Fasting Blood Ammonia Predicts Risk and Frequency of Hepatic Encephalopathy Episodes in Patients With Cirrhosis

BACKGROUND & AIMS: There is controversy over the use of measuring blood levels of ammonia (NH3) in the management of patients with overt hepatic encephalopathy (HE). METHODS: We performed a retrospective analysis of data from a randomized, double-blind study of 178 patients with cirrhosis given glycerol phenylbutyrate (an NH3-lowering agent) or placebo for 16 weeks. Blood samples were collected at baseline and on study days 7 and 14 and NH3 levels were measured. The probabilities of having an HE episode, based on ammonia values, were modeled using binary logistic regression. A Cox proportional model was used to determine the risk of HE episodes in patients with baseline fasting NH3 levels <= 1.5-fold the upper limit of normal (ULN) versus patients with fasting NH3 levels >1.5-fold the ULN. RESULTS: The risk and frequency of HE episodes and HE-related hospitalizations correlated with baseline (mean, 51 +/- 6 mu mol/L; ULN, 35 mu mol/L) and on-study fasting levels of NH3, and increased sharply at levels >1.5-fold the ULN. Regardless of baseline level, NH3 exposure and the relative risk of HE episodes were decreased by glycerol phenylbutyrate. CONCLUSIONS: In analysis of data from a phase 2 study of the effects of glycerol phenylbutyrate in patients with cirrhosis, we found that fasting levels of NH3 in blood can identify patients at risk for HE-related morbidity. Patients with HE might benefit from NH3-lowering therapy.