Integrated plasma proteomic drives Risk stratification and phenotypic trajectory in ILD

Résumé fourni par la source

Introduction: Interstitial lung disease (ILD) encompasses heterogeneous disorders with poorly defined biomarkers, limiting early diagnosis and classification of disease behavior. Aims and objectives: Our aim is to integrate plasma proteomics to reveal ILD biomarkers and predictive risk stratification, uncovering molecular subtypes. Methods: We performed cox regression on proteins to predict future ILD incidence. ILD were grouped to explore their minute subtypes. Validation is performed in an external queue. At the same time, significant genes were verified in basic experiments. Single-cell RNA-seq mapped epithelial-immune dynamics. Results: In 53,018 UK Biobank participants (1019 incident ILD), we analyzed 1,463 plasma proteins (Olink) linked to longitudinal outcomes. Machine learning prioritized biomarkers validated in 120 Chinese patients/controls via mass spectrometry/Olink. Cox regression identified 640 ILD-associated proteins (p<3.42×10⁻⁵). An 11-protein panel predicted ILD (AUC=0.833), with WFDC2 (AUC=0.803) and GDF15 (AUC=0.788) showing early divergence (>10 years pre-diagnosis). Three molecularly distinct ILD subtypes were delineated: 1) "Adhesion-dominant" (SFTPD⁺; reduced FVC decline rate: −1.2%/year), 2) "Hyper-inflammatory" (LGALS9⁺; elevated CRP levels), and 3) "High-risk immunodysregulated" (WFDC2⁺; all-cause mortality was higher than other two groups HR=2.89). scRNA-seq revealed pathogenic Basal/SCGB3A2⁺ epithelial states driving TGF-β/Wnt crosstalk with macrophages. Lung tissue validation confirmed elevated WFDC2 and GDF15 in fibrosis. Conclusions: These findings emphasize that the immune-regulatory ILD subtype marked by WFDC2 elevation represents a high-risk phenotype with accelerated disease progression and distinct molecular pathology.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.