Aim Classifying more patients as rheumatoid arthritis (RA) (2010 American College of Rheumatology/European League Against Rheumatism criteria for RA) may improve treatment outcomes but may cause overtreatment in daily practice. The authors determined the efficacy of initial methotrexate (MTX) plus prednisone treatment in patients with 1987 or 2010 classified RA and undifferentiated arthritis (UA).
Method 610 recent onset RA or UA patients started with MTX 25 mg/week and prednisone 60 mg/day tapered to 7.5 mg/day in 7 weeks. Percentage remissions after 4 months were compared between RA (1987 or 2010 criteria) and UA. Predictors for remission were identified.
Results With the 2010 criteria, 19% more patients were classified as RA than with the 1987 criteria, but similar remission rates were achieved: 291/479 (61%) 2010 classified RA and 211/264 (58%) 1987 classified RA patients (p=0.52), and 79/122 (65%) UA patients (p=0.46). Anticitrullinated protein antibodies (ACPA) positive RA patients achieved more remission (66%) than ACPA negative RA patients (51%, p=0.001), but also had a lower mean baseline Disease Activity Score (DAS) (3.2 vs 3.6, p<0.001). Independent predictors for remission were male sex, low joint counts, DAS and Health Assessment Questionnaire, low body mass index and ACPA positivity.
Conclusion Initial treatment with MTX and a tapered high dose of prednisone results in similarly high remission percentages after 4 months (about 60%) in RA patients, regardless of fulfilling the 1987 or 2010 criteria, and in UA patients. Independent predictors indicate that initiating treatment while disease activity is relatively low results in more remission.
Statistics from Altmetric.com
Starting treatment earlier in the disease course of rheumatoid arthritis (RA) has improved functional and radiological outcome as compared with delayed treatment.1,–,6 New RA classification criteria support this trend,7 but have triggered concerns that some patients may now be misclassified and overtreated as a result.8
Remission has increasingly become a treatment goal in clinical trials, resulting in remission rates that vary between 26% and 42%.9
It is hypothesised that starting treatment already in the phase of undifferentiated arthritis (UA) may prevent progression to classified RA and increase permanent remission rates. However, methotrexate (MTX) monotherapy for patients with probable RA postponed but did not prevent progression to RA. Similar drug free remission rates (about 25%) were achieved in the MTX and placebo groups.10
Since in RA initial combination treatment with prednisone leads to a more rapid clinical improvement and less radiological progression of joint damage than disease modifying antirheumatic drug monotherapy,3 ,11,–,14 treatment with combination possibly in the phase of UA may increase remission and drug free remission rates, as well as improve short-term functional outcome and long-term joint damage progression.
To investigate this, we designed the IMPROVED (Induction therapy with Methotrexate and Prednisone in Rheumatoid Or Very Early arthritic Disease) study, the first clinical trial in patients with UA and early RA, with an induction phase with MTX and a tapered high dose of prednisone, aimed at achieving remission. This trial allowed us to evaluate the effect of classifying patient groups according to the old and the new RA classification criteria and to identify predictors of remission.
Patients and methods
The IMPROVED study is a multicentre, clinical trial in recent onset RA and UA patients. All patients were initially treated for 4 months with MTX 25 mg/week and a tapered high dose of prednisone, starting with 60 mg/day, tapered in 7 weeks to 7.5 mg/day, and continued in this dose up to 4 months. Later, this introduction phase will be followed by a single blind randomised controlled trial where those patients who did not achieve remission will be treated according to two treatment strategies: one starting with a combination of MTX, sulfasalazine, hydroxychloroquine and low dose prednisone and the other with a combination of MTX with adalimumab.
Rheumatologists participating in the Foundation for Applied Rheumatology Research designed and conducted the study. Patients were recruited between March 2007 and September 2010 in 12 hospitals in the Western part of the Netherlands. The Medical Ethics Committee of each participating centre approved the study protocol and all patients gave written informed consent. The IMPROVED trial was registered in the ISRCTN Register (number 11916566) and the EudraCT (number 2006-006186-16).
Patients with RA classified according to the 1987 American College of Rheumatology (ACR) criteria15 with a symptom duration of <2 years and UA, defined as likely to have early RA according to the treating rheumatologist, with at least one arthritic joint and one other painful joint, regardless of symptom duration, were included in the trial. All patients had a Disease Activity Score (DAS)≥1.6.16
Exclusion criteria included previous therapy with disease modifying antirheumatic drugs or corticosteroids, pregnancy or pregnancy wish during the study, malignancy within the last 5 years, bone marrow hypoplasia, elevated liver enzyme levels (aspartate transaminase (AST) and/or alanine transaminase (ALT)>3 timesnormal value), serum creatinine level >150 umol/l or estimated creatinine clearance of <75%, uncontrolled diabetes mellitus, uncontrolled hypertension, heart failure (New York Heart Association class III/IV), alcohol or drug abuse, serious infections in the previous 3 months or chronic infectious disease, opportunistic infections within previous 2 months, active or latent hepatitis B infection, documented HIV infection or AIDS, lymphoproliferative disease and multiple sclerosis. All patients with active tuberculosis (TB) were excluded, as well as UA patients with latent TB. RA patients with latent TB could be included if they started adequate antituberculous therapy (according to local TB specialists) prior to initiation of high dose prednisone.
Reclassification according to the 2010 ACR/EULAR classification criteria
After inclusion was complete the new classification criteria were published. Unless specified otherwise (by adding the year of classification criteria between brackets), ‘RA’ in the text denotes RA classified according to the 2010 criteria, and ‘UA’ denotes not fulfilling the 2010 criteria.
Primary outcomes after 4 months were percentage clinical remission, defined as a DAS<1.6,16 disease activity measured by DAS, functional ability measured by the Health Assessment Questionnaire (HAQ)17 and radiological progression using the Sharp/van der Heijde scoring method (SHS).18
Radiological damage was assessed by two independent readers using SHS, blinded for patient identity and time order of the radiographs.18 Progression was defined as an increase in SHS score of ≥0.5 points. Due to the small distribution of SHS scores, caused by a large proportion of patients without progression, the interobserver and intraobserver intraclass correlation coefficients were not suitable for measuring reliability.19 In 91.5% of patients both readers scored the same progression. In the others, the median (IQR) difference in progression score between readers was 2(2-3). A consensus score was reached for radiographs with inter-reader differences ≥ median difference in progression score (n=41).
Percentage remissions according to ACR/European League Against Rheumatism (EULAR) preliminary definition20 were compared with percentage remissions based on the DAS.
All outcomes were calculated according to the intention-to-treat principle. Percentage remissions in the RA and the UA group were compared using χ2 test. Categorical variables were compared between groups using χ2 test, normally distributed outcome measures using Independent Samples t test and skewed outcome measures using Mann–Whitney U test.
Independent predictors for remission were identified using univariate followed by multivariate logistic regression with achieving or not achieving remission as binominal dependent variable. All available clinical variables were entered in a univariate regression analysis. Using a p value <0.10, significant variables were then entered in the multivariate regression analysis.
Between March, 2007 and September, 2010, 730 patients signed informed consent and were screened for inclusion (figure 1). We included 610 patients: 364 RA patients (1987 classification criteria) or 479 RA patients (2010 classification criteria) and 122 UA patients (ie, not fulfilling the 2010 classification criteria) (table 1).
During 4 months, 12 patients left the trial: two patients because of a revised diagnosis (one osteoarthritis, one lupus), two because of comorbidity, six withdrew consent and two died (figure 1).
RA (1987) patients had a higher mean DAS based on more affected joints, higher erythrocyte sedimentation rate and higher serum C reactive protein than RA (2010) patients. UA patients included fewer female subjects, were less often rheumatoid factor and anticitrullinated protein antibodies (ACPA) positive, and had lower disease activity and HAQ. There was no significant difference between RA (1987 or 2010 classified) and UA patients in baseline damage scores or erosiveness (table 2).
After 4 months, DAS <1.6 was achieved in 58% of the RA (1987), 61% of the RA (2010) (p=0.52) and 65% of the UA patients (p=0.46 compared with RA 2010).
DAS improved more in the RA (2010) group than in the UA group and was similar as in the RA (1987) group, resulting in comparable mean (SD) DAS levels after 4 months: 1.4 (0.9) in UA, 1.6 (0.9) in RA (1987) and 1.5 (0.9) in RA (2010) patients. Also HAQ improved more in the RA patients than the UA patients, resulting in HAQ levels of 0.44 both in UA and RA patients (p=0.96) after 4 months.
Baseline and 4-month radiographs of hands and feet were available for 546 patients. After 4 months, 61 patients (10%) showed radiological progression, without a difference between UA and RA patients. In those with progression the median (IQR) SHS progression was 1 (1–1) point.
Patients who did not achieve remission after 4 months treatment had a higher baseline DAS and higher DAS components, and were more often ACPA negative than patients who did achieve remission (table 3). Of the ACPA positive RA patients, 66% achieved remission compared with 51% of the ACPA negative RA patients (p=0.001). ACPA positive RA patients had a lower baseline DAS (mean (SD) 3.19 (0.89)) than ACPA negative RA patients (mean (SD) 3.64 (0.94), p<0.001). ACPA negative RA patients who achieved remission had a shorter median (IQR) symptom duration (12 weeks) (8–26) than those who did not (20 weeks (10–31), p=0.02). In the whole study population, there was a trend for more remission in patients with shorter symptom duration.
The distribution of joints was different in patients with RA and UA. All RA patients had involvement of small joints (wrists, hands and feet), compared with 94% of the UA patients (p<0.001). Large joint (all other joints) involvement was found in similar percentages of RA and UA patients (73% vs 68%, p=0.22). Patients with large joint involvement had more affected small joints (median (IQR) 10 (6–17) vs 7 (4–11), p<0.001) and achieved remission less often than patients without large joint involvement (57% vs 76%, p<0.001).
Predictors for remission
Significant univariate clinical predictors for achieving remission in the total study population were baseline DAS, HAQ, symptom duration, male sex, ACPA positivity, number of affected small joints, number of affected large joints and body mass index (BMI) (table 4). Fulfilling the 1987 or 2010 classification criteria for RA was not a predictor for remission. In a multivariate regression analysis including baseline DAS and excluding number of affected small and large joints, independent predictors were baseline DAS, HAQ, symptom duration, ACPA positivity, male sex and BMI. In a model including the baseline numbers of affected small and large joints instead of the DAS, the number of affected small and large joints were both predictive, independently of each other. In this analysis, ACPA positivity was not an independent predictor (table 4).
ACR/EULAR preliminary definition of remission
According to the preliminary ACR/EULAR definition,20 157/610 (26%) of the patients achieved remission after 4 months (34 patients could not be defined because of missing data), without a difference between UA and RA patients (29/122 (24%) vs 126/479 (26%), p=0.45).
Mean (SD) DAS after 4 months of patients in ACR/EULAR remission is 0.82 (0.41).
In all, 206/610 (34%) patients did achieve DAS remission but were not in ACR/EULAR remission. They had a median (IQR) tender joint count of 0 (0–1), a median (IQR) swollen joint count of 0 (0–0), a median (IQR) C reactive protein of 5 (3–9) and a mean (SD) Visual Analogue Scale general health of 21(14).
A total of 152/610 (25%) patients achieved remission by both criteria, 201/610 (33%) did not achieve remission according to either and 5/610 (0.8%) patients were in ACR/EULAR remission but not DAS remission, based on arthritis in the feet (not included in the ACR/EULAR remission definition).
The data suggest that the ACR/EULAR definition of remission is more stringent than DAS remission, resulting in lower remission percentages. Clinical and radiological follow-up are needed to show which definition is most adequate.
During 4 months of treatment, 341/610 (56%) of the patients reported one or more adverse events (table 5). There were 16 serious adverse events in 16 (3%) of 610 patients (8 per 100 patient-years). Two patients died: a 70-year-old female subject from a myocardial infarction later found to be caused by giant cell arteriitis (incorrect inclusion due to alternative diagnosis) and an 85-year-old female subject after refusing treatment for pneumonia. Fourteen hospital admissions were reported for patients with bacterial coxarthritis, pneumocystis carinii pneumonia (a patient with pre-existing non-specific interstitial pneumonia), other pneumonia (three patients), viral pneumonitis, urothelial cell carcinoma, surgery for carcinoma of the caecum, diverticulitis, bleeding from a benign intestinal polyp, supraventricular tachycardia, hypertension, peripheral arterial occlusion and pulmonary embolism.
Initial treatment with MTX and a tapered high dose of prednisone results in similar remission rates in 2010 classified and 1987 classified RA patients and in UA patients after 4 months. The majority (90%) of the patients showed no radiological progression after 4 months. Independent predictors for remission were low baseline DAS, low numbers of affected large and small joints, ACPA positivity, male sex and BMI.
The early remission rate of 61% is higher than previously reported in cohorts such as COBRA (Combinatietherapie Bij Reumatoide Artritis) and BeSt (Behandel Strategieën), where patients also received MTX and a tapered high dose of prednisone, combined with sulfasalazine.11 ,3 This is most likely explained by our intentional inclusion of patients with milder disease activity and not (yet) fulfilling the classification criteria for RA. Also, our patients had on average a shorter symptom duration. Thus, the higher remission rate in this study would support the window of opportunity theory. However, earlier inclusion may have overclassified patients who possibly had self-limiting disease.8 Other possible explanations are the initial dose of MTX (25 mg/week compared with 7.5 mg/week in the other cohorts) and the absence of sulfasalazine in the initial drug combination.
The 2010 ACR/EULAR classification criteria were formulated to classify patients earlier in disease course.7 In this study however, the symptom duration of patients classified as RA according to the 1987 or the 2010 criteria is comparable, which might explain why we found no difference in clinical response and remission rates between the groups, even though the 2010 criteria classified 19% more patients.
Also, we found no difference in remission rates between RA and UA patients, although we hypothesised that UA patients, as presumably very early RA, would benefit more from early combination therapy, and despite the fact that UA patients had a lower mean baseline disease activity and were predominantly male subjects. This may be explained by the comparable symptom duration in UA and RA patients. Of the UA patients, 64% had a symptom duration >12 weeks, thus possibly missing the so-called window of opportunity.21 Also, only a few UA patients were ACPA positive, compared with 68% in the RA group, and ACPA positivity in the total study population was found to be a predictor of achieving remission. ACPA negative RA and ACPA negative UA both may represent or include illnesses that do not sufficiently respond to combination therapy with MTX and prednisone and require different treatments.22 Previously, in the PROMPT (Probable rheumatoid arthritis: Methotrexate versus Placebo Treatment) study ACPA negative UA patients did not benefit from treatment with MTX monotherapy.10
The baseline characteristics in this study population suggest that classifying patients as RA by the new classification criteria rests predominantly on numbers of (small) joints involved and ACPA positivity, with UA patients having less joints involved and almost all UA patients being ACPA negative. ACPA negative patients who were still classified as RA had a higher disease activity and a longer symptom duration than ACPA positive RA patients. These characteristics may explain why ACPA negative RA patients achieve less remission than ACPA positive RA patients. It is possible that they might have benefited more from treatment if they were treated earlier.
As shown in previous studies, male patients achieve more remission than female patients.23 Our results show that male sex is an independent predictor of remission and not associated with a lower pain score or tender joint count. Also, a lower BMI was found to be an independent predictor of remission, which may be related to relative underdosing of patients with a high BMI.
The early and intensive treatment with a high dose of MTX and a tapered high dose of prednisone in this study was accompanied by adverse events in more than half (56%) of the patients. Although most adverse events were mild, serious adverse events were reported in 3% of patients. Two older patients died, one from pneumonia that may have been treatment related and on the patient's request remained untreated, and one of a vasculitis related cardiac event. This patient thus was misdiagnosed and, since the lethal event occurred during treatment with the tapered dose of prednisone, possibly underdosed.
In conclusion, initial therapy with MTX and a tapered high dose of prednisone results in high remission percentages (about 60%) both in early RA patients (regardless of classification according to the 1987 or 2010 criteria) and in UA patients after 4 months of treatment. Independent predictors for remission, besides male sex and low BMI, indicate that initiation of treatment while disease activity is relatively low results in more remission, regardless of whether patients fulfil the classification criteria for RA. ACPA negative patients may benefit from early treatment, but on the whole achieve less remission on MTX with prednisone than ACPA positive patients. This may indicate that this subgroup of patients represents a different disease for which the optimal treatment remains to be determined.
We would like to thank all patients as well as the following rheumatologists (other than the authors) who participated in the IMPROVED study group (all locations are in The Netherlands): W M de Beus (Medical Center Haaglanden, Leidschendam); C Bijkerk (Reinier de Graaf Gasthuis, Delft); M H W de Bois (Medical Center Haaglanden, The Hague); M de Buck (Medical Center Haaglanden, Leidschendam); G Collée (Medical Center Haaglanden, The Hague); J A P M. Ewals (Haga Hospital, The Hague); R J Goekoop (Haga Hospital, The Hague); Y P M Goekoop-Ruiterman (Haga Hospital, The Hague); B A M Grillet (Zorgsaam, Terneuzen); J H L M van Groenendael (Franciscus Hospital, Roosendaal); L R Lard (Medical Center Haaglanden, Leidschendam); E T H Molenaar (Groene Hart Hospital, Gouda); M van Oosterhout (Groene Hart Hospital, Gouda); A J Peeters (Reinier de Graaf Gasthuis, Delft); N Riyazi (Haga hospital, The Hague); H K Ronday (Haga hospital, The Hague); A A Schouffoer (Groene Hart Hospital, Gouda); P E H Seys (Lievensberg hospital, Bergen op Zoom); P B J de Sonnaville (Oosterschelde Hospital, Goes); I Speyer (Bronovo Hospital, The Hague); G M Steup-Beekman (Bronovo Hospital, The Hague); and M L Westedt (Bronovo Hospital, The Hague). We would also like to thank all other rheumatologists and trainee rheumatologists who enrolled patients in this study, and all research nurses for their contributions.
Funding The study was designed by the investigators and financially supported by Abbott. Data collection, trial management, data collection, data analysis and preparation of the manuscript were performed by the authors.
Competing interests None.
Patient consent Obtained.
Ethics approval Approval provided by the Medical Ethics Committees of all participating hospitals.
Provenance and peer review Not commissioned; externally peer reviewed.
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.