Abstract
Although there have been several reports regarding the association between hypoxic hepatic injury and clinical outcomes in patients who underwent conventional cardiopulmonary resuscitation (CPR), limited data are available in the setting of extracorporeal CPR (ECPR). Patients who received ECPR due to either in- or out-of-hospital cardiac arrest from May 2004 through December 2018 were eligible. Hypoxic hepatitis (HH) was defined as an increased aspartate aminotransferase or alanine aminotransferase level to more than 20 times the upper normal range. The primary outcome was in-hospital mortality. In addition, we assessed poor neurological outcome defined as a Cerebral Performance Categories score of 3 to 5 at discharge and the predictors of HH occurrence. Among 365 ECPR patients, 90 (24.7%) were identified as having HH. The in-hospital mortality and poor neurologic outcomes in the HH group were significantly higher than those of the non-HH group (72.2% vs. 54.9%, p = 0.004 and 77.8% vs. 63.6%, p = 0.013, respectively). As indicators of hepatic dysfunction, patients with hypoalbuminemia (albumin < 3 g/dL) or coagulopathy (international normalized ratio > 1.5) had significantly higher mortalities than those of their counterparts (p = 0.005 and p < 0.001, respectively). In multivariable logistic regression, age and acute kidney injury requiring continuous renal replacement therapy were predictors for development of HH (p = 0.046 and p < 0.001 respectively). Furthermore age, arrest due to ischemic heart disease, initial shockable rhythm, out-of-hospital cardiac arrest, lowflow time, continuous renal replacement therapy, and HH were significant predictors for in-hospital mortality. HH was a frequent complication and associated with poor clinical outcomes in ECPR patients.