Background: Increasing evidence shows that pediatric atopic dermatitis (AD) differs from adult AD on a biologic level. Broad biomarker profiling across a wide range of ages of pediatric patients with AD is lacking.
Objective: Our aim was to identify serum biomarker profiles in children with AD aged 0 to 17 years and compare these profiles with those previously found in adults with AD.
Methods: Luminex multiplex immunoassays were used to measure 145 biomarkers in serum from 240 children with AD (aged 0-17 years). Principal components analysis followed by unsupervised k-means clustering were performed to identify patient clusters. Patients were stratified into age groups (0-4 years, 5-11 years, and 12-17 years) to assess association between age and cluster membership.
Results: Children aged 0 to 4 years had the highest levels of TH1 cell-skewing markers and lowest levels of TH17 cell-related markers. TH2 cell-related markers did not differ significantly between age groups. Similar to the pattern in adults, cluster analysis identified 4 distinct pediatric patient clusters (TH2 cell/retinol-dominant, skin-homing-dominant, TH1 cell/TH2 cell/TH17 cell/IL-1-dominant, and TH1 cell/IL-1/eosinophil-inferior clusters). Only the TH1 cell/TH2 cell/TH17 cell/IL-1-dominant cluster resembled 1 of the previously identified adult clusters. Although no association with age or age of onset seemed to be found, disease severity was significantly associated with the skin-homing-dominant cluster.
Conclusion: Four distinct patient clusters based on serum biomarker profiles could be identified in a large cohort of pediatric patients with AD, of which 1 was similar to previously identified adult clusters. The identification of endotypes driven by distinct underlying immunopathologic pathways might be useful to define pediatric patients with AD who are at risk of persistent disease and may necessitate different targeted treatment approaches.
Keywords: Atopic dermatitis; biomarkers; cluster analysis; endotypes; pediatric; personalized medicine; principal components analysis.
Copyright © 2021 The Authors. Published by Elsevier Inc. All rights reserved.