Background SWITCH-RA is a global observational study evaluating the effectiveness of switching to an alternative TNFi or to rituximab (RTX) following initial TNFi failure in patients with RA. This subanalysis assessed the reasons for discontinuation of initial TNFi therapy and factors associated with selection of RTX vs alternative TNFi as subsequent therapy.

Methods Reasons for initial TNFi discontinuation and rationale for selection of subsequent therapy were recorded. Items identified as being most strongly associated with selection of RTX or TNFi as second biologic formed the initial set of explanatory variables in a model that was subsequently refined using stepwise variable selection. Results were analysed using multivariate logistic regression.

Results A total of 1107 patients (mean age 55.5 yrs; mean disease duration 8.3 yrs) were analysed. Reasons for discontinuing initial TNFi were: lack of efficacy (n=824 [74%]); intolerance (n=264 [24%]); other (n=19 [2%]). Patients discontinuing due to inefficacy had lower mean (SD) disease duration (7.5 [6.54] yrs) than those discontinuing due to intolerance (10.5 [8.60] yrs) and for other reasons (9.0 [5.69] yrs). Patients discontinuing due to intolerance had lower mean (SD) duration on initial TNFi (18.1 [23.18] mo) vs patients discontinuing due to lack of efficacy (27.2 [26.13] mo) or other reasons (41.9 [37.35] mo). At time of switch, mean baseline DAS28-ESR scores by patient subgroup were: lack of efficacy 5.1 (1.27); intolerance 4.8 (1.35); other 5.0 (1.23) (p=0.0151 for overall association). Overall, 602 patients (54.4%) received RTX and 505 (45.6%) received alternative TNFi as subsequent therapy. Among those patients who switched due to intolerance, both the mean disease duration (12.7 vs 8.6 yrs; p=0.0082) and mean duration of previous TNFi therapy (22.0 vs 14.9 mo; p=0.0468) were significantly greater for the RTX cohort. Factors most clearly associated with selection of RTX were: no risk of lymphoma (OR=7.583 [95% CI 3.948, 14.566]), low frequency of administration (OR=4.216 [2.908, 6.112]), and low risk of infections (OR=4.168 [2.441, 7.120]). Seropositive (RF and/or anti-CCP) patients were more likely to receive RTX over an alternative TNFi (OR=1.761 [1.301, 2.383]).

Conclusions Lack of efficacy was the main reason for discontinuation of an initial TNFi. No risk of lymphoma, low frequency of administration and low risk of infections were the factors most strongly associated with selection of RTX over an alternative TNFi.

Disclosure of Interest J.-E. Gottenberg Consultant for: Abbott, BMS, Pfizer, Roche, Schering-Plough, C. Mpofu Employee of: F. Hoffmann-La Roche Ltd, W. Bensen Grant/Research support from: Abbott, Amgen, BMS, Janssen, Merck, Lilly, Novartis, Pfizer, Proctor and Gamble, Roche, Sanofi-Aventis, Schering, Takeda, UCB, Warner Chilcott, Wyeth, Consultant for: Abbott, Amgen, BMS, Janssen, Merck, Lilly, Novartis, Pfizer, Proctor and Gamble, Roche, Sanofi-Aventis, Schering, Takeda, UCB, Warner Chilcott, Wyeth, Speakers Bureau: Abbott, Amgen, BMS, Janssen, Merck, Lilly, Novartis, Pfizer, Proctor and Gamble, Roche, Sanofi-Aventis, Schering, Takeda, UCB, Warner Chilcott, Wyeth, A. Rubbert-Roth Consultant for: Roche, Chugai, MSD, UCB, F. Irazoque Consultant for: Pfizer, Roche, Janssen, V. Martínez Taboada Grant/Research support from: Schering-Plough, Wyeth-Pharma, Roche, Consultant for: UCB-Pharma, Brystol-Myers Squibb, Roche, Cellerix, Pfizer, C. Chung Employee of: Genentech, Inc, L. Hinsch-Gylvin Employee of: F. Hoffmann-La Roche Ltd, C. Ferri Employee of: University of Modena & Reggio Emilia, P. Emery Consultant for: Pfizer, Merck, Abbott, BMS, Roche, Novartis