Early rheumatoid factor decline predicts subsequent clinical remission in rheumatoid arthritis: the ANSWER cohort

Abstract Objectives To compare early changes in rheumatoid factor (RF) across classes of biological or targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs) and assess whether early RF decline was associated with 12-month clinical remission in rheumatoid arthritis (RA). Methods We retrospectively analysed 1,423 newly initiated b/tsDMARD treatment courses from the multicentre ANSWER cohort that continued for ≥12 months and had RF measured at baseline and 3 months. Early RF decline was defined as a ≥ 20% reduction at 3 months. Multivariable regression assessed factors associated with RF change and 12-month Clinical Disease Activity Index (CDAI) remission. Twelve-month remission rates were also compared across four categories of RF reduction. Results RF reduction differed across classes, with Janus kinase inhibitors (JAKi) showing the greatest decline. Compared with tumour necrosis factor inhibitors (TNFi), JAKi were associated with a greater RF reduction (β = −14.50, 95% CI − 20.04 to − 8.97; P < 0.001). The direction was similar in multiple sensitivity analyses and remained significant in propensity score-matched analyses. Early RF decline was independently associated with 12-month CDAI remission (adjusted OR 1.39, 95% CI: 1.04–1.88; P = 0.029). A significant overall trend towards higher 12-month remission rates was observed across categories of greater RF reduction (P for trend = 0.004). Conclusions Early RF decline differed across b/tsDMARD classes and was associated with 12-month remission. Longitudinal RF dynamics may provide clinically useful prognostic information in RA.