Hypertension-Induced Cerebral Small Vessel Disease Leading to Cognitive Impairment

Abstract

Objective:

Alzheimer's disease and vascular dementia are responsible for more than 80% of dementia cases. These two conditions share common risk factors including hypertension. Cerebral small vessel disease (CSVD) is strongly associated with both hypertension and cognitive impairment. In this review, we identify the pathophysiological changes in CSVD that are caused by hypertension and further explore the relationship between CSVD and cognitive impairment.

Data Sources:

We searched and scanned the PubMed database for recently published literatures up to December 2017. We used the keywords of “hypertension”, “cerebral small vessel disease”, “white matter lesions”, “enlarged perivascular spaces”, “lacunar infarcts”, “cerebral microbleeds”, and “cognitive impairment” in the database of PubMed.

Study Selection:

Articles were obtained and reviewed to analyze the hypertension-induced pathophysiological changes that occur in CSVD and the correlation between CSVD and cognitive impairment.

Results:

In recent years, studies have demonstrated that hypertension-related changes (e.g., small vascular lesions, inflammatory reactions, hypoperfusion, oxidative stress, damage to autoregulatory processes and the blood-brain barrier, and cerebral amyloid angiopathy) can occur over time in cerebral small vessels, potentially leading to lower cognitive function when blood pressure (BP) control is poor or lacking. Both isolated and co-occurrent CSVD can lead to cognitive deterioration, and this effect may be attributable to a dysfunction in either the cholinergic system or the functionality of cortical and subcortical tracts.

Conclusions:

We explore the currently available evidence about the hypertensive vasculopathy and inflammatory changes that occur in CSVD. Both are vital prognostic indicators of the development of cognitive impairment. Future studies should be performed to validate the relationship between BP levels and CSVD progression and between the numbers, volumes, and anatomical locations of CSVD and cognitive impairment.

INTRODUCTION

Hypertension affects more than 60% of individuals aged 65 years old or older and more than 80% of people aged older than 85 years old.[1] Of note, the brain and its functions are early targets of hypertension-induced organ damage. By 2050, 135.5 million people are predicted to be living with dementia worldwide.[2] This will place a substantial burden on the society, economy, and family happiness. In addition, cerebral small vessel disease (CSVD) can manifest as hypertensive vascular lesions and is a major contributor to cognitive impairment and dementia.[3,4] Despite its devastating effects, the pathogenesis of CSVD and the association between CSVD and cognitive impairment remain incompletely understood. Developing a method to identify cognitive decline early in patients with hypertension may provide a unique opportunity to implement preventive therapies before overt dementia develops. Moreover, silent CSVD, which is usually ignored, is characterized by white matter lesions (WMLs), enlarged perivascular spaces (EPVSs), lacunar infarcts (LIs), and cerebral microbleeds (CMBs). The ability to observe these signs on magnetic resonance imaging (MRI) has improved over the past decade with advances in neuroimaging. Therefore, in this review, we aimed to evaluate the pathological mechanisms in CSVD that are affected by hypertension and to further explore the relationship between CSVD and cognitive impairment.

WHITE MATTER LESION

WML or WM hyperintensities are usually located around the ventricles, especially close to the ventricular horns. WML is defined as hyperintensities on fluid-attenuated inversion recovery (FLAIR) and proton density/T2-weighted images, without prominent hypointensity on T1-weighted images.[5] WML refers to the entire spectrum of such lesions from mild perivascular tissue damage surrounding the lipohyalinotic arterioles which result in minimal axonal loss, to more severe ischemic damage, which can cause extensive myelin and axonal loss.[6]

It is generally thought that the most important risk factors for WML progression are age and hypertension.[7] On the one hand, long-term hypertension results in lipohyalinosis of the media and thickening of the vessel walls, and narrowing of the lumen of the arterioles and small perforating arteries that are derived from cortical and leptomeningeal arteries and nourish the deep WM.[8] On the other hand, hypertension increases blood vessel fibrosis, altering the distribution of Type 4 collagen and other extracellular matrix and resulting in stiffening of the vessel walls and a reduction in cerebral blood flow (CBF), especially at times of increased need.[9] It has previously been reported that hypertension modifies the relationship between amyloid and WML, in that patients with either current or past evidence of hypertension have more WML for a given burden of amyloid.[10] High systolic blood pressure (SBP) stiffens aortic roots, favoring upstream left ventricular hypertrophy and downstream WML as a result of an increase in the intensity of the propagation of pulse waves to the smallest cerebral vessels.[11] Furthermore, the deep WM has a watershed-like blood supply and is therefore more vulnerable than other regions’ impairment in CBF. This suggests that hypoperfusion and hypoxia may be early features of the development of WML.[12,13] Interestingly, hypertension influences the autoregulation of CBF by narrowing the range of the autoregulatory process instead of its speed, resulting in a steeper CBF–BP curve.[14]

Both the occurrence and the progression of WML are associated with cognitive decline, dementia, stroke, and mortality.[15] A longitudinal cohort study conducted over 4 years by Uiterwijk et al.[16] suggested that, in hyperintensive patients, the progression of periventricular WML was related to cognitive impairment (especially executive function), whereas there was no association between baseline periventricular WML and cognitive dysfunction. Thus, preventing the progression of WML should be emphasized as a therapeutic goal in hypertensive patients. Moreover, WML-associated reductions in gray matter volume were significantly more substantial and executive function, and memory was worse in uncontrolled hypertensives than in normotensives.[6] In addition, cognitive decline was more associated with the degree of periventricular WML than that of subcortical WML. Intriguingly, the former may disrupt the long associative tracts that connect more distant cortical areas, whereas the latter may cause more limited damage to short cortico-cortical connections.[16] A recent study based on a three-dimensional FLAIR showed that performance on the Mini Mental State Examination (MMSE) was associated with an increase in the total WML volume. However, only parietal WML volumes were independently correlated with Montreal Cognitive Assessment scores.[17] The effects of WML on memory are fully mediated by acetylcholinesterase activity. Data suggest that the effect of WML on the dysfunction of the cholinergic system in Alzheimer's disease (AD) patients with mild cognitive impairment (MCI) depends on the WML distribution.[18] The results of this study suggest that further pharmacological studies are warranted to explore whether WML influences responses to cholinergic treatment. For example, Blume et al.[19] found that, in Parkinson's disease, a higher WML volume was associated with the rapid onset of dementia within the 1^st year of treatment with deep brain stimulation. In contrast, a previous study failed to find a relationship between total WML volume and MMSE performance.[20] Future studies should be performed to validate the relationship between WML volume and cognitive impairment, and these studies should extend their follow-up periods as long as possible.

ENLARGED PERIVASCULAR SPACES

Perivascular spaces surround perforating arterioles and venules as they course from the subarachnoid space through the brain parenchyma and serve as an important drainage system for interstitial fluids and solutes in the brain. EPVS, also known as Virchow–Robin spaces, are most likely to be identified in T2-weighted MRI and are characterized by punctate or linear signal intensities similar to cerebrospinal fluids.[21,22]

Long-term hypertension damages the blood vessels and initiates the expression of hypoxia-sensitive genes (HIF-1α, etc.) and molecular cascades during its hypoxic phase. Inflammation is ultimately induced by the release of cytokines, inflammatory matrix metalloproteinases, and cyclooxygenase-2, and these, in turn, open the blood-brain barrier (BBB) resulting in the induction of the expression of adhesion molecules in endothelial cells and thereby contributing to leukocyte and platelet adhesion and microvascular occlusion.[13,23,24] The disarrangement of the BBB leads to the leakage of plasma components through the BBB into the vessel wall and perivascular space, and this is thought to cause EPVS.[25,26] Although Yao et al.[27] previously suggested that hypertension may promote the development of EPVS throughout the brain including the basal ganglia (BG) and WM in addition to the hippocampus. Multiple recent studies that have focused on the association between ambulatory BP levels and EPVS have found that higher SBP levels are independently associated with EPVS in the BG but not the WM. These results support the notion that EPVS in the BG may be a marker of CSVD.[22,28] The results of a prospective cohort study showed that WM-EPVS was associated with cerebral amyloid angiopathy (CAA) and superficial siderosis.[29] Conversely, a prospective, multicenter, hospital-based study published by Zhang et al.[30] showed that hypertension was an independent risk factor for EPVS in WM but not in BG or hippocampus. Future studies should further explore the relationship between BP levels and EPVS progression.

Patients with a high degree of EPVS at baseline experience greater declines in processing speed and have an increased likelihood of developing vascular dementia.[31] Huijts et al.[32] conducted a 5-year study of 189 patients with a high risk of CSVD and found that an increase in BG-EPVS was associated with a decrease in information processing speed and that this relationship was independent of age and WML. In addition, a study conducted by Arba et al.[33] revealed that the total EPVS number was not associated with cognitive impairment. While a case report published in 2017 suggested that widespread enlarged PVS could potentially cause neurological deficits and that the effect of EPVS on perivascular circulation could lead to focal brain dysfunction.[34] However, in patients with different degrees of CSVD, BG-EPVS seemed to depend on the presence of other markers of CSVD.[35] In addition, Yao et al.[27] surprisingly failed to find any significant association between the load of hippocampal load of EPVS and baseline cognitive performance or incident dementia over an 8-year follow-up period, although functional and structural changes in the hippocampus have long been demonstrated to play critical roles in memory, learning, and cognitive impairments.[36] Hence, the presence of EPVS in different anatomic sites indicates different levels of clinical significance.

LACUNAR INFARCT

LI is defined as infarction lesions of 3–15 mm in diameter that are located in the internal capsule, BG, corona radiata, thalamus, or brainstem and are caused by the occlusion of a perforating artery.[37]

These LIs are rarely fatal during the acute phase, and there is, therefore, a paucity of pathological data regarding this condition. While hypertension is the principal risk factor for stroke, it may be even more important in LI than large artery atherosclerosis and non-LIs with the same clinical severity.[38,39] A growing amount of evidence suggests that the anatomical location of LI is important to its etiology. Hypertension and an increasing WML volume independent of other vascular risk factors are significant risk factors for new LI in the deep WM. However, hyperhomocysteinemia has been associated an increased risk of LI in the BG, and hyperlipidemia always leads to isolated LI in the deep gray nuclei/internal capsule.[40,41] In patients with diabetes, LI is more frequently caused by branch orifice atheromatous disease than hypertensive arteriopathy, the latter of which is the predominant vascular pathology underlying strictly deep/mixed CMB.[42]

LI is usually considered benign and appropriate secondary preventive measures are, therefore, often neglected despite the fact that LI is associated with a higher risk of subsequent stroke and dementia.[37] Indeed, the impact of CSVD on dementia could be much more substantial than the impact of cerebral large vessel disease. A prospective cohort study conducted by Kitagawa et al.[43] revealed that the incidence of dementia was more substantially affected by the presence of LI than carotid stenosis. One key complication associated with this condition might be diffuse cerebrovascular endothelial failure, which leads to BBB damage, local inflammation, and reduced CBF as a result of the loss of autoregulation. A study that focusing on BP and poststroke cognitive impairment found that high home BP and multiple LI were significantly independent predictors for the progression of both cognitive impairment and stroke recurrence.[44] Finally, the Ohasama study revealed that, at a 7-year follow-up appointment, LI was a better indicator than WML of declines in motivation, interest, and reaction to the environment.[45]

CEREBRAL MICROBLEED

CMB, cerebral microhemorrhages, refers to homogeneous, small (<10 mm in diameter), round, or ovoid hypointensities evident on susceptibility-weighted imaging or T2* Gradient-Recall Echo MRI sequences. These signs correspond to areas of hemosiderin deposits that are themselves caused by the prior leakage of blood from small arteries, arterioles, and/or capillaries.[46]

Kwon et al.[47] focused on hypertensive stroke patients and found that exposure of the vascular endothelium to sustained hypertensive stress, particularly during reverse dipping at night, could be a rational explanation for the higher prevalence of CMB. Previous data showed that hypertension increases the expression of the cytokine tumor necrosis factor-α (TNF-α),[48] which is a pivotal regulatory cytokine that is secreted primarily by macrophages/microglia, the main cell types found to underlie CMB in pathologic/autopsy samples. In addition, a higher level of TNF receptor 2 promotes the pathogenesis of CMB.[49] The results of The Atherosclerosis Risk in Communities Study indicated that hypertensive disease (indicated by LI and WML) may contribute to deep or mixed-deep and lobar CMB, whereas CAA may drive the development of lobar-only CMB.[50] Indeed, hypertension may also be a causative factor for future CAA-related hemorrhages, which act synergistically on recurrent stroke in patients with strictly lobar CMB.[51] Alternatively, Jia et al.[52] found that hypertension increased the risk of CMB in the territory near the posterior cerebral artery (in the temporal, parietal, and occipital lobe) and deep and infratentorial locations.

The results of the population-based AGES–Reykjavik Study suggested that hypertensive vasculopathy and the combined effects of hypertensive and CAA play roles in the pathogenesis of cognitive deterioration.[53] They also found that patients with ≥3 CMB exhibited steeper declines in a composite measure of global cognitive function, memory, and speed than did those without CMB. In addition, deep and mixed CMB was associated with memory, whereas strictly lobar CMB was associated with speed and visuospatial executive functions.[53,54] Nevertheless, Heringa et al.[55] reported that there was no difference in cognitive functions between patients without CMB and those with ≥1 or ≥3 CMB. Furthermore, Rabelo et al.[56] also reported that, in mild AD patients, those with amnestic MCI, and cognitively normal elderly subjects, there was no significant difference between groups with and without CMB, indicating that CMB is not a good candidate neuroimaging biomarker for these diseases, especially in their early phases. Longitudinal studies may provide more robust information about CMB progression and its prognostic clinical significance. Multiple CMB (≥3) may disrupt connections between functionally important cortical and subcortical tracts that are critical for cognitive processes, ultimately damaging these neural networks and interfering with cognition. However, the direct impact of CMB on cognitive functions appears to be limited.[55]

CONCLUSION

Studies in the literature have confirmed that multiple hypertensive vasculopathies and changes in inflammatory status play pivotal roles in the pathological mechanisms underlying CSVD. The neuroimaging markers for CSVD, including WML, EPVS, LI, and CMB, have independent or combined effects on cognitive impairment. Therefore, in patients with CSVD, carefully monitoring and treating hypertension may provide a benefit by preventing cognitive impairment. More attention should be paid to this issue, and targeted efforts are needed to increase our understanding of the relationship between BP levels and CSVD progression and between the numbers, volumes, and anatomical locations of CSVD and cognitive impairment.

Financial support and sponsorship

This study was supported by the grants from the Science & Technology Pillar Program of Hebei Province, China (No. 14277787D), and the Science & Technology Plan Program of Hebei Province, China (No. 16397795D).

Conflicts of interest

There are no conflicts of interest.