Objectives To determine the reasons of methotrexate (MTX) discontinuation, frequency of adverse events (AE) and whether the time in inactive disease before MTX withdrawal disease is associated with the risk of disease flare.

Methods Patients with juvenile idiopathic arthritis (JIA) beginning treatment with MTX were prospectively observed in the national JIA biologic register Biologika in der Kinderrheumatologie/Biologics in Paediatric Rheumatology and its follow-up register Juvenile arthritis Methotrexate/Biologics long-term Observation. Inactive disease was defined by a clinical Juvenile Arthritis Disease Activity Score ≤1, flare after MTX discontinuation by reoccurrence of at least moderate disease activity or restart of treatment with a disease-modifying antirheumatic drug .

Results MTX treatment was initiated in 1514 patients after a mean disease duration of 2.1 years (SD=2.8). 40% of the patients experienced oligoarticular onset of JIA. MTX was discontinued in 982 (64.9%) patients. Ineffectiveness (36.9%) and achieving inactive disease (32.1%) were the most common reasons. Among the latter (n=316), 184 (58.2%) patients experienced a flare on follow-up. The likelihood of a flare was a function of time in inactive disease prior to MTX discontinuation (HR 0.95; 95% CI 0.92 to 0.97). Patients with inactive disease for longer than 12 months had a significantly lower flare rate (58 of 119, 48.7%; HR 0.48; 95% CI 0.34 to 0.69). The most frequently reported AE was MTX intolerance, including nausea, aversion and vomiting, accounting for 441 events (13.0 events/100 exposure years) in 307 (20.3%) patients.

Conclusions Patients who spent at least 12 months in inactive disease before MTX discontinuation had a significantly lower flare rate.