Dermot Linden, Delia Dorrian, Michael Glenn, Shikha Tandel, Thea Mawhinney, Karl Welzenbach, Matthew J Rowland, Ming Yang, Laurent Klein, Emanuele Randazzo, Miriam Sindelar, Joerg Schlotterbeck, Alexander Cattini, Ahilanandan Dushianthan, Emma Crossley, Michael McKelvey, Melanie Bailey, John Conlon, Sharon Carr, Judy Bradley, Clifford C Taggart, David A Simpson, Joseph Kidney, Cecilia M O'Kane, Daniel Francis McAuley
Between January 2023 and March 2024, twenty-two participants were recruited and completed all study procedures. BAL IL-8 did not differ between groups at 6 hours (hyperoxia 375.7 pg/mL (IQR 351.1-467.7) vs placebo 331.9 pg/mL (IQR 251.6-459), p=0.33). No between-group differences were observed in BAL cytokines or markers of systemic inflammation. Hyperoxia was associated with an approximately fourfold increase in BAL reduced glutathione (unadjusted p=0.02). Hyperoxia was associated with distinct transcriptional responses at 24 hours with 175 differentially expressed genes (padj <0.05), including upregulation of genes related to extracellular matrix remodelling, platelet activation and endothelial repair.
INTRODUCTION: It is uncertain whether hyperoxia contributes to adverse intensive care unit outcomes. Potential harm is likely mediated by oxidative stress driving pulmonary and systemic inflammation.
METHODS: We conducted a randomised, double-blind, placebo-controlled trial to test the hypothesis that hyperoxia modifies pulmonary and systemic inflammation following inhaled lipopolysaccharide (LPS) challenge. Healthy adult volunteers were randomised (1:1) to receive hyperoxia administered via high-flow nasal oxygen (FiO2 1.0, 60 L/min) or synthetic medical air for 6 hours. Bronchoalveolar lavage (BAL) was performed at 6 hours following inhalation of 50 µg of LPS. Blood was collected at baseline, 6 hours and 24 hours. The primary outcome was BAL interleukin-8 (IL-8). Secondary outcome measures included markers of pulmonary and systemic inflammation. Untargeted metabolomic and lipidomic profiling of BAL fluid and plasma was performed by liquid chromatography-mass spectrometry to investigate oxidative stress. Transcriptomic analysis was also performed on whole blood.
RESULTS: Between January 2023 and March 2024, twenty-two participants were recruited and completed all study procedures. BAL IL-8 did not differ between groups at 6 hours (hyperoxia 375.7 pg/mL (IQR 351.1-467.7) vs placebo 331.9 pg/mL (IQR 251.6-459), p=0.33). No between-group differences were observed in BAL cytokines or markers of systemic inflammation. Hyperoxia was associated with an approximately fourfold increase in BAL reduced glutathione (unadjusted p=0.02). Hyperoxia was associated with distinct transcriptional responses at 24 hours with 175 differentially expressed genes (padj <0.05), including upregulation of genes related to extracellular matrix remodelling, platelet activation and endothelial repair.
DISCUSSION: In a human inhaled LPS challenge model, 6 hours of hyperoxia did not increase markers of pulmonary or systemic inflammation.
TRIAL REGISTRATION NUMBER: NCT05414370.