Deep Phenotyping of Pulmonary Edema and Pulmonary Vascular Permeability in COVID-19 ARDS.

Background Clinical monitoring of pulmonary edema due to vascular hyperpermeability in ARDS poses significant clinical challenges. Presently, no biological or radiological markers are available for quantifying pulmonary edema. Our aim was to phenotype pulmonary edema and pulmonary vascular permeability in patients with COVID-19 ARDS. Methods Transpulmonary thermodilution measurements were conducted in 65 COVID-19 ARDS patients on the day of intubation to determine extravascular lung water index (EVLWi) and pulmonary vascular permeability index (PVPi). In parallel, ventilatory parameters, clinical outcomes, the volume of lung opacity measured by chest CT, and plasma proteomics (358 unique proteins) were compared between tertiles based on the EVLWi and PVPi. Regression models were used to associate EVLWi and PVPi with plasma, radiological, and clinical parameters. Computational pathway analysis was performed on significant plasma proteins in the regression models. Results Patients with the highest EVLWi values at intubation exhibited poorer oxygenation parameters and more days on the ventilator. Extravascular lung water strongly correlated with the total volume of opacity observed on CT(r=0.72), while the PVPi had weaker associations with clinical and radiological parameters. Plasma protein concentrations demonstrated a stronger correlation with PVPi than with EVLWi. The highest tertile of PVPi was associated with proteins linked to the acute phase response (cytokine and chemokine signaling) and extracellular matrix turnover. Conclusions In the clinical setting of COVID-19 ARDS, pulmonary edema (EVLWi) can be accurately quantified through chest CT and parallels deterioration in ventilatory parameters and clinical outcomes. Vascular permeability (PVPi) is strongly reflected by inflammatory plasma proteins.