Synopsis

Keywords: Neurofluids, Brain, Cerebrospinal fluid dynamics

Normal pressure hydrocephalus (NPH) is a brain pathology with enlarged ventricles and is diagnosed as a cerebrospinal fluid (CSF) dynamics disorder based on a delayed ascent of radiotracer over the cerebral convexity and elevated flow through the cerebral aqueduct. In this study, we provide fluid dynamics-based evidence of disruption in CSF oscillation within the cerebral aqueduct in NPH patients. Particularly, we show that the peak volumetric flow is delayed with respect to peak transmantle pressure in NPH patients as compared to healthy control participants.

INTRODUCTION

Cerebrospinal fluid (CSF) disorders, such as hydrocephalus¹ and idiopathic intracranial hypertension², are characterized by a disruption of CSF regulation in the brain. Normal pressure hydrocephalus (NPH) is a paradigmatic example of a CSF dynamics disorder diagnosed by delayed ascent of radiotracer over the cerebral convexity during a nuclear medicine cisternogram^3,4,5 and increased flow through the cerebral aqueduct on phase contrast MRI⁶. The cerebral aqueduct or the aqueduct of Sylvius is the conduit through which CSF flows through the intracranial compartments. The main goal of this study was to evaluate if the CSF flow through the cerebral aqueduct has a unique fluid mechanics signature in NPH.

METHODS

The study included 18 patients, 6 each cognitively unimpaired (CU), Alzheimer’s disease (AD), and NPH groups. The mean age of participants was 78.8±3.8 in CU, 76.2±3.4 in AD, and 75.8±3.3, with 3 female participants in CU, 2 in AD, and 1 in CU. Imaging was performed on a 3T Philips system. The protocol included a real-time echo-planar-based phase contrast acquisition at the cerebral aqueduct and sagittal high spatial resolution highly T2-weighted imaging of the midline structures. The phase contrast data was retrospectively gated using pulse oximeter recording, and the volumetric flow of CSF (density $$$\rho \approx 10^{-3} \text{kg/m}^{3}$$$ and kinematic viscosity $$$ \nu \approx 0.71\times10^{-6} \text{m}^{2}/\text{s}$$$) through the aqueduct calculated for multiple cardiac cycles. The data was averaged to get the flow in one cardiac cycle. The length $$$L$$$ and radius $$$r$$$ of the aqueduct was measured from the high resolution T2-weighted sequence using a multi-planar reformatting tool (Figure 1). The flow data and measurements were then input into a model that calculated pressure across the aqueduct for a cardiac cycle using a boundary layer formulation and axisymmetric Navier-Stokes equations⁷, which outputs non-dimensionalized volume $$$\overline{V}(t)$$$ and pressure $$$\overline{P}(t)$$$ over a cardiac cycle. The quantities with dimensions can be recovered through $$$P(t)=\rho \omega^{2} L_{s} L \overline{P}(t)$$$ and $$$V(t)=\omega V_{s} \overline{V}(t)$$$, where $$$\omega$$$ is the cardiac frequency, $$$V_{s}$$$ is the stroke volume and $$$L_{s}=V_{s}/\pi r^{2}$$$ is the characteristic stroke length. The model has been used to study age and aqueductal dimension related effects on transmantle-pressure⁸. The aqueductal dimensions were measured from brain MRI scans of the individuals as shown in Figure 1. The absolute values of the peak pressure and volume are reported for the caudal and the cranial cycle where the caudal direction was defined to be positive. The difference between the peak pressure and the peak volume during the caudal cycle reported in % cardiac cycle was defined as the phase lag $$$\phi$$$. For the two-sample Welch’s t-test, p-value of 0.01 and less was considered significant.

RESULTS

The cardiac pulsation drives CSF into the brain through the cerebral aqueduct. The pressure difference between the third and fourth ventricles, $$$P(t)=P_{3}(t)-P_{4}(t)$$$, facilitates the volumetric CSF flow $$$ V(t)$$$ in the aqueduct. The time-dependent non-dimensionalized pressure difference $$$\overline{P}(t)$$$ and the volumetric flow $$$\overline{V}(t)$$$ for the groups plotted versus the cardiac cycle are shown in Figure 2. When the pressure difference changes in sign between caudal and cranial cycle, there is a slight time lag between the peak pressure difference and the peak volume as the transition cannot be instantaneous. That quantity defined as the phase lag $$$\phi$$$ is illustrated in the last panel of Figure 2. The absolute peak values for the pressure and volume during the caudal and the cranial cycles are shown in the left and the middle panels of Figure 3, respectively. The phase lag values shown on the right panel of Figure 3 demonstrates significantly higher phase lag in NPH patients compared to the CU with Welch’s t-test value of p<0.01.

DISCUSSION

NPH is a CSF dynamics disorder. The increased phase lag in NPH we report in this study provides a fluid dynamics description of the disorder shedding further light into the physiopathology of NPH. The phase lag could be an important quantity to study in other CSF dynamics disorders as well.

CONCLUSION

NPH is a brain pathology with enlarged ventricles with demonstrated alteration in the brain viscoelastic properties⁹. In this study, we have shown that CSF dynamics is also altered in NPH hinting towards a possible underlying connection with brain viscoelasticity. In addition to the trend of elevated mean volumetric flow through the cerebral aqueduct in NPH, we show that the phase lag between the peak pressure and the peak volumetric flow is significantly higher in NPH. The result motivates further fluid dynamics-based studies in NPH. One interesting direction would be to analyze the phase lag between before and after shunt surgery in cases of NPH to assess if the value normalizes after surgery.

Acknowledgements

We are thankful for Dr. Stephanie Sincomb providing us with the model for the pressure calculations performed in this study and for prompt feedback on pertinent questions.