Real-world evaluation of TNF-inhibitor utilization in rheumatoid arthritis

Abstract

Objectives:To evaluate 12-month treatment patterns, healthcare resource use (HCRU), and costs for patients with rheumatoid arthritis (RA), following initiation of index TNF inhibitors (TNFi) and subsequent biologic DMARDs (bDMARDs).Methods:This was a retrospective cohort analysis of adults with RA newly initiating TNFi in the Truven Marketscan Commercial Claims and Encounters and Medicare Supplemental Databases during 2010-2013. A sub-group of patients who switched to a bDMARD within 12 months post-index and within 180 days of last index TNFi were subsequently evaluated over 12 months. TNFi/bDMARD treatment patterns were characterized as: continuers, no gap >180 days in prescription/administration of index TNFi; discontinuers, gap >180 days; switchers, initiated new bDMARD. Concomitant conventional synthetic DMARD use, co-morbid chronic illnesses, and RA severity were assessed. All-cause/RA-related HCRU and costs were evaluated 12 months post-index.Results:Of 9567 identified patients, 67.2%, 17.3%, and 15.4% were continuers, discontinuers, and switchers, respectively. Switchers had the highest 12-month unadjusted mean all-cause costs of 34,585 vs 33,051 for continuers (p = 0.1158) and 24,915 for discontinuers (p < 0.0001; discontinuers vs continuers, p < 0.0001). RA-related costs comprised 82.8%, 31.4%, and 85.7% of total costs for continuers, discontinuers, and switchers, respectively. Of 764 switchers, 68.2% switched to alternative TNFi (cyclers), the rest to non-TNFi bDMARDs; 36.7% of patients who switched to TNFi switched again (to third-line bDMARD) vs 27.6% (p = 0.0313) of those who switched to non-TNFi bDMARDs. Switchers to non-TNFi bDMARDs had higher mean 12-month all-cause costs of 76,580 compared with 50,689 for switchers to alternative TNFi (p < 0.0001); biologic-Administration visits comprised 78.8% of the greater total RA-related costs of switchers to non-TNFi bDMARDs.Conclusions:Real-world TNFi discontinuation/switching rates correspond to randomized controlled trial non-response rates. TNFi cycling is common and associated with an increased likelihood of switching to third-line bDMARD. Switching to non-TNFi bDMARDs was associated with higher costs, mostly attributed to in-office administrations.