An investigation of CSF dynamics and biochemical composition in normal pressure hydrocephalus utilizing the phase contrast MRI scanning

Abstract

Normal pressure hydrocephalus (NPH) is a known cause of dementia. It is the only type of dementia which can be treated surgically by pulling a ventriculo-peritoneal shunt to drain excess cerebrospinal fluid (CSF) from the ventricles of the brain. In
addition, it is associated with difficulty in walking (which usually precedes dementia) and incontinence (which usually follows dementia). It is important to diagnose this condition properly as there are other dementia types and different neurological conditions which present in a similar way but would not respond to any surgical intervention. A number of different invasive and non-invasive tests are required to differentiate these various conditions from each other and to select the appropriate patients who would respond to shunting. Computerized tomography and simple magnetic resonance imaging (MRI) of the head are widely used clinically as part of the investigations, but these can not be relied upon completely.
Cine Phase contrast MRI scanning is a more sophisticated technique of examining the flow of CSF through different regions of the brain. This study prospectively investigated the role of cine phase contrast MRI scanning in calculating the various CSF flow variables through the cerebral aqueduct and studied their relationship with the patient selection for shunting. The decision for shunting was based on the usual clinical grounds and the treating physician was blinded by the results of the CSF flow characteristics. For analysis of the results patients were divided into two groups depending on whether they had only one or two MRI scans (the second scan was performed after drainage of two hundred ml of CSF via the lumbar drain). There were seventeen patients in the first group (10 males and 7 females). There was no statistical difference between the peak velocity, average velocity, average flow over range, forward volume, reverse volume and net forward volume between shunted (n=6) and un-shunted patients (n=11). The p values for these parameters were 0.567, 0.621, 0.929, 0.416, 0.449
and 0.948, respectively. The second group comprised of twenty one patients.
Since these patients had two MRI scans each, the differences between the above CSF flow parameters were calculated and studied in relation to shunt status. Again the differences could not reach statistical significance between shunted (n10) and un-shunted (n=11) patients. The calculated p values for differences between peak velocities, average velocities, average flows over range, forward volumes, reverse volumes and net forward volumes between shunted and un-shunted patients were 0.224, 0.116, 0.712.0.608, 0.624 and 0.438, respectively. Using stepwise multiple regression to look at the unique contribution of each variable towards the shunt status, the difference between average velocities, difference between reverse volumes and difference between forward volumes achieved statistical significance in that order.
In the second part of the study, biochemical analysis of the CSF samples collected from NPH patients was carried out. The concentration of total protein and different cations were determined in these samples (n=29) and the differences between
various parameters in shunted and un-shunted patients were studied and compared. The difference in the total protein concentration between shunted (n=15) and un-shunted (n=14) patients reached significance (p0.030). This significant was not sustained when the blood contaminated CSF samples (from both shunted and un-shunted patients) were excluded from comparison. There was no significant difference between the concentrations of Nat K, Ca2 Mg2 and trace elements studied (Fe 2 , Pb2 , Zn 2 , Cu2t Mn2 and Se2 ) in the CSF samples of shunted and un-shunted patients. However, on multiple regression analysis Na was shown to have independent unique contribution on the outcome i.e. the
selection of patients for shunting.
It was concluded that although the present study could not establish the precise role of the cine phase contrast MRI scans in the evaluation of NPH patients but conducting a large scale study is more likely to show the statistical significance of
different CSF flow variables. Furthermore, CSF proteomics might be the way forward to identify the true NPH patients and shunt responders.