Proteomic Biomarkers of Survival in Non–Idiopathic Pulmonary Fibrosis Interstitial Lung Disease

Rationale: Although idiopathic pulmonary fibrosis (IPF) has been widely studied, progressive non-IPF interstitial lung disease (ILD) remains poorly understood. Objectives: To identify and validate proteomic biomarkers of non-IPF ILD survival. Methods: High-throughput proteomic data were generated using plasma collected as part of prospective registries at the universities of California and Texas (discovery cohort; n = 676) and in the PRECISIONS multiomic study (validation cohort; n = 616). Proteins associated with 3-year transplant-free survival (TFS) were identified using multivariable Cox proportional hazards regression, and those associated with TFS after adjustment for false discovery were advanced for validation cohort testing. Pathway analysis was performed to identify molecular pathways unique to non-IPF ILD and shared with IPF. Measurements and Main Results: Of 2,925 proteins tested in the discovery cohort, 73 were associated with TFS, with 44 showing sustained TFS association in the validation cohort. The top TFS-associated proteins were amphiregulin (hazard ratio [HR], 2.51; 95% confidence interval [CI], 2.07-3.04), integrin subunit-β6 (HR, 2.46; 95% CI, 1.95-3.10), and keratin 19 (HR, 1.70; 95% CI, 1.47-1.98). All but one validated biomarker showed consistent TFS association across non-IPF ILD subtypes. Pathway analysis identified several molecular pathways shared with IPF, together with three pathways unique to non-IPF ILD. Conclusions: We identified and validated novel prognostic protein biomarkers in non-IPF ILD, most of which showed consistent association across non-IPF ILD subtypes. Although most biomarkers and molecular pathways identified were previously linked to IPF, several were unique to non-IPF ILD, suggesting that unique biology may contribute to progressive non-IPF ILD.