The effects in man of liver disease, renal failure and hepatic microsomal enzyme induction on the elimination kinetics of antipyrine in saliva have been examined. Antipyrine (10 mg/kg) was given orally and assayed in saliva by gas-liquid chromatography. The mean antipyrine half-life from saliva in nine epileptic subjects receiving long term anticonvulsant drug therapy (6 hr +/- 0-9 SEM) was significantly shorter than in twenty normal healthy volunteers (10-7 +/- 0-6). Therapy included phenytoin and phenobarbitone, two drugs known to induce hepatic microsomal enzymes. Five subjects with chronic renal failure exhibited no significant difference in salivary anti-pyrine half-life (11-7 +/- 1-9) compared to the control group, whereas six subjects with chronic liver disease and impaired hepatic function had significantly increased half-life values (42-4 +/- 10). The results suggest that differences in the activity of hepatic microsomal enzymes are reflected by changes in salivary antipyrine elimination kinetics. Chronic renal failure appeared to have no effect on the function of these enzymes.

Antipyrine kinetics were measured on saliva in eight anaesthetists during a period when they were giving general anaesthetics and a period when they were working exclusively in intensive care. During the anaesthesia period there was a reduction in antipyrine half-life and the clearance of antipyrine increased. A nalysis of the data in groups failed to detect these changes because of the wide variation in metabolism between subjects. Exposure to anaesthetic agents under non-scavenging operating theatre conditions appears to enhance hepatic metabolism.

Isolated liver perfusion is a useful technique for the study of a variety of metabolic functions of the liver (Bartosek, Guaitani and Miller, 1973). The composition of the perfusate varies from laboratory to laboratory, but the one which is most widely used contains erythrocytes suspended in a plasma substitute, such as Krebs-Henseleit solution (Krebs and Henseleit, 1932), which may be supplemented vvath albumin and organic nutrients. Studies in our laboratory, using a perfusate such as this, have led to an examination of the [HCO.^'] of Krebs-Henseleit solution. We have found that the nominal [HCO3'] of Krebs-Henseleit, which is designed to approximate rat plasma, is inconsistent with the [HGOg'] measurable in this artificial physiological medium.Solutions prepared according to the formula of Krebs-Henseleit (Table 1) were gassed with carbogen of appropriate O2/CO2 content and analysed for [HCO3'] using the following four methods: