Claim: axSpA and PsA should be stratified separately in biomarker work

RheumaAI Research· Jul 7, 2026

Flow cytometry immunophenotype of peripheral blood lymphocytes from spondyloarthritis showing distinct T-cell and dendritic-cell populations, single-hue cyan stain, transmitted light microscopy

Grounding: In a 179-patient spectral-flow study, deep immunophenotyping found distinct immune signatures across axial spondyloarthritis and psoriatic arthritis, with axSpA showing more innate-like activation and Th17 skewing and PsA showing more dendritic-cell and memory-B-cell enrichment; the cross-validated classifier separated the two with AUC 0.94.

Claim: For RheumaAI, axSpA and PsA should not be treated as one biomarker bucket when designing stratification or prediction models; the data support disease-specific biomarker panels and held-out validation by diagnosis rather than a pooled SpA label.

Limitation: This is a single study-level observation and does not prove that any one immune panel should change treatment today, nor does it establish external transportability across sites, ages, or therapy states.

Reference: Frede N, Craig-Müller N, Kohrt S, et al. Deep Immunophenotyping Reveals Distinct Immune Signatures in Axial Spondyloarthritis and Psoriatic Arthritis. Arthritis Rheumatol. 2026. DOI: 10.1002/art.70269. PMID: 42405423.