Background/Purpose: Despite new effective therapies for Rheumatoid Arthritis (RA), glucocorticoids (GC) are widely prescribed. It is possible that dose and duration of GC therapy may have an impact on the timing of initiation of biologics. The objective of this study is to describe the GC patterns of therapy in patients with RA and its impact on time to biologic initiation in a real-world setting.

Methods: Biologic-naïve patients with early RA (≤1 year) treated with ≥1 csDMARD, who were enrolled in the Corrona registry after June 2008 and had 24 months (m) of follow-up were included in this study. Prospectively collected data on prednisone use reported by treating rheumatologists – was the primary measure to assess the GC use. GC use in milligrams (mg) was stratified into quartiles (Q): 1≤Q1≤4.8, 4.8<Q2≤5.0, 5.0<Q3≤10.0, 10.0<Q4≤60, based on the average dose until the initiation of a biologic or until 24m after enrollment, which ever occurred first. Impact of average GC dose on time to biologic initiation was examined using cumulative incidence rates for each of the GC dose quartiles. Cox-proportional hazards models were used to evaluate the association between time to biologic initiation, GC use and disease activity as measured by clinical disease activity index (CDAI); mean prednisone dose and CDAI were time-varying covariates.

Results: A total of 1,998 patients were included in the study, there were n=1178 who did not take prednisone and of the remaining 820 who were on GC, n=360 initiated a biologic. Overall, 73% were female with an average age of 57 yrs, about 60% were in moderate/high disease activity (CDAI >10) with a mean disease duration of 0.5 yrs at enrollment.  The average daily dose of GC was 7.0 mg from enrollment to 24m. Patients in the lower quartile (1-4.8 mg) compared to the highest quartile (>10 mg) of average GC use had significantly lower pain levels (27.4 vs 45.2), patient global evaluation of disease activity (23.6 vs 42.5),  better functional disability as measured by modified Health Assessment Questionnaire (0.2 vs 0.5) and spent lower number of days (198 vs 322 days) in moderate/high disease activity, defined as CDAI>10 (from enrollment to 24m) (all results with p<0.001). Patients on lower doses of GC (in the lowest quartile) had significantly longer time to initiation of a biologic compared to those who were in the highest quartile of average GC use (19.2m vs 9.2m respectively, p<0.001) (figure). A higher mean GC dose was significantly associated with a shorter time to biologic initiation after adjusting for time-varying CDAI (for every 5mg increase, HR: 1.13, p<0.0001).

Conclusion: In this large real world study of newly diagnosed RA patients, treatment with GCs was common and associated with time to biologic DMARD initiation.  Patients treated with higher GC doses initiated biologics earlier and spent more time in moderate to severe disease activity prior to biologic initiation compared to those treated with lower GC doses.