ABSTRACT: INTRODUCTION: This study is based on the results from a Belgian expanded access program in which patients with active refractory and erosive rheumatoid arthritis (RA) were treated with intravenous infusions of infliximab in combination with methotrexate. The objectives of this study were to evaluate the continuation rate of infliximab and its clinical effect over a 7-year period and to document the reasons for discontinuation. METHODS: Between 2000 and 2001, 511 patients with severe and refractory RA were enrolled and treated with infliximab. After 7 years, apart from routine clinical follow-up, treating rheumatologists were asked to complete a questionnaire designed specifically for the present study to evaluate the current therapy with infliximab, the level of disease activity (Disease Activity Score in 28 joints [DAS28]) and the reasons for infliximab discontinuation. RESULTS: After 7 years, 160 of 511 patients (31%) were still on infliximab treatment. The major reasons for infliximab discontinuation included lack of efficacy (104 patients), adverse events (107 patients) and elective change of therapy (70 patients). The majority of cases of treatment discontinuation for safety reasons occurred during the first 2 years. In contrast, discontinuation due to ineffectiveness showed a more constant rate over the 7-year period. Mean DAS for patients still on treatment with infliximab decreased from 5.7 (standard error [SE] 0.1) at baseline to 3.0 (SE 0.1) at year 4 and remained that low until year 7 (3.0 [SE 0.1]). Low disease activity (defined as DAS28 <3.2) was present in 60.9% of patients, and 45.5% achieved remission (DAS28 <2.6). DAS28 at the time of treatment discontinuation due to ineffectiveness decreased over the 7-year period from 5.6 (SE 0.3) in 2001 to 4.8 (SE 0.3) in 2008. CONCLUSIONS: This observational study revealed that patients who continue to receive infliximab experience sustained clinical benefit. The majority of safety issues occurred during the first 2 years of infliximab therapy. We observed that the DAS at the time of therapy discontinuation showed a trend to decrease over time.