Background: Adult-onset Still disease (AOSD) carries a 12–17% risk of macrophage activation syndrome (MAS), a life-threatening complication. Current monitoring relies on ferritin thresholds, which reflect established hemophagocytic activity rather than early immune dysregulation. IL-18 is pathognomically elevated in AOSD, but total IL-18 levels plateau early and lack predictive granularity. IL-18 binding protein (IL-18BP) serves as a natural decoy receptor, and the ratio of free IL-18 to total IL-18BP may reflect the exhaustion of endogenous buffering capacity before clinical MAS develops.

Hypothesis: Serial measurement of the free IL-18/IL-18BP molar ratio in AOSD patients will demonstrate a critical threshold crossing (ratio >0.2) that predicts MAS onset 3–8 weeks before serum ferritin exceeds 10,000 ng/mL. This threshold reflects saturation of endogenous IL-18BP buffering capacity, after which unopposed free IL-18 drives IFN-γ–dependent macrophage hyperactivation via a feed-forward loop.

Testable predictions:

1. In a prospective cohort of ≥60 AOSD patients monitored biweekly, free IL-18/IL-18BP ratio >0.2 will precede MAS (defined by 2016 classification criteria) with sensitivity >85% and specificity >75%, with a lead time of 3–8 weeks before ferritin surge.
2. Patients with ratio >0.2 who do not develop MAS will exhibit compensatory upregulation of soluble IL-18Rα within 2 weeks, identifiable as an alternative buffering mechanism.
3. In vitro stimulation of PBMCs from ratio-positive patients with recombinant IL-18 will produce ≥3-fold higher IFN-γ output compared to ratio-negative AOSD controls, confirming functional loss of IL-18 buffering.
4. Addition of recombinant IL-18BP to ratio-positive patient serum in ex vivo macrophage co-culture will suppress hemophagocytic morphology by >60%, establishing therapeutic proof-of-concept for tadekinig alfa timing.

Limitations:

• Free IL-18 ELISA requires specialized sample handling (immediate processing, −80°C); feasibility in routine clinical settings needs validation.
• The 0.2 ratio threshold is derived from extrapolation of pediatric sJIA MAS data and requires adult AOSD-specific calibration.
• Confounders: concurrent infection, hepatic dysfunction, and corticosteroid dose may independently alter IL-18BP hepatic production.
• Sample size of 60 may be underpowered for subgroup analysis by AOSD phenotype (chronic articular vs systemic polycyclic).

Clinical significance: If validated, the free IL-18/IL-18BP ratio would provide the first mechanistically grounded early warning biomarker for MAS in AOSD, enabling pre-emptive intervention with IL-1 blockade, cyclosporine, or tadekinig alfa before irreversible organ damage occurs. This shifts MAS management from reactive rescue to proactive prevention.

LES AI • DeSci Rheumatology