B22-09 Linking Serum Mirnas and Plasma Metabolites to the Varied Clinical Manifestations of Adult Asthma

Background Adult asthma is a heterogeneous disease characterized by differences in lung function, inflammation, atopy, and treatment response. Metabolomics captures integrated biochemical effects of genetic and environmental exposures, while circulating microRNAs (miRNAs) regulate gene expression and influence immune and metabolic pathways. Integrating these omic layers may reveal shared and distinct molecular mechanisms underlying different asthma endotypes. Methods We analyzed plasma metabolomic profiles (1,150 metabolites; N = 1,035) and serum small RNA sequencing data (364 miRNAs; N = 974) from asthmatic adults enrolled in the Mass General Brigham Biobank-Longitudinal Lung Function Study (MGBB-LLF). Eleven clinical traits spanning five phenotypic domains—oral corticosteroid (OCS) use, lung function, blood eosinophils, blood neutrophils, and atopy—were evaluated. Generalized linear models adjusted for demographic and clinical covariates were used to identify significant omic-trait associations (FDR q < 0.05). Multi-domain features were defined as those associated with at least two phenotypic domains. miRNA associations were further replicated in the independent COPDGene cohort (N = 538; peripheral whole blood small RNA sequencing) at p < 0.05. Finally, LASSO logistic regression assessed the predictive performance of domain-associated omic features for asthma-related hospitalization. Results We identified 72 significant miRNA-trait associations involving 57 unique miRNAs, predominantly domain-specific, with enrichment in neutrophil- and lung function-related traits. Five miRNAs exhibited multi-domain associations. Of the 72 associations identified in the MGBB-LLF cohort, 29 (40%) were replicated in the COPDGene cohort at p < 0.05. Replicated domains included blood neutrophils, blood eosinophils, and lung function, with hsa-miR-25-3p showing multi-domain replication across lung function and blood neutrophils. Metabolomic analysis identified 685 significant associations involving 394 unique metabolites, including steroid hormones, amino acids, and lipids. Four steroid metabolites (DHEA-S and three pregnenolone derivatives) were associated with all five phenotypic domains. Predictive modeling showed that metabolites outperformed miRNAs in forecasting asthma-related hospitalizations, achieving an AUC of 0.82 (0.77-0.87) for prior-year and 0.68 (0.63-0.72) for two-year predictions (p = 3.96 × 10⁻8), whereas miRNA model did not achieve prediction statistically better than chance (p = 0.115). Conclusion Integrated miRNA and metabolomic profiling reveal domain-specific and cross-domain molecular signatures of asthma heterogeneity. Metabolites, particularly steroid-related pathways, showed strong predictive potential for exacerbation risk, underscoring their value for as biomarkers and potential precision medicine targets in adult asthma. The replication of miRNA associations from an asthma population to a COPD population points to the shared regulatory genomic basis of these diseases. This abstract is funded by: This work was funded by NIH grant R01 HL155742. COPDGene was supported by NHLBI grants U01 HL089897 and U01 HL089856 and by NIH contract 75N92023D00011. The COPDGene study (NCT00608764) has also been supported by the COPD Foundation through contributions made to an Industry Advisory Committee that has included AstraZeneca, Bayer Pharmaceuticals, Boehringer-Ingelheim, Genentech, GlaxoSmithKline, Novartis, Pfizer, and Sunovion.