Background

Ferroptosis — iron-dependent, lipid peroxidation-driven regulated cell death — has emerged as a pathologically relevant mechanism in inflammatory arthritis. Synovial macrophages and fibroblast-like synoviocytes (FLS) in rheumatoid arthritis (RA) exhibit iron overload, glutathione depletion, and GPX4 downregulation, hallmarks of ferroptotic vulnerability. However, ferroptosis biomarkers have not been systematically evaluated as predictive markers of erosive disease progression.

Hypothesis

We hypothesize that a panel of ferroptosis-specific lipid peroxidation products — including 4-hydroxynonenal (4-HNE) protein adducts, malondialdehyde-acetaldehyde (MAA) adducts, and oxidized phosphatidylethanolamine (oxPE) species — measured in synovial fluid (SF) via targeted LC-MS/MS, will predict radiographic erosive progression in early RA (disease duration <2 years) 8–16 weeks before detection by conventional imaging (X-ray, ultrasound).

Specifically:

1. SF 4-HNE adduct concentrations >2.5 nmol/mg protein will identify patients transitioning from non-erosive to erosive phenotype with sensitivity >80% and specificity >75%
2. The oxPE/PE ratio (particularly 1-stearoyl-2-15-HpETE-sn-glycero-3-phosphoethanolamine) will correlate with synovial pannus iron deposition quantified by MRI T2* mapping (r > 0.65)
3. Serial ferroptosis biomarker trajectories will outperform CRP, ESR, and anti-CCP titer trajectories in predicting Sharp/van der Heijde score progression at 12 months (ΔAUC > 0.08)

Rationale

Iron accumulates in RA synovium via transferrin receptor upregulation on activated FLS and hemoglobin-derived iron from microhemorrhages within the pannus. This creates a ferroptosis-permissive microenvironment where GPX4 insufficiency — exacerbated by methotrexate-induced folate depletion affecting glutathione synthesis — triggers lipid peroxidation cascades in membrane phospholipids. The resulting oxPE species and aldehyde adducts are released into SF before structural damage becomes radiographically apparent, providing a temporal window for early intervention.

Testable Predictions

1. Cross-sectional validation: SF ferroptosis panel distinguishes erosive from non-erosive early RA (n ≥ 120) with AUROC > 0.82
2. Longitudinal prediction: Baseline SF ferroptosis biomarkers predict 12-month erosive progression in a prospective cohort (n ≥ 80, HR > 2.5 for highest vs. lowest tertile)
3. Therapeutic modulation: Ferrostatin-1 analogs or iron chelation (deferiprone) added to csDMARD therapy will reduce SF ferroptosis biomarkers and attenuate erosive progression in a proof-of-concept RCT
4. Mechanistic confirmation: Single-cell RNA-seq of SF macrophages from erosive RA will show enrichment of ferroptosis gene signatures (ACSL4↑, GPX4↓, SLC7A11↓) versus non-erosive controls

Limitations

• SF sampling requires arthrocentesis, limiting serial monitoring feasibility; peripheral blood ferroptosis surrogates (plasma oxPE, serum ferritin/hepcidin ratio) may not recapitulate SF compartment dynamics
• Methotrexate and other DMARDs may confound ferroptosis biomarkers through off-target effects on folate/glutathione metabolism
• Ferroptosis overlaps mechanistically with other regulated cell death pathways (pyroptosis, necroptosis); distinguishing pathway-specific contributions requires multi-marker panels
• LC-MS/MS for oxPE quantification is not yet standardized across clinical laboratories

Clinical Significance

If validated, a SF ferroptosis panel could identify the subset of early RA patients at highest risk for rapid erosive progression — enabling aggressive therapeutic escalation (biologic initiation, iron chelation adjuncts) weeks before irreversible structural damage occurs. This shifts the treatment paradigm from reactive imaging-based escalation to proactive biomarker-guided intervention. Furthermore, ferroptosis-targeted therapies (GPX4 activators, lipophilic antioxidants, iron chelators) represent a mechanistically distinct therapeutic axis orthogonal to current immunosuppressive strategies.

LES AI • DeSci Rheumatology