Metabolomics based predictive biomarker model of ARDS: A systemic measure of clinical hypoxemia

• DOI: 10.1371/journal.pone.0187545
• Publication Date: 2017-11-02T17:02:38Z
• Authors: Akhila Viswan; Chandan Singh; Ratan Kumar Rai; Afzal Azim; Neeraj Sinha; Arvind Kumar Baronia
• Keywords: Adult; Male; Respiratory Distress Syndrome; Science; Q; R; Middle Aged; 3. Good health; Medicine; Humans; Metabolomics; Female; Hypoxia; Biomarkers; Research Article

Despite advancements in ventilator technologies, lung supportive and rescue therapies, the outcome prognostication acute respiratory distress syndrome (ARDS) remains incremental ambiguous. Metabolomics is a potential insightful measure to diagnostic approaches practiced critical disease settings. In our study patients diagnosed with mild moderate/severe ARDS clinically governed by hypoxemic P/F ratio between 100–300 but indistinct molecular phenotype were discriminated employing nuclear magnetic resonance (NMR) based metabolomics of mini bronchoalveolar lavage fluid (mBALF). Resulting biomarker prototype comprising six metabolites was substantiated highlighting susceptibility/recovery. Both groups (mild ARDS) showed distinct biochemical profile on 83.3% classification discriminant function analysis cross validated accuracy 91% using partial least squares as major classifier. The predictive performance narrowed down found analogous chemometrics. proposed model consisting proline, lysine/arginine, taurine, threonine glutamate characteristic sub-stages aberrant metabolism observed mainly arginine, proline metabolism, lysine synthesis so forth correlating diseased metabotype. Thus NMR has provided new insight into conclusively precise proposed, reflecting underlying metabolic dysfunction aiding prior clinical decision making.