The Pathology of Alzheimer Disease

A. Epidemiology

AD is the most common dementing illness and can occur at any age over 30. It tends to affect women more than men, even when survival differences are taken into account. It is relatively uncommon before age 65, affecting about 0.1% of the population, but its incidence and prevalence rise rapidly thereafter: about 10% of the over-65 crowd, and as many as half of those over the age of 85, may be affected. The overall prevalence of AD in the U.S. population is approximately 1500 per 100,000 people, making it about 4 times more common than PD, and 150 times more common than either ALS or HD. Mutations affecting the b-amyloid precursor protein or presenilin proteins have been identified in rare families demonstrating early-onset (before the age of 65 in the former, and before the age of 50 in the latter) AD. In addition, the apolipoprotein E4 allele appears to be a major risk factor for late-onset AD, underscoring the probability that late-onset AD represents a complex interplay of multiple genetic as well as environmental factors. Individuals suffering from AD initially demonstrate memory difficulties. As time passes, emotional disturbances, changes in personality, language difficulties, and visuospatial impairments become manifest. Motor disturbances are usually not encountered until very late in the course of the disease.

B. Pathology

The primary neuropathologic alterations in the AD brain are synaptic and neuronal loss, accompanied by the development of amyloid plaques and neurofibrillary degeneration.

Senile plaques

Classical ("neuritic") plaques comprise a central core of amyloid surrounded by degenerating neurites (silver-positive swollen nerve terminals containing tau and ubiquitin reactive paired helical filaments identical to neurofibrillary tangles, described below). With the development of sensitive antibodies to Alzheimer's type “β”- amyloid, it has been shown that neuritic plaques are far outnumbered by "diffuse" plaques: amyloid immunoreactive deposits unaccompanied by neuritic degeneration. While the relationship between diffuse and neuritic plaques is unresolved, only the neuritic type is associated with clinically apparent neurologic dysfunction. Amyloid is also deposited within the media of leptomeningeal and intra-cortical blood vessels (referred to as amyloid or congophilic angiopathy).

Neurofibrillary tangles

NFTs are perinuclear intraneuronal accumulations of paired helical filaments composed of protease-resistant cores and protease-labile outer coats. The core is composed largely, if not entirely, of abnormally phosphorylated tau protein (a low molecular weight microtubule associated protein), while the coat contains a variety of other proteins, including ubiquitin. While antibodies to these various proteins specifically highlight NFTs in tissue sections, these tangles are also well stained by a variety of silver-based histochemical preparations. While it is likely that many of the dystrophic neurites within plaques are directly continuous with the cell bodies of nerve cells showing neurofibrillary degeneration, it has yet to be proven that neuronal damage is directly caused by amyloid.

Neuropil threads

Neuropil threads are tau, ubiquitin, and silver-positive processes scattered throughout the neuropil. They frequently occur in the dendrites of tangle-bearing neurons, and their density often parallels neurofibrillary tangle density. In some regions their formation may precede the development of neurofibrillary tangles.

Neuronal loss

Due to numerous methodological difficulties in the quantitative assessment of neuronal populations within cerebral cortex, studies of neuronal loss have lagged far behind those of SPs and NFTs. However, to understand the proximate cause of cognitive failure in AD, one must understand the type and degree of neuronal dysfunction and death. Recent stereological morphometric analyses have show that the total number of neurons in the superior temporal sulcus (a high order association cortex) is unchanged from the sixth to the tenth decade in non-demented individuals, but is reduced by 41% in patients with AD, with the degree of loss strongly correlating with the duration of dementia. NFTs have also been found to correlate with dementia, although the numbers of neurons lost exceeds the number of NFTs by more than an order of magnitude. By contrast, SP density does not correlate with neuronal loss or duration of disease. Despite this impressive degree of neuronal loss in the superior temporal sulcus, the overall gross appearance and weight of AD brains overlaps significantly with those of age-matched controls, and definitive diagnosis requires histologic examination of brain tissue. Investigations into the use of accurate positional matching and digital subtraction of serially acquired magnetic resonance images to allow determination of rates of global and regional atrophy have demonstrated preferential atrophy in the temporal lobe & hippocampal regions in patients with AD.

Clinicopathologic Correlation

Although generally approached as a global cerebrocortical affliction, careful analyses of AD and control brains have shown evolution of pathologic damage centering on the limbic system. Neuronal damage within this system impairs the transmission of information from the neocortex to the hippocampal formation, and contributes to the personality changes and early cognitive decline seen in AD patients. Neocortical pathology is reflected by a corresponding impairment of higher brain functions. The appearance of a hypokinetic, hypertonic syndrome during the late stages of AD is probably related to involvement of the striatum and substantia nigra.

The problem with many of the pathologic systems for diagnosing AD is that they try to address too many questions at once. Specifically, they attempt to answer the questions: 1. Is this brain normal? and 2. Are the histopathologic features seen within the brain responsible for the patient’s dementia? with one set of criteria. I find it instructive to view amyloid and neurofibrillary degeneration in AD as directly analogous to atherosclerosis and myocardial injury in ischemic heart disease. That is, while one should not consider the presence of amyloid within the cerebral cortex as healthy, and while there are very old people without detectable amounts of amyloid within their brains, the distribution and amount of visible amyloid does not itself produce neuronal dysfunction. Amyloid deposition does appear to increase the likelihood of neurofibrillary degeneration, which is closely associated with neurologic dysfunction. Careful anatomic studies have demonstrated a sequence of neurofibrillary degeneration beginning in the entorhinal cortex (clinically asymptomatic) and progressing through a limbic (hippocampal) stage (where patients demonstrate behavioral changes but are not clearly diagnosable as demented) to a final, neocortical stage (clearly associated with clinical AD). Such a system can only be used at autopsy, and studies are ongoing to try and find more accurate means of diagnosing AD during life, both for prognostic and (hopefully soon) therapeutic reasons.