Background/Purpose: Biologic therapy has improved RA management and enabled some patients to achieve remission. Many clinicians decrease the biologic dose for patients in low disease activity (LDA) or remission. However, it is unclear which patients may flare and if flare contributes to radiographic progression. We assessed, post hoc, whether subjects who flared had a higher incidence of radiographic progression and compared subjects with and without flares.

Methods: PRESERVE (ClinicalTrials.gov: NCT00565409) was a 2-period trial in subjects with moderate RA despite MTX. Period 1 was open-label, single treatment induction; all subjects received etanercept (ETN) 50mg+MTX weekly (QW) for 36 wks. Subjects in LDA or remission (disease activity score for 28 joints [DAS28] ≤3.2) during wks 12 to 36 continued to Period 2, the randomized, double-blind phase to evaluate maintenance of LDA/remission. Subjects were randomized to receive ETN 50mg+MTX QW, ETN 25mg+MTX QW, or placebo+MTX QW to wk 88, when flare and radiographic progression were evaluated. Flare was defined 2 ways: 1) loss of LDA with/without DAS28 change of 0.6; and 2) relapse (DAS28>5.1 or DAS28>3.2 at ≥2 time points). Radiographic progression was evaluated according to 4 levels of stringency, based on modified total Sharp score (mTSS): 1) minimally clinically important difference (change of 5); 2) smallest detected difference (change of 2.3); 3) mTSS change >0.5; and 4) mTSS change >0.0. Demographics and baseline (BL) disease characteristics were compared for subjects with vs without flare, defined as loss of LDA and DAS28 change of 0.6. Analysis of covariance and chi-square test were used to compare continuous and categorical outcomes, respectively.

Results: Age, race, BMI, and disease duration did not differ significantly for flare vs non-flare subjects, total N=531. BL DAS28 was higher for flare than non-flare: mean (SD) 4.37 (0.45) vs 4.27 (0.45), respectively, p=0.046. Other BL disease characteristics were similar between the groups. When flare was defined as relapse, significantly more flare than non-flare subjects exhibited all 4 degrees of radiographic progression (table). When flare was defined as loss of LDA with/without DAS28 change of 0.6, there was no significant difference in radiographic progression for flare vs non-flare, but numerically more subjects with flare progressed. This was the trend for all treatments; the numbers were too small to analyze. Numerically more placebo subjects progressed, regardless of flare status or progression category (data not shown).

Conclusion: Using relapse as a rigorous definition of flare, radiographic progression occurs in significantly more flare vs non-flare subjects. This demonstrates that radiographic progression is a consequence of flare, especially with biologic withdrawal. Patients should be closely monitored if biologic therapy is dosed down.