Rheumatoid arthritis (RA) is a chronic, inflammatory autoimmune disease (1) (2–4)characterized by joint invasion, predominantly affecting small joints in the distal extremities with symmetric distribution and recurrent episodes (5). In early stages, patients primarily present with redness, swelling, heat, pain, and stiffness in affected joints, while moderate to advanced stages manifest varying degrees of joint deformity accompanied by bone and skeletal muscle atrophy (6). During the chronic phase, synovial hypertrophy and hyperplasia occur within the joint cavity, forming pannus that destroys bone tissue and subsequently causes progressive joint dysfunction (7–9). More than 50% of RA patients develop varying degrees of joint bone destruction within two years of onset, ultimately leading to disability or even death (10). Research indicates that early diagnosis and timely treatment of RA can effectively prevent or significantly alleviate joint damage in 90% of RA patients (11).

Currently, the clinical diagnosis of rheumatoid arthritis primarily relies on the diagnostic criteria jointly established by the American College of Rheumatology (ACR) and the European League Against Rheumatism (EULAR) in 2010 (12), as shown in Table 1. This scoring system evaluates the number of affected joints, detects serological antibodies (RF, anti-CCP antibodies), acute phase reactants (CRP, ESR), and records the duration of synovitis, significantly improving the sensitivity and specificity of RA diagnosis (13–15). However, the sensitivity for early RA diagnosis remains unsatisfactory at approximately 70% (16). Therefore, it is necessary to explore the incorporation of new indicators to meet the diagnostic requirements for early RA.

The 14-3–3 proteins (products of the YWHA gene family) constitute a widely distributed and highly conserved family of acidic eukaryotic proteins with molecular weights of approximately 29–33 kd. First discovered by Moore et al. (17) in 1967 from neuronal proteins extracted from bovine brain tissue, this family comprises seven isoforms: α/β, γ, ϵ, σ, ξ, θ/τ, and η (18). The 14-3–3 proteins interact with over 200 intracellular ligands through “amphipathic grooves” (19, 20), participating in the regulation and control of multiple functions and disease mechanisms, including cell cycle regulation, transcriptional regulation, signal transduction, immune inflammation, and tumorigenesis (21, 22). Notably, the concentration of 14-3-3η protein in synovial fluid is five times higher than in serum (23), suggesting that 14-3-3η protein likely originates from synovial fluid.

Kilani et al. first reported the correlation between 14-3-3η protein and RA in 2007 (24). Their research team found that RA patients had significantly higher levels of 14-3-3η protein in both synovial fluid and serum compared to healthy individuals. This report also indicated a correlation between 14-3-3η protein and matrix metalloproteinases (MMPs). Subsequent studies revealed that as a pro-inflammatory mediator, 14-3-3η protein can participate in the regulation of various biological activities (25). Through MAPK/ERK, SAPK/JNK, and JAK-STAT signaling cascade pathways, it leads to upregulation of interleukin-1β, interleukin-6, tumor necrosis factor-α, joint-degrading factor MMP-9, and receptor activator of nuclear factor κB ligand, thereby promoting inflammatory joint damage. Additional reports indicate that extracellular 14-3-3η protein possesses ligand activity, activating macrophages and utilizing signal cascade effects to upregulate pro-inflammatory factors, effectively inducing inflammatory factor activation and leading to inflammation and even joint destruction (26).

Currently, most diagnostic studies on 14-3-3η protein for RA are based on European and American populations, with limited and conflicting research data available for the Chinese population. The study subjects are predominantly confirmed RA patients with disease duration exceeding one year, while data on early RA patients within one year of onset remain insufficient. This study aims to compare the diagnostic value of 14-3-3η protein with commonly used laboratory RA diagnostic indicators for early RA diagnosis and evaluate its diagnostic performance. This research can further provide clinical research value for the diagnosis and treatment of early RA patients in the Chinese population.

This study demonstrated that the 14-3-3η protein offers significant advantages in sensitivity (88.1%) and negative likelihood ratio (0.14) for early rheumatoid arthritis (RA) diagnosis compared to other biomarkers. Both 14-3-3η and anti-CCP antibody showed higher sensitivity and larger area under the curve (AUC) in both early and established RA groups than other indicators. RF and CRP exhibited high specificity but insufficient sensitivity, resulting in limited diagnostic value based on AUC. Although ESR showed some diagnostic relevance across all metrics, its performance was not prominent. IgM demonstrated low sensitivity and unsatisfactory AUC, rendering it unsuitable as an independent diagnostic marker for RA. After propensity score matching, IgA, IgG, C3, and C4 failed to show consistent statistical significance across both early and established RA groups compared to disease control (DC) and healthy control (HC) groups, and were therefore excluded from further diagnostic evaluation.

Pertsinidou et al. reported that the positivity rates of common diagnostic biomarkers varied significantly across RA patients of different ages and genders (27). Consistent with this, chi-square and ANOVA tests in our study revealed statistically significant differences in age and sex among the four participant groups (P < 0.001). To mitigate potential confounding effects from these variables, we applied propensity score matching to balance baseline characteristics and constructed logistic regression models to adjust for age and sex biases.

Maksymowych et al. reported that at a serum cut-off value of ≥0.19 ng/mL, 14-3-3η protein yielded a sensitivity of 77% and specificity of 93% for established RA (28). When applied to early RA, these values were 64% and 93%, respectively. In our study, both early and established RA groups showed higher sensitivity (88.1% and 88.6%, respectively) but lower specificity (81.9% and 83.4%) compared to those earlier findings. These discrepancies may be attributable to geographic, ethnic, and clinical differences, including treatment history among participants. Variations in assay manufacturers and antibody specifications may also contribute, especially since 14-3-3η has not yet been standardized as a commercial in vitro diagnostic marker.

Naides et al. observed that among 124 RA patients, 15 were anti-CCP negative, and 10 of those (67%) were positive for 14-3-3η, leading the authors to suggest that 14-3-3η may have superior clinical utility over anti-CCP (29). Our ROC analysis also indicated high sensitivity of 14-3-3η in both early and established RA, and the lowest negative likelihood ratio (0.14) in the early RA group, which is consistent with Naides’ findings. However, in our study, anti-CCP antibody demonstrated the highest specificity, Youden index, and positive likelihood ratio in both early RA (96.4%, 0.81, and 21.25, respectively) and established RA (98.2%, 0.87, and 44.50, respectively), along with the lowest negative likelihood ratio (0.11). Although 14-3-3η showed the highest sensitivity in early RA, its specificity and AUC were lower than those of anti-CCP. In established RA, anti-CCP and 14-3-3η had comparable sensitivity, but anti-CCP showed better specificity and higher AUC. Therefore, it remains uncertain whether 14-3-3η offers greater clinical utility than anti-CCP. Nonetheless, 14-3-3η demonstrates excellent sensitivity, especially in early RA.

Maksymowych et al. reported no correlation between 14-3-3η and CRP or ESR (P > 0.05) (30), a finding corroborated in our study for both early and established RA groups. The lack of correlation, combined with high sensitivity and AUC values of 14-3-3η in both RA groups, supports its potential as an independent diagnostic marker for RA.

As a retrospective study, our research is limited by non-randomized case selection, which may introduce selection bias. Future studies should incorporate synovial fluid 14-3-3η levels across different RA stages, along with imaging data such as X-ray of hands and wrists, swollen joint count (SJC28), tender joint count (TJC28), and other clinical indices to further evaluate the diagnostic utility of 14-3-3η throughout the disease course.

Serum 14-3-3η protein shows considerable promise as a novel diagnostic biomarker for RA. It complements existing biomarkers, improves diagnostic accuracy, and facilitates early RA detection, warranting broader clinical attention and application. Further multi-center studies with larger sample sizes are needed to validate its diagnostic utility, accounting for geographic, ethnic, environmental, and therapeutic variables, to scientifically establish its clinical value through large-scale data analysis.