BACKGROUND & AIMS: Hepatic encephalopathy (HE) is now thought to be caused by cerebral oedema although the precise pathogenesis is uncertain. We hypothesised that if ammonia is a key factor, induced hyperammonaemia would lead to transient changes in brain water distribution and metabolite concentration, detectable by diffusion tensor imaging (DTI) and magnetic resonance spectroscopy (MRS). METHODS: Thirteen cirrhotic patients being evaluated for liver transplantation were challenged with 54 g of equal parts of threonine, serine, and glycine. Conventional magnetic resonance imaging was performed to exclude structural lesions and localise regions of interest. DTI was used to generate white matter apparent diffusion coefficient (ADC) maps and proton MRS to measure brain metabolite concentrations before and after the challenge. RESULTS: The challenge caused a mean (±SD) rise in blood ammonia of 58 (±41) μmol/L, which was accompanied by a significant 9% increase in ADC (p=0.004). Increased ADC significantly correlated with blood ammonia (r=0.58, p=0.04). The change in ammonia levels also correlated with the increase in glutamine levels (r=0.78, p=0.002). Myo-inositol concentration decreased significantly by 0.7 (±0.7)mMol/L between scans and this correlated with the mean difference in ADC (r=0.59, p<0.04). CONCLUSIONS: These results show that ammonia can directly drive changes in brain water distribution as a mechanism for cerebral oedema development. Since cerebral astrocytes contain glutamine synthetase, our MRS data suggest intracerebral formation of glutamine from ammonia. The rapid decrease in myo-inositol indicates that this organic osmolyte plays a protective role in HE by release from astrocytes in order to maintain cell volume.
BACKGROUND & AIMS: Hepatic encephalopathy (HE) is now thought to be caused by cerebral oedema although the precise pathogenesis is uncertain. We hypothesised that if ammonia is a key factor, induced hyperammonaemia would lead to transient changes in brain water distribution and metabolite concentration, detectable by diffusion tensor imaging (DTI) and magnetic resonance spectroscopy (MRS). METHODS: Thirteen cirrhotic patients being evaluated for liver transplantation were challenged with 54 g of equal parts of threonine, serine, and glycine. Conventional magnetic resonance imaging was performed to exclude structural lesions and localise regions of interest. DTI was used to generate white matter apparent diffusion coefficient (ADC) maps and proton MRS to measure brain metabolite concentrations before and after the challenge. RESULTS: The challenge caused a mean (±SD) rise in blood ammonia of 58 (±41) μmol/L, which was accompanied by a significant 9% increase in ADC (p=0.004). Increased ADC significantly correlated with blood ammonia (r=0.58, p=0.04). The change in ammonia levels also correlated with the increase in glutamine levels (r=0.78, p=0.002). Myo-inositol concentration decreased significantly by 0.7 (±0.7)mMol/L between scans and this correlated with the mean difference in ADC (r=0.59, p<0.04). CONCLUSIONS: These results show that ammonia can directly drive changes in brain water distribution as a mechanism for cerebral oedema development. Since cerebral astrocytes contain glutamine synthetase, our MRS data suggest intracerebral formation of glutamine from ammonia. The rapid decrease in myo-inositol indicates that this organic osmolyte plays a protective role in HE by release from astrocytes in order to maintain cell volume.