Introduction: Addressing unwanted solutes within the cerebrospinal fluid (CSF) has potential to help treat many CNS diseases including: subarachnoid hemorrhage (SAH), meningitis and leptomeningeal metastases. Pre-clinical models of CSF dynamics [1] do not accurately mimic humans and state-of-the art imaging lacks visualization at the resolution or duration required. Neurapheresis, a novel platform for CSF filtration-based therapeutics, has undergone extensive optimization and safety testing and is now in human clinical studies. Here, we present a platform for Neurapheresis device optimization.

Methods: CSF space geometry was defined by T2-weighted MR image segmentation of a healthy-adult. Small anatomic structures were manually added as described previously [2]. The intracranial CSF space included the cortical subarachnoid space, intraventricular regions and cisterns. To define CSF flow oscillations, phase-contrast MRI was obtained along the spine. These measurements were boundary conditions for a bi-phasic computational fluid dynamics simulation with fluid consisting of blood and CSF. A dual-lumen catheter was incorporated, simultaneously aspirating CSF from the lumbar space and returning fluid to the mid-thoracic at flow rates up to 2 ml/min. To help validate the computational results, an MRI compatible, optically clear, 3D-printed CSF phantom with subject-specific CSF flow oscillations was constructed. To simulate SAH, aqueous fluorescein sodium (5 mL bolus; 66.4 mM) was injected within the intracranial subarachnoid space. Tracer spread was recorded by time-lapse over multiple hours. Digital image subtraction was applied to quantify spatial-temporal tracer distribution.

Results: In vitro results provided visualization of tracer spread with high spatiotemporal resolution throughout the entire CSF system under various protocols. Numerical modeling provided detailed quantification of the CSF flow field including steady-streaming flow dynamics and spatial-temporal distribution of blood within the CSF.

Conclusions: The provided platform can allow parametric testing to better understand CSF dynamics and optimizing Neurapheresis for current and future indications.

Patient Care: This research may lead to optimization of Neurapheresis, with improved CSF filtration protocols in terms of device geometry and filtration rate, direction and/or location.

Learning Objectives: 1) Understand the normally passive nature of the CSF space and how active circulation and targeted interventions can be used for CSF filtration as well as drug circulation. 2) Presentation of detailed quantification of CSF space 3D structure, volume and connectivity for 38 distinct regions within the intracranial cisterns, ventricles and spinal and intracranial subarachnoid space. 3) Conceptual understanding of how in vivo imaging can be used to define numerical and experimental models of the human CSF system.

References: 1. Khani M, Xing T, Gibbs C, Oshinski JN, Stewart GR, Zeller JR, Martin BA (2017), "Nonuniform Moving Boundary Method for Computational Fluid Dynamics Simulation of Intrathecal Cerebrospinal Flow Distribution in a Cynomolgus Monkey." J Biomech Eng, 139. https://www.ncbi.nlm.nih.gov/pubmed/28462417 2. Sass LR, Khani M, Natividad GC, Tubbs RS, Baledent O, Martin BA (2017), "A 3D subject-specific model of the spinal subarachnoid space with anatomically realistic ventral and dorsal spinal cord nerve rootlets." Fluids Barriers CNS, 14: 36. https://www.ncbi.nlm.nih.gov/pubmed/29258534