Flumazenil for hepatic coma in patients with liver cirrhosis: an Italian multicentre double-blind, placebo-controlled, crossover study.

Several factors suggest that endogenous benzodiazepines and gamma-amino-butyric acid may be involved in pathophysiology of hepatic encephalopathy (HE). Contrasting opinions exist on the therapeutic efficacy of flumazenil in the treatment of HE. This study was planned to assess the efficacy of flumazenil by a double-blind, placebo-controlled, crossover design in a large and selected population of cirrhotic patients in stage 4a HE admitted to intensive care units over a 4-year period. Out of 236 patients selected for the study, 132 received flumazenil, whereas 131 patients received placebo. Improvement of the neurological score was documented in 31 patients (23%) of flumazenil group and in two patients (1.5%) of placebo group (p < 0.001) during the first study period, whereas during the crossover period, improvement of the neurological score was documented in seven patients (5.3%) of the flumazenil group and in none of the placebo group (p = 0.022). Improvements in EEG tracings were observed in 44 patients (33.3%) of flumazenil group and in five patients (3.8%) of placebo group (p < 0.001) during the first study period; during the crossover period, improvements in EEG tracings were observed in 10 patients (7.5%) of the flumazenil group and in two patients (1.5%) of the placebo group (p = 0.040). The presence of benzodiazepines was detected in the serum of three responders and in two non-responders. The presence of diazepam and NN-desmethyl diazepam was documented in two responders; prior intake of synthetic diazepam was later confirmed in these patients. The results of our study suggest that flumazenil is beneficial only in a selected subset of cirrhotic patients with severe HE; the applicability of this treatment to unselected patients with hepatic coma or to cirrhotic patients with less severe HE still remains to be determined.