Background Recent studies suggested that in rheumatoid arthritis (RA) patients discontinuing their prior tumor necrosis factor inhibitor (TNFi), treatment persistence was greater in those switching to a disease-modifying antirheumatic drug (DMARD) with a new mechanism-of-action (MOA) than in those cycling to a second TNFi.
Objectives To evaluate in RA patients real-world treatment persistence and clinical outcomes associated with TNFi cycling vs switching to a new MOA DMARD.
Methods This retrospective analysis of RA patients used the US clinical JointMan® database, an evidence-based rheumatology software application that rheumatologists use to systematically document patient-level data in routine clinical practice. Adult RA patients were included if they had a physician prescription record of a TNFi from April 1, 2010 through March 31, 2015 and a subsequent DMARD that was a different TNFi (TNFi cycling cohort) or new MOA DMARD (new MOA DMARD switching cohort). Patients were indexed on the date of the subsequent TNFi or new MOA DMARD from a prior TNFi during the pre-index period. Treatment persistence was assessed, defined as neither discontinuing nor switching the drug post-index before the end of the observational period. Change in Clinical Disease Activity Index (CDAI) at 1 year post-index and its correlation with treatment persistence were evaluated. Kaplan-Meier survival analysis measured time to nonpersistence. Cox proportional hazards regression and ordinary least-squares regression were used to identify independent factors predicting time to nonpersistence and 1-year CDAI change, respectively.
Results Included were 613 patients (mean age 56.5 years; 78.5% women; 82.9% white; baseline CDAI 22.7). After discontinuing a prior TNFi, 54.2% (n=332) of patients cycled to another TNFi and 45.8% (n=281) switched to a new MOA DMARD. During follow-up, 36.7% (n=225) of patients were persistent with their index drug, 34.1% (n=209) switched from their index drug to another drug, and 29.2% (n=179) discontinued their index drug. Compared with TNFi cyclers, new MOA DMARD switchers were more likely to be persistent at 1 year (45.2% vs 29.5%; P<0.0001) and to persist for a longer period of time (Figure 1). After adjusting for covariates, TNFi cyclers were more likely to be nonpersistent than new MOA DMARD switchers (adjusted hazard ratio 1.511; P<0.001). Reduction in CDAI at 1 year was lower in TNFi cyclers than in new MOA DMARD switchers (−4.73 vs −7.64; P=0.027); however, this effect was not statistically significant when baseline CDAI was controlled for in the regression model (−5.83 vs −6.39; P=0.607). Overall, treatment persistence was associated with a significantly higher reduction in CDAI (−8.69 vs −3.16; P<0.001).
Conclusions This real-world study suggests that after discontinuing their prior TNFi, RA patients who switched to a new MOA DMARD demonstrated better treatment persistence than those cycling to another TNFi. Persistence with RA treatment was associated with better clinical outcome.
Acknowledgement Research sponsored by Sanofi and Regeneron Pharmaceuticals, Inc. Editorial assistance provided by I. Ouwehand, PhD, of Excerpta Medica, funded by Sanofi and Regeneron Pharmaceuticals, Inc.
Disclosure of Interest W. Wei Shareholder of: Sanofi, Employee of: Sanofi, K. Knapp Employee of: Arthritis Northwest and Discus Analytics, Inc., L. Wang Consultant for: Sanofi and Regeneron Pharmaceuticals, Inc., Employee of: STATinMED research, C. Chen Shareholder of: Regeneron Pharmaceuticals, Inc., Employee of: Regeneron Pharmaceuticals, Inc., G. Craig Shareholder of: Discus Analytics, Inc., Employee of: Arthritis Northwest and Discus Analytics, Inc., K. Ferguson Shareholder of: Discus Analytics, Inc., Employee of: Arthritis Northwest and Discus Analytics, Inc., S. Schwartzman Shareholder of: Discus Analytics, Inc., Consultant for: Discus Analytics, Inc.