D2.102 - Adult Gut Microbiota and Metabolomics in Different Phenotypes of Asthma

Bronchial asthma is a common chronic inflammatory airway disease with an increasing prevalence that significantly impacts human life and health. The pathogenesis of asthma is highly complex, with diverse clinical phenotypes, and remains inadequately understood. 

This study was conducted in the Respiratory Medicine Outpatient Clinic and the Health Check-up Department of the Second Affiliated Hospital of Xi'an Jiaotong University. Asthma (AS) patients and healthy controls (HC) who met the inclusion and exclusion criteria were enrolled, including 55 asthma patients and 17 healthy controls. The asthma group was further divided into eosinophilic asthma (EA) and non-eosinophilic asthma (NEA), allergic asthma (AAS) and non-allergic asthma (NAAS), low FeNO (LF), intermediate FeNO (IF), and high FeNO (HF) asthma groups, and mild-to-moderate asthma (MA) and severe asthma (SA) groups. Fecal samples were collected from all study subjects, and the gut microbiome differences among the groups were analyzed using 16S rRNA sequencing. Metabolite expression in each group was analyzed using non-targeted metabolomics with gas chromatography-mass spectrometry.

In comparison to the HC group,the relative abundances of Clostridiaceae and Prevotella were significantly increased in the AS group, Further analysis using a random forest machine learning algorithm identified a set of 6 microorganisms as a predictive model, with ROC curve analysis indicating good diagnostic performance for these 6 microorganisms, with an AUC of 0.812. Differential microbiota analysis revealed that Enterobacteriales and Prevotella were higher in the EA group.The relative abundances of Micrococcales and Faecalicatena were significantly higher in the AAS group. Further analysis using a random forest machine learning algorithm identified 2 microorganisms as a predictive model, with ROC curve analysis indicating good diagnostic performance for these 2 microorganisms, with an AUC of 0.777.In the IF group, the relative abundances of Eubacteriales,Prevotellaceae,and Vescimonas were higher.Dfferential microbiota analysis revealed that, compared with the SA group, the relative abundances of Enterococcaceae,and Actinomycetaceae were significantly higher in the MA group.AS and HC groups: A random forest model constructed using the identified differential metabolites selected 6 metabolites as a predictive model, with ROC curve analysis indicating good diagnostic performance for these 6 metabolites, with an AUC of 0.782.AAS and NAAS groups:  A random forest model constructed using the identified differential metabolites selected 3 metabolites as a predictive model, with an AUC of 0.742.Different FeNO groups, with 45 differential metabolites identified, with an AUC of 0.986.MA and SA groups: A random forest model constructed selected 4 metabolites as a predictive model,  with an AUC of 0.933.

There is significant dysbiosis of gut microbiota and alterations in metabolites among different phenotypes of asthma patients.