To determine whether aggressive treatment initiated early in the course of rheumatoid factor (RF)–positive or RF-negative polyarticular juvenile idiopathic arthritis (JIA) can induce clinical inactive disease within 6 months.
To determine whether aggressive treatment initiated early in the course of rheumatoid factor (RF)–positive or RF-negative polyarticular juvenile idiopathic arthritis (JIA) can induce clinical inactive disease within 6 months.
Between May 2007 and October 2010, a multicenter, prospective, randomized, double-blind, placebo-controlled trial of 2 aggressive treatments was conducted in 85 children ages 2–16 years with polyarticular JIA of <12 months' duration. Patients received either methotrexate (MTX) 0.5 mg/kg/week (maximum 40 mg) subcutaneously, etanercept 0.8 mg/kg/week (maximum 50 mg), and prednisolone 0.5 mg/kg/day (maximum 60 mg) tapered to 0 by 17 weeks (arm 1), or MTX (same dosage as arm 1), etanercept placebo, and prednisolone placebo (arm 2). The primary outcome measure was clinical inactive disease at 6 months. An exploratory phase determined the rate of clinical remission on medication (6 months of continuous clinical inactive disease) at 12 months.
By 6 months, clinical inactive disease had been achieved in 17 (40%) of 42 patients in arm 1 and 10 (23%) of 43 patients in arm 2 (χ2 = 2.91, P = 0.088). After 12 months, clinical remission on medication was achieved in 9 patients in arm 1 and 3 patients in arm 2 (P = 0.053). There were no significant interarm differences in adverse events.
Although this study did not meet its primary end point, early aggressive therapy in this cohort of children with recent-onset polyarticular JIA resulted in clinical inactive disease by 6 months and clinical remission on medication within 12 months of treatment in substantial proportions of patients in both arms.
Juvenile idiopathic arthritis (JIA) encompasses a group of diseases of unknown etiology, defined by the International League of Associations for Rheumatology as having in common arthritis in ≥1 joint that persists for ≥6 weeks and begins before 16 years of age, with other conditions excluded (1). With a prevalence of ∼1 per 1,000 children in the US, JIA is the most common pediatric rheumatic illness and the cause of acquired childhood disability (2, 3). During the last 20 years, the advent of a host of immune response modifiers (biologic agents) that directly inhibit the action of proinflammatory mediators has revolutionized the treatment and expected outcome of JIA (4–8) such that extended periods of clinically quiescent disease may now be induced.
Newly published guidelines from the American College of Rheumatology (ACR) provide some guidance for the initiation and safety monitoring of drugs commonly used in JIA, including biologic agents (9). However, it remains unclear as to exactly when in the course of the disease and in what combination these treatments should be started to produce optimal outcomes. At present, it cannot be predicted with confidence which children with JIA will have a less favorable outcome, although some risk factors have been identified (9–11). The polyarticular (both rheumatoid factor [RF]–positive and RF-negative) categories comprise ∼30% of all patients with JIA, and the majority of these children continue to receive combinations of multiple medications for many years (12, 13); disease-free periods without medication of >1 year are uncommon (14).
Investigations in adult rheumatoid arthritis have demonstrated improved outcomes, including less radiographic progression of joint damage, when aggressive treatment is started early in the disease course (15–18). Thus, many rheumatologists now believe there is a “window of opportunity” early in the disease during which aggressive therapy has a profound long-term effect (18–20).
To date, there have not been any randomized, double-blind, placebo-controlled trials of biologic agents in children with recent-onset JIA in which the primary end point is clinical inactive disease (21). The trial described herein was designed to determine whether 2 aggressive treatment regimens started early in the course of polyarticular JIA result in clinical inactive disease within 6 months of initiation. In an exploratory phase, we investigated the potential of the treatments to induce clinical remission on medication (6 continuous months of clinical inactive disease while receiving treatment) within 12 months of initiation.
Patients ages 2–16 years were recruited from 15 sites in the US. All patients were diagnosed as having active juvenile polyarthritis (RF positive or negative) of <12 months in duration (1); however, patients without psoriasis who had a first-degree relative with psoriasis were allowed to enroll. None of the patients had previously received biologic therapy. The only disease-modifying antirheumatic drug (DMARD) allowed was methotrexate (MTX) at a dosage of ≤0.5 mg/kg/week (maximum 40 mg) started no earlier than 6 weeks prior to enrollment. Eligible patients were permitted to have received up to 2 intraarticular corticosteroid injections before or up to 2 weeks after baseline, and oral prednisolone for up to 4 weeks, but must have discontinued corticosteroids at least 1 week prior to enrollment. Patients with past or current JIA-associated uveitis were excluded. Female patients who had reached puberty were tested for pregnancy throughout the study.
This multicenter, randomized, double-blind, placebo-controlled study compared the efficacy of 2 aggressive treatment regimens to induce clinical inactive disease within 6 months of therapy initiation and clinical remission on medication by 12 months. The study consisted of a 6-month pivotal phase and an exploratory phase that lasted up to 12 months after enrollment (Figure 1).
A nonstratified blocked randomization scheme (with a block size of 10 in a 1:1 ratio across all centers) was used to assign patients to 1 of 2 treatment arms. Patients were allocated to study arms at the baseline visit by an electronic data capture system after the screening visit and inclusion/exclusion case report forms were completed. Pharmacists at each institution were given password-protected access to the electronic data capture system, which permitted them to determine to which arm each patient had been assigned.
Treatment arm 1 medications were open-label MTX administered subcutaneously at a dosage of 0.5 mg/kg/week (maximum of 40 mg/week), blinded etanercept administered subcutaneously at a dosage of 0.8 mg/kg/week (maximum of 50 mg/week), and blinded oral prednisolone administered at a dosage of 0.5 mg/kg/day (maximum 60 mg/day) and tapered to 0 over the first 4 months of therapy. Treatment arm 2 medications included open-label MTX administered as in arm 1, blinded etanercept placebo administered subcutaneously every week, and blinded oral prednisolone placebo administered daily and tapered to 0 over 4 months. Etanercept, etanercept placebo, and shipping supplies for these medications were provided by Amgen.
Patients in each arm received 1 mg/day of folic acid and were allowed to receive a single nonsteroidal antiinflammatory drug as concomitant therapy during the study period. Up to 2 intraarticular corticosteroid injections within 2 weeks after the baseline visit were allowed. The use of other antiinflammatory or antirheumatic therapies was not allowed during the 12-month study period.
The exploratory phase lasted up to 12 months after enrollment. Patients in whom end points were not met in the pivotal phase were given open-label treatment arm 1 medications in the exploratory phase. Up to 2 intraarticular corticosteroid injections were allowed in the exploratory phase.
Study visits occurred at screening, at baseline, and at 1, 2, 4, 5, 6, 7, 8, 10, and 12 months. After the baseline visit, the following assessments were performed at every visit: joint examination, conducted in a blinded manner by a trained certified joint assessor (whose only contact with the patient was to conduct the joint assessment) to determine the number of joints with active arthritis and the number of joints with limited range of motion; physician's global assessment of disease activity (on a Likert-like scale of 0–10) assessed by the treating pediatric rheumatologist; parent's global assessment of the patient's overall well-being (on a Likert-like scale of 0–10); Childhood Health Assessment Questionnaire (C-HAQ) (22); evaluation of vital signs; review of concomitant medications; and review of medication diary card for medication compliance. Laboratory tests to assess medication safety and disease activity included complete blood cell count, erythrocyte sedimentation rate (ESR), chemistry panel, urinalysis, and pregnancy test for girls who had reached puberty. Adverse events (AEs) were assessed using the National Cancer Institute's Common Toxicity Criteria version 3 (23), and those of grade 3 or higher were recorded. All infections requiring systemic therapy were documented and followed up until resolution.
The primary efficacy end point in the pivotal phase was the attainment of clinical inactive disease at 6 months. Clinical inactive disease is defined as no joints with active arthritis; no fever, rash, serositis, splenomegaly, or generalized lymphadenopathy attributable to JIA; no active uveitis; ESR in the normal range in the laboratory where tested; and a physician's global assessment of disease activity score of 0 (21). A secondary end point in the pivotal phase was the achievement of at least an ACR Pediatric 70 level of response (24) at the 4-month visit. The ACR Pediatric 70 criteria are defined as ≥70% improvement from baseline in a minimum of 3 of the 6 variables in the ACR Pediatric core set (physician's global assessment of disease activity, parent's assessment of overall well-being, the number of joints with active arthritis, the number of joints with limited range of motion, the C-HAQ score, and the ESR), with no more than 1 of the remaining variables worsening by >30% from baseline (24). Patients in whom ACR Pediatric 70 improvement was not achieved were considered treatment failures in the pivotal phase and were treated with open-label arm 1 medications and placed in the exploratory phase of the trial.
The end point in the exploratory phase was the attainment of clinical remission on medication, defined as 6 continuous months of clinical inactive disease while receiving medication (21). Patients in whom clinical inactive disease was achieved at any time point and who subsequently experienced a disease flare were discontinued from the study. A flare was defined as worsening of ≥3 of any of the 6 ACR core set variables by the following amounts from the previous visit: worsening of physician's global assessment of disease activity by ≥2 units on a Likert scale of 0–10; worsening of parent/patient assessment of overall well-being by ≥2 units on a Likert scale of 0–10; increase of ≥2 in the number of joints with active arthritis; increase of ≥2 in the number of joints with limitation of motion; increase of a minimum of +0.125 in the C-HAQ score; and increase in the ESR from a normal value to an abnormal value.
Safety was assessed in all patients who received ≥1 dose of study medication.
This study was approved by the National Institute of Arthritis and Musculoskeletal and Skin Diseases, by an independent Data and Safety Monitoring Board, and by the institutional review boards of all participating sites and was conducted in compliance with the Declaration of Helsinki. Patients of sufficient age provided assent, and written informed consent was obtained from the patients' parents or legal guardians. The study was conducted between May 2007 and October 2010.
Based on preliminary data from both adult and pediatric studies (4–6), we estimated that clinical inactive disease would be attained in 60% of the patients in arm 1 and 20% of those in arm 2 at the 6-month visit in the pivotal phase. Using an alpha error of 0.05 and a beta error of 0.1, NQuery Advisor version 5.0 yielded a necessary evaluable sample size of 36 patients per arm, or 72 in total. Based on previous trials, we estimated a nonevaluable rate of 15%, thus yielding a needed enrollment of 86 patients.
Baseline demographic and disease characteristics and the frequencies of patients in whom the primary end point was attained were compared between treatment groups by Wilcoxon's rank sum test for continuous numeric or ordinal data, or the likelihood ratio chi-square or Fisher's exact test for nominal data. Logistic regression was used to identify independent variables that were predictive of the outcome in the pivotal phase. The intent-to-treat last observation carried forward approach was used for missing data and for patients who withdrew from the study.
A total of 92 patients were assessed for eligibility, and 87 patients were randomized, 2 of whom were found to be ineligible shortly after enrollment and were excluded from study participation. One patient was excluded prior to receiving study medication, and the other was excluded after receiving a single dose of medication. The baseline demographic and disease characteristics of the 85 patients who received medication are summarized in Table 1. Forty-two patients were randomized to arm 1, and 43 were randomized to arm 2. Sixty-three female patients and 22 male patients with a mean age of 10.5 years and a mean disease duration of 5.1 months at the baseline visit were included. The groups were well balanced except for the mean number of joints with active arthritis and the ESR. Although the mean ESR was significantly higher in arm 2, there was no difference in the proportion of patients with an elevated ESR between the 2 treatment arms.
|Arm 1 (n = 42)||Arm 2 (n = 43)|
|Sex, no. (%)|
|Female||29 (69.0)||34 (79.1)|
|Male||13 (31.0)||9 (20.9)|
|Race, no. (%)|
|White||35 (83.3)||38 (88.4)|
|Black||4 (9.5)||1 (2.3)|
|Other||3 (7.1)||4 (9.3)|
|Age at baseline, years||9.9 ± 4.6||11.1 ± 4.1|
|Duration of symptoms, months||4.9 ± 0.5||5.2 ± 0.6|
|No. of joints with active disease||18.3 ± 11.0||25.5 ± 14.4†|
|No. of joints with limited motion||13.6 ± 11.8||16.3 ± 13.2|
|C-HAQ disability index score||1.1 ± 0.8||1.3 ± 0.7|
|Parent's assessment of well-being||5.6 ± 2.1||5.2 ± 2.8|
|Physician's assessment of disease activity||7.0 ± 1.8||7.1 ± 1.9|
|ESR, mm/hour||29.0 ± 21||44.6 ± 30‡|
|Elevated ESR, no. (%)||20 (47.6)||27 (62.8)|
|Positive for RF, no. (%)||14 (33.3)||17 (39.5)|
|Positive for ANAs, no. (%)||33 (78.5)||25 (59.5)§|
|Positive for anti-CCP, no. (%)||14 (35)¶||16 (39)¶|
|Previous treatment with MTX, no. (%)||6 (14.2)||4 (9.3)|
|Previous treatment with prednisolone, no. (%)||2 (4.7)||2 (4.6)|
Figure 2 shows a flow diagram of patient progress through the phases of the trial. By the 4-month visit, 2 patients in arm 1 had discontinued due to AEs, and the remaining 40 were evaluated. Of these 40 patients, an ACR Pediatric 70 response was achieved in 30, who continued in the double-blind phase; this level of response was not achieved in 10 patients, who were started on open-label medication. Of the 43 patients in arm 2, 3 discontinued due to inadequate response to therapy prior to the 4-month visit. Of the 40 patients evaluated in arm 2, an ACR Pediatric 70 response was achieved in 19; this level of improvement was not achieved in 21 patients, who were started on open-label medications. At the 6-month visit (the end of the pivotal phase of the trial), an additional 2 patients in arm 1 had discontinued the study. Of the 28 remaining patients, clinical inactive disease was achieved in 17. The 11 patients in whom clinical inactive disease was not achieved were entered into the open-label exploratory phase. Among the 19 patients in arm 2 who remained in the double-blind phase after the 4-month visit, clinical inactive disease was achieved in 10 (1 of whom discontinued at the 6-month visit due to fear of needles). Eight of the patients in whom clinical inactive disease was not achieved subsequently entered the open-label exploratory phase.
All 17 patients in arm 1 in whom clinical inactive disease was achieved at 6 months remained in the double-blind phase until the 12-month visit, and all continued to have clinical inactive disease. Clinical remission on medication was attained in 9 of these patients. Of the 9 patients in arm 2 who continued in the double-blind phase after 6 months, 1 discontinued prior to the 12-month visit due to disease flare. The remaining 8 patients were evaluated, and 7 were found to have clinical inactive disease; clinical remission on medication was attained in 3 of them. Figure 2 provides additional details regarding the outcomes in those who entered the open-label exploratory phase.
Table 2 summarizes the outcome data for all patients. An ACR Pediatric 70 response was attained in a significantly higher proportion of patients in arm 1 than patients in arm 2 at the 4-month visit, and, therefore, a significantly higher proportion of patients in arm 1 remained in the double-blind phase. There was no significant difference between arm 1 and arm 2 in the proportion of patients in whom clinical inactive disease was achieved by the 6-month visit. Of those patients in whom an ACR Pediatric 70 response was achieved at 4 months, clinical inactive disease was achieved in 73% in arm 1 and 58% in arm 2 between 7 and 12 months. Of those patients in whom an ACR Pediatric 70 response was not achieved at 4 months, clinical inactive disease was achieved in 50% between 7 and 12 months, but it lasted a median of only 1 month. The difference in the proportion of patients in each arm in whom clinical remission on medication was attained by 12 months approached significance, but the numbers were small (P = 0.053).
|Arm 1 (n = 42)||Arm 2 (n = 43)||χ2||P|
|Met ACR Pedi 70||30 (71)||19 (44)||6.46||0.011|
|Did not meet ACR Pedi 70||12 (28)||24 (55)|
|Clinical inactive disease achieved||17 (40)||10 (23)||2.91||0.088|
|Clinical inactive disease not achieved||25 (60)||33 (77)|
|Month 12/end of study|
|Clinical remission on medication achieved||9 (21)||3 (7)||–||0.053†|
|Clinical remission on medication not achieved||33 (79)||40 (93)|
Figure 3 displays the proportions of patients in each arm in whom ACR Pediatric 70 levels of improvement and clinical inactive disease were achieved at each study visit during the 6-month pivotal phase and at the 12-month/end of study visit. Arm 1 showed a consistently higher proportion of patients whose disease improved as compared to arm 2 across all visits for all levels of improvement, except for clinical inactive disease at the 2-month visit. There were a total of 440 visits (246 in arm 1 and 194 in arm 2), excluding baseline visits, during which assessment for clinical inactive disease was conducted under double-blind conditions. Ninety-four (38%) of the assessments in arm 1 showed the patient to have clinical inactive disease compared to 44 (22%) of the assessments in arm 2. Sixteen (59%) of the 27 patients in whom clinical inactive disease was achieved at 6 months did not maintain clinical inactive disease during the next 6 months, due to disease flare in 3 (all in arm 2) and due to an elevation of the ESR, an increase in the physician's global assessment to 1 (on a 0–10 scale), or the identification of a joint with active disease by blinded joint assessment in 13 (9 in arm 1 and 4 in arm 2).
With the exception of parent's assessment of overall well-being in arm 2, all 6 of the core set variables used to determine the level of ACR improvement showed statistically significant improvements from baseline as early as 4 months after initiation of treatment (data not shown).
Because 2 potential predictors of outcome were significantly higher in arm 2 than in arm 1 at the baseline visit, logistic regression was used to determine if either the number of joints with active arthritis or the ESR influenced the likelihood of attaining clinical inactive disease at the 6-month visit. The resulting odds ratios (ORs) were well below statistical significance (P = 0.745 for ESR; P = 0.378 for number of joints with active disease). Specifically, the ORs were 0.997 (95% confidence interval [95% CI] 0.979–1.016) for the ESR, and 0.982 (95% CI 0.944–1.022) for the number of joints with active disease.
We also used logistic regression to determine if antinuclear antibody, RF, or anti–cyclic citrullinated peptide status or disease duration at the baseline visit was predictive of the primary outcome. Only disease duration was significantly predictive; the shorter the disease duration at baseline, the more likely it was that clinical inactive disease would be achieved at 6 months (OR per month 1.324, P < 0.011). Thus, for each month sooner after disease onset aggressive treatment was started, the patient was 1.324 times more likely to achieve clinical inactive disease at 6 months. RF positivity had no influence on the primary outcome when the 2 arms were combined in the regression analysis. Clinical inactive disease was achieved at 6 months in 9 (29%) of the 31 patients who were RF positive compared to 19 (35%) of the 54 patients who were RF negative (P = 0.56).
There was no difference between treatment arms in the number of patients who received intraarticular corticosteroid injections within 2 weeks of baseline or in the exploratory phase.
A total of 8 patients (2 in arm 1 and 6 in arm 2) discontinued the study due to insufficient therapeutic effect or a disease flare after clinical inactive disease had been achieved.
Three patients experienced serious AEs (SAEs). Two events occurred during open-label dosing. These were pneumonia, which resulted in the patient being withdrawn from the study, and a single psychotic event that resolved with tapering of prednisolone. The latter patient continued in the study. A septic hip joint was identified at the last study visit in a patient who had discontinued open-label therapy 2 months earlier. The hip had not undergone intraarticular steroid injection, and no organism was isolated. All 3 SAEs resolved without sequelae, and none was unexpected.
In addition to the SAEs mentioned above, there were a total of 5 grade 3 (severe) events in 2 patients in arm 1 (1 elevation of transaminase levels and 1 peritonsillar abscess), none of the patients in arm 2, and 3 patients receiving open-label medication (1 low white blood cell count, 1 elevation of transaminase levels, and 1 adjustment reaction). Only the 2 patients with transaminase level elevations were withdrawn from the study.
Four additional patients with AEs that were less severe than grade 3 were withdrawn from the study. These included 1 patient with worsening of a pre-existing, recurrent herpes simplex virus infection, 1 patient with pneumonia, and 2 patients with persistent elevations in transaminase levels, who were withdrawn as per protocol. The recurrent herpes simplex virus infection returned to its prestudy status, while the pneumonia and transaminase level elevations resolved completely.
The rates of infections that required systemic therapy were low and did not differ significantly based on the treatment the patient was receiving when the infection occurred. Eighteen infections occurred during 247 months of MTX monotherapy (0.87/year); 16 occurred during 297 months of MTX and etanercept therapy (0.65/year), and 17 occurred during 360 months of MTX, etanercept, and prednisolone therapy (0.57/year).
Irreparable damage to the joints and surrounding tissue is known to occur early in the course of JIA in perhaps as many as 60% of patients with polyarticular disease (25–27). The rate of radiographic wrist destruction in these children progresses over the time during which the disease remains active, but is most pronounced during the first year following the onset of symptoms (26). Furthermore, clinically important osteoporosis can occur, with some patients failing to recover normal bone mineralization even if arthritis resolves, increasing the bone fracture risk (28–30). These well-known findings underscore the urgency of bringing new-onset disease under control as rapidly as possible.
This study is the first randomized, double-blind, placebo-controlled trial to use aggressive treatment regimens as the initial therapy in polyarticular JIA, with clinical inactive disease as soon as 6 months as the primary outcome measure. Both treatment arms consisted of aggressive treatment: higher-dose subcutaneous MTX in both arms, with the addition of etanercept as a first-line agent plus 4 months of tapered prednisolone in the more aggressive arm. While there was a trend toward a higher rate of induction of clinical inactive disease in the more aggressive treatment arm (etanercept, MTX, and prednisolone), the difference was not statistically significant. Despite this lack of statistical significance, the results presented here have a high degree of clinical relevance for several reasons. Both treatment regimens demonstrated dramatic effectiveness in these severe categories of childhood arthritis. Overall, clinical inactive disease was induced in 32% of patients by 6 months and in 66% by 12 months.
An important finding of this trial is the clinically relevant impact of timing of therapy on the achievement of clinical inactive disease. The chance of achieving clinical inactive disease increased by 1.324 for each month earlier a patient was treated. By comparison, the average time to achievement of clinical inactive disease in the published literature is 13 months (12, 14). Achievement of clinical inactive disease is undoubtedly a high bar to attain. Although a third of the patients in this study achieved clinical inactive disease by 6 months, this state was not continuously maintained in 60% of the patients. This loss of clinical inactive disease was usually due to minimal elevations in the ESR, an increase in the physician's global assessment to 1 on a 10-point scale, or the identification of a joint with active disease by blinded joint assessment.
Comparison of the results of this study to those of most other trials of biologic agents in JIA is difficult for a number of reasons. Other studies included patients who were not treated early in the disease course, included patients whose disease had failed to respond to prior MTX therapy, or were retrospective reports of uncontrolled, nonrandomized treatment (4–8, 12–14). The results of a recent trial of early aggressive therapy in 60 children with polyarticular JIA who were naive for DMARDs, the Aggressive Combination Drug Therapy in Very Early Polyarticular JIA (ACUTE) study (31), were similar to the findings of the present study. The ACUTE study was a 54-week, multicenter, randomized, open-label trial that compared 3 treatments: MTX, MTX with infliximab, and MTX with sulfasalazine and hydroxychloroquine (31). Patients had a mean disease duration of 1.9 months and a mean physician's global assessment of 5.5 (on a visual analog scale of 0–10); 37% were positive for antinuclear antibody and 2% were RF positive. At 6 months, modified clinical inactive disease was achieved in 60% of the patients in the MTX and infliximab arm, 30% in the combination arm, and 5% in the MTX only arm. Although the ACUTE study patient sample, concomitant medications, and protocol design were different from those of the present trial, the results provide additional evidence of the benefit of early aggressive therapy.
In the present trial, achievement of ACR Pediatric 70 improvement by 4 months was an important predictor of clinical inactive disease at 6 months. In 50% of the patients in whom an ACR Pediatric 70 response was not reached at 4 months and who were consequently started on open-label medication, clinical inactive disease was achieved between 7 months and 12 months. However, it was short-lived, lasting a median of only 1 month. This may suggest that the durability of clinical inactive disease is greatest when it is achieved early in the treatment course.
There are several possible reasons why the difference in the proportion of patients in whom clinical inactive disease was attained between the 2 treatment arms was not statistically significant. The initial assumption that the primary end point of clinical inactive disease at 6 months would be achieved in 60% of patients in the more aggressive treatment arm proved too optimistic. This may be due to the high disease activity and high proportion of patients who were RF positive in this trial.
The effectiveness of MTX administered subcutaneously at a dosage of 0.5 mg/kg/week was slightly greater than had been anticipated (23% observed versus 20% expected). The excellent response to subcutaneous MTX suggests a new standard dosage and route for maximal effectiveness of MTX in treating polyarticular JIA.
A crucial observation is that there were few SAEs, and that there were no clinically important differences in grade 3 or higher AEs or infections requiring systemic treatment between the treatment arms. While this study has limited power to detect rare side effects, these treatments appear to have acceptable short-term safety profiles. Studies with larger subject-years of followup, such as a disease-specific, consolidated registry, will provide more valid assessments of safety for these treatment approaches.
This trial has several shortcomings that may have influenced the results. The study was limited to only 2 of the 7 categories of JIA: RF-positive and RF-negative polyarticular JIA. The validity of extrapolating the findings of this study to other JIA categories is unknown. Uveitis was an exclusion criterion for this trial, thus precluding conclusions regarding the therapeutic utility of the treatment regimens studied here on JIA-associated eye disease. While enrolled patients were referred to as newly diagnosed, to make recruitment feasible we permitted the enrollment of patients who had had symptoms for up to 12 months and those who had previously received small doses of MTX and prednisolone. The finding that patients with shorter disease duration at enrollment were more likely to attain clinical inactive disease at 6 months is undoubtedly one of the more clinically important findings of this trial. This suggests, however, that had enrollment been limited to those with disease of shorter duration, the results may have been different.
The ACR recently adopted the Provisional Criteria for Defining Clinical Inactive Disease in Select Categories of JIA (32). The newly adopted criteria contain a slight modification to those used in this study, i.e., the ESR may be elevated if not attributable to JIA, and the duration of morning stiffness cannot exceed 15 minutes. Neither the cause for elevation of the ESR nor the duration of morning stiffness was collected in this trial, and, while the impact on our findings is likely negligible, we are unable to state this with certainty. While MTX monotherapy administered subcutaneously at a dosage of 0.5 mg/kg/week may no longer be considered aggressive therapy by some pediatric rheumatologists, it is still an aggressive approach for many. This dosage and route are not the standard of care for children with polyarticular JIA. Rather than start at a low dosage or start with an oral route of administration and then increase the dosage and/or switch to a subcutaneous route of administration, we chose to use the maximally effective dosage and route of administration of MTX at the beginning of treatment.
This study represents the largest, most homogeneous cohort of children with new-onset polyarticular JIA treated in a blinded, placebo-controlled trial using a standardized protocol. Due to the substantial proportion of patients in whom clinical inactive disease was achieved after only 6 months of aggressive early treatment, and in whom clinical remission on medication was achieved by 12 months, the treatments used in this study may serve as a template for new standards of care for polyarticular JIA. Followup of this cohort is under way to determine the durable, long-term benefits of this treatment approach.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Wallace had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Wallace, Giannini, Lovell.
Acquisition of data. Wallace, Giannini, Spalding, Hashkes, O'Neil, Zeft, Szer, Ringold, Brunner, Schanberg, Sundel, Milojevic, Punaro, Chira, Gottlieb, Higgins, Ilowite, Kimura, Hamilton, Johnson, Lovell.
Analysis and interpretation of data. Wallace, Giannini, Huang, Lovell.
The investigators wish to thank all of the patients and their families for their dedicated participation in this study. The authors thank the following core study research team members: Morty Cohen, Kim Gama, Audrey Hendrickson, Susan Jacob, Elena Mano, Anne Murphy, and Nora Singer. The authors also thank the following site study team members: Stacy Ardoin, Aimee Baker, Shawna Baker, Imelda Balboni, Lilliana Barillas-Arias, Tara Barker, Samantha Bell, Joseph Benavides, Heather Benham, John Bohnsack, Bonny Bowen, Ann Clark, Bronte Clifford, Irene Borras Coughlin, Sonya Crook, Fatma Dedeoglu, Lauren Dickey, Anne Eberhard, Helen Emery, Maria Espinosa, Kimberly Fluker, Jennifer Frankovich, Tracy Fuelling, Robert Fuhlbrigge, Sharon Goodwin, Alisa Gotte, Tom Griffin, Alexei Grom, Kathleen Haines, Heather Hanson, Theresa Harris, Marla Hashiguchi, Kristen Hayward, Michael Henrickson, Shirley Henry, Aimee Hersch, Sarah Holland, Joyce Hsu, Jennifer Huggins, Patricia Irigoyen, Jim Jarvis, Kathleen Kenney-Riley, Susan Kim, Patricia Lee, Tzielan Lee, Suzanne Li, Katherine Madson, Bernadette McNally, Esi Morgan-Dewitt, Jackie Morrill, Su-Ellen Mortland, Daphne Nayar, Marilynn Orlando, Julisa Patel, Karin Peterson, Egla Rabinovich, Jennifer Rammel, Kathy Redmond, Mary Ellen Riordan, Jenny Rossette, Kandice Roush, Ann Rutherford, Christy Sandborg, Robert Sheets, Abi Siva, Erin Smith, Mary Beth Son, Charles Spencer, Ann Stevens, Janalee Taylor, Tracy Ting, Jennifer Turner, Katherine Tuthill, Heather Van Mater, Emily vonScheven, Jennifer Weiss, Jolene Wesley, Katharine Willcoxon, Janet Wootton, and Justine Zasa. The authors would also like to thank Target Health for providing and partially subsidizing the electronic data capture system and electronic database.