Rheumatoid arthritis (RA) is one of the most prevalent chronic inflammatory diseases that primarily affects the joints; however, it should be considered as a systemic syndrome. Therapeutic advances in the past decade have improved outcomes [1]. Despite advances in treatment with biological and targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs), the frequency of remission with a combination of conventional synthetic DMARDs and b/tsDMARDs was 49 % in early RA and 19 % in established RA [2]. This highlights the need for personalized treatment strategies based on biological markers.

Type I interferons (IFNs) play a crucial role in RA pathogenesis. High IFN-α serum levels are associated with higher disease activity and the presence of autoantibodies [3,4]; however, IFN-α positivity is not associated with achievement of CDAI remission at 24 weeks [4]. Regarding associations with treatment, one study reported that a higher type I IFN beta/alpha ratio was associated with response to TNF inhibitors [5], suggesting that the balance between different type I IFNs may be more clinically relevant. Additionally, one study observed a significantly greater decrease in DAS28 in type I IFN low signature patients than in the type I IFN high signature group in patients with RA treated with rituximab [6]. Furthermore, a study has shown that elevated IFN-α levels correlate with therapeutic resistance in early RA, potentially through epigenetic reprogramming mechanisms [7], highlighting their usefulness in predicting outcomes.

Meanwhile, previous investigations have revealed the nuanced and sometimes contradictory effects of type II IFNs in RA. One study observed that plasma IFN-γ was significantly positively correlated with DAS28-ESR [8], suggesting a pro-inflammatory role of IFN-γ. However, the ratio of IFN-γ and interleukin (IL)-4 mRNA expression in whole blood was significantly increased by treatment with infliximab [9]. Animal models have shown divergent findings. A study using a collagen-induced arthritis model demonstrated that IFN-γ triggered arthritis by enhancing local inflammatory processes in joints [10]. Meanwhile, another study in this model showed that IFN-γ deficiency accelerates disease onset and severity [11]. Also, IFN-γ exerts anti-inflammatory, chondroprotective, and anti-osteoclastogenic effects in murine antigen-induced arthritis through a mechanism that involves the regulation of chemokine synthesis and local neutrophil recruitment [12], suggesting regulatory functions of IFN-γ. These investigations highlight that their roles appear contradictory and context-dependent.

The role of IFNs in the production of autoantibodies in the pathogenesis of RA focuses not only on rheumatoid factor (RF) and anti-citrullinated protein antibodies (ACPA), but also on autoantibodies against carbamylated proteins (anti-CarP). ACPA-positive early RA patients showed higher serum IFN-γ levels and upregulated IFN-γ response genes than ACPA-negative early RA patients [13]. In addition, significant correlations between IFN signature genes and the levels of ACPA and anti-CarP in chronic/established RA and early RA have been reported [14]. These reports highlight the importance of IFN in autoantibody production.

Our previous study suggested that the balance between different cytokine networks may be more important than individual cytokine levels in determining treatment response [15]. This shift from single-biomarker approaches to network-based analyses could offer further understanding of the immunological heterogeneity observed in patients with RA. The IFN paradox, in which the same molecular pathways can be both protective and pathogenic depending on the context, exemplifies the challenges in developing biomarker-based treatment strategies. Therefore, we hypothesized that the profiles of IFNs and autoantibodies contribute to the prediction of outcomes in RA. In this study, we analyzed the association of IFN profiles and autoantibodies with multiple cytokine and chemokine levels.