suMMARY A fatal case associated with severe hyperammonaemia is described in which no urea cycle enzyme deficiency could be found. This prompted further investigation of blood ammonia levels in neonates admitted to the premature baby unit at Hammersmith Hospital. 102 specimens were taken from 42 babies within the first 3 weeks of life; the babies had a variety of clinical conditions. The mean ammonia level was 94.5 ,mol/l (132.3 ±g/ 100 ml) (range 32-255 ,umol/l (44.8-357 jig/ 100 ml), SD ± 41.0). These results, although higher than the range for older babies in hospital,were not as high as in the baby with severe hyperammonaemia. Serial levels in 10 babies suggested that the range of blood ammonia levels was greatest in the first 2 weeks of life and narrowed considerably after this period. Great care is needed in collecting blood samples and measuring them if accurate results are to be obtained.In 1977 a fatal case associated with severe hyperammonaemia was seen at Hammersmith Hospital. A plasma ammonia level taken 3 hours before death was 3000 ,umol/l (4200 ,ug/100 ml) and cerebrospinal fluid taken shortly before death showed a level of 4000 tumol/l (5600 Fg/100 ml). A liver biopsy was taken immediately after death, but no enzyme defect of the urea cycle could be found, and investigations for organic aciduria did not show an inborn error in branch-chain amino-acid metabolism. The clinical details are given below.Three cases of severe hyperammonaemia in the neonatal period were reported by Pollack et al.l and 5 more were reported by Ballard et al.2 In none of these babies was an enzyme defect observed, and no explanation for the cause could be found, but it was suspected that hyperammonaemia in the neonatal period might not be a rare occurrence.
Case historyA boy was born normally to a 22-year-old primigravida at 35 weeks' gestation after spontaneous onset of labour. The Apgar score at one minute was 9. The birthweight was 2-58 kg (10th centile) and, apart from slight grunting, no abnormalities were noted initially. Milk feeds were started via nasogastric tube at 2 hours. The grunting continued for 14 hours and at this stage the baby was noted to be irritable but no cause for this was apparent.At 24 hours his skin became mottled and he developed gasping respirations; blood sugar, urea and electrolytes, calcium, magnesium, and cerebrospinal fluid were all normal. The only abnormal finding was a metabolic acidosis. At 26 hours the baby had generalised convulsions which could not be controlled satisfactorily. In seeking an explanation for the deteriorating condition, a metabolic disorder was considered and a plasma ammonia level estimated with the result already mentioned above.However, the urea cycle enzyme activities measured in the liver biopsy (taken shortly after death) by the method of Brown and Cohen,3 as described by Levin,4 did not show any defect. The analysis of urine for organic acids by gas chromatographic and mass spectrometric methods of Chalmers et al.5 also failed to show an inborn error of orga...