Abstract

In acute respiratory distress syndrome (ARDS), regional aeration is often gravity-dependent, with positive end-expiratory pressure (PEEP) recruiting the lung dorsally. Although recruitability can be assessed globally, our aim was to determine the impact of PEEP on regional recruitability and regional strain. To achieve a large representation of recruitability, we studied two preclinical porcine models of acute lung injury [(ALI) 19 symmetrical and 10 asymmetrical ALI], 20 patients with ARDS of mixed etiology (mixed ARDS), and 15 with COVID-19 ARDS. All study subjects underwent a single-breath derecruitment maneuver from high-to-low PEEP to quantify recruitability using the recruitment-to-inflation ratio (R/I). The regional effects of PEEP on strain were assessed using electrical impedance tomography (EIT). Symmetrical ALI animals had the highest R/I (1.39 [1.04-1.66]), followed by mixed ARDS (1.06 [0.70-1.23]), COVID-19 ARDS (0.66 [0.51-0.98]), and asymmetrical ALI (0.45 [0.22-0.85]). Dorsal regions had the highest recruitability (P = 0.001), and differences between dorsal and ventral regions were higher in recruitable subjects. Increasing PEEP decreased ventral dynamic strain (P < 0.01), with varying effects on dorsal dynamic strain. A paradoxical increase in dorsal dynamic strain associated with ventral hyperinflation could be observed across all groups, but more frequently in the less recruitable subjects. It was predicted by the EIT ventral-to-dorsal shift in ventilation normalized to the change in dorsal lung volume (P < 0.001). In animals and patients with varying recruitability, a higher global R/I is associated with a higher effect on the dorsal versus ventral R/I. PEEP can paradoxically increase dorsal strain due to ventral overdistention, and this is detectable by EIT.