Premium
Noninferiority Trials to Evaluate Drug Effects in Rheumatoid Arthritis
Author(s) -
Rothwell Rebecca,
Nikolov Nikolay P.,
Maynard Janet W.,
Levin Gregory
Publication year - 2020
Publication title -
arthritis and rheumatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 4.106
H-Index - 314
eISSN - 2326-5205
pISSN - 2326-5191
DOI - 10.1002/art.41257
Subject(s) - medicine , rheumatoid arthritis , placebo , randomized controlled trial , physical therapy , clinical trial , rheumatology , margin (machine learning) , clinical endpoint , intensive care medicine , alternative medicine , pathology , machine learning , computer science
Objective The increased availability of highly effective treatments in rheumatoid arthritis ( RA ) necessitates a reexamination of study designs evaluating new treatments. We undertook this study to discuss possible specifications and considerations of noninferiority ( NI ) trials assessing drug effects in RA . Methods We focused on the use of approved tumor necrosis factor inhibitors ( TNF i) as potential active controls and reviewed previous placebo‐controlled studies. We summarized the similarities in baseline characteristics and study design of the historical placebo‐controlled studies used. After performing meta‐analyses to estimate the effects of TNF i on symptoms, physical function, and radiographic progression in RA , we proposed NI margins and evaluated the feasibility of NI trials in this therapeutic setting. Results We determined that an NI trial comparing an experimental treatment to a TNF i using the symptomatic end point of the American College of Rheumatology 20% improvement criteria response can feasibly provide evidence of a treatment effect, with a 12% absolute difference as one possible appropriate NI margin. For change from baseline in the Health Assessment Questionnaire disability index score, reasonable margins range from 0.10 to 0.12. In evaluating radiographic progression, an appropriate margin and the corresponding feasibility of the trial are dependent on the selected active control and the expected variability in progression. Conclusion Active‐controlled studies in RA with justified NI margins can provide persuasive evidence of treatment effects on symptomatic, functional, and radiographic end points. Such studies can also provide reliable, controlled safety data and relevant information for treatment decisions in clinical practice. Thus, we recommend considering NI designs in future clinical trials in RA .