Article Text


Post-traumatic hydrocephalus: influence of craniectomy on the CSF circulation
  1. Academic Neurosurgical Unit, Addenbrooke's Hospital, Cambridge CB2 2QQ, UK
  1. Dr Marek Czosnyka, Academic Neurosurgical Unit, Box 167, Addenbrooke's Hospital, Cambridge CB2 2QQ, UK email MC141{at}MEDSCHL.CAM.AC.UK

Statistics from

Post-traumatic ventricular dilatation may have a wide range of aetiological factors: starting from neuronal loss due to head trauma and possible secondary ischaemic insults, to obstruction of CSF circulation resulting in hydrocephalus. It is important to differentiate between post-traumatic hydrocephalus and brain atrophy before considering placement of a shunt. Making this decision can be facilitated by measurement of the resistance to CSF outflow.1 However, the pattern of the CSF circulation may change dramatically after a cranioplasty resulting from a previous decompressive craniectomy for refractory intracranial hypertension after head injury. The effect of the skull and dura on CSF hydrodynamics has been explored experimentally: the resistance to CSF outflow after craniectomy decreases twofold and brain compliance (expressed using the pressure-volume index, PVI) increases.2 This problem is important clinically as the following case illustrates:

CT and infusion studies. (A) Scan performed on admission. (B) After right frontal lobectomy and bone flap removal. (C) Four months after injury, before cranioplasty. (F) Infusion test demonstrated low resistance to CSF outflow and increased brain compliance. ICP=mean intracranial pressure; AMP= pulse amplitude of ICP waveform. Constant infusion rate of 1.5 ml/min is indicated by a thick horizontal line. X axis=time. (D) Five months after injury, after cranioplasty. (G) Infusion test demonstrated grossly increased resistance to CSF outflow and normal brain compliance. (E) One month after shunting: normalisation of ventricles. Bicauduate index decreased from 33% to 21% with a decrease in the 3rd ventricle diameter ( from 13 mm to 8 mm).

A 44 year old man fell downstairs and was admitted with a Glasgow coma score (GCS) of 4. Brain CT disclosed an intracerebral haematoma, which required a right frontal lobectomy and decompressive craniectomy to control raised intracranial pressure. Five months later he remained severely disabled with deteriorating GCS and increasing spasticity. Brain CT showed a progressive ventricular dilatation with widening of the cortical sulci. Cranioplasty had been delayed because of persistent problems with infections.

The first lumbar computerised infusion test3 was performed 5 months after injury to study the patient's CSF circulation. The opening pressure was low (5 mm Hg) with a very low pulse amplitude. An infusion of normal saline at a rate of 1.5 ml/min increased the intracranial pressure (ICP) to a plateau of 12.2 mm Hg within 22 minutes. The calculated resistance to CSF outflow was normal (5 mm Hg/ml/min) and the pressure-volume index was increased to 28 ml (figure).

It is important to mention here that the normal range for the pressure-volume index, calculated from the constant rate infusion (as an inverse of elastance coefficient3) is different from the values obtained by the bolus injection.1 Values below 13 ml indicate a tight brain, from 13 ml to 23 ml normal compliance, and above 23 ml hypercompliant brain. A slow constant rate infusion tests global compliance of the craniospinal axis whereas a fast bolus volume load probably tests compartmental compliance of the container into which the extra volume is added.

This pattern of CSF circulation with low or normal resistance to CSF outflow, increased brain compliance, and very few vasogenic waves is characteristic of cerebral atrophy.3

Cranioplasty was carried out as his deterioration was attributed to the “syndrome of trephined” where the brain sinks in, particularly with erect posture and dehydration producing deterioration in conscious level and focal signs. However, 1 month later, there had been no progress in the patient's condition and repeat CT again suggested progressive ventriculomegaly. The infusion study was repeated. The opening ICP was not dramatically different (10 mm Hg) to the previous study. However, the pulse amplitude (1.5 mm Hg) was increased, and the calculated resistance to CSF outflow was greatly increased to 20 mm Hg/ml/min, with a normal pressure-volume index of 15 ml. Such a pattern is specific for hydrocephalus. After this test the patient was shunted with a Codman Medos programmable valve (setting 120 mm H2O) ventriculoperitoneal shunt with remarkable clinical improvement, the GCS rose to 14, he began to talk and his spasticity in his arms decreased dramatically. It is obvious why the pressure-volume compensatory reserve (PVI) decreases after cranioplasty2, but the interpretation of an increase in the resistance to CSF is not immediately apparent. Two explanations are possible:

The patient had developed an acute hydrocephalus, possibly as a result of traumatic subarachnoid haemorrhage. Craniectomy was a factor allowing compensation of CSF circulation in the early stages. It is difficult to explain what is the nature of such compensation. Shapiroet al 2 attempted to offer an interesting but conceptually difficult hypothesis that the time constant (resistance to CSF outflow×compliance of cerebrospinal space) of cerebrospinal system hydrodynamics has a tendency to remain constant. Therefore, a mechanistic increase in compliance after craniectomy tends to be followed by a decrease in the resistance to CSF outflow. This process may be reversed after cranioplasty—that is, a decrease in PVI may be followed by an increase in the resistance to CSF outflow.

The second possible scenario is more important for clinical management. A large craniectomy may facilitate irreversible ventricular enlargement over weeks or months.4 Thus, after cranioplasty, the expanded ventricles may, via the cerebral mantle, obstruct the lumen of the cortical subarachnoid space and increase the resistance to CSF outflow.

This case demonstrates that when the CSF circulation is studied in patients with a large craniectomy the CSF outflow resistance cannot be taken reliably as a guide for shunting. Overnight ICP monitoring or CSF infusion study should be performed after cranioplasty, when CSF circulatory reserve decreases dramatically. Moreover, a prolonged period without a bone flap may encourage ventricular dilatation.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.