The efficacy of the benzodiazepine antagonist flumazenil has been assessed clinically in a double blind, randomised, placebo-controlled multicentre study in patients with grade I-III portosystemic encephalopathy. In an ancillary study reported here the effect of flumazenil on the electroencephalogram (EEG) was analysed in 32 patients who had EEG grading according to protocol. Following the baseline observation period, patients were randomised to receive (at 1 min interval) 3 sequential bolus injections of flumazenil (0.4, 0.8 and 1 mg) or placebo followed by infusions of flumazenil (1 mg/h) or placebo for 3 h. Patients were monitored for 5 h after infusion. A positive response was defined as 1 point improvement in EEG grade. After independent analysis of the EEG gradings 5 out of 17 (29%) flumazenil treated patients showed an improvement in EEG grading (3 after bolus, 2 during follow-up) compared to 2 out of 15 (13%) placebo treated patients (1 after bolus, 1 during follow-up) (95% confidence interval of difference: -12% to + 50%). Of the 5 EEG responders after flumazenil, 3 also had an improvement in clinical PSE grading (none after bolus, 2 during infusion, 1 during follow-up), compared to neither of the 2 EEG responders after placebo. EEg responders did not differ from non-responders with respect to Child-Pugh score, basal EEG, PSE grade and positivity for benzodiazepines. In conclusion, treatment perspectives for flumazenil in portosystemic encephalopathy appear to be present for only a minority of patients; however, this study yields no support for a major role of benzodiazepine antagonists in the treatment of hepatic encephalopathy.
BACKGROUND: Portal systemic encephalopathy (PSE) is a complex neuropsychiatric syndrome associated with hepatic failure. Small scale studies have shown the benzodiazepine receptor antagonist flumazenil to be effective in ameliorating PSE.
AIMS: To determine the efficacy of flumazenil in patients with non-comatous mild to moderate PSE (stages I to III) due to severe chronic liver disease.
PATIENTS: 49 male and female adults without symptoms of severe bleeding and sepsis and who screened negative for benzodiazepine in both blood and urine, were included in the study.
METHODS: Patients were randomised to receive either three sequential bolus injections of flumazenil (0.4, 0.8, and 1 mg) or placebo at one minute intervals, followed by intravenous infusions of either flumazenil (1 mg/h) or placebo for three hours. Clinical PSE grading and vital signs were assessed hourly during baseline and post-treatment periods and half hourly during treatment. The main outcome measures were improvement in group average PSE score and reduction of two points in individual PSE score (clinically relevant improvement).
RESULTS: The mean average improvement in the PSE score in the subjects treated with flumazenil was not statistically significantly different from placebo. However, for patients showing clinically relevant improvement, the difference between flumazenil and placebo was statistically significant (seven of 28 v none of 21; p = 0.015). Flumazenil was well tolerated.
CONCLUSIONS: A subgroup of patients with PSE resulting from chronic liver disease may benefit from the administration of flumazenil.
