SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Objective

To compare the efficacy of combination therapy with methotrexate (MTX) and hydroxychloroquine (HCQ), MTX and sulfasalazine (SSZ), and MTX, HCQ, and SSZ in patients with rheumatoid arthritis (RA).

Methods

RA patients (n = 171) who had not previously been treated with combinations of the study medications were randomized to receive 1 of the 3 treatment combinations in this 2-year, double-blind, placebo-controlled protocol. HCQ was given at a dosage of 200 mg twice a day. The dosage of MTX was accelerated from 7.5 mg/week to 17.5 mg/week in all patients who were not in remission. Similarly, the dosage of SSZ was escalated from 500 mg twice a day to 1 gm twice a day in patients who were not in remission. The primary end point of the study was the percentage of patients who had a 20% response to therapy according to the American College of Rheumatology (ACR) criteria at 2 years.

Results

Intent-to-treat analysis revealed that patients receiving the triple combination responded best, with 78% achieving an ACR 20% response at 2 years, compared with 60% of those treated with MTX and HCQ (P = 0.05) and 49% of those treated with MTX and SSZ (P = 0.002). Similar trends were seen for the ACR 50% response, with 55%, 40%, and 29% of patients in the 3 treatment groups, respectively, achieving these results at 2 years (P = 0.005 for the triple combination group versus the MTX and SSZ group). All combination treatments were well-tolerated. Fourteen patients (evenly distributed among the 3 groups) withdrew from the protocol because of symptoms that were potentially related to the study medication.

Conclusion

The triple combination of MTX, SSZ, and HCQ is well-tolerated, and its efficacy is superior to that of the double combination of MTX and SSZ and is marginally superior to that of the double combination of MTX and HCQ.

Rheumatoid arthritis (RA) is a common disease that causes substantial morbidity in most patients and premature mortality in some (1–4). Over the last decade, significant progress has been made with regard to new approaches to treating this systemic disease that targets synovial tissue. One of the new approaches has been the almost universal use of combinations of disease-modifying antirheumatic drugs (DMARDs) to treat selected patients. A decade ago, DMARDs were rarely used in combination, whereas now, 99% of rheumatologists use them to treat a growing percentage of their RA patients (5).

The proliferation in the number of combinations used and their popularity was initially fueled by the realization that monotherapy with DMARDs is often ineffective. More recently, data have become available from a number of studies demonstrating the effectiveness of combinations of DMARDs in early RA (6–8) and in patients with RA who have had suboptimum responses to methotrexate (MTX) (9–15).

In one of the first studies to show the effectiveness of combinations of DMARDs, we demonstrated the superior efficacy of the triple combination of MTX, sulfasalazine (SSZ), and hydroxychloroquine (HCQ) over both MTX alone and the double combination of SSZ and HCQ (10). An obvious question from the results of our first trial was whether the double combinations with MTX were sufficient or whether all 3 drugs were important for the observed success. For this reason, and because the combination of MTX and HCQ is the one most commonly used in practice (5), we designed a trial to address this question.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

This study was conducted by the Rheumatoid Arthritis Investigational Network (RAIN), which brings rheumatologists at the University of Nebraska Medical Center (UNMC) together with rheumatologists elsewhere in Nebraska as well as in Iowa, South Dakota, Minnesota, California, and Illinois for the purpose of investigator-initiated clinical studies of RA. All of the participating physicians were involved not only in enrolling patients and collecting data, but also in developing the study protocols. Enrollment occurred at 7 centers from 1994 to 1998 and ranged from 10 to 49 patients per center.

Selection of patients

Patients whose cases were being followed at the Rheumatology Clinic of UNMC and at the private offices of the physicians in the RAIN network, and who met the entry criteria, were asked to participate in the present study. The protocol was approved by the Institutional Review Board at UNMC, as well as Institutional Review Boards at the individual sites. All patients gave their written informed consent.

The criteria for entry into the study were age 19–80 years, RA fulfilling the criteria of the American College of Rheumatology (ACR; formerly, the American Rheumatism Association) (16), disease duration >6 months, and active disease, with at least 3 of the following 4 features: erythrocyte sedimentation rate (ESR) >28 mm/hour, duration of morning stiffness ≥45 minutes, ≥8 tender joints, and ≥3 swollen joints. Patients were not eligible for the study if they had any of the following criteria: previous combination therapy with any of the medications studied in this protocol, stage IV disease, allergy to any of the study drugs, women of childbearing age who were not using adequate contraception, or significant liver, renal, hematologic, pulmonary, or cardiovascular disease.

With regard to previous MTX therapy, 2 groups of patients were enrolled in this trial. The first group of patients had not previously been treated with MTX; the second group of patients had responded to MTX suboptimally and entered the study at 17.5 mg/week of MTX orally. At randomization, patients were stratified according to previous MTX therapy.

Study design

We enrolled 171 patients in this 2-year, randomized, double-blind, placebo-controlled trial. The randomization and distribution of medication was handled centrally by the University of Nebraska Pharmacy; equal numbers of cards with each group assignment were mixed, drawn, and placed in sequentially numbered envelopes that were opened as the patients were enrolled. The randomization was designed to be blocked for every 6 patients.

Patients were treated with either MTX and HCQ, MTX and SSZ, or all 3 drugs. Each patient received 3 bottles: bottle A contained MTX (a generous gift from Mylan Pharmaceuticals, Morgantown, WV), bottle B contained HCQ or a matching placebo (Plaquenil; a generous gift from Sanofi-Winthrop, New York, NY), and bottle C contained SSZ or a matching placebo (Azulfidine; a generous gift from Pharmacia & Upjohn, Kalamazoo, MI). The placebos were supplied by the pharmaceutical companies and were identical to the active medications.

Patients who had had suboptimum responses to MTX at a dosage of 17.5 mg/week entered the study at this dosage of MTX. If the patient had not previously received MTX, the drug was started at 7.5 mg/week and increased to 12.5 mg/week at 2 months and to a maximum of 17.5 mg/week at 4 months if ACR remission criteria (17) had not been met at these assessments. Similarly, SSZ was started at 500 mg twice a day and increased to 1 gm twice a day at 6 months unless remission had been achieved. The dosage of HCQ was 200 mg twice a day and was kept constant throughout the study.

The patients were evaluated every 2 months for the first 6 months and then every 3 months until the end of the 2-year study by their physicians, who were unaware of the treatment assignment. If at the 1-year evaluation, after 6 months of maximum therapy, patients did not meet the ACR criteria for 20% improvement (18), we considered their treatment to be ineffective. If they had improved by ≥20%, we evaluated them every 3 months for the duration of the 2-year study.

All data were collected by physicians and study coordinators at each study site. Data collection forms were sent by fax to UNMC.

Evaluation criteria

The primary outcome variable was a 20% response to treatment, according to the ACR improvement criteria (18), at 2 years. To be considered a successful completer, patients must have achieved this level of improvement by 1 year and maintained it until 2 years. The main analysis was an intent-to-treat analysis. If patients withdrew early because of possible medication side effects or inefficacy, they were considered to have completed the study as a treatment failure. For the ACR core criteria, we utilized a modified joint count in which 38 joints were scored (19) and the Arthritis Impact Measurement Scales 2 (AIMS2) (20). Since ESRs were required only at the beginning and end of the protocol, interim ACR 20% responses were based on patients meeting 2 of the remaining 4 below-the-line requirements (modified ACR 20% response).

The individual components of the ACR core set of disease activity measures, as well as the duration of morning stiffness, are also reported. Additionally, results for ACR 50% and ACR 70% responders are reported.

Monitoring of toxicity

An ophthalmologic examination was performed on all patients every 6 months. Complete blood cell counts (including platelet counts) and measurement of serum levels of aspartate aminotransferase and albumin were performed monthly on all patients throughout the study. ESRs were measured at the beginning and at the end of the trial.

Patients were excluded from further study if their treating physician thought that the results of their laboratory tests justified withdrawal. Patients were also withdrawn from the study if they had any other symptoms of toxicity that were considered to limit further therapy with any of these medications.

Concurrent therapy

We permitted concurrent therapy with systemic corticosteroids if the dosage remained stable throughout the study period and the patient took no more than 10 mg of prednisone (or its equivalent) per day. We also permitted the use of nonsteroidal antiinflammatory medications, both as regular therapy and on an as-needed basis. Joint injections were permitted, but were limited to 2 injections during the 2-year trial (no more than 1 injection in any joint), and were not allowed within 6 weeks prior to any evaluation time point.

Statistical analysis

The primary end point was successful completion of the 2-year trial, having achieved an ACR 20% response at 1 year and maintaining this response for the remainder of the trial. The distribution of treatment failures was compared among the treatment groups by chi-square analysis. When the overall results of the chi-square analysis indicated significant differences in the distribution of failures across the 3 groups, pairwise comparisons were made using a chi-square test. A Bonferroni correction was used to account for multiple comparisons, resulting in an adjusted significance level of 0.0167.

To adjust the treatment comparison for possible demographic and baseline confounding factors, a logistic regression analysis was used to compare the odds of ACR 20% response failure at 2 years for each of the 2-drug groups relative to the 3-drug group. Baseline characteristics, including age, sex, disease duration, steroid dosage, tender joint count, swollen joint count, pain, patient's global assessment, physician's global assessment, ESR, AIMS2, and duration of morning stiffness, were included in the regression model. Terms were omitted in a step-down manner until all terms in the model were significant at the 0.10 level. Additionally, to best depict the number of patients who continued taking active treatment throughout the trial, a Kaplan-Meier plot was created (21). The log-rank test was used to compare treatment groups.

An analysis of variance (ANOVA) model was used to compare the average change in outcome over the treatment period among the 3 groups for the individual components of the ACR core set of disease activity measures as well as for the duration of morning stiffness. Tukey's post hoc comparison was made among all pairs of the treatment groups in which the average response differed significantly among treatment groups, based on an overall F test. The ANOVA models were also adjusted for demographic/patient characteristics, including age, sex, steroid use, and disease duration, where the continuous covariance had been categorized by quartiles. Possible interactions between the treatment group and the demographic/patient characteristics were also investigated. Results of the unadjusted models are presented when adjusted and unadjusted results were in agreement.

RESULTS

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Of the 171 patients who entered the trial, 58 were randomly assigned to receive MTX and HCQ, 55 to receive MTX and SSZ, and 58 to receive all 3 drugs. The distributions of age, disease duration, rheumatoid factor positivity, sex, and steroid usage were roughly balanced across the treatment groups (Table 1). Fourteen patients discontinued the study because of symptoms that were potentially related to the study medication, and 49 patients completed the study before 2 years because of lack of efficacy. Eleven patients stopped early for a variety of reasons (4 at the request of the patient, 5 because of protocol violations/lost to followup, and 2 at the request of their physicians), and their last observation (ACR responder status or individual parameter value) was carried forward. The remaining 97 patients completed the 2-year trial, having achieved a minimum of an ACR 20% response at 1 year and maintained it until the 2-year point.

Table 1. Characteristics of the study patients*
 MTX and HCQMTX and SSZMTX, HCQ, and SSZAll patients
  • *

    MTX = methotrexate; HCQ = hydroxychloroquine; SSZ = sulfasalazine.

No. of patients585558171
Age, years, mean (range)50.9 (28–68)52.5 (25–71)48.9 (26–66)50.9 (25–71)
% female78847679
Disease duration, years, mean ± SD7.9 ± 10.05.8 ± 5.96.9 ± 8.46.9 ± 6.8
% rheumatoid factor positive88888988
% taking steroids (mean dosage, mg/day)71 (4.4)56 (3.1)50 (3.6)59 (3.7)

Since medication dosages were escalated based on the absence of clinical remission, the mean dosages for the patients who completed the trial did not differ among the 3 groups. The MTX dosage ranged from 16.9 to 17.3 mg/week. The SSZ dosage ranged from 1.86 to 1.88 gm/day. Dosages of the study medications were escalated to their maximum in >85% of the patients in all treatment arms.

Toxicity

Fourteen patients were withdrawn from this 2-year trial by their treating physicians because of a variety of side effects that were potentially related to the study medications. These patients were evenly distributed among the 3 treatment groups. In the MTX and SSZ group, 5 patients were withdrawn, 1 each because of headaches, rash, and pneumonia and 2 because of gastrointestinal distress. In the MTX and HCQ group, 5 patients were withdrawn, 1 each because of weight loss, gastrointestinal distress, possible HCQ-related changes in the eye, myocardial infarction, and lobar pneumonia. In the group that received all 3 drugs, 4 patients were withdrawn, 1 each for gastrointestinal distress, transient changes on liver function studies, mild neutropenia, and non-Hodgkin's lymphoma.

Treatment outcomes

The primary outcome variable of this trial was an ACR 20% response at 2 years. An intent-to-treat analysis revealed that 35 of the 58 patients in the MTX and HCQ group, 27 of the 55 patients in the MTX and SSZ group, and 45 of the 58 patients in the triple therapy group achieved an ACR 20% response at 2 years (Figure 1). The probability of treatment failure differed significantly among the groups (P = 0.007). Based on pairwise comparisons relative to triple therapy, the probability of failure was significantly greater for the MTX and SSZ group (P = 0.002) and marginally significant for the MTX and HCQ group (P = 0.05). A Bonferroni adjustment for multiple comparisons would require a P value of 0.0167 for significance. Also shown in Figure 1 are the results for the ACR 50% and ACR 70% responses. Two patients in each treatment group met the ACR criteria for clinical remission (17) at 2 years.

thumbnail image

Figure 1. Patients with American College of Rheumatology (ACR) 20%, 50%, and 70% responses at 2 years, by treatment group. MTX = methotrexate; HCQ = hydroxychloroquine; SSZ = sulfasalazine; NS = not significant.

Download figure to PowerPoint

Figure 2 presents a Kaplan-Meier plot showing the time to treatment failure for patients in each of the 3 groups. The difference among these curves was significant (P = 0.006 by log-rank test).

thumbnail image

Figure 2. Kaplan-Meier plots of patients with ACR 20% responses to the assigned study treatment. Estimated response at 24 months: 77% for those taking MTX, HCQ, and SSZ, 60% for those taking MTX and HCQ, and 46% for those taking MTX and SSZ. See Figure 1 for definitions.

Download figure to PowerPoint

The individual parameters of the ACR core set of disease activity measures as well as the duration of morning stiffness are shown in Table 2, representing the findings at study entry. The 3 groups were roughly balanced in terms of these parameters at study entry. To assure that differences in these parameters at study entry did not confound treatment differences, logistic regression analysis was performed, adjusting for the following variables: age, sex, disease duration, steroid dosage, tender joint count, swollen joint count, pain, patient's global assessment, physician's global assessment, ESR, and AIMS2. The results of the adjusted model were consistent with the results of the unadjusted model.

Table 2. Initial values for the ACR core set of disease activity measures as well as duration of morning stiffness, by treatment group*
 MTX and HCQMTX, HCQ, and SSZMTX and SSZ
  • *

    Values are the mean ± SD. ACR = American College of Rheumatology; MTX = methotrexate; HCQ = hydroxychloroquine; SSZ = sulfasalazine; ESR = erythrocyte sedimentation rate; AIMS2 = Arthritis Impact Measurement Scales 2.

Tender joint count (maximum 38)15.7 ± 8.219.7 ± 9.215.6 ± 7.4
Swollen joint count (maximum 38)21.1 ± 8.324.0 ± 8.819.1 ± 7.9
Pain (0–10 scale)4.6 ± 2.25.3 ± 2.15.2 ± 2.0
Global assessment
 Patient's (0–10 scale)5.0 ± 2.35.1 ± 2.24.6 ± 2.3
 Physician's (0–10 scale)5.3 ± 1.55.5 ± 1.55.3 ± 1.5
ESR (mm/hour)28.5 ± 20.330.1 ± 21.034.1 ± 26.5
AIMS22.2 ± 1.32.4 ± 1.52.3 ± 1.2
Morning stiffness (minutes)116 ± 94133 ± 91116 ± 83

Table 3 shows the mean change in the ACR core parameters as well as in the duration of morning stiffness at the end of the study, by treatment group. In all cases, the improvement in the group treated with all 3 drugs was greater than that in either of the other 2 treatment groups. These differences reached statistical significance for a number of comparisons, including the tender joint count, ESR, and duration of morning stiffness for triple therapy versus MTX and HCQ, and the swollen joint count, AIMS2, and duration of morning stiffness for triple therapy versus MTX and SSZ.

Table 3. Changes in values for the ACR core set of disease activity measures as well as duration of morning stiffness, by treatment group*
 MTX and HCQMTX, HCQ, and SSZMTX and SSZP
MTX + HCQ vs. all 3 drugsMTX + SSZ vs. all 3 drugsMTX + HCQ vs. MTX + SSZ
  • *

    Values are the mean ± SD. See Table 2 for definitions.

  • Statistically significant.

Tender joint count (maximum 38)−10.0 ± 9.6−14.0 ± 8.6−10.2 ± 8.90.050.080.99
Swollen joint count (maximum 38)−14.0 ± 10.0−17.0 ± 8.9−13.0 ± 9.90.210.040.70
Pain (0–10 scale)−1.8 ± 2.6−2.7 ± 3.0−1.8 ± 3.40.300.300.99
Global assessment
 Patient's (0–10 scale)−2.1 ± 3.2−2.5 ± 2.6−1.5 ± 3.10.800.190.51
 Physician's (0–10 scale)−2.7 ± 2.3−3.2 ± 1.9−2.3 ± 2.60.470.110.67
ESR (mm/hour)−3.7 ± 21−13.0 ± 18−6.8 ± 190.030.210.70
AIMS2−0.94 ± 1.31−1.6 ± 1.4−0.71 ± 1.030.080.020.73
Morning stiffness (minutes)−59.2 ± 103.3−109.3 ± 86.4−53.2 ± 89.50.010.0060.94

With respect to previous treatment with MTX, 2 groups of patients were enrolled in this trial (see Patients and Methods). A subanalysis of these groups is shown in Figure 3. Similar trends remain in these smaller groups of patients, with more patients responding to triple therapy than to either of the double therapies. This difference reached marginal statistical significance in favor of triple therapy over MTX and HCQ for those without previous MTX exposure, after adjusting for multiple comparisons, and the difference reached statistical significance for triple therapy over MTX and SSZ for those who entered the trial as suboptimum responders to MTX therapy.

thumbnail image

Figure 3. Patients with ACR 20%, 50%, and 70% responses, by treatment group, for A, those who had not previously been treated with MTX, B, those who had previously been treated with MTX at study entry. See Figure 1 for definitions.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

The treatment of RA with the combination of MTX, HCQ, and SSZ has been called triple therapy. This combination was originally shown to be more effective than either MTX alone or the combination of SSZ and HCQ in a study we reported in 1996 (10). An obvious question that arose from our initial trial was whether all 3 drugs were necessary in this combination, or whether patients treated with the double combinations of MTX and either SSZ or HCQ would do just as well. The relevance of this question was highlighted by the fact that the double combination of MTX and HCQ is by far the combination most commonly prescribed by clinicians in the US (5). The results of the present trial corroborate the results of our initial combination trial, in which we demonstrated the efficacy of the triple-DMARD combination therapy (MTX, HCQ, and SSZ), and provide evidence that all 3 drugs in combination are better than the MTX double combinations. Furthermore, as in our initial trial, the present study showed that this combination of medications was well-tolerated, and study withdrawals were similar in all 3 groups.

Two other groups of investigators recently reported on the efficacy of the triple combination of MTX, SSZ, and HCQ (7, 8). Both studies were open, but randomized, trials in patients with early RA. In the Finnish RA trial (8), patients were randomized to receive either triple therapy plus prednisolone or SSZ plus prednisolone at the discretion of the treating physician. The main end point of that trial was remission at 2 years, and patients randomized to triple therapy were more likely to have achieved remission at 2 years (odds ratio 2.7). Importantly, those authors have also reported that patients treated with all 3 drugs were less likely to develop a C1–C2 subluxation (22). In the second trial, Turkish patients were randomized to receive single, double, or triple therapy (7). Primary end points included an ACR 50% response, remission according to the ACR criteria, and no radiographic progression. In all cases, double therapy was better than single therapy, and triple therapy was better than double therapy, except for the criterion of no radiographic progression, which favored triple therapy but did not reach statistical significance for the comparison between double and triple therapy. The open nature of these trials is an obvious weakness, but importantly, the radiographic outcomes in both trials were blinded.

An important and much-studied group of patients with RA is the group who have had only a partial or a suboptimum response to MTX. A number of studies have enrolled this type of patient and have randomized patients to receive active therapy or placebo in addition to their baseline MTX therapy (9, 11, 13–15). Trials designed in this way have demonstrated the superiority of cyclosporine (9), etanercept (14), infliximab (15), and leflunomide (13) over placebo therapy in patients who continue to take MTX. These trials have been instrumental in gaining Food and Drug Administration approval of these medications for RA or for RA suboptimum responders. An obvious weakness of these trials is the fact that no comparisons with active therapy have been done.

In our trial, a subset of the patients who were enrolled were defined as MTX suboptimum responders, that is, patients who continued to have active disease despite treatment with MTX at a dosage of 17.5 mg/week. These patients responded well in our trial (Figure 3). Those who received all 3 drugs had the best results (71% achieving ACR 20% response). This magnitude of response compares favorably with the response reported for groups of such patients treated with etanercept (71% achieving ACR 20% response at 1 year) (14), infliximab (42–50% achieving ACR 20% response at 1 year) (15), cyclosporine (48% achieving ACR 20% response at 1 year) (9), and leflunomide (51% achieving ACR 20% response at 1 year) (13).

This and other trials continue to demonstrate the superiority of combination therapy in the treatment of RA. Trials in early disease have consistently shown that combination therapy is superior to monotherapy (6–8, 23). In patients who have had suboptimum responses to MTX, 5 different therapies have now been shown to be more effective than MTX monotherapy (9–15). However, many critical questions remain. Should combination DMARD therapy be used in all patients with early RA? Which of the 5 effective therapies is best for MTX suboptimum responders? And, perhaps most importantly, how can we predict which patients will best respond to each therapy? Obviously, trials to address these questions are needed. In early disease, a trial that compares combination therapy with a rapid step-up approach (which is how many clinicians currently treat these patients) would be attractive. In the MTX suboptimum responders, trials directly comparing the 5 therapies that have been shown to be effective are critically needed.

The triple combination of MTX, SSZ, and HCQ is well-tolerated and relatively inexpensive. The superior efficacy of this combination over single or double therapy has now been demonstrated in 2 trials of patients with early RA (7, 8) and in 2 trials of patients with more established RA (10). Direct comparison of this therapy with other effective, but more expensive, alternatives would seem prudent.

Acknowledgements

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

We gratefully acknowledge the significant contributions of our RA patients; their attitudes and approach to life despite this chronic disease continue to be an inspiration to all of us. Without their active cooperation, clinical studies to answer critical questions in the treatment of RA would, of course, not be possible. We would also like to thank Lucie Case and Renee Crosby for their excellent secretarial assistance. Finally, we would like to thank the Albert G. and Bernice F. Hansen Foundation for funding, and Pharmacia & Upjohn, Mylan Pharmaceuticals, and Sanofi-Winthrop for supplying the study medications and placebos.

REFERENCES

  1. Top of page
  2. Abstract
  3. PATIENTS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES
  • 1
    Hochberg MC. Adult and juvenile rheumatoid arthritis: current epidemiologic concepts. Epidemiol Rev 1981; 3: 2744.
  • 2
    Scott DL, Symmons DPM, Coulton BL, Popert AJ. Long-term outcome of treating rheumatoid arthritis: results after 20 years. Lancet 1987; 1: 110811.
  • 3
    Pincus T, Brooks RH, Callahan LF. Prediction of long-term mortality in patients with rheumatoid arthritis according to simple questionnaire and joint count measures. Ann Intern Med 1994; 120: 2634.
  • 4
    Wolfe F, Mitchell DM, Sibley JT, Fries JF, Bloch DA, Williams CA, et al. The mortality of rheumatoid arthritis. Arthritis Rheum 1994; 37: 48194.
  • 5
    Mikuls TR, O'Dell J. The changing face of rheumatoid arthritis therapy: results of serial surveys. Arthritis Rheum 2000; 43: 4645.
  • 6
    Boers M, Verhoeven AC, Markusse HM, van de Laar MA, Westhovens R, van Denderen JC, et al. Randomised comparison of combined step-down prednisolone, methotrexate and sulphasalazine with sulphasalazine alone in early rheumatoid arthritis. Lancet 1997; 350: 30918.
  • 7
    Calguneri M, Pay S, Caliskaner Z, Apras S, Kiraz S, Ertenli I, et al. Combination therapy versus monotherapy for the treatment of patients with rheumatoid arthritis. Clin Exp Rheumatol 1999; 17: 699704.
  • 8
    Möttönen T, Hannonen P, Leirisalo-Repo M, Nissila M, Kautiainen H, Korpela M, et al. Comparison of combination therapy with single-drug therapy in early rheumatoid arthritis: a randomised trial. Lancet 1999; 353: 156873.
  • 9
    Tugwell P, Pincus T, Yocum D, Stein M, Gluck O, Kraag G, et al. Combination therapy with cyclosporine and methotrexate in severe rheumatoid arthritis. N Engl J Med 1995; 333: 13741.
  • 10
    O'Dell JR, Haire CE, Erikson N, Drymalski W, Palmer W, Eckhoff PJ, et al. Treatment of rheumatoid arthritis with methotrexate alone, sulfasalazine and hydroxychloroquine, or a combination of all three medications. N Engl J Med 1996; 334: 128791.
  • 11
    O'Dell JR, Haire C, Erikson N, Drymalski W, Palmer W, Maloley P, et al. Efficacy of triple DMARD therapy in patients with RA with suboptimal response to methotrexate. J Rheumatol 1996; 23 Suppl 44: 724.
  • 12
    O'Dell J, Paulsen G, Haire C, Palmer W, Wees S, Eckhoff J, et al. Combination DMARD therapy with methotrexate-sulfasalazine-hydroxychloroquine in rheumatoid arthritis: continued efficacy with minimal toxicity at 5 years [abstract]. Arthritis Rheum 1998; 41 Suppl 9: S132.
  • 13
    Kremer JM, Caldwell JR, Cannon GW, Genovese M, Cush JJ, Bathon J, et al. The combination of leflunomide and methotrexate in patients with active rheumatoid arthritis who are failing on methotrexate treatment alone: a double-blind placebo controlled study [abstract]. Arthritis Rheum 2000; 43 Suppl 9: S224.
  • 14
    Weinblatt ME, Kremer JM, Bankhurst AD, Bulpitt KJ, Fleischmann RM, Fox RI, et al. A trial of etanercept, a recombinant tumor necrosis factor receptor:Fc fusion protein, in patients with rheumatoid arthritis receiving methotrexate. N Engl J Med 1999; 340: 2539.
  • 15
    Lipsky PE, van der Heijde DM, St Clair EW, Furst DE, Breedveld FC, Kalden JR, et al, for the Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy Study Group. Infliximab and methotrexate in the treatment of rheumatoid arthritis. N Engl J Med 2000; 343: 1594602.
  • 16
    Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988; 31: 31524.
  • 17
    Pinals RS, Masi AT, Larsen RA, the Subcommittee for Criteria of Remission in Rheumatoid Arthritis of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee. Preliminary criteria for clinical remission in rheumatoid arthritis. Arthritis Rheum 1981; 24: 130815.
  • 18
    Felson DT, Anderson JJ, Boers M, Bombardier C, Furst D, Goldsmith C, et al. American College of Rheumatology preliminary definition of improvement in rheumatoid arthritis. Arthritis Rheum 1995; 38: 72735.
  • 19
    Egger MJ, Huth DA, Ward JR, Reading JC, Williams HJ. Reduced joint count indices in the evaluation of rheumatoid arthritis. Arthritis Rheum 1985; 28: 6139.
  • 20
    Meenan RF, Mason JH, Anderson JJ, Guccione AA, Kazis LE. AIMS2: the content and properties of a revised and expanded Arthritis Impact Measurement Scales health status questionnaire. Arthritis Rheum 1992; 35: 110.
  • 21
    Kaplan E, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958; 53: 45781.
  • 22
    Neva MH, Kauppi MJ, Kautiainen H, Luukkainen R, Hannonen P, Leirisalo-Repo M, et al, the FIN-RACo Trial Group. Combination drug therapy retards the development of rheumatoid atlantoaxial subluxations. Arthritis Rheum 2000; 43: 2397401.
  • 23
    Kirwan JR the Arthritis and Rheumatism Council Low-Dose Glucocorticoid Study Group. The effect of glucocorticoids on joint destruction in rheumatoid arthritis. N Engl J Med 1995; 333: 1426.