The current understanding of undifferentiated arthritis (UA) suggests that it represents an early, heterogeneous phase of inflammatory joint disease influenced by genetic susceptibility and environmental triggers, leading to immune dysregulation and loss of tolerance (1). This phase is often marked by the emergence of autoantibodies and can progress to various defined rheumatic conditions, such as rheumatoid arthritis (RA), spondyloarthritis (SpA) or connective tissue diseases (CTD), or may resolve spontaneously (1–3). UA is typically defined as inflammatory arthritis that does not fulfill classification criteria for specific rheumatic diseases, reflecting significant clinical variability and uncertainty in long-term outcomes (2, 3).

In recent years, there has been growing interest in identifying diagnostic and prognostic biomarkers capable of predicting, at early stages, which patients presenting with inflammatory musculoskeletal symptoms (with or without autoantibodies) will progress to established rheumatic diseases. Several serum and tissue biomarkers have been explored in this setting, particularly in early RA. Acute-phase reactants such as C-reactive protein (CRP) and serum amyloid A (SAA), synovial and cartilage-derived molecules including matrix metalloproteinases—most notably MMP-3—and inflammatory proteins such as calprotectin have been associated with synovial inflammation, disease activity, and structural progression in early inflammatory arthritis (4, 5). In addition, a range of autoantibodies beyond rheumatoid factor (RF) and anti–citrullinated protein antibodies (ACPA), including anti-carbamylated protein (anti-CarP) and anti-PAD4 antibodies, as well as cartilage degradation products such as COMP and CTX-II, have shown potential value in predicting progression from UA to RA and long-term radiographic damage (4–6). However, despite encouraging findings in selected cohorts, most of these biomarkers have demonstrated limited incremental value over established clinical, serological, and imaging tools, with inconsistent performance across independent populations and disease stages, supporting the ongoing need to identify novel biomarkers for early disease stratification.

Galectins represent a conserved family of β-galactoside-binding proteins, characterized by their specific recognition of poly-N-acetyllactosamine glycoconjugates through evolutionarily conserved carbohydrate recognition domains (CRDs) (7). This family includes at least 15 mammalian members, each containing one or two ~130-amino-acid CRDs, and exhibits dual immunomodulatory capabilities. By cross-linking glycoconjugates, individual galectins can exert either pro-inflammatory or anti-inflammatory effects, thereby modulating various stages of both acute and chronic inflammatory responses. This functional diversity arises from their broad cellular distribution and their ability to interact with distinct molecular partners within immune signaling pathways (7, 8).

Galectin-1 (Gal-1), a 14 kDa galectin capable of forming homodimers, is prominently expressed in key immune cell populations, including inflammatory macrophages, activated T cells, regulatory T cells, and tolerogenic dendritic cells. Its potent anti-inflammatory effects emerge through the coordinated suppression of both arms of the immune response, which have been extensively characterized in experimental models (9, 10). Thus, it can induce apoptosis in activated lymphocytes (particularly Th1 and Th17 cells), promote Th2 polarization, and inhibit the production of inflammatory mediators such as IL-2, IFN-γ, and TNF-α. Gal-1 has been proposed as a potential diagnostic biomarker for various inflammatory autoimmune diseases, including systemic lupus erythematosus (SLE), RA, Sjögren’s syndrome or systemic sclerosis (11). In the context of RA, elevated circulating levels of Gal-1 have been reported compared to healthy controls (11, 12). This increase may reflect an endogenous attempt to counterbalance ongoing inflammation, given Gal-1’s known immunosuppressive properties (8–10). However, despite its promising profile, further studies are needed to validate its clinical utility. In this regard, this work aims to determine if Gal-1 serum levels can be proposed as a biomarker for early diagnosis in patients with pre-RA/UA, or for monitoring disease activity and treatment response.

This study provides the first longitudinal evaluation of serum Gal-1 levels in patients initially diagnosed with UA over a two-year follow-up. Our results indicate that baseline Gal-1 serum levels did not predict progression from UA to a definite diagnosis (RA, SpA, or CTD), arthritis severity, or treatment requirements. Longitudinal analyses also did not support Gal-1 as a biomarker of disease activity.

Previous studies have highlighted the immunomodulatory and anti-inflammatory properties of Gal-1 in autoimmune diseases, suggesting its potential as a biomarker in RA or SLE (11). While most studies reported elevated serum Gal-1 levels compared to healthy controls in both early (12) and established RA (15–17), some found no significant differences (18–20). In SLE, our recent study demonstrated elevated Gal-1 levels compared with controls (18), whereas no differences were observed in psoriatic arthritis (PsA), consistent with the findings of Bably et al. (21). To our knowledge, no previous studies have assessed Gal-1 modulation across different follow-up stages in an overall UA cohort.

In our cohort, UA patients who subsequently developed a chronic inflammatory rheumatic disease (RA, CTD, or SpA) within two years exhibited similar Gal-1 levels at baseline and throughout follow-up, with no significant variation across diagnostic outcomes. No differences were observed between RA and persistent UA, in line with our previous study in the overall early arthritis population (12), although a tendency toward lower levels was noted in UA, which represented only 34% of the sample. These findings may reflect the very early stage of disease in this population, where inflammatory mechanisms are less active and Gal-1 levels remain relatively low. Supporting this notion, we recently reported that Gal-1 concentrations are influenced by disease duration and may be associated with disease severity in long-term RA patients (18). In that study, patients with immune-mediated inflammatory diseases (IMIDs), including RA, SLE, and PsA with longer disease duration (>5 years) showed significantly higher plasma Gal-1 levels than patients with shorter disease duration.

Nonetheless, UA patients who later evolved to SpA presented slightly lower Gal-1 levels compared to other diseases, although differences were not statistically significant. This observation may reflect distinct immunopathological pathways and differential regulation of Gal-1 in SpA compared with RA or CTD. Consistently, previous studies have reported higher Gal-1 levels in RA than in SpA or PsA, both in serum and synovial fluid (18, 21, 22). The divergent immunoregulatory roles of galectins across IMIDs may underlie these findings. Gal-1 primarily exerts anti-inflammatory effects by inhibiting Th1 and Th17 responses and promoting Treg expansion, but it can also favor Th2 polarization and humoral activation, enhancing B-cell differentiation and antibody production (8–10). This dual behavior may contribute to elevated levels in diseases dominated by adaptive and humoral immune activation, such as established RA or SLE, where Gal-1 may act as a compensatory regulatory molecule. Conversely, the relatively low Gal-1 levels observed in SpA could reflect the predominance of innate and IL-23/IL-17–driven immune pathways (23), in which Gal-1 expression and signaling appear less central (8–10).

In this study, baseline serum Gal-1 levels were not significantly associated with other disease parameters or severity indicators, including seropositivity or the need for intensive treatment. Similarly, other studies have failed to demonstrate a link between Gal-1 levels, radiographic progression, seropositivity, or response to DMARD therapy (16, 19, 22). Gal-1 levels were independent of disease activity over the two-year follow-up, despite improvements in DAS-28 scores, agreeing with previous reports in early and established RA (12, 16, 17, 19, 21) and early SpA (22). In contrast, Méndez-Huergo et al. (15) reported a positive correlation in two independent cohorts of 80 RA patients with long-standing, high-activity disease, consistent with our recent study in IMIDs (18), in which only late-stage patients showed an association between Gal-1 levels and disease activity.

Our study has several limitations. First, this was a single-center study, limiting the generalizability of the results, as the cohort was predominantly white European and patient management may have been influenced by local clinical practices. Second, the limited sample size and uneven distribution of patients across final diagnostic groups may have reduced statistical power. Third, the follow-up period was limited to 24 months, which although generally appropriate for early diagnosis in RA, axSpa or PsA (24–26), may be insufficient to fully capture delayed diagnostic transitions in other inflammatory rheumatic diseases such as peripheral SpA (27) or CTD (28), where prodromal or clinically subtle phases may be prolonged beyond two years. Finally, patient samples were collected many years ago (some over 20 years), which may have impacted Gal-1 detection, as prolonged frozen storage can significantly reduce measurable levels, as previously reported by our group (12). We attempted to mitigate this effect by adjusting for storage duration in our analyses.

Future studies should be conducted in larger, prospective, multicenter cohorts using standardized ELISA techniques to validate these findings. Additionally, the observed trend toward lower Gal-1 levels in SpA patients warrants further investigation in larger samples, as the current cohort may not fully capture potential associations or predictive patterns.

In conclusion, although Gal-1 is involved in immune-regulatory pathways relevant to RA, SpA, and SLE pathogenesis, our findings indicate that it has limited value in predicting which UA patients will progress to established chronic inflammatory rheumatic disease. Baseline and longitudinal Gal-1 levels did not correlate with disease activity, prognosis, or diagnostic outcomes. Notably, a trend toward lower Gal-1 levels was observed in patients evolving to SpA, although the small sample size limits interpretation. These findings suggest variable regulation of Gal-1 across distinct pathogenetic mechanisms rather than reflecting overall inflammatory burden.