Abstract 1623

Introduction. Microdialysis allows measurements of metabolites e.g. glucose, lactate and glycerol in subcutaneous adipose tissue. The metabolites are analysed in small sample volumes, and the analysis is performed at bedside with immediate display of results. The aim of this investigation was to evaluate the reliability, safety and overall usefulness in neonates with hypoglycaemia.

Material and methods. Eight newborn infants (mean birth weight 3951 g, range 2720-5140) with pronounced hypoglycaemia (S-glucose < 2.2 mmol/L at > 6 h of age) were studied. The etiological backgrounds were IDM (n = 5), neonatal stress (n = 2) and panhypopitutarism (n = 1). A 30 mm double lumen plastic cannula with a semi-permeable membrane is implanted in the abdominal subcutaneous adipose tissue and perfused with 0.3 µL/min of Ringer's solution. The catheter mimics the function of a capillary blood vessel, and the concentrations in the outgoing dialysate reflect the corresponding levels in the adipose tissue. The microdialysis continued until the infant had stable glucose levels as judged by the attending physician (range 48 - 172 h). Each sample was collected over 30 min and correlated to intermittent blood samples. To investigate the dynamics of the procedure, e.g. the time period between insertion of the device until reliable glucose measurements were obtained, five healthy adult volunteers were also studied.

Results. All catheters were inserted with minimal trauma and without any complications. No local infections were observed. In one infant, the catheter was replaced due to suspiciously low readings. The inserting trauma results in an initial rise in glycerol and low glucose levels. The rise in glycerol were higher in infants than in adults, indicating that the inserting trauma was more severe in infants than in adults. Glucose and lactate seemed to be reliable after 4-6 h both in infants and in adults and glucose levels after 3-5 h. We analysed 649 microdialysis samples collected between 5-48 h after insertion of the catheters in infants obtaining various anti-hypoglycemic treatment modalities (glucose, somatostatin and glucagon). False low glucose concentrations were found in 18 samples (< 1 %). Five of these were due to sample volumes below a critical cut-off, while 13 samples contained no dialysate volume at all. All the newborn infants displayed an extreme variability in the glucose levels. There was a strong correlation, even in the blood glucose range below 2.5 mmol/L (n = 10), between dialysate and blood glucose levels (r=0.969, P<0.001).

Conclusion. Microdialysis is safe, atraumatic and reliable within a wide range of glucose concentration. The technique is especially useful for monitoring the effects of ongoing treatment.