A0961
Title: Integrated analysis of imaging and RNA-seq to characterize smoking-related lung disease phenotypes
Authors: Fenghai Duan - Brown University School of Public Health (United States) [presenting]
Abstract: The goal is to identify biological distinctions across smoking-related phenotypes by integrating clinical, imaging, and bronchial epithelial RNA-seq data. Using k-means clustering, participants are grouped based on CT imaging features, and their clinical phenotypes are analyzed. Bronchial epithelial RNA-seq data are further examined through differential gene expression, gene set enrichment, and variation analyses. Three distinct clusters are identified: Preserved, interstitial predominant, and emphysema predominant. Compared to the preserved cluster, the interstitial and emphysema clusters showed poorer lung function, lower exercise capacity, and worse quality of life. They also experienced faster declines in function, greater emphysema progression, more respiratory events, and higher mortality. The emphysema cluster had the most severe outcomes, followed by the interstitial cluster. Transcriptomic analysis indicated that severe disease stages were linked to heightened inflammatory responses, especially through the TNF-alpha pathway, while milder stages showed upregulation of T-cell-related genes. Using quantitative CT imaging, we identified three subgroups among individuals with a history of heavy smoking. Differences in airway gene expression suggest a connection between clinical severity and inflammation, possibly mediated by the TNF-alpha pathway.
