(1155) Additive Prognostic Value of Lung Inhomogeneity in Patients With ARDS

Introduction: In acute respiratory distress syndrome (ARDS), lung inhomogeneity acting as “stress concentrators” and yielding heterogeneous distribution of lung stress has been recognized as a key mechanism contributing to ventilator induced lung injury (VILI) and increased mortality, besides transpulmonary driving pressure (DPL) reflecting global Lung stress. However, each component's relative contribution is unclear. This study aimed to address whether and to what extent lung inhomogeneity would add to DPL in estimating the risk of mortality in patients with ARDS.

Methods: In this prospective study from two university hospitals, patients with ARDS receiving invasive mechanical ventilation and monitored by esophageal manometry were investigated. Respiratory mechanics including DPL were measured early after intubation at tidal volume (Vt) of 6.5 ± 0.8 ml/kg of predicted body weight (PBW). Lung inhomogeneity was estimated by the difference (PL,ER-ES) between inspiratory transpulmonary pressure (PL,IN) calculated from the elastance ratio (PL,ER) and esophageal pressure, reasonably reflecting the intraparenchymal difference in PL,IN between nondependent and dependent lung regions. The primary outcome was mortality at 60 days.

Results: Of 181 patients available for analysis, 63 (34.8%) died by Day 60. On multivariate Cox regression analysis, DPL (adjusted hazard ratio [HR],1.134; 95% CI, 1.040-1.237; P=0.004) and PL,ER-ES (adjusted HR,1.431; 95% CI, 1.288-1.590; P< 0.001) were independently associated with 60-day mortality after adjustment for Sequential Organ Failure Assessment (SOFA) (P=0.007) and Vt/PBW (P=0.13). In subgroup of patients with a lower DPL ( < 10), only PL,ER-ES (adjusted HR, 1.59; 95% CI, 1.336-1.915; P < 0.001) remained independently predictive of 60-day mortality. The receiver operating characteristics (ROC) analyses showed that a multifunctional index, incorporating DPL and PL,ER-ES, presented significantly higher area under ROC curve than DPL alone (0.82 vs 0.76, P=0.03) for predicting 60-day mortality.

Conclusions: In patients with ARDS, Lung inhomogeneity assessed by PL,ER-ES adds to DPL in estimating the risk of mortality (and then of VILI), with the impact of the PL,ER-ES on mortality being three times as large as that of the DPL.