Parenteral nutrition is life saving in patients with intestinal failure but liver dysfunction is commonly encountered, especially in neonates. Although abnormal liver function tests associated with short-term parenteral nutrition are usually benign and transient, liver dysfunction in both children and adults receiving long-term parenteral nutrition can progress to end-stage liver disease and liver failure. The aetiology of parenteral nutrition-associated liver disease is complex and multifactorial, with a range of patient, disease and nutrition-related factors implicated. Sepsis is of particular importance, as is the lack of enteral nutrition and overfeeding with intravenous glucose and/or lipid. Deficiencies of a number of amino acids including choline and taurine have also been implicated. Management of hepatic dysfunction in parenteral nutrition should initially focus on preventing its occurrence. Sepsis should be managed appropriately, enteral nutrition should be encouraged and maximised where possible and parenteral overfeeding should be avoided. Provision of parenteral lipid should be optimised to prevent the adverse effects of both deficiency and excess, and cyclical rather than continuous parenteral feeding should be administered. There is some evidence of benefit in neonates from oral antibiotics to prevent intestinal bacterial overgrowth and from oral ursodeoxycholic acid, but less to support their use in adults. Similarly, data to support widespread use of parenteral choline or taurine supplementation are lacking at present. Ultimately, severe parenteral nutrition-associated liver disease may necessitate referral for small intestine and/or liver transplantation.

This work is aimed at demonstrating the influence of insulin in the triggering of hepatic cholestasis in young gilt-head sea bream larvae reared under experimental conditions. The results are based on an immunological assay of insulin and an ultrastructural study of the yolk syncytial layer and the liver. Compared to fed larvae, fasting gilt-head sea bream larvae displayed a substantial increase in insulin level correlated with dysfunction of the yolk syncytial layer and hepatic histopathology. This correlation was verified in larvae reared in sea water containing insulin. Ultrastructural observations suggested that insulin affects lipoprotein secretion by the yolk syncytial layer. The failure of plasmatic lipoproteins would then cause dysfunction of the biliary lipid secretion mechanisms and, hence, trigger hepatic cholestasis. These results in gilt-head sea bream show for the first time the influence of insulin in the triggering of hepatic cholestasis during the early development of a vertebrate.

Four hundred and thirty-four twins delivered from 220 women at Notre-Dame Hospital were studied during a period of 11 years (1969–1979). The maternal, fetal, and neonatal outcome was compared before and after 1974, the year ultrasonography and other changes in perinatal care were introduced in our institution. Early diagnosis occurred more frequently after 1974, together with increased antenatal hospitalization. Preeclampsia and hepatic cholestasis occurred in 19.5% and 9.5% of women, respectively. Preterm delivery occurred in 42.2% of the cases. The corrected fetal mortality rate was 30.2/1,000 and corrected perinatal mortality rate was 74.9/1,000. About 90% of the neonatal deaths occurred in infants born before 36 weeks. Internal version and complete extraction of the second twin increased the neonatal mortality sixfold compared to spontaneous delivery. Availability of ultrasound examination significantly reduced preterm delivery.