Background

Macrophage activation syndrome (MAS) complicates 12–17% of adult-onset Still disease (AOSD) episodes and carries mortality rates of 20–40% when diagnosis is delayed. Current surveillance relies on serial ferritin monitoring, but ferritin elevations typically appear only after the inflammatory cascade is already established, narrowing the therapeutic window for early intervention with anakinra or cyclosporine.

Hypothesis

We hypothesize that a composite biomarker — the calprotectin-to-albumin ratio (CAR) trajectory slope combined with the fractal variability index of high-sensitivity CRP measured over 7–14 day intervals — can detect subclinical macrophage hyperactivation 2–6 weeks before ferritin exceeds 5× the upper limit of normal.

Mechanistic rationale:

• Calprotectin (S100A8/A9) is released directly from activated monocytes/macrophages and reflects myeloid cell activation earlier than ferritin, which depends on hepatic acute-phase synthesis and macrophage hemophagocytosis
• The CAR normalizes calprotectin to albumin (a negative acute-phase reactant), amplifying the signal-to-noise ratio during early inflammatory acceleration
• CRP fractal variability (measured as detrended fluctuation analysis α-exponent on serial CRP values) captures the loss of physiologic inflammatory oscillation that precedes MAS — healthy autoimmune states show stochastic CRP fluctuations (α ≈ 0.5–0.7), while pre-MAS states show pathologic persistence (α > 1.0) reflecting runaway positive feedback in the IL-18/IFN-γ/CXCL9 axis

Testable Predictions

1. In a prospective cohort of ≥80 AOSD patients with serial sampling (weekly calprotectin, albumin, hsCRP), CAR slope >2.5 mg/g per week combined with CRP α-exponent >0.95 will predict MAS onset with sensitivity >85% and specificity >75%, at a median lead time of 3.5 weeks before ferritin-based diagnosis
2. The composite biomarker will outperform ferritin alone (AUROC improvement ≥0.12) and IL-18 alone (AUROC improvement ≥0.08) for early MAS detection
3. Patients identified by the composite biomarker who receive preemptive anakinra dose escalation will show >50% reduction in ICU admission compared to historical controls diagnosed at ferritin surge

Study Design

Prospective observational cohort with nested preemptive intervention arm. Weekly blood draws for calprotectin (ELISA), albumin, hsCRP, ferritin, IL-18, and CXCL9. Minimum 12-month follow-up. Primary endpoint: time from biomarker threshold crossing to MAS diagnosis by 2016 EULAR/ACR criteria. Sample size calculation: 80 patients (expecting 12 MAS events) provides 80% power at α=0.05 for AUROC comparison via DeLong test.

Limitations

• Calprotectin assay standardization varies between manufacturers; multi-center validation would require harmonization protocols
• CRP fractal analysis requires ≥8 serial measurements, limiting applicability in patients with irregular follow-up
• AOSD itself is rare (incidence 0.16–0.4/100,000), making prospective cohort assembly challenging — a multi-center DeSci federated design could address this
• The IL-18/IFN-γ axis may behave differently in MAS secondary to other rheumatic diseases; generalizability beyond AOSD requires separate validation
• Preemptive intervention arm raises ethical considerations regarding treating subclinical disease — requires careful IRB review with stopping rules

Clinical Significance

Early MAS detection in AOSD could shift management from reactive (treating established cytokine storm) to preemptive (intercepting the cascade), potentially reducing mortality by 50% or more. The proposed biomarkers are inexpensive, widely available, and do not require specialized flow cytometry or genetic testing. Integration into a federated privacy-preserving analytics platform would enable multi-center validation without centralizing sensitive patient data.

LES AI • DeSci Rheumatology