Objective: We try to screen out novel protein biomarkers with diagnostic value in bronchoalveolar lavage fluid (BALF), so as to supplement bronchoscopy biopsy for early-stage lung adenocarcinoma.

Materials and Methods: Liquid chromatography MS analysis (LC-MS/MS) with data-independent acquisition strategy was carried out in BALF samples. Differentially expressed proteins (DEPs) were identified by Limma function. All the DEPs were submitted to Gene Oncology (GO) annotation. DEPs associated with tumor characteristics were selected to randomForest algorithm, and a further combination derived from the biomarkers was developed to predict the malignant nodule by randomForest. Early lung adenocarcinoma tissue proteome was acquired from CPTAC public database to verify prognostic significance of concerned proteins. Area under the curve (AUC), sensitivity, specificity, accuracy and F1 score were used to evaluate diagnostic performance of biomarkers.

Results: Patients in the LUAD group were nearly all invasive adenocarcinoma. Patients in the LUBN group mainly included pulmonary tuberculosis. DEPs including 1 upregulated and 10 downregulated proteins in malignancy were identified. Most of downregulated proteins in LUAD group involved ubiquitylation-dependent proteolysis, tumor cell apoptosis and cell cycle. The only upregulated protein was MUC1 which participated in alveolar epithelial cells regeneration. UBQLN1 showed well AUC in both the training and validation cohorts. Furthermore, low levels of UBQLN1 in lung adenocarcinoma tissue conferred poor survival according to the public database.