By using proton nuclear magnetic resonance to measure urinary lactate: creatinine (L:C) ratio, it may be possible to make an early diagnosis of hypoxic–ischaemic encephalopathy and start neuroprotective treatment. Ninety eight babies were studied in Taiwan (New England Journal of Medicine1999;341:328–35) and the mean L:C ratio on urine collected within 6 hours of birth was 0.09 in 58 normal babies, 0.19 in 24 asphyxiated babies who did not develop clinical hypoxic–ischaemic encephalopathy, and 17.0 in 16 babies who did. An L:C ratio of 0.64 or greater was 94% sensitive and 100% specific for hypoxic–ischaemic encephalopathy.

Children with laryngeal symptoms often have gastro-oesophageal reflux (Journal of Pediatric Surgery1999;34:1053–6). In Montreal, Canada around 60% of young children with a diagnosis of stridor, laryngomalacia, or laryngitis had significant reflux on pH monitoring, and the symptoms resolved with medical treatment of the reflux in over 80% of them. Reflux was not common in children with recurrent otitis media.

Evidence of a neuronal basis to dyslexia continues to accumulate. A necropsy study in Boston, Massachusetts, USA (Annals of Neurology 1999;46:189–96) has shown that, whereas normal brains have larger neurons in the left primary visual cortex than in the right, dyslexic brains show no such asymmetry.

A study in Brussels (Journal of Pediatrics1999;135:34–8) has added to the evidence that smoking in pregnancy affects infant behaviour. Newborn babies and 3 month old infants of smoking mothers tolerated a greater sound intensity before arousing from sleep than did those of non-smoking mothers. The researchers suggest that this lack of arousal could be a factor in sudden infant death syndrome.

Long term follow up of 410 children in Pittsburgh, USA who underwent adenoidectomy or adenotonsillectomy for recurrent or persistent otitis media (Journal of the American Medical Association1999; 282:945–53) showed some improvement in the first year after adenotonsillectomy but the rate of postoperative morbidity, especially after adenotonsillectomy, led the authors to advise against either operation as a first resort.

A test developed at the Mayo Clinic (Lancet1999;354:901–5) may prove a useful general screening test for lipid storage disorders. Fibroblasts were incubated with a fluorescent analogue of lactosylceramide (called BODIPY-LacCer in short). In high concentrations the fluorescence changes from green to red, and in the sphingolipidoses the substance accumulates in lysosomes whereas in normal cells it is found in the Golgi apparatus. A positive test, therefore, consists of showing red lysosomes on fluorescence microscopy. On testing 26 lipid storage disease cell lines (10 different diseases) and 20 control cell lines the sensitivity was 96% and specificity 90%. At present the test requires special equipment but it may be possible to make it simpler and to do it on white blood cells rather than cultured fibroblasts.

Hepatitis C virus (HCV) infection acquired through blood transfusion may be less serious in children than in adults. In Munich, Germany (New England Journal of Medicine1999;341:866–70) 458 people who had been given blood during cardiac surgery as children (mean age 2.8 years) before the introduction of routine HCV screening for blood donors were assessed 10–27 years later. Sixty seven (15%) had HCV antibodies but only 37 (55%) of those had persisting HCV viraemia. The other 30 appeared to have cleared the infection spontaneously. Seventeen of the 37 had liver biopsy but only three showed evidence of progressive liver disease and all three had other possible causes (two with heart failure and one with prior hepatitis B infection). Since the introduction of HCV screening for blood donors, most childhood infection will be perinatal and the natural history of such infection remains unknown. It is also possible that longer follow up will reveal more cases of chronic liver disease in blood transfusion related infection.

Studies on transgenic mice may give a clue to human IgA nephropathy (Nature Medicine1999;5:1018–25). These mice are deficient in uteroglobin (UG), a steroid inducible, cytokine-like, anti-inflammatory, and immunomodulatory protein that is normally secreted by mucosal epithelial cells. They develop almost all the features of human IgA nephropathy. UG binds with fibronectin and the heteromer prevents fibronectin aggregation and inhibits the formation of IgA–fibronectin complexes and their binding to glomerular cells. It also prevented injected IgA being deposited in the glomeruli of UG deficient mice. Studies on the role of UG in human disease seem indicated.

Did you know that after a group A β haemolytic streptococcal infection some children become pale faced with black ears and black rings around their eyes, and develop a craving for bamboo shoots? Well they don't, but an unknown proportion of them might develop PANDAS (paediatric autoimmune neuropsychiatric disorders associated with streptococcal infections). This remains unproved as a clinicopathological entity but it is proposed as a possible example of molecular mimicry caused by streptococcal antibodies cross reacting with brain tissue. Now research at the US National Institutes of Health (Lancet1999;354:1153–8) has shown, in a highly selected group of 29 children fitting the definition of PANDAS, that plasma exchange or intravenous immunoglobulin may improve obsessive compulsive symptoms, and plasma exchange—but not immunoglobulin—may improve tic disorder. The authors emphasise that these treatments should not be given to other children with obsessive compulsive disorder or tic disorders except as part of approved clinical trials.

Five babies aged 51–127 days with complete Di George syndrome received allogeneic, cultured thymus tissue transplants and two survived (New England Journal of Medicine1999;341:1180–9). The thymus tissue was taken from 2–35 day old infants undergoing cardiac surgery and transplanted into the quadriceps muscles of the Di George syndrome babies. T cell proliferative responses developed after transplantation in four of the five infants. The two infants who survived received their transplants at ages 90 and 63 days and were alive and well at ages 5 years 9 months and 13 months, respectively.

Common acute lymphoblastic leukaemia (ALL) usually presents in children aged 2–4 years. In about a quarter of cases DNA analysis (but not conventional cytogenetic analysis) shows a translocation [t (12;21) (p 13; q 22)] fusing the TEL gene on chromosome 12 with the AML1 gene on chromosome 21. It has been shown in four pairs of monozygotic twins that this translocation is present neonatally and has arisen from a single cell clone in each pair. There must therefore have been a change in a single cell of one of the twins in utero with subsequent colonisation of the other twin. Now research in London and Milan (Lancet1999;354:1499–503; see also commentary,Ibid: 1486–7) has shown that at least six of 12 children who presented with ALL at age 2–5 and carried this translocation also had the translocation detectable in stored neonatal blood spots. The implication is that in many, possibly most, children with ALL the disease originates in utero. There is presumably a second event that occurs postnatally to determine the onset of clinical disease because the concordance of ALL in young monozygotic twins is only around 5%. What causes the translocation in utero or the progression to leukaemia in postnatal life is unknown.