Cerebrospinal Fluid Flow

Abstract

Circulation of cerebrospinal fluid and interstitial fluid around the central nervous system and through the brain transports not only those water-like fluids but also any solutes they carry, including nutrients, drugs, and metabolic wastes. Passing through brain tissue primarily during sleep, this circulation has implications for neurodegenerative disorders including Alzheimer's disease, for tissue damage during stroke and cardiac arrest, and for flow-related disorders such as hydrocephalus and syringomyelia. Recent experimental results reveal several features of this flow, but other aspects are not fully understood, including its driving mechanisms. We review the experimental evidence and theoretical modeling of cerebrospinal fluid flow, including the roles of advection and diffusion in transporting solutes. We discuss both local, detailed fluid-dynamic models of specific components of the system and global hydraulic models of the overall network of flow paths.

Keywords: biological fluid mechanics; cerebrospinal fluid; glymphatic system; viscous flow.

Figure 1

The anatomy of cerebrospinal and interstitial fluid circulation in the central nervous system. Large chambers, such as the ventricles and subarachnoid space, are contiguous with small chambers that penetrate brain tissue, such as PVSs and nerve sheaths, allowing for global circulation and solute transport. In some chambers, fluid flows freely; in others, it moves slowly through porous media. Illustration by Dan Xue.

Figure 2

Measuring flow in the surface perivascular space of a live mouse via particle-tracking. (a) Micron-scale fluorescent tracer particles (green) injected at the back of the skull are swept along by flowing cerebrospinal fluid (CSF) (blue) and pass through the perivascular space surrounding a surface artery (red). Automated particle-tracking measurements of tracer particles’ positions and velocities make it possible to overlay pathlines colored according to the instantaneous velocity of each. (b,c) Averaging many such measurements over time results in a mean flow field. Solid squares indicate regions where ≥15 measurements were made; open squares indicate regions where <15 measurements were made (see also Supplemental Video 1). (d) The instantaneous root-mean-square (RMS) CSF velocity pulses in synchrony with the cardiac cycle [as determined by electrocardiogram (ECG)] and shows little correlation with the respiratory cycle. Imaging provided by A. Ladrón-de-Guevara and M. Nedergaard; analysis by Y. Gan and D.H. Kelley.

Figure 3

Shapes of surface PVSs. (a) In vivo images showing cross sections of surface PVSs in mouse (top, bottom) and human (middle). Images adapted with permission from (top) Mestre et al. (2018b) (CC BY 4.0); (middle) Bedussi et al. (2018) (CC BY-NC 4.0); and (bottom) Schain et al. (2017) (CC BY 4.0). (b) A simple model parameterizing cross-sectional shapes of PVSs with a circle and an ellipse allows the normalized hydraulic resistance for many such shapes to be calculated, revealing that resistance has minimum values (circles) at a particular elongation (with other parameters held constant) and that in vivo shapes have nearly the same elongation (triangles). Curve colors match border colors for the corresponding shapes in panel a. Panel adapted from Tithof et al. (2019) (CC BY 4.0). (c) In vivo images showing cross sections of human periarterial (top) and perivenous (bottom) spaces. (d) Advection–diffusion equations in domains built from in vivo images show that solute transport is faster in periarterial spaces than perivenous spaces, if the pressure drop is the same across each, because periarterial spaces are bigger. Panels c and d adapted from Vinje et al. (2021) (CC BY 4.0). Abbreviations: BV, blood vessel; PVS, perivascular space; SAS, subarachnoid space.

Figure 4

Modeling flow and transport in brain extracellular space. (a) Jin et al. (2016) studied cerebrospinal fluid flow from periarterial to perivenous spaces on a hexagonal lattice estimating vessel locations in primate brain tissue and modeled the extracellular space with a geometric reconstruction that recapitulated the structure of tissue imaged by Kinney et al. (2013). Panel adapted from Jin et al. (2016) (CC BY-NC-SA 4.0). (b) Holter et al. (2017) simulated flow and transport in domains created by segmenting the 3D images of Kinney et al. (2013). Panel adapted from Holter et al. (2017) (CC BY 4.0). (c) Penetrating arterioles (red) and ascending venules (blue) in cortex extend roughly perpendicular to the brain surface, allowing for 2D modeling. Data from Pablo Blinder. (d) Taking arterioles and venules to be line sources and line sinks, respectively, and normalizing by the tissue permeability, Schreder et al. (2022) modeled flow and hydraulic resistance in both primate (not shown) and mouse brain tissue. Curves show streamlines. Panel adapted from Schreder et al. (2022); copyright 2022 the authors.

Figure 5

Cerebrospinal fluid (CSF) flow outside the brain. (a) Using functional magnetic resonance imaging (fMRI) in humans, Fultz et al. (2019) pulsed the magnetic field to excite water molecules in the field of view, and then measured as unexcited molecules were swept into the field of view by the flow of CSF. The raw signal grew earliest and most strongly in slices at the edge of the field of view, placed just below the fourth ventricle, indicating bulk fluid flow there. (b) Sánchez et al. (2018) developed an asymptotic fluid mechanical model for flow in the spinal subarachnoid space, driven by pressure fluctuations at its junction with the skull. The model predicts an axial streaming flow at second order, which varies depending on the eccentricity of the subarachnoid space, as shown here in cross section. (c) Lawrence et al. (2019) used the flow model of Sánchez et al. (2018) to simulate solute transport in the spinal subarachnoid space, with application to drug delivery. Here $x$ is the axial coordinate, $s$ is the azimuthal coordinate, $T$ is time, and red curves show the concentration after summing over radius and azimuth. An initial bolus of solute is dispersed by the pulsing and streaming flow. Panels adapted with permission from (a) Fultz et al. (2019), copyright 2019 the authors; (b) Sánchez et al. (2018), copyright 2018 Cambridge University Press; and (c) Lawrence et al. (2019), copyright 2019 Cambridge University Press.