Lung Injury and Its Prognostic Significance in Acute Liver Failure

Abstract

Hypoxemia is a feared complication of acute liver failure, and high oxygen requirements will frequently lead to removal of patients from the transplant list. As data regarding the prevalence and outcome of acute respiratory distress syndrome in acute liver failure are scant and hypoxemia being a commonly encountered systemic complication, we analyzed radiological, gas exchange, and ventilator data in consecutive patients admitted with acute liver failure. Acute liver failure patients receiving mechanical ventilation admitted between January 2007 and February 2011 were included. Patients were categorized according to the Berlin definition as: no acute respiratory distress syndrome, acute respiratory distress syndrome (PaO2/FIO2 < 300 mm Hg), and subdivisions of mild, moderate, and severe acute respiratory distress syndrome (200-300 mm Hg, 100-200 mm Hg, and < 100 mm Hg, respectively). Chest radiographs were independently assessed by two observers for the presence or absence of acute respiratory distress syndrome. Absence of left atrial pressure elevation was based on combined hemodynamic and echocardiographic assessment. Two hundred acute liver failure patients were admitted during the study period of whom 148, median age 39 years (16-74 yr), were included. Thirty-one (21%) had acute respiratory distress syndrome (17 mild acute respiratory distress syndrome [12%], 9 moderate acute respiratory distress syndrome [12%], and 5 severe acute respiratory distress syndrome) within the first 72 hours following admission. Acute respiratory distress syndrome patients required higher positive end-expiratory pressure (7 vs 6 vs 10 vs 15 cm H2O for no, mild, moderate, or severe acute respiratory distress syndrome, p = 0.014), had reduced respiratory system compliance (34 vs 29 vs 30 vs 23 L/cm H2O, p = 0.028), and an increased number of ventilator days (no acute respiratory distress syndrome, 10 d; mild acute respiratory distress syndrome acute lung injury, 12 d; moderate acute respiratory distress syndrome, 23 d; severe acute respiratory distress syndrome, 22 d; p = 0.097). Duration of liver intensive therapy unit stay (p = 0.175), survival (p = 0.877), inotrope requirements (p = 0.495), need for extracorporeal renal support (p = 0.565), and severity of organ failure scores were not affected. Extravascular lung water index had a moderate sensitivity of 65% and specificity of 77% for the prediction of acute respiratory distress syndrome. The prevalence of lung injury is relatively low in acute liver failure, where 21% fulfilled acute respiratory distress syndrome criteria. Overall presence of acute respiratory distress syndrome appeared to have a limited impact on outcome.