Recapitulation of physiologic and pathophysiologic pulsatile CSF flow in purpose-built high-throughput hydrocephalus bioreactors

Background: Hydrocephalus, an accumulation of cerebrospinal fluid (CSF) in the ventricles of the brain, is often treated via a shunt system to divert the excess CSF to a different compartment; if left untreated, it can lead to serious complications and permanent brain damage. It is estimated that one in every 500 people are born with hydrocephalus. Despite more than 60 years of concerted efforts, shunts still have the highest failure rate of any neurological device requiring follow-up shunt revision surgeries and contributing to the $2 billion cost of hydrocephalus care in the US alone. The absence of a tested and validated long-term in-vitro model that can incorporate clinically relevant parameters has limited hypothesis-driven studies and, in turn, limited our progress in understanding the mechanisms of shunt obstruction in hydrocephalus. Testing clinical parameters of flow, pressure, shear, catheter material, surface modifications, and others while optimizing for minimal protein, cellular, and blood interactions has yet to be done systematically for ventricular catheters. Several studies point to the need to not only understand how cells and tissues have occluded these shunt catheters but also how to stop the likely multi-faceted failure. For instance, studies show us that tissue occluding the ventricular catheter is primarily composed of proliferating astrocytes and cells of the macrophage lineage. Cell reactivity has been observed to follow flow gradients, with elevated levels of typically pro-inflammatory interleukin-6 produced under shear stress conditions greater than 0.5 dyne/[Formula: see text]. But also, that shear can shift cellular attachment. The Automated, In vitro Model for hydrocephalus research (AIMS), presented here, improves upon our previous long-term in vitro systems with specific goals of recapitulating bulk pulsatile cerebrospinal fluid (CSF) waveforms and steady-state flow directionality relevant to ventricular catheters used in hydrocephalus.

Methods: The AIMS setup was developed to recapitulate a wide range of physiologic and pathophysiologic CSF flow patterns with varying pulse amplitude, pulsation rate, and bulk flow rate with high throughput capabilities. These variables were specified in a custom-built user interface to match clinical CSF flow measurements. In addition to flow simulation capabilities, AIMS was developed as a modular setup for chamber testing and quality control. In this study, the capacity and consistency of single inlet resin chambers (N = 40), multidirectional resin chambers (N = 5), silicone chambers (N = 40), and PETG chambers (N = 50) were investigated. The impact of the internal geometry of the chamber types on flow vectors during pulsatile physiologic and pathophysiologic flow was visualized using Computational Fluid Dynamics (CFD). Dynamic changes in ventricular volume were investigated by combining AIMS with MRI-driven silicone model of a pediatric patient's ventricles. Parametric data were analyzed using one-way analysis of variance (ANOVA) or repeated measures ANOVA tests. Non-parametric data were analyzed using Kruskal-Wallis test. For all tests, a confidence interval was set at 0.95 (α = 0.05). In a subset of experiments, AIMS was also tested for its capability to measure the flow of florescent microspheres through the holes of unused and explanted ventricular catheters.

Results: The analysis of peak amplitude through chambers indicated no statistically significant differences between the chamber batches. This high throughput setup was able to reproduce clinical measurements of bulk CSF flow tested in up to 50 independent pump channels such that there was no exchange of solution or flow interference between adjacent channels. Physiologic and pathophysiologic clinical measurements of CSF flow patterns were recapitulated in all four chamber types of the AIMS setup with and without augmented compliance. The AIMS setup's automated priming feature facilitated constant fluid contact throughout the study; no leaks or ruptures were observed during short- (up to 24 h) or long-term (30 days) experiments. Finally, qualitative microscopy long-exposure image capture revealed microsphere movement under steady-state and pulsatile flow of spheres moving into the shunt catheter.

Conclusion: AIMS successfully simulates clinical measurements of physiologic and pathophysiologic CSF pulsation amplitude and frequency, as exemplified using clinical data of CSF exiting an externalized ventricular drain in four distinct chamber types, as well as flow patterns from a valve. This provides a promising platform for investigating the direct interaction between CSF, immune cells, and shunt hardware under relevant flow conditions when both the source of bulk flow and pulsatility are coupled. The implementation of this system in conjunction with a previously reported three-dimensional hydrogel scaffold in future work will enhance our understanding of shunt-related complications and improve treatment strategies by reducing the obstruction rate.