1. Predicting persistence of atopic dermatitis in children using clinical attributes and serum proteins.
- Author
-
Lauffer, Felix, Baghin, Veronika, Standl, Marie, Stark, Sebastian P., Jargosch, Manja, Wehrle, Julius, Thomas, Jenny, Schmidt‐Weber, Carsten B., Biedermann, Tilo, Eyerich, Stefanie, Eyerich, Kilian, and Garzorz‐Stark, Natalie
- Subjects
BLOOD proteins ,ATOPIC dermatitis ,DISEASE exacerbation ,AGE distribution ,K-means clustering - Abstract
Background: Atopic dermatitis (AD) is the most common inflammatory skin disease in children, with 30% of all those diagnosed developing chronic or relapsing disease by adolescence. Such disease persistence cannot yet be predicted. The aim of the present study was to predict the natural course of AD using clinical parameters and serum proteins. Methods: Sera of 144 children with AD (age 0‐3 years) were analyzed for IgE and 33 cytokines, chemokines, and growth factors. Patient disease course until the age of 7 years was assessed retrospectively. Unsupervised k‐means clustering was performed to define disease endotypes. Identified factors associated with AD persistence at the age of 7 years were validated in children with AD in an independent cohort (LISA Munich; n = 168). Logistic regression and XGBoosting methods followed by cross‐validation were applied to predict individual disease outcomes. Results: Three distinct endotypes were found in infancy, characterized by a unique inflammatory signature. Factors associated with disease persistence were disease score (SCORAD), involvement of the limbs, flexural lesion distribution at the age of 3 years, allergic comorbidities, and disease exacerbation by the trigger factors stress, pollen exposure, and change in weather. Persistence was predicted with a sensitivity of 81.8% and a specificity of 82.4%. Factors with a high impact on the prediction of persistence were SCORAD at the age of 3 years, trigger factors, and low VEGF serum levels. Conclusion: Atopic dermatitis in infancy comprises three immunological endotypes. Disease persistence can be predicted using serum cytokines and clinical variables. [ABSTRACT FROM AUTHOR]
- Published
- 2021
- Full Text
- View/download PDF