The idea of removing almost half of a child’s brain sounds gross and capable of generating distinct psychological barriers to its use as treatment. Nevertheless, the operation has been performed for over half a century. Some 30 to 40 years ago it fell into disrepute because of the complications (obstructive hydrocephalus, superficial haemosiderosis, and intracranial haematoma) which occurred in up to a third of patients at that time. Improved surgical technique has, however, made the operation safe. It is now considered when seizures arise from one structurally abnormal hemisphere especially when there is hemiplegia and/or visual field defect. Cognitive and behavioural problems are common in these children. Children with Rasmussen syndrome without such disabilities may be considered. A report of 33 children operated upon at Great Ormond Street Hospital between 1991 and 1997 (

) has shown that hemispherectomy can give very good results.

The children varied in age from 4 months to 17 years (median 4.25 years) and were followed up for between 1 and 8 years. They were divided into three groups according to pathology; developmental (16; 10 hemimegencephaly, 2 polymicrogyria, 2 focal cortical dysplasia, 1 diffuse cortical dysplasia, 1 microdysgenesis), acquired (11; 6 middle cerebral artery infarct, 3 after encephalitis or trauma, 1 hemiconvulsion-hemiplegia-epilepsy, 1 perinatal ischaemia) and progressive (6; 4 Rasmussen encephalitis, 2 Sturge-Weber syndrome).

Overall, 17 children were seizure free after operation and only three had their seizures reduced by less than 75%. (Almost all had had many seizures daily before surgery.) Seizures stopped in 9/11 children with acquired pathology, 3/6 with progressive pathology, and 5/16 with developmental pathology. Freedom from seizures or >75% reduction in seizures was achieved in 10/11 (acquired), 6/6 (progressive), and 14/16 (developmental). Hemiplegia remained the same in 22 children, was worse in 6, and improved in 5. There was no significant cognitive deterioration after operation and four children showed cognitive improvement. None lost language. Twelve children had behavioural difficulties before operation and these difficulties improved postoperatively in eleven.

In a specialist centre hemispherectomy may benefit carefully selected children.