To systematically analyze literature with the aim of examining whether rheumatoid factor (RF) is a predictor of response to tumor necrosis factor (TNF) antagonists in rheumatoid arthritis (RA).

A systematic review and meta-analysis of observational studies were conducted. All studies on the association of baseline RF (titer and/or status) and response to any TNF antagonists, or with enough information to estimate this association were included. Qualitative analysis and meta-analysis using random-effects approach by type of outcome response and RF test was performed. Risk of publication bias was also evaluated.

The systematic review included 18 studies of 4163 identified articles, involving 5703 patients with homogeneous baseline characteristics. The most common outcome to assess response was European League Against Rheumatism (EULAR) response criteria, normally merging good and moderate categories as response. The weighted mean difference (WMD) of baseline IgM RF titer in meta-analysis was higher in the non-responders group [−101.58 (95% CI −156.58,−46.59) I²=0.0]. Combined odds ratios (ORs) of positive IgM RF, positive IgA RF, and positive IgG RF to achieve good/moderate response were 1.08 (0.80, 1.47), I²=40.9%; 0.83 (0.39, 1.73), I²=39.8%, and 1.30 (0.48, 3.51), I²=62.9%, respectively. We did not find an association between a positive IgM RF and EULAR good response or remission.

This meta-analysis does not support baseline IgM RF titer as a predictor of response to TNF antagonists in RA. However, this conclusion is hampered by high heterogeneity in the studies included in this meta-analysis.