PET/CT-Guided Personalized Mechanical Ventilation to Minimize Ventilator-Induced Lung Injury

Ventilator-induced lung injury (VILI) is an inflammatory process that threatens patients requiring ventilatory support such as the 17 million surgical or the 751,000 septic patients/year in the US. Our long-term goal is to understand and advance methods to detect and prevent VILI. Excessive lung strain (=change in volume/initial volume) is understood as a fundamental mechanism of VILI. Recent data show that clinical outcomes during mechanical ventilation are influenced by ventilator settings and importantly associated with a lung mechanics measure related to global lung strain (driving pressure). Yet, knowledge is scant on the effects of ventilatory settings on the spatial distribution of strains in lungs of size comparable to those of humans and on the ensuing lung injury. In the previous period of this project, we advanced Computed Tomography (CT)-based registration methods to show that a current clinical standard ventilatory strategy does not prevent deterioration of local lung strain and aeration in supine animal and human lungs, with progressive derecruitment of dependent lung and increased strain in the well-aerated nondependent lung. We challenged current paradigms as we found local strains lower than proposed global injurious thresholds. Using Positron Emission Tomography (PET), we showed that local 18F-FDG uptake is a marker of neutrophilic inflammation and that it changes directly with the interaction of local lung strain and blood volume (higher exposure to inflammatory mediators) in a model of protective mechanical ventilation and endotoxemia. This implied the novel concept of the two-hit mechanism as the main factor for local VILI at usual clinical strains. We also found that homogeneous distributions of strain and aeration prevented that mechanical deterioration, suggesting that lung expansion homogenization could protect from VILI. Our exciting pilot imaging data indicate tidal local capillary closure with current ventilatory settings providing novel evidence for endothelial injury in presumably protective settings. We hypothesize that regional lung mechanical deterioration characterized by an increase in strain and aeration heterogeneity with current clinical ventilatory strategies produces changes in pulmonary blood volume conducive to endothelial injury, and composes the substrate for inflammation and early VILI in large heterogeneously inflated lungs. We will test this hypothesis with respiratory-gated PET/CT by using CT-derived lung strains and aeration to ascertain global mechanics measures best indicative of the positive end-expiratory pressure (PEEP) leading to homogenous stretch distributions in sheep. We will use such measures to personalize PEEP and assess its effect on: (a) strain deterioration, and the spatial relation of this deterioration with tidal capillary closure, endothelial damage, and lung injury in clinically relevant 48h sheep studies; and (b) local distributions of strain and aeration in septic patients. At the end of this project, we will have tested a strategy translatable to humans to minimize VILI; ascertained relations between global respiratory mechanics, endothelial and parenchymal injury to enhance lung protection; and established imaging methods to assess local VILI.