ObjectivesAnti-citrullinated protein antibodies (ACPAs) form immune complexes with citrullinated proteins binding toll-like receptor (TLR) 4, which has been proposed as a mediator of rheumatoid arthritis (RA). NI-0101 is a first-in-class humanised monoclonal antibody blocking TLR4, as confirmed by inhibition of in vivo lipopolysaccharide-induced cytokine release in healthy volunteers. This study was design to confirm preclinical investigations supporting a biomarker-driven approach for treatment of patients with RA who present positive for these immune complexes.MethodsPlacebo-controlled, double-blind, randomised (2:1) trial of the tolerability and efficacy of NI-0101 (5 mg/kg, every 2 weeks for 12 weeks) versus placebo in ACPA-positive RA patients with inadequate response to methotrexate. Efficacy measures included Disease Activity Score (28-joint count) with C reactive protein (DAS28-CRP), European League Against Rheumatism (EULAR) good and moderate responses, and American College of Rheumatology (ACR) 20, ACR50 and ACR70 responses. Subgroup analyses defined on biomarkers were conducted. Pharmacokinetics, pharmacodynamics and safety were reported.Results90 patients were randomised (NI-0101 (61) and placebo (29)); 86 completed the study. No significant between-group difference was observed for any of the efficacy endpoints. Subgroup analyses using baseline parameters as covariants did not reveal any population responding to NI-0101. Treatment-emergent adverse events occurred in 51.7% of patients who received placebo versus 52.5% for NI-0101.ConclusionsWe demonstrate for the first time that in RA, a human immune-mediated inflammatory disease, blocking the TLR4 pathway alone does not improve disease parameters. Successful targeting of innate immune pathways in RA may require broader and/or earlier inhibitory approaches.