AB0294 PERSISTENCE WITH ABATACEPT VERSUS TUMOR NECROSIS FACTOR-INHIBITORS FOR RHEUMATOID ARTHRITIS COMPLICATED BY POSITIVE ANTI-CYCLIC CITRULLINATED PEPTIDE/RHEUMATOID FACTOR OR OTHER POOR PROGNOSTIC FACTORS

Abstract

Background:Rheumatoid arthritis (RA) treatment usually begins with a non-biologic disease-modifying antirheumatic drug (DMARD), followed by a biologic DMARD (including abatacept or tumor necrosis factor-inhibitors [TNFis]) in non-responsive patients.1Since, treatments are switched if disease activity does not improve, it is valuable to understand treatment persistence and switch patterns in RA patients with poor prognostic factors in a real-world setting.Objectives:To assess 12-month treatment persistence in early-line abatacept versus TNFi treated patients with RA complicated by poor prognostic factors.Methods:We performed a multi-center retrospective medical record review of adult RA patients with poor prognostic factors treated at 6 United States clinics. Patients were treated with abatacept or TNFi as the first biologic treatment at the clinic. Poor prognostic factors included positive anti-cyclic citrullinated peptide antibodies (ACPA+), positive rheumatoid factor antibodies (RF+), increased C-reactive protein levels, elevated erythrocyte sedimentation rate levels, or presence of joint erosions. TNFis included adalimumab, etanercept, infliximab (and their biosimilars), certolizumab pegol, or golimumab. Data were collected from first biologic treatment for ≥1 year. Patients with Crohn’s disease, ankylosing spondylitis, ulcerative colitis, psoriatic arthritis, or anal fistula were excluded. Demographic, disease, and treatment information (start, stop, reason for discontinuation) was abstracted. Treatment persistence (continuation of index treatment with gap ≤60 days) at 12 months and time to discontinuation were reported. Multivariate logistic and Cox regressions were used to compare 12-month persistence and risk of discontinuation between abatacept and TNFi, controlling for demographic and clinical characteristics (age, sex, Charlson comorbidity index [CCI], RA duration), baseline utilization, and clinic. Findings among a subgroup of ACPA+ and/or RF+ patients are reported.Results:Data on 265 patients (100 abatacept, 165 TNFi) were collected, including 163 ACPA+ and/or RF+ patients (55 abatacept, 108 TNFi). Overall, abatacept patients were older than TNFi patients (67.0 vs. 60.3 years, p<0.001), but there were no statistically significant differences in gender, comorbidities, or duration of treatment at the clinic. At 12 months, 83.0% of abatacept patients were persistent vs. 66.1% of TNFi patients (p=0.003). Persistence was similar among ACPA+ and/or RF+ patients (83.6% vs. 64.8%, p=0.012). Median time to discontinuation was 1,423 days for abatacept vs. 690 days for TNFi (p=0.014) (961 days vs. 581 days among ACPA+ and/or RF+ patients, p=0.048) (Figures 1,2). In the adjusted analysis, risk of all-cause discontinuation was statistically significantly higher among TNFi than abatacept patients (1.7 [95% CI: 1.1-2.6], p=0.012). The odds of TNFi patients being persistent at 12 months was 51% lower than abatacept patients, although not statistically significant (p=0.071). More TNFi than abatacept patients discontinued index treatment due to disease progression (27.3% vs. 12.0%, p=0.003). Adjusted analyses showed that TNFi patients had a statistically significantly higher risk of discontinuing index treatment due to disease progression (3.4 [95% CI: 1.6-7.2], p=0.001).Figure 1.Time to discontinuation of index treatment among all patients (N=265)Conclusion:In a real-world setting, RA patients with ACPA or RF positivity or other poor prognostic factors are less likely to discontinue abatacept compared with TNFi and are more likely to be persistent on their early line treatment. This difference may be explained by the lower proportion of patients discontinuing abatacept due to disease progression.