Background/Purpose:  Monotherapy with etanercept (ETN) may be a viable therapeutic option for maintenance of patients with rheumatoid arthritis (RA) who prefer to eliminate potential burdens or side effects of combination therapy.  Our objective was to compare persistence of low disease activity (LDA)/remission in patients with RA who were on combination therapy with ETN and a conventional synthetic disease-modifying antirheumatic drug (csDMARD; most commonly methotrexate) who then either discontinued the csDMARD or continued on combination therapy.

Methods:  This analysis included RA patients from the Corrona registry during 10/1/2001–8/31/2017.  ETN monotherapy (Mono) patients were initially treated with ETN + csDMARD combination therapy, achieved Clinical Disease Activity Index (CDAI) LDA/remission (score ≤10), and discontinued the csDMARD (index visit).  The index visit for the comparator combination therapy (Combo) group (patients who continued on combination therapy) was selected as the date that had a similar time interval from the initiation visit as the Mono group.  Propensity score (PS) matching (1:2 without replacement) was used to ensure balanced groups and included variables selected a priori (baseline CDAI, RA duration, duration in LDA/remission before index) and variables not originally balanced between the Mono and Combo groups.  Cox regression was used to compare persistence in LDA/remission between the groups, overall and at specific time points, adjusted for any covariates that remained imbalanced after PS matching.  Patients were censored if disease activity increased to moderate or severe, ETN was discontinued, or a csDMARD was reinitiated (in patients in the Mono group).

Results: We identified 182 Mono and 403 Combo patients. After matching, 120 Mono and 207 Combo patients were eligible (45 on Mono had >3 years ETN before index date and could not be matched).  After PS matching, characteristics at index visit were similar between groups except prednisone dose (mean [standard deviation] 6.2 [2.6] for Mono, 4.7 [3.4] for Combo; standardized difference 0.488).  Models indicated that a substantial proportion of both Mono and Combo groups remained in LDA/remission through 24 months after the index date (Table).  The overall persistence in LDA/remission between Mono and Combo groups was not statistically different (P = 0.057).

Conclusion: Approximately 3 of 4 patients with RA remained on ETN Mono and maintained LDA/remission for 2 years after csDMARD discontinuation.  Although Combo persistence was numerically greater than Mono, the high level of persistence with ETN Mono following achievement of LDA/remission and discontinuation of csDMARD suggests that ETN Mono may be a viable option for patients who cannot adhere to or tolerate csDMARDs.