Approximately one-third to half of patients with established RA will experience deterioration in symptom control in the year following discontinuation of disease-modifying anti-rheumatic drugs (DMARDs). Particularly for older DMARDS, such as parenteral gold and anti-malarials, reinstitution has been suggested to be less effective in controlling RA disease activity. Though this phenomenon is not as well studied for methotrexate and unstudied for newer cytokines inhibitors, there is concern that reinstating an effective DMARD after discontinuation may not achieve the previous level of efficacy. For this reason, DMARDs are generally continued through long periods of symptom-free remission. Here, Buch et al (Rheumatology; 43, 2004: 243) examine relapse of RA after discontinuing infliximab and efficacy after restarting the drug.
Methods: 17 of the 24 subjects enrolled in the ATTRACT (Anti-TNF Therapy in RA with Concomitant Therapy) Trial in Leeds, UK who received infliximab were followed into the extension phase of the trial. In the first two years of the trial, patients with RA were randomized to placebo or 3mg/kg or 10mg/kg doses of infliximab administered every 4 or 8 weeks (comprising four infliximab treatment groups). Methotrexate of at least 12.5 mg per week was continued in all groups and was continued into the extension phase after infliximab therapy was discontinued. At 24 months, infliximab therapy was stopped, and subjects were followed for relapse. At relapse, defined as a 20% or greater deterioration in ACR composite score, subjects were restarted on infliximab at 3mg/kg at standard dosing schedule and reassessed at nine months for change in ACR response.
Results: All 17 patients flared after discontinuation of infliximab. Mean time to flare was between 13.5 to 20 weeks. A statistically non-significant trend in greater time to flare was observed in the two groups previously treated with 10mg/kg of infliximab compared with the two groups previously treated with 3mg/kg of infliximab. 14 of the 17 patients were retreated with at least 9 months of 3 mg/kg of infliximab after relapse. 12 of these 14 achieved comparable ACR-(n) responses on retreatment as observed on nave treatment. The remaining two patients had a less efficacious response on retreatment. No infusion reactions or drug-specific toxicities were noted on retreatment.
Conclusion: Continuation of anti-TNF therapy is required to maintain remission in established RA. Reinstituting therapy after discontinuation and relapse is of comparable efficacy to response in nave subjects.
Editorial Comments: This small study addresses important issues that have not been previously investigated in the context of TNF inhibitors, namely whether continuous TNF therapy is required in TNF responsive patients and whether discontinuous TNF therapy can achieve the same level of efficacy as continuous therapy. This question has practical implications owing to the cost and toxicities related to TNF therapy. Unfortunately, this imperfectly designed and underpowered study cannot reliably support the stated conclusions. Importantly, it is impossible to refute an induction effect with the use of TNF inhibition, since no non-TNF treated patients are included in this extension sub-analysis. In other words, among the members of this subgroup, the two years of initial infliximab may have made latter responses better on retreatment than matched subjects that had not received TNF inhibition during that period. Despite this, these preliminary results are worth noting and suggest areas for further investigation.