Personalized Prediction of Atopic Dermatitis Disease Course Across Diverse Settings
Funded Grant
Overview
Affiliation
View All
Overview
description
PROJECT SUMMARY Atopic dermatitis often presents in early childhood and is characterized by a heterogeneous disease course: while some children improve by adolescence, many others continue to have persistent disease and develop comorbid allergic, neuropsychiatric, and metabolic conditions. Prognostic information is needed for patient counseling and to improve the efficiency of clinical trials evaluating whether new treatments could impact long- term outcomes. We propose a series of aims designed to predict which children with early-onset disease are more likely to develop persistent disease and comorbid conditions. First, using machine learning methods, we will develop prognostic models for persistent atopic dermatitis in late childhood using a standardized set of easy-to-measure early life predictors in diverse birth cohorts from Uganda, Ecuador, and the United Kingdom. We will compare model performance across cohorts and evaluate whether a single, generalized model for persistent atopic dermatitis could be useful in diverse settings. Second, we will create another set of models for each cohort adding more detailed predictors such as genetic risk scores, environmental variables and bio samples that vary across cohorts and may be more costly to measure. We will compare the performance of the models with an expanded set of predictors to the models that use a more limited set of standardized predictors within each cohort to better understand the nature of data needed for atopic dermatitis risk prediction across diverse settings. Finally, we will expand the prognostic models to additionally capture atopic dermatitis comorbidities, including other allergic diseases such as asthma and allergies; neuropsychiatric conditions such as anxiety, depression, and attention-deficit hyperactivity disorder; and metabolic disorders such as diabetes, dyslipidemia, obesity and hypertension. The results will offer useful prognostic information for patients, address a critical need for atopic dermatitis research among diverse populations, and enable targeted enrollment to improve the efficiency of clinical trials. This is particularly timely because the FDA has endorsed the use of tools designed to enroll patients most likely to demonstrate disease persistence, and there are many new systemic treatments under development for atopic dermatitis that are being tested for children as young as 6 months old.
