ULTRAPEEP: Lung ultrasound for the assessment of lung recruitment during esophageal pressure-guided PEEP in ARDS

Background: Whereas the importance of low tidal volume to avoid ventilator-induced lung injury (VILI) in patients with ARDS is well known, several uncertainties still exist regarding how to set positive end-expiratory pressure (PEEP). Many approaches have been considered, but no one showed a clear effectiveness in terms of outcome. Recently a ventilator strategy using esophageal pressure to estimate the transpulmonary pressure has been proposed by Talmor and colleagues. Although they found an improvement in arterial oxygenation, it was not explored whether the increase in oxygenation was due to lung recruitment. Objectives: The aims of this study were: i) to evaluate the differences between the PEEP level and the end-expiratory transpulmonary pressure (PLEEO) by setting PEEP according to ARDS Network and to esophageal pressure (Pes); ii) to assess whether the Pes-guided PEEP is associated with an improvement in oxygenation; iii) to show that Pes-guided PEEP increases lung recruitment estimated by lung ultrasound score (LUS); iv) to determine whether the new setting of PEEP is associated to a change in plasmatic cytokines as markers of VILI. Methods: 15 patients with moderate and severe ARDS were enrolled. For the first 2 hours, PEEP was set according to the Acute Respiratory Distress Syndrome Network standard-of-care recommendations (phase A). It was then adjusted according to measurements of Pes for the following 2 hours (phase B) to maintain a positive PLEEO. PEEP levels were finally returned equal to phase A for the last 2 hours (phase C). The primary end point was the improvement in lung recruitment assessed with lung ultrasound at the end of phase B. Measurements and Main Results: Lower PEEP levels and PLEEO were found during phase A compared to phase B [median PEEP was 10 [10.0-12.0] vs 16 cmH2O [15.5-19.5]; median PLEEO was -3 [-4.0 to -0.5] vs 2 cmH2O [2-4], p < 0.02]. Arterial oxygenation improved in phase B compared to phase A [median PaO2:FiO2 during phase A was 149 [120 to 166] vs 166 [153 to 177] during phase B, p < 0.05]. The median LUS was 18 during phase A [11 to 27] and decreased to 15 [11 to 18] during phase B. A decrease in LUS ≥4 (indicating lung recruitment) was found in 10 of 15 patients. Although differences in oxygenation and LUS from phase B to phase C were not significant, the clinical parameters at the end of phase C tended to be similar to phase A. The analysis of plasmatic cytokines revealed no difference between phases. Conclusions: In patients with moderate and severe ARDS, PEEP-induced LUS reduction and increase in oxygenation seems to indicate that setting PEEP according to the Pes-guided method results in a greater alveolar recruitment than setting PEEP according to ARDS Network strategy. Further investigations are needed to confirm these results and to exclude the presence of PEEP-induced alveolar overdistension.