2020 Volume 6 Pages 12-22
Vascular dementia (VaD) is regarded as the second most common type of dementia but recent surveys suggest vascular cognitive impairment (VCI) is even more frequent in most populations worldwide. White matter hyperintensities (WMHs) detected on brain MRI are the radiological signature of cerebral small vessel disease (SVD), which is highly characteristic in subcortical ischaemic vascular dementia (SIVD), the most prevalent subtype of VaD. Vascular risk factors are strongly associated with WMHs and development of post-stroke VaD or dementia. The gliovascular unit (GVU) has a critical role in SIVD. Therefore, any component of the GVU could be a therapeutic target for VaD. When considering pharmacological approaches, more attentions ought to be paid onto pleiotropic effects of existing drugs. Autonomic dysfunction is highly prevalent in VaD patients and is a treatable factor to protect GVU from VaD pathologies. As non-pharmacological approaches for VaD, environmental enrichment (EE) and physical exercise training, particularly limited EE rather than full-time EE, have been proved to preserve GVU integrity in VaD. Glial responses, especially clasmatodendrosis, would be a novel therapeutic target for VaD. Animal models of VaD are useful to demonstrate pathophysiology, explore and establish safe and effective treatments. Nevertheless, pathophysiological substrate of VaD is heterogeneous. Multimodal combination treatments targeting GVU, implementing both pharmacological and non-pharmacological interventions, would be a promising interventional strategy to counter vascular dementia.
Current trends suggest a high burden of cerebrovascular disorders worldwide that is concomitant with increased frequency of vascular cognitive impairment (VCI). There is no cure for VCI or for vascular dementia (VaD), the second most common type of dementia among the ageing population (1). It is therefore timely that safe and effective interventional strategies, which delay onset, slow down cognitive decline in patients with cerebrovascular disorders, or reduce the burden of VaD are implemented urgently. Subcortical ischaemic vascular dementia (SIVD) is the most prevalent subtype of VaD, where elderly exhibit disability and deficits in cognitive function over long periods of their remaining lives (2)(3). SIVD is primarily characterised by cerebral small vessel disease (SVD), which is described by a variety of pathologies including lacunar infarcts, microinfarcts, microbleeds and white matter (WM) lesions (4)(5)(6). It is thought that diffuse WM changes linked to SIVD largely result from a chronic hypoperfusive state or cerebrovascular insufficiency during ageing (7)(8). While the pathological mechanisms that lead to the progression of VCI because of SIVD or post-stroke patients who develop dementia remain unclear, it is urgent to target strategies, which restore or maintain brain perfusion. In this article, we will review the latest findings regarding pathogenesis and potential novel interventional strategies for VCI and VaD.
It is now abundantly clear that reducing risk of vascular disease reduces risk of cognitive impairment or dementia. Several longitudinal studies suggest that the strongest predictors for development of dementia after stroke was the presence of cardiovascular risk factors. In our large longitudinal prospective cohort of the Cognitive Function After Stroke (CogFAST) study (9) in elderly stroke survivors, who predominantly had small infarcts or SVD type of changes, we found that during the mean follow-up period of ~4 years, up to 25% of subjects developed dementia after their first episode of stroke. At baseline, none of them had confirmed to have dementia. Neuroimaging and post-mortem investigations revealed that more than 75% who developed dementia could be diagnosed to have VaD, lacking significant Alzheimer type of pathology. The study remarkably revealed that the strongest predictors of dementia after stroke was multiple (two or more) cardiovascular risk factors including hypertension, hyperlipidemia and diabetes mellitus. This fact implies that risk to developing dementia after stroke is strongly related to the presence of vascular risk factors and VaD type of pathology. We also found that cognitive processing speed and performance on measures of attention were significantly associated with greater white matter hyperintensities (WMH) volumes, particularly in the frontal lobe regions and were worse in those with hypertension (10). We further found that the presence of an APOE epsilon4 allele was associated with greater progression of cognitive decline in stroke survivors. This also has implications for interventions aimed at the secondary prevention of dementia in stroke patients (11). However, in clinical settings, extensive treatment or control of vascular risk factors in old age is essential to prevent or slow down the progression of VCI and prevent VaD (3).
We note that cerebral SVD manifests in several cortical and subcortical insults. However, microvascular abnormalities are key to SVD. Microvascular changes including arteriolosclerosis, intimal thickening, fibroid necrosis, hyalinization, and enlarged perivascular spacing occurring during ageing but these lesions are at their peak or most severe in VaD. These changes are also prominent in post-stroke survivors, who develop dementia (PSD) and those who do not (PSND). While it is difficult to differentiate the burden of these lesions between PSD and PSND, it is noteworthy that they both exhibit higher total vascular pathology scores (12) compared to normal ageing control subjects (13). With the microvascular network, even capillaries are affected particularly in the rarefied white matter (WM) in both PSD and VaD (14). Collapsed and string microvessels, microaneurysm-like structures and tortuous vessels were frequently observed in patients with PSD and VaD. In terms of tissues changes, greater frontal WM volumes separate post-stroke survivors who develop dementia or PSD versus those who remain stable (PSND). Among post-stroke patients, we also noted that the prevalence of microinfarcts was greater in PSD compared to PSND subjects, suggesting the presence of cerebral microinfarcts is one of the predictors of dementia after stroke (9).
There is evidence that SVD pathology may also be exacerbated by autonomic dysfunction. Our recent study also revealed detrimental effects of autonomic dysfunction on severity of VaD and other dementias (13). At least 40% of dementia patients, including VaD patients, exhibit some form of autonomic dysfunction in life (15). The proposed mechanism of autonomic dysfunction on vascular pathology is that recurrent episodic hypotension promotes chronic cerebral hypoperfusion. This then causes ischaemic damage or rarefaction of the WM, and eventually leads to disconnection of the WM and impaired cognitive function (16). In our study, autonomic dysfunction was associated with higher burden of SVD changes, particularly in the deep frontal WM in patients with age-related dementia including VaD (13). We emphasize the importance of screening autonomic function for elderly dementia, especially VaD patients in clinics. The most critical finding was that in some cases, autonomic dysfunction is likely ‘asymptomatic’ (17). Thus, there are dementia patients who have no apparent clinical symptoms of autonomic dysfunction e.g. repeated falls and syncope. These individuals could be only proved to have dysautonomia by clinical examination. However, ‘asymptomatic’ autonomic dysfunction also has the same adverse effects on cerebrovascular pathology as symptomatic dysautonomia. Therefore, ‘asymptomatic’ autonomic dysfunction is also a risk to develop SVD and WM damage, similar to symptomatic dysautonomia. Treatments to improve autonomic dysfunction, even for preciously detected ‘asymptomatic autonomic dysfunction’ cases, implementing lifestyle intervention or pharmacological approaches or combination of both would ameliorate SVD pathology and contribute to prevent or slow down cognitive decline in patients diagnosed with VaD and have features of autonomic dysfunction. Collectively, these findings strongly implicate that ischemic insults or chronic cerebral hypoperfusion cause or worsen ageing associated SVD and VaD.
White matter hyperintensities (WMH), detected on brain magnetic resonance imaging (MRI) of T2-weighted imaging (T2WI) and fluid-attenuation inversion recovery (FLAIR) imaging, are associated with cognitive dysfunction in patients with SVD, stroke and VaD (18), most probably due to loss of white matter integrity. In our longitudinal CogFAST cohort, volume of WMHs were associated with higher mortality rate, shorter time to dementia onset, and were also an independent predictor of survival to dementia (19). These findings indicate that severity of WMHs is related to loss of white matter integrity and impairment in functional connectivity within cortical-subcortical circuits to instigate dementia and reduced life expectancy after stroke.
The pathophysiological substrates of WMH are heterogeneous (20). One of the key features of WMH is loss of myelin rather than axonal degeneration. We have previously reported severe myelin loss in the frontal white matter of VaD patients compared to other types of age-related dementias (5), suggesting disruptions of frontal-subcortical circuits are more profound in VaD compared to other dementias (21). This observation is consistent with finding from a study showing decreased myelin proteins in the WM in SVD and VaD (22). Loss of oligodendrocytes as another substrate of WMH (20) is frequently observed in SVD and VaD patients (23).
Other substrates of WMH are inflammatory changes, evident in form of gliosis including astrogliosis, causing gliovascular unit (GVU) dysfunction and blood brain barrier (BBB) disruption (20). The GVU (Figure 1) controls and maintains the functions of the BBB. The cellular components of the GVU in the white matter (WM), consists of endothelial cell, pericyte, astrocyte, microglia and oligodendrocyte (23). In physiological conditions, these cells are interacting each other to maintain brain homeostasis through BBB. Loss or damage of microvascular, mural and glial cells caused by ischemic injury or chronic cerebral hypoperfusion results in dysfunction of GVU (24) and increased BBB permeability, which may promote diffuse WM changes and cognitive decline (25). Cilostazol, a phosphodiesterase type III inhibitor, restored BBB damage in an experimental stroke model in mice (26). EE also showed protective effect on BBB in a rat model of chronic cerebral hypoperfusion (27). Therefore, protection of GVU or its cellular components, especially in the WM, would be a key strategy to preserve BBB and WM integrity, as well as to prevent subsequent strokes and cognitive decline.
The GVU comprised of various cellular components, plays a critical role to control and maintain functions of the blood brain barrier (BBB). Abbreviations; AQP4, Aquaporin-4; BM, basement membrane.
Astrocytes, one of the major cellular components of the GVU, maintain brain homeostasis by interacting with blood vessels via astrocytic end-feet. Perivascular end-feet of astrocytes have a critical role to regulate e.g. electrolytes, amino acids and water homeostasis in brain (28). The water homeostasis in brain is regulated by water channel family aquaporins (29). Aquaporin-4 (AQP4) is present on astrocytic end-feet surrounding small vessels in brain and exchanging water between vascular lumen and brain parenchyma. AQP4 also play a role in pathologic conditions, e.g. reduce oedema formation after cerebral ischemia (30).
Astrocytes are sensitive to various stimuli and transform to reactive cells as activated astrogliosis, often observed in response to ischaemic insults (31). Previous reports suggest that reactive astrocytes caused elevated reactive oxygen species (ROS) and induced cellular inflammation (32), which progresses to more widespread brain injury (33). Therefore, targeting of astrocytes which may return reactive astrocytes to a quiescent phenotype (31) would represent a therapeutic target for VaD. Recent report showed that the opioid analgesic oxycodone suppressed detrimental reactive astrogliosis by inhibiting nuclear factor kappa-B (NF-κB) signaling in reactive astrocytes (34).
A phenomenal morphological change in astrocytes is described as clasmatodendrosis, a cluster of damaged astrocytes (clasmatodendrocytes) with hypertrophic cell bodies and loss of or retracted processes (35). We previously found the presence of severe clasmatodendrosis in the deep WM of post-stroke survivors who developed dementia, and also had greater WM hyperintensity (WMH) volumes (19). Similar astrocytopathy was observed in patients with hereditary cerebrovascular disease, cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) (36). CADASIL is also characterized by strategic WMH detected on brain MRI. Therefore, clasmatodendrosis is one of the key pathophysiological markers of progressing WM lesions. In addition to clasmatodendrosis, abnormal distribution of AQP4, characterised by aggregation of AQP4 at the periphery of GFAP-positive astrocytes/clasmatodendrocytes, was evident in post-stroke survivors who developed dementia with greater WMH compared with age-matched normal control subjects
(19). These observations strongly implicate that clasmatodendrosis and AQP4 dislocation in VaD patients is caused by vascular changes, pathological alterations in GVU and BBB disruption.
For the past decades, experimental animal models of cerebral hypoperfusion (37), which restrict cerebral blood flow by occluding or narrowing cervical and cerebral arteries, have been extensively analysed in mice (38), (39), (40), rats (41), (42), (43), (44), (45), (46), gerbils (47), as well as non-human primates (19), aiming to elucidate the pathological mechanisms of developing SVD and VaD. In addition to these models, bilateral common carotid artery stenosis (BCAS) in mice, as a mouse model of chronic cerebral hypoperfusion, has been established (48), (49). BCAS mouse model exhibited BBB dysfunction (50), changes in glial cells such as: loss of oligodendrocytes, astrogliosis and microglial proliferation, particularly in the WM, as well as hippocampal pathology (51). As BCAS model characterises similar cerebrovascular pathologies to human VaD patients, it has been widely used and has been accepted as one of the most useful rodent models of VaD (52), (53), (54).
The BCAS mouse model has been very useful to uncover pleiotropic effects of pharmacological agents, role of genes and cell therapy against SVD and VaD. For example, BCAS has been utilised to evaluate efficacies of an anti-platelet agent cilostazol, phosphodiesterase type 3 (PDE3) inhibitor (55), (56); anti-hypertensive drug: telmisartan, an angiotensin II type 1 receptor blocker (57); an antibiotic agent: minocycline, tetracycline (58); an angiogenic peptide: adrenomedullin (59); a gene silencer: silent information regulator 2 homolog 1 (SIRT1) (60), (61) and bone marrow derived mononuclear cells (62). All of these compounds and agents showed protective effects on cerebrovascular pathologies, WM changes and cognitive decline induced by BCAS, suggesting these are promising interventional strategies to ameliorate VaD pathologies. Notably, a recent study proved that cilostazol, a phosphodiesterase type 3 (PDE3) inhibitor, slowed down cognitive decline in patients with mild cognitive impairment (63), and has been successfully applied for a prospective multicenter clinical trial (64).
Our past studies have also successfully reproduced WM changes similar to VCI and VaD patients in the long-term version of the BCAS model (65). We explored the effects of differing degrees of environmental enrichment (EE) on WM pathological changes. Long-term BCAS in mice caused WM damage characterised by WM atrophy, WM disintegration and loss of oligodendrocytes, resulted in cognitive dysfunction. These unfavourable changes were attenuated by EE more effectively by limited exposure to EE (only 3 hours a day) rather than full-time (24 hours a day) exposure to EE. Thus EE, especially limited amount of EE and physical exercise training, appears a safe and effective interventional strategy for VaD patients with extensive WM pathology.
Physical activity or exercise has been shown as a strategy to slow down cognitive impairment in humans with VCI (66). Previous experimental evidence a rat model of chronic cerebral hypoperfusion showed that environment-induced physical exercise and cognitive stimulation as well as social interactions, which are incorporated in EE and physical exercise training, exhibit beneficial effects on cognitive dysfunction (27), (67), (68). These studies reported that EE preserved BBB integrity, increased brain plasticity, enhanced neurogenesis, and increased synaptogenesis. Furthermore, combination of EE and physical exercise enhanced neurogenesis (69) and upregulation of genetic expressions associated with neurogenesis, synaptic plasticity, neuroprotection and intact memory function (70), resulted in preserved cognitive function (71). EE also has been reported to have favourable effects on motor function recovery after experimental stroke (72), (73). Thus, EE and even moderate physical exercise training could be safely and consistently encouraged in man and therefore an effective interventional strategy for patients with SVD, VaD and stroke.
Our research group has also successfully replicated astrocytopathy, including reactive astrocytes and clasmatodendrosis with aberrant distribution of AQP4, similar to VaD patients in the BCAS model (74) as well as in a non-human primate model of cerebral hypoperfusion (19). Long-term BCAS in mice caused astrogliosis, clasmatodendrosis with AQP4 dysfunction in the WM. AQP4 dislocation and clasmatodendrosis may enhance WM pathology due to the disturbance of water homeostasis. We found that limited regime of EE not only reduced astrocytopathy but ameliorated other aspects of GVU functions (74). Thus, glial responses, especially clasmatodendrosis, would be a new aspect to explore pathogenesis and establish novel treatment for VaD. This suggests limited exposure to EE preserves BBB function via maintenance of GVU integrity. We surmise that long-term BCAS is a relatively reliable model of VaD, which accurately replicates several features of the pathophysiology of WM changes as evident in VaD patients. EE, especially limited amount of EE and physical exercise training, appears a safe and effective interventional strategy to attenuate unfavourable effects of cerebral hypoperfusion and slow down subsequent cognitive decline in VaD.
We have reviewed pertinent findings regarding the pathogenesis of VaD and its potential therapeutic targets. Gliovascular disruption caused by ischemic insults or chronic cerebral hypoperfusion has been suggested to play critical role to develop VaD pathology in brain. Therefore, each component of GVU would be a therapeutic target for VaD. Firstly, strict control of VRFs is essential to protect GVU from VaD pathology. It is also important to explore and establish safe and effective strategies. Pharmacological approaches, which value more for pleiotropic effects of existing drugs e.g. antiplatelet agents such as cilostazol, or antihypertensive drugs such as angiotensin II receptor antagonists, would be ideal in terms of safety point of view and cost effectiveness. Secondary, autonomic dysfunction, even asymptomatic dysautonomia, should be highly considered as a treatable factor to ameliorate small vessel disease (SVD) pathology in patients with VaD. Finally, non-pharmacological approaches comprising EE, physical exercise training, social interactions and mental activity would be effective interventional strategies for VaD. Nonetheless, we ought to be aware that the pathophysiology of VaD is heterogeneous and therefore, just one intervention may not be sufficiently efficacious. Large-scale prospective clinical trials in well-characterised cohorts would be required, a combination of multimodal interventions implementing pharmacological and non-pharmacological interventions (Figure 2) would be promising interventional strategies to prevent VaD, and likely to impact on processes involved in other ageing-related dementias.
Each component in both pharmacological and non-pharmacological approaches has been proposed to have protective effects on gliovascular unit (GVU) and eventually counter VaD. Abbreviations; ARB, angiotensin II receptor blocker; BMMNCs, bone marrow derived mononuclear cells; GVU, gliovascular unit; PDE3 inhibitor, phosphodiesterase type III inhibitor; SIRT1, silent information regulator 2 homolog 1; VaD, vascular dementia.
The authors declare no conflict of interest
We are grateful to the patients, families, and clinical house staff for their cooperation in the investigation of this study. We also appreciate the cooperation of the Newcastle Brain Tissue Resource (NBTR) directors and staff in assisting us with this study. We are thankful to Janet Slade and Arthur Oakley for the expert technical assistance and for assisting us in managing and screening the cohort. We specially thank Neil Hamilton and Sandra Hogg for excellent technical assistance with respect to the daily care of mice and animal husbandry.
YH and RNK declare no conflict of interests concerning this study. Our work is supported by grants from the Medical Research Council (MRC, G0500247), Newcastle Centre for Brain Ageing and Vitality (BBSRC, EPSRC, ESRC and MRC, LLHW), and Alzheimer’s Research UK (ARUK, PG2013-022). Tissue for this study was collected by the Newcastle Brain Tissue Resource, which is funded in part by a grant from the UK MRC (G0400074), by the Newcastle NIHR Biomedical Research Centre in Ageing and Age Related Diseases award to the Newcastle upon Tyne Hospitals NHS Foundation Trust, and by a grant from the Alzheimer’s Society and ART as part of the Brains for Dementia Research Project. YH was supported by SENSHIN Medical Research Foundation, Osaka, Japan and The Great Britain Sasakawa Foundation, London, United Kingdom.
