• basal ganglia;
  • glutamine;
  • high energy phosphates;
  • magnetic resonance spectroscopy

Abstract: Background/Aims: In vivo magnetic resonance spectroscopy can be used to study cerebral metabolism non-invasively. We aimed to correlate 1H and 31P magnetic resonance spectral abnormalities in the brains of patients with subclinical hepatic encephalopathy. Methods: Eighteen patients were studied at 1.5T, with combined 1H and 31P magnetic resonance spectra obtained from multiple voxels in the cerebral cortex and basal ganglia. Peak area ratios of choline, glutamine/glutamate, relative to creatine in the 1H spectra and percentage phosphomonoesters, phosphodiesters and βNTP signals relative to total 31P signals in the 31P spectra were measured. Results: Six patients did not complete the full examination –31P results are available from 12 patients only. Relative to creatine, there were reductions in choline and elevations in glutamine/glutamate, varying across the brain with choline significantly reduced in occipital cortex (p<0.05) and glutamine/glutamate most significantly elevated in temporo-parietal cortex (p<0.0001). Percentage phosphomonoester (p<0.05), phosphodiester (p<0.05) and βNTP (p<0.005) signals were significantly decreased in basal ganglia spectra. No correlation was found between the magnitude of 1H and 31P MRS changes, except between percentage phosphodiester decrease and glutamine/glutamate to creatine increase in occipital cortex. Conclusion: The results of this study point to a multifactorial aetiology for this condition.