Clinical Utility of DTI‐ALPS in Identifying Dysfunction in the “Glymphatic” System

For centuries, neurologists, neurosurgeons, and neuroanatomists have been intrigued by the mechanisms that entail waste clearance in the brain, especially given the lack of traditional lymphatic channels. Early researchers identified "prelymphatic" pathways using electron microscopy on tissue samples, but the dynamic movement of fluids in the brain remained poorly understood.1 It was not until 2012 that in vivo tracking of tracers injected into the cisternal spaces of rodents uncovered a previously unknown pathway, leading to a groundbreaking shift in neuroscience.2 This discovery, the "glymphatic" system, opened new avenues for investigating how the brain clears toxic metabolic byproducts, such as amyloid-beta, a key contributor to neurodegenerative diseases. In response to this breakthrough, there was immediate interest in developing MRI sequences capable of detecting this pathway in the human brain, with direct clinical implications. This led to the formation of the "Imaging Neurofluids" study group within the International Society for Magnetic Resonance in Medicine (ISMRM) in 2019, which brings together international researchers to advance our understanding of these pathways through multimodal imaging techniques.3 However, the glymphatic system is not the only proposed mechanism for waste clearance in the brain. In 2008, Carare and colleagues identified an alternative pathway, which they termed the intramural periarterial drainage pathway.4 Tracers injected directly into the brain parenchyma of rodents were quickly observed in the outer walls of arterioles, suggesting that waste products like amyloid-beta drain toward the arterial walls. This contrasts with the glymphatic pathway, where amyloid-beta and interstitial fluid drain along venous perivascular spaces. Another model proposes that fluid exchange occurs just beneath the cortical surface, where the vascular pulsatility of pial arteries in the subarachnoid space facilitates the mixing of fluids, with interstitial fluid flowing out toward the pial arteries along a favorable concentration gradient.5 One of the challenges in understanding brain waste clearance is unraveling the mechanisms of fluid movement within the complex interstitial spaces of the brain parenchyma.6 Is fluid movement governed by bulk flow, convective flow, or diffusion-based processes? This is a crucial question for MRI researchers aiming to develop methods to measure such movement. Another challenge is determining the precise nature of the perivascular spaces—do fluids and solutes move primarily along periarteriolar spaces, or do they flow more through perivenous spaces? Additionally, understanding which region of the brain must be examined is key: fluid exchange primarily occurs in the cortex, but also across the ventricular ependymal wall, where CSF from the ventricles enters the brain parenchyma. One of the most promising MRI techniques to date is intrathecal gadolinium-enhanced dynamic imaging.7 This method, is based on administering a small amount of gadolinium based contrast agent in the spinal thecal sac. Gadolinium enhancement is then tracked over time within the brain parenchyma. Delayed gadolinium clearance over the entire brain have been identified in specific neurological disorders. However, while proponents of this technique argue for its safety, it is not widely accepted for clinical research. As with other MRI sequences, intrathecal gadolinium-enhanced imaging faces challenges related to spatial and temporal resolution, as well as the need for subject preparation and long acquisition times, making it less popular among researchers. Arterial spin labelling, 2D-phase contrast MRI, intravoxel incoherent motion based DTI and intravenous based gadolinium enhancement methods have been used but are not ready for clinical use. In 2017, the diffusion tensor image analysis along the perivascular space (DTI-ALPS) index was proposed as a method to identify disruptions in perivascular fluid movement, particularly in the deep white matter.8 The index assumes that the direction of perivascular spaces is in the x-direction (toward the cortex) on an axial plane, near the ventricular wall in the deep white matter. Large 5 mm regions of interest are placed in clearly defined locations on DTI-color Fractional Anisotropy (FA) maps. The acquisition protocol uses a standard clinical DTI with two b-values (0 and 1000 mm2/second), making it suitable for retrospective data analysis. The index is calculated as a ratio between diffusivities across the x, y, and z axes. Interpretation of the index is straightforward: a high index suggests greater diffusivity toward the cortex, while a low index indicates no preference for direction in the selected region of interest. Though it has become popular for studying "glymphatic dysfunction" in various neurological conditions, the DTI-ALPS index remains aspecific, as similar alterations in fluid motion can be seen in a variety of neurological disorders. Moreover, changes in fluid motion in small regions of the brain do not reflect the brain's overall clearance capacity. If all diseases alter the DTI-ALPS index, it would suggest that all neurological disorders require the same treatment aimed at correcting fluid flow and reducing toxic waste deposits. This suggests a need for further investigation into the true significance of the index. The DTI-ALPS index has yet to be validated and may reflect other features of white matter that are measured by other DTI-based indices, such as fractional anisotropy, radial diffusivity, axial diffusivity, and mean diffusivity. The positioning of region of interest is questionable. Optimal b-values have not been calculated. The inventors acknowledge that scanner type, imaging plane, and basic sequence parameters can affect measurements.9 We encourage the MRI research community to develop methods to validate the DTI-ALPS index to ensure it truly reflects altered fluid flow in the context of brain clearance capacity, rather than merely serving as an index without clear clinical significance. There is a need to develop clinically feasible MRI-based techniques to better understand brain waste clearance. To address the issue of low temporal resolution, techniques such as continuous arterial spin labeling may prove useful. For spatial resolution, efforts are being made to improve the detection of intraparenchymal slow flow using diffusion tensor-based intravoxel incoherent motion imaging. While promising, these techniques are still in their early stages of development, and widespread clinical adoption is yet to be seen. Accurately assessing waste clearance pathways in neurological disorders is essential for deepening our understanding of diseases with unclear etiology. Such advancements could lead to novel treatment strategies and drug delivery methods. As a clinical neuroradiologist, I would be thrilled to have a reliable MRI method that could directly assess waste clearance dysfunction in real time. And I am sure I would not be the only one eager for such a development!