Clinical-pathologic correlations in vascular cognitive impairment and dementia

Abstract

The most common causes of cognitive impairment and dementia are Alzheimer's disease (AD) and vascular brain injury (VBI), either independently, in combination, or in conjunction with other neurodegenerative disorders. The contribution of VBI to cognitive impairment and dementia, particularly in the context of AD pathology, has been examined extensively yet remains difficult to characterize due to conflicting results. Describing the relative contribution and mechanisms of VBI in dementia is important because of the profound impact of dementia on individuals, caregivers, families, and society, particularly the stability of health care systems with the rapidly increasing age of our population. Here we discuss relationships between pathologic processes of VBI and clinical expression of dementia, specific subtypes of VBI including microvascular brain injury, and what is currently known regarding contributions of VBI to the development and pathogenesis of the dementia syndrome. This article is part of a Special Issue entitled: Vascular Contributions to Cognitive Impairment and Dementia edited by M. Paul Murphy, Roderick A. Corriveau and Donna M. Wilcock.

Figure 1. Pathologic changes in CADASIL

(A) Deposition of granular basophilic material within the vascular media (arrow) that can be observed on hematoxylin and eosin/luxol fast blue stained sections of cerebral cortical white matter. (B) In concert, smooth muscle breakdown (highlighted by brown smooth muscle actin immunohistochemistry (arrow) is present in conjunction with arterolosclerosis as demonstrated by increased collagen (green staining from Gomori trichrome). (C) Immunohistochemistry for Notch 3 extracellular domain highlights medial deposition (arrow) in sclerotic subcortical white matter arterioles, (D) which corresponds to granular osmiophilic material deposits identified ultrastructurally. Scale bars in A–C = 100 microns; scale bar in D = 2 microns. H&E/LFB, hematoxylin and eosin/Luxol fast blue; GT, Gomori trichrome; EM, electron microscopy.

Figure 2. Distribution of individual summary neuropathology scores for three functional groups in the ACT Study

(A) Higher cognitive function: categorization of “not dementia” and performance on the CASI of 91 or higher(upper four quintiles of ACT cohort) within 2 years of death (n=107), (B) Lower cognitive function: categorization of “not dementia” and performance on the CASI of 90 or lower (lowest quintile of ACT cohort) within 2 years of death (n=71), and (C) Dementia: diagnosis of “dementia” (n=158). Diagnosis of “dementia” was made according to DSM-IV criteria. Each bar is the summary neuropathology score for an individual, which converts consensus neuropathologic results into numerical scores of 0, 1, 2, or 3 and then sums these values for AD neuropathologic change (AD), LBD, and μVBI for a maximum score of 9. For AD neuropathologic change (blue), consensus grades of “not AD,” “low AD neuropathologic change,” “moderate AD neuropathologic change,” and “high AD neuropathologic change” were converted to 0, 1, 2, and 3, respectively. For μVBI, the number of microvascular lesions observed by consensus protocol were grouped as 0 for no MVL, 1 for 1 MVL, 2 for two MVLs, and 3 for three or more MVLs. For LBD, 0 is for no LB observed, 1 for LB in the substantia nigra or locus ceruleus, 2 for LB in the limbic system, and 3 for cerebral cortical LB in consensus tissue sections.

Figure 3. μVBI in the ACT Study

(A) The cumulative frequency distribution of summary neuropathology scores for the same three functional groups as described in Figure 2. ANOVA had P < 0.0001 and corrected repeat comparisons had P < 0.05 for every paired comparison. (B) Plot of the number of MVLs in consensus screening sections from the subset of individuals in panel A with the corresponding APOE genotypes. Data are mean ± SEM except for the APOE 2/3 “Not dementia Low CASI” group (n=2) and APOE 2/3 “Dementia” group (n=8) which plot only the mean because of small sample sizes. Two-way ANOVA for the entire data set of only APOE 3/3 and APOE 3/4 had P < 0.0001 for cognitive function or dementia group, but was not significant for APOE genotype or interaction. (C) Plot of the number of MVLs in consensus screening sections or (D) concentration of cerebral cortical Aβ₄₂ from ACT participants categorized with “not dementia” and performance on the CASI of 91 or higher (upper four quintiles of ACT cohort) within 2 years of death and who have APOE genotypes 2/3 (n=16), 3/3 (n=51), or 3/4 (n=22). ANOVA was not significant (NS) for number of MVLs but had P < 0.0001 for cerebral cortical concentration of Aβ₄₂.