By Susan Gauthier, DO, MSc
Associate Professor of Clinical Neurology, Weill Cornell Medical College
SYNOPSIS: Perivascular space (PVS) abnormalities were demonstrated in two cohorts of patients with neuromyelitis optica spectrum disorder (NMOSD) as compared to healthy controls. Magnetic resonance imaging provided visualization of PVS enlargement, and quantification of reduced flow in NMOSD patients, both of which were found to be associated with disease outcomes.
SOURCE: Cacciaguerra L, Carotenuto A, Pagani E, et al. Magnetic resonance imaging evaluation of perivascular space abnormalities in neuromyelitis optica. Ann Neurol 2022;92:173-183.
Studies have suggested that perivascular space (PVS) abnormalities, such as dilation or reduced flow, may relate to the pathogenesis of several neurodegenerative diseases. Specifically, alterations in the PVS have been implicated as a potential cause of reduced clearance in the recently described “glymphatic system” of the central nervous system (CNS). Neuromyelitis optica spectrum disorder (NMOSD), an autoimmune disease caused by autoantibodies targeting the aquaporin-4 (AQP4) water channel, is thought to damage astrocytes, which highly express AQP4 and, importantly, are membrane-bound within the PVS. NMOSD is classically characterized by severe relapses of optic neuritis and myelitis. However, recent studies have suggested that a global neurodegenerative process can be observed (i.e., gray matter loss and cognitive impairment). In this study, Cacciaguerra et al sought to determine if PVS abnormalities are present in patients with NMOSD and if these findings may relate to clinical and magnetic resonance imaging (MRI) structural outcomes. The investigators hypothesize that this pathway may contribute to neurodegenerative changes observed in NMOSD, as has been suggested for other neurological diseases.
This study included two cohorts of well-matched NMOSD patients and healthy controls (HC), represented as exploratory and validation datasets. The investigators visually evaluated the size of PVS on axial T2-weighted images within the basal ganglia and white matter of the centrum semiovale. PVS enlargement was assessed according to the established Potter’s score.1 A score equal to 0 corresponds to the absence of enlarged perivascular spaces, a score of 1 to 1-10, a score of 2 to 11-30, a score of 3 to 21-40, and a score of 4 to > 40 enlarged perivascular spaces per region. Flow rate or diffusion along PVS was measured with diffusion tensor imaging (DTI), in association with susceptibility-weighted imaging for vessel visualization. Diffusion tensor imaging along perivascular spaces (DTI-ALPS index) was calculated as the ratio of water diffusivity parallel to vessels, and diffusivities perpendicular to white matter tracts. This approach has been used previously to measure flow within the PVS in several other studies.
The same two regions of interest (ROI) were identified in the left hemisphere of all subjects and represented both white matter fiber projections and associated fibers. Susceptibility MRI for vessel visualization was only available in the exploratory MRI, thus the investigators used these scans to direct placement of ROI in the validation dataset. They demonstrated a high intraclass correlation (ICC = 0.97) comparing manual vs. automated ROI placement in the exploratory dataset, providing support for the use of this approach in the validation set.
NMOSD patients had a lower cortical volume as compared to HC in both exploratory and validation datasets, confirming previous observations of a neurodegenerative component to this disease. Combining the datasets, enlargement of PVS within white matter was appreciated in NMOSD patients as compared to HC. In both exploratory and validation datasets, the automated DTI-ALPS index was statistically lower in NMOSD subjects compared to HC.
Several correlations were found to be significant between enlargement of PVS and lower gray matter structural measures as well as impairment on the Paced Auditory Serial Addition Test, a measure of cognitive processing speed. Interestingly, no correlations were found between DTI-ALPS index and PVS scores. In a pooled analysis, regression models were implemented to determine the association of the various MRI features with the disease classification and clinical disability, as measured by expanded disability status scale (EDSS). The investigators found that lower DTI-ALPS index, deep gray matter volume, and cortical volume were independently associated with having a diagnosis of NMOSD; the final model including all three variables had an adjusted model R2 = 0.62. In the second model, a higher number of myelitis, a lower DTI-ALPS index, and higher lesion volume all were associated with a higher EDSS (final model adjusted R2 = 0.55). These findings were not reproduced when processing speed was considered as the clinical outcome.
COMMENTARY
This study highlights recent research regarding the potential importance of PVS, wherein these spaces may function to provide essential drainage or clearance for the CNS. It is thought that accumulation of protein aggregates and, potentially, a reduced density of AQP4 water channels may contribute to “glymphatic impairment” in neurodegenerative diseases. Results of this study demonstrate that PVS abnormalities are detected in NMOSD, and impaired flow or function may have implications on disease outcomes.
As imaging technology advances, the “glymphatic impairment” hypothesis can be explored further and validated among different neurological diseases. Although the specific mechanisms leading to PVS abnormalities among various neurodegenerative diseases are unknown and likely to vary, this work underscores the potential for a common mechanistic pathway leading to neuronal damage and could have widespread implications for novel therapeutic strategies to prevent neurodegeneration.
REFERENCE
- Potter GM, Chappell FM, Morris Z, Wardlaw JM. Cerebral perivascular spaces visible on magnetic resonance imaging: Development of a qualitative rating scale and its observer reliability. Cerebrovasc Dis 2015;39:224-231.
