Introduction Hyperammonaemia occurs in patients with advanced cirrhosis, which is also associated with a major redistribution of blood flow (hyperdynamic circulation). Previous investigations have sought to explain hyperammonaemia on the basis of metabolic adaptation (enzymatic derangements). We hypothesised that hyperammonaemia is centrally linked to the haemodynamic disturbances caused by portal hypertension. To test this hypothesis, we developed a theoretical model, which predicted arterial ammonia levels when organ blood flow is modulated.

Methods Assumptions on individual organ fluxes of ammonia (across the gut, liver, muscle, brain and kidney) were based on published arterio-venous differences and tracer kinetic data. In order to study the role of organ blood flow in isolation, we assumed that hepatic detoxification function and ammonia production were normal. A wide range of conditions was investigated (increased cardiac output, splanchnic vasodilation, low to high porto-systemic shunt fraction). In addition, we used scenarios of organ blood flow corresponding to Child Pugh A, B and C. Finally, we considered the effect of lowering ammonia release in the renal vein.

Results Hyperammonaemia developed when the fraction of gastrointestinal blood shunted was more than 65%. The influence of redistribution of organ blood flow accounted for <10% of the increase in arterial levels. Decrease in ammonia release in the renal vein may be sufficient to compensate for the increase in arterial ammonia due to shunted flow in Child Pugh A and B, but is insufficient in stage C.

Conclusion Portosystemic shunting may be centrally involved in the generation of hyperammonaemia in liver cirrhosis. This can be partially attenuated by increased urinary ammonia excretion. In patients with cirrhosis, decreased hepatic functions and metabolic changes may further increase the arterial ammonia levels.

Competing interests None declared.