Sporadic late-onset AD (commonest).

 Familial late-onset AD (uncommon).

 Familial early-onset AD (rare).

 Associated with Down syndrome.

 Associated with other degenerative disease.

MACROSCOPIC APPEARANCES

The brain shows atrophy and the weight is usually in the range of 900–1200 g. There is shrinkage of cerebral gyri and widening of sulci, most prominently in the medial temporal regions (particularly the hippocampus) but also in the frontal and parietal regions. Generally, the occipital lobe and the motor cortex are relatively spared. This pattern of atrophy occurs in several dementing diseases and is not specific for AD (Fig. 31.2).

In slices of fixed brain, the cortical mantle may appear thinned (Fig. 31.3). The white matter is of normal color and texture, but reduced in volume. There may be significant dilatation of the ventricular system, especially of the temporal horn of the lateral ventricles. In the midbrain, the substantia nigra is usually normally pigmented but can appear pale. The locus ceruleus is often paler than normal. Cerebral infarcts or hemorrhages may be encountered and can be related to cerebrovascular amyloid deposition or coexistent arteriosclerotic disease.

Diffuse Aβ deposits

Diffuse deposits of Aβ peptide are seen on immunostaining as loose structures with irregular, ill-defined margins. In this form of deposit, the majority of protein is not aggregated as amyloid filaments (Table 31.4). Diffuse deposits are the main type of plaque seen in normal aging. Some diffuse Aβ deposits are described as fleece- or lake-like. Diffuse deposits may be seen in the subpial region in the cerebral cortex.

Focal Aβ deposits

Amyloid plaques are extracellular proteinaceous deposits composed of Aβ peptide, largely as amyloid filaments. Neuronal processes that traverse the plaque region show variable abnormalities and are termed plaque-associated dystrophic neurites. Plaques associated with abnormal neurites are termed neuritic plaques.

There are three main types of focal plaque deposit:

It is believed that there is progression from diffuse, through primitive to classic and finally burnt-out plaques, but there is no direct evidence for this. Plaques are widely distributed in the brain in AD. The neocortex and hippocampus are always involved. As disease progresses, plaques may also be present in the basal ganglia (where the great majority are diffuse), hypothalamus, the tegmentum of the midbrain and pons, the cerebellum (where all are diffuse), and the subcortical cerebral white matter.

Neuritic changes in plaques

Neuritic plaques include tau-immunoreactive dystrophic neurites (see Neuritic abnormalities in AD, below, and Fig. 31.11). In some neuritic plaques there are dystrophic neurites that contain chromogranin A and ubiquitin, but not tau protein; of the sparse neuritic plaques that may be present in the brains of cognitively normal elderly subjects, this form predominates. A variety of other plaque amyloid-related proteins can be demonstrated immunohistochemically, including apolipoprotein E (apoE), α1-antichymotrypsin, serum amyloid-P protein, growth factors, heparin sulfate, and complement factors.

Neurofibrillary tangles

Neurofibrillary tangles are neuronal inclusions composed largely of filamentous aggregates of hyperphosphorylated tau proteins that are variably ubiquitylated and glycated. In sections stained with hematoxylin and eosin, intracellular NFTs are faintly basophilic and extracellular NFTs appear eosinophilic (Fig. 31.16). In sections stained by silver impregnation, or when immunostained, several morphologic forms of NFT can be identified, the shape of the NFT probably being determined by that of the neuron containing it. A multi-stage model of NFT formation has been proposed (Fig. 31.17). Ultrastructural investigation reveals that NFTs are composed of paired helical filaments (PHFs) with a maximum diameter of 20 nm and a periodic narrowing to 10 nm every 80 nm (Fig. 31.18). A small proportion of filaments is straight, with a diameter of 15 nm. Detailed examination of PHF preparations shows that the filaments have a dense core region with a surrounding fuzzy coat.

NFTs are readily detected by antisera directed against phosphorylated tau protein (Fig. 31.19). Many NFTs are immunoreactive for ubiquitin or P62 (Fig. 31.20). Neurofibrillary tangles can be seen in elderly brains in low density and restricted distribution as well as in a variety of other conditions. Hence, they are not specific to AD (Table 31.5). In AD, the density of NFTs is closely related to the severity of dementia.

Neuritic abnormalities in AD. There are two main forms:

Perisomatic granules. Immunostaining for ubiquitin reveals densely-labeled round bodies adjacent to pyramidal neurons representing distended, retracted synaptic boutons. These are termed perisomatic granules (Fig. 31.21).

Tangle-associated neuritic clusters in AD. Tau immunostaining may show shows filamentous aggregates in the pyramidal cell region of the hippocampus. Although these superficially resemble plaque neurites, there is no associated focal accumulation of Aβ. These clusters represent ingrowth of tau-containing cell processes into a region previously occupied by a NFT (Fig. 31.22).

Neuronal and synaptic loss in AD. A 30–40% loss of neocortical neurons can be demonstrated in advanced AD, particularly in young-onset patients. The neuronal loss is associated with astrocytic gliosis and, in some cases, cortical microvacuolation, the latter often termed status spongiosus (Fig. 31.23). This pattern of vacuolation is coarser than that typically seen in prion disease and is largely confined to the outer cortical layers. In AD, synaptic loss of 30–50% can be demonstrated by quantitation of synapse-related proteins in affected cortical regions. The most widely used marker is synaptophysin, a glycoprotein associated with synaptic vesicles. The degree of synaptic loss correlates well with clinical scores of the severity of dementia.

Glial pathology in AD. Reactive astrocytes occur in and around neuritic plaques, and in areas of neuronal loss and cortical microvacuolation. Microglial activation occurs in relation to plaque amyloid. This close association has led to speculation that microglia are responsible for amyloidogenic processing of Aβ peptide. Aβ-induced activation of microglia may also lead to secretion of cytokines which are potentially neurotoxic.

White matter pathology in AD. Pallor of myelin staining of central hemispheric white matter is quite common in AD. In some cases this is related to microvascular changes of coincidental arteriolosclerosis. In other cases pallor of myelin occurs in the absence of vascular abnormalities.

Subcortical involvement in AD. Many subcortical regions are involved by plaques, NFTs, or NTs in AD (Table 31.6). Some regions, such as the dorsal raphe nucleus, are affected at an early stage of disease. Involvement of the nucleus basalis of Meynert is especially important, as this is the cholinergic projection nucleus to the cerebral cortex. Cell loss from this nucleus results in a severe cholinergic deficit in the cerebral cortex in AD.

Pathologic staging of AD

The evolution of pathologic changes in AD is fairly predictable and was subdivided into several stages by Braak and Braak, originally using silver staining and subsequently on the basis of immunohistochemistry for hyperphosphorylated tau protein.

Plaque stages correlate poorly with the severity of dementia (Fig. 31.24) and are:

NFT stages correlated well with the severity of dementia (Fig. 31.25). The Braak staging of neurofibrillary degeneration is incorporated in the 2012 National Institute on Aging–Alzheimer’s Association (NIA-AA) guidelines for the neuropathologic assessment of Alzheimer’s disease neuropathologic criteria for diagnosis of AD (p. 625). The BrainNet Europe Consortium has validated a scheme incorporating immunohistochemistry for hyperphosphorylated tau instead of silver staining for staging of neurofibrillary changes in AD (Fig. 31.26).

AD pathology in the cognitively normal elderly

The histologic changes that affect AD patients may be found in a restricted distribution or low density in cognitively normal elderly individuals. Plaques occur in the cortex with increased frequency in aging. In normal aging, the plaques are mainly diffuse. There may be small numbers of neuritic plaques, most associated with ubiquitin- and chromogranin-immunoreactive neurites, that do not contain tau protein. NFTs may be seen in small numbers in the hippocampus and entorhinal cortex in the cognitively normal elderly. This corresponds to Braak stages I–III (Fig. 31.25).

Pathologic diagnostic criteria for AD

Difficulties in making a pathologic diagnosis of AD occur because the histologic changes are not entirely pathognomonic and overlap those in the cognitively normal elderly. The diagnosis of AD should be restricted to cases with both plaques and NFTs in the hippocampus and neocortex (extending at least to Braak stage 4), and a clinical history of dementia.

Several different criteria have been proposed for the pathologic diagnosis for AD:

Combining these three scores (‘ABC scoring’) allows the pathologist to allocate a probability that AD-associated abnormalities accounted for the patient’s dementia in life. AD lesions seen in the post-mortem brain from cognitively normal elderly people are considered pathologic rather than a part of a normal aging process. The likelihood that clinical dementia has been caused by AD lesions in the brain is stratified on the basis of the post-mortem neuropathologic findings, as follows (Table 31.11):

Patients who have plaques and NFTs in a very restricted density or distribution or who are not clinically demented are best classified as having ‘Alzheimer’s disease neuropathologic change’, as this does not pre-judge the nature of their disease.

Patients who have a restricted distribution of tau immunoreactive structures not meeting criteria for a specific tauopathy and not fitting with a Braak stage can be designated as having ‘Stage + neurofibrillary change’.

In applying earlier NIA–Reagan criteria, it became apparent that many laboratories had moved away from use of silver-stained preparations in evaluating AD pathology and that alternate approaches were desirable. The BrainNet Europe group published staging criteria based on evaluation of tau-labeled histologic sections from four brain regions (Figs 31.27–31.30).