Analysis Identifies Predictors of Disease Flare for Patients With RA in Remission

Among patients with early rheumatoid arthritis (RA) achieving clinical remission, both patient function and MRI-detected bone erosion were independent predictors of disease flare 6 and 12 months after treatment withdrawal.

According to a post hoc analysis of the AVERT study published in Arthritis Research & Therapy, these factors could help identify patients with early RA—defined as active for 2 years or less—who achieved clinical remission as candidates for successful treatment withdrawal.

The 24-month AVERT phase 3b trial was published in 2015, with patient registration beginning in 2010. For the 12-month double-blind treatment period, patients aged 18 and older with early RA were randomized to weekly subcutaneous abatacept (Orencia) 125 mg, methotrexate (MTX), or subcutaneous abatacept 125 mg plus MTX. Patients with low disease activity at month 12 were included in the subsequent 12-month treatment withdrawal period.

Disease flare was defined as a doubling of Tender 28-Joint Count (TJC[28]) and Swollen 28-Joint Count (SJC[28]), increase in Disease Activity Score in 28 joints (DAS28[CRP]) of at least 1.2 times relative to the beginning of withdrawal, or investigator’s judgment of RA flare.

This post hoc analysis included 172 patients who had achieved remission. During the withdrawal period, 100 patients experienced a disease flare by month 6, and 13 more patients experienced a flare by month 12, with a total of 113 (66%) patients experiencing a disease flare within a year of treatment withdrawal.

“Although the tapering or discontinuation of bDMARDs [biologic disease-modifying antirheumatic drugs] is often recommended in patients with sustained remission, complete withdrawal of RA therapy may be possible in some patients without inducing disease flares,” the authors wrote. “Modern imaging techniques, soluble biomarkers, and physician/patient-reported measures offer the potential to predict such flares.”

The analysis also showed that higher Health Assessment Questionnaire–Disability Index (HAQ-DI) and MRI-detected bone erosion, bone edema, synovitis, and combined inflammation scores during withdrawal were potential predictors of flare (P ≤ .01).

In a multivariable analysis, high scores for HAQ-DI and MRI-detected bone erosion were confirmed as independent predictors of flare at 6 and 12 months post-withdrawal (P < .01).

“The ability of MRI to detect subclinical joint inflammation may explain our observation that MRI, but not laboratory measures of disease activity such as CRP or clinical measures such as SJC(28) or TJC(28), predicted risk of flare,” the authors wrote. “As more data on predictors of flare after treatment taper or withdrawal are collected, a combination of clinical and imaging factors may be defined for the accurate identification of patients suitable for treatment withdrawal or those who would be at risk of flare.”

However, the authors noted that the cost of an MRI scan solely to predict a disease flare would need to be justified by potential savings in bDMARD usage and the potential to avoid unnecessary treatment.

At the same time, they also said defined cut-off scores for HAQ-DI and MRI-detected bone erosion could potentially help both clinicians and patients with RA in remission decide when to stop treatment.

Reference

Ahmad HA, Baker JF, Conaghan PG, et al. Prediction of flare following remission and treatment withdrawal in early rheumatoid arthritis: post hoc analysis of a phase IIIb trial with abatacept. Arthritis Res Ther. Published online February 16, 2022. doi:10.1186/s13075-022-02735-8