Introduction
Rheumatoid arthritis (RA), a chronic autoimmune disorder affecting approximately 0.5–1% of the global population, has witnessed remarkable scientific progress over the past five years (2021–2026). Advances span disease biology—from non-coding genetic architectures to synovial microenvironment characterization—alongside clinical innovations including targeted synthetic disease-modifying antirheumatic drugs (tsDMARDs), precision biomarkers, and evolving treat-to-target strategies. Concurrently, the field confronts substantial challenges: safety signals necessitating risk stratification for Janus kinase (JAK) inhibitors, persistent non-response heterogeneity, implementation barriers for biosimilars, and unmet needs in special populations. This review synthesizes key pathogenesis insights, therapeutic advances, and critical treatment challenges from recent literature, emphasizing evidence-based strategies to optimize RA management.
Recent Advances in Pathogenesis
Genetic and Epigenetic Architecture Beyond HLA
Network-based genome-wide association studies (GWAS) have transcended traditional HLA-centric risk models by identifying 14 statistically significant gene modules comprising 193 non-HLA genes that distinguish anti-cyclic citrullinated peptide antibody-positive/rheumatoid factor-positive (CCP+/RF+) from CCP−/RF+ RA endotypes1. Heritability analysis revealed a 0.31 difference (p=0.03) between these subtypes, with key modules implicating interferon signaling and extracellular matrix remodeling (Module 1), complement system activation featuring membrane attack complex components (Module 5), transforming growth factor-β/bone morphogenetic protein (TGF-β/BMP) pathways (Module 2), and endoplasmic reticulum protein folding machinery (Module 10)1. These modules demonstrated discriminatory power for inadequate responses to methotrexate and tumor necrosis factor inhibitors (TNFi) when validated against synovial tissue single-cell RNA-sequencing (scRNA-seq) data, positioning non-HLA modular biology as a foundation for precision therapeutic selection1.
Functional dissection of noncoding variants using massively parallel reporter assays in activated T helper cells identified enhancer variants whose allelic combinations were exclusive to RA patients, with CRISPR-Cas9 validation confirming enhancer-target gene interactions6. Complementing genetic studies, long noncoding RNA ZNF667-AS1 was shown to regulate JAK/STAT signaling by sponging miR-523-3p, thereby suppressing interleukin-6 (IL-6), IL-17, and tumor necrosis factor-α (TNF-α) production in lipopolysaccharide-stimulated chondrocytes and reducing CD4+ IL-17+ T helper 17 cell generation in vivo7.
Cytokine Networks and JAK/STAT Signaling Defects
Systems immunology profiling across 194 RA patients and 41 healthy controls revealed paradoxical attenuation of cytokine-induced JAK/STAT signaling in multiple immune cell subsets despite chronic inflammation2. Interferon-α (IFNα)→phosphorylated STAT1 (p-STAT1) signaling was significantly reduced (p<0.01) in monocytes, B cells, naïve and memory T cells, with monocyte granulocyte-macrophage colony-stimulating factor (GM-CSF)+IL-2→p-STAT5 signaling showing the most marked decreases (p=2.27×10⁻²¹)2. Higher disease activity correlated inversely with IFNα→p-STAT5 and IL-10→p-STAT1 activation but positively with IL-6→p-STAT1/3 signaling in central memory CD4⁻ T cells2. Critically, effective treatment with methotrexate or TNFi normalized these signaling defects, suggesting aberrant JAK/STAT responsiveness is a reversible consequence of chronic inflammation rather than a fixed immune cell phenotype2.
JAK1 emerged as a convergence point for multiple inflammatory mediators in RA, participating in signaling downstream of IL-6, oncostatin M, TNF (indirectly), and γc-receptor cytokines in synovial fibroblasts, macrophages, and endothelial cells5. Ex vivo studies demonstrated that baricitinib (JAK1/2 inhibitor) effectively suppressed oncostatin M-induced JAK signaling and subsequent IL-6, monocyte chemoattractant protein-1 (MCP-1), and interferon-γ-induced protein 10 (IP-10) expression in RA synovial fibroblasts, though it failed to alter spontaneous cytokine release from intact synovial membrane cells, highlighting stimulus-dependent and cell-type-specific effects8.
Synovial Microenvironment and Mucosal Origins
Spatial transcriptomics of synovial biopsies revealed distinct cellular organization patterns distinguishing seropositive and seronegative RA25. Seropositive RA tertiary lymphoid organ (TLO) structures exhibited organized B and T cell recruitment via CXCL13/CXCL12/CCL19 chemokine gradients, with macrophage-enriched areas co-localizing with activated CD55+/HLA-DRA high fibroblast populations25. In contrast, seronegative samples showed increased dendritic cell presence and fibronectin/integrin-mediated fibroblast interactions25. These spatial architectures provide a molecular blueprint for understanding endotype-specific treatment responses and synovial tissue heterogeneity25.
The mucosal origins hypothesis gained further support from epidemiologic and mechanistic data linking respiratory exposures, obesity, diet, and microbiome composition to RA initiation at mucosal barriers4. Anti-posttranslationally modified protein antibodies (AMPAs) and abnormal glycosylation emerged as additional biomarkers extending beyond conventional anti-CCP and RF serology4. Functional genomics implicated loss of specific macrophage populations and synovial fibroblast proliferation as critical cellular drivers converging mucosal dysbiosis with joint-specific inflammation4.
Biomarkers and Patient Stratification
Multi-Omics Approaches
Mendelian randomization integrating proteome and metabolome-wide association data identified 32 proteins and 13 metabolites causally associated with RA risk3. Bayesian model averaging prioritized bromodomain-containing protein 2 (BRD2), p38 mitogen-activated protein kinase, and nuclear factor-κB1 (NF-κB1) as top candidate proteins with colocalization evidence supporting shared genetic variants between RA and nine proteins plus one metabolite3. Functional enrichment implicated natural killer cell-mediated cytotoxicity, p38 MAPK cascade, and NF-κB signaling as therapeutic targets validated through differential expression analyses between RA patients and healthy controls3.
High-dimensionality flow cytometry and synovial scRNA-seq demonstrated that ACPA-positive and ACPA-negative RA exhibit distinct immune signatures, with differential B cell, T follicular helper cell profiles in peripheral blood and markedly different CD4 T cell proinflammatory cytokine production (IFN-γ, IL-17, TNF-α) in synovium37. These endotype-specific gene signatures reinforce the need for precision therapeutic selection tailored to serologic and molecular phenotypes37.
Imaging Biomarkers
In a 5-year prospective study of 60 RA patients in remission (Disease Activity Score in 28 joints-erythrocyte sedimentation rate [DAS28-ESR] <2.6), first-year magnetic resonance imaging (MRI) erosion progression independently predicted treatment change (odds ratio [OR] 1.2, 95% confidence interval [CI] 1.0–1.3) and radiographic progression26. Patients with radiographic progression exhibited significantly higher MRI erosion progression (p=0.03) and bone marrow edema (p=0.04), establishing Rheumatoid Arthritis MRI Scoring System (RAMRIS) erosion and edema scores as sensitive biomarkers for identifying at-risk remission patients26.
MRI-detected tenosynovitis demonstrated 85% sensitivity for RA in 1,211 early arthritis patients (88% in ACPA-positive, 82% in ACPA-negative RA), significantly exceeding sensitivity in psoriatic arthritis (65%, p=0.001) and other inflammatory arthropathies32. Intermetatarsal bursitis, present in 69% of early RA patients, independently associated with local synovitis (OR 1.69, 95% CI 1.12–2.57) and tenosynovitis (OR 2.83, 95% CI 1.80–4.44), with size reductions paralleling synovitis improvement during DMARD therapy38. These juxta-articular inflammatory features expand the RAMRIS imaging phenotype and provide treatment-responsive biomarkers.
Therapeutic Advances and Clinical Evidence
JAK Inhibitors: Efficacy and Safety Paradigm Shift
Head-to-head trials established superior efficacy of selective JAK inhibitors over conventional biologics. In SELECT-CHOICE, upadacitinib 15 mg demonstrated significantly greater DAS28-C-reactive protein (CRP) improvement versus abatacept in biologic-experienced patients (change −2.52 vs. −2.00; difference −0.52, 95% CI −0.69 to −0.35, p<0.001), with clinical remission rates of 30.0% versus 13.3% (p<0.001) at week 1212. Filgotinib 200 mg achieved DAS28-CRP <2.6 in 54% versus 46% for adalimumab at week 52 (p=0.024)12.
However, the ORAL Surveillance trial (4,362 RA patients ≥50 years with ≥1 cardiovascular risk factor) fundamentally altered the risk-benefit landscape1213. Tofacitinib failed non-inferiority versus TNFi for major adverse cardiovascular events (MACE: 3.4% vs. 2.5%; hazard ratio [HR] 1.33, 95% CI 0.91–1.94) and cancer excluding non-melanoma skin cancer (4.2% vs. 2.9%; HR 1.48, 95% CI 1.04–2.09), with lung cancer the most common malignancy1213. The U.S. Food and Drug Administration (FDA) issued Boxed Warnings for all JAK inhibitors in September 2021, emphasizing increased risks of serious cardiovascular events, malignancy, blood clots, and death, with recommendations to reserve JAK inhibitors for patients with inadequate response or intolerance to one or more TNFi13.
JAK-STAT pathway dysregulation sustains both inflammatory and thrombotic processes central to cardiovascular disease in RA, manifesting as elevated IL-6/IL-1β/TNF-α, reduced IL-10, and overexpression of prothrombotic proteins including protein kinase Cε on activated platelets14. While JAK inhibitors theoretically reduce cardiovascular events through anti-inflammatory mechanisms, their prothrombotic potential necessitates rigorous clinical monitoring, particularly in high-risk populations14.
Biosimilars and IL-6 Pathway Inhibition
Seven adalimumab biosimilars demonstrated therapeutic equivalence to reference adalimumab across pivotal trials, with American College of Rheumatology 20% improvement (ACR20) response rates of 68.7–82.7% and comparable safety profiles15. Switching studies consistently showed maintained efficacy, safety, and immunogenicity after transition from reference product15. However, none received FDA "interchangeability" designation, constraining automatic pharmacy substitution15. Implementation barriers—pricing, reimbursement policies, and acceptance concerns—limited biosimilar uptake despite lower costs relative to reference products15.
Tocilizumab biosimilars (CT-P47, BAT1806) met equivalence endpoints with DAS28-ESR improvements of −3.01 versus −3.00 for reference tocilizumab, with maintained efficacy through week 5212. Olokizumab, a direct IL-6 inhibitor, demonstrated non-inferiority to adalimumab in the CREDO program, with 96% of completing patients electing long-term continuation12.
Guidelines and Treat-to-Target Strategies
The ACR 2021 guideline comprises 44 recommendations (7 strong, 37 conditional) emphasizing early DMARD initiation with methotrexate as first-line therapy, treat-to-target strategies aiming for remission or low disease activity (ACR-European Alliance of Associations for Rheumatology [EULAR] criteria), and monitoring every 1–3 months with treatment adjustment at 3 and 6 months if targets are not achieved2241. EULAR 2022 recommendations converge on these principles, with updated guidance specifying that JAK inhibitors should only be used after consideration of cardiovascular disease and cancer risk factors, glucocorticoids should be tapered and discontinued (not merely dose-reduced), and dose reduction is permissible only after sustained remission is achieved2342.
Treatment Challenges and Unmet Needs
Non-Response Heterogeneity and Failed Novel Mechanisms
Despite therapeutic advances, substantial proportions of patients demonstrate inadequate responses or treatment failure. Multiple novel mechanisms failed to demonstrate efficacy in recent trials. The ContRAst program evaluating otilimab (GM-CSF inhibitor) enrolled 3,851 patients across three trials; while ContRAst-1 and ContRAst-2 met primary ACR20 endpoints in methotrexate-inadequate responders and conventional/biologic DMARD-inadequate responders, ContRAst-3 failed in biologic/JAK-inhibitor-experienced patients, leading to program discontinuation12. Bruton's tyrosine kinase (BTK) inhibitors showed mixed results: evobrutinib failed its primary endpoint with unexpectedly high placebo response (49.5%)12, while fenebrutinib demonstrated dose-dependent efficacy comparable to adalimumab but unclear development status12. IL-1 inhibitor natrunix failed to meet its primary endpoint (ACR50 at week 12), with study conduct irregularities compromising data interpretation12.
Safety, Comorbidities, and Special Populations
Comorbidities—cardiovascular disease, serious infections, lymphomas, and non-melanoma skin cancers—significantly impact RA outcomes and mortality, necessitating therapeutic choice balancing disease control against comorbidity-specific risks16. RA patients demonstrate 1.5–2-fold elevated cardiovascular risk driven by chronic inflammation, dyslipidemia, and endothelial dysfunction; JAK inhibitors carry heightened thrombotic and cardiovascular risk, while TNFi reduce cardiovascular events16.
Elderly RA patients (≥65 years), underrepresented in randomized trials, experience higher serious infection rates, particularly with concurrent glucocorticoids17. Non-TNFi agents (abatacept, tocilizumab) show better retention in elderly cohorts, while JAK inhibitors warrant heightened infection surveillance in this age group17. Mental health outcomes with JAK inhibitor therapy showed clinically meaningful Short Form-36 Mental Component Score improvements (pooled mean reduction 4.95 points, 95% CI 4.41–5.48), though incremental benefit over comparators was modest18.
Access, Cost, and Implementation Barriers
Despite biosimilar availability, real-world effectiveness data remain limited, and uptake has been constrained by non-biological factors including pricing structures, formulary restrictions, and clinician/patient acceptance barriers15. Implementation of precision medicine approaches integrating multi-omics biomarkers, single-cell transcriptomics, and functional enhancer mapping faces standardization and clinical translation challenges19.
Future Directions and Emerging Therapeutics
Selective TYK2 and IL-23 Inhibition
Tyrosine kinase 2 (TYK2), mediating immune responses to IL-12, IL-23, and interferon-α, represents an emerging therapeutic target with potential advantages over pan-JAK inhibition19. Deucravacitinib, an allosteric TYK2 inhibitor, achieved ACR20 responses of 52.9% (6 mg) and 62.7% (12 mg) versus 31.8% placebo in psoriatic arthritis, with notably zero serious adverse events, herpes zoster, opportunistic infections, or MACE—suggesting potential safety advantages over pan-JAK inhibitors20.
IL-23 inhibition. IL-23 inhibitors are effective in psoriatic arthritis and related Th17-driven diseases, but this success has not translated to rheumatoid arthritis. In a randomized RA trial, ustekinumab and guselkumab did not demonstrate sufficient efficacy to establish IL-23 pathway blockade as a viable RA treatment strategy. Accordingly, IL-23 inhibition should be viewed as informative for comparative arthritis biology rather than as an emerging RA therapy at present30.
Tolerance-Inducing Therapeutics
Emerging approaches include tolerogenic vaccines aiming for autoantigen-specific immune tolerance, T cell therapies using polyclonal or chimeric antigen receptor (CAR)-engineered regulatory T cells (Tregs), and IL-2 therapies to expand immunosuppressive Tregs in vivo35. These modalities represent a paradigm shift from broad immunosuppression toward selective suppression of pathogenic autoantigen-specific responses while preserving protective immunity35.
Evidence demonstrates that synovial Tregs in RA exhibit impaired programmed cell death protein 1 (PD-1) signaling compromising their suppressive function, while RA synovial macrophages show reduced PD-1 ligand 1 (PD-L1) expression contributing to Treg dysfunction29. Engaging PD-1 on synovial Tregs restores their ability to suppress inflammatory T cells, identifying PD-1 pathway engagement as a strategy to rejuvenate Treg-mediated immune tolerance29.
Prevention in Preclinical RA
The preclinical RA phase spans from genetic/environmental risk factors through autoimmune processes (ACPA/RF production detectable years before clinical arthritis) to clinically suspect arthralgia without evident synovitis34. The presence of both ACPA and RF increases RA progression risk up to 10-fold34. Combination of symptoms, laboratory biomarkers, and imaging findings provides optimal approaches for selecting target at-risk populations for prevention interventions, establishing rationale for early intervention windows in preclinical RA34.
Conclusion
RA research over the past five years has elucidated non-HLA genetic architectures, cytokine network dysregulation, and synovial microenvironment heterogeneity that distinguish disease endotypes and predict treatment responses. Therapeutic advances include JAK inhibitors with superior efficacy offset by cardiovascular and malignancy risks necessitating careful patient selection, biosimilars expanding access despite implementation barriers. Imaging biomarkers (RAMRIS, tenosynovitis, intermetatarsal bursitis) enable precision monitoring, while multi-omics stratification supports endotype-tailored therapy. Persistent challenges—non-response heterogeneity, failed novel mechanisms, comorbidity management, special population needs, and health equity gaps—underscore the complexity of RA pathophysiology and the necessity for integrated clinical, regulatory, and health systems approaches to optimize outcomes for all patients.