Despite recent innovations in imaging and genomic screening promotes advance in diagnosis and treatment of lung adenocarcinoma (LUAD), there remains high mortality of LUAD and insufficient understanding of LUAD biology. Our previous study performed an integrative multi-omic analysis of LUAD, filling the gap between genomic alterations and their biological proteome effects. However, more detailed molecular characterization and biomarker resources at proteome level still need to be uncovered. In this study, a quantitative proteomic experiment of patient-derived benign lung disease samples was carried out. After that, we integrated the proteomic data with previous dataset of 103 paired LUAD samples. We depicted the proteomic differences between non-cancerous and tumor samples and among diverse pathological subtypes. We also found that up-regulated mitophagy was a significant characteristic of early-stage LUAD. Additionally, our integrative analysis filtered out 75 potential prognostic biomarkers and validated two of them in an independent LUAD serum cohort. This study provided insights for improved understanding proteome abnormalities of LUAD and the novel prognostic biomarker discovery offered an opportunity for LUAD precise management.