Exploring the Varied Clinical Presentation of Pediatric Asthma Through the Metabolome

This study aimed to characterize the plasma metabolomic signatures of children with mild to moderate asthma and to assess how these signatures correlate with clinical phenotypes and endotypes.The analysis included 953 participants (13 ± 2.1 years) from the Childhood Asthma Management Program (CAMP) and 1155 participants (9.2 ± 1.9 years) from the Genetic Epidemiology of Asthma in Costa Rica Study (GACRS) cohorts.Plasma samples were analyzed for 589 metabolites using high-resolution liquid chromatography–tandem mass spectrometry methods. Linear models were used to determine associations between these metabolites and 22 clinical asthma traits across 5 phenotypic domains (airway hyperresponsiveness [AHR], atopy, lung function [LF], blood eosinophils [B-EOS], and blood neutrophils [B-NEU]). Metabolomic risk scores (MRSs) were calculated to represent either a positive or negative association for each clinical domain within the previously identified metaboendotypes (endotypes 1 through 5) in each cohort.The 2 study cohorts, while generally similar, had differences in B-EOS and LF, although LF z scores, which account for age, were similar. Many participants had elevated total serum immunoglobulin E and positive skin prick tests as evidence of atopy. There was a significant association between 154 unique metabolites and at least 1 of 22 clinical traits. Histamine, kynurenine, nudifloramide, niacinamide, taurine, PC(P-36:0)/PC(O-36:1), 12,13-diHOME, azelate, and sebacate were significantly associated with differences in B-EOS, B-NEU, and T2 vs non–T2-driven responses. The fatty acid and benzoate metabolites were significantly associated with AHR, LF, and B-EOS, suggesting the potential role of fatty acids and environmental exposures, such as industrial pollutants, on asthma. Vitamin A was also noted to have a strong association with asthma traits. The MRSs by domain help characterize each metaboendotype with its unique metabolic signature and inflammatory profile.The authors demonstrate how these extensive analyses of the metabolome could be used to identify metabolites that may be clinically significant or be associated with specific signs or symptoms of different asthma clinical profiles. This could lead to better classification of the phenotypic heterogeneity of pediatric asthma and, also, potentially result in selecting more personalized treatment options.This large, multicohort study adds to the body of literature illustrating the biologic complexity of pediatric asthma while offering to further clarify disease pathogenesis. Analysis of the metabolome suggests that distinct biochemical pathways underlie the dysregulated inflammation in asthma in each patient and thus potentially further fine-tune characterization of the different pediatric asthma endotypes and phenotypes. How factors such as dietary and/or environmental exposures could specifically influence asthma inflammation and how to meaningfully modify these exposures would warrant further study. This study also highlights the importance of large, diverse cohorts in unraveling disease heterogeneity across populations. Ultimately, in this era of rapidly emerging new biologic therapies for asthma, it remains to be seen how all of these metabolomic data could help select a particular targeted therapy for each pediatric patient with asthma.