Pulmonary inflammation decreases with ultra-protective ventilation in experimental ARDS under VV-ECMO: a positron emission tomography study.

BACKGROUND: Experimentally, ultra-protective ventilation (UPV, tidal volumes [VT ] < 4 mL.kg-1 ) strategies in conjunction with veno-venous extracorporeal membrane oxygenation (VV-ECMO) are associated with lesser ventilator-induced lung injuries (VILI) during acute respiratory distress syndrome (ARDS). However, whether these strategies reduce lung inflammation more effectively than protective ventilation (PV) remains unclear. We aimed to demonstrate that a UPV strategy decreases acute lung inflammation in comparison with PV in an experimental swine model of ARDS.

METHODS: ARDS was induced by tracheal instillation of chlorhydric acid in sedated and paralyzed animals under mechanical ventilation. Animals were randomized to receive either UPV (VT 1 mL.kg-1 , positive end-expiration pressure [PEEP] set to obtain plateau pressure between 20 and 25 cmH2 O and respiratory rate [RR] at 5 min-1 under VV-ECMO) or PV (VT 6 mL.kg-1 , PEEP set to obtain plateau pressure between 28 and 30 cmH2 O and RR at 25 min-1 ) during 4 h. After 4 h, a positron emission tomography with [11 C](R)-PK11195 (ligand to TSPO-bearing macrophages) injection was realized, coupled with quantitative computerized tomography (CT). Pharmacokinetic multicompartment models were used to quantify regional [11 C](R)-PK11195 lung uptake. [11 C](R)-PK11195 lung uptake and CT-derived respiratory variables were studied regionally across eight lung regions distributed along the antero-posterior axis.

RESULTS: Five pigs were randomized to each study group. Arterial O2 partial pressure to inspired O2 fraction were not significantly different between study groups after experimental ARDS induction (75 [68-80] mmHg in a PV group vs. 87 [69-133] mmHg in a UPV group, p  = 0.20). Compared to PV animals, UPV animals exhibited a significant decrease in the regional non-aerated compartment in the posterior lung levels, in mechanical power, and in regional dynamic strain and no statistical difference in tidal hyperinflation after 4 h. UPV animals had a significantly lower [11 C](R)-PK11195 uptake, compared to PV animals (non-displaceable binding potential 0.35 [IQR, 0.20-0.59] in UPV animals and 1.01 [IQR, 0.75-1.59] in PV animals, p  = 0.01). Regional [11 C](R)-PK11195 uptake was independently associated with the interaction of regional tidal hyperinflation and regional lung compliance.

CONCLUSION: In an experimental model of ARDS, 4 h of UPV strategy significantly decreased lung inflammation, in relation to the control of VT -derived determinants of VILI.