BACKGROUND/AIMS To measure cerebral benzodiazepine receptor binding using 11C-flumazenil positron emission tomography in patients with stable chronic hepatic encephalopathy, who were also characterised by proton magnetic resonance spectroscopy. METHODS Six abstinent patients of mean age 61 years with alcohol related cirrhosis and grade I–II hepatic encephalopathy and 11 matched healthy volunteers were studied. Each patient's encephalopathy was defined according to clinical, psychometric, electroencephalographic, and magnetic resonance spectroscopy criteria. Using positron emission tomography, the brain volume of distribution of 11C-flumazenil was obtained; this reflects benzodiazepine receptor availability. Proton magnetic resonance spectra were acquired at 1.5 T using a multivoxel technique; peak area ratios were calculated for choline, glutamine/glutamate, N-acetylaspartate, and creatine resonances. RESULTS The mean volume of distribution of 11C-flumazenil was significantly higher in the cortex, cerebellum, and the basal ganglia in the patients compared with controls (p<0.001). In the patient group, the mean glutamine/glutamate to creatine ratio was significantly increased and the mean choline to creatine ratio was significantly decreased in all brain areas, compared with healthy volunteers. However, the N-acetylaspartate to creatine ratio was unchanged compared with controls. CONCLUSIONS The spectroscopy results reflect the cerebral metabolic derangement associated with hepatic encephalopathy. Stable grade I–II chronic hepatic encephalopathy in alcohol related cirrhosis may be associated with increased cerebral benzodiazepine receptor availability. However, a direct effect of previous chronic exposure to alcohol cannot be excluded. |