The understanding of cerebrospinal fluid (CSF) dynamics was constantly refined for the last three decades [1-6].
In summary, CSF is physiologically produced by active secretion from cerebral arterial blood. The major site of this process is the choroid plexuses of the ventricular system , but the extrachoroidal production of CSF is responsible for a significant amount of the total CSF formation . CSF production rate can be assessed by means of continuous CSF drainage techniques  or perfusing the subarachnoid spaces with a tracer and subsequently measuring its dilution . Unfortunately, both techniques have the limitation of averaging all possible dynamic components of CSF production. Therefore, although the normal average rate of CSF production in human beings is accepted to be 0.35 mL min-1, very little is known about dynamic changes in its production rate and their clinical significance. CSF secretion is supposed to be proportional to brain metabolism and tends to decrease with age .
CSF flows from the lateral and third ventricles through the aqueduct of Sylvius to reach the fourth ventricle. Passage through the narrow aqueduct is pulsatile, and its flow velocity can be detected by phase-contrast magnetic resonance imaging (MRI) techniques. The study of the characteristics of CSF pulsations is promising, and several MRI techniques have gained increasing interest for the modelling and the diagnosis of CSF dynamics disturbances .
Normally, CSF exits the fourth ventricle through the foramina of Magendie and Luschka, flows freely through the basal cisterns upwards towards the superior sagittal sinus and downwards towards the lumbar subarachnoid space. The spinal subarachnoid space accounts for a significant part of the compensatory reserve of the system, with a compliant venous network that can accommodate acute changes in intracranial volumes displacing CSF caudally. This volume-buffering mechanism is based on the free circulation of CSF fluid cancelling out all intracranial pressure (ICP) gradients and therefore protecting the brain from the risk of herniation. If the normal pathways of CSF circulation are viable - such as in the case of communicating forms of hydrocephalus - patients can tolerate acute rises in ICP up to 60 mmHg without any subsequent adverse effect.
Re-absorption of CSF fluid into the venous compartment takes place predominantly through the arachnoid granulations of the sagittal sinus . The resistance to CSF outflow has been assessed in normal subjects as ranging from 6 to 10 mmHg mL-1 min-1 .
In cases of disturbed CSF re-absorption through the physiological pathways, a secondary component of CSF re-absorption is the leakage directly into the brain parenchyma. This phenomenon can sometimes be visualized as a periventricular hypo-density along the horns of the lateral ventricles .
CSF dynamics disturbances: the need for specific diagnostic tools
Hydrocephalus comprises a miscellaneous group of disorders of CSF dynamics, leading to an excessive accumulation of CSF within the brain .
Aetiology varies widely. Although idiopathic forms are relatively common, hydrocephalus can be secondary to congenital malformations or acquired diseases such as trauma, haemorrhage, infection or tumours. Pathogenesis is also extremely variable. Non-communicating forms are due to narrowing or obstruction of the normal CSF flow pathways within the brain, while communicating forms can be primarily related to CSF re-absorption functional disorders, as in the case of ‘normal pressure hydrocephalus'. Clinical presentation also varies widely, from the paediatric to the geriatric manifestations of the disease. All forms of hydrocephalus are unified by the presence of a disorder in the normal CSF dynamics. Ideally, disequilibrium at any level between the secretion of CSF and its re-absorption, an abnormal bulk or pulsate flow of CSF or an insufficient volume-buffering compensatory reserve can lead to hydrocephalus and - with some distinction to be made - to its spinal counterpart, syringomyelia .
Standard brain imaging can often be misleading in the diagnosis of hydrocephalus. While the enlargement of cerebral ventricles is not necessarily of hydrocephalic nature - as in the frequent case of ventriculomegalia associated with cortical atrophy - even the finding of slit ventricles does not strictly exclude hydrocephalus or a shunt malfunction. Moreover, while shunting dramatically improves CSF dynamics, changes in ventricular size are often subtle or absent even in shunt responders. Whenever proper expertise is available, individualized assessment of CSF dynamics can assist the neurosurgeon in the diagnosis of hydrocephalus and the prognostication of shunting, providing a rationale for programmable valve management and excluding shunt malfunction in non-responders . In patients undergoing a third ventriculostomy, postoperative CSF dynamics diagnostics might also provide valuable prognostic information (Fig. 1) [19,20].
CSF dynamics diagnostics: computerized infusion studies
Comprehensive computerized tools for clinical evaluation are now commercially available . Computerized infusion tests (http://www.neurosurg.cam.ac.uk/icmplus/) were developed to compensate for the limitation of Katzman's lumbar infusion method, especially the inadequate accuracy of estimation of the resistance to CSF outflow. ICP and arterial pressure signal processing and model analysis are implemented in a user-friendly interface to allow the estimation of cerebrospinal dynamics variables such as CSF outflow resistance, brain compliance and pressure-volume index, estimated sagittal sinus pressure, CSF formation rate, compensatory reserve and cerebral vasoreactivity. Those variables can assist in the prognostication of normal pressure hydrocephalus and in the diagnosis of idiopathic intracranial hypertension. The technique is also helpful in the assessment of shunt malfunction, including posture-related over-drainage and proximal or distal shunt obstruction .
The procedure is performed in awake patients (Fig. 2). Access to the subarachnoid space is usually obtained via two lumbar needles  which are connected to an infusion pump  and to a pressure transducer  via a stiff saline-filled tube. The CSF pressure, zeroed at the level of external acoustical meatus, yields a measure of the ICP. In shunted patients, the lumbar access is substituted by two 25-G needles connected to the shunt antechamber, providing a direct measure of ICP. Signals are converted in a digital format and recorded on a laptop for further analysis . After measuring ICP baseline, the subarachnoid space is infused with mock CSF at a constant rate. Briefly, ICP rises until a plateau defines the equilibrium at which the mock CSF infusion flow is re-absorbed. The infusion is painless, and the whole procedure takes about 30 min.
Although infusion studies are listed as part of the management of hydrocephalus by recent normal pressure hydrocephalus (NPH) guidelines , individualized CSF dynamics assessment is not yet common practice, with a handful of leading neurosurgical units being laudable rarities in Europe. This can be explained by the fact that infusion studies are time-consuming procedures, entailing waveform collection and analysis, requiring some degree of biomechanical expertise and specialized training. However, infusion studies are excellent decision-making tools, and even hiring a dedicated neuroscientist should be considered a cost-effective investment for most neurosurgical units.
CSF infusion studies: in vivo shunt evaluation
The common symptoms of dizziness, headache and nausea are the usual presenting features of a patient with suspected shunt malfunction. As the clinical presentation is very unspecific, further testing is needed for a diagnosis of shunt malfunction. A standard computed tomography (CT) does not provide any valuable information about the shunt performance (Fig. 3).
A CSF withdrawal test can temporarily ease the symptoms, thus suggesting an ex juvantibus diagnosis of shunt malfunction. Unfortunately, symptoms are often capricious and unpredictable, and the placebo effect is never easy to exclude. It follows that repeating a CSF withdrawal test the next day is often a reasonable option. Setting up a CSF constant flow drainage is an alternative, which in turn carries a greater risk of shunt infection. Still, the diagnosis will be based on the patient's symptoms and on his/her ability to communicate them. Also, while a CSF withdrawal test could help with the diagnosis of a shunt occlusion, this test neither assesses siphoning and shunt over-drainage nor helps in the case of proximal occlusion (Fig. 4).
On the other hand, a computerized CSF infusion study can provide a complete diagnosis of shunt malfunction in 1-h time, comparing the invivo findings with the expected performance of the specific shunt model . Such a test can also be performed on outpatients, avoiding unnecessary admissions and virtually eliminating the possibility of useless revisions.
CSF infusion studies: NPH
NPH is a clinical syndrome first described by Hakim and Adams in 1965 , characterized by gait apraxia, incontinence and dementia, and diagnosed predominantly among people over the age of 60 yr. Although there are many guidelines for NPH management, there is no randomized blind trial justifying them .
The interest about NPH was recently revived by Silverberg and colleagues , who postulated the new nosological entity of CSF circulatory failure, with features of Alzheimer disease (AD) and NPH. Patients with both diseases have reduced CSF production and decreased CSF turnover, which may be associated with a decreased ability to clear amyloid beta-peptides (Aβ) and tau protein from the brain. They hypothesize that low-flow shunting might improve the CSF turnover and the clearance of those potentially toxic metabolites. A trial evaluating the safety and efficacy of shunting as a treatment for AD is now underway.
Meanwhile, the debate was accelerated by a provocative paper by Stein and colleagues  of Pennsylvania University. Using the tools of health economics, they conclude that all patients suspected of having NPH should be shunted without resorting to any screening tests except for simple scanning to confirm the presence of hydrocephalus. However, in our opinion those conclusions are a bit far-fetched. Shunting is a purely mechanical treatment, and without preoperative CSF infusion studies it is even difficult to know whether CSF dynamics have been normalized or not . In addition, infusion studies provide valuable prognostication of the shunting procedure, a CSF re-absorption resistance (Rcsf) >18 mmHg having a 92% positive predictive value for clinical improvement following shunting . According to the experience gained in the ‘Dutch trial', the best strategy is to test CSF dynamics of every patient with clinical features compatible with NPH. Patients should be shunted if Rcsf is >18 mmHg mL−1 min−1 or, when Rcsf is lower, if CT findings are typical for NPH and there is no or limited cerebrovascular disease .
1. Katzman R, Hussey F. A simple constant-infusion manometric test for measurement of CSF absorption. I. Rationale and method. Neurology
2. Marmarou A. A theoretical model and experimental evaluation of the cerebrospinal fluid
1973, Drexel University, Philadelphia, PA.
3. Ekstedt J. CSF hydrodynamic studies in man. Method of constant pressure CSF infusion. J Neurol Neurosurg Psychiatry
4. Marmarou A, Shulman K, Rosende RM. A non-linear analysis of CSF system and intracranial pressure
dynamics. J Neurosurg
5. Borgesen SE, Gjerris F. The predictive value of conductance to outflow of CSF in normal pressure hydrocephalus
6. Giulioni M, Ursino M. Impact of cerebral perfusion pressure and autoregulation on intracranial dynamics: a modeling study. Neurosurgery
7. Davson H. Formation and drainage of the CSF in hydrocephalus
. In: Shapiro K, Marmarou A, Portnoy H, eds. Hydrocephalus
. New York, USA: Raven Press, 1984: 112-160.
8. Oshio K, Watanabe H, Song Y, Verkman AS, Manley GT. Reduced cerebrospinal fluid
production and intracranial pressure
in mice lacking choroid plexus water channel Aquaporin-1. FASEB J
9. Ekstedt J. CSF hydrodynamic studies in man. Normal hydrodynamic variables related to CSF pressure and flow. J Neurolog Neurosurg Psychiatry
10. Pappenheimer JR, Heisey SR, Jordan EF, Downer JC. Perfusion of the cerebral ventricular system in unanaesthetized goats. Am J Physiol
11. Albeck MJ, Skak C, Nielsen PR, Olsen KS, Borgesen SE, Gjerris F. Age dependency of resistance to cerebrospinal fluid
outflow. J Neurosurg
12. Wagshul ME, Chen JJ, Egnor MR, McCormack EJ, Roche PE. Amplitude and phase of cerebrospinal fluid
pulsations: experimental studies and review of the literature. J Neurosurg
13. Davson H, Welch K, Segal MB. The Physiology and Pathophysiology of Cerebrospinal Fluid
. New York USA: Churchill Livingstone Inc, 1987.
14. Albeck MJ, Borgesen SE, Gjerris F, Schmidt JF, Sorensen PS. Intracranial pressure
and cerebrospinal fluid
outflow conductance in healthy subjects. J Neurosurg
15. Gjerris F., Borgesen S.E. Pathophysiology of CSF circulation. In: Crockard A, Hayward A, Hoff JT, eds. Neurosurgery The Scientific Basis of Clinical Practice
. Cambridge, MA: Blackwell, 1992: 146-174.
16. Czosnyka M, Czosnyka Z, Momjian S, Pickard JD. Cerebrospinal fluid
dynamics. Physiol Meas
17. Chang HS, Nakagawa H. Hypothesis on the pathophysiology of syringomyelia based on simulation of cerebrospinal fluid
dynamics. J Neurol Neurosurg Psychiatry
18. Czosnyka M, Whitehouse H, Smielewski P, Simac S, Pickard JD. Testing of cerebrospinal compensatory reserve in shunted and non-shunted patients: a guide to interpretation based on an observational study. J Neurol Neurosurg Psychiatry
19. Drake J, Chumas P, Kestle J et al.
. Late rapid deterioration after endoscopic third ventriculostomy: additional cases and review of the literature. J Neurosurg
20. Nishiyama K, Mori H, Tanaka R. Changes in cerebrospinal fluid
hydrodynamics following endoscopic third ventriculostomy for shunt-dependent noncommunicating hydrocephalus
. J Neurosurg
21. Smielewski P, Czosnyka M, Steiner L, Belestri M, Piechnik S, Pickard JD. ICM+: software for on-line analysis of bedside monitoring data after severe head trauma. Acta Neurochir Suppl
22. Czosnyka ZH, Czosnyka M, Pickard JD. Shunt testing in-vivo
: a method based on the data from the UK shunt evaluation laboratory. Acta Neurochir Suppl
23. Marmarou A, Bergsneider M, Relkin N, Klinge P, Black PM. Development of guidelines for idiopathic normal-pressure hydrocephalus
: introduction. Neurosurgery
24. Czosnyka Z, Czosnyka M, Richards HK, Pickard JD. Laboratory testing of hydrocephalus
shunts - conclusion of the UK Shunt evaluation programme. Acta Neurochir (Wien)
25. Hakim S, Adams RD. The special clinical problem of symptomatic hydrocephalus
with normal cerebrospinal fluid
pressure. Observations on cerebrospinal fluid
hydrodynamics. J Neurol Sci
26. Esmonde T, Cooke S. Shunting for normal pressure hydrocephalus
(NPH). Cochrane Database Syst Rev
27. Silverberg GD, Mayo M, Saul T, Rubenstein E, McGuire D. Alzheimer's disease, normal-pressure hydrocephalus
, and senescent changes in CSF circulatory physiology: a hypothesis. Lancet Neurol
28. Stein SC, Burnett MG, Sonnad SS. Shunts in normal-pressure hydrocephalus
: do we place too many or too few? J Neurosurg
29. Pickard JD, Spiegelhalter D, Czosnyka M. Health economics and the search for shunt-responsive symptomatic hydrocephalus
in the elderly. J Neurosurg
30. Boon AJ, Tans JT, Delwel EJ et al.
. Dutch normal-pressure hydrocephalus
study: prediction of outcome after shunting by resistance to outflow of cerebrospinal fluid
. J Neurosurg
31. Tans JT, Boon AJ, Dutch NPH Study Group. How to select patients with normal pressure hydrocephalus
for shunting. Acta Neurochir Suppl