Systematic Pharmacovigilance and Genetic Validation of Medications Associated With Organizing Pneumonia : Integrating FAERS Data With Multiomics Analysis

Wu, Wenhao and Zhen, Huang and Li, Guangming and Cai, Shanshan and Fan, Xinyuan (2026) Systematic Pharmacovigilance and Genetic Validation of Medications Associated With Organizing Pneumonia : Integrating FAERS Data With Multiomics Analysis. International journal of clinical practice, 2026.

Abstract

Background Drug-induced organizing pneumonia (DIOP) is increasingly recognized, but systematic data on associated drugs and their biological mechanisms are limited. Methods We analyzed DIOP reports from the FDA Adverse Event Reporting System (FAERS) (2004–2024). Disproportionality analysis (ROR, PRR, BCPNN, and MGPS) was performed to detect signals. Time-to-onset (TTO) was characterized by using Weibull analysis. Furthermore, to explore biological plausibility, we performed a post-GWAS multivariate analysis using sparse canonical correlation analysis (sCCA) and proteome-wide association studies (PWAS) to investigate shared genetic architecture between identified drug targets and lung injury phenotypes (ILD, IPF, CRP, and TNF). Results Analysis of 3912 DIOP reports identified strong signals across diverse categories: antineoplastics (e.g., bleomycin and decitabine), immunotherapies (e.g., pembrolizumab), and anti-inflammatory agents (e.g., mesalazine and amiodarone). Genetic analysis revealed significant overlaps: FKBP1A (sirolimus target) and JAK1/3 (tofacitinib target) were genetically associated with ILD and TNF, suggesting shared pathways between drug targets and lung fibrosis. PWAS further validated functional links, identifying a strong positive association between GSR (nitrofurantoin target) and CRP, and between EGFR (brigatinib target) and inflammatory phenotypes. Conclusion This study integrates real-world pharmacovigilance data with genetic validation. The findings not only highlight high-risk drugs requiring clinical vigilance but also suggest a shared genetic architecture between drug targets and lung injury phenotypes, providing supportive evidence for the biological plausibility of these associations.