Abstract Background Idiopathic pulmonary fibrosis (IPF) remains a progressive and fatal interstitial lung disease with limited therapeutic options and marked heterogeneity in clinical trajectories. The biological determinants of long-term survival in IPF are incompletely understood, and non-invasive biomarkers reflecting lung pathology are critically needed to guide precision medicine approaches. We leveraged paired bronchial lavage fluid (BALF) and lung tissue proteomics data to identify protein signatures associated with long vs. short survival and evaluated cross-compartment concordance as a strategy for non-invasive biomarker discovery. Methods Participants with IPF (N = 50; 25 long survival, 25 short survival) were enrolled. Untargetedbiased proteomics was performed on BALF (N = 50) and lung tissue (N = 10). Protein intensities were normalized to housekeeping proteins, log2 transformed, and median centered. Linear models with empirical Bayes moderation were used to test the association between protein expression and long vs. short survival, adjusting for age, sex, and smoking status. Overlapping survival-associated proteins were defined as present in both compartments with q 0.1 and concordant directionality. Protein set enrichment, drug repurposing, and cross-compartment correlation of matched sample (N = 9; 630 shared proteins) analyses were performed. Non-negative matrix factorization (NMF) was used to identify latent proteomic subtypes associated with survival. Results Multiple BALF and lung tissue proteins showed significant association with long survival. In BALF, THBS1 (log2FC= -7.2, q = 3.9E-5) was the strongest survival associated protein, while ORM1 (log2FC= -9.9, q = 4.1E-8) and FAM81B (log2FC=7.5, q = 3E-7) were the top tissue associated proteins. Nine proteins demonstrated concordant cross-compartment association with survival (APCS, ARPC2, CTSZ, NAMPT, PEBP1, PFN1, RAP1B, FASN, and THBS1). Drug repurposing analyses highlighted the availability of several FDA-approved drugs selectively targeting significant survival associated proteins with the potential to be repurposed for IPF management. Cross-compartment analysis revealed a subset of proteins with strong positive (e.g., ACO1, CPNE1) or inverse (e.g., A2M, APCS) BALF-lung tissue correlations. NMF identified two BALF and three lung tissue proteomic clusters. Increased expression of proteins in lung tissue compartment 1 (containing THBS1) were negatively correlated (r= -0.66, p = 0.04) with long survival after adjustment for age, sex, and smoking status and enriched for proteins involved in platelet aggregation and the antibacterial humoral response. Conclusions We identified proteomic signatures of long survival in IPF and highlighted BALF proteins reflecting lung tissue biology, supporting their utility as minimally invasive translational biomarkers. Future longitudinal studies are warranted to validate and refine these signatures for non-invasive biomarker stratification and candidate therapeutic targeting strategies to improve IPF management and outcomes. This abstract is funded by: NIH
Wilson et al. (Fri,) studied this question.