Cerebrovascular Disease and Mechanisms of Cognitive Impairment
2012; Lippincott Williams & Wilkins; Volume: 43; Issue: 9 Linguagem: Inglês
10.1161/strokeaha.112.655803
ISSN1524-4628
Autores Tópico(s)Cerebrovascular and Carotid Artery Diseases
ResumoHomeStrokeVol. 43, No. 9Cerebrovascular Disease and Mechanisms of Cognitive Impairment Free AccessReview ArticlePDF/EPUBAboutView PDFView EPUBSections ToolsAdd to favoritesDownload citationsTrack citationsPermissionsDownload Articles + Supplements ShareShare onFacebookTwitterLinked InMendeleyReddit Jump toSupplemental MaterialFree AccessReview ArticlePDF/EPUBCerebrovascular Disease and Mechanisms of Cognitive ImpairmentEvidence From Clinicopathological Studies in Humans Raj N. Kalaria, FRCPath Raj N. KalariaRaj N. Kalaria From the Centre for Brain Ageing and Vitality, Institute for Ageing and Health, Newcastle University, Newcastle upon Tyne, UK. Originally published9 Aug 2012https://doi.org/10.1161/STROKEAHA.112.655803Stroke. 2012;43:2526–2534Other version(s) of this articleYou are viewing the most recent version of this article. Previous versions: January 1, 2012: Previous Version 1 IntroductionCerebrovascular disease (CVD) is an important cause of disability and dementia. Accumulating evidence from clinical, neuroimaging, and pathological studies indicates a durable link between CVD and cognitive impairment. To distinguish this relationship, over the years various disorders have been progressively refined to include arteriosclerotic dementia, multi-infarct dementia, vascular dementia (VaD), subcortical ischemic vascular dementia, and vascular cognitive impairment (VCI).1–4 VCI is attributed to conditions resulting from a variety of cerebrovascular lesions or impaired brain perfusion,5 but the pathological diagnosis of VCI, however, requires systematic evaluation of potentially relevant clinical and other phenotypic features. Although vascular pathology substrates that explain cognitive dysfunction have been narrowed down, debate still abounds as to what is the gold standard for confirming the clinical diagnosis of these conditions. In the late 1960s, the seminal clinicopathological studies in Newcastle provided evidence that destruction of brain tissue by ischemic injury is associated arteriosclerotic dementia.6 Cognitive impairment was attributed to cerebral softening with loss of a relatively large volume (∼100 mL) of tissue that was reported to be overdiagnosed clinically in comparison to Alzheimer disease (AD).7,8 These views are now disputed. Nonetheless, identifying the pathological substrates of CVD causing cognitive and behavior dysfunction has been a lasting challenge.In this review, I discuss available clinicopathological evidence to link cerebrovascular changes to cognitive impairment and highlight key substrates as well as consider relevant mechanisms. The contribution of Alzheimer type of pathology to CVD-related impairment9 is not discussed in detail although this remains an intriguing area of overlap with the selected topic. Previously published articles in English until December 2011 derived from various popular databases including PubMed of specific interest were searched online. Several search terms were used initially to obtain wider coverage of articles from studies in which cognitive function was related to neuroimaging and physiological and pathological measures. However, using key terms such as cognitive impairment, dementia, stroke, vascular, AND pathology yielded 541 articles. Further refining the search terms to infarcts, small-vessel disease and white matter, AND pathology, I found 85 articles. Of these the majority provided limited qualitative or descriptive data. To focus on the topic of interest and make a case for/against which vascular changes or lesions relate to cognitive impairment, the search was narrowed down to reviewing clinicopathological studies only. To evaluate the reported findings, I finally used 21 articles, which reported minimal neuropsychometric data sets and quantitative or semiquantitative pathological results on various changes linked to VaD or VCI. It is clear from this exercise that intracranial small vessel disease (SVD) rather than large strategic cortical lesions is a major factor in VCI. Lacunar infarcts and multiple microinfarcts within the lenticular nuclei, thalamus, and frontal white matter (WM) occur frequently (>50%) in CVD. Neocortical microinfarcts best correlate with impairment but they also multiply in the presence of cerebral amyloid angiopathy or focal microvascular amyloid deposition. Diffuse WM changes predominantly associated with disruption of corticocortical pathways represent deep WM demyelination and damaged axons. Diffuse WM changes result from a chronic hypoxic state generated by reduced perfusion. Hippocampal sclerosis and neuronal atrophy are also important elements related to dementia. Inflammatory and oxidative stress mechanisms precede apoptotic neuronal death but it is unclear if autophagy occurs extensively. Familial disorders such as cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) characterized by severe disintegration of frontal WM exhibit cortical neuron apoptosis, presumably contributing to interruption of frontosubcortical circuits. An important characteristic of VCI is executive dysfunction, but little is known about the vascular basis of neuronal integrity in the frontal lobe.Types of Vascular Change and Mechanisms Contributing to Cognitive ImpairmentClinical observation and MR neuroimaging, in particular, provide vital correlative information on where to look but cannot provide a foolproof system for correct diagnosis. Precise pathological examination is necessary to affirm accurate dementia diagnosis and elucidate disease mechanisms. However, unlike neuroimaging, pathological practice is labor-intensive and can only achieve a limited screen, but what is examined has to be decisive and representative.3 Despite such aspirations, identifying the neuropathological substrates of VaD/VCI per se is complicated by the heterogeneous distribution and localization of lesions and the coexistence of neurodegenerative changes (Table 1). Several factors are involved in defining in cerebrovascular phenotype and relating them to degree of impairment.10,11 Vascular lesions include the type and size of vessels, origin and nature of vascular occlusion, and diffuse or focal integrity of the vessel wall components as critical variables in perfusion. The degree of parenchymal changes relative to cognitive domains and dysfunction is dictated by the presence of hemorrhage, distribution of arterial territories, and the anatomic location (cortical versus subcortical), size, multiplicity, laterality, and age of the lesions besides genetic influences and previous existence of systemic vascular disease. Arteriolar changes including intimal thickening, fibroid necrosis, and lipohyalinosis (Figure) may be quantified as the sclerotic and hyalinosis indices. The progression of cellular events and manner of cell death, for example, apoptosis, autophagy occurring within the parenchyma after infarction, is important in assessing the severity of the lesions. Other variables to be noted include size of the incomplete infarct or perifocal hypoperfused (penumbra) zone and the presence of edema, pyknotic neurones, bulbous axons, shrunken oligodendrocytes, hemosiderin, neutrophils, lipid-laden macrophages, and intensity of astrocytosis or microgliosis. For the majority of these variables, semiquantitative methods with inherent limitations are the norm. However, there is still lack of critical standardized methods to assess lesion burden.Table 1. Types of Vascular and Parenchymal Changes in Neuropathological Cases Associated With VaDPathological FeaturePredominant LocationFrequencyAssociation With CIAtheromasCarotid artery bifurcation and internalHighWeakAtheromatous and occlusive diseaseCircle of Willis, proximal branches of MCA, ACA, PCAHighModerateComplete infarctions (macroscopic), arterial territorial infarctionsCortical and subcortical regionsModerateWeakLacunar infarctsWM, basal ganglia, thalamusModerateModerateCystic infarctsWM, basal ganglia, thalamusModerateUnknownSmall or microinfarctsCortical and subcorticalHighStrongHyalinosis, lipohyalinosis, fibroid necrosisWM, cortical and subcortical grey matterModerateUnknownCribriform change, perivascular spacingWM, basal ganglia, internal and external capsulesHighStrongDemyelination and oligodendrocyte changesWMHighStrongGliosis: astrocytosis and microgliosisWM, cortical and subcorticalVariableModerateCerebral amyloid angiopathyCorticalModerateModerateIntracerebral hemorrhagesCortical, subcortical and lobarLowModerateMicrospongy form changeNeocortical layer I–IIModerateUnknownLaminar necrosis, gliosisNeocortical ribbonLowUnknownHippocampal atrophy and sclerosisCA1–CA4ModerateStrongAlzheimer type of pathology (concomitant)Hippocampus, neocortexLowStrongData summarized from several references.10 Percent cases are averaged from ≥2 reported studies. Cystic infarcts (possibly also lacunar) with typically ragged edges were admixed in both cortical and subcortical structures.VaD indicates vascular dementia; CI, cognitive impairment; MCA, middle cerebral artery; ACA, anterior cerebral artery; PCA, posterior cerebral artery; WM, white matter.Download figureDownload PowerPointFigure. Scheme showing progressive vascular and parenchymal changes linked to different but interconnected mechanisms. Microangiopathies occurring as nonamyloid, for example, hypertensive type (A) involve vascular degenerative processes progressing from loss of smooth muscle cells, wall thickening, fibroid necrosis to hyalinosis, and collagenosis. Cerebral amyloid angiopathies (CAAs; B) involve deposition of fibrillar protein, loss of smooth muscle cells, and intimal thickening. Both processes may lead to common parenchymal changes involving gray matter (GM) and white matter (WM) microinfarctions and rarefaction. A series of molecular and perivascular cellular stages is apparent manifested by activation of the endothelium, concomitant breach of the blood–brain barrier and perivascular glial responses.23,58,59,94,95 Microhemorrhages or bleeds (B, D) within the tissue may be caused by leakage from microvessels53 through both mechanisms include microaneurysms and rupture of walls due to deposition of fibrillar proteins or iron.96 Increased perivascular spacing occurs because of reduction in arterial vascular tone and lack of perivascular solute drainage.97 The WM changes ensue due to a chronic hypoxic state and decline in oligodendrocytes.51,55A, Vessel wall thickening, fibroid necrosis, lipohyalinosis, or hyalinization. B, arteriolar and focal Aβ deposits, severe CAA, and CAA-related microhaemorrhage. C, Microinfarcts in putamen and WM. D, CAA-related rarefaction and microhemorrhages. Aβ indicates amyloid β protein; BBB, blood–brain barrier; CD68, cluster of differentiation marker 68 for microglia; EC, endothelial cell; ECM, extracellular matrix; eNOS, endothelial nitric oxide synthase; HIF-1α, hypoxia inducible factor 1α; NV, neurovascular; PVS, perivascular space; SMC, smooth muscle cell; VEGF, vascular endothelial growth factor; WM, white matter.Cerebrovascular changes may be conveniently grouped under large- and small-vessel domains. Large-vessel and cardiac embolic events involving atherosclerosis, plaque rupture, intraplaque hemorrhage, thrombotic occlusion and embolism, dissection, and dolichoectasia may lead to macroinfarcts. Accompanying hemodynamic events cause border-zone or watershed lesions (approximately 5 mm) as wedge-shaped regions of pallor and rarefaction extending into the WM. These incomplete or subinfarctive changes, however, suspected to be of clinical importance, are not uniformly described in VCI.3 Stenosis arising from atherothromboembolism associated with major arterial territories may be admixed in cortical and subcortical regions12 and rarely extend in smaller vessels beyond the circle of Willis.13 Clinically, thromboembolic events are considered to be responsible for up to 50% of all ischemic strokes, whereas SVD causes 25% of the infarcts14 but similar frequencies are not evident in autopsy series of dementia cases.15 The occlusion of internal carotid arteries and those of the circle of Willis are thought to explain ≤15% of VaD,12 but there is no widely held consensus on correlation of % stenosis and degree of impairment in CVD.16 In VaD/VCI arterial disease is considered more significant than venous disease. However, venous adventitial fibrosis has been linked to cognitive abnormalities.17 Laminar necrosis and complicated angiopathies such as fibromuscular dysplasia, arterial dissections, granulomatous angiitis, collagen vascular disease, and giant-cell arteritis are rarer in CVD with dementia.14SVD, Microangiopathy, and the Blood–Brain BarrierSmall-vessel anomalies leading to microinfarcts and diffuse WM changes are common in VCI.18 These entail arteriolosclerosis, lipohyalinosis, fibrinoid necrosis, microatheromas, segmental arterial disorganization, and cerebral amyloid angiopathy (CAA; Figure). Although the initiating factors causing the microangiopathy may be different, end-stage pathology invariably involves replacement of vascular smooth muscle with collagenous or other nontensile fibrillar material in both sporadic and familial cases. Small vessels of the brain10 including perforating arterioles and intracerebral end-arteries are causal in lacunar infarcts (lacunes or cystic lesions generally 50% of casesHippocampal injury or scarring in >50%HAAS32Aging230; 74–97Lacunar infarcts and MIMixed and ATPCASIMicrovascular lesions including MIs associated dementiaMIs more common with lacunar infarcts; cortical MIs common with subcortical MIsGeneva University Study33Neuropsychiatric disease45; 63–100Cortical MISome ATP; macroinfarcts excludedCDRCombined cortical MI highly associated with CIPeriventricular and deep WM demyelination additional factorsReligious Orders Study71Aging153; 76–93Cerebral macroinfarcts and MISeveral ATPCERADMI increased odds of dementia; MI in 29%, of whom 49% had macroscopic infarctionsLack of direct interaction between ATP and infarctions; MI related to perceptual speed not global cognitive scoreGeneva University Study86Neuropsychiatric disease72; 63–100Microvascular pathology and MISome ATPCDRMicrovascular pathology, cortical MI (40%) strongly related to cognitive dysfunctionWM demyelination related to lower CDRs; thalamic and BG lacunes and Aβ-protein staging also predicted cognitive statusGeneva University Study87Neuropsychiatric disease47; 73–101Cortical MISome ATP, Braak IIICDRCortical MI strongly associated compared with Aβ stagingPeriventricular demyelination to a lesser degree contributes to dementia in those at high riskReligious Orders Study36Aging148; 80–95Subcortical infarcts including MISome ATPCERAD plus RUSH batteryMI (24%) alone not related to impairment or overall cognitive dysfunctionSubcortical infarcts (>60%) add to ATP and worsen memory functionAdult Changes in Thought Study88Aging211; 70–90Macroinfarction and MISome ATP and LBsCASIMI (>1 in 62%) was strongly associated with dementiaOther markers of significance were Braak stage and presence of cortical LBsBaltimore Aging Study89Aging179; 67–96Macroinfarction and MISome ATPDSM-III-R and CDRMI, independent of macroscopic infarcts, significant cause of dementiaATP adds to the burden of dementia depending on location of infarctsHAAS90Aging443; 75–102Macroinfarcts and MISome ATPCASIMI (34%) sole or dominant lesion associated with dementiaATP worsen outcome in presence of MICambridge City Over 75s91Aging213; 76–100Microvascular pathology and MIMajor ATPDSM-IV and CERADMI marginally more common in those (4%) with diagnosis of VaDATP and microvascular pathology very commonReligious Orders Study92Aging425; 80–98MI pathologySome ATPCERAD plus RUSH batteryMI (30%) common in demented; cortical MI found to be associated with lower cognition, semantic memory, perceptual speed, and visuospatial abilitiesSubcortical MI higher than cortical MI; ATP did not differ by presence of MICOGFAST67CVD55; 76–98Macroinfarcts and MIFew ATPCAMCOGMI high in demented stroke survivorsMajority (70%) of demented had Braak Stage III or lowerTable summarizes recent community-based and cohort studies in which small lesions or MI was assessed. In previous clinicopathological studies, lacunar lesions were associated with VaD28,93 or CVD with dementia.39HAAS indicates Honolulu-Asia Aging study; COGFAST, cognitive function after stroke; CVD, cerebrovascular disease; MI, microinfarcts; ATP, Alzheimer type of pathology; LBs, Lewy bodies; SIVD, Subcortical Ischemic vascular dementia; DSM-IV-R, Diagnostic and Statistical Manual of Mental Disorders, 4th Edition, Revised; MTS, Mental Test Scores; ADDTC, Alzheimer Disease Diagnostic and Treatment Centers; CASI, Cognitive Abilities Screening Instrument; CERAD, Consortium to Establish a Registry for Alzheimer Disease; CDR, Clinical Dementia Rating; RUSH, Rush Alzheimer Disease Center; CAMCOG, Cambridge Cognitive Assessment; CI, cognitive impairment; WM, white matter; VaD, vascular dementia; CAA, cerebral amyloid angiopathy.Microinfarction as a Key Factor in the VCIMicroinfarcts are described as attenuated lesions of indistinct shape occurring in both cortical or subcortical regions involving a small vessel at its core but are foci with pallor, neuronal loss, axonal damage (WM), and gliosis (Figure). They appear robustly associated with cognitive impairment and predict poor outcome in the elderly with CVD.30–32 Multiple rather than single microinfarcts (50–500 μm diameter) are more correlated with dementia also independent of Alzheimer pathology (Table 2).10 Among these is the neocortical rather than subcortical microinfarcts and to a lesser extent periventricular demyelination predict progression of cognitive deficits.33,34 Microinfarcts in the watershed regions35 and elsewhere,36 however, can aggravate the degenerative process as indicated by worsening impairment in AD. Microinfarcts most probably denote an undiscovered surrogate of an underlying microvascular disease.10CAA is a cause of intracerebral and lobar hemorrhages37 and it commonly occurs in AD.38 However, CAA is also an independent factor in approximately 10% of VCI cases in the general absence of Alzheimer pathology.20,39–46 The correlation between CAA and microinfarcts shows these are in tandem important substrates of cognitive decline.47,48 Conversely, neocortical microinfarcts associated with severe CAA appear the primary pathological substrate in a proportion of CVD cases.47,49 Amyloid β protein accumulation within or juxtaposed to the vasculature may lead to degeneration of both larger perforating arterial vessels as well as cerebral capillaries. The weakening integrity of the vessel walls lowers perfusion with resultant cortical microinfarctions.47,49,50 In turn, hypoperfusion may also interfere with the arterial pulsations and with the interstitial fluid pressure, leading to reduced perivascular clearance of amyloid β. Changes in the hemodynamics, for example, hypotension in the presence of CAA, has also been indicated as a key factor in the genesis of cortical watershed microinfarcts.49WM Changes in VCISubcortical leukoencephalopathy is a common pathological change in SVD (Table 1) and in VaD.51 Both the sclerotic changes in the medullary arteries and the WM changes in subcortical VaD are most prominent in the frontal lobe.52 These are accompanied by vacuolization and widening of the perivascular spaces.27 Vascular stenosis caused by collagenosis may induce chronic ischemia or edema in the deep WM leading to capillary loss and more widespread effects.53,54 WM rarefaction is largely attributed to myelin loss and axonal abnormalities resulting from vascular insufficiency (oligemia) and a chronic hypoxic state55 rather than Wallerian degeneration secondary to cortical loss of neurons reflect these lesions.18,56,57 However, microarrays have indicated several molecular pathways are involved in WM changes.58 The vascular basis of WM disease also comprises pallor or swelling of myelin, shrinkage of oligodendrocytes, and accumulation of degenerated myelin basic protein. The shrinkage of oligodendrocytes appears a primary event toward cell death with eventual loss of number in the perifocal infarcted WM.51 Although mild to moderate ischemic injury activates the myelin repair system, prolonged ischemia is thought to damage oligodendrocyte precursor cells, which do not retain the ability to compensate for the myelin loss resulting in unsuccessful remyelination. Reactive astrogliosis or microgliosis usually signals repair but it is unclear whether diffuse axonal loss occurs in the rarefied deep WM.59Hippocampal Atrophy and SclerosisHippocampal neurones are highly vulnerable to disturbances in the cerebral circulation or hypoxia caused by systemic vascular disease. SVD leads to hippocampal and brain atrophy and can be apparent in >50% of cases.31,45 The presence of focal microinfarcts and scars versus diffuse or segmental (CA1, subiculum) neuron loss and astrocytosis varies considerably with age and not all agree this is necessarily of vascular origin.60–62 The loss of CA1 neurons in ischemic VaD has been related to lower hippocampal volume and memory score63 but conflicting reports of neuronal loss in SVD64 suggest other mechanisms are involved. Selective hippocampal neuronal shrinkage does not only appear an important substrate for VaD but also delayed dementia after stroke in the absence of any neurodegenerative pathology.65 This provides evidence for a vascular basis for hippocampal neurodegeneration and fits with neuroimaging data that medial temporal lobe atrophy is not restricted to AD. The simplest mechanistic explanation for the atrophy is that the neuronal or dendritic arbor results in subsequent loss in connectivity, which contributes to brain structural and functional changes. Such reasoning is consistent with the finding that concentrations of soluble synaptophysin were found to be decreased in VaD besides AD.66Overlap With AD PathologyTissue changes characterizing AD including amyloid β plaques and neurofibrillary pathology occur more often in cases of cerebrovascular disease than in normal aging elderly.67,68 Several aging series studies reveal similar pathological results indicating that a combination of Alzheimer type of changes and microvascular disease worsen cognitive outcomes.69 The Nun Study was one of early reports that suggested that individuals with neurofibrillary pathology and one or 2 lacunar infarcts experienced a steeper drop in cognitive function compared with those with no infarcts.70 The relationship between AD pathology and infarctions is debated but it is clear that the presence of one or more infarctions independently increases the odds of dementia beyond their additive effect.71 Thus, the presence of even marginally increased burden of amyloid deposits and τ pathology above normal aging and insufficient for pathological diagnosis of AD72 adds to the tissue processes, which result in cognitive impairment due to vascular causes.Neurotransmission in VCIThe neurochemical basis of cognitive decline in CVD is poorly understood. The perivascular nerve plexus21 is most vulnerable, yet only few transmitter-specific changes reflecting neurovascular pathology have been described in subtypes of VaD. Cortical cholinergic synaptic activity was reduced in multi-infarct dementia.73,74 However, loss of cholinergic function was only evident in patients with VaD with concurrent Alzheimer pathology.75 Conversely, a novel increase in cholinergic activity in the frontal cortex was revealed in infarct dementia.76 Despite these changes, there do not appear to be pronounced effects in the cholinergic cell bodies of the basal forebrain in VaD.77,78 There was also loss of glutamatergic synapses, assessed by vesicular glutamate transporter 1 concentrations, in the temporal cortex of VaD,79 but preservation of these in the frontal lobe suggests a role in sustaining cognition and protecting against dementia after a stroke. Identification of the morphological equivalents of these changes in types of pyramidal cells in the frontal lobe would be relevant. Other studies have reported deficits in monoamines including dopamine and 5-hydroxytryptamine in the basal ganglia and neocortex in VaD.74 To compensate for the loss,80 5-hydroxytryptamine(1A) and 5-hydroxytryptamine(2A) receptors are apparently increased in the temporal cortex in multi-infarct but not subcortical VaD. Such findings, albeit fragmentary, reveal distinctions between the neurochemical pathology of VaD subtypes and suggest possibilities of pharmacological manipulation with novel therapies in VaD.Clinicopathological Features of Importance in Familial SVD as Models of VCIStudies on the familial forms of SVD of the brain such as CADASIL have largely concentrated on describing the clinical phenotypes.81,82 Compared with older subcortical VaD types, there is much greater progression of pathological changes in CADASIL indicated by profoundly increased sclerotic index in WM and subcortical vessels, which is accompanied by tortuosity and myelin degradation causing lack of drainage of the interstitial fluid and perivascular spaces. Apoptotic loss of cerebral vascular smooth muscle cells leading to wall thickening and fibrosis in small- and medium-sized penetrating arteries is causal in subcortical infarction and cribriform change.83 Lack of arteriolar wall and endothelial integrity likely reduce both blood flow and volume in affected frontotemporal WM and subcortical gray matter structures with effects on the hemodynamic reserve by decreasing the vasodilatory response.84 Neuronal apoptosis, predominantly in neocortical layers III and V, appears a crucial cell death mechanism in CADASIL.83 The extensive demyelination and axonal damage in the underlying WM contribute to cortical atrophy and affects frontal lobe cognitive functions.What More Can Be Done?The heterogeneity of CVD compels understanding of the neuropathological substrates and mechanisms of VCI. SVD causing cortical and subcortical microinfarcts appears most robustly related to cognitive impairment. Identifying the protein or molecular substrates of subcortical microvascular disease would be important in defining precise mechanisms. Diffuse WM changes trigger hypoxic and angiogenic signals but it is unclear what prevents increase in microvascularity to enhance perfusion. Concomitant hippocampal atrophy including sclerosis is an important feature of VCI but we lack information on the neuronal substrates of frontal lobe dysfun
Referência(s)