CSF shunt removal in children with hydrocephalus.

The possibility to remove a previously inserted CSF shunt device in hydrocephalic children is a well known though rare event for paediatric neurosurgeons. A retrospective analysis of our experience with a series of 850 children affected by non tumoral hydrocephalus shows that obvious shunt independence could be demonstrated in 3.2% of the patients (27 cases). The time interval between the CSF shunt insertion and removal ranged between 8 months and 12 years (mean: 8 years). Parameters analysed to search for any predictive elements were age at surgery, aetiology, type of prosthesis utilised, time interval between insertion and removal of the shunt, number of the possible revisions. The results of the study suggest that the highest incidence of shunt independence is reached in subjects operated on in early infancy, as 24 of 27 removed shunts were in patients operated on under 6 months of age, and the remaining in 2 children treated when less than 2 years old. Such a finding could be explained on the grounds of a delay in maturation of the CSF absorption mechanisms followed by a late normalisation in these patients. As regards to aetiology, 41% of the 27 patients considered in this series were affected by a post-haemorrhagic hydrocephalus, which was progressive, as demonstrated by serial neuroradiological examinations and echo-Doppler cerebral studies at the time of the surgical treatment. In six children the hydrocephalus was associated with myelomeningocele. Five patients had aqueduct stenosis and 2 communicating hydrocephalus. The types of CSF shunting system we utilised did not play any role in determining or facilitating shunt independence. No correlation was observed with the need and the number of shunt revisions. The role of the interval time between the insertion and the removal of the shunt was not analysable, because of the possible acquisition of the shunt independence prior to its demonstration at the moment of the surgical revision of the CSF shunt (elective lengthening because of the physiological body growth) or to the radiological demonstration of CSF shunt device disconnection. The same constraint prevents the evaluation of the actual overall incidence of shunt independence in shunted hydrocephalic children, as some of them could have harboured a non-functioning CSF shunt device, though unnoticed.