Background/Purpose: The Childhood Arthritis and Rheumatology Research Alliance (CARRA) Registry began enrolling children with juvenile idiopathic arthritis (JIA) in July 2015. The large number of children with prevalent JIA in the Registry and inclusion of medication logs which document medication use prior to enrollment provide a unique opportunity to describe patterns of medication use among children with JIA.

Methods: Participants were enrolled at >55 centers in the US and Canada.  Children with polyarthritis, including the ILAR categories of extended oligoarticular JIA and polyarticular JIA (RF positive and negative), were included in these analyses if they had >12 months of data available since diagnosis. Data were obtained from Registry medication logs that contained the patient’s medication use history, including start and stop dates. Patient date of diagnosis and medication start and stop dates were imputed if month or day were missing. Longitudinal treatment maps were developed to visualize patient’s individual treatment courses and identify common usage patterns among them. To generate the treatment maps, each patient’s disease duration was divided into 60-day intervals over the first two years since diagnosis. Medication use for each interval was classified by the predominant medication usage pattern during that interval, including non-biologic and/or biologic DMARDs, or no treatment. Patients were sorted by initial and subsequent medication use to determine the most common patterns of medication use.

Results: 853 patients were included, the majority of whom had RF negative polyarticular JIA (71%; Table). 50% of children with RF negative polyarthritis 66% of children with RF positive polyarthritis received a biologic DMARD, most commonly a tumor necrosis factor alpha inhibitor (TNFi). Among children with extended oligoarticular JIA, 35% of children did not receive non-biologic or biologic DMARD treatment. The most common pattern of medication use in this group was methotrexate monotherapy (oral or subcutaneous). Among children with polyarticular JIA (RF positive or negative), the most common sequence was methotrexate monotherapy with addition of biologic DMARD, most commonly TNFi. Less common patterns were biologic DMARD monotherapy and initial therapy with biologic and non-biologic DMARDs. Use of a second or third TNFi and non-TNFi biologics was rare during the first 2 years of treatment.

Conclusion: Although there is variability in treatment approaches to non-systemic polyarthritis in children, patterns of early medication use can be identified by longitudinal mapping of individual patients’ medication data. These data may be used to assess how treatment approaches change over time as new therapies become available.

Table. Patient Characteristics and Medication Use During the First 24 Months After JIA Diagnosis

*IQR: Interquartile range

**Methotrexate, sulfasalazine, leflunomide