Pulmonary pulsatility quantified by electrical impedance tomography in severe acute respiratory distress syndrome patients undergoing extracorporeal membrane oxygenation support

Background The cardiac-related pulsatility signal from electrical impedance tomography (EIT) correlates with stroke volume in mechanically ventilated patients with acute respiratory distress syndrome (ARDS). However, in swine models, regional pulsatility amplitude was also shown to increase with downstream flow obstruction. We aimed to investigate the relationship between regional pulsatility and pulmonary hemodynamics in a cohort of severe ARDS patients on veno-venous extracorporeal membrane oxygenation (ECMO). Methods We reanalysed data obtained from 20 ARDS patients receiving ECMO support. EIT was recorded 30 min after adjusting ECMO blood flow to target three ranges of mixed venous oxygen saturation (SvO(2)) (70-75 %; 75-80 %; >80%), applied in random order. Ventilation was protective with PEEP 15[12-16] cmH(2)O, Vt 4 +/- 1 ml/kg PBW. Quality of EIT tracings allowed for separate modelling of pulsatility and ventilation in 16/20 patients. EIT units were calibrated to millilitres of tidal volume (ml*) using synchronised tidal volume measurements, allowing for between-patient comparisons. Mixed-effects modelling was employed to account for repeated measurements. Results Across blood flow steps, pulsatility amplitude was directly related to stroke volume (SV) (beta = 0.28 (0.06 - 0.5) ml*/mL, p = 0.014) and systolic pulmonary artery pressure (PAPs) (beta = 0.47 (0.14 - 0.81) ml*/mmHg, p = 0.008) and inversely related to mixed venous oxygen tension (PvO(2)) (beta = -0.41 (-0.79--0.03) ml*/mmHg, p = 0.039). Changes in pulsatility had an 80% concordance rate with changes in SV and 83% with PAPs. At each ECMO blood flow step, there was a decrease in ventral lung pulsatility during inspiration (p < 0.01 for all steps). Moderate to strong correlations were observed between dorsal pulsatility and pulmonary artery pressure (rho = 0.72, p = 0.001 at Low SvO(2); rho = 0.59, p = 0.017 at Intermediate SvO(2); rho = 0.81, p < 0.001 at High SvO(2)). Conclusion n severe ARDS patients on ECMO, pulsatility amplitude reflects stroke volume changes induced by positive intrathoracic pressures and mixed venous saturation targets. However, downstream flow obstruction appears to be the leading determinant in the dorsal lung and may be useful to monitor right heart loading in patients with ARDS.