Background: Asthma is a heterogeneous disease with diverse and poorly defined phenotypes, especially in children.
Objective: We aimed to classify childhood asthma phenotypes using unsupervised cluster analysis based on type 2 (T2) biomarkers and to evaluate their clinical characteristics and outcomes.
Methods: We retrospectively analyzed 614 pediatric patients. Hierarchical clustering was performed using four variables: age, absolute eosinophil count (AEC), eosinophil cationic protein (ECP), and total immunoglobulin E (IgE). Clinical characteristics and two-year clinical outcomes were compared across clusters. The effect of age was examined using Analysis of Covariance (ANCOVA) and age-tertile subgroup analyses.
Results: Three distinct clusters were identified. Cluster 2, the T2-high asthma group (n = 157; median age: 8.0 years), was characterized by male predominance (69.4%), the highest levels of T2 biomarkers (AEC, ECP, IgE, FeNO; all P < 0.001), airway hyperresponsiveness (BDR, P = 0.024; PC20, P < 0.001), and reduced lung function (FEV1, P = 0.002). In contrast, Cluster 1 (n = 252; median age: 4.0 years) showed the highest exacerbation and steroid use rates but relatively low T2 biomarker levels. Cluster 3 (n = 205; median age: 12.0 years) had moderate T2 levels and the lowest exacerbation burden. After adjusting for age, Cluster 2 maintained 4-6 folds higher T2 biomarker levels compared to other clusters (all P <0.001). These cluster-specific differences were not observed in the age tertile subgroup analysis.
Conclusions: The identified school-age T2-high cluster in childhood asthma exhibits distinct immunological and clinical phenotypes, characterized by high airway hyperresponsiveness, atopic features, and decreased lung function.
扫码关注我们
求助内容:
应助结果提醒方式:
