The extracellular amino acid content was measured in the parietal cortex in portacaval and sham operated rats, using the brain dialysis technique. The amino acid content of the perfusate was determined for 10 min before and during stimulation with potassium chloride. Basal levels of aspartate, glutamine, glycine, methionie, valine, phenylalanine and leucine were 2-to 6-fold higher in the PC-shunted as compared to the sham operated rats. For glutamate, taurine, and GABA no differences were observed between the two groups. After KCl stimulation the release of glutamate and GABA increased significantly in both groups. For GABA this rise was approximately twice as high in the PC-shunted rats (+300%, P less than 0.01) as in the sham operated rats (+150%, P less than 0.01 as compared to basal). In the sham operated, but not in the PC-shunted rats, methionine and valine levels rose significantly (+200%, P less than 0.05) and glutamine release decreased (-50%, P less than 0.05). These findings suggest that the brain metabolism of amino acids is altered after a portacaval shunt. This could in turn alter the neurotransmission and partly explain the low spontaneous motor activity seen in these animals.

No abstract

Ventilation-perfusion relationships were studied in patients with non-alcoholic liver cirrhosis. Spirometry was essentially normal but the transfer factor of the lung (DLCO) was reduced by an average 34% of predicted. Arterial oxygen tension (PaO2) ranged from normal down to 6.9 kPa. Varying degrees of ventilation-perfusion (VA/Q) abnormalities (multiple inert gas elimination technique) were observed with increased dispersion of the perfusion distribution (log SDQ, 0.90; range 0.32-1.71; upper normal limit, 0.60) and the presence of both regions of low VA/Q ratios (between 0.1 and 0.005) (mean 4.1%; range 0-18.8%) and shunt (VA/Q ratios below 0.005) (mean 3.9%; range 0.19.8%). There was a close similarity between measured and calculated PaO2 in normoxaemic patients, but calculated values exceeded measured PaO2 in hypoxaemic patients. The difference between calculated and measured PaO2 correlated inversely to DLCO (r = 0.65, p less than 0.05). An inverse correlation was also noted between DLCO and the sum of shunt and low VA/Q regions (r = 0.87, p less than 0.001). It is concluded that hypoxaemia in non-alcoholic liver cirrhosis patients can be accounted for by intrapulmonary shunting and VA/Q mismatch, and possibly a "diffusion-perfusion" defect in patients with more severe gas exchange impairment.