Manganese toxicity in children receiving long-term parenteral nutrition.

BACKGROUND: In patients receiving long-term parenteral nutrition (PN), cholestatic disease and nervous system disorders have been associated with high blood concentrations of manganese. In such patients, the normal homoeostatic mechanisms of the liver and gut are bypassed and the requirement for this trace element is not known; nor has it been certain whether hypermanganesaemia causes the cholestasis or vice versa. We explored the direction of effect by serial tests of liver function after withdrawal of manganese supplements from children receiving long-term PN. We also examined the relation between blood manganese concentrations and brain lesions, as indicated by clinical examination and magnetic resonance imaging (MRI). METHODS: From a combined group of 57 children receiving PN we identified 11 with the combination of hypermanganesaemia and cholestasis; one also had a movement disorder. Manganese supplements were reduced in the first three and withdrawn in the remainder. MRI was done in two of these children. We also looked at manganese concentrations and MRI scans in six children who had received PN for more than 2 years without developing liver disease. FINDINGS: In the hypermanganesaemia/cholestasis group, four of the 11 patients died. In the seven survivors baseline whole-blood manganese was 615-1840 nmol/L, and after 4 months it had declined by a median of 643 nmol/L (p < 0.01). Over the same interval total bilirubin declined by a median of 70 mumol/L (p < 0.05). Two of these children had movement disorders, one of whom survived to have an MRI scan; this showed, with T1 weighted images, bilateral symmetrically increased signal intensity in the globus pallidus and subthalamic nuclei. Such changes were also seen in five other children--one from the hypermanganesaemia/cholestasis group and four of six in the long-term PN group without liver disease (in all of whom blood manganese was above normal). INTERPRETATION: The cholestasis complicating PN is multifactorial, but these results add to the evidence that manganese contributes. In view of the additional hazard of basal ganglia damage from high manganese levels in children receiving long-term PN, we recommend a low dose regimen of not more than 0.018 mumol/kg per 24 h together with regular examination of the nervous system.