EIT-assessed ventilation–perfusion responses to prone positioning and inhaled nitric oxide in mechanically ventilated COVID-19-related ARDS: a prospective sequential study

Background Prone positioning (PP) and inhaled nitric oxide (iNO) are widely utilized rescue therapies for severe hypoxemia during invasive mechanical ventilation. However, their combined effects on regional ventilation-perfusion (V/Q) matching remain incompletely characterized. This study utilized single-plane saline-contrast electrical impedance tomography (EIT) to achieve real-time bedside visualization and evaluate the relative spatial mechanisms of these interventions in patients with COVID-19-related ARDS.Methods We conducted a prospective sequential physiological study in mechanically ventilated patients with moderate-to-severe COVID-19-related acute respiratory distress syndrome (ARDS). Relative regional ventilation, perfusion, and V/Q matching distributions were monitored across four predefined, fixed sequential conditions: supine position ventilation (SPV), SPV with iNO (20 ppm for 1 h; SPV + iNO), prone position ventilation after stabilization (PPV), and PPV with an additional 1 h of iNO (PPV + iNO). Primary analyses focused on prespecified within-patient contrasts, supplemented by exploratory difference-in-differences (DiD) analyses to evaluate position-dependent vascular responses.Results Twenty-eight patients completed all study phases. Hemodynamics and conventional respiratory mechanics remained remarkably stable across all four conditions. Systemic oxygenation improved progressively; both SPV + iNO and PPV significantly increased PaO2 and PaO2/FiO2 compared to SPV alone, while the addition of iNO during PPV (PPV + iNO) yielded further oxygenation improvements and reduced FiO2 requirements. Within the EIT-assessed thoracic slice, SPV + iNO produced the most widespread V/Q improvements, significantly reducing the global non-perfused fraction (-7.28%), wasted ventilation (-8.14%), wasted perfusion (-5.89%), and the global inhomogeneity of perfusion (GIper). Transitioning to PPV similarly reduced the global non-perfused fraction (-8.76%) and GIper, while ventral-to-dorsal ventilation distribution remained stable. Notably, while PPV + iNO improved systemic oxygenation, it did not produce additional significant reductions in global EIT V/Q metrics, and locally increased the dorsal low V/Q compartment (+ 4.01%). DiD analyses revealed that while body position did not significantly modify the acute systemic oxygenation response to iNO, it significantly altered EIT-derived V/Q responses, indicating an attenuated capacity for iNO to improve regional V/Q matching in the imaged slice when administered in the prone position.Conclusions Both prone positioning and iNO improved oxygenation in moderate-to-severe COVID-19-related ARDS. Within the EIT-assessed thoracic slice, these improvements were accompanied by better regional V/Q matching, which was mainly perfusion-related rather than ventilation-related and suggests limited recruitability of the imaged region; the effect of iNO on regional V/Q matching was attenuated in the prone position. As EIT samples a single thoracic cross-section, these are localized findings and should not be generalized to the whole lung. Trial registration ClinicalTrials.gov, NCT05715762. Registered 08 February 2023.