(1374) Validation of Volatile Metabolites of Hyperoxia-Induced Oxidative Stress in a Bench-to-Bedside Study

Introduction: Oxygen therapy is a cornerstone treatment in the critical care setting. While undoubtedly lifesaving, hyperoxia may inadvertently increase pulmonary oxidative stress (OS), ultimately causing oxidative injury. Previous clinical studies have found various volatile organic compounds (VOCs) to be associated with OS, particularly hydrocarbons. However, concerns remain whether these VOCs truly originate from OS. This study aimed to identify VOCs of hyperoxia-induced OS in a controlled experimental in vitro setting and confirm their detection in critically ill patients.

Methods: Human alveolar cells (A549) were cultured and seeded in a glass environment to minimize volatile plastic contaminates. Cells were exposed to ~100% O2 for 24h or 21% as a control, after which the headspace air was sampled and supernatants were collected. Captured VOCs were analyzed by gas-chromatography and mass-spectrometry. Wilcoxon-test with false discovery correction was used to assess differences in VOCs between conditions, and a minimum of two-fold change was considered biologically relevant. Inflammation and injury by OS were assessed by IL-8 and LDH. Volatilome data of an existing observational cohort of critically ill patients on invasive ventilation >24h was used to assess the translatability of experimentally identified VOCs.

Results: Hyperoxia exposure of A549 cells was accompanied by a marked increase in IL-8 and LDH in cell supernatant (P < 0.001). One VOC, 1-dodecene, an alkene, was detected with a significant and biologically relevant increase after hyperoxia exposure. Heptane, octane, decane and undecane, four VOCs previously associated with OS that could be detected using a gas standard, did not increase after hyperoxia exposure. The identified VOC from the in vitro model, 1-dodecene, could not be detected in the cohort of critically ill patients.

Conclusions: In this bench-to-bedside study on hyperoxia-induced OS, no translatable VOCs of OS were found. Moreover, several VOCs previously associated with OS in clinical studies were found not to increase after hyperoxia exposure, despite clear signs of OS-induced inflammation and injury. These findings further challenge the validity of previous clinically detected VOCs of OS.