Measuring Clinical Response and Remission in Juvenile Idiopathic Arthritis: Measures of Disease Activity

Measures of Disease Activity

The definitions of disease flare, improvement, and remission currently in use for describing disease status in JIA are discussed below. Because the measures of disease flare and improvement are relative measures which cannot be compared between patients, recent work (also discussed below) has examined whether the Disease Activity Score (DAS), a continuous measure of disease activity used in the assessment of adult rheumatoid arthritis, can be applied to JIA.

Improvement

Prior to the development of the pediatric core set and the American College of Rheumatology (ACR) Pediatric 30 response criteria in 1997, there had been no single, uniform definition of improvement for use in clinical trials of JIA.^[10] Previous response criteria focused on single outcome measures, including the percentage improvement in number of active joints, physician preference, and overall improvement in physician global assessment (ranging from any improvement to 30% improvement), which led to difficulty in comparing study outcomes.^[11] While the ACR Pediatric 30 was initially designed to distinguish between active treatment and placebo, it is now used as the primary outcome measure for trials of biologic agents and second line therapies and remains the only prospectively validated measure of disease activity in JIA. Though not formally prospectively evaluated, the ACR Pediatric 20, ACR Pediatric 50, ACR Pediatric 70, and ACR Pediatric 90 measures are also used as outcome measures in pediatric trials, and are defined as 20%, 50%, 70%, 90% improvement respectively in a minimum of three variables in the core set with worsening of one variable by no more than 30% (Table 2).^[12]

While the ACR Pediatric 30 was a significant step towards creating a standardized outcome measure in pediatric rheumatology, these measures primarily assess relative efficacy within the context of clinical trials. These measures are of less utility in quantifying response, tracking patient progress longitudinally, and describing an individual's disease state at a specific moment in time.

The DAS was developed in 1993 as a measure of disease activity in adults with rheumatoid arthritis.^[13] The goal of this project was to provide a standardized, composite outcome measure of disease activity, which would be represented as a continuous score. The DAS equation incorporates the Richie Articular Index (a joint count that grades the tenderness of each joint on a scale of 0-3),^[14] a swollen joint count (based on a 44 joint count), erythrocyte sedimentation rate (ESR), and patient assessment of general health to provide a value between 0 and 10 (Table 3). Following the creation of the DAS, additional equations have been developed, including the DAS28, which incorporates a swollen and tender joint count (based on a 28 joint count) in place of the Richie Articular Index. While the original DAS was developed in a cohort of patients with early rheumatoid arthritis, both the DAS and the DAS28 subsequently have been found to be equally applicable to patients with longer standing rheumatoid arthritis and to perform comparably to the adult ACR improvement criteria in clinical trials.^[15] The DAS has also provided the basis for the European League Against Rheumatism (EULAR) response criteria.^[16] The DAS has the advantages of being both a continuous measure of disease activity over time and an absolute measure of disease state and may also be used for measuring patients' response to treatment, to facilitate the communication of the disease status of an individual patient between providers, and to provide a quality measure of treatment.^[17]

Given the attributes of the DAS, a recent investigation was performed to measure the concordance between the ACR Pediatric 20, ACR Pediatric 30, DAS, and DAS28 in a cohort of 75 children with polyarticular JIA who were receiving methotrexate or a TNFα inhibitor.^[18*] Each patient's DAS score was dichotomized, with patients who met the EULAR criteria for moderate or good response categorized as treatment responders, and the remaining patients categorized as nonresponders. Concordance was then assessed for each combination of outcome measures. With the ACR Pediatric 30 used as the standard for response, the highest concordance (κ = 0.71) was found between the DAS and ACR Pediatric 30. Less good concordance (κ = 0.55) was found between the DAS28 and the ACR Pediatric 30, likely due to incorporation of the 28 joint count, which does not include temporomandibular joints, ankle joints, and foot joints. The agreement between the DAS and DAS28 and the ACR Pediatric 20 was less good (κ = 0.5 and κ = 0.58, respectively). Furthermore, as compared to the ACR Pediatric 30, the DAS showed a sensitivity and specificity of 93% and 80% in differentiating responders from nonresponders, while the DAS28 was found to have a sensitivity of 90% and a specificity of 66%. Overall, these data indicate that the DAS has promise in distinguishing responders and nonresponders in therapeutic trials of polyarticular JIA. It remains to be determined, however, whether the DAS is responsive enough to be used as a continuous measure of disease activity in children and how it will correlate with additional measures of disease activity, including the other ACR response measures and remission. The DAS does not incorporate systemic features and it may not be appropriately sensitive for children with oligoarticular disease, given the small number of active joints. Further research will be needed to determine whether it can be modified for use in these disease subtypes.

Flare

A preliminary definition of disease flare based on items from the pediatric core set was published in 2002 (Table 2).^[19] This definition was derived from data from a cohort of 51 children with polyarticular JIA enrolled in a randomized clinical trial of etanercept. The investigators tested different combinations of items from the pediatric core set and found that 40% worsening in two of six core set items without improvement in more than one core set variable by 30% or more performed well within this study cohort. If the DAS is further validated in pediatric trials, its continuous scoring system may also be a useful measure for defining and quantifying disease flare in children with polyarticular JIA.

Remission

While the primary goal in the management of JIA is the achievement and maintenance of remission, defined as the absence of all active disease, the measurement and comparison of remission rates between study populations has been complicated by the lack of a standardized definition. A consensus-based, preliminary definition of clinical remission for oligoarticular, polyarticular JIA (rheumatoid factor positive and negative), and systemic JIA was developed in 2004 and provides definitions for inactive disease, remission on medication, and remission off of medication (Table 4).^[20]

These definitions were retrospectively applied to a cohort of 437 children with JIA who were treated at centers in the United States or Italy between the years 1980 and 2000 and who were followed-up for a minimum of 4 years.^[21] The investigators reported that 89% of these children achieved one or more episodes of inactive disease during the study period. Thirty-six percent of these episodes of inactive disease lasted 2 years or longer and only 6% lasted for as long as 5 years. Forty-four percent of patients in the cohort experienced one or more episodes of remission off of medication. The median duration of active disease prior to the first episode of inactive disease was 20 months, with children with oligoarticular disease experiencing the shortest median duration of active disease prior to inactive disease and those with rheumatoid factor-positive disease experiencing the longest duration (17 months and 29 months, respectively).

The investigators also calculated the proportion of each patient's total disease course that was spent with active versus inactive disease. For the entire study population, it was determined that the median proportion of inactive disease was 40% of the total disease course. While more than half of children with oligoarticular JIA spent a minimum 58% of their disease course in inactive disease, children with polyarticular disease (rheumatoid factor positive and negative), and systemic JIA spent 66% or more of their disease course with active disease. As expected, patients with rheumatoid factor-positive polyarticular disease spent the smallest proportion of time with inactive disease (a median of 16% of the total disease duration).

Although these definitions provide an important first step towards the standardization of reporting of remission in JIA and the use of remission as an endpoint in therapeutic trials, they have not been prospectively validated. A recent study of adult patients with rheumatoid arthritis^[22*] who were classified as being in remission based on physician exam reported that 96% of these patients had evidence of active synovitis on gadolinium-enhanced MRI and 46% had evidence of bone marrow edema. Similar studies are needed in JIA to determine whether achievement of remission based on these preliminary definitions is indeed representative of inactive disease and correlates with the prevention of additional joint damage, improved quality of life, and improved function.