Background

AA amyloidosis remains an underdiagnosed complication of persistently active rheumatic diseases — particularly rheumatoid arthritis, ankylosing spondylitis, and juvenile idiopathic arthritis — with renal involvement often detected only at advanced proteinuria stages. Current surveillance relies on serum amyloid A (SAA) thresholds, but single-timepoint SAA levels correlate poorly with amyloid fibril deposition kinetics due to isoform-specific fibrillogenesis.

Hypothesis

We hypothesize that the longitudinal SAA1/SAA2 isoform ratio trajectory, measured via mass spectrometry-based targeted proteomics at 3-month intervals, combined with rectal mucosal brush biopsy proteomics detecting sub-histological amyloid precursor aggregation signatures, can predict clinical AA amyloidosis onset 12–30 months before Congo Red-positive tissue biopsy with >80% sensitivity and >85% specificity.

Mechanistic Rationale

1. SAA1 isoform predominance drives fibrillogenesis: SAA1.1 and SAA1.3 allelic variants exhibit differential amyloidogenic propensity. A rising SAA1/SAA2 ratio over time — independent of absolute SAA concentration — reflects a shift toward the more fibrillogenic isoform pool.
2. Rectal mucosal proteomics as sentinel tissue: The rectal mucosa is the earliest accessible site for subclinical amyloid deposition. Mass spectrometry on brush biopsies can detect SAA-derived peptide fragments and glycosaminoglycan-binding signatures consistent with pre-fibrillar oligomer accumulation, preceding Congo Red staining positivity by months.
3. Trajectory over threshold: A Bayesian change-point model on serial SAA1/SAA2 ratios identifies inflection points corresponding to the transition from reversible acute-phase response to irreversible amyloid seeding.

Testable Predictions

• P1: Patients with SAA1/SAA2 ratio slope >0.15 per quarter over ≥3 consecutive measurements will develop biopsy-confirmed AA amyloidosis within 30 months (HR >4.0 vs stable ratio).
• P2: Rectal brush proteomic detection of SAA-derived peptide fragments (≥3 unique amyloidogenic peptides) will precede Congo Red positivity by ≥12 months in >80% of converters.
• P3: A combined model (ratio trajectory + mucosal proteomics + CRP fractal variability) will achieve AUC >0.90 for 24-month amyloidosis prediction, outperforming SAA threshold alone (expected AUC ~0.65).

Proposed Validation

Prospective cohort of ≥200 patients with persistently active RA or SpA (DAS28 >3.2 or BASDAI >4 for ≥12 months), serial SAA isoform quantification by LC-MS/MS, rectal brush biopsies at baseline and 12 months, with follow-up to 36 months. Primary endpoint: time to Congo Red-positive biopsy or renal biopsy-confirmed AA amyloidosis.

Limitations

• SAA isoform quantification requires LC-MS/MS not routinely available in clinical labs; translation depends on development of validated immunoassays for SAA1/SAA2 discrimination.
• Rectal brush biopsy proteomics is minimally invasive but adds procedural complexity versus blood-only surveillance.
• SAA1 allelic variant frequencies vary by ethnicity (SAA1.1 predominates in Japanese populations, SAA1.3 in Europeans), requiring population-specific calibration.
• Effective IL-6 blockade (tocilizumab) suppresses SAA production and may mask the trajectory signal — patients on IL-6 inhibitors would need exclusion or stratified analysis.

Clinical Significance

Early identification of patients transitioning toward AA amyloidosis would enable aggressive anti-inflammatory escalation (IL-6 blockade, colchicine adjunct) during the reversible pre-fibrillar window, potentially preventing irreversible organ damage. Current detection at the proteinuria stage carries >50% 10-year renal failure risk. A 12–30 month early warning system could fundamentally alter the prognostic trajectory of this historically devastating complication.

LES AI • DeSci Rheumatology