Background: The biological mechanisms underlying the differential response to abatacept in patients with rheumatoid arthritis (RA) are unknown. Here, we aimed to identify cellular, transcriptomic, and proteomic features that predict resistance to abatacept in patients with RA. Methods: Blood samples were collected from 22 RA patients treated with abatacept at baseline and after 3 months of treatment. Response to treatment was defined by the European League Against Rheumatism (EULAR) response criteria at 3 months, and seven patients were classified as responders and the others as non-responders. We quantified gene expression levels by RNA sequencing, 67 plasma protein levels, and the expression of surface molecules (CD3, 19, and 56) by flow cytometry. In addition, three gene expression data sets, comprising a total of 27 responders and 50 non-responders, were used to replicate the results. Results: Among the clinical characteristics, the number of monocytes was significantly higher in the non-responders before treatment. Cell type enrichment analysis showed that differentially expressed genes (DEGs) between responders and non-responders were enriched in monocytes. Gene set enrichment analysis, together with single-cell analysis and deconvolution analysis, identified that Toll-like receptor 5 (TLR5) and interleukin-17 receptor A (IL17RA) pathway in monocytes was upregulated in non-responders. Hepatocyte growth factor (HGF) correlated with this signature showed higher concentrations in non-responders before treatment. The DEGs in the replication set were also enriched for the genes expressed in monocytes, not for the TLR5 and IL17RA pathway but for the oxidative phosphorylation (OXPHOS) pathway. Conclusions: Monocyte-derived transcriptomic features before treatment underlie the differences in abatacept efficacy in patients with RA. The pathway activated in monocytes was the TLR5 and IL17RA-HGF signature in the current study, while it was the OXPHOS pathway in the replication set. Elevated levels of HGF before treatment may serve as a potential biomarker for predicting poor responses to abatacept. These findings provide insights into the biological mechanisms of abatacept resistance, contributing valuable evidence for stratifying patients with RA.Keywords: Abatacept; IL17RA; MYD88; Monocyte; Multi-omics; OXPHOS; Rheumatoid arthritis; TLR5.