Abstract Rationale Metabolic alteration in cancer patients identified by various tumor markers has revolutionized treatment and diagnostics for various malignancies. Mass spectrometry-based metabolomics is considered a promising field for developing novel tumor biomarkers. Among metabolites, amino acids are some of the most suitable biomarkers for cancer research because they are directly involved in the metabolic pathways that feed tumor cells. Previously published studies have compared the metabolites within blood, sputum, and exhaled breath condensate in individuals with cancer against a control population without underlying malignancy. This study uses bronchoalveolar lavage (BAL) fluid to analyze differences between the amino acid profiles of the healthy and the affected lung of patients with known non-small cell lung cancer (NSCLC). Methods We performed a metabolomic analysis of BAL fluid from 14 patients with NSCLC using high-performance liquid chromatography - mass spectrometry. Bronchoscopy was performed in each patient under sedation and BAL samples were obtained from the segmental bronchus leading to the tumor and from the corresponding bronchus of the contralateral healthy lung, which served as a control. Based on the amino acid profiles, we developed a diagnostic model for NSCLC, and its performance was evaluated using Receiver Operating Characteristic (ROC) analysis. Results A total of 32 amino acids were detected in the BAL fluid. Among these, 10 amino acids showed statistically significant differences (p 0.05) between the cancerous and contralateral healthy lungs: valine, lysine, ornithine, asparagine, trans-4-hydroxyproline, carnosine, 1-methylhistidine, homocitrulline, alanine, and 2-aminobutyric acid. These combinations of amino acids were used to construct diagnostic models for NSCLC based on ROC analysis (Fig.). The following combinations of four amino acids - Ornithine-Lysine-Carnosine-Valine; Ornithine-Lysine-Asparagine-2-Aminobutyric Acid; Ornithine-Lysine-Homocitrulline-Valine; and 1-Methylhistidine-Asparagine-trans-4-Hydroxyproline-Alanine - demonstrated the highest positive predictive value for NSCLC (0.88-0.93) with a cutoff of 0.41-0.8 with high sensitivity and specificity. Conclusion BAL fluid in patients with NSCLC shows characteristic differences in the amino acid profiles between healthy and affected lungs even within the same individual. These differences may be attributable to distinctive differences in metabolism in tumor cells when compared to normal tissue and may serve as biomarkers of the disease that can be used as an effective method for early diagnosis of NSCLC. This abstract is funded by: None
Danilevskaya et al. (2026) studied this question.