Background Observational studies are critical in assessing medication safety and effectiveness. Non-random assignment can potentially lead to channeling bias. Understanding the characteristics of patients receiving similar treatments is important for addressing these biases prior to performing comparative effectiveness analyses. Abatacept (ABA) was one of the first efficacious non-TNF-targeting biologic treatments available. To date, little is known about the patients receiving it.
Objectives To assess characteristics of patients receiving ABA as their first, second, or third or more biologic compared to all other biologics.
Methods Participating RA patients in the National Data Bank for Rheumatic Diseases (NDB) provided treatment and other characteristics (demographic including social, RA-related and comorbidities) through self-reported biannual questionnaires. Only responses from 2005-2012 were used for greater comparability since that is when ABA was available. Initiators of biological therapies were stratified by type of biologic (ABA vs. all others) and by number exposed to previously.
Results A total of 6,682 RA patients initiated at least one new biologic after 2005. A total of 272 received ABA without any prior biologic experience: they were older, had more experience with smoking, diabetes, cancer, hypertension, and other comorbidities, although they used less prednisone in comparison (see Table). Some trends were found in the 1st to 4th biologic used: increasing HAQ, cormorbidities, pain, RA duration, and prednisone use (all p<0.01). ABA patients with no prior biologics were less exposed to MTX. When ABA was the 2nd biologic, patients were most likely to have prior exposure to infliximab; this is in contrast to 1st biologic users who started etanercept and adalimumab before infliximab. When ABA was the 3rd biologic, the anti-TNFs were the most frequent biologic previously administered. Few ABA patients had previous exposure to rituximab compared to other biologics.
Conclusions There appears to be a trend of channeling bias with ABA use in our RA cohort. ABA users were generally older and had more comorbidities including a history of cancer when they had less other biologic experience. Early intravenous ABA use often followed intravenous infliximab use. Careful examination of baseline characteristics will be critical in future comparative effectiveness analyses.
Disclosure of Interest: K. Michaud Grant/research support from: ACR Rheumatology Research Foundation, S. Pedro: None Declared, T. Simon Employee of: Bristol-Myers Squibb, F. Wolfe Grant/research support from: NDB receives grant funding from BMS, Amgen, UCB, Pfizer, and AstraZeneca.