Table_3_Serum Cytokine Profiling Identifies Axl as a New Biomarker Candidate for Active Eosinophilic Granulomatosis With Polyangiitis.doc
Background: Eosinophilic granulomatosis with polyangiitis (EGPA) prognosis is generally favorable and is treated with combined corticosteroids/immunosuppressor(s) therapy. However, disease flares increase the number of clinical visits. Therefore, discovering new serum biomarkers for early identification of active EGPA is crucial.
Objective: To identify reliable serum biomarkers to measure EGPA activity.
Methods: The expression of 160 proteins was compared in sera from 15 inactive and 13 active EGPA patients by antibody-based microarray. Network-based analysis identified patterns in the different groups. Differentially expressed proteins (DEPs) in active disease were identified, and the correlation between their serum levels and clinical parameters was assessed. DEPs were further analyzed for GO enrichment and KEGG pathways. Finally, DEP marker candidates were validated by ELISA and Bio-plex as well as against a second cohort of 22 inactive and 18 active EGPA patients.
Results: The active group presented higher peripheral and sputum eosinophil counts, FeNO, and FEV1 (% predicted) (P < 0.05). Network-based analysis showed scattered expression patterns in active subjects, but no significant bias in inactive subjects. Significant differences were observed in serum levels of 19 candidate markers, all of which were higher in active EGPA (P < 0.05). KEGG analysis indicated that DEPs were mainly involved in the MAPK, PI3K-Akt, RAS and Rap1 related pathways. Nine out of 19 candidate markers were positively correlated with peripheral eosinophil counts including FGF-7, SCF, GDNF, β-NGF, IGFBP-4, Axl, PIGF, Insulin, NT-4, ErbB3, OPN and BMP-4 (r = 0.693, r = 0.692, r = 0.687, r = 0.683, r = 0.671, r = 0.606, r = 0.571, r = 0.570, r = 0.516, respectively; P < 0.05), while two, CD14 and MCP-3, were negatively correlated (r = −0.644 and r = −0.515; P < 0.05). The higher expression of Axl, OPN, HCC-4, GDNF, and MCP-3 in active EGPA subjects was confirmed by ELISA and Custom Multiplex Bio-plex analyses.
Conclusion: The serum protein profiles were significantly different between active and inactive EGPA. The expression of the candidate proteins correlated with peripheral blood eosinophil count. Serum Axl, OPN, HCC-4, GDNF, and MCP-3 levels were consistently higher in active EGPA, independent of the assessment methods. Finally, Axl had the largest AUC, indicating that this cytokine may serve as novel biomarker for the diagnosis of active EGPA.