Systematic review and meta-analysis of clinical trials and observational studies based on a sensitive search. Meta-regression was used to explore causes of heterogeneity. Unpublished data of clinical trials provided by the authors were also included.
The electronic search captured 3221 references and 422 meeting abstracts. By hand search, four additional articles were also identified. A total of 23 studies meet the purpose of the study and were included in the review. RF positivity at starting predicts better ACR20 [OR, 1.95 (1.24, 3.08)], ACR50 [OR, 5.38 (2.50, 11.60)] and EULAR response [OR, 3.52 (1.66, 7.45)] in 14 studies with RTX, and better ACR20 [OR, 1.51 (1.21, 1.90)] in 6 studies with TCZ. In 3 studies with ABT, no association was found between response and RF [OR 1.36 (0.97, 1.90)]. No asymmetries in the funnel plots or significant variables were found in the meta-regression.
In RA, RF positivity predicts better response to RTX and TCZ but not to ABT.