Objective: To evaluate the use of a randomized, double blind, drug withdrawal design as a means to test the efficacy of longterm therapy with antirheumatic drugs.
Methods: We evaluated 286 patients with rheumatoid arthritis (RA) treated with amiprilose hydrochloride for 1-3 years, with response, with or without other antirheumatic therapy, in a double blind, 12 week withdrawal study that compared patients randomized to continue amiprilose therapy vs patients randomized to placebo. The primary efficacy variable was preventing a predefined degree of clinical reactivation, or flare; the statistical tests of success were a difference in the proportion of flares and in the mean time to flare.
Results: Thirty percent of patients taking amiprilose and 43% of placebo patients experienced flare (p = 0.026). Patients taking amiprilose had a longer flare-free interval compared to placebo patients (p = 0.027), with the time to reactivation or flare becoming statistically different 73 days after withdrawal.
Conclusion: Placebo controlled withdrawal designs are useful as evidence to support the longterm effectiveness of therapy in a proportion of patients with RA.