Design: Retrospective multicentre cohort study (ANSWER)
Population: 1,423 biologic or targeted synthetic DMARD courses with RF at baseline and 3 months
Primary outcome: Association of early RF decline with 12-month CDAI remission
Effect: Early RF decline linked to remission (adjusted odds ratio 1.39, 95% CI 1.04-1.88); JAK inhibitors greatest decline

Rheumatoid factor is a diagnostic marker in rheumatoid arthritis, but its value for tracking treatment response has been uncertain. This analysis of 1,423 biologic or targeted synthetic DMARD courses in the multicentre ANSWER cohort asked whether an early change predicts outcome.

An early rheumatoid factor decline, a fall of at least 20% by three months, was independently associated with achieving clinical remission at 12 months (adjusted odds ratio 1.39, 95% CI 1.04 to 1.88), and remission rates rose steadily across greater degrees of RF reduction. The size of the early decline differed by drug class, with JAK inhibitors producing the greatest fall, significantly more than TNF inhibitors.

The practical use is as an early read on trajectory. A clear RF drop by three months is a favourable prognostic sign, while its absence may prompt earlier reassessment of the regimen, within a treat-to-target approach rather than as a standalone switch rule.

ANSWER cohort analysis of 1,423 biologic or targeted synthetic DMARD treatment courses.
An early (3-month) rheumatoid factor fall of at least 20% predicted 12-month remission (adjusted odds ratio 1.39).
Remission rates rose with greater degrees of RF reduction.
JAK inhibitors produced the greatest early RF decline, more than TNF inhibitors.
Use an early RF drop as a favourable prognostic sign within treat-to-target, not a standalone switch rule.

Why it matters

It gives an inexpensive early signal of who is responding, potentially sharpening treat-to-target decisions before 12 months.

The statistics, in plain English

An odds ratio of 1.39 is a modest independent association, so RF change adds prognostic information but does not by itself determine outcome; this is retrospective cohort data, so it describes a correlation rather than proving RF monitoring improves care.