Drs. Lovell, Schneider, Ilowite, Wallace, Gedalia, Giannini, and Stein have received consulting fees or honoraria (less than $10,000 per year). Dr. Reiff has received consulting fees or honoraria (more than $10,000 per year).
Long-term safety and efficacy of etanercept in children with polyarticular-course juvenile rheumatoid arthritis
Article first published online: 26 MAY 2006
Copyright © 2006 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 54, Issue 6, pages 1987–1994, June 2006
How to Cite
Lovell, D. J., Reiff, A., Jones, O. Y., Schneider, R., Nocton, J., Stein, L. D., Gedalia, A., Ilowite, N. T., Wallace, C. A., Whitmore, J. B., White, B., Giannini, E. H. and Pediatric Rheumatology Collaborative Study Group (2006), Long-term safety and efficacy of etanercept in children with polyarticular-course juvenile rheumatoid arthritis. Arthritis & Rheumatism, 54: 1987–1994. doi: 10.1002/art.21885
- Issue published online: 26 MAY 2006
- Article first published online: 26 MAY 2006
- Manuscript Accepted: 27 FEB 2006
- Manuscript Received: 13 OCT 2004
- Immunex, a wholly owned subsidiary of Amgen
- Wyeth Research
Previous studies showed that etanercept treatment in patients with polyarticular-course juvenile rheumatoid arthritis (JRA) provided rapid clinical improvement that was sustained for up to 2 years. The goal of our study was to provide data on safety and efficacy after 4 years of etanercept treatment in patients with JRA.
Patients with active polyarticular-course JRA who participated in an efficacy study continued etanercept treatment in an open-label extension. Safety was assessed by measuring rates of serious adverse events (SAEs) and serious infections. Efficacy was assessed using the American College of Rheumatology (ACR) Pediatric 30 criteria for improvement and standard measures of disease activity. (The ACR Pediatric 30 criteria are defined as improvement of ≥30% in at least 3 of 6 core response variables used to assess disease activity, with no more than 1 variable worsening by ≥30%.)
Of the 69 patients who enrolled in the original efficacy study, 58 patients (84%) enrolled in the extension, 34 patients received etanercept treatment for ≥4 years, and 32 of these received complete efficacy assessments. The rate of SAEs was 0.13 per patient-year, and the rate of serious infections was 0.04 per patient-year, in a total etanercept exposure of 225 patient-years. Eighty-two percent of patients who received corticosteroids at any time during the extension were able to decrease their dosage to ≤5 mg/day prednisone equivalent. Of the 32 patients with complete efficacy data who received etanercept for ≥4 years, 94% achieved an ACR Pediatric 30 response and 78% achieved an ACR Pediatric 70 response at the last study visit.
Etanercept offers an acceptable safety profile in children with polyarticular-course JRA and provides significant improvement in disease manifestations that are sustained for ≥4 years.
Until recently, the treatment of children with juvenile rheumatoid arthritis (JRA) relied on physical and occupational therapy, nonsteroidal antiinflammatory drugs (NSAIDs), corticosteroids, and disease-modifying antirheumatic drugs (DMARDs), primarily methotrexate (MTX) (1). Treatment with MTX provides satisfactory disease control for some patients with JRA; however, the response to MTX is inadequate in some patients, even at higher dosages, and lack of tolerability limits its usefulness in others (2, 3).
Options for treating JRA were dramatically improved by the introduction of etanercept, a biologic response modifier. Etanercept is a fully human, soluble, tumor necrosis factor (TNF) inhibitor that binds soluble TNF, which prevents it from interacting with cell-surface TNF receptors, thereby blocking the physiologic effects that lead to inflammation and joint damage in patients with JRA. Etanercept is currently the only biologic response modifier that has received Food and Drug Administration approval for the indication of polyarticular-course JRA. Etanercept also is efficacious in providing symptomatic relief and functional improvement in other rheumatic diseases, including rheumatoid arthritis (RA), psoriatic arthritis, and ankylosing spondylitis (4–8).
In a previous randomized clinical trial in 69 patients with polyarticular-course JRA, etanercept demonstrated significant clinical benefit and was well tolerated (9). Patients received open-label etanercept for 3 months, and those who responded were randomized to receive either blinded etanercept or placebo until disease flare occurred or 4 months elapsed, whichever happened first. Eighty percent of patients who received etanercept for a total of 7 months had ≥30% improvement from baseline in at least 3 of 6 core response variables used to assess disease activity, with no more than 1 variable worsening by ≥30% (the American College of Rheumatology [ACR] pediatric 30 criteria for improvement ), compared with 35% of patients who received etanercept for 3 months followed by placebo.
Fifty-eight of these patients with JRA chose to enter an ongoing open-label extension study. Results at ∼2 years showed that etanercept continued to be well tolerated and that efficacy was sustained (11). We now report longer-term safety and efficacy results for the patients who remain in the long-term extension, most of whom have received etanercept for at least 4 years.
PATIENTS AND METHODS
Role of the funding source.
Immunex, a wholly owned subsidiary of Amgen, was responsible for data collection and site monitoring. The study design and data analysis were a combined effort of representatives of the Pediatric Rheumatology Collaborative Study Group and Immunex. The named authors wrote the manuscript with assistance from Susan Myers of Amgen. The authors and Amgen agree with the publication and content of this manuscript.
Patients, ages 4–17 years, with active polyarticular-course JRA (systemic, polyarticular, or pauciarticular onset) were enrolled in the original, 2-part efficacy study of etanercept (9). Eligibility criteria have been previously described (9, 11). In summary, all patients had very active disease and had failed MTX therapy at the time of enrollment into the efficacy study.
Patients who were receiving MTX at the screening for the original efficacy study had to undergo a washout of at least 2 weeks before receiving etanercept. From the start of the original efficacy study through year 1 of the extension, the prednisone dosage had to remain stable (the maximum dose permitted was 0.2 mg/kg/day or 10 mg/day, whichever was less). After 1 year of the extension, the use and dosages of JRA medications, including corticosteroids, NSAIDs, and pain medications, could be adjusted without restriction. MTX could be added, if necessary, at a dose of 10–20 mg/m2/week. Intraarticular injections were recorded for the first year of the extension but not thereafter.
This study is an ongoing, multicenter extension to assess the long-term safety and efficacy of etanercept in patients with JRA. Patients received 0.4 mg/kg of open-label etanercept twice weekly (to a maximum 25 mg per dose). Patients were assessed at months 1, 2, and 3, then every 3 months during the first year of the extension, and approximately every 4–6 months during the following years.
The institutional review boards at each site approved the protocols. Each patient's parent or legal guardian provided written informed consent before the start of the efficacy study and the extension.
Analyses of safety included all available data on all patients, from the beginning of the original efficacy study through the database lock in the continuing long-term extension. Serious adverse events (SAEs) were defined as events that were fatal or life threatening, resulted in a persistent or significant disability or incapacity, were a congenital anomaly or birth defect, or required or prolonged inpatient hospitalization. The SAEs were classified using COSTART (Coding Symbols for Thesaurus of Adverse Reaction Terms) terminology. SAE reports were collected during the study and for 30 days after discontinuation of etanercept. Reports of serious infections, defined as infections requiring hospitalization or intravenous antibiotics, were also collected.
For rates per patient-year of SAEs and serious infections, the duration of exposure to etanercept was calculated for all patients from the beginning of the original efficacy study and excluded time off etanercept therapy between the efficacy study and the extension, where applicable.
Efficacy analyses were limited to data on the 58 patients who entered the long-term extension, and results were calculated relative to these patients' baseline values in the original efficacy study. Exposure to etanercept during the efficacy study was not included because it varied (∼3–7 months) depending upon randomization during the efficacy study (9). Only the observed values during the extension were analyzed for efficacy; no imputations or estimation methods were used for missing values. These missing data are primarily a reflection of patient withdrawals over time. In addition, at various time points, data on 1 or 2 patients were missing (data not collected) for various efficacy parameters. In each instance in which this occurred, a footnote is provided and the number of observations listed in the tables is reflective of the actual number of observations for that parameter.
Efficacy was assessed using the ACR Pediatric 30, 50, and 70 criteria (10). Other efficacy assessments included patient's assessment of pain (using a visual analog scale), the articular severity score (sum of severity scores for swelling, tenderness, limitation of motion [LOM], and pain on motion), the duration of morning stiffness (first 2 years only), the number of patients with 0 swollen joints, and the number of patients with 0 tender joints. Joint assessments were performed as follows: 74 joints were assessed for tenderness and/or pain on motion, 71 joints were assessed for LOM, and 66 joints were assessed for swelling.
In Table 1, demographic characteristics, disease history, and previous JRA therapy at the beginning of the original efficacy study are presented for the 58 patients who enrolled in the extension. The mean ± SD duration of JRA at the baseline of the efficacy study for these patients was 5.9 ± 3.2 years (median 5.5 years, range 0.7–12.3 years).
|Baseline of efficacy study|
|Patients who entered the extension study (n = 58)†||Patients in the extension study for ≥4 years (n = 32)|
|Age, mean ± SD years||10.4 ± 3.8||10.6 ± 3.8|
|Female, no. (%)||39 (67)||26 (81)|
|Race, no. (%)|
|White||43 (74)||27 (84)|
|African American||4 (7)||1 (3)|
|Hispanic||9 (16)||2 (6)|
|Other||2 (3)||2 (6)|
|JRA onset type, no. (%)|
|Pauciarticular||5 (9)||2 (6)|
|Polyarticular||34 (59)||24 (75)|
|Systemic||19 (33)||6 (19)|
|Duration of JRA, mean ± SD years||5.9 ± 3.2||5.9 ± 3.2|
|Rheumatoid factor positive, no. (%)||13 (23)||8 (27)|
|Previous MTX, no. (%)||58 (100)||32 (100)|
|Concomitant therapy at enrollment|
|NSAIDs, no. (%)||56 (97)||32 (100)|
|Corticosteroids, no. (%)||22 (38)||13 (41)|
|Steroid dosage, mean mg/day||5.7||5.0|
None of the 58 patients was receiving MTX at the baseline of the extension study. After the first year of treatment in the extension study, MTX could be added to the treatment regimen. At 2 years in the extension study, 8 of 47 patients remaining in the extension (17%) were receiving MTX (mean dose 13.4 mg/week), at 3 years, 10 of 43 patients remaining in the extension (23%) were receiving MTX (mean dose 15.5 mg/week), and at 4 years, 13 of 38 patients remaining in the extension (34%) were receiving MTX (mean dose 15.8 mg/week). Use of other DMARDs during the extension study was minimal (2 patients received leflunomide and 1 patient received hydroxychloroquine).
Twenty-three of 58 patients (40%) were receiving low-dose systemic corticosteroids at the baseline of the extension study. During the extension, 16 of 23 patients (70%) discontinued corticosteroid therapy, and 4 other patients decreased their daily dose to ≤3 mg of prednisone equivalent. Another 11 patients initiated therapy with corticosteroids during the extension; 9 of these 11 patients subsequently were able to discontinue corticosteroid therapy. Overall, 28 of 34 patients (82%) who received corticosteroids at any time during the extension were able to decrease their dosage to ≤5 mg/day of prednisone equivalent. Eleven of 58 patients (19%) received intraarticular injections during the first year of the extension.
Patient retention during etanercept treatment.
At the end of the original efficacy study, 58 patients chose to enter the open-label extension (Figure 1). Thirty-four of the 58 patients (59%) remain in active followup in the extension study. Efficacy data are available for 4 years of etanercept treatment in the extension study for 32 of these 34 patients.
Patient retention and discontinuation because of lack of efficacy were examined by disease onset type. Of the 58 patients who enrolled in the extension study, 34 (59%) had polyarticular onset, 19 (33%) had systemic onset, and 5 (9%) had pauciarticular onset of disease (Table 1). Six patients withdrew from the extension study because of lack of efficacy (Figure 1). Of these 6 patients, 2 had polyarticular onset, 3 had systemic onset, and 1 had pauciarticular onset of disease. Of the 34 patients remaining in the study, 22 (65%) had polyarticular onset, 9 (26%) had systemic onset, and 3 (9%) had pauciarticular onset of disease.
Safety data were calculated for all 69 patients who enrolled in the original efficacy study, from the start of that study through the last patient's visit in the extension study before the data cutoff date for this analysis. The total exposure to etanercept during this time was 225 patient-years. In Table 2, rates of SAEs and serious infections in each year of etanercept treatment are provided, and range from 0 to 0.20 per patient-year. Overall, the rate of SAEs was 0.13 per patient-year.
|Year of etanercept treatment*||Serious adverse events†||Serious infections, events/patient-year‡|
|Patient-years||No. of events||Events/patient-year|
During the original efficacy study and the extension study that followed, 8 patients (12%) had serious infections, for a rate of 0.04 infections per patient-year (Table 2). The exposure-adjusted rates of serious infections did not increase over time with continuing etanercept treatment. Three of the 8 serious infections have been previously described (9, 11): gastrointestinal infection in a 15-year-old boy (which occurred during the efficacy study; the patient remained in the study), aseptic meningitis secondary to an infection with varicella-zoster virus in a 13-year-old boy (who withdrew from the study), and sepsis in an 8-year-old girl (who withdrew from the study). The remaining 5 serious infections are described below (some of which were previously described as SAEs ).
A 7-year-old girl developed a wound infection the day after treatment for a knife cut to her hand for which she received 4 stitches. She was hospitalized for 2 days because of fever and increased swelling and tenderness of the hand and received intravenous antibiotics; etanercept was temporarily discontinued. The blood culture result was negative. She was discharged on a regimen of oral antibiotics, recovered completely, and remained in the study. A 17-year-old girl developed a herpes zoster infection. She was hospitalized and treated with intravenous acyclovir; etanercept was temporarily discontinued. The event resolved and she remained in the study. A 16-year-old boy developed appendicitis and withdrew from the study. A 15-year-old girl developed a postoperative wound infection following chin implant surgery due to temporomandibular arthritis. The implant was removed, the patient recovered completely, and withdrew from the study. Finally, a 14-year-old girl developed a dental abscess that resulted in a tooth extraction. She was hospitalized for 4 days and treated with antibiotics. She recovered completely, and withdrew from the study.
No lupus-like diseases, demyelinating events, malignancies, or deaths occurred.
Efficacy data were assessed for the 58 patients who entered the extension study, and were compared with their baseline values at the beginning of the original efficacy study.
Disease activity at the beginning of the efficacy study and yearly during the open-label extension is shown in Table 3. Patient's assessment of pain, physician's and patient's/parent's global assessment, Childhood Health Assessment Questionnaire scores, joint counts, and the articular severity score improved from the baseline of the efficacy study, and the improvements were sustained through year 4 of treatment with etanercept in the extension.
|Baseline of original study (n = 58)||Years in long-term extension†|
|1 (n = 52)||2 (n = 47)||3 (n = 41)||4 (n = 32)|
|Patient's assessment of pain, 0–10 scale|
|Mean ± SEM||3.7 ± 0.3||1.3 ± 0.2||1.6 ± 0.3||1.7 ± 0.4||1.7 ± 0.4|
|Physician's global assessment, 0–10 scale|
|Mean ± SEM||6.4 ± 0.3||2.4 ± 0.3||2.3 ± 0.3||2.0 ± 0.3||1.9 ± 0.4|
|Patient's/parent's global assessment, 0–10 scale|
|Mean ± SEM||4.8 ± 0.3||2.1 ± 0.3||1.9 ± 0.3||2.1 ± 0.3||2.1 ± 0.4|
|Mean ± SEM||6.7 ± 1.1||2.2 ± 0.5||1.6 ± 0.5||0.7 ± 0.2||0.6 ± 0.2|
|Mean ± SEM||42.4 ± 3.6||23.4 ± 3.2||–||–||–|
|Mean ± SEM||1.5 ± 0.1||0.8 ± 0.1||0.7 ± 0.1||0.6 ± 0.1||0.6 ± 0.1|
|Total joints with active disease, no.|
|Mean ± SEM||30.1 ± 1.9||10.2 ± 1.9||9.3 ± 1.8||6.6 ± 1.5||6.7 ± 1.9|
|LOM + painful/tender joints|
|Mean ± SEM||13.8 ± 1.7||4.9 ± 1.4||3.4 ± 1.2||3.5 ± 1.1||2.4 ± 0.9|
|Articular severity score|
|Mean ± SEM||107 ± 8||49 ± 7||50 ± 8||44 ± 9||34 ± 8|
|Duration of morning stiffness, minutes|
|Mean ± SEM||87 ± 13||20 ± 6||–||–||–|
The proportion of patients demonstrating response according to the ACR Pediatric 30 criteria at yearly time points during the extension are shown in Figure 2. Values shown are relative to the baseline of the efficacy study. ACR Pediatric 30 responses are shown for 2 groups of patients: all patients who entered the extension (n = 58) and those patients with efficacy data who completed 4 years of treatment in the extension (n = 32).
The proportion of patients in whom JRA disease activity measures were normal is shown in Table 4. The time points shown are at the baseline of the original efficacy study and at years 2 and 4 of the extension.
|Disease activity measure||At baseline of efficacy study||At 2 years of long-term extension||At 4 years of long-term extension|
|No. of active joints = 0||0/58 (0)||14/47 (30)||12/32 (38)|
|No. of joints with LOM and pain and/or tenderness = 0||3/58 (5)||26/47 (55)||20/32 (63)|
|No. of joints with LOM = 0||1/58 (2)||7/47 (15)||7/32 (22)|
|Physician global assessment = 0||0/58 (0)||9/47 (19)||9/32 (28)|
|Patient/parent global assessment = 0||2/58 (3)||12/47 (26)||5/31 (16)|
|Patient assessment of pain = 0||4/58 (7)||9/47 (19)||7/32 (22)|
|C-HAQ = 0||2/58 (3)||14/47 (30)||11/32 (34)|
|CRP in normal range||9/58 (16)||30/46 (65)||25/32 (78)|
|ESR in normal range†||19/58 (33)||NA||NA|
This study examined the safety and efficacy after 4 years of treatment with etanercept in patients with polyarticular-course JRA. Approximately half of the patients with JRA who enrolled in the original efficacy study were still receiving open-label etanercept after 4 years in the extension study. The rate of SAEs did not increase with exposure to etanercept; the overall rate of SAEs was 0.13 per patient-year. The observed rate of serious infections (0.04 per patient-year) was less than the rate reported for adult patients with RA (0.096 per patient-year) (12). This cohort of patients with JRA is the first to be treated for a long period with a biologic agent; infection rates in patients with JRA treated with other TNF inhibitors are not yet published.
A comparison of the rate of serious infections obtained during this study with rates reported for healthy children is not possible because rates of infection for children are not expressed in patient-years. In addition, rates of infection in patients with JRA are confounded because many of these patients do not receive routine childhood immunizations (due to immunosuppressive treatments for JRA) and they often receive other medications, such as corticosteroids and MTX, which predispose them to infection.
Patients with JRA showed sustained improvements in disease activity measures in this extension study. ACR Pediatric 30 scores observed at year 1 were sustained through year 4 of treatment with etanercept. The trend toward increased ACR Pediatric 30 scores over time is evident in patients who completed 4 years of etanercept treatment as well as for all patients who entered the extension, although interpretation of the data is limited by the small sample size. The addition of MTX may well have contributed to the sustained improvements seen in these patients. Recent data from the Trial of Etanercept and Methotrexate with Radiographic Patient Outcomes (TEMPO) study demonstrated that efficacy improved when adults with RA were treated with a combination of etanercept and MTX (13).
It is important to note that all patients in this trial had severe disease. In fact, at the baseline of this study, the patients had the longest mean duration of disease and the most severely active disease of any in the trials of biologic agents performed to date in patients with polyarticular-course JRA. In addition, all patients had demonstrated an inadequate response to MTX prior to enrolling in this study. The ability to maintain excellent disease control on a consistent basis for at least 4 years in this severely affected group of JRA patients represents an important clinical advance, even though some patients required the addition of low to moderate dosages of MTX to their treatment regimen.
There is no standardized validated criterion for remission in patients with JRA; however, it is useful to assess the proportion of patients whose disease activity indices returned to normal or 0. Patients who achieve scores of 0 on measures of disease activity have achieved maximal improvement for that measure. In this study, few or no patients had disease activity scores of 0 at baseline, whereas after 4 years of etanercept therapy, 63% had 0 joints with LOM with pain and/or tenderness, and 28% had physician's global assessment scores of 0. Etanercept treatment (with the addition of MTX in some patients) provided significant clinical improvement in patients with JRA, as illustrated by the proportion of patients who achieved scores of 0 in each disease activity measure.
An additional indication of efficacy is the tapering of corticosteroid use. More than 80% of patients who received corticosteroids during the extension were able to discontinue corticosteroids or decrease their dosage to ≤5 mg/day during treatment with etanercept.
Although it is impossible to assess whether some of the efficacy results observed in this study may have been spontaneous improvements in the underlying disease, it is likely that spontaneous improvements were a relatively small proportion of observed improvements, given the persistence of disease in this severely affected cohort of patients.
Interpretation of the long-term efficacy results in this study is subject to the usual limitations of a completer analysis. Comparison of ACR Pediatric 30 scores of all patients who entered the extension with those who completed 4 years of treatment in the extension showed a similar pattern, indicating that the results were not greatly affected by patient dropout.
In this study, etanercept provided sustained clinical benefit when used for up to 4 years in patients with JRA. Although there remains a need to carefully monitor infections in patients with JRA who are receiving etanercept, rates of SAEs and serious infections did not increase over time in this study. Overall, etanercept offers significant clinical benefit with an acceptable safety profile in patients with polyarticular-course JRA.
The authors would like to acknowledge the significant contribution made to this study by the following members of the Pediatric Rheumatology Collaborative Study Group: B. Bernstein, MD, G. D. Cawkwell, MD, PhD, B. Feldman, MD, MSc, FRCPC, B. Gottlieb, MD, MS, Brent Graham, MD, R. Laxer, MD, FRCP, Judyann C. Olson, MD, Murray Passo, MD, A. Reed, MD, B. Shaham, MD, Mandel Sher, MD, D. Sherry, MD, and E. D. Silverman, MD. The authors thank the following study site coordinators for assistance with the study: Nicola J. Bradford, RN, MS, Barbara Feldman, MS, PT, OTR, Karen Felty, RN, Anne Johnson, BS, Norma Liburd, RN, Cathy Mobley, RN, Eileen Pagano, RN, MS, and Jannalee Taylor, RN. The authors thank Susan Myers, MSc, for assistance with writing the manuscript.
- 1Treatment of juvenile rheumatoid arthritis. In: KoopmanWJ, editor. Arthritis and allied conditions: a textbook of rheumatology. 14th ed. Philadelphia: Lippincott Williams & Wilkins; 2001. p. 1294–310., .
- 11Pediatric Rheumatology Collaborative Study Group. Long-term efficacy and safety of etanercept in children with polyarticular-course juvenile rheumatoid arthritis: interim results from an ongoing multicenter, open-label, extended-treatment trial. Arthritis Rheum 2003; 48: 218–26., , , , , , et al, and the
- 13Trial of Etanercept and Methotrexate with Radiographic Patient Outcomes (TEMPO) study investigators. Therapeutic effect of the combination of etanercept and methotrexate compared with each treatment alone in patients with rheumatoid arthritis: double-blind randomised controlled trial. Lancet 2004; 363: 675–81., , , , , , et al, for the