomized to CSA and 27 to MTX. During this trial period, the treatment was discontinued by 16 patients taking CSA (mainly because of loss of efficacy) and by 4 taking MTX. At month 24, the probability (± SEM) of survival on treatment was 0.273 ± 0.09 for CSA and 0.852 ± 0.07 for MTX. Of the 6 CSA patients who completed the trial, 4 (66.7%) were ACR 20% responders, and 3 (50%) were both ACR 50% and ACR 70% responders. Of the 23 completers in the MTX arm, 21 (91.3%) were ACR 20% responders, 18 (78.3%) were ACR 50%, and 10 (43.5%) were ACR 70% responders. The treatment was not responsible for severe adverse events. Radiography showed a slow progression in the damage score and number of eroded joints in both treatment groups.

Conclusion

Stepping down to single agent maintenance therapy following 6 months of combination treatment with CSA and MTX in early RA was only successful with MTX. Because this treatment did not prevent some radiographic progression, other approaches (e.g., step-up approach) may be more appropriate in early RA.

INTRODUCTION

Because most patients with rheumatoid arthritis (RA) have a poor long-term outcome on traditional treatments (1, 2), therapy is rapidly becoming more aggressive. The current trend is to intervene early, before the development of irreversible joint damage, using drugs that are as much targeted to the patient as possible. There are clearly many unresolved questions concerning the most indicated therapy for any individual patient, but it is now universally recognized that traditional treatment with a single disease-modifying antirheumatic drug (DMARD) is often inadequate (3). The concurrent use of 2 or more DMARDs is a more aggressive and widely used therapeutic approach based on solid pathophysiologic and clinical grounds (4–7). Of the many DMARD combinations studied, the association of cyclosporin A (CSA) and methotrexate (MTX) seems to be one of the most promising (8–12).

In this study, patients younger than 65 years with early, active nonerosive RA were first treated with the CSA/MTX combination for 6 months, and then randomly switched to maintenance therapy with only 1 of the 2 drugs for an additional 18 months. The goals of the trial were to assess the feasibility and outcome of the so-called step-down approach (13) and to evaluate the safety of the combination in early RA.

PATIENTS AND METHODS

Study design. This open trial was divided into 2 consecutive stages: a first 6-month phase during which the patients were treated with CSA and MTX in combination, and a subsequent 18-month period during which they were stepped down to receive just 1 of the drugs as maintenance therapy. The first phase was open and uncontrolled; the second was carried out under randomized, single-blind conditions (the assessor was unaware of the therapy).

Patient selection. Patients with RA defined on the basis of the revised American College of Rheumatology (ACR; formerly, the American Rheumatism Association) criteria (14) were eligible for the study if they were aged 18–65 years and fulfilled the following inclusion criteria: disease duration between 6 months and 2 years and active disease. Active disease was defined by the presence of at least 3 of the following: 5 or more tender joints, 5 or more swollen joints, morning stiffness ≥45 minutes, erythrocyte sedimentation rate (ESR) ≥30 mm/hour, and no evidence of definite radiographic joint erosions in the hands or feet (stage I according to Steinbrocker's classification). The exclusion criteria were the presence of other active diseases; pregnancy; serum creatinine values ≥1.2 mg/dl or creatinine clearance values <70 ml/minute; proteinuria >0.5 g per 24 hours; increased serum potassium, uric acid, and transaminases; increased blood urea nitrogen; hemoglobin <9.0 g/dl; white blood cell count <3,000 cells/mm3; platelet count <100,000 cells/mm3; arterial hypertension not controlled by antihypertensive treatment; previous treatment with CSA, MTX, or immunosuppressive agents; treatment with corticosteroids at a dosage >10 mg/day; and other treatments or living habits (e.g., alcohol or drug abuse) contraindicating the use of CSA or MTX. Women of child-bearing age not using an adequate contraceptive method were also excluded. Before entering the study, the patients were fully informed and gave their written consent.

All of the patients completing the first 6-month stage of the trial were considered eligible for the second phase (months 6–24) and were allocated to 1 of the 2 groups according to a randomization list created at study entry.

Treatment regimens and monitoring. The starting dosages were CSA at 3 mg/kg/day divided into 2 daily administrations separated by an interval of 12 hours and MTX at 7.5–10 mg (for patients less than 60 kg in weight) or 12.5–15 mg (for patients more than 60 kg in weight) administered intramuscularly once per week. The dosages could be reduced in the case of unwanted effects. The CSA dose had to be reduced by 1 mg/kg/day if serum creatinine increased to 30% or more above baseline value, if serum potassium rose above the normal range, if arterial hypertension developed and persisted despite specific therapy, if serum transaminases increased to 100% or more above upper normal limit, or if whole blood trough CSA levels were higher than 200 ng/ml. The MTX dose had to be reduced to a minimum of 5 mg/week in the case of serum creatinine values above the upper normal limit, the abnormal presence of blood or proteins in the urine, serum transaminase values of 100% or more above the upper normal limit, hemoglobin values below 9 g/dl, a white blood cell count of <3,000 cells/mm3, or a platelet count of <100,000 cells/mm3. If these abnormalities had not resolved after 2 weeks on the lower dosage or with appropriate treatment, the responsible agent(s) had to be discontinued. If the number of tender and swollen joints and the patient evaluation of pain did not improve by at least 20% after 2 months of therapy, the dosage of 1 or both agents could be increased: CSA up to a maximum of 4 mg/kg/day and MTX to 20 mg/week. For ethical reasons, the patients who discontinued the study therapy could be started on another DMARD treatment as indicated by their clinical condition.

Corticosteroids were allowed and their dosage modified as required, provided the dosage did not exceed 10 mg/day. Local corticosteroid injections could not be given during the 2 weeks preceding a joint assessment. The nonsteroidal antiinflammatory drugs allowed were nimesulide, naproxen, indomethacin, and ketoprofen; their dosage could be reduced but not increased during the study period. Other medications could be given if required, preferably avoiding drugs capable of interfering with the study drugs.

Clinical assessment. Patients were evaluated at entry, after 15 days, and then after 1, 2, 4, 6, 9, 12, 15, 18, 21, and 24 months. They were clinically assessed by the same rheumatologist (NB) throughout the study period as follows: the number of tender and swollen joints (using a 68-joint count), the duration of morning stiffness (in minutes), patient assessment of pain on a 100-mm visual analog scale (VAS), patient overall assessment of disease activity on a 100-mm VAS, physician overall assessment of disease activity on a 100-mm VAS, disability as measured by the Italian version of the disability index of the Health Assessment Questionnaire (15), the ESR measured using the Westergren method (mm/hour), and C-reactive protein. In addition to being evaluated individually, these measures were also used to compute the disease activity score modified for the 28-joint count (16), and to define a patient as a responder to treatment according to the ACR 20%, 50%, and 70% criteria (17) and according to the European League Against Rheumatism (EULAR) criteria (18).

Radiographs of the hands, wrists, and feet were taken at baseline and after 12 and 24 months. They were evaluated by 2 independent, trained radiologists (AC and FP) who were blinded to the clinical and pharmacologic data, and who mutually agreed on the score to be assigned to each joint. The readings were carried out using the 32-joint damage score (DS) of the Larsen–Dale method (19) and an erosion score based on the number of eroded joints (EJC).

Laboratory assessments and adverse event monitoring. At each visit, the patients underwent a complete physical examination (including arterial blood pressure and pulse rate) and laboratory assessment; at study entry, their medical history was recorded and they underwent chest radiography and electrocardiography. The laboratory tests included a complete blood cell count; platelet count; the measurement of serum bilirubin, transaminases, creatinine, uric acid, electrolytes, blood urea nitrogen, whole blood trough CSA levels (using a specific monoclonal antibody); and urinalysis. Possible adverse events were monitored by asking questions aimed at identifying any therapy-related problem that had occurred since the last visit. Concurrent medications were recorded at each visit.

Statistical analysis. The mean changes with their 95% confidence intervals between baseline and end of the open-label period (month 6) were calculated for all of the clinical variables and tested by means of a paired student's t-test. The differences in the same variables between the end of the open-label period and the end of the randomized period (month 24) were compared within group by means of a paired student's t-test and between groups by means of an unpaired student's t-test. The differences in categorical data were evaluated using the chi-square test. The radiographic findings were analyzed by comparing the mean scores at baseline and month 12, and at month 12 and month 24, within group (paired t-test) and between groups (unpaired t-test). Bonferroni correction was applied when required. The analysis was based on the intention-to-treat approach using the data collected at the last protocol visit in the case of premature study withdrawal (last observation carried forward; LOCF). This method was used because most of the patients who discontinued the study therapy were started on other DMARDs, and so an evaluation at the end of the trial period would have yielded misleading results.

The survival on treatment analysis was carried out using the Kaplan–Meier method.

RESULTS

Months 0–6 (combination period).Patients and drug dosages. A total of 62 patients were enrolled in the trial, but 5 withdrew their consent before receiving a single dose of the study therapy and were, therefore, not included in the data analysis. The baseline clinical and demographic characteristics of the 57 evaluable patients are shown in Table 1. During this first period of the trial, 7 of the 57 patients (12.3%) discontinued the study therapy because of adverse events (Table 2). The mean (±SD) baseline dosages of CSA and MTX were 2.7 ± 0.2 mg/kg/day and 9.8 ± 0.6 mg/week, respectively; at the end of the period, they were 2.3 ± 0.8 mg/kg/day and 9.3 ± 2.8 mg/week. These differences were not statistically significant. During this stage of the trial, the CSA dosage was reduced in 11 patients and increased in 1 and the MTX dosage was reduced in 7 patients and increased in 6.

Table 1. Baseline demographic and clinical characteristics of the 57 patients who took at least 1 dose of the study therapy*
CharacteristicValue
  • *

    Values are given as mean ± SD (range). ESR = erythrocyte sedimentation rate.

Sex (F/M)53/4
Age (years)49.5 ± 11.6 (20.2–65.1)
Disease duration (years)0.8 ± 0.5 (0.5–2.0)
No. of tender joints16.9 ± 7.9 (5–39)
No. of swollen joints13.3 ± 6.2 (5–36)
Morning stiffness (minutes)129.6 ± 85.6 (20–420)
ESR (mm/hour)47.3 ± 23.9 (5–110)
Disease activity score/28 joints5.4 ± 0.8 (3.8–7.3)
Table 2. Adverse events responsible for the early withdrawal of 11 patients from the study*
Adverse eventTime of withdrawalStudy therapyOutcome
  • *

    CSA = cyclosporin A; MTX = methotrexate.

Periocular edemaMonth 2CSA + MTXResolution
Panniculitis in the lower limbsMonth 1CSA + MTXResolution
Epigastric pain, vision reductionMonth 3CSA + MTXResolution
BreathlessnessMonth 3CSA + MTXResolution
Cognitive disorder due to diffuse brain hypoperfusionMonth 2CSA + MTXImprovement
Epigastric pain, nauseaMonth 4CSA + MTXResolution
AnemiaMonth 4CSA + MTXResolution
Increased creatinine, vomiting, diarrheaMonth 4CSA + MTXResolution
Interstitial pulmonary diseaseMonth 7MTXResolution
Increased creatinine (80% more than baseline)Month 9CSAResolution
Arterial hypertension and increased transaminasesMonth 12CSAResolution
MyalgiaMonth 15CSAResolution

Response. At the end of this first phase, the mean study outcome values all showed a highly significant improvement ranging from 37% (ESR) to 78.5% (morning stiffness) (Table 3). Of the 50 patients who completed this phase, 42 (84.0%) could be defined as ACR 20% responders, 30 (60.0%) as ACR 50% responders, 23 (46.0%) as ACR 70% responders, and 42 (84.0%) as good or moderate responders according to the EULAR criteria. Indirect evidence of the clinical response was the reduction in the mean corticosteroid dosage from 2.2 ± 3.1 to 1.4 ± 2.3 mg/day.

Table 3. Variations in mean values of disease activity measures and corticosteroid dosages in 57 patients during the combination period*
VariableBaseline mean ± SDChange at 6 month Mean ± SD (95% CI)Improvement %P
  • *

    Intention-to-treat analysis using the last observation carried forward method was used for withdrawals. CI = confidence interval; VAS = visual analog scale; HAQ = Health Assessment Questionnaire; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein.

No. of tender joints16.9 ± 7.9−11.88 ± 9.66 (−14.39,−9.37)74.4<0.01
No. of swollen joints13.3 ± 6.2−9.98 ± 7.45 (−11.92,−8.05)75.2<0.01
Pain (mm, VAS)62.3 ± 22.5−39.19 ± 27.41 (−46.31,−32.07)62.9<0.01
Morning stiffness (minutes)129.6 ± 85.6101.76 ± 95.89 (76.86,126.66)78.5<0.01
Patient's global assessment (mm, VAS)65.6 ± 24.4−38.89 ± 32.10 (−47.23,−30.56)59.3<0.01
Physician's global assessment (mm, VAS)56.1 ± 25.0−39.79 ± 29.92 (−47.56,−32.02)70.9<0.01
Disability (Disability Index of HAQ)1.5 ± 0.8−1.01 ± 0.88 (−1.24,−0.79)66.7<0.01
ESR (mm/hour)47.3 ± 23.9−17.19 ± 23.56 (−23.31,−11.07)36.4<0.01
CRP (mg/dl)2.8 ± 3.2−2.11 ± 3.33 (−2.98,−1.24)75.0<0.01
Disease activity score5.4 ± 0.8−1.99 ± 1.43 (−2.36,−1.61)37.0<0.01
Corticosteroid dose (mg/day)2.2 ± 3.1−0.81 ± 2.5 (−1.45,−0.17)37.0<0.02

Tolerability. In addition to the adverse events responsible for early withdrawal from the study (Table 2), a large number of minor adverse events occurred in 44 patients (77.2%) (Table 4). The mean value of serum creatinine (SD) was 0.7 (0.2) mg/dl at baseline and 0.9 (0.2) at the final evaluation (P < 0.01). The increase in serum creatinine of 60% from the baseline value recorded in 1 patient quickly resolved after therapy discontinuation. Mean values of systolic and diastolic arterial pressure (SD) were 80.9 (8.7) and 132.2 (25.7) mm Hg at baseline, and 82.9 (8.3) and 132.2 (24.6) mm Hg at the last scheduled visit (differences not significant). An increase in arterial blood pressure developed in 8 (14%) patients and required medication in 5 cases. No patient stopped the study therapy because of arterial hypertension.

Table 4. Adverse events not responsible for study withdrawals throughout the trial period*
Adverse eventMonth 0–6Month 6–24
CSA + MTX (n = 57)CSA (n = 22)MTX (n = 27)
  • *

    CSA = cyclosporin A; MTX = methotrexate.

Nausea/vomiting19 (33.3%)1 (4.5%)12 (44.4%)
Increased serum transaminase17 (29.8%)5 (22.7%)16 (59.3%)
Arterial hypertension13 (22.8%)5 (22.7%)0
Increased serum creatinine11 (19.3%)3 (13.6%)1 (3.7%)
Hypertrichosis5 (8.8%)1 (4.5%)0
Hair loss4 (7.0%)00
General malaise3 (5.3%)00
Oral ulcer3 (5.3%)01 (3.7%)
Gingival hypertrophy3 (5.3%)00
Epigastric pain3 (5.3%)00
Rash/skin itching3 (5.3%)00
Muscular cramps2 (3.5%)00
Headache/dizziness2 (3.5%)1 (4.5%)1 (3.7%)
Others11 (19.3%)1 (4.5%)4 (14.8%)

Months 6–24 (single agent period).Patient and drug doses. Fifty patients completed the 6-month combination period and were, therefore, eligible for the second stage of the study. One patient withdrew because of personal reasons, so 49 were allocated to 1 of the 2 groups: 22 to receive CSA and 27 to receive MTX. During the second period of the trial, 18 patients (14 taking CSA and 4 taking MTX) prematurely discontinued the therapy and 2 of the CSA patients were lost to followup because they moved to a distant location. Consequently, only 29 patients were still taking the study therapy (6 taking CSA and 23 taking MTX) at month 24, when the overall probability (±SEM) of survival on treatment was 0.592 ± 0.07 (0.273 ± 0.09 for CSA and 0.852 ± 0.07 for MTX; P < 0.001).

Starting from baseline, the mean time (±SD) on study treatment for the patients in the CSA group who prematurely discontinued therapy was 13.1 ± 1.0 months. The reasons for premature withdrawal were adverse events in 4 cases (3 taking CSA and 1 taking MTX; Table 2), lack of efficacy in 12 (9 taking CSA and 3 taking MTX), and other causes in 4 patients (all taking CSA).

At the start of the second trial period, the mean dosages of CSA and MTX were 2.5 ± 0.4 mg/kg/day and 9.8 ± 1.0 mg/week, respectively; at the end of the study, they were 2.5 ± 0.4 mg/kg/day and 10.1 ± 2.2 mg/week. During this period, the CSA dosage was reduced in 6 patients and increased in 4 and the MTX dose was reduced in 4 patients and increased in 12.

Response. The results shown in Table 5 reflect the higher rate of therapy discontinuations in the CSA group, in which the mean values of all of the efficacy variables significantly worsened; no significant differences were observed in the MTX group. According to the LOCF analysis, of the 22 CSA patients, 9 (40.9%) were ACR 20% responders, 8 (36.4%) were ACR 50% responders, 6 (27.3%) were ACR 70% responders, and 11 (50.0%) were good or moderate EULAR responders. Of the 27 MTX patients, 24 (88.9%) were ACR 20% responders, 19 (70.4%) were ACR 50% responders, 11 (37.9%) were ACR 70% responders, and 25 (92.6%) were EULAR responders. Of the 6 CSA patients who completed the trial, 4 (66.7%) were ACR 20% responders and 3 (50.0%) were both ACR 50% and ACR 70% responders. Of the 23 MTX patients who completed the trial, 21 (91.3%) were ACR 20% responders, 18 (78.3%) were ACR 50% responders, and 10 (43.5%) were ACR 70% responders.

Table 5. Variations in mean values of disease activity measures and corticosteroid dosage in 49 patients during the single-agent period (months 6–24)*
VariableCSA (n = 22)MTX (n = 27)CSA versus MTX P
Month 6 mean ± SDChange at month 24 mean ± SD (95% CI)PMMonth 6 mean ± SDChange at month 24 mean ± SD (95% CI)P
  • *

    Intention-to-treat analysis using the last observation carried forward method was used for withdrawals. CSA = cyclosporin A; MTX = methotrexate; CI = confidence interval; VAS = visual analog scale; HAQ = Health Assessment Questionnaire; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein.

  • Within group.

  • Between groups.

No. of tender joints4.3 ± 5.2+4.68 ± 9.73 (0.62,8.75)0.0354.8 ± 5.7−0.56 ± 7.48 (−3.38,2.27)NSNS
No. of swollen joints3.2 ± 4.0+5.73 ± 10.34 (1.40,10.05)0.0173.1 ± 4.6+1.11 ± 6.95 (−1.51,3.73)NSNS
Pain (mm, VAS)22.2 ± 20.1+17.27 ± 27.63 (5.73,28.82)0.00821.1 ± 18.4−0.41 ± 21.37 (−8.47,7.66)NS0.001
Morning stiffness (minutes)17.3 ± 29.3+58.86 ± 27.63 (20.56,97.16)0.00733.0 ± 59.0+2.22 ± 76.78 (−26.74,31.19)NSNS
Patient's global assessment (mm, VAS)23.4 ± 17.2+16.14 ± 29.22 (3.92,28.35)0.01726.8 ± 25.0−6.18 ± 28.62 (−16.98,4.61)NS0.007
Physician's global assessment (mm, VAS)12.4 ± 12.4+22.41 ± 22.52 (12.99,31.82)<0.00115.8 ± 22.3−1.33 ± 24.18 (−10.45,7.79)NS0.003
Disability (Disability Index, HAQ)0.53 ± 0.53+0.38 ± 0.58 (0.13,0.624)0.0060.43 ± 0.45+0.18 ± 0.70 (−0.89,0.44)NSNS
ESR (mm/hour)29.2 ± 17.4+10.59 ± 19.30 (2.52,18.66)0.01829.4 ± 13.2−1.30 ± 18.76 (−8.37,5.78)NSNS
CRP (mg/dl)0.8 ± 1.1+0.49 ± 1.14 (0.02,0.96)0.0560.6 ± 1.3+0.24 ± 1.69 (−0.40,0.88)NSNS
Disease activity score3.2 ± 1.2+0.96 ± 1.52 (0.33,1.60)0.0073.5 ± 1.0−0.31 ± 1.27 (−0.79,0.71)NS0.012
Corticosteroid dose (mg/day)1.8 ± 2.6+1.60 ± 3.51 (0.10,3.11)0.0491.2 ± 2.0+0.05 ± 2.46 (−0.88,0.98)NS0.009

Tolerability. Four patients (3 taking CSA and 1 taking MTX) experienced adverse events leading to premature study withdrawal (Table 2); other adverse events occurred in 11 CSA and 24 MTX patients (Table 4). In the CSA group, the mean value of serum creatinine (SD) was 0.8 (0.2) mg/dl at the start and 0.9 (0.2) at the final evaluation (difference not significant). An increase of creatinine serum value of 80% from baseline, which resolved quickly after therapy discotinuation, was responsible for study withdrawal in 1 CSA patient. The mean values of systolic and diastolic arterial pressure (SD) were 85.0 (8.9) and 133.4 (32.8) mm Hg at the start and 84.1 (10.0) and 137.0 (17.2) mm Hg at the last scheduled visit (differences not significant). In addition to the 5 patients who were started on antihypertensive therapy during the first 6 months, another CSA patient developed arterial hypertension requiring medication.

Radiographic evaluation. Because the radiographs of 4 patients were not available, only 19 CSA and 26 MTX patients were radiologically evaluable. At baseline, the mean DSs (±SEM) in the CSA and MTX groups were 5.1 ± 1.5 and 5.0 ± 1.7, respectively, and the mean EJCs (±SEM) were 1.1 ± 0.3 and 0.7 ± 0.3. The variations of these scores at month 12 and 24 are shown in Table 6. At month 12 in the CSA group, the mean DS had increased to 12.4 ± 3.2 and the mean EJC to 3.2 ± 1.1. At month 12 in the MTX group, the corresponding figures were 10.4 ± 2.5 and 2.2 ± 0.8. A slight increase in the mean values of the radiographic scores was recorded at month 24: in the CSA group the mean DS and EJC were 15.6 ± 4.7 and 4.3 ± 1.6, respectively; in the MTX group, they were 15.0 ± 3.4 and 3.0 ± 1.0, respectively. There was no significant difference between the 2 treatment groups at month 12 or month 24. Radiographic progression occurred in patients with no apparent radiologic damage at baseline and in patients with already established anatomic changes, as shown by the progression recorded in 3 of the 9 CSA patients and in 6 of the 18 MTX patients whose DS and EJC baseline values were 0.

Table 6. Variations in mean values of the radiographic scores between baseline and month 12 and months 12 and 24*
TimeCSA (n = 19)MTX (n = 26)
DSEJCDSEJC
  • *

    CSA = cyclosporin A; MTX = methotrexate; DS = damage score; EJC = eroded joint count; CI = confidence interval.

  • Statistically significant difference (Bonferroni correction).

Baseline; mean ± SEM5.1 ± 1.51.1 ± 0.35.0 ± 1.70.7 ± 0.3
Change at month 12; mean ± SEM (95% CI) and P value+7.3 ± 3.2+2.2 ± 1.0+5.2 ± 2.0+1.5 ± 0.7
(1.12,13.51)(0.23,4.08)(1.37,9.10)(0.09,2.91)
P = 0.03P = 0.04P = 0.02P = 0.05
Change at month 24; mean ± SEM (95% CI) and P value+3.2 ± 2.0+1.1 ± 0.6+5.2 ± 1.5+0.9 ± 0.3
(−0.65,7.07)(−0.14,2.35)(2.23,8.07)(0.31,1.46)
P = 0.1P = 0.1P < 0.01P < 0.01

DISCUSSION

This study was intended to evaluate safety and efficacy of early intervention with CSA/MTX in patients younger than 65 years with active, nonerosive RA, and to probe the feasibility of the step-down approach to therapy. The results of the first stage of this trial seem to suggest that combination therapy with CSA and MTX is a feasible treatment in early RA. The overall tolerability and safety of the therapy proved to be good. Although adverse events (Table 4) occurred in most of the patients, they were usually mild and transient, and those (Table 2) that led to early study withdrawal of 7 patients (12.3%) resolved after discontinuation of the therapy. At the end of this period of the study, 84.0% of the 50 remaining patients were ACR 20% responders, 60.0% were ACR 50% responders, and 46.0% were ACR 70% responders. Because of the nonrandomized setting, these figures do not allow any conclusions to be drawn as to the efficacy of this combination therapy. Because MTX and CSA have been shown to control rheumatoid joint inflammation through different mechanisms of action (20–23), their combination should lead to good disease control. This has been proven to be true for the treatment of RA refractory to MTX alone (8–11), but not for early, poor-prognosis RA, where 1 study has shown that the combination is not better than monotherapy (12).

The randomized phase of this trial, which was the primary part of the study, shows that after 6 months of CSA/MTX combination therapy, stepping down to single agent maintenance therapy was successful only when MTX was used. Most of the patients in the CSA group had to discontinue the drug, and the likelihood of remaining on this treatment was only 0.273 ± 0.09, as compared with 0.852 ± 0.07 for the MTX group. It must be said that because of the low number of patients in the CSA arm who completed the trial (only 6), any conclusion should be considered qualitative. Loss of efficacy, which occurred in 9 of the 14 CSA withdrawals, was the main reason for therapy discontinuation, but adverse events (3 cases) and events unrelated to the therapy (4 cases) also played a role. In the MTX group, loss of efficacy was responsible for discontinuation in 3 patients, and there was 1 adverse event. The LOCF analysis of the disease activity measures recorded at the last visit was consistent with a tendency toward a decrease in efficacy in the CSA patients.

Tolerability and safety were good in this second period of the trial: all of the adverse events responsible for study withdrawals resolved after discontinuation of the therapy and, although a large number of minor adverse events were recorded, they were generally mild and well tolerated. These results suggest that if CSA/MTX combination therapy is started in early, active nonerosive RA with the intention of employing a step-down approach, MTX is the agent to be used as maintenance monotherapy after 6 months of the combination.

The radiologic data added some interesting information. Radiographs were taken at the twelfth rather than the sixth month because the effects of DMARDs on bone and cartilage tend to be delayed, and 6 months may have been too short a period to detect any changes. Articular damage measured by the DS and EJC was seen in both treatment groups at month 12 and 24. Because there was no control group, the progression recorded at month 12 only tells us that some joint deterioration developed, despite combination therapy and good control of articular symptoms.

The 24-month evaluation showed a further progression of radiographic damage in both treatment groups. Although the increase in radiographic scores from the 12-month values was greater in the MTX group than in the CSA group, progression was comparable in the 2 groups. Because most of the patients in the CSA group discontinued the study therapy long before the end of the trial (the mean time of treatment in the CSA withdrawals was about 13 months starting from baseline), it seems that well-conducted maintenance treatment with MTX had the same impact on radiographic progression as prematurely stopped maintenance treatment with CSA. This could have been due to a temporary tail effect of the 6-month CSA/MTX combination therapy. However, it must be said that, given the differences of the DS and EJC of the 2 groups recorded between months 12 and 24 and the low number of patients, this study had virtually no power to detect significant differences in the radiographic scores. The anatomic joint deterioration continued steadily throughout the trial but between months 12 and 24 it was particularly slow, with a mean number of new eroded joints of only 1 in both treatment groups. In conclusion, the radiologic data suggest that some joint deterioration occurred despite combination therapy, that this deterioration was mild between months 12 and 24, and that no difference was seen between CSA and MTX.

More aggressive combination treatment of severe RA early in the course of disease (i.e., starting less than 2 years after diagnosis) could lead to better clinical control and prevention of long-term disability (24). There are two basic approaches to the use of combinations of classic DMARDs in early RA: the step-up approach, which consists of adding another drug in the case of patients who fail to respond to conventional monotherapy, and the more aggressive step-down approach, which means that patients are given drug combinations at diagnosis and then switched to maintenance therapy with just 1 drug. This study addressed the issue of early intervention with combination CSA/MTX therapy in patients younger than 65 years with active, nonerosive RA, followed by a step-down to single-agent maintenance treatment. The results suggest that such a combination was safe and probably effective in improving inflammation-related joint symptoms. Because most of the patients continuing on CSA alone experienced disease symptom worsening, MTX is the drug to which patients should be switched following a 6-month combination period. Radiographic damage occurred despite the combination therapy; it was comparable in the 2 groups, but overall it seemed to be mild. Further studies are needed to know whether the ability of both CSA and MTX to slow structural joint damage is additive or synergistic when the 2 agents are used in association (25–28). Taken together, these results do not seem to be in favor of the CSA/MTX combination therapy and the step-down approach used in this trial. Other approaches, such as the step-up approach, may be more appropriate in early aggressive RA.

Acknowledgements

We are grateful to Dr. Ornella Della Casa-Alberighi and Dr. Roberto Ferrara for their constructive suggestions and continuous and competent help.

Ancillary