090 Serum proteomic endotypes are predictive for response to treatment of atopic dermatitis

Atopic dermatitis (AD) is a well-studied chronic inflammatory skin disease. However, analysis of the disease’s impact on serum proteome has not been well investigated. With this study we aimed for comprehensive characterization disease and its change after systemic treatment. Global spectrometric profiles have generated from 61 AD patients before (BL) 3 months initiation dupilumab treatment (DPL) as 30 healthy controls (HC) using LC-MS/MS technology. Endotype detection performed model-based clustering. Predictability endotypes assessed linear support vector machines. Differential abundance testing conducted models. Proteomic at baseline showed considerable heterogeneity were robustly classifiable into two BL1 BL2. membership could be predicted with median balanced accuracy 0.96 (LQ=0.94, UQ=0.98). Although clinically indistinguishable baseline, differences in regard to longer-term response therapy measured by e.g. EASI 18M (p=2.43E-03), IGA (p=6.19E-03) SCORAD 21M (p=2.15E-02). respect HC revealed 588 829 proteins significantly altered abundances detected direct comparison between BL2 amounted 164. The observed alterations affect related regulation immune system (IL1R, IL6R, IL13RA1, TNFRSF21), organization extracellular matrix (TNC, collagens, integrins) central carbon metabolism (ENOA, G6PD, TKT, TALDO). In conclusion, report putative proteomic that enable accurate prediction atopic dermatitis.