Relation between tag position and degree of visualized cerebrospinal fluid reflux into the lateral ventricles in time-spatial labeling inversion pulse magnetic resonance imaging at the foramen of Monro

The results of our study show a substantial variation in the amount of observed reflux
from the third ventricle into the lateral ventricles based upon the position of the
CSF tag in Time-SLIP imaging. For demonstration of reflux into the lateral ventricles,
we found that the optimum tag location is with the top (superior) border of the tag
at the level of the FMSP.

We postulate the explanation of these findings is that the majority of CSF exchanged
between the third and lateral ventricles during normal pulsation is from fluid located
in the cephalad portion of the third ventricle. Fluid tagged in the more caudal portion
of the third ventricle contributes less to the overall exchanged pool of fluid. An
alternative hypothesis would be that fluid closer to the foramen experiences a higher
reflux velocity resulting in an apparent increase in reflux height.

Modern investigation into the normal flow of CSF has suggested a much more complex
circulation than was historically thought 12]–15]. In vivo imaging techniques have promoted the idea of a pulsatile nature of CSF flow
as opposed to the classic theory of bulk flow from the choroid plexus of the ventricular
system outward into the subarachnoid space 13], 16]. The pulsatile nature results in bidirectional flow through the ventricular system
which can drive fluid from the subarachnoid space back as far as the lateral ventricles
17]. This system drives the regulation of intracranial pressure and the clearance of
metabolites, many of which are pro-inflammatory or toxic to the brain substance 1], 2]. Since the brain lacks true lymphatic vasculature, the majority of interstitial fluid
clearance is through the CSF 1], 2].

Disturbance of the normal pulsatile nature of CSF flow has been shown to correlate
with many diseases of the CNS. Recent studies have confirmed decreased reflux from
the third ventricle into the lateral ventricles in conditions such as aqueductal stenosis
18], noncommunicating hydrocephalus 9], and normal pressure hydrocephalus 11]. It is also possible that alterations in normal CSF physiology play a role in the
pathogenesis of other diseases of the CNS. The application of techniques described
in this study may prove useful in further elucidating both the normal physiologic
flow of CSF, as well as in assessment of underlying pathologic disturbances.

The advent of MRI sequences sensitive to flow effects has greatly advanced our understanding
of normal and abnormal CSF flow patterns. One of the most commonly used MRI pulse
sequences is phase-contrast imaging (PC-MRI) 7], 8]. Phase-contrast MRI relies on the detection of complex phase changes in moving fluid
7]. An advantage of PC-MRI is the ability to quantify CSF flow 7], 11]. A primary disadvantage of PC-MRI, however, is that turbulent CSF flow 11], 19], 20] and bulk CSF flow 11], 21] cannot be reliably visualized. Time-SLIP imaging allows the noninvasive visualization
of CSF displacement without relying on velocity-dependent phase changes. The principle
underlying time-SLIP imaging is visualization of fluid displacement by selective magnetic
tagging of protons. This technique can be applied to visualize the amount of fluid
displacement from the third ventricle into the lateral ventricles. The bidirectional
flow through the foramen of Monro related to the cardiac and respiratory cycle results
in the mixing of tagged protons in the third ventricular CSF refluxing back into the
lateral ventricles. This technique allows improved characterization of turbulent flow
and CSF stasis relative to phase-contrast techniques.

We have shown that the amount of visualized CSF reflux into the lateral ventricles
in Time-SLIP imaging is dependent on the positioning of the CSF tag. To the authors’
knowledge, there are no existing publications describing the distance dependence of
Time-SLIP tag placement on the amount of visualized CSF reflux. Our results suggest
caution be exercised when interpreting a lack of reflux as pathologic when the tag
position is not optimal. We have also illustrated that inappropriate tag positioning
may continue to show CSF reflux, but the amount of reflux will be underestimated.
The information presented herein is essential for future studies assessing the physiologic
and pathologic variability of this phenomenon.