Project Details

Defining clinically relevant childhood asthma phenotypes using routinely available laboratory data and clustering methods

Ferdinand Cacho, MD, MPH

Summary

BACKGROUND: Asthma, a common chronic childhood disease, is highly heterogeneous with variable presentations and treatment responses; however, it is typically treated as a single entity with non-specific treatments, even though a substantial sub-group of children do not respond, limiting insights into subtype-specific medication effects. Defining asthma phenotypes is essential to understanding its pathogenesis and advancing targeted therapies to improve asthma care and outcomes. GAP: Current approaches to asthma phenotyping rely on pre-defined clinical categories and overlook data-driven methods, limiting exploration of asthma heterogeneity and its associations with clinical outcomes and treatment responses. HYPOTHESIS: We hypothesize that 1) Distinct childhood asthma phenotypes can be identified using cluster-based analysis of CBC biomarker data, & 2) These phenotypes will differentiate key clinical features of asthma, such as symptoms, lung function, and medication use. METHODS: We will analyze data from a national birth cohort to identify distinct complete blood cell count-based phenotypes using clustering analysis. Then, we will examine associations between these phenotypes and clinical characteristics through multivariable regression models. RESULTS: Pending. IMPACT: The proposed research leverages an innovative methodology to identify asthma subtypes using widely accessible laboratory data from a routine pediatric test and aims to bridge the gap between standard clinical testing and personalized asthma management. The goal is to enable more precise asthma subtype assessment to predict therapeutic responses and guide asthma treatment in children.