Oxidative Lipidomic Biomarkers During Mechanical Ventilation in Critically Ill Patients (OXY-BAL).

This proposal will elucidate an approach for precise and safe oxygenation levels and correlation with the oxidative lipid biomarker profile of optimal oxygenation. The aim of this study is to identify and determine the levels of oxidized lipids and lipid mediators following exposure to oxygen supplementation during mechanical ventilation by oxidative lipidomics. Alveolar cellular membrane and lung lining fluid are the primary targets of therapeutic oxygen and comprised of proteins, cholesterol, and lipids. Among these components, lipids are more susceptible to oxidation producing downstream mediators that facilitate inflammatory and pro-fibrotic pathways. Polyunsaturated Fatty Acids (PUFAs) account for 30% of the lipid components and contain n-3 fatty acids (i.e eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA)) and n-6 fatty acids ( arachidonic acid (AA) and Dihomo-γ-linolenic acid (DGLA)). In the presence of reactive oxygen species and activation of the metabolic enzymes (e.g. cyclooxygenases and lipoxyegenases), n-3 and n-6 PUFAs undergo peroxidation to form "oxidized forms" and are further metabolized into numerous proinflammatory and proresolving lipid mediators. Many of these lipid mediators (Prostaglandin E2, PGE2; 11-hydroxy--eicosatetraenoic acid 11-HETE, 5-hydroxy-eicosapentaenoic acid 5-HEPE) perpetuate inflammatory mechanisms known to mediate HALI45. Additionally, some of these mediators known as specialized proresolving mediators (SPMs) dampen inflammation while promoting pro-resolution actions on tissue inflammation after injury (e.g maresins, Resolvin D5). Lipidomics is one of the most sensitive approaches for studying lipid profiles in clinical research. Given the access to lung samples (bronchoalveoar fluid; BALF) in the setting of acute hypoxemic respiratory failure in the ICU, this provides strong rationale to utilize this largely unexplored method to guide clinical practice of optimal oxygenation goals.

We propose a prospective cohort study to evaluate oxidized lipid biomarkers in the BALF of patients with acute hypoxemic respiratory failure. Inclusion criteria are 1) Adults ≥ 18 years with acute hypoxemic respiratory failure 2) with expected MV => 48 hours. Exclusion criteria are 1) MV for procedures like EGD, bronchoscopy , colonoscopy;2) Chronic respiratory failure prior to admission; 3) Hemoptysis, diffuse alveolar hemorrhage; 3) Primary team prohibits bronchoscopy for a specific clinical indication or safety; 4) Medication use such as N acetyl cysteine; aspirin, fish oil supplements. Control subjects will be those with MV without acute hypoxemic or hypercarbic respiratory failure (e.g. MV due to stroke or seizures). Informed consent will be taken. All consenting patients will have a two sequential BALF collections by collected from patients on day 1 (≤6 hours of MV if initial FiO2 use is ≥0.5 for 90 minutes) and day 2 (≥48 hours after the first BALF). We have specifically chosen these time points because Day 1 will allow us to define initial oxidized n-6 PUFA precursor levels and their relationship to clinical outcomes. Since markers of lipid peroxidation are at the highest on Day 2, these Day 2 samples will allow us to define the trajectory of oxidized n-6 PUFA and associate these with clinical outcomes.

Following the BAL collection, it will be stored at -80 and Liquid chromatography-mass spectrometry will be used for oxidative lipidomic analysis of ~130 oxidized lipid species. Patient data will be abstracted from the chart at time of enrollment. Clinical data variables will include: 1) demographics; 2) medical history 3) hourly FiO2 and SpO2 4) Sequential Organ Failure Assessment (SOFA) scores for disease severity; 5) daily arterial blood gas for PaO2/FiO2 ratio 6) daily vital signs 7) clinical laboratory studies (blood counts for hemoglobin, electrolyte panel, lactic acid levels; 8) medications; and 9) outcomes measures (ventilator-free days, survival). More importantly, we will note co-morbidities and treatment related variables that may independently modify oxidized lipids and serve as confounders. Such smoking status, active lung or breast cancer, quantity and nature of tube feeds, use of propofol (a lipid emulsion) for sedation, and utilization of immune modulators such as steroids, will be noted and will be adjusted for univariate and multivariate analysis.