Treatment effectiveness and treatment patterns among rheumatoid arthritis patients after switching from a tumor necrosis factor inhibitor to another medication

Machaon MK Bonafede,1 Jeffrey R Curtis,2 Donna McMorrow,1 Puneet Mahajan,3 Chieh-I Chen4 1Outcomes Research, Truven Health Analytics, Cambridge, MA, 2Division of Clinical Immunology and Rheumatology, University of Alabama at Birmingham, Birmingham, AL, 3Health Economics and Outcomes Research, Sanofi, Bridgewater, NJ, 4Health Economics and Outcomes Research, Regeneron Pharmaceuticals, Inc., Tarrytown, NY, USA Objectives: After treatment failure with a tumor necrosis factor inhibitor (TNFi), patients with rheumatoid arthritis (RA) can switch to another TNFi (TNFi cyclers) or to a targeted disease-modifying antirheumatic drug (DMARD) with a non-TNFi mechanism of action (non-TNFi switchers). This study compared treatment patterns and treatment effectiveness between TNFi cyclers and non-TNFi switchers in patients with RA. Methods: The analysis included a cohort of patients from the Truven Health Analytics ­MarketScan Commercial database with RA who switched from a TNFi (adalimumab, certolizumab pegol, etanercept, golimumab, or infliximab) either to another TNFi or to a non-TNFi targeted DMARD (abatacept, tocilizumab, or tofacitinib) between January 1, 2010 and September 30, 2014. A claims-based algorithm was used to estimate treatment effectiveness based on six criteria (adherence, no dose increase, no new conventional therapy, no switch to another targeted DMARD, no new/increased oral glucocorticoid, and intra-articular injections on <2 days). Results: The cohort included 5,020 TNFi cyclers and 1,925 non-TNFi switchers. Non-TNFi switchers were significantly less likely than TNFi cyclers to switch therapy again within 6 months (13.2% vs 19.5%; P<0.001) or within 12 months (29.7% vs 34.6%; P<0.001) and significantly more likely to be persistent on therapy at 12 months (61.8% vs 58.2%; P<0.001). Non-TNFi switchers were significantly more likely than TNFi cyclers to achieve all six of the claims-based effectiveness algorithm criteria for the 12 months after the initial switch (27% vs 24%; P=0.011). Conclusion: Although the absolute differences were small, these results support switching to a non-TNFi targeted DMARD instead of TNFi cycling when patients with RA require another therapy after TNFi failure. Keywords: rheumatoid arthritis, biologic, switching, tumor necrosis factor inhibito