Encyclopedia
Please note this is a comparison between Version 1 by Marco Parillo and Version 2 by Lindsay Dong.
Brain spaces around (perivascular spaces) and alongside (paravascular or Virchow–Robin spaces) vessels have gained significant attention due to the advancements of in vivo imaging tools and to their crucial role in maintaining brain health, contributing to the anatomic foundation of the glymphatic system. In fact, it is widely accepted that peri- and para-vascular spaces function as waste clearance pathways for the brain for materials such as ß-amyloid by allowing exchange between cerebrospinal fluid and interstitial fluid. Visible brain spaces on magnetic resonance imaging are often a normal finding, but they have also been associated with a wide range of neurological and systemic conditions, suggesting their potential as early indicators of intracranial pressure and neurofluid imbalance.
- perivascular spaces
- paravascular spaces
- Virchow–Robin spaces
- PVS
- glymphatic system
- neurofluids
- magnetic resonance imaging
- MRI
1. Introduction
Perivascular spaces are generally described as fluid-filled areas surrounding arterioles, capillaries, and venules in the brain that allow for the passage of fluid or particles [1]. Durand-Fardel and Pestalozzi first reported the existence of brain perivascular spaces in 1842 and 1849, respectively [1]. Rudolf Virchow and Charles Robin also reported spaces surrounding the perforating vessels in the basal ganglia and hemispheric white matter in postmortem brain specimens visible to the naked eye in 1851 and 1859, respectively [2][3][2,3].
Despite being described in the human brain more than 150 years ago, perivascular spaces have gained significant attention in recent years due to advancements of in vivo imaging tools. Moreover, evidence from both clinical and pre-clinical studies indicates that perivascular spaces play a crucial role in maintaining brain health, and their dysfunction is linked to various neurological disorders. The introduction of the glymphatic system hypothesis favored confusion in the literature about the definition and use of terms in describing spaces around and along vessels in the brain’s microanatomy, often used as synonyms (e.g., Virchow–Robin spaces, perivascular spaces (PeVS), and paravascular spaces (PaVS)). It is widely accepted that PaVS function as waste clearance pathways for the brain for materials such as ß-amyloid by allowing an exchange between cerebrospinal fluid (CSF) and interstitial fluid and forming the anatomic foundation of the glymphatic system [4][5][4,5].
2. Anatomy
The anatomy of PeVS and PaVS is intricate and remains a matter of debate. In humans, the cortical surface of the brain and the intracranial perforating arterioles are covered by the pia mater [6][7][9,10]. Arterioles that traverse the base of the brain and enter the basal ganglia are enveloped by two layers of leptomeninges along their length, forming a space between them that communicates with the subarachnoid space [8][9][11,12]. Arterioles in the cortex and venules have a single leptomeningeal membrane that adheres closely to the vessel wall without an outer layer, permitting that the spaces surrounding these vessels communicate with the subpial space directly and, somehow, indirectly with the subarachnoid space [10][11][12][13,14,15]. These spaces can be defined as PaVS and correspond to the “Virchow–Robin spaces”. They are bound internally by the pial sheath and externally by the basement membrane of astrocytes (glia limitans) on the arterial side and the vein wall and the glia limitans on the venous side [13][14][16,17]. The glymphatic model suggests that CSF flows through para-arterial spaces, interacts with interstitial fluid and solutes, and then is removed from the brain through para-venous spaces [15][18]. There are also other spaces within the arterial tunica media, located between the middle layers of the basement membrane of arterial smooth muscle cells [14][17]. These structures are called PeVS, and they serve as channels for interstitial fluid flow in the opposite direction to that of blood flow (also known as the intramural periarterial drainage pathway) towards the cervical lymph nodes through major cerebral arteries in the neck [13][16][16,19]. The venous vasculature does not contain PeVS [14][17]. In summary, PeVS are spaces around the blood vessels (between the basement membrane of the pial sheath and the basement membrane of the arteries), while PaVS are spaces alongside the blood vessels (outside the vessel wall) (Figure 1).Figure 1. Schematic drawing of paravascular and perivascular spaces currently believed micro-anatomy along and around basal arterioles, cortical perforating arterioles, veins, and venules. SAS: subarachnoid space; SPS: subpial space; ○: paravascular space; ●: perivascular space; *: vessel lumen; dotted line: arterioles endothelium; dash-dot line: vein and venules endothelium; dashed line: meningeal membrane; black line: glia limitans; grey line: meningeal membrane plus glia limitans. (A) The brain’s basal arterioles originate from the subarachnoid space and are surrounded by two leptomeningeal membranes. The inner membrane is in close proximity to the arteriolar wall, while the outer membrane is connected to the pia mater, allowing the basal para-arteriolar spaces to have direct communication with the SAS. The perivascular space is another space within the arterial tunica media. (B) In cortical arterioles, where there is only one leptomeningeal membrane closely applied to the vessel wall and no outer layer, the para-arteriolar space is believed to communicate with the SPS rather than the SAS. The perivascular space is another space within the arterial tunica media. (C) Veins and venules have only one leptomeningeal membrane that is closely attached to the vessel wall without any outer layer, and thus it is believed that the paravenous space communicates with the SPS rather than the SAS. The venous vasculature does not contain perivascular space.
3. Imaging
Based on the anatomy previously described, PeVS and PaVS are not distinguishable on clinical imaging. Although PVS are located around both arteries and veins, recent studies utilizing 7 Tesla MRI have indicated that the majority of PVS visible with MRI in the centrum semiovale are para-arterial rather than para-venous [8][17][18][11,21,22], but this does not exclude the presence of venular pathology (e.g., venous collagenosis) [19][23]. The cause of para-venous spaces being less visible at MRI could be related to the smaller size of para-venous spaces and differences in the quantity and/or composition of the para-venous fluid [20][24].
PVS may contain proteinaceous material, including extracellular matrix, fibrin/fibrinogen, and hemosiderin deposits, in addition to interstitial fluid [21][22][25,26]. However, it is still unclear how these deposits affect the signal and visibility of PVS on MRI in vivo [23][24][27,28]. Hence, the densitometric and signal intensity characteristics of PVS are typically comparable to those of CSF (Figure 2) [20][24].
Figure 2. Typical appearance of symmetrically enlarged PVS at the level of the basal ganglia (type I) with densitometric and signal intensity characteristics typically comparable to those of cerebrospinal fluid (black arrows). (A) Unenhanced computed tomography shows hypodense PVS. (B) A 1.5 T magnetic resonance imaging (MRI) fluid attenuated inversion recovery image shows hypointense PVS. (C) A 1.5 T MRI apparent diffusion coefficient image shows no diffusion restriction. (D) A 1.5 T MRI T2-weighted turbo spin echo image shows hyperintense PVS with a microscopic central vessel inside, known as the “vessel sign”. (E) A 1.5 T MRI T1-weighted spin echo image shows hypointense PVS.
PVS appear on computed tomography (CT) scans as well-circumscribed fluid-density spaces without contrast enhancement or calcifications. On MRI, PVS display hypointensity on T1-weigthed and fluid-attenuated inversion recovery (FLAIR) images, as well as hyperintensity on T2-weighted images, with no signs of contrast enhancement or mass effect; the signal of the surrounding brain is typically normal. Since PVS are communicating compartments, they do not exhibit restricted diffusion on diffusion-weighted imaging (DWI) [25][29]. A high-resolution MRI can reveal a microscopic central vessel within the PVS, known as the “vessel sign” [26][30]. Indeed, MRI has greater sensitivity in detecting PVS than CT, especially using T2-weighted and cisternographic sequences [25][29], and contrast agent administration is not required (avoiding the problems associated with gadolinium deposits [27][28][31,32]). Regarding morphology, when the penetrating artery is parallel to the imaging plane, PVS seems to have a striped appearance, whereas when they are perpendicular, PVS appear rounded or ovoid [25][29]. In terms of location, PVS are usually observed symmetrically in certain areas of the brain, such as the basal ganglia (including the lentiform nucleus, along the anterior commissure, and in the internal and external capsules), located just above the anterior perforated substance, and the centrum semiovale, running towards the lateral ventricles from the external aspect of the white matter with the highest concentration (PVS-to-white-matter ratio) in the subinsular white matter [29][33]. They are also found in the hippocampus, midbrain, pons, and, in some cases, in cerebellar white matter [1]. PVS are commonly categorized into three types based on their location.
Enlarged and giant PVS may show surrounding T2/FLAIR signal abnormalities, indicating gliosis [30][40]. Giant PVS are most commonly found in the mesencephalothalamic region (type III location), where they can cause a mass effect on the aqueduct of Sylvius and finally lead to obstructive hydrocephalus (Figure 3A,B) [31][32][33][41,42,43]. While headaches are the most common symptom, other non-specific neurological symptoms such as memory loss, seizures, dizziness, syncope, visual, balance, or concentration impairment may also be present. In addition, even very large PVS can be completely asymptomatic [34][35][44,45]. A specific type of tumefactive PVS, referred to as “type IV”, has been identified in the subcortical white matter of the anterior superior temporal lobe. These giant PVS typically have an elevated perilesional FLAIR signal with no mass effect that indicates gliosis and are commonly related to a branch of the middle cerebral artery and a focal region of cortical absence or thinning (Figure 3C,D) [36][37][46,47].
Figure 3. A 1.5 T magnetic resonance imaging appearance of giant PVS in T2 turbo spin echo images (A,C) and fluid attenuated inversion recovery (FLAIR) images (B,D). (A,B) Patient with giant PVS in the right mesencephalothalamic region (type III) that cause a mass effect on the aqueduct of Sylvius. (C,D) Patient with a right anterior temporal lobe enlarged perivascular space (type IV) characterized by an elevated perilesional FLAIR signal with no mass effect, which indicates gliosis.
4. Quantification
4.1. Visual Scoring
Different types of visual rating systems have been implemented, including the Potter scoring system [38][49]. This system categorizes PVS into five grades based on their number in the basal ganglia and centrum semiovale (grade 0 = none; grade 1 = 1–10; grade 2 = 11–20; grade 3 = 21–40; grade 4 ≥ 41 corresponding to the “état criblé” or “status cribrosum” described by Durand-Fardel in the basal ganglia [1]). Each hemisphere is scored separately, and the higher score of the two is used (Figure 4).
Figure 4. The 1.5 T magnetic resonance imaging T2 turbo spin echo images in different patients show the Potter scoring system based on counting visible PVS. Each hemisphere is scored separately for PVS, and the higher score of the two is used, as indicated at the bottom. (A–D) Axial slices at centrum semiovale level (type II). (E–H) Axial slices at basal ganglia level (type I), where H corresponds to the “état criblé” or “status cribrosum” described by Durand-Fardel in the basal ganglia. Note also the presence of a lacune in (D) (black circle) and that PVS in the centrum semiovale can be observed as they approach the cortex; their dilation increases as they reach the inner edge of the cortex; however, they cannot be seen as they pass through the cortex.
4.2. Automatic and Semi-Automatic Segmentation and Morphometry
Advances in technology have led to the emergence of various automatic and semi-automatic methods for segmenting and quantifying PVS. Once the PVS are segmented, various metrics and morphological features can be computed from segmentation masks [20][39][24,52]. The PVS volume is an indicator of the amount of fluid within the PVS [29][33]. The mean cross-sectional diameter can be utilized to differentiate PVS from other cerebral small vessel disease-related lesions [40][53]. Linearity represents the similarity of a particular PVS cluster to the tubular morphology [41][54], while solidity reflects the shape complexity of PVS, where a lower solidity indicates a more tortuous and less compact shape [20][24].5. Associations
5.1. Physiological
It is widely known that the visibility of PVS increases with age, especially in the basal ganglia, centrum semiovale, and hippocampus [42][66]. A recent study documented low PVS volumes in the basal ganglia and centrum semiovale in healthy subjects up to the age of 35 despite high Potter visual scores in the white matter of young subjects; also, it was suggested that in the basal ganglia, a Potter grade 1 and a Potter grade 3 could be considered physiological until the age of 50 and from the age of 70 onward, respectively, while in the centrum semiovale, a Potter grade 4 could be considered physiological from the age of 50 [43][67]. A higher PVS burden has been associated with the male sex, and it is also seen in healthy adolescents [29][44][33,65]. A circadian dynamic nature of the amount of MRI-visible PVS (higher PVS burden is visible on MRI at later times of the day) with a predominantly asymmetric distribution has been suggested. This could be associated with circadian fluctuations in respiration or/and blood pressure (which are known regulators of perivascular flow) or may be linked to the circadian control of flow through aquaporin-4 (a water channel that facilitates fluid transport from the PVS to the cerebral parenchyma) [29][33].5.2. Pathological
It is widely known that the visibility of PVS in the basal ganglia (but not in the centrum semiovale) increases with hypertension [42][66], and hypertension was found to be related to a low DTI-ALPS index [45][70]. Historically, MRI biomarkers of cerebral small vessel disease have included enlarged PVS along with other indicators such as recent small subcortical infarcts, lacunes, white matter hyperintensities (WMH) of presumed vascular origin, cerebral microbleeds, and brain atrophy, which are often present in association with stroke or cognitive decline.6. Differential Diagnosis
Enlarged PVS (diameter between 3–5 and 15 mm) most frequently need to be differentiated from lacunes and WMH as they are similar in signal characteristics and they often coexist in the context of cerebral small vessel disease (Figure 5A–F) [25][46][29,104]. When PVS appear notably enlarged (e.g., giant PVS with diameter ≥ 15 mm), causing mass effect, and presenting cystic configurations, they must be distinguished from other cystic pathologies of possible congenital, vascular, inflammatory, neoplastic, or hereditary nature (Figure 5G–L) [47][48][105,106].Figure 5. The 1.5 T magnetic resonance images show the main differential diagnosis for enlarged PVS in the context of cerebral small vessel disease (A–F) and 2 examples of differential diagnosis for giant PVS (G–L) using T2 turbo spin echo images (A,D,G,J), FLAIR images (B,E,H,K) and apparent diffusion coefficient images (C,F,I,L). (A,B) Patient with a small area of altered signal in the left centrum semiovale (black circle) characterized by hyperintensity in T2 and FLAIR images and diffusion restriction, therefore compatible with a recent subcortical infarction. (D–F) Patient with a small area of altered signal in the right centrum semiovale (empty black arrow) characterized by hyperintensity in the T2 image, hyperintense rim in the FLAIR image, and no diffusion restriction, therefore compatible with a lacune. Note also the white matter hyperintensities in centrum semiovale bilaterally. (G–I) Patient with a dysembryoplastic neuroepithelial tumor in the right medial temporal lobe (thick black arrow), with a distinct “bubbly” cystic appearance associated with T2/FLAIR signal abnormality and no diffusion restriction. (J–L) Patient with a left parieto-occipital neurenteric cyst characterized by a slightly hyperintense content in FLAIR images and no diffusion restriction.
6.1. Recent Small Subcortical Infarct
It is an infarct that occurred within the territory of a perforating arteriole in a time span of a few weeks. It typically has a maximum diameter of ≤20 mm in the axial plane, exhibits restricted diffusion on DWI, and increased signal intensity on FLAIR and T2-weighted sequences [25][29].
6.2. Lacune of Presumed Vascular Origin
It is characterized by a subcortical, round, or ovoid cavity filled with fluid that typically ranges from 3 mm to 15 mm in diameter. It is commonly related to a prior small subcortical infarct or hemorrhage that occurred within the territory of a perforating arteriole [25][29]. In FLAIR images, lacunes typically exhibit a central hypointensity resembling that of CSF, surrounded by a hyperintense rim. Nevertheless, the presence of a rim is not always detectable, and PVS may also have a hyperintense rim when passing through an area of WMH. When the suppression of fluid signal in the central cavity is not achieved with FLAIR imaging, it may have a clear CSF-like intensity on T1-weighted and T2-weighted sequences. When lacunes are numerous, the radiological picture is described as “état lacunaire”, a possible substrate for multi-infarct vascular dementia [49][107].
