Interstitial fluid flow under the microscope: is it a future drug target for high-grade brain tumours such as glioblastoma?

What and where is interstitial fluid flow in the brain?

IFF is the movement of fluid within the space between cells in tissue, or interstitial space. In the brain, this space consists of a network of extracellular matrix co-mingled with a dense, process-rich network of astrocytes, microglia, and neurons. Fluid moves within the interstitial space as it drains toward open structures in and around the brain, and in 1998, Geer and Grossman observed that these bulk fluid flow pathways correlated with paths of tumour cell dissemination[1]. In fact, flow pathways such as the white matter tracts and ventricles are known impediments to treating glioblastoma.

Why is IFF important and how is it relevant in cancer?

Tumour progression is marked by an increase in fluid accumulation and solid stress due to influx of blood and other fluids into the tissue as well as accumulation of cells and extracellular matrix. The resulting heightened pressures (in patients undergoing standard of care treatment, 3mmHg) can be 2–10 times those of the adjacent healthy tissue, and this pressure gradient causes fluid flow from the tumour into the surrounding tissue[2], [3]. In vitro, applying IFF through engineered hydrogel systems stimulates brain tumour cell invasion both alone and towards gradients of serum [4], [5]. In other cancers, this fluid flow is also known to activate cells in the tissue around the tumour, which further supports disease progression[6].

What knowledge do we need to understand IFF in the context of brain cancers?

Currently, there are numerous gaps in our knowledge of fluid flow in the brain in general, let alone fluid flow in brain cancers. We are only just beginning to understand the nature of how fluid moves into, around, and out of the brain, and many previous findings remain highly debated. Additionally, the underlying physiological state of an organism has been shown to affect the rate, production and movement of cerebrospinal fluid within the brain, including age, sex, disease state, circadian rhythm, etc. Though we know that there are changes to the bulk flows, definitive IFF rates, what it means prognostically, full repertoire of effects on tumour cells and on other cells surrounding the tumours, the relationship to bulk flows, vascularization, and cerebral pressure is still murky.

Why can’t we get there right now?

Currently, major impediments to the implementation of IFF as a target include a lack of understanding of 1) its nature in non-tumour conditions, 2) the variability of its magnitudes and patterns with tumours, 3) the diversity of responses to IFF on a cellular and tissue level and 4) its complex effects on drug delivery and response. These can be remedied using methods (such as imaging) and inclusion in in vitro studies. Though we have made strides with these techniques, we have only scratched the surface of this ubiquitous force. However, tools for in vitro study (including hydrogel based systems employing microfluidics) and in vivo imaging (including MRI or intravital imaging) have allowed us to begin better assessing IFF, including the elucidation of several potential targets.

Known molecular targets (Figure 1)

Figure 1: Potential targets for interstitial fluid flow-mediated changes to glioblastoma.

A) Targets on glioma cells include the receptor CXCR4 (which can be targeted with AMD3100 (Plerixafor)) or its ligand CXCL12 as mediators of chemotactic invasion and CD44 and its substrate hyaluronic acid. B) Fluid flow in the brain can be targeted via multiple exploratory targets including (clockwise from top right): Targeting of sites of drainage including subdural and meningeal lymphatics; targeting sites of CSF generation and movement in the choroid plexus lining the ventricles; targeting the neurovascular unit via blood vessels reducing leakage and the astrocytes increasing or altering fluid uptake via Aquaporins. (Brain image generated via Biorender).

IFF does not passively move tumour cells but in fact stimulates pathways overlapping with some known targets in GBM. There are two primary mechanisms identified as mediators of IFF triggered invasion: 1) Autologous Chemotaxis is a mechanism whereby a cell stimulates its own migration via a self-secreted chemokine. When there is no convective force from fluid flow, the chemokine forms a diffuse, uniform cloud around the cell; however, if there is fluid flow, the chemokine is transpo rted downstream of the cell forming a gradient along which the cell migrates. To mediate this mechanism, a cell must express both the chemokine and its receptor. For GBM this has been identified as the ligand-receptor pair CXCL12 and CXCR4, the latter of which is overexpressed in many cancers and correlates with worsened survival. This flow-mediated invasion via CXCL12-CXCR4 has been shown both in vitro and in vivo[7]. Interestingly, this mechanism may be dependent on a number of microenvironmental factors including the rate of IFF and the density of tumour cells [8]. 2) Mechanotransduction occurs when a cell detects mechanical changes to its surroundings, leading to signaling events and downstream cell behaviors such as motility. Generally, IFF will deform the extracellular matrix and therefore the receptors bound to this matrix. In glioblastoma the primary receptor currently implicated in this mechanism is the hyaluronan receptor CD44, which is also correlated with worsened survival[5]. These two mechanisms of flow-stimulated invasion may act on distinct cell populations, suggesting dual-targeted therapies may be best for reducing glioblastoma invasion.

Other receptors and receptor-ligand pairs have been identified in other cancer to elicit flow-mediated invasion, such as CCR7-CCL21 and integrins [2]. There is an opportunity for more exhaustive testing and potential repurposing of therapies within the context of GBM. And although these are pathways that have been examined, there are other potential biomarkers and targets yet to be identified. Similarly, current research on tumour stiffness and detection of mechanical stress has identified targets that may have overlapping targets with IFF in the brain [9]. Fully integrating these varied approaches creates potential for multi-targeted therapeutics.

It is also possible to directly and therapeutically alter IFF itself by altering major bulk flow mediators within and around the brain. These targets may be attractive because they are outside of the tumour and thus may have more predictable molecular targets. Recent identification of new pathways of fluid movement and drainage from the brain offer unique opportunities to alter fluid flow. However, we do not yet understand the downstream effects of such changes and offer warnings that side effects (such as edema or seizure) could ensue when altering overall interstitial fluid pressures and fluid movement within this sensitive organ.

Sites of drainage, such as the meningeal and cribriform lymphatics, subarachnoid villi, and perivenous spaces, could offer a means to manipulate fluid flow at the outlet. This technique has been implemented in peripheral cancers (such as targeting or reducing lymphangiogenesis) but is a promising therapeutic strategy against GBM[10], [11]. Alternately, sites of influx, such as the blood vasculature and the choroid plexus, would allow alteration of incoming fluid to the brain and reduce overall fluid flow. Vascular targets such as VEGFR2 have been tried with glioblastoma with poor results due to increases in invasion post-therapy[12]. Strategically altering flow throughout or within particular areas of the brain may impact GBM outcomes, but more research is necessary to understand the complex anatomical-physiological interactions. Sites of regulation within the brain include the glymphatic system[13], the ependymal cell layers[14], and glial cells. Targeting of any of these systems could result in vast changes to fluid flow though, their impact on IFF is yet unknown. Targets on astrocytes within the glymphatic system, such as aquaporin-4, overlap with identified targets in GBM and thus IFF could play a role in this therapy[15].

Expert Opinion: Where are we headed with this research?

Intratumoural heterogeneity is a major impediment to understanding and treating individual tumours. These differences are both found within the tumour cells themselves, but also within the tissue surrounding the tumours. Recent mapping of IFF in tumours in vivo using MRI has shown high levels of differential flow and mass transport in and around implanted models of GBM[16]. These flows are different in both magnitude and direction and can vary quite dramatically in adjacent tissue regions. Interestingly, in mice it appears that flow from the tumour may be governed more by the surrounding tissue structure than individual aspects of the tumour itself. This syncs well with data in patients suggesting that proximity to various structures (such as white matter tracts) are important prognostic factors. Related to delivery of drugs, it may be advantageous to have more or less IFF depending on localization within the tumour. Based on the heterogeneity of the tissue flows that have been observed, it is likely that drug delivery and distribution is subject to these heterogeneous flow patterns and thus a major area for opportunity for strategic application of current therapies. Additionally, almost all of the current research has focused on the role of flow in invasion of tumour cells with little progress in the role of flow on other cellular behaviours such as cellular differentiation, genetic instability, metabolism, proliferation, and therapeutic response.

In total, IFF is an untapped wealth of potential drug research. Since we are not exploring how cells (any cells) respond in the presence of IFF we may be ignoring an entire set of genes, proteins, and other druggable targets that are only identifiable in the presence of flow. With the recent advent of imaging modalities we can begin to actually determine where in tumours IFF is heightened and use that information to discover new targets for therapy. I believe that this area of research coupled with better and simpler to implement in vitro models is the next step in using IFF as a biomarker and as a druggable target in GBM.